Selected accession A0A9L0TTC1; ordinary UniProt record is TrEMBL and cites genome sequencing/submission rather than a gene-specific functional experiment. Human evidence is summarized in the paired human investigation. Sequence comparison records the exact target gaps and transfer limits. The targeted Europe PMC search found no decisive gene-specific horse functional assay warranting a separate horse Edison investigation; general omics/association hits were not counted as mechanistic validation.
The substrate donor is acyl-CoA and the acceptor is the sn-1 hydroxyl of glycerol-3-phosphate. Lysophosphatidic acid enters several glycerolipid routes, so pathway participation is broader than the one catalyzed reaction. The 2023 structural work reports an N-terminal catalytic domain and associated C-terminal domain, and replaces a conventional multi-pass-transmembrane picture with an amphipathic surface/N-terminal loop-helix membrane association mechanism. The UniProt record links this reaction to direct experimental sources; the acyl-CoA statement is therefore not an ARBA corroboration. Papers mainly describing another GPAT paralog may include GPAT1 comparator assays, so title alone cannot justify discarding their GO annotations.
Sources: PMID:36522428(https://pubmed.ncbi.nlm.nih.gov/36522428/), PMID:18238778(https://pubmed.ncbi.nlm.nih.gov/18238778/), PMID:19075029(https://pubmed.ncbi.nlm.nih.gov/19075029/).
Deep research. just deep-research-falcon HORSE GPAM --fallback perplexity-lite failed (falcon: "Failed to authenticate"; perplexity fallback: HTTP 401 insufficient quota). No horse deep-research file exists; literature context comes from the human GPAM review, its falcon report (genes/human/GPAM/GPAM-deep-research-falcon.md), the human Swiss-Prot record and PMID:36522428.
Identity of the benchmark accession. A0A9L0TTC1 (790 aa) is NOT the canonical horse GPAM product. The UniProt horse reference proteome labels it "Isoform of A0A9L0TCA8"; A0A9L0TCA8 (828 aa, same length as human Q9HCL2) is the gene-centric canonical entry. Ensembl gene ENSECAG00000014742 has nine protein-coding transcripts; A0A9L0TTC1 is ENSECAT00000105322 (the Ensembl-canonical transcript is ENSECAT00000137388 = A0A9L0TA25, 900 aa). A0A9L0TTC1 lacks an in-frame 38-residue segment equivalent to human 370-407, which is conserved in human, mouse, rat and canonical horse GPAM and which in the human cryo-EM structure contains resolved catalytic-domain secondary structure (helix 380-390, strand 396-400, helices 401-403, 406-415). The HXXXXD motif, all six CoA-binding residues and the mitochondrial-localization region are retained. Activity of this isoform is untested. See RESULTS-rereview.md.
Evolutionary grounding. PANTHER PTHR12563:SF16 (same subfamily as human GPAM). PAINT node PTN000970957 (Eumetazoa) carries IBD for GO:0004366, GO:0005739, GO:0006650 and GO:0019432; horse descends from it. No plasma-membrane term at any metazoan node; the plasma-membrane IEA derives from IPR028354 (GPAT_PlsB), shared with bacterial PlsB.
Changes vs previous review.
- gene_symbol corrected from the accession to GPAM.
- Description rewritten (project-independent; records the isoform status).
- GO:0006629 lipid metabolic process: ACCEPT -> MODIFY to GO:0019432 + GO:0008654 (correct-but-general rule; differs from the human review, which accepted it).
- GO:0031966 mitochondrial membrane: KEEP_AS_NON_CORE -> MODIFY to GO:0005741 (previous reason argued it was too general).
- GO:0004366, GO:0005741, GO:0008654, GO:0016024: ACCEPT, with gene-level/isoform caveat; GO:0005886 plasma membrane: REMOVE (bacterial PlsB mapping); GO:0016746/GO:0016747: MODIFY to GO:0004366.
- Support now cites PAINT/subfamily placement and feature retention (RESULTS-rereview.md) and the human Swiss-Prot curated text instead of a pairwise-identity statement; the older RESULTS.md is kept but no longer cited.
- core_functions adds triglyceride and phospholipid biosynthetic process as directly_involved_in; one suggested question and one experiment on the isoform.
Next relative. Canonical horse GPAM A0A9L0TCA8 has no review in the repo; its folder would be genes/HORSE/GPAM__A0A9L0TCA8. The PTHR12563 family review lists A0A9L0TTC1 as an SF16 representative; subfamily placement agrees, but the canonical A0A9L0TCA8 would be a better representative.
Modules. modules/triacylglycerol_biosynthesis.yaml (GPAM is the animal G3P-acylation annoton, grounded on human Q9HCL2).