GPAT2 (donkey, A0A8C4KXU5) notes

2026-10-21 — Full re-review, independent of the ProtNLM2 predictions

Scope: full review of the 800-residue donkey GPAT2 entry A0A8C4KXU5 (Ensembl ENSEASP00005003122.2, gene
ENSEASG00005002382.2), the full-length model of the same gene as the 718-residue entry A0A9L0ILJ0 reviewed
in genes/EQUAS/GPAT2/. ProtNLM2 output was not opened or used.

Setup and deep research

• just fetch-gene EQUAS A0A8C4KXU5 --alias GPAT2__A0A8C4KXU5: 14 GOA rows, all IEA (Ensembl Compara
  GO_REF:0000107, TreeGrafter GO_REF:0000118, combined GO_REF:0000120, InterPro2GO, UniProt SubCell).
• just deep-research-falcon EQUAS GPAT2__A0A8C4KXU5 --fallback perplexity-lite failed: falcon
  "Failed to authenticate"; perplexity-lite HTTP 401 insufficient quota. No deep-research file exists.
  The review relies on the cached mouse primary papers below, read directly.

Evolutionary grounding

• A0A8C4KXU5, mouse Q14DK4, human Q6NUI2 and rat D3ZI76 are all PANTHER PTHR12563:SF15 (GPAT2,
  mitochondrial); GPAT1/GPAM (mouse Q61586, human Q9HCL2) is SF16
  (interpro/panther/PTHR12563/PTHR12563-entries.csv).
• Cached PAINT (PTHR12563-paint.tsv): GPAT1/GPAT2 node PTN000970957 (taxon:6072) carries GO:0004366,
  GO:0005739, GO:0006650, GO:0019432. No node carries GO:0034587 piRNA processing or GO:0003841 AGPAT
  activity. GO:0006631 fatty acid metabolic process is only on bacterial PlsB node PTN000285870.
• QuickGO for mouse Q14DK4: GPAT activity IDA/EXP; AGPAT activity IDA (PMID:22905194); PA biosynthesis IMP
  (PMID:22905194); piRNA processing IMP (PMID:23611983); mitochondrion IDA; mitochondrial outer membrane
  ISS/ISO. IBAs on mouse match the PAINT node terms only.
• Own analysis (GPAT2__A0A8C4KXU5-bioinformatics/RESULTS.md): 83.75% identity to mouse over 99.88% of
  its length (one unpaired mouse residue, 363); HXXXXD = HKSLLD at 205-210; both mouse TM segments align
  gap-free. Similar for human (88.04%) and rat (84.5%).

Literature (mouse Gpat2; nothing on donkey)

• PMID:17689486 and ["Confocal microscopy confirmed a mitochondrial location."] (abstract only).
• PMID:22905194; the authors hedge on AGPAT: ["These findings suggest that
  the increase in PA may be due to an AGPAT activity for this heterologously-expressed protein, but we
  cannot exclude the possibility that the protein has dual enzymatic activities."]
• PMID:22285183
• PMID:23611983 and ["These data suggest that the acyltransferase activity of GPAT2 is dispensable
  for the biogenesis of piRNA."]

Decisions

• GPAT activity: ACCEPT (core). PA biosynthetic process: ACCEPT (most specific catalytic process term).
• AGPAT activity: KEEP_AS_NON_CORE; crude-membrane activity in overexpressing cells, authors hedge, PAINT
  did not propagate it. Not removed because it rests on a curated mouse IDA.
• piRNA processing: ACCEPT. PAINT absence is read as a decision that it is not ancestral to GPAT1+GPAT2;
  the transfer stays within SF15, and the catalytic-motif-independent requirement plus MILI binding at the
  site of processing meet the participation (scaffold) bar. Single study, mouse GS cells.
• mitochondrion / mitochondrial membrane / membrane: MODIFY to GO:0005741 mitochondrial outer membrane.
• TAG biosynthesis, glycerol-3-phosphate metabolic process: KEEP_AS_NON_CORE. Fatty acid metabolic process:
  MARK_AS_OVER_ANNOTATED. lipid metabolic process, phospholipid biosynthetic process, acyltransferase
  activity (two rows): TOO_GENERAL.
• Everything above is inferred for donkey from mouse experiments plus subfamily membership; no donkey
  GPAT2 function has been measured.