Non-canonical Notch signaling represents an ancestral mechanism to regulate neural differentiation

Abstract

Background: Cellular differentiation is a critical process during development of multicellular animals that must be tightly controlled in order to avoid precocious differentiation or failed generation of differentiated cell types. Research in flies, vertebrates, and nematodes has led to the identification of a conserved role for Notch signaling as a mechanism to regulate cellular differentiation regardless of tissue/cell type. Notch signaling can occur through a canonical pathway that results in the activation of hes gene expression by a complex consisting of the Notch intracellular domain, SuH, and the Mastermind co-activator. Alternatively, Notch signaling can occur via a non-canonical mechanism that does not require SuH or activation of hes gene expression. Regardless of which mechanism is being used, high Notch activity generally inhibits further differentiation, while low Notch activity promotes differentiation. Flies, vertebrates, and nematodes are all bilaterians, and it is therefore unclear if Notch regulation of differentiation is a bilaterian innovation, or if it represents a more ancient mechanism in animals.

Results: To reconstruct the ancestral function of Notch signaling we investigate Notch function in a non-bilaterian animal, the sea anemone Nematostella vectensis (Cnidaria). Morpholino or pharmacological knockdown of Nvnotch causes increased expression of the neural differentiation gene NvashA. Conversely, overactivation of Notch activity resulting from overexpression of the Nvnotch intracellular domain or by overexpression of the Notch ligand Nvdelta suppresses NvashA. We also knocked down or overactivated components of the canonical Notch signaling pathway. We disrupted NvsuH with morpholino or by overexpressing a dominant negative NvsuH construct. We saw no change in expression levels for Nvhes genes or NvashA. Overexpression of Nvhes genes did not alter NvashA expression levels. Lastly, we tested additional markers associated with neuronal differentiation and observed that non-canonical Notch signaling broadly suppresses neural differentiation in Nematostella.

Conclusions: We conclude that one ancestral role for Notch in metazoans was to regulate neural differentiation. Remarkably, we found no evidence for a functional canonical Notch pathway during Nematostella embryogenesis, suggesting that the non-canonical hes-independent Notch signaling mechanism may represent an ancestral Notch signaling pathway.

Keywords: Cellular differentiation; Evolution; Nematostella vectensis; Notch; Nvnotch.

Figure 1

Nvnotch and Nvdelta embryonic expression. Expression of Nvdelta(A,B) and Nvnotch(C,D) is shown at early gastrula (A,C) and late gastrula (B,D) stages. Nvdelta is expressed in a “salt and pepper” expression pattern at early gastrula (A), and ubiquitously expressed at late gastrula (B), though there are cells enriched for Nvdelta in the late gastrula (B, arrows). Clusters of cells distributed in a “salt and pepper” pattern express Nvnotch in the early gastrula stages (C). By late gastrula, Nvnotch appears to have low-level ubiquitous expression (D). Images are lateral views taken from a superficial focal plane; oral is to the left.

Figure 2

Activation of Nvnotch suppresses NvashA expression. Images of animals stained for NvashA by in situ hybridization are shown (A-F). All images are lateral views with oral to the left. The relative focal plane is indicated to the left of each row of images. Animals with control wild-type Notch activity (A,D), with Notch activity reduced by injection of a Nvnotch morpholino (MO) (B,E), and with Notch activity overactivated by overexpression (OE) of the Nvnotch intracellular domain (Nvnicd) (C,F) are shown. (G) Quantitative (q)PCR analysis of the relative expression of NvashA is compared in animals with reduced Notch activity (DAPT, Nvnotch MO, Nvdelta MO) and increased Notch activity (Nvnicd OE), and to animals injected with a control MO. The red rectangle indicates a relative fold change of −1.5 to 1.5, which we consider to correspond with no change in expression level. (H) Quantification of the average number of NvashA-positive cells counted in the aboral domain (see Methods). N ≥20 animals counted for each treatment.

Figure 3

Nvdelta activates Nvnotch activity to suppress NvashA. (A-B) Shown are Aboral views of NvashA expression in control (A) and Nvdelta overexpressing (OE) (B) animals. Phenotypic classes were scored as no expression, weak, wild-type (WT) levels, and strong expression. The key is shown in the image and bars at the base of each image represent the percentage of animals in each phenotypic class. (C) Relative fold change of NvashA and previously identified Nvasha neural gene targets and in animals overexpressing Nvdelta (light grey bars) and animals that are overexpressing Nvdelta and treated with DAPT (dark grey bars). Red rectangle denotes relative fold change −1.5 to 1.5, which corresponds to no change in relative expression.

Figure 4

Nvnotch suppresses neurogenesis by regulating NvashA expression. (A) Relative expression levels of NvashA target genes in animals with overexpressing (OE) Nvnicd (dark blue bars), overexpressing Nvnicd and NvashA (light blue bars), animals treated with DAPT (dark orange bars), and DAPT treated animals injected with the NvashA morpholino (MO) (light orange bars). Red rectangle represents relative fold change –1.5 to 1.5, which corresponds to no change in relative expression. Each treatment was repeated at least three times. (B-G) Aboral views of mRNA in situ images from two NvashA neural target genes are shown. Animals with overactive Nvnotch(B,E), control (C,F), and both overactive Nvnotch and overactive NvashA(D,G) are shown. Animals in (B-G) were quantified into phenotypic classes based on having no, weak, wild-type (WT)-like, or strong expression levels. The key is shown in the image and bars at the base of each image represent the percentage of animals in each phenotypic class.

Figure 5

DAPT treatment increases NvashA expression in the planula larva. (A-C) Forty-eight hours post fertilization (hpf) animals either treated with control DMSO (A) or with DAPT (B-C). (A) NvashA expression in control animals is detected in the developing pharynx (arrow), in the endoderm (arrowhead), and in the ectoderm (inset). (B) Treatment with DAPT increases NvashA expression in each tissue. (C) Quantitative (q)PCR analysis reveals a three-fold increase in the relative levels of NvashA in DAPT-treated animals. (D-F) Seventy-two hpf animals either treated with control DMSO (D) or with DAPT (E-F). (D) NvashA expression in control animals is detected in the developing pharynx (arrow) and in the endoderm. (E) Treatment with DAPT increases NvashA expression in each tissue. (F) qPCR analysis reveals a three-fold increase in the relative levels of NvashA in DAPT-treated animals. The key in (C) and (F) shows that animals were grown in normal 1/3X sea water (black line between time intervals) or in the presence of DAPT (red line between time intervals). Animals in (A,B,D,E) were quantified into phenotypic classes based on having no, weak, wild-type-like, or strong expression levels. The key is shown in the image and bars at the base of each image represent the percentage of animals in each phenotypic class. Red box in (C and F) indicates the region between 0 and 1.5-fold change, which we consider to indicate no change in expression. All animals are shown in a lateral view with the oral side to the left.

Figure 6

Nvnotch regulates “salt and pepper” differentiation genes. Relative fold change of “salt and pepper” genes in animals following treatment with DAPT (blue bars), injection with Nvnicd (dark orange bars), injection with Nvnicd and NvashA (light orange bars), or NvashA alone (green bars). Red rectangle denotes relative fold change −1.5 to 1.5, which indicates no change in relative expression. “salt and pepper” differentiation genes are suppressed by Nvnotch activity while genes with broad expression domains are unaffected by any of our treatments. OE, overexpressing.

Figure 7

Nvnotch does not regulate Nvhes expression in the Nematostella embryo. (A) Average relative fold change of Nvhes gene expression in animals injected with Nvnotch morpholino (MO; orange bars), treated with DAPT (blue bars), injected with Nvnicd:venus (dark purple bars), injected with Nvdelta:venus (light purple bars), or a control MO (grey bars. Red rectangle covers the region where the relative fold change ratio is equal to −1.5 to 1.5 and corresponds to no change in relative expression level. (B-E) Lateral views of late stage gastrula expressing Nvhes2(B-C) or Nvhes3(D-E). Oral is to the left. Deep focal plane is shown and superficial focal plane is shown in inset. We observed no discernable difference in Nvhes2 or Nvhes3 expression by in situ analysis between wild-type and Nvnotch MO injected animals. We scored N >80 embryos for each treatment. OE, overexpressing.

Figure 8

Nvhes2 and Nvhes3 overexpression does not repress NvashA expression. (A-F) Lateral views of embryos expressing NvashA; oral is to the left. There is no discernable difference in NvashA expression in control (A-B), Nvhes2 overexpressing (OE) (C-D), or Nvhes3 (E-F). N >65 scored for each experiment. (G) Relative fold change of NvashA in embryos treated overexpressing Nvhes2 or Nvhes3. (H) Average relative fold change of NvashA, neural genes, and Nvhes genes in animals injected with the NvsuH morpholino (MO; dark grey bars) or a dominant negative NvsuH (DN; light grey bars). Red rectangle denotes relative fold change −1.5 to 1.5, which indicates no change in relative expression. Each injection was repeated three times.