Differential gene regulation in DAPT-treated Hydra reveals candidate direct Notch signalling targets

Abstract

In Hydra, Notch inhibition causes defects in head patterning and prevents differentiation of proliferating nematocyte progenitor cells into mature nematocytes. To understand the molecular mechanisms by which the Notch pathway regulates these processes, we performed RNA-seq and identified genes that are differentially regulated in response to 48 h of treating the animals with the Notch inhibitor DAPT. To identify candidate direct regulators of Notch signalling, we profiled gene expression changes that occur during subsequent restoration of Notch activity and performed promoter analyses to identify RBPJ transcription factor-binding sites in the regulatory regions of Notch-responsive genes. Interrogating the available single-cell sequencing data set revealed the gene expression patterns of Notch-regulated Hydra genes. Through these analyses, a comprehensive picture of the molecular pathways regulated by Notch signalling in head patterning and in interstitial cell differentiation in Hydra emerged. As prime candidates for direct Notch target genes, in addition to Hydra (Hy)Hes, we suggest Sp5 and HyAlx. They rapidly recovered their expression levels after DAPT removal and possess Notch-responsive RBPJ transcription factor-binding sites in their regulatory regions.

Keywords: Axis formation; Hydra; Nematocyte differentiation; Notch pathway; Wnt pathway.

Fig. 1.

Overview of the experimental and analysis workflow.Hydra polyps were treated with either DAPT or DMSO (control) for 48 h. Thereafter, total RNA for sequencing was collected at three time points. The sample 0 h was taken immediately after 48 h of DAPT treatment. This is also the time point at which DAPT was removed from the samples and total RNA was collected 3 and 6 h after DAPT removal. Six biological replicates for each treatment were collected and processed at the same time point. Pairwise differential gene expression analysis by DESeq2 was performed between DAPT- and DMSO-treated samples for each of the three collection time points. This analysis revealed 831 Notch-responsive genes (NR genes) after 48 h of DAPT treatment (0 h). For these genes we characterized the expression at time points 3 h and 6 h. For 666 NR genes single-cell expression data from homeostatic polyps were available (Siebert et al., 2019) and was used to elucidate expression pattern and cell-state-specific expression using hierarchical cluster and non-negative matrix factorization (NMF) analysis. Additionally, motif enrichment was performed for the set of NR genes.

Fig. 2.

Differential expression of NR genes post DAPT treatment. (A) Expression dynamics of differentially expressed NR genes and time points for the recovery of original expression levels. For both up- and down-regulated genes, ∼35% recover original expression within the first 3 h (‘Recovered at 3h’, grey), ∼25% recover within 6 h (‘Recovered at 6h’, green), and ∼30–40% do not recover original expression within the time course of the experiment (‘Not recovered’, yellow). The remaining genes (∼5%, ‘Other’, purple) behave irregularly, e.g. recovered after 3 h, deregulated again after 6 h. (B) Heatmap highlighting expression differences of all 831 NR genes. The colour key refers to the log₂ fold change values. The cyan line in the small diagram indicates the distribution of z-scores. Clustering of NR genes by their log2Foldchange for each time point revealed upregulated (blue) and downregulated (red) genes. No value (white background) means the gene was not differentially expressed at that particular time point and thus had control expression levels. We identified sets of genes that recover their expression by 3 h (e.g. HyHes, differentially expressed at 0 h, thereafter back to control expression level), genes that recover expression by 6 h, and genes that do not recover expression during the course of the experiment, i.e. at 6 h after DAPT removal (e.g. the post-mitotic nematocyte gene markers CnASH, NOWA and Spinalin). A fourth set includes genes that are differentially expressed at 0 h and 6 h, but not at the 3 h time point (e.g. CnGSC).

Fig. 3.

NR gene expression in homeostatic polyps based on single-cell expression data. Expression data and cell state annotations were retrieved from Siebert et al. (2019). Hierarchical clustering was performed for 666 NR genes using average expression values for each annotated cell state. The colour key refers to cell state expression values. The green line in the small diagram indicates the distribution of z-scores. This revealed expression in nematoblast/nematocyte-specific genes (violet, red, blue and yellow cluster), ectodermal epithelial cell genes including battery cell genes (black), endodermal epithelial cell genes including tentacle genes (grey), genes ubiquitously expressed across a wide range of cell states (cyan) and genes with a sporadic expression (green). Nematoblast/nematocyte genes constituted 47% of the NR genes. i, cell of the interstitial lineage; nb, nematoblast; ecEp, ectodermal epithelial cell; enEP, endodermal epithelial cell; en, endoderm; ec, ectoderm.

Fig. 4.

Hierarchical clustering of NR genes expressed in the nematocyte lineage. (A) NR genes expressed in cells of the nematocyte lineage were clustered separately to reveal their expression in the differentiation states of nematogenesis. This revealed a set of genes only expressed in mature nematocytes (cyan cluster), genes mainly expressed in cell state nb5 (blue), genes mainly expressed in nb6 (red), genes mainly expressed in nb8 (green) and genes expressed ubiquitously in stages nb4 through nb8 (black). Almost all of these genes were downregulated upon DAPT treatment. (B) The majority of both mature nematocyte genes and nematoblast genes did not recover their expression 6 h after DAPT removal (yellow). This includes POU4, Dickkopf3, NOWA and Spinalin. Furthermore, genes are represented that recovered after 3 h (grey), 6 h (green) or that had an irregular recovery profile (magenta).

Fig. 5.

Homeostatic expression of nematoblast marker genes and proteins. (A) t-Distributed stochastic neighbour embedding (t-SNE) representation showing the interstitial cell state expression of HyZic, PCNA, CnASH, NOWA and spinalin. Cluster labels are provided for cell states of the nematoblast lineage according to Siebert et al. (2019). nb, nematoblast; nem, nematocyte; ISC, interstitial stem cell. Blue dots indicate cells expressing the respective genes. PCNA is expressed in proliferating cells, nematoblast cell states nb1 and nb2. HyZIC is mainly expressed in nb2. This is in accordance with previously published work indicating HyZIC expression in proliferating nematoblasts. CnASH is expressed in nb5, 6, 7, 8, representing post-mitotic nematoblasts lacking PCNA expression. This is in complete agreement with previous work (Lindgens et al., 2004). NOWA encoding an outer capsule protein, is expressed in nb5 and nb7, spinalin, encoding a protein occurring inside the capsule, is expressed in nb4, 5, 6, 7 and 8 and in mature nematocytes (nem), all representing post-mitotic nematoblast stages. (B) Laser confocal microscopic sections of co-immunofluorescence staining with anti-HyZIC, anti-CnASH and anti-NOWA antibodies, in merged images DNA stain DAPI (blue), CnASH (green), HyZIC (red), NOWA (red). Anti-NOWA antibody delineates capsules (upper panel, middle image and red in merged). Co-staining with anti-CnASH antibody indicates signal in cytoplasm of capsule containing cells (upper panel, left hand image and green in merged). Capsule containing CnASH-positive cells (lower panel, left hand side and merged image green) are not stained with anti-HyZIC antibody (lower panel, middle image and merged image red); C, capsule; N, nucleus. Scale bars: 20 µm. (C) Schematic summary of gene expression in the nematoblast lineage indicating a differentiation pathway from interstitial stem cell precursors (ISC/nb) via proliferating PCNA and HyZIC expressing amplifying nematoblast precursors (nb1, nb2) via post-mitotic nematoblasts not expressing PCNA (nb3, nb4) to capsule forming CnASH expressing nematoblasts (nb5, 6, 7, 8), t-SNE representation of cells with clusters labeled by cell state as presented in Siebert et al. (2019) with permission. Images are representative of three experiments.

Fig. 6.

NR gene subset with expression in epithelial cells. Non-nematoblast NR genes were clustered separately to determine their expression in epithelial cell states. This revealed sets of genes that are most strongly expressed in endodermal tentacle cells (red cluster), ectodermal basal disc cells (yellow), ectodermal battery cells (black), body column ectoderm cells (cyan), body column endoderm cells (green) and all epithelial cells (grey). The analysis also revealed smaller gene sets expressed in endodermal foot cells, endodermal head cells or ectodermal head cells. Tentacle, battery and head-specific genes were mainly downregulated upon DAPT treatment whereas the genes in the remaining clusters were mainly upregulated. The colour key refers to cell state expression values. The green line in the small diagram indicates the distribution of z-scores.

Fig. 7.

Recovery time of NR genes. (A) The majority of epithelial NR genes were upregulated upon DAPT treatment. 50% of the upregulated ectodermal-specific NR genes recover expression within the first 3 h post-DAPT removal (light green) and include the apoptosis-involved gene DAD1. SP5 on the other hand, which is expressed in both epithelia, is downregulated and recovers expression also within the first 3 h. (B) Head-specific fNR genes, including tentacle, battery and ectodermal and endodermal head genes, are mostly downregulated upon DAPT treatment. In contrast, foot-specific genes, including endodermal foot genes and basal disc genes are mostly upregulated. Many of these genes play a predominant role in patterning.

Fig. 8.

Motif enrichment analysis of NR gene promoter regions. (A) Workflow of motif enrichment analysis. Putative promoter regions were identified using a previously published ATAC-seq dataset generated using whole wild-type Hydra (Siebert et al., 2019). NR gene promoter regions were defined as ATAC-seq peaks that fell within 5 kb upstream of the transcription start site of an NR gene. Using HOMER, NR gene promoters were compared against control peaks that were not associated with NR genes to identify significantly enriched (FDR≤0.05) transcription factor-binding motifs. (B) Notch/RBPJ-binding motifs were significantly enriched in the putative promoters of genes that were downregulated upon DAPT treatment and recovered rapidly following inhibitor removal. (C) Pou and Pax transcription factor binding motifs were significantly enriched in the putative promoters of genes that were downregulated upon DAPT treatment and did not recover their expression over the course of the RNA-seq experiment. Plots of normalized ATAC-seq read density in the 5 kb upstream of (D) HyAlx and (E) SP5 demonstrate the presence of predicted RBPJ-binding sites in the putative promoters of NR genes. Red bars indicate predicted instances of Notch-binding motifs.