RRM2B (p53R2) review notes

UniProtKB: Q7LG56 (RIR2B_HUMAN). Gene: RRM2B / P53R2. 351 aa. EC 1.17.4.1.

Core biology (grounded in UniProt + cached PMIDs)

RRM2B encodes the p53-inducible small subunit of ribonucleotide reductase (p53R2).
It is a paralog of the cell-cycle-regulated small subunit RRM2, but unlike RRM2 it is
induced by p53/TP53 in response to DNA damage and is present in resting/quiescent cells
[PMID:10716435, PMID:11517226]. Together with the large catalytic subunit RRM1 it forms an
active ribonucleotide reductase (RNR) that reduces ribonucleoside diphosphates (NDPs) to
2'-deoxyribonucleoside diphosphates (dNDPs) — the rate-limiting de novo step in dNTP supply
for DNA synthesis and repair [PMID:11517226, PMID:16376858].

• EC 1.17.4.1 / RHEA:23252: "a 2'-deoxyribonucleoside 5'-diphosphate + [thioredoxin]-disulfide
• H2O = a ribonucleoside 5'-diphosphate + [thioredoxin]-dithiol" (UniProt CATALYTIC ACTIVITY).
• Cofactor: binds 2 iron ions per subunit; contains a diiron-tyrosyl free radical center
  required for catalysis [PMID:19728742 (crystal structure); PMID:11517226 (iron-tyrosyl radical)].
  UniProt Fe-binding residues: 100, 131, 134, 194, 228, 231; ACT_SITE 138.

Subcellular location

RNR is a cytosolic process. Pontarin et al. PMID:18997010 used three independent methods
and found R1/R2/p53R2 reside in the cytosol, arguing against the earlier nuclear-translocation
model and supporting a primary function of p53R2 for mtDNA replication. The earlier reports
described translocation cytoplasm->nucleus after DNA damage [PMID:11719458, PMID:12615712];
UniProt SUBCELLULAR LOCATION: "Cytoplasm. Nucleus. Note=Translocates from cytoplasm to nucleus
in response to DNA damage." HPA IDA supports cytosol + nucleoplasm.
Treat cytosol as the core location; nucleus/nucleoplasm/cytoplasm accepted as observed locations.
Mitochondrion IEA (GO_REF:0000108, inferred from GO:0006264) is not a documented physical
localization of the protein — RNR is cytosolic and supplies dNTPs that are imported into
mitochondria; mark over-annotated.

Disease / physiology

• MTDPS8A/8B (mitochondrial DNA depletion syndrome 8, encephalomyopathic +/- tubulopathy):
  RRM2B mutations cause severe mtDNA depletion; Rrm2b-/- mouse recapitulates this. p53R2 has a
  crucial role in dNTP supply for mtDNA synthesis, distinct from DNA repair PMID:17486094.
• PEOA5: autosomal dominant progressive external ophthalmoplegia with multiple mtDNA
  deletions PMID:19664747.
• RCDFRD: rod-cone dystrophy, deafness, Fanconi renal dysfunction PMID:32827185.
  These establish the physiological role: providing dNTPs for mitochondrial DNA replication/
  synthesis in post-mitotic/quiescent tissues (skeletal muscle high expression).

Annotation review reasoning

• MF GO:0004748 (ribonucleoside-diphosphate reductase activity, thioredoxin disulfide as
  acceptor): CORE. Direct IDA [PMID:16376858 catalytic activity; PMID:12615712], IBA, ISS, IEA
  all agree. Accept.
• GO:0016491 oxidoreductase activity (IEA, InterPro): correct but far too general given the
  specific GO:0004748 is present -> MARK_AS_OVER_ANNOTATED (parent of the specific term).
• BP dNTP synthesis: GO:0009263 (deoxyribonucleotide biosynthetic process), GO:0009265
  (2'-deoxyribonucleotide biosynthetic process), GO:0009185 (ribonucleoside diphosphate metabolic
  process) — all correct core/near-core process terms for RNR. Accept the specific ones; the
  ComplexPortal IDA on 0009265 and 0009185 are supported by PMID:16376858. GO:0009185 is somewhat
  general (metabolic vs biosynthetic) but correct.
• DNA repair GO:0006281 / GO:0000731 (DNA synthesis involved in DNA repair): supported by
  IDA PMID:11719458 and the p53-checkpoint literature. Accept as core/near-core (supplies dNTPs
  for repair). NAS on 0000731 from PMID:10716435 accepted.
• GO:0006264 mitochondrial DNA replication (IMP PMID:17486094; IEA): supported — p53R2 supplies
  dNTPs for mtDNA synthesis; loss causes mtDNA depletion. Accept (KEEP_AS_NON_CORE style: it is the
  in vivo physiological consequence; RRM2B itself doesn't replicate DNA but supplies precursors).
  Keep as core process — this is arguably the most important physiological role.
• Cell-cycle regulation terms: GO:1901992 (positive regulation of mitotic cell cycle phase
  transition, IEA/ARBA), GO:0010971 (positive regulation of G2/M transition, IDA PMID:10716435),
  GO:0070318 (positive regulation of G0 to G1 transition, IDA PMID:11517226 / IEA). PMID:10716435
  shows p53R2 induction causes G2/M arrest (a checkpoint), which is not the same as "positive
  regulation of G2/M transition". These are indirect/over-interpreted -> MARK_AS_OVER_ANNOTATED.
  GO:0070318 IDA is cited to PMID:11517226 which is an in-vitro RNR activity paper about resting
  cells; does not directly demonstrate positive regulation of G0->G1 -> MARK_AS_OVER_ANNOTATED.
• GO:0014075 response to amine (IEA from rat ortholog, GO_REF:0000107): weakly supported,
  electronic transfer from a rat paralog; peripheral -> MARK_AS_OVER_ANNOTATED.
• protein binding GO:0005515 (IPI): bare, uninformative. 12615712 = TP53 (functional/regulatory),
  19015526 = ATM/MDM2 (regulation of p53R2 stability). Per policy do NOT REMOVE bare protein-binding
  IPIs -> MARK_AS_OVER_ANNOTATED (keep the functional context in reason).
• identical protein binding GO:0042802 (IPI): RNR small subunits form homodimers (structural
  ferritin-like fold; UniProt INTERACTION lists Q7LG56-Q7LG56). Large-scale screens (25416956,
  25502805, 31515488). Uninformative for function -> MARK_AS_OVER_ANNOTATED.
• Cytosol/cytoplasm/nucleus/nucleoplasm: cytosol (IBA/IDA/TAS/HPA) core; others accepted as
  observed locations. Mitochondrion IEA over-annotated (see above).

core_functions (author-supplied, strictly validated)

• MF: GO:0004748.
• Process: GO:0009263 (deoxyribonucleotide biosynthetic process) — the de novo dNTP supply role.
• Process: GO:0006264 (mitochondrial DNA replication) — the key physiological role (dNTP supply
  for mtDNA synthesis; disease-defining).
• Location: GO:0005829 (cytosol).

QA re-review 2026-07-23 (conservative, no changes)

Independent QA pass over the completed 45-annotation review. Outcome: no edits made —
no problems met the bar for a confident conservative fix.

Checks performed:
- uv run ai-gene-review validate -> ✓ Valid, no warnings.
- protein binding (GO:0005515) x3 IPI and identical protein binding (GO:0042802) x3 IPI
are all MARK_AS_OVER_ANNOTATED (none ACCEPTed), each retained-not-removed per curation
policy with the functional context (TP53, ATM/MDM2, ORC4/RNF41, homodimer) noted in the
reason. Compliant.
- Core function correctly captured and not over-generalized: two core_functions entries both
keyed on MF GO:0004748 (ribonucleoside-diphosphate reductase activity) — (1) de novo dNTP
supply via GO:0009263, (2) dNTP supply for mtDNA replication via GO:0006264 — both located to
cytosol (GO:0005829). MF is in the MF branch, both process ids in BP, location in CC: no
wrong-branch author-supplied ids.
- All PMIDs cited in supporting_text are cached but abstract-only (full_text_available: false);
validation confirms every supporting_text is a verbatim substring of the cached abstract.
- GOA/review completeness: GOA has 46 annotation rows, review has 45. The single delta is a
legitimate WITH/FROM duplicate — GO:0005515 IPI PMID:19015526 has two rows differing only in
partner (MDM2 UniProtKB:Q00987, ATM UniProtKB:Q13315), correctly collapsed to one review
annotation whose summary already names both partners. No annotation is missing.
- Description is clean project-independent biology (no this review/curation/PN framing).

Considered but deliberately left alone (not confident improvements):
- Several non-core observed locations (nucleus, nucleoplasm, cytoplasm) use ACCEPT with reasons
that describe them as "non-core observed location". KEEP_AS_NON_CORE would be marginally more
literal, but ACCEPT-with-note is a consistent, acceptable convention here; not changed.
- Cell-cycle IDA terms (GO:0010971 positive reg G2/M; GO:0070318 positive reg G0->G1) are
MARK_AS_OVER_ANNOTATED with sound arguments (cited papers show G2/M arrest / in-vitro resting
cell dNTP supply, not the annotated positive regulation). A stronger REMOVE is disallowed for
experimental annotations whose full text we have not read; over-annotated is the correct
conservative call. Left as-is.