id: MODULE:eukaryotic_mismatch_repair
title: Human DNA mismatch repair — MutSalpha/MutSbeta mismatch recognition, RFC-PCNA-directed MutLalpha incision, EXO1 excision with RPA protection, and POLD resynthesis with LIG1 ligation; MSH2/MSH6/MSH3/MLH1/PMS2/RFC1-5/PCNA/EXO1/HMGB1/RPA1-3/POLD1-4/LIG1
description: >-
  Post-replicative DNA mismatch repair (MMR) removes base-base mismatches and
  small insertion/deletion loops that escape the proofreading of replicative
  DNA polymerases. In human cells a mismatch is recognised by one of two MutS
  homologue heterodimers that share MSH2: MutSalpha (MSH2-MSH6) binds
  base-base mismatches and one- or two-nucleotide insertion/deletion loops,
  and MutSbeta (MSH2-MSH3) binds larger insertion/deletion loops and has only
  a limited role in base-base mismatch repair. MutSalpha is pre-enriched on
  chromatin through binding of the MSH6 PWWP domain to H3K36me3. In an
  ATP-dependent step the mismatch-bound MutS complex recruits MutLalpha
  (MLH1-PMS2). MutLalpha is a latent endonuclease whose active site lies in
  PMS2; it is activated by the mismatch, MutSalpha, ATP and the PCNA sliding
  clamp that replication factor C (RFC1-5) loads at the strand break or
  primer terminus. Because RFC loads PCNA in a fixed orientation, the
  activated MutLalpha nicks the discontinuous (newly synthesised) strand,
  which is how the daughter strand is identified. Exonuclease 1 (EXO1),
  activated by MutSalpha, removes the error-containing tract 5' to 3' from a
  nick, and excision stops after the mispair is removed through the combined
  action of MutLalpha and RPA, the heterotrimeric single-stranded DNA-binding
  protein (RPA1-RPA3) that coats and protects the gap. The chromatin protein
  HMGB1 stimulates MutSalpha-activated excision. The gap is filled by DNA
  polymerase delta (POLD1-POLD4) working with PCNA and RFC, and DNA ligase 1
  (LIG1) seals the remaining nick. Loss of MSH2, MSH6, MLH1 or PMS2 causes
  microsatellite instability and Lynch syndrome. Mismatch avoidance by
  polymerase proofreading, MMR-dependent signalling of DNA damage, meiotic
  crossover resolution and the specialised mutagenic MMR of antibody genes
  are outside this module.
status: DRAFT
scope: CONCRETE
evidence:
  - source_id: PW:0000662
    title: mismatch repair pathway
    statement: >-
      The module boundary follows the RGD Pathway Ontology term and its
      diagram. The PW definition names Msh proteins (MutS counterparts), Mlh
      and Pms proteins (MutL counterparts), Exo1 excision, Rpa protection of
      single-stranded DNA, and resynthesis and ligation by DNA polymerase
      delta and DNA ligase 1 with the Pcna clamp and the Rfc clamp loader.
      The diagram draws Exo1, Lig1, the Mlh and Msh members, the Rfc complex
      and the pold complex associated with Pcna. Derived through
      projects/RGD_PATHWAY_ONTOLOGY (pathways/PW_0000662.yaml).
  - source_id: GO:0006298
    title: mismatch repair
    statement: >-
      skos:closeMatch for PW:0000662 in the project seed (RGD page GO xref, PW
      definition xref and lexical match). It is the module's process concept.
  - source_id: PMID:16143102
    title: Reconstitution of 5'-directed human mismatch repair in a purified system.
    statement: >-
      Human MMR has been reconstituted from purified components, and the
      reconstituted protein set defines the module's parts.
    supporting_text: The reconstituted system includes MutSalpha or MutSbeta, MutLalpha, RPA, EXO1, HMGB1, PCNA, RFC, polymerase delta, and ligase I.
  - source_id: PMID:16873062
    title: Endonucleolytic function of MutLalpha in human mismatch repair.
    statement: >-
      MutLalpha is a latent endonuclease activated by the mismatch, MutSalpha,
      RFC, PCNA and ATP; its incision is biased to the nicked strand.
    supporting_text: We show that human MutLalpha is a latent endonuclease that is activated in a mismatch-, MutSalpha-, RFC-, PCNA-, and ATP-dependent manner.
  - source_id: PMID:33510387
    title: Mispair-bound human MutS-MutL complex triggers DNA incisions and activates mismatch repair.
    statement: >-
      Current model of the human reaction: MutSalpha recruits MutLalpha, which
      nicks the daughter strand 5' to the mismatch; EXO1 excises from that
      nick; POLD fills the gap with PCNA and RPA, and LIG1 seals it.
    supporting_text: MutSα recruits MutLα to form a mismatch-bound complex, which initiates MMR by nicking the daughter strand 5' to the mismatch.
  - source_id: PMID:10748105
    title: Identification of a second MutL DNA mismatch repair complex (hPMS1 and hMLH1) in human epithelial cells.
    statement: >-
      MutLbeta (MLH1-PMS1) exists in human cells, but this study shows only
      complex formation and MLH1-dependent stability, not a repair activity;
      MutLbeta is therefore recorded as context, not as a part.
    supporting_text: Our data provide evidence that MutL proteins form two different complexes, MutL-alpha (hPMS2 and hMLH1) and MutL-beta (hPMS1 and hMLH1).
  - source_id: PMID:10615123
    title: 'MLH3: a DNA mismatch repair gene associated with mammalian microsatellite instability.'
    statement: >-
      MLH3 binds MLH1 (MutLgamma) and a dominant-negative MLH3 causes
      microsatellite instability, supporting at most a minor or redundant
      somatic MMR role; MutLgamma is recorded as context, not as a part.
    supporting_text: Cells in culture stably expressing a dominant-negative MLH3 protein exhibit microsatellite instability.
  - source_id: PMID:11292842
    title: 'The interacting domains of three MutL heterodimers in man: hMLH1 interacts with 36 homologous amino acid residues within hMLH3, hPMS1 and hPMS2.'
    statement: >-
      MLH3, PMS1 and PMS2 compete for the same MLH1 interaction region, so
      MutLbeta and MutLgamma draw on the MLH1 pool that MutLalpha needs.
    supporting_text: Furthermore, we show that there is competition for the interacting domain in hMLH1 among the three other MutL homologues.
notes: >-
  Derived from the RGD Pathway Ontology term PW:0000662 (mismatch repair
  pathway) through the RGD_PATHWAY_ONTOLOGY project (seed in
  projects/RGD_PATHWAY_ONTOLOGY/data/module_seeds.yaml, tier
  related_module_only; mapped GO process mismatch repair, skos:closeMatch; the
  PW page also cross-references KEGG map03430). RGD curates 24 human genes to
  PW:0000662: EXO1 HMGB1 LIG1 MLH1 MLH3 MSH2 MSH3 MSH6 PCNA PMS1 PMS2 POLD1
  POLD2 POLD3 POLD4 RFC1 RFC2 RFC3 RFC4 RFC5 RPA1 RPA2 RPA3 RPA4. Twenty-one are
  modelled as participants; MLH3, PMS1 and RPA4 are context (decisions 1 to 3).
  The parts follow the human reaction reconstituted from purified proteins
  (PMID:16143102, PMID:16873062), which uses exactly the RGD set minus those
  three. Molecular functions were taken from each protein's experimental GO
  annotations in QuickGO (November 2026); departures are stated on the
  annotons (PCNA processivity factor activity is IBA-supported; MutSbeta uses
  the parent term mismatched DNA binding). Gene reviews exist only for HMGB1
  and POLD2.

  Boundary decisions. (1) MutLbeta (MLH1-PMS1) is context, not a part. It forms
  in human cells (PMID:10748105), but no repair activity has been shown for it,
  PMS1 is not needed in the reconstituted reaction, and PMS1 germline variants
  are not established Lynch syndrome alleles. (2) MutLgamma (MLH1-MLH3) is
  context, not a part. MLH3 carries the conserved endonuclease motif, and a
  dominant-negative MLH3 causes microsatellite instability (PMID:10615123),
  but its established role is meiotic crossover formation, and any somatic MMR
  role is minor or a backup for MutLalpha; it would be a variant of the
  incision step only if purified human MutLgamma were shown to support
  repair. MLH3, PMS1 and PMS2 compete for the same MLH1 region
  (PMID:11292842). (3) RPA4 is excluded. It is an RPA2 paralogue that forms an
  alternative RPA complex that does not support DNA replication, and there is
  no evidence that it acts in MMR; the canonical RPA1-RPA3 heterotrimer is the
  RPA of the reconstituted reaction (PMID:9430682). (4) HMGB1 is included as an
  accessory annoton in the excision step, with positive regulation of mismatch
  repair as its process (accepted in the HMGB1 gene review), because it
  stimulates MutSalpha-activated excision but is not an essential catalytic
  participant. (5) EXO1-independent excision is context. Excision from a
  MutLalpha nick without EXO1, by POLD strand displacement or another
  nuclease, is described in the literature (PMID:33510387 notes unidentified
  nucleases) but has no defined human participant set, so it is not modelled
  as an alternative route and is recorded as a knowledge gap. (6) PCNA and RFC
  are modelled as one strand-discrimination step that feeds both incision and
  resynthesis; PCNA is not duplicated in the resynthesis step. (7)
  Out of scope: polymerase proofreading (DNA replication proofreading),
  MMR-dependent DNA damage signalling and apoptosis in response to alkylating
  agents, meiotic mismatch repair and crossover resolution, somatic
  hypermutation and class-switch recombination (which use MutSalpha, EXO1 and
  polymerase eta mutagenically), and the MCM9 helicase and WRN, which modulate
  MMR but are not RGD members. MSH4 and MSH5 are meiotic MutS homologues and
  are not MMR participants.

  Related modules. No human DNA replication, PCNA, RFC or polymerase delta
  module exists in modules/ (searched for replication, PCNA, RFC, polymerase,
  ligase, clamp and repair). Drosophila modules on the shared machinery exist
  by name only: dmel_dna_polymerase_complexes, dmel_elg1_rfc_like_complex and
  dmel_9_1_1_checkpoint_clamp. The bacterial analogue is
  muth_independent_mismatch_repair (MutS, MutL endonuclease directed by the
  beta clamp, UvrD, SSB and exonucleases), which shares the mapped GO process
  but not the taxon; the two are not linked as variants because their
  participants are not orthologous at the helicase and excision steps.

  GO-CAM. No GO-CAM model in gocams/index.tsv contains a mismatch repair
  activity for any member. The seed's best GO-CAM lead
  (gomodel:5fce9b7300002336, SPModule-TREM1-DAP12) overlaps only through
  HMGB1 as an extracellular receptor ligand and is unrelated; EXO1 and RPA1
  appear only in DNA end-resection and homologous recombination models. No
  gocam_associations are recorded.

  Observations for upstream review. (a) PMS1 carries contributes_to DNA
  endonuclease activity (IDA, PMID:33510387) in human GOA, but the full text of
  that paper assays human MLH1-PMS2 and mentions PMS1 only as the yeast MutLalpha
  subunit (yeast MLH1-PMS1), so the row may be a yeast-human naming confusion.
  (b) PCNA carries purine-specific mismatch base pair DNA N-glycosylase
  activity (IDA, PMID:11005803), but the abstract of that paper reports PCNA
  binding to MSH2-MSH6 and MSH2-MSH3 and no glycosylase assay; this looks like
  a mis-selected term. (c) MLH1 and PMS2 carry meiotic mismatch repair (IDA,
  ComplexPortal) from PMID:16873062, a study of somatic MMR in human cell
  extracts; only the abstract was checked. (d) LIG1 has mismatch repair only by
  TAS and POLD subunits have no experimental mismatch repair row, although
  both are required in the reconstituted reaction (PMID:16143102). PANTHER
  family ids and PTNs are deliberately not asserted.
module:
  id: eukaryotic_mismatch_repair
  label: Human DNA mismatch repair
  module_type: BIOLOGICAL_PROCESS
  concepts:
    - preferred_term: mismatch repair
      term:
        id: GO:0006298
        label: mismatch repair
      description: >-
        The GO process mapped (skos:closeMatch) to PW:0000662. Here the
        human, nick-directed, post-replicative reaction.
  context:
    taxa:
      - preferred_term: Homo sapiens
        term:
          id: NCBITaxon:9606
          label: Homo sapiens
    cellular_components:
      - preferred_term: nucleus
        term:
          id: GO:0005634
          label: nucleus
  parts:
    - order: 1
      role: mismatch recognition
      node:
        id: mismatch_recognition
        label: Mismatch recognition by a MutS homologue heterodimer
        module_type: MOLECULAR_FUNCTION
        description: >-
          MSH2 partners with MSH6 (MutSalpha) or MSH3 (MutSbeta). The two
          complexes have overlapping but distinct specificities: MutSalpha
          for base-base mismatches and small loops, MutSbeta for larger
          insertion/deletion loops. Both carry PCNA-binding motifs in their
          MSH6 or MSH3 subunit.
        variant_sets:
          - id: muts_heterodimer
            label: MutS homologue heterodimer
            axis: mismatch type (base-base mismatches and small loops versus larger insertion/deletion loops)
            variant_kind: REALIZER
            selection: ONE_OR_MORE
            evidence:
              - source_id: PMID:8942985
                statement: >-
                  MSH2-MSH3 and MSH2-MSH6 bind mismatched DNA with overlapping
                  but not identical specificity.
                supporting_text: Analysis of the mismatched nucleotide-binding specificity of the hMSH2-hMSH3 and hMSH2-hMSH6 protein complexes showed that they have overlapping but not identical binding specificity.
              - source_id: PMID:16143102
                statement: In the reconstituted reaction MutSbeta is the preferred sensor for insertion/deletion mispairs.
                supporting_text: In this system, MutSbeta plays a limited role in repair of base-base mismatches, but it processes insertion/deletion mispairs much more efficiently than MutSalpha, which efficiently corrects both types of heteroduplexes.
            variants:
              - id: mutsalpha_variant
                label: MutSalpha (MSH2-MSH6)
                module_type: PROTEIN_COMPLEX
                annotons:
                  - id: mutsalpha_mismatch_sensor
                    label: MutSalpha ATP-dependent mismatch sensor
                    participant:
                      selector_type: PROTEIN_COMPLEX
                      protein_complex:
                        preferred_term: MutSalpha (MSH2-MSH6)
                        term:
                          id: GO:0032301
                          label: MutSalpha complex
                        active_units:
                          - id: mutsalpha_msh2
                            label: MSH2 subunit
                            role: Shared MutS-homologue subunit; ATPase and mismatch-binding partner in both MutSalpha and MutSbeta.
                            participant:
                              selector_type: GENE_PRODUCT
                              gene_product:
                                preferred_term: MSH2
                                term:
                                  id: UniProtKB:P43246
                                  label: DNA mismatch repair protein Msh2
                          - id: mutsalpha_msh6
                            label: MSH6 subunit
                            role: >-
                              Mismatch-contacting subunit that confers base-base specificity; its
                              PWWP domain reads H3K36me3 and its N-terminal motif binds PCNA.
                            participant:
                              selector_type: GENE_PRODUCT
                              gene_product:
                                preferred_term: MSH6
                                term:
                                  id: UniProtKB:P52701
                                  label: DNA mismatch repair protein Msh6
                            function:
                              preferred_term: histone H3K36me3 reader activity
                              term:
                                id: GO:0140003
                                label: histone H3K36me3 reader activity
                              evidence:
                                - source_id: PMID:23622243
                                  statement: The MSH6 PWWP domain binds H3K36me3 and recruits MutSalpha to chromatin.
                                  supporting_text: Here, we show that an epigenetic histone mark, H3K36me3, is required in vivo to recruit the mismatch recognition protein hMutSα (hMSH2-hMSH6) onto chromatin through direct interactions with the hMSH6 PWWP domain.
                    function:
                      preferred_term: ATP-dependent DNA damage sensor activity
                      term:
                        id: GO:0140664
                        label: ATP-dependent DNA damage sensor activity
                    processes:
                      - preferred_term: mismatch repair
                        term:
                          id: GO:0006298
                          label: mismatch repair
                    role_description: >-
                      Binds a base-base mismatch or a one- or two-nucleotide loop and, after
                      ATP binding, recruits MutLalpha to form the initiation complex;
                      ATP-dependent release from the mismatch later gives EXO1 access. MSH2
                      and MSH6 carry this activity as contributes_to (IDA) in human GOA.
                    evidence:
                      - source_id: PMID:33510387
                        statement: Sliding of MutSalpha is needed for final mismatch removal rather than for initiation.
                        supporting_text: Using functional MMR assays and systems preventing proteins from sliding, we show that sliding of human MutSα is required not for MMR initiation, but for final mismatch removal.
              - id: mutsbeta_variant
                label: MutSbeta (MSH2-MSH3)
                module_type: PROTEIN_COMPLEX
                annotons:
                  - id: mutsbeta_loop_sensor
                    label: MutSbeta insertion/deletion loop recognition
                    participant:
                      selector_type: PROTEIN_COMPLEX
                      protein_complex:
                        preferred_term: MutSbeta (MSH2-MSH3)
                        term:
                          id: GO:0032302
                          label: MutSbeta complex
                        active_units:
                          - id: mutsbeta_msh2
                            label: MSH2 subunit
                            role: Shared MutS-homologue subunit.
                            participant:
                              selector_type: GENE_PRODUCT
                              gene_product:
                                preferred_term: MSH2
                                term:
                                  id: UniProtKB:P43246
                                  label: DNA mismatch repair protein Msh2
                          - id: mutsbeta_msh3
                            label: MSH3 subunit
                            role: Loop-specificity subunit; carries an N-terminal PCNA-binding motif.
                            participant:
                              selector_type: GENE_PRODUCT
                              gene_product:
                                preferred_term: MSH3
                                term:
                                  id: UniProtKB:P20585
                                  label: DNA mismatch repair protein Msh3
                    function:
                      preferred_term: mismatched DNA binding
                      term:
                        id: GO:0030983
                        label: mismatched DNA binding
                    processes:
                      - preferred_term: mismatch repair
                        term:
                          id: GO:0006298
                          label: mismatch repair
                    role_description: >-
                      Binds insertion/deletion loops of one or more extra nucleotides
                      (+G, +CA, +(CA)5 in the cited gel-shift assay) but not a G/T
                      mismatch. MSH3 carries mismatched DNA binding and its children
                      dinucleotide insertion or deletion binding and dinucleotide repeat
                      insertion binding as contributes_to (IDA, PMID:8942985), and a NOT
                      annotation to guanine/thymine mispair binding; the parent term is
                      used because no experimentally annotated term covers all loop sizes.
    - order: 2
      role: PCNA loading by RFC (strand discrimination signal)
      node:
        id: pcna_loading
        label: RFC loading of the PCNA clamp at the strand discontinuity
        module_type: BIOLOGICAL_PROCESS
        description: >-
          RFC binds PCNA with ATP, opens it and closes it around primer-template
          or nicked DNA in a fixed orientation. The oriented clamp both activates
          the MutLalpha endonuclease and directs it to the discontinuous strand,
          and it is reused as the processivity factor for POLD in resynthesis.
          PCNA also binds MSH6 and MSH3.
        annotons:
          - id: rfc_clamp_loader
            label: RFC (RFC1-RFC5) DNA clamp loader
            participant:
              selector_type: PROTEIN_COMPLEX
              protein_complex:
                preferred_term: replication factor C (RFC1-RFC5)
                term:
                  id: GO:0005663
                  label: DNA replication factor C complex
                active_units:
                  - id: rfc_rfc1
                    label: RFC1 large subunit
                    role: Large subunit; DNA-binding and clamp-interacting.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RFC1
                        term:
                          id: UniProtKB:P35251
                          label: Replication factor C subunit 1
                  - id: rfc_rfc2
                    label: RFC2 subunit
                    role: AAA+ ATPase small subunit.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RFC2
                        term:
                          id: UniProtKB:P35250
                          label: Replication factor C subunit 2
                  - id: rfc_rfc3
                    label: RFC3 subunit
                    role: AAA+ ATPase small subunit.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RFC3
                        term:
                          id: UniProtKB:P40938
                          label: Replication factor C subunit 3
                  - id: rfc_rfc4
                    label: RFC4 subunit
                    role: AAA+ ATPase small subunit.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RFC4
                        term:
                          id: UniProtKB:P35249
                          label: Replication factor C subunit 4
                  - id: rfc_rfc5
                    label: RFC5 subunit
                    role: AAA+ ATPase small subunit.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RFC5
                        term:
                          id: UniProtKB:P40937
                          label: Replication factor C subunit 5
            function:
              preferred_term: DNA clamp loader activity
              term:
                id: GO:0003689
                label: DNA clamp loader activity
              targets:
                - preferred_term: PCNA
                  term:
                    id: UniProtKB:P12004
                    label: DNA sliding clamp PCNA
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              RFC2-RFC5 carry DNA clamp loader activity as contributes_to (IDA,
              PMID:12930902, an assay of the RFC2-5 core shared with the
              CTF18-RFC-like complex). RFC is required for MutLalpha activation.
            evidence:
              - source_id: PMID:16873062
                statement: RFC is one of the four factors needed to activate MutLalpha incision.
                supporting_text: We show that human MutLalpha is a latent endonuclease that is activated in a mismatch-, MutSalpha-, RFC-, PCNA-, and ATP-dependent manner.
          - id: pcna_clamp
            label: PCNA sliding clamp
            participant:
              selector_type: PROTEIN_COMPLEX
              protein_complex:
                preferred_term: PCNA homotrimer
                term:
                  id: GO:0043626
                  label: PCNA complex
                active_units:
                  - id: pcna_protomer
                    label: PCNA protomer
                    role: Forms the homotrimeric ring; binds PIP-box motifs in MSH6, MSH3, MLH1, PMS2 and POLD.
                    stoichiometry: "3"
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: PCNA
                        term:
                          id: UniProtKB:P12004
                          label: DNA sliding clamp PCNA
            function:
              preferred_term: DNA polymerase processivity factor activity
              term:
                id: GO:0030337
                label: DNA polymerase processivity factor activity
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              The oriented clamp activates and directs MutLalpha incision and is
              the processivity factor for POLD during resynthesis. Human GOA has no
              experimental molecular function for PCNA that captures the MMR role
              (its experimental MF rows are bindings, plus a purine-specific
              mismatch DNA N-glycosylase row that looks erroneous); the processivity
              factor activity used here is IBA-supported. GO has no term for
              activation of the MutLalpha endonuclease by PCNA.
            evidence:
              - source_id: PMID:11005803
                statement: PCNA binds MSH2-MSH6 and MSH2-MSH3, and the interaction is needed before resynthesis.
                supporting_text: When human MSH3 or MSH6 peptides containing the PCNA binding motif were added to a human cell extract, mismatch repair activity was inhibited at a step preceding DNA resynthesis.
    - order: 3
      role: strand-directed incision
      node:
        id: mutlalpha_incision
        label: MutLalpha (MLH1-PMS2) endonuclease nicking of the daughter strand
        module_type: REACTION
        description: >-
          The MutS-MutLalpha initiation complex, with oriented PCNA, makes
          incisions on the discontinuous strand, chiefly 5' to the mismatch,
          giving EXO1 an entry point close to the error. The endonuclease
          active site is the PMS2 DQHA(X)2E(X)4E motif.
        annotons:
          - id: mutlalpha_endonuclease
            label: MutLalpha latent DNA endonuclease
            participant:
              selector_type: PROTEIN_COMPLEX
              protein_complex:
                preferred_term: MutLalpha (MLH1-PMS2)
                term:
                  id: GO:0032389
                  label: MutLalpha complex
                active_units:
                  - id: mutlalpha_mlh1
                    label: MLH1 subunit
                    role: Shared MutL-homologue subunit; ATPase and dimerisation partner of PMS2, PMS1 and MLH3.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: MLH1
                        term:
                          id: UniProtKB:P40692
                          label: DNA mismatch repair protein Mlh1
                  - id: mutlalpha_pms2
                    label: PMS2 catalytic subunit
                    role: Carries the endonuclease active-site motif.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: PMS2
                        term:
                          id: UniProtKB:P54278
                          label: Mismatch repair endonuclease PMS2
            function:
              preferred_term: DNA endonuclease activity
              term:
                id: GO:0004520
                label: DNA endonuclease activity
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              MLH1 carries DNA endonuclease activity as contributes_to (IDA,
              PMID:33510387) and PMS2 as contributes_to (IBA). MutLalpha also
              terminates EXO1 excision once the mismatch is removed.
            evidence:
              - source_id: PMID:16873062
                statement: The endonuclease active site is in PMS2.
                supporting_text: The probable endonuclease active site has been localized to a PMS2 DQHA(X)(2)E(X)(4)E motif.
              - source_id: PMID:16873062
                statement: Incision is directed to the nicked strand.
                supporting_text: Incision of a nicked mismatch-containing DNA heteroduplex by this four-protein system is strongly biased to the nicked strand.
    - order: 4
      role: excision of the error-containing tract
      node:
        id: exo1_excision
        label: EXO1 5'-to-3' excision from the nick, stimulated by HMGB1
        module_type: REACTION
        description: >-
          EXO1, activated and made processive by MutSalpha, hydrolyses the
          incised strand 5' to 3' from a nick and past the mismatch. Excision
          is terminated by MutLalpha and by RPA after the mispair is removed.
          HMGB1 stimulates MutSalpha-activated excision. EXO1-independent
          excision, which explains the weak mutator phenotype of EXO1 loss, is
          recorded in notes as context and not modelled as a route.
        annotons:
          - id: exo1_exonuclease
            label: EXO1 double-stranded DNA 5'-3' exonuclease
            participant:
              selector_type: GENE_PRODUCT
              gene_product:
                preferred_term: EXO1
                term:
                  id: UniProtKB:Q9UQ84
                  label: Exonuclease 1
            function:
              preferred_term: double-stranded DNA 5'-3' DNA exonuclease activity
              term:
                id: GO:0051908
                label: double-stranded DNA 5'-3' DNA exonuclease activity
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              The only exonuclease in the reconstituted human reaction. EXO1
              carries this activity by IDA (PMID:10608837, PMID:11842105) and
              mismatch repair by IDA and IMP.
            evidence:
              - source_id: PMID:14636568
                statement: MutSalpha activates EXO1 and excision stops after the mispair is removed.
                supporting_text: In the presence of RPA, ATP, and a mismatch, MutSalpha activates 5'-->3' excision by EXOI, and excision terminates after removal of the mispair.
              - source_id: PMID:11809771
                statement: EXO1 is the excision activity of human nuclear extracts.
                supporting_text: In systems comprised of pure components, the 5' to 3' hydrolytic activity of hEXOI is activated by hMutSalpha in a mismatch-dependent manner.
          - id: hmgb1_excision_cofactor
            label: HMGB1 stimulation of MutSalpha-activated excision
            participant:
              selector_type: GENE_PRODUCT
              gene_product:
                preferred_term: HMGB1
                term:
                  id: UniProtKB:P09429
                  label: High mobility group protein B1
            function:
              preferred_term: DNA binding, bending
              term:
                id: GO:0008301
                label: DNA binding, bending
            processes:
              - preferred_term: positive regulation of mismatch repair
                term:
                  id: GO:0032425
                  label: positive regulation of mismatch repair
            role_description: >-
              Accessory chromatin protein that binds MutSalpha and is required
              before excision in depleted extracts; it acts with RPA to stimulate
              excision. Its DNA bending activity (IMP, PMID:19223331) was shown in
              a transcriptional setting, and how it acts in MMR is not known, so the
              annoton carries the regulatory process accepted in the HMGB1 review
              rather than a core MMR process.
            evidence:
              - source_id: PMID:15014079
                statement: HMGB1 binds MutSalpha and acts before excision.
                supporting_text: Evidence is also presented that HMGB1 physically interacts with MutSalpha and is required at a step prior to the excision of mispaired nucleotide in mismatch repair.
              - source_id: PMID:16143102
                statement: RPA and HMGB1 stimulate excision with complementary roles.
                supporting_text: RPA and HMGB1 play similar but complementary roles in stimulating MutSalpha-activated, EXO1-catalyzed excision in the presence of a mismatch
              - source_id: file:human/HMGB1/HMGB1-ai-review.yaml
                statement: The HMGB1 review accepts positive regulation of mismatch repair (IDA, PMID:15014079).
    - order: 5
      role: single-stranded gap protection
      node:
        id: rpa_gap_protection
        label: RPA coating of the excision gap
        module_type: MOLECULAR_FUNCTION
        annotons:
          - id: rpa_ssdna_binding
            label: RPA heterotrimer single-stranded DNA binding
            participant:
              selector_type: PROTEIN_COMPLEX
              protein_complex:
                preferred_term: replication protein A (RPA1-RPA3)
                term:
                  id: GO:0005662
                  label: DNA replication factor A complex
                active_units:
                  - id: rpa_rpa1
                    label: RPA1 (RPA70) subunit
                    role: Main ssDNA-binding subunit; its zinc-finger motif is required for MMR in vitro.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RPA1
                        term:
                          id: UniProtKB:P27694
                          label: Replication protein A 70 kDa DNA-binding subunit
                  - id: rpa_rpa2
                    label: RPA2 (RPA32) subunit
                    role: ssDNA-binding and regulatory (phosphorylated) subunit.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RPA2
                        term:
                          id: UniProtKB:P15927
                          label: Replication protein A 32 kDa subunit
                  - id: rpa_rpa3
                    label: RPA3 (RPA14) subunit
                    role: Structural subunit of the trimerisation core.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: RPA3
                        term:
                          id: UniProtKB:P35244
                          label: Replication protein A 14 kDa subunit
            function:
              preferred_term: single-stranded DNA binding
              term:
                id: GO:0003697
                label: single-stranded DNA binding
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              Coats the excision gap, stimulates excision, helps terminate it past
              the mismatch, and supports resynthesis by POLD. RPA1, RPA2 and RPA3
              each carry single-stranded DNA binding (IDA) and mismatch repair (IMP,
              PMID:9430682).
            evidence:
              - source_id: PMID:9430682
                statement: Wild-type RPA is required for mismatch repair in vitro.
                supporting_text: We report for the first time that wild type RPA is required for a mismatch repair reaction in vitro.
              - source_id: PMID:16143102
                statement: RPA has a distinct role in excision termination.
                supporting_text: RPA has a distinct role in facilitating MutLalpha-mediated excision termination past mismatch.
    - order: 6
      role: gap-filling DNA synthesis
      node:
        id: pold_resynthesis
        label: DNA polymerase delta resynthesis with PCNA
        module_type: REACTION
        annotons:
          - id: pold_polymerase
            label: DNA polymerase delta (POLD1-POLD4) gap filling
            participant:
              selector_type: PROTEIN_COMPLEX
              protein_complex:
                preferred_term: DNA polymerase delta holoenzyme (POLD1-POLD4)
                term:
                  id: GO:0043625
                  label: delta DNA polymerase complex
                active_units:
                  - id: pold_pold1
                    label: POLD1 catalytic subunit
                    role: Polymerase and 3'-5' proofreading exonuclease.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: POLD1
                        term:
                          id: UniProtKB:P28340
                          label: DNA polymerase delta catalytic subunit
                    function:
                      preferred_term: DNA-directed DNA polymerase activity
                      term:
                        id: GO:0003887
                        label: DNA-directed DNA polymerase activity
                  - id: pold_pold2
                    label: POLD2 (p50) subunit
                    role: Non-catalytic B-subunit bridging POLD1 and POLD3; contributes to the holoenzyme's DNA polymerase activity (POLD2 gene review core function).
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: POLD2
                        term:
                          id: UniProtKB:P49005
                          label: DNA polymerase delta subunit 2
                  - id: pold_pold3
                    label: POLD3 (p66) subunit
                    role: Carries a PIP box that tethers the holoenzyme to PCNA.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: POLD3
                        term:
                          id: UniProtKB:Q15054
                          label: DNA polymerase delta subunit 3
                    function:
                      preferred_term: protein-macromolecule adaptor activity
                      term:
                        id: GO:0030674
                        label: protein-macromolecule adaptor activity
                  - id: pold_pold4
                    label: POLD4 (p12) subunit
                    role: Small subunit; degraded after DNA damage. No experimental molecular function in human GOA.
                    participant:
                      selector_type: GENE_PRODUCT
                      gene_product:
                        preferred_term: POLD4
                        term:
                          id: UniProtKB:Q9HCU8
                          label: DNA polymerase delta subunit 4
            function:
              preferred_term: DNA-directed DNA polymerase activity
              term:
                id: GO:0003887
                label: DNA-directed DNA polymerase activity
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              PCNA-dependent POLD fills the excision gap. POLD1 carries this
              activity by IMP (PMID:16762037). Polymerase epsilon is not part of
              the reconstituted reaction.
            evidence:
              - source_id: PMID:33510387
                statement: POLD fills the gap with PCNA and RPA, and LIG1 seals the nick.
                supporting_text: The DNA gap is filled by DNA polymerase δ in concerted reactions with PCNA and RPA, and the nick is ligated by ligase I.
              - source_id: file:human/POLD2/POLD2-ai-review.yaml
                statement: The POLD2 gene review covers the accessory subunit's role in the polymerase delta holoenzyme.
    - order: 7
      role: nick ligation
      node:
        id: lig1_ligation
        label: DNA ligase 1 sealing of the repaired strand
        module_type: REACTION
        annotons:
          - id: lig1_ligase
            label: LIG1 DNA ligase
            participant:
              selector_type: GENE_PRODUCT
              gene_product:
                preferred_term: LIG1
                term:
                  id: UniProtKB:P18858
                  label: DNA ligase 1
            function:
              preferred_term: DNA ligase activity
              term:
                id: GO:0003909
                label: DNA ligase activity
            processes:
              - preferred_term: mismatch repair
                term:
                  id: GO:0006298
                  label: mismatch repair
            role_description: >-
              Seals the final nick. LIG1 carries DNA ligase activity by IDA and IMP;
              the ATP-dependent child term DNA ligase (ATP) activity is IBA-only in
              human GOA. Its mismatch repair annotation is TAS (Reactome).
            evidence:
              - source_id: PMID:16143102
                statement: Ligase I is the ligase of the reconstituted human reaction.
                supporting_text: The reconstituted system includes MutSalpha or MutSbeta, MutLalpha, RPA, EXO1, HMGB1, PCNA, RFC, polymerase delta, and ligase I.
  connections:
    - source: mismatch_recognition
      target: mutlalpha_incision
      connection_type: POSITIVELY_REGULATES
      description: Mismatch-bound, ATP-loaded MutS recruits MutLalpha and is required to activate its endonuclease.
    - source: pcna_loading
      target: mutlalpha_incision
      connection_type: POSITIVELY_REGULATES
      description: Oriented PCNA activates MutLalpha and directs incision to the discontinuous strand.
    - source: mutlalpha_incision
      target: exo1_excision
      connection_type: PROVIDES_INPUT_FOR
      description: The MutLalpha nick 5' to the mismatch is the entry point for EXO1.
    - source: mismatch_recognition
      target: exo1_excision
      connection_type: POSITIVELY_REGULATES
      description: MutSalpha activates EXO1 and confers processivity on it.
    - source: rpa_gap_protection
      target: exo1_excision
      connection_type: NEGATIVELY_REGULATES
      description: RPA bound to the gap displaces EXO1 and helps terminate excision past the mismatch.
    - source: exo1_excision
      target: pold_resynthesis
      connection_type: PROVIDES_INPUT_FOR
      description: The RPA-coated single-stranded gap is the template for resynthesis.
    - source: pcna_loading
      target: pold_resynthesis
      connection_type: POSITIVELY_REGULATES
      description: PCNA is the processivity clamp for POLD.
    - source: pold_resynthesis
      target: lig1_ligation
      connection_type: PROVIDES_INPUT_FOR
      description: The nick left after gap filling is sealed by LIG1.
knowledge_gaps:
  - gap_statement: >-
      How the mismatch-bound MutS-MutLalpha complex communicates with the strand
      discontinuity that may lie hundreds of nucleotides away (MutS sliding,
      DNA looping, or MutLalpha recruitment by PCNA at the primer terminus) is
      not resolved, and nor is the strand signal on the leading strand, where
      no Okazaki-fragment ends exist (ribonucleotide-removal nicks have been
      proposed).
    gap_kind:
      - BIOLOGY
  - gap_statement: >-
      The identity and relative weight of EXO1-independent excision in human
      cells are unknown: whether it relies on strand displacement by POLD from
      MutLalpha nicks, on another nuclease, or on both, and why EXO1 loss gives
      only a mild mutator phenotype whereas MSH2, MLH1 or PMS2 loss gives a
      strong one.
    gap_kind:
      - BIOLOGY
  - gap_statement: >-
      Whether MutLgamma (MLH1-MLH3), which carries a conserved endonuclease
      motif, or MutLbeta (MLH1-PMS1) contributes any catalytic activity to
      somatic human mismatch repair, beyond competing with PMS2 for MLH1, has
      not been shown with purified human proteins.
    gap_kind:
      - BIOLOGY
