Pnkd (mouse, Q69ZP3) review notes

Deep research

just deep-research-falcon mouse Pnkd --fallback perplexity-lite was run on 2026-10-20 and
failed: falcon failed to authenticate ("Exception: Failed to authenticate") and the
perplexity-lite fallback returned HTTP 401 (insufficient quota). No deep-research file exists.
The review is based on the cached primary literature (PMID:21487022, PMID:25730884,
PMID:22214848 abstract only, PMID:18614015), the UniProt record and the GOA rows.

Evolutionary placement

• UniProt: PANTHER PTHR11935:SF116 "HYDROLASE PNKD-RELATED"; metallo-beta-lactamase
  superfamily, glyoxalase II family; InterPro IPR017782 (Hydroxyacylglutathione hydrolase) and
  Pfam HAGH_C. The PNKD subfamily is distinct from the HAGH (glyoxalase II) subfamily.
• Zinc ligands (by similarity to human HAGH Q16775) H172, H174, D176, H177, H229, D253, H291.
  I checked the sequence: residues 166-180 read VAILCTHKHWDHSGG, so the THxHxDH motif is
  intact. The protein is therefore plausibly catalytically competent.
• PAINT: IBA thiolester hydrolase activity (GO:0016790) from PTN004322580 and IBA mitochondrion
  from PTN000940186. PAINT did NOT propagate hydroxyacylglutathione hydrolase activity
  (GO:0004416) to the PNKD clade, although HAGH carries it. I treat that omission as deliberate:
  the PNKD subfamily is not inferred to be a glyoxalase II.
• Isoform 2 (MR-1S-like short isoform) lacks residues 143-385, i.e. the whole
  metallo-beta-lactamase domain, so it cannot carry the enzymatic activity.

Experimental evidence: what is established in mouse vs inferred from human

Mouse (experimental):
- Pnkd is neuronal and widely expressed in brain; one isoform is membrane-associated
PMID:22214848 (abstract only).
- Pnkd mutant knock-in/transgenic mice show reduced striatal extracellular dopamine and
enhanced dopamine release on caffeine/ethanol PMID:22214848. These are organismal/circuit
phenotypes; the IMP rows (regulation of dopaminergic transmission, dopamine metabolism,
posture) are acts_upstream_of_or_within and are not evidence of direct participation.
- Pnkd-L co-immunoprecipitates RIM1 and RIM2 from mouse frontal cortex; Pnkd localises to
presynaptic terminals by immunogold EM, absent in Pnkd-/-; RIM1/2 levels reduced in Pnkd-/-
cortex; Pnkd-/- has weaker synaptic strength and increased paired-pulse facilitation
PMID:25730884.
- Overexpressed (human) PNKD suppresses synaptic vesicle exocytosis in cultured hippocampal
neurons PMID:25730884. Note the tension with the KO,
which also weakens synaptic strength; the authors attribute the KO effect to RIM
destabilisation.
- Pnkd KO and mutant transgenic cortex have ~20% lower glutathione PMID:21487022.
- MitoCarta mass-spec (mouse tissues) detects Pnkd in mitochondria; per PMID:21487022 the
peptides support the short isoform PMID:21487022.

Human (experimental, transferable via 1:1 orthology in the same PANTHER subfamily):
- Purified human PNKD-L hydrolyses S-D-lactoylglutathione at an "extremely low but consistent"
rate, judged not physiological; human PNKD-L fails to rescue Drosophila hagh-null SLG
hydrolysis PMID:21487022. Authors propose an unknown alpha-hydroxythioester substrate.
- Human PNKD-L localises to the cell membrane and late endosomes in SH-SY5Y and Cos7 and does not
colocalise with mitochondrial markers PMID:21487022; another study (Ghezzi 2009, not cached)
reported mitochondrial localisation of the L and S isoforms and ER/Golgi/PM for the M isoform.
- N-terminal cleavage of PNKD-L (~15-30 aa) is blocked by disease mutations PMID:21487022.

Decisions

• GO:0016790 thiolester hydrolase activity (IBA): ACCEPT, core; inferred (not assayed in mouse).
• GO:0004416 HAGH activity (IEA InterPro2GO): MARK_AS_OVER_ANNOTATED; in vitro trace activity
  only on human PNKD-L, not physiological, PAINT declined to propagate it.
• GO:0051596 methylglyoxal catabolic process (IEA): REMOVE; human PNKD-L does not rescue SLG
  metabolism in vivo.
• Protein binding IPI (RIM1/RIM2): REMOVE per GO:0005515 policy; no specific MF term for RIM
  binding; interaction itself is real.
• Mitochondrion: ACCEPT (MitoCarta HDA, IBA, human EXP); likely isoform-dependent.
• Plasma membrane / membrane: ACCEPT/MODIFY to plasma membrane.
• Presynaptic cytosol IDA: ACCEPT (mouse immunogold).
• Cytoplasm: KEEP_AS_NON_CORE (short isoform cytoplasmic; general term).
• ER, Golgi (ISO from human PNKD-M overexpression): KEEP_AS_NON_CORE.
• Complex IV assembly ISO: UNDECIDED. The human source annotation is no longer present in
  current QuickGO for Q8N490 (checked 2026-10-20), and no cached publication supports it.
• Negative regulation of neurotransmitter secretion: KEEP_AS_NON_CORE (overexpression-based; IGI
  with RIM; regulatory role mechanism unclear).
• Dopaminergic transmission, dopamine metabolism, posture IMP: KEEP_AS_NON_CORE for the
  synaptic transmission term; MARK_AS_OVER_ANNOTATED for dopamine metabolic regulation and
  posture (organismal phenotypes of a knock-in mutant).