XB-ART-38291Dev Biol October 15, 2008; 322 (2): 355-67.
Show Gene links Show Anatomy links
Hairy2-Id3 interactions play an essential role in Xenopus neural crest progenitor specification.
Loss of function studies have shown that the Xenopus helix-loop-helix transcription factor Hairy2 is essential for neural crest formation and maintains cells in a mitotic undifferentiated state. However, its position in the genetic cascade regulating neural crest formation and its relationship with other neural crest regulators remain largely unknown. Here we find that Hairy2 is regulated by BMP, FGF and Wnt and that it is only required downstream of BMP and FGF for neural crest formation. We show that Hairy2 overexpression represses neural crest and upregulates neural border genes at early stages while it expands a subset of them in later embryos. We show that Hairy2 downregulates Id3, another essential HLH neural crest regulator, through attenuation of BMP signaling. Knockdown and rescue experiments indicate that Id3 protein, which physically interacts with Hairy2, negatively regulates Hairy2 activity. However, Id3 is required to allow Hairy2 to promote neural crest formation. Together, our results provide evidence that Hairy2 acts downstream of FGF and BMP signals at the neural border to maintain cells in an undifferentiated state, and that Hairy2-Id3 interactions play an essential role in neural crest progenitor specification.
PubMed ID: 18721802
Article link: Dev Biol
Species referenced: Xenopus
Genes referenced: ag1 bmp4 bmpr1a dct fgf8 fgfr4 foxd3 hes1 hes2 hes4 hes6.1 hes6.2 hoxb9 hoxc9-like id3 krt12.4 msx1 myc myod1 myt1 neurod1 nog notch1 nrp1 otx2 pax3 pax8 ranbp2 six1 six3 snai2 sox10 sox15 sox2 sox3 sox9 st14 tbxt trpc2 tubb2b wnt7b wnt8a zic1
Morpholinos: fgf8 MO1 hes4 MO1 hes4 MO2 id3 MO2 msx1 MO3 pax3 MO4 zic1 MO1
Article Images: [+] show captions
|Fig. 1. Signaling pathways and transcriptional factors involved in the regulation of Hairy2 expression in NC. (A) Hairy2 is activated by attenuation of BMP (tBr 50 pg), activation of Wnt (dnGSK3β 250 pg) or FGF (FGF8a 25 pg). Respective induction: (a) 72% n = 27; (b) 50% n = 25; (c) 64% n = 22. Activation of BMP (Ca-Alk3 50 pg), inhibition of Wnt (GSK3β 250 pg) or FGF (δR4 250 pg) has the opposite effect. Respective inhibition: (d) 82% n = 31; (e) 58% n = 17; (f) 62% n = 19. Activation (Su(H)Ank-GR 1 ng) or repression (Su(H)DBM-GR 1 ng) of Notch signaling does not influence Hairy2 expression (g, h; unaffected 92% n = 26 and 85% n = 4, respectively). (i) Q-PCR analysis of animal cap explants derived from embryos injected radially with ICD (250 pg) or Su(H)dbm (250 pg) mRNA together with noggin (50 pg)+ Wnt8 (50 pg) mRNA. (B) Expression of Hairy2 in Msx-1 (a, b), Pax3 (c, d), and Zic1 (e, f) overexpressing or depleted (20 ng MO) embryos. Overexpression of each of these factors (125–250 pg mRNA) expands Hairy2 expression (Respective induction: (a) 57%, n = 77; (c) 63%, n = 30; (e), 85% n = 55). Depletion of Msx1 reduces Hairy2 expression on the injected side (b, 49% with mild decrease, n = 70). Depletion of Pax3 (d) and Zic1 (f) inhibited Hairy2 expression (respective inhibition: d, 63% n = 59; f, 65%, n = 29). (C) FGF8a injections with CoMO expanded Hairy2 (a) and Snail2 (e) (Expansion in 92%, n = 49 and 55%, n = 20 respectively). In presence of Msx1 MO, the expansion caused by activation of FGF signaling was still observed for Hairy2 (b, 73% expanded, n = 30) but not for Snail2 (f, 45% decreased, n = 20). Activation of Wnt signaling by Wnt7b with CoMO expands Hairy2 (c, 83% expanded, n = 18) and Snail2 (g, 78% of expansion, n = 17). In the presence of Pax3 MO, Wnt7b still expanded Hairy2 expression (d, 93% expansion, n = 14) while Snail2 is decreased (h, 62% reduced, n = 13). All embryos are shown in dorso-anterior view. β-galactosidase was used as a lineage tracer (light blue staining) and the injected side is oriented to the right. (D) NC inducing signals and neural border factors involved in Hairy2 regulation. The dashed line indicates that Msx1 contributes to a lesser extend than Pax3 and Zic1 to Hairy2 expression.|
|Fig. 2. Hairy2 is required for NC induction downstream of FGF and BMP. Embryos were injected either with mRNA encoding an activated form of FGFR4 (torso fused FGFR4, tFGFR4) and a control MO or with tFGFR4 mRNA and Hairy2 MOs. tFGFR4 and CoMO injections expanded Snail2 (a) and FoxD3 (c) (Expansion in 62% and 43%, n = 24 and n = 2 8 respectively). Hairy2 MOs blocked the ability of tFGFR4 to expand Snail2 (b) and FoxD3 (d) (61%, n = 28 and 63% decreased, n = 27 respectively). Attenuation of BMP signaling by tBR with CoMO expanded Snail2 (e) and FoxD3 (g) (expansion in 68%, n = 19 and 60%, n = 20 respectively). In the presence of Hairy2 MOs, tBR no longer induced expansion of Snail2 (f, decreased in 35%, n = 20) nor FoxD3 (h, decreased in 73%, n = 19). Wnt signaling with CoMO expanded both Snail2 (i, 84%, n = 19) and FoxD3 (k, 83%, n = 18). Co-injection of Wnt7b and Hairy2 MOs still expanded both Snail2 (j, 78% expansion, n = 18) and FoxD3 (l, 100% expansion, n = 19). All embryos are shown in dorso-anterior view. β-galactosidase was used as a lineage tracer (red staining) and the injected side is oriented to the right.|
|Fig. 3. Hairy2 overexpression represses NC marker at early stages but expands a subset of them in later embryos. (A) Snail2 expression is initially repressed by Hairy2-GR overexpression (500 pg mRNA) and increased in later embryos (a, stage 13, 65% decreased, n = 17; b, stage 15, 58% decreased, n = 24; c, stage 21, 27% reduced n = 22; d, stage 27, 53% increased n = 19). Dorsal view (a, b) or high magnification of the head region are shown (c, d). β-galactosidase was used as a lineage tracer. (e) Q-PCR analysis of Snail2 expression in animal caps injected with noggin+ Wnt8 (control) or injected with noggin+ Wnt8 and Hairy2-GR (250 pg/blastomere) and harvested at different stages. Snail2 is repressed by Hairy2 during neurula stages but is progressively upregulated at later stages. (B) Injection of 500 pg of Hairy2-GR mRNA increases Sox10 (a) but decreases Sox9 (b) and Trp2 (c) (78% induced for Sox10, n = 29; 60% inhibited n = 20 for Sox9 and 62%, n = 21 for Trp2) at tailbud stages. Lateral views with control and injected sides are shown. β-galactosidase was used as a lineage tracer. (C) Q-PCR analysis of animal caps derived from embryos injected with noggin+ Wnt8 mRNA and cultured until the equivalent of stage 28. Hairy2-GR (250 pg/blastomere) represses Sox9 but increases Snail2 and Sox10 while Hairy2 depletion (5 ng/blastomere) reduces all markers. (D) Temporal ability of Hairy2 activity to activate late NC development. Note that Hairy2-GR ability to increase Sox10 is only observed when DEX is added before st14. (E) Eight-cell-stage embryos injected radially at the animal pole with Hairy2-GR (250 pg/blastomere) (b, d) display reduced pigmentation. Cranial cartilages were also reduced compared to uninjected controls (a, c). Respective inhibitions: (b) 58%, n = 24; (d) 54%, n = 21. (F) Extracts from stage 32 (a) or 45 (b) embryos injected radially with Hairy2-GR were analysed using anti-GFAP antibody. GAPDH was used as loading control. Note that GFAP is first reduced and later increased by Hairy2 overexpression.|
|Fig. 4. Hairy2 affects the expression of neural border genes. Hairy2-GR mRNA overexpression (500 pg) expands Msx1 (b, 64% n = 22), Pax3 (d, 62% n = 17) and strongly increases Zic1 (f, g, 78% n = 14), SoxD (i, j, 100% n = 11). Injection of Hairy2 MOs (10 ng) reduces Msx1 (a, 60% n = 26) and expands Pax3 (c, 58% n = 31) and Zic1 (e, 53% n = 17). SoxD appears unaffected (h, 80% normal n = 20). Embryos are shown in dorsal (a–f, h, i) or lateral view (g, j), with inset showing control sides. β-galactosidase was used as a lineage tracer.|
|Fig. 5. Hairy2 negatively regulates Id3 by repression of BMP4. (A) At neurula stage, Hairy2 and Id3 expression overlaps in the NC region as shown by double in situ hybridisation. Id3 is in purple and Hairy2 in dark blue (a–b). Higher magnifications is shown (a’–b’). (B) Hairy2-GR overexpression (500 pg) decreases Id3 (a, 52%, n = 25). Knockdown of Hairy2 increases Id3 (b, 56%, n = 32). Overexpression of a Hairy2 DNA-binding mutant (Hairy2DBM-GR 500 pg) does not affect Id3 (c, 95% unaffected, n = 22). In NC induced animal caps, Hairy2 MOs (5 ng/blastomere) upregulates Id3 while Hairy2-GR overexpression (250 pg/blastomere) reduces Id3. Hairy2DBM-GR (250 pg/blastomere) has no effect on Id3 (d, e). Id3 overexpression or depletion does not affect Hairy2 expression significantly (f, g, 90% n = 19 and 80% n = 24, respectively). (C) Hairy2-GR overexpression inhibits (a, 56%, n = 18), while its knockdown increases BMP4 expression in the NC (b, 45%, n = 22). Overexpression of Hairy2DBM-GR (500 pg) has no effect on BMP4 expression in NC (c, 80% normal, n = 15). Q-PCR analysis of animal caps induced to NC injected with Hairy2 MOs, Hairy2-GR or Hairy2DBM-GR (d, e). Injection of CA-Alk3 (25 pg) activates Id3 (f, 43%, n = 16) and blocks Hairy2 ability to repress Id3 in embryos (g, 77% normal, n = 25). All embryos are shown in dorso-anterior view. β-galactosidase was used as a lineage tracer (light blue staining) and the injected side is oriented to the right.|
|Fig. 6. Id3 interacts and blocks Hairy2 activity in NC. (A) Overexpressed Flag-tagged Hairy2 but not Hairy1, Hes2 or Hes6 interacts with HA-tagged Id3 in immunoprecipitation assays. (B–C) Mapping of the domains of Hairy2 and Id3 required for their interaction. Note that only the Hairy2 deletion mutants containing the Orange domain are able to interact with Id3 (B) and that the C-terminal part of the HLH domain of Id3 is required for Hairy2 interaction (C). (D) Co-injection of Id3-GR (1 ng) rescues Snail2 inhibition by Hairy2-GR (70% rescued, n = 26)(a, b). Q-PCR analysis of Snail2 in animal caps induced to form NC injected radially with Hairy2-GR (250 pg), Id3-GR (500 pg) or both (250 pg and 500 pg respectively. A control western blot showing that in each condition similar levels of overexpressed proteins Myc-Hairy2-GR and Flag-Id3-GR have been produced is shown (c). Co-injection of Id3-GR (1 ng) blocks the ability of Hairy2-GR to induce SoxD (d, 89% ectopic n = 18; e, 74% rescued n = 19). Dorso-anterior views with injected side to the right (a, b) or lateral views (d–g) of neurula embryos are shown. β-galactosidase was used as a lineage tracer.|
|Fig. 7. Id3 is required for Hairy2 to promote NC formation. (A) Hairy2-GR (500 pg) injection induces Sox10 (a; 55% increased, n = 24) in tailbuds and upregulates SoxD (e, 88%, n = 16), Pax3 (h, 65%, n = 32), Msx1 (k, 59%, n = 27) and Zic1 (n, 63%, n = 32) in neurula embryos. Hairy2-GR and Id3 MO co-injection(15 ng) decreases Sox10 (60% inhibition, n = 27) in embryos but still induces SoxD (g, 70% increased n = 20), Pax3 (j, 62%, n = 29), Msx1 (m, 49, n = 35) and Zic1(p, 67%, n = 33). Control minus DEX shows that the Id3 MO efficiently inhibits Sox10 (50% inhibited, n = 15;) and has no effect on SoxD (f, 76% normal n = 17) Pax3 (i, 70%, n = 23), Msx1 (l, 85%, n = 21) and Zic1(o, 65%, n = 23). Hairy2-GR was also unable to induce Sox10 with Id3 MO in NC induced animal explants (d). (B) Id3-GR (1 ng) partially rescues Snail2 and Sox10 in Hairy2 (10 ng) depleted embryos (a, b, d, e) and in NC induced animal caps explants analysed by Q-PCR (c, f). Respective inhibition and rescue: (a) 68% inhibited, n = 26; (b) 89% rescued, n = 27; (d) 56% inhibited, n = 18; (e) 75% rescued; n = 30. Dorso-anterior views of embryos with injected side to the right (a, b) or lateral views (d, e) of embryos are shown. β-galactosidase was used as a lineage tracer.|
|Figure S1. Hairy2 depletion using morpholinos targeting both Hairy2a and b pseudolleles efficiently inhibits NC formation. (A) Hairy2 MOs efficiently and specifically inhibits the translation of both Hairy2 pseudoalleles. Design of the Hairy2a and Hairy2b MOs (a). GFP fluorescence of embryos injected with 500 pg of Hairy2a-GFP or Hairy2b-GFP mRNA alone (b,c) or in combination with 5 ng of the corresponding MO (d,e). No effect was observed with Hairy2b MO 5mismatch (f). Hairy2a+b MOs have no effect on GFP translation (g). (B) Hairy2 MOs (10 ng MO) effciently inhibits Snail2 expression (76%, n = 47) in neurula embryos (st14-15) (a). The phenotype of Hairy2 knockdown is rescued (67% normal, n = 29) by coinjection of a low dose (50 pg) of Hairy2-GR mRNA lacking the MOs binding site (b). Hairy2 MOs (10ng) blocks the expression of Slug, Sox10, Sox9 and Trp2 in tailbud embryos (respective inhibition: 53%, n = 26 ; 57%, n = 17 ; 75%, n = 24 and 69% n = 16 respectively)(c-f). (C) Hairy2 inhibitory phenotype on NC is not due to an increase in primary neurons nor to cell fate changes. Note that in Hairy2 MOs (10ng total) injected embryos, the expression of neuronal (a, MyT1 n = 24; b, N-tubulin n = 14; c, NeuroD n = 27), neural plate (d, Sox3 n = 22; e, Nrp1 n = 22), mesodermal (f, MyoD n = 14), placodal (g, Six1 n = 18; h, Six3 n = 31; I, Pax8 n = 19), antero-posterior patterning markers (j, Otx2 n = 21; k, HoxB9 n = 20; l, XAG1 n = 22) or epidermal (m, Epi. k n = 34) is unaffected. Dorso-anterior views with injected side to the right (Ba,b,C) or lateral views (Bc,f) of embryos are shown. β-galactosidase was used as a lineage tracer.|
|Figure S2. Hairy2 and Hairy2-GR mRNA injection (500 pg) into the equatorial zone of 4- cell stage embryos represses the mesodermal marker Xbra. Respective inhibition: a, 94% n = 34 ; b, 64% n = 46. DEX was added at stage3. Embryos at gastrula stage are viewed from the ventral side. β-galactosidase was used as a lineage tracer.|
|Figure S3. Injection of mouse Hes1-GR mRNA (500 pg) like Xenopus Hairy2 mRNA downregulates Snail2 at neurula stage and promotes Sox10 at stage 28. Indicated phenotype was observed in: a, 73% inhibited, n = 15; b, 64% increased, n = 14. Embryos are viewed from the dorsal (a) or lateral side (b). β-galactosidase was used as a lineage tracer.|
|Figure S4. Hairy2 overexpression does not induce major cell fate changes in the early embryo. (A) In neurula embryos, Hairy2-GR mRNA (500 pg) injection represses the expression of the MyT1, N-tubulin and NeuroD neuronal markers (a, 50% n = 22; b, 55% n = 18; c, 52% n = 27). No significant change is observed on the expression of neural plate (Sox3), mesodermal (MyoD), placodal (Six1, Six3, Pax8) and antero-posterior patterning markers (Otx2, HoxB9, XAG1). All unaffected: d, n = 15; e, n = 34; f, n = 14; g, n = 17; h, n = 24; i, n = 25; j, n = 16; k, n = 25; l, n = 20. Hairy2-GR mRNA (500pg) represses the epidermal marker Epi. K. (m, 64% n = 26 and expands transiently Sox2 at neurula stages (n, 66% increase n = 18) but have no effect on Sox2 at tailbud stages (o, 100% unaffected n = 19). (B) Hairy2-GR overexpression (500 pg) decreases N-tubulin (a, 75% n = 12; b, 53% n = 19; c, 55% n = 20; d, 50% n = 12) and does not affect Sox3 and Six1 expression in stage 13 to 26-27 embryos. (e-l, all unaffected; n = 20). Embryos are shown in dorso- anterior views with the injected side to the right (a-n) or in lateral view with inset showing the control side (o). β-galactosidase was used as a lineage tracer. (C) Q-PCR analysis of stage 14 NC induced animal caps derived from embryos injected with Hairy2 MO5mis (5 ng/blastomere), Hairy2 MOs (5 ng/blastomere) or Hairy2-GR mRNA (250 pg/blastomere). Note that both Hairy2 gain or loss of function reduces Slug (a) without significantly affecting Six1, Sox3 or N-tubulin expression. Ep. keratin expression which is strongly reduced in NC induced explants, appears also unaffected by Hairy2(c).|
|Figure S5. Hairy2 and Id3 MO double knockdown leads to a more dramatic reduction of Sox10 expression in NC than individual single knockdown. (A) Embryos unilaterally injected with Hairy2 MOs (10 ng), Id3 MO (15 ng) or both MOs (25 ng total). Note that Sox10 expression is dramatically inhibited in double knockdown (c) (83% totally inhibited, n=24) while residual expression is clearly seen in single knockdown (50% reduced, n=14 for hairy2 MO and 45%, n=17 for Id3 MO). Lateral views of control and injected sides of tailbud embryos are shown. β-galactosidase was used as a lineage tracer. (B) Q-PCR analysis of NC induced animal cap explants from embryos injected radially with Hairy2 MO (5 ng/blastomere), Id3 MO (10 ng/blastomere) or both (15 ng total/blastomere). Note that both Hairy2 and Id3 MO inhibit Sox10 expression and that the reduction observed is more severe in the double knockdown.|
|dct gene expression in Xenopus laevis embryos, NF stage 28, as assayed by in situ hybridization. Lateral view: anterior right, dorsal up.|