Correct anteroposterior patterning of the zebrafish neurectoderm in the absence of the early dorsal organizer

Abstract

Background: The embryonic organizer (i.e., Spemann organizer) has a pivotal role in the establishment of the dorsoventral (DV) axis through the coordination of BMP signaling. However, as impaired organizer function also results in anterior and posterior truncations, it is of interest to determine if proper anteroposterior (AP) pattern can be obtained even in the absence of early organizer signaling.

Results: Using the ventralized, maternal effect ichabod (ich) mutant, and by inhibiting BMP signaling in ich embryos, we provide conclusive evidence that AP patterning is independent of the organizer in zebrafish, and is governed by TGFβ, FGF, and Wnt signals emanating from the germ-ring. The expression patterns of neurectodermal markers in embryos with impaired BMP signaling show that the directionality of such signals is oriented along the animal-vegetal axis, which is essentially concordant with the AP axis. In addition, we find that in embryos inhibited in both Wnt and BMP signaling, the AP pattern of such markers is unchanged from that of the normal untreated embryo. These embryos develop radially organized trunk and head tissues, with an outer neurectodermal layer containing diffusely positioned neuronal precursors. Such organization is reflective of the presumed eumetazoan ancestor and might provide clues for the evolution of centralization in the nervous system.

Conclusions: Using a zebrafish mutant deficient in the induction of the embryonic organizer, we demonstrate that the AP patterning of the neuroectoderm during gastrulation is independent of DV patterning. Our results provide further support for Nieuwkoop's "two step model" of embryonic induction. We also show that the zebrafish embryo can form a radial diffuse neural sheath in the absence of both BMP signaling and the early organizer.

Figure 1

Two independent ways of inhibiting BMP signaling result in massive ectopic expression of chordin. (A-E) BMP signaling detected by immunohistochemistry using an anti phospho-Smad1/5/8 antibody; green dots indicate nuclei containing the phosphorylated Smad. (F-J) In situ hybridization with chd probe. (K-O) Expression of the epidermal marker p63. Wild-type embryos (A,F,K) are shown with in lateral view with ventral side to the left. Untreated ich embryos (B,G,L), and ich embryos injected with bmp2bMO (C,H,M), or with the two βcatMOs (D,I,N), or with all three MOs (E,I,O) are shown in lateral views. All embryos are at ~70% epiboly.

Figure 2

Inhibition of BMP signaling in ich embryos does not result in organizer induction. In situ hybridization with boz (A,B,C), gsc (D,E,F) and chd (G,H,I) probes in uninjected ich and bmp2bMO-injected ich embryos demonstrate that while impaired BMP signaling results in the transcriptional derepression of chd at the margins of the embryo (H,I), other characteristic organizer genes such as boz and gsc are not induced (C and F). (In A-C dome staged embryos are shown from an animal pole view, D-F and G-I present lateral views of 50% epiboly and 70% epiboly staged embryos, respectively. For wild-type embryos, dorsal is on the left.)

Figure 3

Embryos inhibited in both BMP signaling and organizer function form neuroectoderm with correct AP pattern at the end of gastrulation and this patterning can be modulated by similar mechanisms as in wild-type embryos. Single in situ hybridization with cyp26 (A,D,G,J,M,P,S,V,Y) or hoxb1b (B,E,H,K,NQ,T,W,Z) probes or double in situ hybridization with both probes (C,F,I,L,O,R,U,X,A') is shown for wild-type embryos (A-C), untreated ich embryos (D-F), or ich embryos treated with bmp2bMO (G-I), the two β-cat MOs (J-L), bmp2bMO plus the two β-cat MOs (M-O), or ich embryos injected with antivin mRNA (P-R), antivin mRNA plus bmp2bMO (S-U), SU5402 (V-X), or SU5402 plus bmp2bMO (Y-A'). Wild-type embryos [A-C] are shown in dorsal views, while ich embryos are shown in lateral views. The neuroectodermal and tailbud expression domains of cyp26 in wild-type embryos are marked with arrow and star, respectively. All embryos are at ~100% epiboly.

Figure 4

AP neurectodermal patterning is retained at 11 hpf in embryos lacking BMP signaling and an active organizer. Single in situ hybridization with rx3 (A,D,G,J,M) or fkd3 (B,E,H,K,N) probes or double in situ hybridization with both probes (C,F,I,L,O) is shown for wild-type embryos (A-C), untreated ich embryos (D-F), or ich embryos treated with bmp2bMO (G-I), the two β-cat MOs (J-L), or bmp2bMO plus the two β-cat MOs (M-O). In wild-type embryos, rx3 demarcates the eye-field (A,C) and fkd3 the diencephalon (arrows) and rhombomere 5 (arrowheads) of the hindbrain (B,C).

Figure 5

Germ layer segregation in 'ciuffo' embryos. Sections through the posterior protrusions of 24hpf 'ciuffo' embryos stained with characteristic germ layer markers show that neuroectoderm specific markers. islet1 (A) and krox20 (B) are expressed in the outermost layer of the protrusion. The endodermal marker gata5 (C) is expressed in the innermost layers, whereas the mesodermal marker myoD (B) can be observed in between. Sections were made of 24 hpf ich embryos that had been injected with βcatMO1 and βcatMO2, hybridized as whole mounts with the indicated probes.

Figure 6

Patterning of the zebrafish neurectoderm is achieved by the parallel action of posteriorizing "transformative" signals and "activating" BMP-antagonists. (A-J) Neurectodermal regions are color coded as indicated with the most anterior territory represented in orange and the most posterior in blue. The BMP signaling gradient is indicated by grey shading, so that a brownish shading superimposed on the colored regions represent areas of the embryo unable to form neurectoderm due to the absence of an "activating" signal. Embryos of indicated genotype and treatment are diagrammed at 80-90% epiboly. See text for details.