Twist (Sp-Twist, A7Z0B7) — curation notes

Identity

• UniProt A7Z0B7 (TrEMBL, 204 aa), SubName "Twist", derived from the third-party-annotation
  record EMBL BK006287 / protein DAA06084.1; RefSeq NP_001099179.1, GeneID 753325. The UniProt RX
  line cites Gitelman 2007 PMID:17567594, a phylogenetic survey of the vertebrate twist family
  (abstract only in cache; it does not report any sea urchin experiment).
• The DAA06084 protein itself was annotated and deposited by Wu, Yang & McClay 2008
  PMID:18495103, who found the S. purpuratus gene in silico
  PMID:18495103. The 204-aa length matches the UniProt sequence, so A7Z0B7 is the S. purpuratus
  twist gene of the McClay-lab GRN papers with HIGH confidence. SpTwist and LvTwist are 92%
  identical PMID:18495103.
• Domains (UniProt DR): bHLH IPR011598 (residues 109-160; PROSITE PS50888), TWIST1-type bHLH
  IPR047093 (CDD cd11412 bHLH_TS_TWIST1), PANTHER PTHR23349 subfamily SF50. The C-terminal
  sequence ends in ...ERLSYAFSVWRMEG..., i.e. the Twist-family "WR motif"
  PMID:18495103.
  Single twist gene in the sea urchin genome PMID:24598159.
• No GO-CAM contains this gene (gocams/index.tsv has no STRPU rows).

IMPORTANT organism caveat

• All functional data on sea urchin Twist come from Lytechinus variegatus (LvTwist), not from
  S. purpuratus: Wu et al. 2008 PMID:18495103 and Saunders & McClay 2014 PMID:24598159. For
  S. purpuratus the only facts are the in-silico gene identification and the 92% identity.
• twist is not a node of the Oliveri/Tu/Davidson 2008 S. purpuratus skeletogenic GRN
  PMID:24598159; the Sp model lists alx1, ets1, tbr, tel, erg, hex,
  foxb, dri as drivers of the differentiation genes PMID:18413610. Rafiq 2014
  (PMID:24496631, abstract) and Shashikant 2018 (PMID:30264451, full text) do not mention twist.
• Consequence for the review: NEW rows are kept to the minimum that the L. variegatus
  perturbation data support, are recorded as IMP on PMID:18495103, and every such row states
  that the evidence is from the congener (orthology is secure: single-copy gene, 92% identity).

Expression (L. variegatus)

• Low maternal/early level; rises from blastula, dips at mesenchyme blastula, rises again in
  gastrula PMID:18495103.
• Spatially: vegetal plate at late hatched blastula, then PMCs PMID:18495103;
  PMC expression persists through gastrulation, unlike snail. Later also in SMC territory
  PMID:18495103.

Place in the skeletogenic (micromere/PMC) GRN

• Upstream: Alx1 is required for twist expression; Ets1 is not PMID:18495103.
• Downstream/feedback: Twist is needed to maintain alx1 expression after ingression, not to
  initiate it PMID:18495103
  PMID:18495103. Directness untested
  PMID:18495103.
• Differentiation genes: msp130, sm50, sm30 transcripts fall in Twist morphants
  PMID:18495103 PMID:18495103.
• Saunders & McClay 2014 place twist among the 10 proximal EMT regulators downstream of
  alx1/ets1/tbr PMID:24598159 and define an alx1-twist-snail positive-feedback
  sub-circuit for de-adhesion PMID:24598159 PMID:24598159.

Perturbation phenotypes (L. variegatus, two non-overlapping morpholinos; mRNA overexpression)

• Summary PMID:18495103.
• Ingression delayed, not abolished PMID:18495103; cell-autonomous requirement shown by
  micromere transplantation PMID:18495103. In the 2014 five-assay dissection, twist morphant PMCs make the laminin hole, move,
  constrict and keep polarity but fail to de-adhere PMID:24598159.
• No skeleton PMID:18495103; PMC
  syncytial fusion fails cell-autonomously PMID:18495103. Overexpression gives extra/mis-positioned spicules.
• Specification-state maintenance PMID:18495103.
• SMC derivatives also affected (pleiotropic, later role) PMID:18495103.

Decisions for the GOA review

• The 8 GOA rows are all electronic (5 IBA from PAINT node PTN002360678 = twist family, 3 IEA).
  The IBA MF/CC/BP rows (GO:0000981, GO:0000977, GO:0005634, GO:0006357) describe a bHLH TF and
  are accepted; sea urchin evidence for transcriptional regulation is indirect (knockdown lowers
  alx1 and biomineralization transcripts) and from L. variegatus, so no activator/repressor
  sign is asserted. GO:0032502 developmental process (IBA) and GO:0046983 protein dimerization
  (InterPro) are kept as non-core: true by family but uninformative.
• NEW (all IMP, PMID:18495103, L. variegatus caveat recorded): GO:0010718 positive regulation of
  EMT (delayed/incomplete ingression, de-adhesion sub-circuit); GO:0070169 positive regulation of
  biomineral tissue development (skeleton-minus, reduced msp130/sm50/sm30, PMC fusion failure);
  GO:0001710 mesodermal cell fate commitment (maintenance of the PMC specification state via the
  Alx1-Twist feedback loop; same term as the sibling Hex/Erg state-stabilization reviews).
• Not proposed: positive regulation of transcription by RNA polymerase II (no direct target or
  cis-regulatory evidence); any muscle/pigment cell term (secondary, under-explored); mesodermal
  cell fate specification (twist is activated after alx1/ets1 and acts in maintenance, not
  initiation).