Twist1 (X-twi) is a class B basic helix-loop-helix transcription factor that binds E-box DNA motifs as a homodimer or as a heterodimer with E proteins, and carries a conserved C-terminal WR domain (Twist box) that mediates protein-protein interactions. In Xenopus it is expressed from early gastrula in axial and lateral plate mesoderm and, from neural plate stage 14, in the cranial neural crest, where it is maintained through migration into the pharyngeal arches. In the crest it is activated downstream of the neural plate border specifier Pax3, later than the earliest crest specifiers, and it is required for normal crest specifier expression (snai2, sox10); in its absence cells remain in a border-like state. It also biases cranial crest toward chondrogenic ectomesenchyme (Sox9, head cartilage) at the expense of glial fate. Part of this activity works through direct WR-domain binding to Snail1/Snail2, which inhibits Snail2 chromatin recruitment and activity and is regulated by GSK3-beta phosphorylation. Twist is an ancestral mesodermal mesenchyme determinant; its expression in the neural crest is thought to reflect co-option during vertebrate origins, and in the tunicate Ciona ectopic Twist converts a pigment-cell lineage into migratory mesenchyme.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000977 RNA polymerase II transcription regulatory region sequence-specific DNA binding | IBA GO_REF:0000033 | ACCEPT | Summary: Twist1 is a bHLH factor whose basic region recognises E-box (CANNTG) motifs in Pol II regulatory regions as a homodimer or E-protein heterodimer. The IBA reflects a PAINT node covering the Twist/bHLH clade with extensive experimental grounding in mouse and human orthologs. Reason: Sequence-specific E-box binding is the conserved molecular basis of Twist activity and the frog protein has an intact TWIST1 bHLH domain. No Xenopus-specific DNA-binding or ChIP data were found, but nothing argues for loss of the activity, so the phylogenetic inference is accepted. Supporting Evidence: PMID:23443570 Twist possesses a basic domain that can interact with core Ebox sequence file:XENLA/twist1/twist1-uniprot.txt Efficient DNA binding requires dimerization with another bHLH |
| GO:0000977 RNA polymerase II transcription regulatory region sequence-specific DNA binding | IEA GO_REF:0000118 | ACCEPT | Summary: TreeGrafter IEA duplicate of the IBA E-box/Pol II regulatory region DNA-binding annotation. Reason: Same assertion as the IBA, consistent with the conserved bHLH domain; accepted. Supporting Evidence: PMID:23443570 Twist possesses a basic domain that can interact with core Ebox sequence |
| GO:0000981 DNA-binding transcription factor activity, RNA polymerase II-specific | IBA GO_REF:0000033 | ACCEPT | Summary: Pol II-specific DNA-binding transcription factor activity inherited from the Twist/bHLH clade. Reason: Twist is a sequence-specific bHLH transcription factor; this is the core MF. In the frog neural crest it also acts by binding other transcription factors (Snail1/2), captured separately as a NEW transcription regulator inhibitor activity, but the DNA-binding TF activity remains the primary molecular function. Direction (activation vs repression) is context dependent in the family, so the neutral term is appropriate. Supporting Evidence: PMID:23443570 Twist possesses a basic domain that can interact with core Ebox sequence PMID:23443570 domain that mediates homodimerization or dimerization with E12/E47 |
| GO:0000981 DNA-binding transcription factor activity, RNA polymerase II-specific | IEA GO_REF:0000120 | ACCEPT | Summary: Combined IEA (InterPro IPR047093 TWIST1 bHLH and PANTHER) duplicate of the IBA. Reason: Consistent with domain architecture and the IBA; accepted. Supporting Evidence: PMID:23443570 Twist possesses a basic domain that can interact with core Ebox sequence |
| GO:0005634 nucleus | IBA GO_REF:0000033 | ACCEPT | Summary: Nuclear localisation inferred phylogenetically for the Twist clade. Reason: A bHLH transcription factor that binds chromatin (Twist reduces Snail2 chromatin recruitment in frog embryos) acts in the nucleus. Endogenous frog protein localisation has not been imaged, but the inference is sound. Supporting Evidence: file:XENLA/twist1/twist1-uniprot.txt SUBCELLULAR LOCATION: Nucleus. PMID:23443570 Twist diminishes recruitment of Snail2 to chromatin |
| GO:0005634 nucleus | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location keyword mapping to nucleus. Reason: Consistent with the IBA and the transcription factor function. Supporting Evidence: file:XENLA/twist1/twist1-uniprot.txt SUBCELLULAR LOCATION: Nucleus. |
| GO:0006357 regulation of transcription by RNA polymerase II | IBA GO_REF:0000033 | ACCEPT | Summary: Regulation of transcription by RNA polymerase II, inherited from the Twist/bHLH clade. Reason: Correct general process for a Pol II-specific sequence-specific transcription factor. Xenopus perturbations change expression of crest genes (sox9, sox10, snai2, foxd3), although direct targets in frog are not defined, so the general unsigned term is the appropriate level. Supporting Evidence: PMID:23443570 domain that mediates homodimerization or dimerization with E12/E47 PMID:23443570 strongly suggesting that the loss of gene expression reflects altered cell specification as opposed to the death of specific cell populations |
| GO:0006357 regulation of transcription by RNA polymerase II | IEA GO_REF:0000118 | ACCEPT | Summary: TreeGrafter IEA duplicate of the IBA regulation of Pol II transcription annotation. Reason: Same as IBA; accepted. Supporting Evidence: PMID:23443570 domain that mediates homodimerization or dimerization with E12/E47 |
| GO:0014029 neural crest formation | TAS PMID:15242799 Cardiac neural crest ablation alters Id2 gene expression in ... | MODIFY | Summary: TAS from a cardiac neural crest ablation/Id2 paper (abstract-only in cache; the abstract does not mention Twist, which was presumably used as a crest marker). The underlying biology is, however, well supported by later Xenopus loss-of-function data (Lander et al. 2013). Reason: Twist1 is genuinely required for cranial neural crest development in Xenopus: NC-targeted MO reduces snai1/2 and strongly reduces sox10, expands the zic1 border domain, without apoptosis or proliferation changes, and is rescued by MO-resistant Twist. Network placement: twist1 is downstream of the border specifier Pax3 (immediate-early, Zic1-independent), switches on at stage 14 after the earliest specifiers, and its loss stalls cells in a border state; it is therefore a (late, cranial-restricted) neural crest specifier, not a border specifier. GO:0014029 neural crest formation is defined as formation of the specialised ectodermal region between neural and non-neural ectoderm, which is closer to the border layer, so, consistent with the sibling sox10 review, the annotation is refined to the child neural crest cell fate specification. The reference itself is weak (see reference_review) but removal is not justified given strong independent evidence. Proposed replacements: neural crest cell fate specification Supporting Evidence: PMID:23443570 Importantly, the effects of Twist depletion can be rescued by a form of Twist that cannot be targeted by the MO PMID:23443570 suggesting that the cells that did not express NC markers were stalled in a NPB state and/or adopted alternative fates, such as placodes PMID:23443570 strongly suggesting that the loss of gene expression reflects altered cell specification as opposed to the death of specific cell populations PMID:23443570 Twist expression is first detectable in NC precursor cells at stage 14, considerably later than several other NC specifiers, including Snail1, Snail2, Sox9 and Foxd3, indicating that it is unlikely to be a regulatory input into the initial expression of these factors PMID:24360906 Twist1 was neither activated by Zic1 alone, nor downregulated by Zic1 knock-down in vivo, suggesting that twist is regulated by Pax3 only. file:XENLA/twist1/twist1-deep-research-falcon.md Both too little and ectopic excess Twist disturbed normal neural-crest development: its effect is dosage- and context-dependent. |
| GO:0032502 developmental process | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Very general developmental process term from the bHLH clade PAINT node. Reason: Developmental process is correctly inherited from the broad bHLH clade node (Twist, Hand, Paraxis/Scleraxis and relatives), which spans mesodermal as well as neural crest roles, so it is kept as a true but non-specific annotation rather than converted into a single neural crest term. Twist1's specific roles are captured separately: restraint of Snai2 during neural crest fate specification (GO:0140416, GO:0014036) and steering cranial crest toward chondrogenic ectomesenchyme (GO:0048701), both in core_functions and as distinct parts of the neural crest gene regulatory network module (modules/neural_crest_gene_regulatory_network.yaml). Mesodermal developmental roles (the ancestral Twist role) are plausible but lack Xenopus functional data. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: PANTHER:PTN002360678 SUPPORTS TRANSFER Broad bHLH clade node (Twist, Hand, Paraxis/Scleraxis and relatives). A developmental role is correctly inherited, but at a level too general to say what Twist1 does; the gene-specific role is neural crest specification and cranial ectomesenchyme. Supporting Evidence: PMID:23443570 Importantly, the effects of Twist depletion can be rescued by a form of Twist that cannot be targeted by the MO PMID:23443570 suggesting that the cells that did not express NC markers were stalled in a NPB state and/or adopted alternative fates, such as placodes PMID:23443570 At stage 28, injected embryos exhibited reduced expression of Sox9 in the branchial arches, where it marks developing chondrocytes (Fig. 1f) and by stage 43 profound defects in cartilage formation and morphogenesis were observed |
| GO:0032502 developmental process | IEA GO_REF:0000118 | KEEP AS NON CORE | Summary: TreeGrafter IEA duplicate of the general developmental process IBA. Reason: Same assessment as the IBA row: developmental process is correctly inherited at the bHLH family level and is kept as a true but non-specific annotation. The gene-specific roles (Snai2 restraint in neural crest specification; cranial ectomesenchyme) are captured in core_functions and in the neural crest gene regulatory network module rather than by narrowing this broad row. Supporting Evidence: PMID:23443570 Importantly, the effects of Twist depletion can be rescued by a form of Twist that cannot be targeted by the MO PMID:23443570 suggesting that the cells that did not express NC markers were stalled in a NPB state and/or adopted alternative fates, such as placodes PMID:23443570 At stage 28, injected embryos exhibited reduced expression of Sox9 in the branchial arches, where it marks developing chondrocytes (Fig. 1f) and by stage 43 profound defects in cartilage formation and morphogenesis were observed |
| GO:0046983 protein dimerization activity | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro HLH-domain mapping to protein dimerization activity. Reason: The HLH domain mediates homodimerization and heterodimerization with E proteins (E12/E47), which is required for efficient DNA binding; this is a real, structurally grounded activity of the bHLH domain. The more specific heterodimerization term is not used because frog-specific E-protein dimerization was not tested. Supporting Evidence: PMID:23443570 domain that mediates homodimerization or dimerization with E12/E47 file:XENLA/twist1/twist1-uniprot.txt Efficient DNA binding requires dimerization with another bHLH |
| GO:0140416 transcription regulator inhibitor activity | IDA PMID:23443570 Interactions between Twist and other core epithelial-mesench... | NEW | Summary: Through its C-terminal WR domain (Twist box) Xenopus Twist1 binds Snail1 and Snail2 directly, reduces Snail2 recruitment to chromatin, and (when phosphorylated at GSK3-beta site S148) blocks Snail2-induced ectopic neural crest formation. Reason: This is a second, frog-demonstrated molecular activity that explains the stoichiometry-sensitive gain- and loss-of-function phenotypes and links Twist to Snail2 in the crest EMT/specifier unit. Direct binding (GST pulldown), reduced chromatin occupancy of the partner, and functional inhibition of the partner's activity together meet the definition of transcription regulator inhibitor activity better than a bare binding term. The same WR-domain mechanism (inhibition of Runx2, Sox9) is documented for mammalian Twist1. Supporting Evidence: PMID:23443570 Glutathione S-transferase (GST)-pulldown assays indicate that the interaction between Twist and Snail2 is direct PMID:23443570 Twist diminishes recruitment of Snail2 to chromatin PMID:23443570 Snail2 expression induces ectopic NC formation in this assay, and while TwistS148D potently blocked its effects, TwistS148A did not PMID:23443570 Collectively, our data suggested a model in which Twist binds to and inhibits the activity of Snail proteins |
| GO:0048701 embryonic cranial skeleton morphogenesis | IMP PMID:23443570 Interactions between Twist and other core epithelial-mesench... | NEW | Summary: Twist1 acting in cranial neural crest biases crest toward chondrogenic (ectomesenchymal) fate; NC-targeted depletion reduces Sox9 in the branchial arches and causes profound head cartilage defects at stage 43, with a reciprocal increase in Foxd3+ presumptive glia; gain of function has the opposite effect. Reason: Participation: Twist1 is a transcription factor acting within the cranial crest cells that build the head skeleton, determining their chondrogenic fate, so it does part of the work rather than being merely required. The finding is reproduced with the S148A phospho-mutant (malformed facial cartilage) and is conserved: mouse Twist1 null embryos have malformed branchial arches and facial primordia, and mouse Twist1 carries this term. Tetrapod and Ciona data tie Twist specifically to ectomesenchyme. This is the main derivative-level role of the gene in the crest and is kept separate from specification. Supporting Evidence: PMID:23443570 At stage 28, injected embryos exhibited reduced expression of Sox9 in the branchial arches, where it marks developing chondrocytes (Fig. 1f) and by stage 43 profound defects in cartilage formation and morphogenesis were observed PMID:23443570 suggesting that in the absence of Twist, cranial NC cells that would normally give rise to cartilage might instead adopt glial fates PMID:23443570 mice have severe defects in cephalic neural tube closure and malformed branchial arches and facial primordium PMID:23135395 In tetrapods, it appears that only the cephalic neural crest expresses Twist and produces ectomesenchyme |
Loading supporting content…
Download this section (compressed HTML)Q: GO has no neural plate border specification/formation term, and GO:0014029 neural crest formation is defined as formation of the border ectoderm region. The project proposes one (neural plate border formation) and the obsoletion of GO:0014029. Is the separation clean for late crest specifiers such as twist1 (Pax3-dependent, onset stage 14, loss expands zic1), whose loss leaves cells in a border state?
Q: Twist is classed as a crest specifier in Xenopus and zebrafish but its early crest expression is lost in amniotes. Should the specification annotation (GO:0014036) be restricted to anamniote Twist orthologs, with the ectomesenchyme/cranial skeleton role as the conserved vertebrate function?
Q: GO lacks a term for neural crest-derived ectomesenchyme formation or cranial crest fate diversification (cartilage vs glia). Is a term such as 'ectomesenchymal cell fate commitment' warranted for Twist, given the Ciona a9.49 reprogramming result?
Q: Which genes are direct Twist1 (Twist-E protein heterodimer) targets in the Xenopus cranial crest, and does Twist regulate sox9 directly or only by restraining Snail2?
Q: Does Twist1 have a functional role in Xenopus mesoderm (its first and ancestral expression domain), or is that role redundant with other bHLH factors in frog?
Experiment: CUT&RUN for tagged Twist1 and E47 in dissected cranial neural folds and arch crest (stages 17-28), combined with Twist morphant/F0 CRISPR RNA-seq.
Hypothesis: Twist1 directly occupies E-box enhancers of chondrogenic ectomesenchyme genes (e.g. sox9, runx2) in Xenopus cranial crest.
Type: genomic binding assay
Experiment: F0 CRISPR twist1.L/.S knockout or hormone-inducible dominant-negative Twist applied after stage 14, assaying sox10/snai2 at neurula and sox9/cartilage versus foxd3/glia at stage 28-45.
Hypothesis: Twist1 is needed in the crest for ectomesenchyme but not for crest induction per se.
Type: genetic perturbation
Experiment: Test lamprey Twist expression timing and knockdown in cranial crest; compare with Ciona a9.49 Twist misexpression and amphioxus neural border Twist expression.
Hypothesis: Twist co-option into the crest was the step that conferred ectomesenchymal potential in vertebrates.
Type: comparative evolutionary analysis
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)