NQO2 (P16083, NQO2_HUMAN) — review notes

Gene identity and core function

NQO2 = NRH:quinone oxidoreductase 2 / quinone reductase 2 (QR2); UniProt recommended name
Ribosyldihydronicotinamide dehydrogenase [quinone], EC 1.10.5.1. A 231-aa cytosolic FAD
flavoprotein, homodimer, member of the NAD(P)H:quinone oxidoreductase family (paralog of NQO1).

Defining biochemical distinction from NQO1: NQO2 does not use NAD(P)H efficiently as the
electron donor. Instead it uses dihydronicotinamide riboside (NRH) (and other reduced
N-ribosyl/N-alkyl dihydronicotinamides) as the reducing co-substrate.
PMID:10945627
UniProt MISCELLANEOUS: "Uses dihydronicotinamide riboside (NRH) rather than NAD(P)H as an electron donor."

• Catalysis: two-electron reduction of quinones to hydroquinones (quinols), NRH + quinone + H+ →
  N-ribosylnicotinamide + quinol (RHEA:12364; EC 1.10.5.1; GO:0001512). FAD-dependent.
• Cofactors: FAD (flavin) and one structural Zn2+ per subunit (UniProt COFACTOR).
• Biological role: quinone reduction/detoxification (produces less-toxic hydroquinones for
  conjugation), though notably QR2 can in some cases generate more reactive species than the
  parent quinone PMID:18254726. Also implicated in prodrug bioactivation (CB1954, mitomycin C).

Pharmacology — the MT3 melatonin site and resveratrol

QR2 is the long-sought cytosolic "MT3" melatonin binding site.
PMID:18254726
Melatonin is a competitive inhibitor of QR2 (vs the dihydronicotinamide donor), binding in the
active site (not allosteric): PMID:18254726 and X-ray structures show
PMID:18254726.

QR2 is also one of the highest-affinity molecular targets of resveratrol (an inhibitor here,
in contrast to cao-1 where resveratrol is a substrate):
PMID:18254726.

So the melatonin binding and resveratrol binding IDA annotations reflect inhibitor/ligand
binding at the active site, well-evidenced (co-crystals, ITC/kinetics) but pharmacological rather
than the core catalytic function → KEEP_AS_NON_CORE.

Provenance of the resveratrol binding IDA (GO:1905594 → PMID:18254726): In the annotated
paper the direct evidence is a binding assay, not a structure — ITC gave a Kd of ~39 nM
PMID:18254726, plus IC50
inhibition assays; resveratrol served as a positive control there. The crystal structure of the
QR2:resveratrol complex is from an earlier paper — Buryanovskyy et al. 2004 (PDB 1SG0; PMID:15350128),
who also first measured the Kd (~35 nM) by intrinsic tryptophan fluorescence — which Calamini et al.
reuse for molecular replacement/docking. Buryanovskyy 2004 is a candidate reference to add (it is the
primary structural + first-binding source but is absent from NQO2's GOA reference set).

Annotation review decisions (26 GOA annotations)

Molecular function — catalytic

• GO:0001512 dihydronicotinamide riboside quinone reductase activity (IDA PMID:18254726; IEA) →
  ACCEPT (core). Exactly matches EC 1.10.5.1 / RHEA:12364.
• GO:0003955 NAD(P)H dehydrogenase (quinone) activity (IBA) → MODIFY → GO:0001512. This term
  maps to EC 1.6.5.2 (NAD(P)H donor); NQO2 uses NRH, not NAD(P)H. Family IBA propagated the
  NQO1-type activity. Replace with the NRH-specific term.
• GO:0016661 oxidoreductase activity, acting on other nitrogenous compounds as donors (IDA
  PMID:10945627) → MODIFY → GO:0001512. Maps to EC 1.7.-.- (nitrogenous donors), a different
  branch than NQO2's EC 1.10.5.1; the specific NRH:quinone reductase term is the correct classification.
• GO:0016491 oxidoreductase activity (IDA PMID:10945627) → KEEP_AS_NON_CORE. Correct root
  term but uninformatively general.
• GO:0009055 electron transfer activity (IDA PMID:10945627, PMID:18254726) → MARK_AS_OVER_ANNOTATED.
  This term denotes electron-carrier proteins in electron transport chains; QR2 is a soluble
  two-electron quinone reductase, and its electron flow is already captured by the quinone reductase MF.

Molecular function — cofactor/ligand binding

• GO:0071949 FAD binding (IDA) → ACCEPT (core cofactor; flavoprotein).
• GO:0008270 zinc ion binding (IDA) → ACCEPT (structural Zn2+, 1/subunit per UniProt).
• GO:0031404 chloride ion binding (IDA) → KEEP_AS_NON_CORE. Not a recognized functional
  cofactor (absent from UniProt COFACTOR); likely a crystallographic ion.
• GO:1904408 melatonin binding (IDA) → KEEP_AS_NON_CORE (MT3 site; active-site competitive inhibitor).
• GO:1905594 resveratrol binding (IDA) → KEEP_AS_NON_CORE (nanomolar inhibitor binding).
• GO:0042803 protein homodimerization activity (IPI PMID:18254726) → ACCEPT (genuine homodimer;
  informative, unlike bare protein binding). Non-core.

Molecular function — protein binding (5× IPI, GO:0005515)

• PMID:16189514, 25416956, 31515488, 32296183 = large-scale interactome (Y2H) screens;
  PMID:24606901 = HSCB/HSC20 LYR-motif cochaperone study. All → KEEP_AS_NON_CORE. protein binding
  is uninformative (project guideline: avoid); none identifies a functional partnership that revises
  the core-function picture.

Biological process

• GO:1901662 quinone catabolic process (IDA PMID:18254726) → ACCEPT. Captures the quinone
  reduction/detoxification role (main BP).

Cellular component

• GO:0005829 cytosol (IDA GO_REF:0000052; also TAS Reactome, IBA, IDA PMID:10945627) → ACCEPT (core).
  QR2 is cytosolic.
• GO:0005737 cytoplasm (IEA) → ACCEPT (general parent).
• GO:0070062 extracellular exosome (HDA PMID:23533145, PMID:19056867) → KEEP_AS_NON_CORE.
  High-throughput exosome proteomics detection; not a functionally informative localization for a
  cytosolic enzyme.

Enzyme-function / Rhea comparison

• Core MF GO:0001512 carries EC=1.10.5.1, RHEA:12364, MetaCyc 1.10.99.2-RXN, Reactome R-HSA-8936519 —
  a fully-resolved, reaction-specific term (unlike cao-1, where no reaction-level term existed).
• The IBA GO:0003955 carries EC=1.6.5.2 (NAD(P)H) — the wrong cofactor branch for NQO2; this is the
  crux of the MODIFY.
• Contrast with cao-1: there resveratrol is the substrate of a stilbenoid dioxygenase; here
  resveratrol is a high-affinity inhibitor of a quinone reductase — same molecule, opposite role,
  and a good illustration of why resveratrol binding alone (causal-role, ligand-agnostic) underspecifies biology.

Evidence re-review, 2026-09-20

Read all 26 annotation reviews with the primary reference abstracts, available structural/enzymatic evidence and PAINT PTHR10204 (GO:0003955 at PTN000024521). The older notes and existing OpenScientist report overstate the donor exclusion: PMID:35517822 explicitly describes NADH/NADPH usage as extremely inefficient and its Table 1 reports measurable human NADH turnover. Retain the broad NAD(P)H reaction as non-core while NRH-dependent chemistry remains defining. This distinction does not establish efficient cellular NAD(P)H use. The same paper reports that CB1954 activation in its cells required exogenous NRH. PMID:9367528 confirms strong NRH preference and shows that adding the NQO1 C terminus does not erase this preference. Current GO:0009055 includes enzymatic redox transfer; its two IDA rows should be accepted. The existing report was reused and its categorical conclusion flagged as disputed, with its exact recommendation retained for provenance. Generic interaction labels are uninformative; the unresolved HSC20/NQO2-specific binding experiment remains UNDECIDED. New sources: PMID:35517822(https://pubmed.ncbi.nlm.nih.gov/35517822/), PMID:9367528(https://pubmed.ncbi.nlm.nih.gov/9367528/).