Actl7a (rat, Q641W9) review notes

2026-10-28 session

• just fetch-gene rat Actl7a resolved Q641W9 (Swiss-Prot ACL7A_RAT, RGD:1304697, GeneID 298017).
  34 GOA rows; two identical nucleus ISS rows (Q9QY84) collapsed to one in the seeded review (33).
• Only one paper is cited by GOA for rat: PMID:21278383 (full text cached). The rat material in it is
  the Tes co-IP from rat testis extracts PMID:21278383. Immunofluorescence used mouse testes; pulldowns and the crystal
  structure used human Arp7A.
• Consequence: the rat cytoplasm EXP row (UniProt, PMID:21278383) has no rat cytoplasmic
  observation behind it; the cytoplasmic observation is GFP-human-Arp7A in HeLa ("data not shown")
  or mouse sections. Recorded as a WRONG_EVIDENCE_CODE source defect (CHECK). The claim itself is
  true (perinuclear theca is_a cytoskeleton, part_of cytoplasm).
• The IDA protein-containing complex row is genuinely rat (co-IP in rat testis): ACCEPT.
• Mouse donor (Q9QY84, MGI:1343051) annotations checked in QuickGO 2026-10-28: perinuclear theca IDA
  PMID:41169243; male germ cell nucleus IDA and motile cilium IDA PMID:12672658 (MGI); acrosome
  assembly IMP PMID:32923619 + PMID:36734600; spermatid development IMP PMID:37667331; fertilization
  IMP PMID:32923619 + PMID:35921706; Golgi colocalizes_with IDA PMID:21278383.
• Decisions follow the human ACTL7A review for consistency:
• acrosomal vesicle -> MODIFY to perinuclear theca (protein is in the subacrosomal layer, on the
  nuclear side of the inner acrosomal membrane, not inside the acrosome) PMID:21278383. WRONG_TERM source defects (CHECK) on the ISS (UniProt) and ISO (RGD) rows.
• male germ cell nucleus ISO -> KEEP_AS_NON_CORE.
• nucleus -> MODIFY to male germ cell nucleus (already on the gene by ISO; only germ cells express
  it).
• cytoplasm, cytoskeleton -> MODIFY to perinuclear theca (+ Golgi apparatus for cytoplasm).
• Golgi apparatus -> KEEP_AS_NON_CORE (root organelle; transient association during acrosome
  biogenesis; mouse IDA source holds).
• motile cilium -> UNDECIDED (only the 2003 abstract mentions tails; full text not cached; 2023
  study reports a connecting-piece pool, which GO does not place in the flagellum).
• acrosome assembly, spermatid development, perinuclear theca, structural molecule activity,
  protein-containing complex -> ACCEPT.
• fertilization -> KEEP_AS_NON_CORE (knock-out necessity; downstream of the sperm-head structural
  role, no fertilization step performed by Actl7a).
• ND biological_process -> REMOVE (superseded by process rows), per rat-batch precedent.
• NEW: GO:0030274 LIM domain binding (ISS from human, PMID:21278383). The Tes-binding segment
  (human 28-45 = rat 33-50, LQTASLRDGPAKRAVWVR) is identical; rat co-IP shows the complex in rat
  testis. This is an MF, not a process. No NEW process terms (actin cytoskeleton organization, egg
  activation) were added: the evidence for them is mouse/human mutant phenotype only.