XB-ART-38869Dev Biol 2008 Nov 01;3231:114-29. doi: 10.1016/j.ydbio.2008.08.007.
Show Gene links Show Anatomy links
A new role for the Endothelin-1/Endothelin-A receptor signaling during early neural crest specification.
The neural crest is induced at the border of the neural plate in a multistep process by signals emanated from the epidermis, neural plate and mesoderm. In this work we show for the first time the existence of a neural crest maintenance step which is dependent on signals released from the mesoderm. We identified Endothelin-1 (Edn1) and its receptor (Ednra) as key players of this signal and we show that Edn1/Ednra signaling is required for maintenance of the neural crest by a dual mechanism of cell specification and cell survival. We show that: (i) Ednra is expressed in prospective neural crest; (ii) loss-of-function experiments with antisense morpholino or with specific chemical inhibitor suppress the expression of early neural crest markers; (iii) gain-of-function experiments expand the neural crest territory; (iv) epistatic experiments show that Ednra/Edn1 is downstream of the early neural crest gene Msx1 and upstream of the late genes Sox9 and Sox10; and (v) Edn1/Ednra signaling inhibits apoptosis and controls cell specification of the neural crest. Together, our results provide insight on a new role of Edn1/Ednra cell signaling pathway during early neural crest development.
PubMed ID: 18775422
Article link: Dev Biol
Species referenced: Xenopus
Genes referenced: bcl2 dct ece1 edn1 ednra ednrb foxd3 krt12.4 msx1 ranbp2 snai2 sox10 sox2 sox9
Morpholinos: ednra MO1
Article Images: [+] show captions
|Fig. 1. Xenopus laevis Endothelin-1 receptor type A features and expression pattern. (A) Molecular phylogenetic analysis of Ednra shows the relationship with Xenopus ETAX, Xenopus Ednrb and with Ednra from different vertebrate species. The protein sequences of different Ednra were analyzed using ClustalW 1.81, and an unrooted tree was constructed by neighbour-joining analysis. Xenopus laevis Ednrb, Danio rerio EdnrA1, and Danio rerio EdnrA2 branches are collapsed 50%. (B) RT-PCR analysis of temporal expression pattern of Ednra, Ppet-1, and ECE-1 during different developmental stages. Total RNA isolated from stages 1, 10.5, 12, 14, 18, and 21 was isolated, retrotranscribed and amplified as described in Materials and methods. Contamination of genomic DNA was examined by experiments without RT reactions (−RT) using total RNA from stage 18 embryos. EF1α expression was used as loading control. (C–L) Expression pattern of Xenopus laevis Ednra. (C) Ednra transcripts are first detected from late gastrula stage (St. 12.5) in the lateral domains of neural plate (arrows). (D) During neurulation Ednra is expressed in the prospective cephalic (asterisk) and trunk neural crest (arrows). (E) In situ hybridization for FoxD3. (F) Double in situ hybridization for Ednra (purple) and Sox2 (turquoise). (G) Transverse section of (F). Asterisk, prospective neural crest. (H) Tailbud-stage (St. 24) embryo showing expression in the migrating cephalic neural crest (arrowheads), trunk region (arrow), crest cells migration into dorsal fin (white arrowheads), and otic vesicle (OtV). (I) Frontal section of (H), Ednra expression in the otic vesicle (OtV) and the neural crest hyoid migratory stream (arrowheads). NT, neural tube; n, notochord. (J) A St. 30-embryo showing Ednra expression in the branchial arches (black arrowheads), cells into the dorsal fin (white arrowheads), and otic vesicle (OtV). OpV, optic vesicle. (K) Section showing Ednra expression in the dorsomedial and ventromedial region of the otic vesicle (arrowheads). (L) Horizontal section showing Ednra expression (arrowheads) in the first and second pharyngeal arches surrounding the central mesodermal core (asterisks).|
|Fig. 2. Ednra is required for neural crest development. (A–I) Efficiency of Ednra antisense morpholino oligonucleotide. (A) EdnraMO inhibits in vitro translation of Ednra in a dose-dependent fashion. Arrow indicates Ednra protein product. (B–I) EdnraMO inhibits in vivo expression of Ednra–GFP. (B,C) Embryos injected with mRNA encoding Ednra–GFP (1 ng/embryo) showing GFP fluorescence. (D,E) Embryos injected with Ednra–GFP mRNA (1 ng/embryo) and CoMO (30 ng/embryo). (F–I) Embryos injected with Ednra–GFP mRNA (1 ng/embryo) and EdnraMO (F–G, low dose, (l) 10 ng/embryo; H–I, high dose (h), 20 ng/embryo). No embryo shows GFP fluorescence at high dose. White arrows indicate the injected side. (J–S) Analysis of the formation of neural crest derivatives. Normal melanocyte formation (J, arrows) and Trp2 expression (M, arrows). EdnraMO-injected embryos show inhibition in melanocyte development (K) or Trp2 expression (N). The melanocyte formation was completely rescued by the coinjection of EdnraMO and Ednra′ mRNA (L). Induction of melanocytes in vitro by conjugating animal caps with prospective paraxial mesoderm cultured until the equivalent of St. 38. Control conjugates show melanocytes (O, arrows). Conjugates prepared with a 10 μM BQ123-soaked bead show no melanocyte formation (P). EdnraMO-injected (Q, arrowhead) embryos present a reduction of Meckel's and ceratohyal cartilages. (R) Schematic representation of EdnraMO effects on Xenopus head cartilages. M, Meckel's cartilage; Ir, infrarrostral; CH, ceratohyal; BH, basihyal; CB, ceratobranchial. (S) Normal cartilage formation was completely rescued by the coinjection of EdnraMO and Ednra′ mRNA.|
|Fig. 3. Ednra is required for early neural crest specification. (A–I) Effect of EdnraMO on neural crest specification. Arrows indicate the injected side. (A'–D') Transverse sections of embryos at the level of cephalic neural crest. (A, B) Ednra-depleted embryos fail to express neural crest markers Snail2 (A) and FoxD3 (B). (A, B, insets) The injected side is recognized by the fluorescence of the lineage tracer FLDx. (C, D) Expression of neural plate marker Sox2 and epidermal marker XK81a is expanded on the EdnraMO-injected side. (E) Embryo processed by double in situ hybridization for Sox2 and XK81a genes showing the reduction of prospective neural crest domain in the injected side. (F) Injection of control morpholino (CoMO) showed no effect. (G) Coinjection of EdnraMO and Ednra′ mRNA rescues Snail2 expression. (H) Single injection of Ednra′ mRNA expanded the Snail2 expression domain. (I) Quantification of rescue experiments. Results are expressed as percentage of embryos normally expressing Snail2 for each treatment (yellow bars), and as percentage of embryos showing reduced Snail2 expression (purple bars). Two different concentrations of Ednra′ mRNA were coinjected for rescue experiments (0.5 ng/embryo and 0.9 ng/embryo). (J–N) Ednra participates in the early neural crest specification. Ednra-injected embryos show increased expression of Snail2 (J) and FoxD3 (K). Expression of the neural plate marker Sox2 (L) and epidermal marker XK81a (M) are reduced on the injected side. (N) Double in situ hybridization for Sox2 and XK81a shows the enlargement of prospective neural crest territory in the injected side. Broken line, dorsal midline; brackets indicate the width of the neural plate (C, L), the width of neural crest domain (E, N) or the width of the neural plate plus the neural crest domain (D, M).|
|Fig. 4. Temporal requirement of Edn1/Ednra signaling for neural crest induction. Stage 12 (A–J), 14 (K–R) or 16 (S) embryos were grafted on the right neural fold with an Edn1 peptide-soaked bead or with the specific Ednra-inhibitor BQ123. Embryos were cultured until stage 14 (A–J) or 18 (K–S), when the expression of marker genes was analyzed. Arrowheads indicate the grafted side. Red circles indicate the position of Edn1- or BQ123-soaked beads. (A–D, K–M) Edn1 peptide increases the expression of neural crest markers Snail2 (A, K) and FoxD3 (B, L). The marker Sox2 (C) and XK81a are reduced (D). (S) No changes in the expression of FoxD3 were observed when Edn1-soaked beads were grafted into stage-16 embryos. (E–H, N–P) BQ123 leads to a reduction in the expression of neural crest markers FoxD3 (E, N) and Snail2 (F, O) on the treated side. BQ123 treatment produces an increase in the expression of Sox2 (P). (I–J, Q–R) Control embryos grafted with BSA-soaked bead. No effect on the expression of neural crest (I, Q; FoxD3) or neural plate (J, R; Sox2) markers is observed.|
|Fig. 5. Edn1/Ednra signaling is necessary for neural crest maintenance. (A) Schematic representation of transverse section mid-neurula embryos and the different neural crest explants dissected out for the experiments. Blue, neural crest; red, mesoderm: s, paraxial mesoderm; IM, intermediate mesoderm; orange, notochord; yellow, endoderm. Neural crest explants (NC) were prepared by dissection of the neural crest (− M), or by including the underlying mesoderm (+ M). NC explants dissected at stage 16-embryos express Snail2 (B), and FoxD3 (H). NC(− M) explants dissected from stage 16-embryos and cultured until equivalent stage 22 have lost Snail2 (C) and FoxD3 (I) expression. NC(+ M) explants taken from stage 16-embryos and cultured until stage 22, express Snail2 (D) and FoxD3 (J) markers. NC(− M) explants cultured in 10 μM Edn1 peptide until equivalent stage 22 show positive Snail2 (F) and FoxD3 (L) expression. In NC(+ M) explants cultured in 10 μM BQ123 the expression of Snail2 (G) and FoxD3 (M) was inhibited. Control animal cap explants were cultured until equivalent stage 22 in 10 μM Edn1 peptide. No expression of Snail2 (E) or FoxD3 (K) was induced.|
|Fig. 6. Analysis of Ppet-1 and ECE-1 expression. (A) RT-PCR analysis of the expression of Ednra, Ppet-1 and ECE-1 in neural crest and mesodermal explants. Explants were dissected out from stage-16 embryos (see Materials and methods). E, stage-16 embryo; AC, animal caps; NC, neural crest explants; M, mesodermal explants. EF1α, loading control. (B–C) Expression pattern of Xenopus laevis Preproendothelin-1 and Endothelin Converting Enzyme-1. (B) Preproendothelin-1 (Ppet-1) transcripts are detected by whole-mount in situ hybridization during mid-neurula stage (St. 16) in the mesoderm. (B') Transversal section of (B), Ppet-1 is expressed in somite (s) and lateral mesoderm (asterisk). The broken line indicates the tissues that were dissected out for the isolation of neural crest explants containing the mesoderm underlying the neural crest (see Fig. 5). (C) Endothelin Converting Enzyme-1 (ECE-1) is expressed during mid-neurula stage (St. 16) in the dorsal ectoderm and mesoderm. (C') Transversal section of (C). s, somite; n, notochord.|
|Fig. 8. The Edn1/Ednra signaling control on the maintenance of neural crest specification can be dissociated from apoptosis. (A) Snail2GR-injected neural crest (NC) explant dissected out and fixed at stage 16. In control explants the Snail2GR inducible construct was not activated. Explants show the normal apoptotic pattern. (B) Control animal cap explants cultured until equivalent stage 22 in 10 μM Edn1 show no inhibition of normal apoptosis. NC(− M) or NC(+ M) explants from Snail2GR-injected embryos (C–J) and XBcl2-injected embryos (K–R) were dissected at stage 16 as indicated in Fig. 5A. Snail2GR construct was activated at stage 16 by adding dexamethasone (10 μM final concentration). Explants were processed for TUNEL staining (C–F, K–N) or in situ hybridization (G–J, O–R). (C–F, K–N, S) All Snail2GR- and XBcl2-injected explants show a marked inhibition of apoptosis, regardless the presence or not of mesodermal tissue into the explant, or the incubation with 10 μM Edn1 peptide or 10 μM BQ123. The expression of FoxD3 was lost in Snail2GR-injected and XBcl2-injected (G, O; respectively) NC(− M) explants. In Snail2GR-injected and XBcl2-injected NC(+ M) explants the expression of FoxD3 marker was maintained (H, P; respectively). Snail2GR-injected and XBcl2-injected NC(− M) explants incubated in 10 μM Edn1 show the expression of FoxD3 (I, Q; respectively). Snail2GR-injected and XBcl2-injected NC(+ M) explants incubated in 10 μM BQ123 failed to express FoxD3 marker (J, R). (S) Quantification of apoptosis in Snail2GR and XBcl2-injected neural crest explants. Results are expressed as number of apoptotic nuclei/explant ± S.D. Control refers to the TUNEL staining of NC explants taken from stage 16-embryos. Every condition or treatment in Snail2GR-injected explants was significantly different from control explants (, P < 0.001). XBcl2-injected NC(− M) and NC(+ M) explants were statistically different from control, P < 0.001. () XBcl2-injected NC(− M)+Edn1 and NC(+ M) + BQ123 explants were statistically different from control, P < 0.01.|
|Fig. 9. Ednra in the genetic cascade that specifies the neural crest. Ednra lies downstream Msx1 and upstream Sox9–Sox10. (A–O) Representative examples of embryos injected with the indicated reagents. The graphs at the right side represent the percentage of embryos with the indicated phenotypes; the letter in each column correspond to the treatment described for each figure in panels A–O. (A, B) Embryos injected with Msx1-GR show increased expression of Ednra and neural crest marker FoxD3. (C) Msx1-GR construct failed to increase the expression of FoxD3 marker when was coinjected with EdnraMO. (C') EdnraMO leads to a reduction of FoxD3 expression in the injected side (arrowhead). (D, E) Msx1-dominant negative inducible construct (HDMsx1-GR) inhibited the expression of Ednra and FoxD3. (F) The coinjection of dominant negative HDMsx1-GR construct and Ednra mRNA rescue the expression of FoxD3 in the neural crest. (F') Ednra overexpression leads to an increased FoxD3 expression in the injected side (arrowhead). (G) No changes in the expression of Msx1 was observed in embryos microinjected with Ednra mRNA. (H, I) Ednra mRNA-microinjected embryos show expanded expression of Sox9 and Sox10 in the injected side (arrowhead). (J–L) EdnraMO-microinjected embryos show no effect on Msx1 (J), but inhibition of Sox9 (K) and Sox10 (L) expression in the injected side (arrowhead). (M–O) Sox9 mRNA and Sox10 mRNA rescue FoxD3 expression in EdnraMO-injected embryos. (N') Expansion of FoxD3 in embryo injected with Sox9. (O') Expansion of FoxD3 in embryo injected with Sox9.|
|Fig. 10. Edn1/Ednra signaling is required for neural crest migration. (A–C) One dorsal blastomere of a 4–8-cell embryo was injected with 20 ng/embryos EdnraMO (A). EdnraMO-injected side (B) show arrested neural crest migration. Note FoxD3-expressing neural crest cells accumulated lateral to the hindbrain. (D, G) Embryos were grafted before the onset of neural crest migration (St. 18) with Edn1- or BQ123-soaked beads, fixed between St. 21–23, and the expression of Snail2 or FoxD3 analyzed. The midline (upper line) and the leading edge of migration are indicated by broken lines. Embryos grafted with Edn1-soaked beads show an increased neural crest migration (E). Embryos grafted with BQ123-soaked bead show inhibition of neural crest migration (H). (C, F, I) Control side showing normal neural crest migration.|
|Fig. S4. . (A, A') EdnraMO injection produces no changes in cell proliferation. The mitotic nuclei were visualized by whole-mount immunostaining using anti-Phospho-Histone H3 antibody. White arrows indicate the injected side. (B) Transverse histological section of a TUNEL stained EdnraMO-injected embryo. Arrowhead indicates the injected side. Apoptotic nuclei are found in the superficial tissues and in the deep layer of the ectoderm. (C) The injection of EndraMO (20 ng) and Ednra′ mRNA (1 ng) was able to reduce the apoptosis in the injected side of stage 16 embryos. Results are expressed as percentage of embryos showing increased TUNEL labeling in the injected side ± S.D. Control embryos were injected only with lineage tracer.|
|ednra (endothelin receptor type A) gene expression in Xenopus laevis embryo via in situ hybridization, NF stage15, dorsal view, anterior right.|
|ednra (endothelin receptor type A) gene expression in Xenopus laevis embryo via in situ hybridization, NF stage 24, lateral view, anterior right.|
|ednra (endothelin receptor type A) gene expression in Xenopus laevis embryo via in situ hybridization, NF stage 30, lateral view, anterior right.|