STX3 (donkey, A0A9L0J901) notes

Sequence and record

• Unreviewed TrEMBL entry, 289 aa, "Syntaxin-3"; PANTHER PTHR19957:SF34 (SYNTAXIN-3);
  Syntaxin N-terminal Habc region, t-SNARE (SNARE motif) domain 191..253 and a C-terminal
  transmembrane segment 264..288 (tail-anchored, type IV)
  [file:EQUAS/STX3/STX3-uniprot.txt "Single-pass type IV membrane protein"].
• 15 GOA rows, all IEA (TreeGrafter, InterPro2GO, ARBA and GO_REF:0000120 combined).
• Donkey vs human Q13277 (isoform A, 289 aa): 92.39% identity over the full length, no
  unpaired segments [file:EQUAS/STX3/STX3-bioinformatics/RESULTS.md "with 92.39% identity across 289 paired residues"].
  The donkey sequence corresponds to the canonical, membrane-anchored isoform A, not the
  retina-specific isoform B or the nuclear isoform 3.

Ortholog biology

• Human STX3: apical plasma-membrane t-SNARE mediating fusion of apical vesicles; loss causes
  microvillus inclusion disease and retinal dystrophy
  [file:EQUAS/STX3/STX3-bioinformatics/human_Q13277-uniprot.txt "Apical receptor involved in membrane fusion of apical vesicles"].
• Partners SNAP25 and VAMP2 (by similarity) [same file, "Interacts with SNAP25 (By similarity)"].
• Mouse Stx3 (Q64704) has SynGO IDA annotations to postsynapse and to exocytic insertion of
  neurotransmitter receptor to postsynaptic membrane, and MGI IDA to dendrite
  [file:EQUAS/STX3/STX3-bioinformatics/mouse_Q64704-uniprot.txt "C:postsynapse; IDA:SynGO."].
  The underlying SynGO paper (PMID:23395379) is not cached, so it is cited only through the
  mouse UniProt cross-reference.

Decisions

• Core: helical-bundle zippering fusogenic activity (t-SNARE), SNARE binding, SNARE complex,
  plasma membrane, vesicle fusion, exocytosis.
• Broad terms (membrane, vesicle-mediated transport, membrane fusion): TOO_GENERAL.
• Cytoplasm (ARBA): MARK_AS_OVER_ANNOTATED - an integral membrane protein, not a soluble
  cytoplasmic protein.
• Postsynaptic/dendrite/cell-projection terms and intracellular protein transport: non-core.

ProtNLM2 (donkey40 snapshot)

• GO:0016020 membrane -> LSP (GOA has plasma membrane).
• GO:0016192 vesicle-mediated transport -> LSP (GOA has vesicle fusion and exocytosis).
• Non-GO: subcellular_location "Membrane" (not scored).

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

Deep research: just deep-research-falcon EQUAS STX3 --fallback perplexity-lite failed
(falcon "Failed to authenticate"; perplexity-lite HTTP 401 insufficient quota). No
deep-research file exists; this re-review uses cached literature fetched with just fetch-pmid
(PMID:18683220, 20060037, 39408994, 7598732, 24726755, 33974130, 29475951, 23395379).

Identity of the benchmark accession (correction)

• The earlier notes called A0A9L0J901 the canonical, membrane-anchored isoform A. That is
  not right. Syntaxin 3A and 3B share exons 1-8 and differ in the C-terminal half of the SNARE
  motif and the transmembrane anchor, encoded by alternative exons 9A/B, 10A/B and 11A/B
  PMID:18683220.
• New segment analysis (STX3-bioinformatics/compare_isoforms_rereview.py,
  RESULTS-rereview.md): A0A9L0J901 is 100% identical to mouse syntaxin 3B over residues
  226-262 (48.6% to 3A) but 92.6% identical to the 3A anchor segment (36% to 3B). It is a
  3B-SNARE-half + 3A-anchor combination, not one of the described forms 3A/3B/3C/3D
  PMID:7598732. Same length as 3A (289 aa); Habc, full SNARE motif and tail anchor intact;
  no catalytic residues apply (SNARE).
• It is Ensembl transcript ENSEAST00005047497 (not canonical) of gene ENSEASG00005007989.
  UniProt's gene-centric canonical for donkey STX3 is A0A8C4LHZ8 (289 aa, syntaxin 3A form);
  A0A8C4PQW9 (287 aa) is the syntaxin 3B form. No donkey STX3 entry is reviewed.
• The 92% identity to human isoform A reported in RESULTS.md is real, but it averages near
  identity outside the alternative exon with ~49% identity inside it.

Evolutionary grounding

• PANTHER PTHR19957:SF34 (SYNTAXIN-3), same subfamily as human Q13277, mouse Q64704, rat
  Q08849 (interpro/panther/PTHR19957/PTHR19957-entries.csv). No PANTHER family review
  (*-review.yaml) exists for PTHR19957, so there is no subfamily disagreement to report.
• PAINT IBD nodes (PTHR19957-paint.tsv): PTN000461349 (syntaxin family: endomembrane system,
  SNARE complex, SNAP receptor activity, intracellular protein transport, vesicle fusion);
  PTN000461524 (plasma-membrane syntaxins incl. Q13277: plasma membrane, exocytosis);
  PTN002717040 (Stx3 lineage, donor MGI:103077 only: postsynapse, exocytic insertion of
  neurotransmitter receptor). Donkey TreeGrafter rows descend from these.

Decisions (changes from earlier review)

• Core activity is now grounded in the syntaxin 3B liposome fusion assay
  PMID:18683220, relevant because
  the benchmark carries the 3B SNARE-half exon, plus the apical 3A role PMID:24726755.
• endomembrane system: ACCEPT -> KEEP_AS_NON_CORE (broad, peripheral to PM fusion).
• Postsynaptic terms now cite Jurado et al. 2013 PMID:23395379 rather than UniProt GO lines.
• Other actions unchanged: SNARE binding, fusogenic activity, plasma membrane, exocytosis,
  vesicle fusion, SNARE complex ACCEPT; membrane / vesicle-mediated transport / membrane fusion
  TOO_GENERAL; cytoplasm MARK_AS_OVER_ANNOTATED (the soluble nuclear Stx3S form is a different
  transcript PMID:29475951); intracellular protein transport, postsynapse, receptor insertion,
  cell projection KEEP_AS_NON_CORE.
• No NEW terms: ribbon-synapse terms are 3B-specific and the benchmark mixes exons.

Relative

• Same-gene canonical product A0A8C4LHZ8 (donkey syntaxin 3A) has no review; next folder would
  be genes/EQUAS/STX3__A0A8C4LHZ8. Human STX3 (Q13277) and mouse Stx3 (Q64704) also have no
  review in the repo.
• Modules: modules/synaptic_vesicle_exocytosis.yaml and modules/synaptic_vesicle_cycle.yaml
  mention syntaxins; check before linking (STX3 is the ribbon-synapse syntaxin, not the
  conventional-synapse STX1).