Cas9-mediated excision of Nematostella brachyury disrupts endoderm development, pharynx formation and oral-aboral patterning

Abstract

The mesoderm is a key novelty in animal evolution, although we understand little of how the mesoderm arose. brachyury, the founding member of the T-box gene family, is a key gene in chordate mesoderm development. However, the brachyury gene was present in the common ancestor of fungi and animals long before mesoderm appeared. To explore ancestral roles of brachyury prior to the evolution of definitive mesoderm, we excised the gene using CRISPR/Cas9 in the diploblastic cnidarian Nematostella vectensis Nvbrachyury is normally expressed in precursors of the pharynx, which separates endoderm from ectoderm. In knockout embryos, the pharynx does not form, embryos fail to elongate, and endoderm organization, ectodermal cell polarity and patterning along the oral-aboral axis are disrupted. Expression of many genes both inside and outside the Nvbrachyury expression domain is affected, including downregulation of Wnt genes at the oral pole. Our results point to an ancient role for brachyury in morphogenesis, cell polarity and the patterning of both ectodermal and endodermal derivatives along the primary body axis.

Keywords: Brachyury; Cnidarian; Endoderm; Mesoderm; Nematostella; Pharynx.

Fig. 1.

Nematostella oral-aboral axis and gene expression domains. (A) Evolutionary relationships among metazoa (Dunn et al., 2014). (B) A Nematostella juvenile, showing the oral-aboral axis and tissue layers. (C) Gene expression domains as defined by Röttinger et al. (2012): lateral view on the left; oral view on the right.

Fig. 2.

Nvbra expression in uninjected control and Nvbra/Cas9 embryos. In situ hybridization showing Nvbra expression in (A) uninjected embryos and (B) sibling embryos injected with Nvbra gRNAs and Cas9 (Nvbra/Cas9 embryos). Most control embryos show characteristic staining around the oral pole. Most Nvbra/Cas9 embryos show no staining; some show a smaller region of staining, whereas only a few show the normal staining pattern.

Fig. 3.

Pharyngeal development and ectodermal cell polarity, but not cnidocyte differentiation of surface ectodermal cells, are disrupted after Nvbra excision. (A-D) Control embryos were injected with Cas9 only. (A,B) At 48 h post-fertilization (hpf) the blastopore is visible (arrow) and a well-defined epithelial endoderm has formed (dotted line). Ectodermal cells protrude into the archenteron to form the pharynx rudiment (arrowhead). (C,D) By 72 hpf the ectoderm has extended well into the blastocoel (arrowhead). (E-H) In Nvbra/Cas9 embryos, the pharynx fails to form. (E,F) At 48 hpf, the blastopore is visible (arrow) but the endoderm is thicker than in controls (compare dotted lines in B and F) and appears disorganized. No pharyngeal ectoderm extends into the archenteron. (G,H) By 72 hpf, the blastopore is still evident but no pharynx has formed. The endoderm is highly disorganized with cells filling the blastocoel, and embryos appear rounded when compared with the elongated phenotype of control embryos. (I-L) Cas9 control embryos develop cnidocytes normally. At 48 hpf (I) and 72 hpf (J), cnidocytes (red, anti-mcol4) are abundant throughout the ectoderm. Control embryos (72 hpf) also exhibit ectodermal expression of (K) Lgl and highly polarized expression of (L) aPKC (arrowhead) in the apical cortex of ectodermal cells. (M-P) Cnidocytes are present in the ectoderm of Nvbra/Cas9 embryos at both (M) 48 hpf and (N) 72 hpf. (O) Lgl is still basolateral and restricted to the ectodermal cells; there is abrupt cessation of Lgl staining at the blastopore (arrow, compare with K). Expression of aPKC (P) is no longer apically restricted, spreading into the basal regions of ectodermal cells (arrowhead). Ratios in the bottom left corner indicate the number of embryos showing the indicated phenotype (wild type or abnormal)/total number of embryos counted; only embryos with proper ectoderm formation were counted in order to exclude dead or clearly abnormal embryos. Phenotypes of 72 hpf Nvbra/Cas9 embryos are quantified in more detail in Fig. S5. The blastopore is on the left of each image, indicated by an asterisk in the first column. Images A-H and K,L,O,P are single optical sections; images I,J,M,N are 3D rendered from z-stacks.

Fig. 4.

qPCR of genes of the blastoporal gene regulatory network. Bars indicate relative levels of expression of genes at 24 and 48 hpf. Samples were normalized to ribosomal protein P0. Reductions in expression are shown as the negative reciprocal of the expression level, in order to facilitate visualization. Asterisks indicate significant differences (*P<0.05).

Fig. 5.

Gene expression in control and Nvbra/Cas9 embryos. In situ hybridization of 48 hpf control (left columns) and Nvbra/Cas9 embryos (right columns). In all images, oral is towards the left. Genes are organized according to their normal expression domain at 24 hpf (Fig. 1C). Bars to the right of each figure indicate the proportion of each phenotype observed (see key). Numbers of embryos scored for each panel are in Table S2.

Fig. 6.

Effects of Nvbra on the blastoporal gene regulatory network. Gene regulatory relationships in the endomesodermal GRN described by Röttinger et al. (2012). Because Nvbra is a transcription factor, we assume that Nvbra can directly affect only genes in its expression domain (the central ring). Effects on domains outside the central ring are assumed to be mediated by signaling molecules; indirect effects are indicated by the broken lines between regions of the embryo. Dotted arrows indicate interactions inferred from qPCR. Solid arrows indicate interactions inferred from both qPCR and in situ hybridization. Genes analyzed by in situ hybridization alone are not included. Upper panel: 24 h. Lower panel: 48 h. Genes are arrayed according to their normal expression domain at 24 h. Genes closer to the top of the diagram are transcribed earlier in development; those at successively lower positions are transcribed later (Röttinger et al., 2012).