Genetic ablation of neural crest cell diversification

Abstract

The neural crest generates multiple cell types during embryogenesis but the mechanisms regulating neural crest cell diversification are incompletely understood. Previous studies using mutant zebrafish indicated that foxd3 and tfap2a function early and differentially in the development of neural crest sublineages. Here, we show that the simultaneous loss of foxd3 and tfap2a function in zebrafish foxd3(zdf10);tfap2a(low) double mutant embryos globally prevents the specification of developmentally distinct neural crest sublineages. By contrast, neural crest induction occurs independently of foxd3 and tfap2a function. We show that the failure of neural crest cell diversification in double mutants is accompanied by the absence of neural crest sox10 and sox9a/b gene expression, and that forced expression of sox10 and sox9a/b differentially rescues neural crest sublineage specification and derivative differentiation. These results demonstrate the functional necessity for foxd3 and tfap2a for neural crest sublineage specification and that this requirement is mediated by the synergistic regulation of the expression of SoxE family genes. Our results identify a genetic regulatory pathway functionally discrete from the process of neural crest induction that is required for the initiation of neural crest cell diversification during embryonic development.

Fig. 1.

foxd3^zdf10;tfap2a^low double mutant embryos lack NC derivatives. (A) Live larvae at 4 days post-fertilization (dpf), lateral views with either transmitted (left) or reflected (right) light. Wild-type zebrafish exhibit three NC-derived pigment cell types: black melanophores (left, arrows), yellow xanthophores (left, arrowheads) and iridescent iridophores (right, white arrows). foxd3^zdf10 and tfap2a^low single mutants develop essentially normal pigment patterns by 4 dpf. Jaw structures protrude ventrally in both foxd3^zdf10 and tfap2a^low single mutants (asterisks). foxd3^zdf10;tfap2a^low double mutants completely lack NC-derived pigment cells except for occasional head xanthophores. Jaw structures appear to be missing (asterisk). (B) Craniofacial cartilage development revealed by Alcian Blue staining at 4 dpf, ventral views with anterior to the left. The wild-type larval head skeleton of zebrafish consists of the dorsal neurocranium (ne), as well as upper (mandibular, m) and lower (hyoid, h, and ceratobranchial, ce) jaw structures. In both foxd3^zdf10 and tfap2a^low single mutants, mandibular and hyoid structures are disorganized and the ceratobranchial elements are absent, whereas the neurocranium remains intact. Double mutants lack upper and lower jaw structures, and all but the most posterior portion of the neurocranium. (C) Immunostaining with 16A11 monoclonal (anti-Hu) antibody at 3 dpf, lateral views, anterior to the left. DRG neurons are found in each trunk segment of wild-type embryos (arrowheads) and enteric neurons populate the gut tube (arrows). DRG neurons are absent in foxd3^zdf10 single mutants (asterisks) and present in reduced numbers in tfap2a^low single mutants (arrowheads), whereas enteric neurons are reduced in number in both backgrounds (arrows). DRG and enteric neurons are absent in double mutant embryos (asterisks). (D) tyrosine hydroxylase (th) expression in sympathetic neurons at 48 hpf in wild-type embryos (arrowhead). Sympathetic neurons are absent in foxd3^zdf10 and tfap2a^low single mutants, as well as in double mutants (asterisks).

Fig. 2.

Failure of NC sublineage specification in foxd3^zdf10;tfap2a^low double mutants. (A-D) Lateral views with anterior to the left. (A) mitf expression at 24 hpf (arrowheads). foxd3^zdf10;tfap2a^low double mutant embryos completely lack NC-derived melanoblast mitf expression. (B) xanthine dehydrogenase (xdh), diagnostic of xanthophore precursors, at 30 hpf. (C) endothelin receptor B (ednrb1), presumed to be expressed by all chromatophore precursors, at 30 hpf and (D) leukocyte tyrosine kinase (ltk), reported to be expressed by iridophore precursors, at 30 hpf. (E) Dorsal views of dhand expression, anterior to left, ventral of the caudal hindbrain, at 48 hpf. Sympathetic neuron precursors are labeled in wild-type embryos (arrowheads), as are the more ventral and laterally localized enteric neuron precursors (arrows). In the majority of foxd3^zdf10 embryos, both sympathetic and enteric neuron progenitors are absent (asterisk) (Stewart et al., 2006). In tfap2a^low mutants, sympathetic neuron precursors are absent whereas a small number of enteric precursors are present (arrow) (Knight et al., 2003). In foxd3^zdf10;tfap2a^low double mutants, both sympathetic and enteric neuron precursors are absent (asterisk). (F-J) Lateral (F-I) and dorsolateral views (J), anterior to the left. (F) dlx2 expression at 24 hpf. dlx2 expression (arrowheads) is absent in the branchial arches of double mutant embryos, but is retained in the forebrain (arrows). (G) dlx3 expression at 30 hpf; (H) dhand expression at 30 hpf. NC expression of both genes is absent in double mutant embryos with dlx3 expression maintained in non-NC otic vesicle (asterisks). (I) tbx1 expression in the branchial arches (asterisk) at 30 hpf is diagnostic of the endodermal component of the arches. tbx1 expression is normal in all experimental embryos. (J) endothelin 1 (edn1) expression in the mesodermal component of the branchial arches at 30 hpf. Branchial mesoderm appears to develop normally in foxd3^zdf10 and tfap2a^low single mutants, as well as in foxd3^zdf10;tfap2a^low double mutant embryos.

Fig. 3.

Defective expression of NC SoxE family genes in foxd3^zdf10;tfap2a^low double mutant embryos. (A,B,D-F) Dorsal views with anterior to the top of 6, 5 and 9 somite stage (s) embryos. (C) Lateral views with anterior to the left of 30 hpf embryos. (A) NC sox9b expression is severely depleted in foxd3^zdf10;tfap2a^low double mutants at the 6 somite stage, but is retained in the non-NC-derived otic placode (asterisks). (B) NC sox9b expression remains reduced in both single mutants, whereas NC expression is undetectable in double mutants by the 9 somite stage. (C) NC sox9a expression in the branchial arches (arrows) is absent in double mutant embryos. (D) There are slight reductions in NC sox10 expression in foxd3^zdf10 and tfap2a^low single mutants at the 5 somite stage, whereas in double mutant embryos NC sox10 expression is much more reduced. (E) By the 9 somite stage, NC sox10 expression is undetectable in double mutants. (F) In contrast to SoxE gene expression, NC foxd3 expression is maintained in foxd3^zdf10;tfap2a^low double mutant embryos, demonstrating that NC induction occurs.

Fig. 4.

Misexpression of SoxE genes differentially rescues NC sublineage specification and differentiation in foxd3^zdf10;tfap2aMO embryos. (A,E,H) sox9a and sox9b (sox9a/b) RNA co-injection with tfap2a morpholino (tfap2aMO) into foxd3^zdf10 mutant embryos (foxd3^zdf10-tfap2aMO). (A) Live 4 dpf larva, dorsal view, anterior to the left. Melanophore and xanthophore development is rescued in foxd3^zdf10-tfap2aMO embryos co-injected with sox9a/b RNA (arrows; compare with wild-type embryo, D). (E) Alcian Blue staining of a 4 dpf larva, vental views with anterior to the left. Misexpression of sox9a/b in foxd3^zdf10-tfap2aMO embryos results in rescue of the neurocranium, as well as of the mandibular and hyoid jaw structures. (H) dlx2 expression at 24 hpf, dorsal views, anterior to the left. Specification of NC craniofacial progenitors occurs in foxd3^zdf10-tfap2aMO embryos co-injected with sox9a/b RNA, based on dlx2 expression (arrows). (B,F,G,I,J) foxd3^zdf10-tfap2aMO embryos co-injected with sox10 RNA. (B) Live 4 dpf larva, dorsal view, anterior to the left. Consistent with a role in development of non-ectomesenchymal NC derivatives, sox10 misexpression rescues melanophore and xanthophore development in foxd3^zdf10-tfap2aMO embryos (arrows; compare with wild-type embryo, D). (F) Alcian Blue staining of 4 dpf larva, ventral view, anterior to the left. sox10 misexpression does not rescue cranial cartilage structures in foxd3^zdf10-tfap2aMO embryos. (G) 16A11 immunoreactivity, 3 dpf, dorsal view. Dorsal root ganglion neuron development is rescued by sox10 misexpression (arrowheads). (I) dlx2 expression at 24 hpf, dorsal view with anterior to the left. sox10 RNA injection also does not rescue NC dlx2 expression in foxd3^zdf10-tfap2aMO embryos (asterisks), indicating that sox10 function cannot rescue the specification of NC precursors for craniofacial cartilages. (J) th expression at 48 hpf, lateral view, anterior to the left. sox10 misexpression rescues sympathetic neuron development (arrow). (C) Live 4 dpf larva, dorsal view, anterior to the left. Morpholino-mediated knockdown of p53 in foxd3^zdf10-tfap2aMO embryos rescues melanophore, xanthophore (arrows; compare with wild-type embryo, D) and iridophore (not visible) development. (K-M) sox10-responsive NCCs persist in foxd3^zdf10-tfap2aMO embryos. 33 hpf embryos, lateral view, anterior to left. (K) Wild-type embryo with abundant melanophores. (L) In foxd3^zdf10-tfap2aMO embryos injected with a heat shock-inducible sox10 construct but not exposed to heat shock, melanophores fail to develop. (M) In foxd3^zdf10-tfap2aMO embryos injected with a heat shock-inducible sox10 construct and heat shocked at 24 hpf, robust melanogenesis occurs.