The activity of Pax3 and Zic1 regulates three distinct cell fates at the neural plate border

Abstract

In Xenopus, the neural plate border gives rise to at least three cell populations: the neural crest, the preplacodal ectoderm, and the hatching gland. To understand the molecular mechanisms that regulate the formation of these lineages, we have analyzed the role of two transcription factors, Pax3 and Zic1, which are among the earliest genes activated in response to neural plate border-inducing signals. At the end of gastrulation, Pax3 and Zic1 are coexpressed in the neural crest forming region. In addition, Pax3 is expressed in progenitors of the hatching gland, and Zic1 is detected in the preplacodal ectoderm. Using gain of function and knockdown approaches in whole embryos and animal explants, we demonstrate that Pax3 and Zic1 are necessary and sufficient to promote hatching gland and preplacodal fates, respectively, whereas their combined activity is essential to specify the neural crest. Moreover, we show that by manipulating the levels of Pax3 and Zic1 it is possible to shift fates among these cells. These findings provide novel information on the mechanisms regulating cell fate decisions at the neural plate border.

Figure 1.

Molecular characterization of neural plate and NPB cells. (A) Dorsal view (anterior to top) of a schematic representation of an early neurula stage Xenopus embryo highlights the position of the neural plate, nonneural ectoderm, and intervening NPB cells that include the prospective HG, NC, and PE. (B) Schematic representation of a transverse section through a neurula stage embryo in which the different components of the ectoderm are color coded according to A. en, endoderm; so, somite; no, notochord. (C) Expression of neural plate- (Sox2) and NPB (Snail2, Xhe, and Six1)-specific genes by whole-mount in situ hybridization. Dorsal views, anterior to top. (D) Double in situ hybridization showing the relationship of Snail2 (NC; arrowhead) and Xhe (HG; arrow) expression domains. Lateral view, anterior to right. (E) Transverse section through neurula stage embryos illustrating the relative position of the presumptive HG (Xhe; arrows), NC (Snail2; arrowheads), and PE (Six1; bracket). (F) Schematic representation of cross-sections through neurula stage embryo showing the domain of expression of the Sox2, Xhe, Snail2, and Six1 (purple).

Figure 2.

Regulation of NPB cell fates by Bmp attenuation and Wnt signaling. Embryos at the two-cell stage were injected in the animal pole with increasing doses of Noggin (N) mRNA (0.1, 0.4, 1.0, and 2.0 ng) alone or in combination with Wnt1 (+Wnt) mRNA (100 pg). Animal explants were dissected at the blastula stage, cultured for 5 or 10 h, and analyzed by real-time RT-PCR for the expression of four putative NPB specifiers, Pax3, Zic1, Msx1, and Snail1 (A), and a number of cell type-specific genes: Snail2, Sox8, Twist, and cMyc for NC (B); Xhe and Crisp for HG (C); Six1 and Eya1 for PE (D); Sox2 and Ncam for neural plate (E); Xcg for cement gland (F); and Keratin for epidermis (G). Un, uninjected animal explant. Each value was normalized to the level of EF1α expression.

Figure 3.

Expression of Pax3 and Zic1 at the NPB. (A) Developmental expression of Pax3 and Zic1 at the NPB in stage-matched embryos (the stages are indicated in the top right corner). Dorsal views, anterior to top. (B) Transverse section through neurula stage embryos illustrating the relative position of Pax3 and Zic1 in the deeper layer of the ectoderm (arrowheads). Pax3 is also detected in the superficial layer of the ectoderm (arrows), and Zic1 expression domain expands more laterally (bracket) than Pax3. (C) Schematic representation of cross-sections through neurula stage embryo showing the domain of expression of the Pax3 and Zic1 (purple). (D) Xhe and Pax3 are coexpressed in the HG at the tailbud stage. Frontal views.

Figure 4.

Regulation of cell fates at the NPB by Fgf8 signaling. (A) Embryos injected with increasing amounts of Fgf8a mRNA (0.1, 0.5, 5, and 50 pg) in one blastomere at the two-cell stage and analyzed by whole-mount in situ hybridization for Pax3, Zic1, Snail2, Xhe, and Six1 expression. Pax3, Zic1, Snail2, and Xhe are expanded on the injected side by lower doses of Fgf8 (0.1 or 0.5 pg; white arrows). The expression of these genes is expanded on the contralateral side (yellow arrows) and repressed on the injected side (black arrows) for higher doses of Fgf8a (5 or 50 pg). Six1 expression is repressed unilaterally or bilaterally for intermediate (0.5 pg) or high (5 and 50 pg) doses of Fgf8a, respectively. (B) Embryos injected with Fgf8a-specific morpholino (Fgf8aMO) in the marginal zone of four-cell stage embryos exhibit a strong reduction of Pax3, Snail2, Xhe, and Six1 expression at the neurula stages (black arrows). In the absence of Fgf8a function, Zic1 anterior expression domain is expanded (white arrow), whereas posteriorly Zic1 is reduced (black arrow). In A and B, embryos hybridized with Pax3, Zic1, and Snail2 are viewed from the dorsal side, anterior to top. Anterior views are shown for embryos hybridized with Xhe and Six1. In all panels, the injected side is on the right.

Figure 5.

Pax3 is sufficient and required to promote HG fate. (A) Embryos injected with Pax3GR and treated with dexamethasone at the gastrula stage hatch several hours earlier than their uninjected siblings. (B) Stage 17 embryo that received unilateral injection of Pax3GR at the two-cell stage and treated with dexamethasone (+Dex) exhibit a strong ectopic expression of Xhe on the injected side (right side), whereas sibling embryos cultured in the absence of dexamethasone (−Dex) are unaffected. Dorsal views, anterior to top. (C) In animal explants, Pax3GR strongly induces Xhe and Crisp expression 4 h after dexamethasone treatment (+Dex). Pax3-mediated induction of Xhe occurs independently of protein synthesis (+CHX). Another HG-specific gene, Crisp, is significantly reduced in the absence of protein synthesis. (D) Embryos injected with Pax3 morpholino (Pax3MO) exhibit a strong reduction of Snail2 and Xhe expression at the neurula and tailbud stages, respectively (arrows). Sox2 and Zic1 expression were expanded in Pax3-depleted embryos (arrows). Embryos hybridized with Snail2, Sox2, and Zic1 are viewed from the dorsal side, anterior to top. For Xhe staining, anterior view is shown. In all cases, the injected side is indicated by an arrow. (E) Noggin+Wnt activates expression of Snail2 and Xhe in animal explants, an activity that is inhibited by coinjection of Pax3MO.

Figure 6.

Distinct thresholds of Pax3 activity regulate HG and NC fates. (A) Embryos injected with increasing amounts of Pax3GR mRNA in one blastomere at the two-cell stage and treated with dexamethasone at the gastrula stage were analyzed by whole-mount in situ hybridization for Xhe and Snail2 expression at stage 17. Although Snail2 is expanded laterally and anteriorly by lower doses of Pax3GR, Snail2 is repressed for higher doses of Pax3GR. Ectopic expression of Xhe is proportional to the amount of Pax3GR mRNA injected. Dorsal views, anterior to top. In all panels, the injected side is on the right. (B) In animal explants, the induction of Snail2 and Xhe as determined by real-time RT-PCR depends on different amounts of injected Pax3GR. Zic1GR strongly synergizes with Pax3 to induce Snail2. The concentrations of Pax3GR mRNA injected are in nanograms. A single dose of Zic1GR mRNA (0.25 ng) was coinjected as indicated.

Figure 7.

Zic1 promotes the expression of Six1 an activity repressed by Pax3. (A) In animal explants, Zic1GR is sufficient to activate Six1 expression as determined by real-time RT-PCR. Pax3GR strongly represses Zic1-mediated Six1 induction to promote NC (Snail2) and HG (Xhe) fates. The concentrations of Zic1GR mRNA injected are in nanograms. A single dose of Pax3GR mRNA (0.25 ng) was coinjected as indicated. (B) Embryos injected with Zic1 morpholino (Zic1MO) exhibit a strong reduction of Snail2 and Six1 (arrows). The neural plate expression of Sox2 is expanded, whereas Sox2 PE expression domain is reduced (arrow). In these embryos lacking Zic1 function, Xhe expression is altered (arrow), presumably as a consequence of reduced Pax3 expression (arrow). Embryos are viewed from the dorsal side, anterior to top (Snail2 and Pax3). In Six1, Sox2, and Xhe staining, anterior views are shown, dorsal to top. The injected side is indicated by an arrow. (C) Noggin-mediated induction of Six1 is inhibited in the absence of Zic1 function (+Zic1MO). Noggin+Wnt activates expression of Snail2 and Xhe in animal explants, an activity that is inhibited by coinjection of Zic1MO.

Figure 8.

Model for the specification of NPB cells. In response to Bmp, Wnt, and Fgf signaling, the transcription factors Pax3 and Zic1 are activated at the NPB. The activity of these two factors regulates the specification of three NPB cell populations. When Pax3 and Zic1 reach an appropriate balance, NC is induced (A). Variations in the relative activity of these two transcriptional regulators at the NPB can disrupt this balance, either toward Pax3 to promote HG formation (B) or toward Zic1 to promote PE fate (C).