foxd3-a (Xenopus laevis, Q9DEN4) — curation notes

Project context: projects/NEURAL_CREST_ORIGINS.md, Tier 1 (candidate vertebrate-specific
neural crest specifier). Homeolog: foxd3-b (Q9DEN3), no review in this repo.

Identity

• FoxD3-A, aka XFD-6 / xFKH6 / FKH-6; 371 aa; forkhead (winged-helix) domain residues ~92-192.
  Intronless gene; Xenopus orthologue of zebrafish fkd6, chick CWH-3, mammalian HFH-2/genesis
  PMID:11335115.
• Historical Xenopus papers (2001-2007) do not distinguish L/S homeologs; GOA attaches the
  experimental annotations to Q9DEN4 (foxd3-a) only. Deep research flags this explicitly
  [file:XENLA/foxd3-a/foxd3-a-deep-research-falcon.md "The historical experiments generally studied “Xenopus FoxD3/XFD-6” and did not resolve modern X. laevis homeologs"].
  Per project rule, noted, not "fixed" by assertion.

Lines of evidence (project framework)

1. Expression

• Maternal + zygotic; organizer at gastrula, then presumptive neural crest from late gastrula,
  overlapping Zic-r1 and similar to Slug PMID:11493569.
• Premigratory NC; switched off as NC migrates/differentiates (UniProt tissue specificity).
• Also expressed in pluripotent blastula animal-pole cells and lost as competence is lost
  PMID:25931449.
  PMID:30144418.

2. Molecular activity

• Sequence-specific DNA-binding transcriptional repressor; repression needs the winged-helix
  domain PMID:11335115.
• Repression requires a C-terminal eh1/GEH motif that binds Groucho/TLE (Grg4 = tle4, O42478)
  PMID:17138566.
  No activation domain found PMID:17138566.
  eh1/GEH is ancestral to the FoxD subclass PMID:17138566.
• Nuclear PMID:17138566.
• Pioneer→repressor bimodality is shown in zebrafish NC (not Xenopus)
  PMID:30513303,
  PMID:30513303.
  No GO MF term for "pioneer"; not asserted here.

3. Direct targets / enhancers

• Negative autoregulation PMID:11335115.
• No Xenopus ChIP/direct NC target defined. Organizer Nodal regulation is indirect (proposed
  derepression of an unknown inhibitor) [file:XENLA/foxd3-a/foxd3-a-deep-research-falcon.md "The relevant endogenous inhibitor and any direct FoxD3-bound regulatory element were not identified in these experiments."].
• Network position: downstream of border specifiers (Zic) and upstream of / parallel to Slug
  PMID:11493569.
  Downstream of SoxE in explants PMID:30144418.

4. Gain of function

• Sufficiency (Sasai): PMID:11493569
• Conflicting high-dose result (Pohl): PMID:11335115
  — consistent with a dose-sensitive repressor/maintenance factor (excess FoxD3 blocks NC
  progression, and overexpressing cells "do neither differentiate nor migrate").
• Chick: vertebrate FoxD3 (incl. Xenopus) induces ectopic NC; amphioxus FoxD does not
  PMID:24252777.

5. Loss of function

• Dominant-negative FoxD3delN blocks NC without suppressing Sox2; rescued by Slug
  PMID:11493569.
  Caveat: dominant-negative, "FoxD3 (or its closely related factor)".
• Organizer: MO knockdown blocks axis and Nodal expression PMID:17092955.
• Mouse: required for epiblast/ES maintenance PMID:12381664
  and NC progenitor maintenance PMID:18367558;
  restricts mesenchymal fates PMID:21228004.

6. Conservation / outgroups

• Amphioxus FoxD is not expressed at the neural plate border; its regulatory region drives
  mesoderm, not NC PMID:18562679.
• Protein-level innovation: N-terminal 39-aa motif PMID:24252777.
• Repressor activity + DNA specificity are ancestral PMID:24252777.
  So: ancestral MF (forkhead repressor, Groucho recruitment), vertebrate-specific
  deployment (cis) + NC-inducing competence (N-terminal motif, trans).

Network-layer placement

FoxD3 is a neural crest specifier (layer 3), not a border specifier: it is activated
at the border by Zic/Wnt inputs and is required for Slug induction by Zic, and it is not
expressed at the amphioxus border. It also has a competence/multipotency face (blastula
expression, mouse epiblast and NC progenitor maintenance), which places it at the junction
between NC specification and the retained blastula potency programme. GO process choice:
GO:0014034 neural crest cell fate commitment (existing IMP) is the right level for Xenopus
(covers specification + determination; Sasai describes "determination"). Mouse maintenance
data argue for GO:0019827 stem cell population maintenance-type annotation in mouse, but
Xenopus has no direct LOF evidence for that; raised as a question, not a NEW annotation.

The Spemann organizer / mesoderm role (Steiner 2006, Yaklichkin 2007) is real but is a
separate, earlier deployment; kept as non-core.

Decisions summary

• MF: accept IBA/IEA DNA-binding TF; NEW GO:0001227 repressor activity (Pol II) supported by
  Yaklichkin/Pohl; protein binding (tle4) → MODIFY to GO:0001222 transcription corepressor binding.
• BP: neg reg transcription → GO:0000122 for Yaklichkin row; NC fate commitment ACCEPT (core);
  mesoderm formation KEEP_AS_NON_CORE; neurogenesis ±regulation MARK_AS_OVER_ANNOTATED
  (indirect, opposite-sign phenotypes; not where FoxD3 does work).