A Molecular, Spatial, and Regulatory Atlas of the Hydra vulgaris Nervous System

Abstract

Hydra vulgaris, a cnidarian with a simple nerve net, is an emerging model for developmental, regenerative, and functional neuroscience. Its genetic tractability and capacity for whole-system imaging make it well suited for studying neuron replacement, regeneration, and neural circuit function. Here, we present the most comprehensive molecular and spatial characterization of the H. vulgaris nervous system to date. Using single-cell RNA sequencing, we identified eight neuron types, each defined by distinct neuropeptide expression, and further resolved these into fifteen transcriptionally distinct subtypes with unique spatial distributions and morphologies. To investigate the gene regulatory networks underlying neuronal differentiation, we applied trajectory inference, identified key transcription factors, and performed ATAC-seq on sorted neurons to map chromatin accessibility. All datasets are available through an interactive, user-friendly web portal to support broad use by the research community. Together, these resources provide a foundation for uncovering molecular mechanisms that govern nervous system development, homeostasis, and regeneration in H. vulgaris.

Figure 1.. Single cell RNA sequencing reveals fifteen transcriptionally distinct neuron subtypes comprising the Hydra vulgaris nervous system.

(A) The Hydra body is a radially symmetric, hollow tubular animal arranged around an oral-aboral axis. The oral end consists of the hypostome and tentacles (collectively referred to as the “head”), while the aboral end comprises the peduncle and basal disk. The body is formed by two epithelial monolayers, the endoderm and ectoderm, separated by an extracellular matrix. Neurons reside within the interstitial spaces between epithelial cells, forming two distinct nerve nets: one in the ectodermal layer and another in the endodermal layer (Keramidioti et al., 2024). Interstitial stem cells (ISCs), which give rise to neurons, are confined to the ectoderm, but intermediate neural progenitors migrate to generate endodermal neuron subtypes. (B) Single-cell RNA sequencing (scRNA-seq) was performed using two methods: Chromium Single-Cell Gene Expression (10x Genomics) (29,671 cells from this study) and Drop-Seq (5,400 cells from Siebert et al., 2019). A UMAP representation displays cell clusters annotated by their respective cell states: ISC (interstitial stem cell), progen (progenitor) prec (precursors), ec (ectodermal neurons), and en (endodermal neurons). (C) Dot plot showing the expression of neuropeptide genes across Hydra vulgaris neuron clusters. Dot color represents the average expression level of each gene within a cluster, while dot size indicates the percentage of cells in the cluster expressing the gene. The ec1, ec2, ec3, ec4, and ec5 neuron types have distinct neuropeptide expression patterns. (D) Schematic outlining the classification process for H. vulgaris neurons. First, neurons were identified in the scRNA-seq dataset by the expression of elav. Cells were further categorized into either the ectodermal or endodermal nerve net based on the expression of the endodermal marker budhead. Ectodermal neuron types were defined by distinct neuropeptide expression profiles, and ectodermal neuron subtypes were classified based on spatially restricted domains (see main text for detailed information on the classification scheme).

Figure 2.. Spatial organization of ectodermal neuron subtypes in revealed by fluorescence in situ hybridization (FISH).

Confocal images show FISH labeling of twelve distinct ectodermal neuron subtypes using subtype-specific molecular markers. (A–G) Oral neuron subtypes: ec1C (G007791), ec1E (G003886), ec4A (RFamide C, G017226), ec4B (RFamide B, G017227), ec1B (G016169), ec2 (G010335), and ec3C (G026484). Arrow in panel G indicates expression in the tentacles. (H–I) Body column neuron subtypes: ec1A (G026993) and ec3B (G004106). (J–L) Aboral neuron subtypes: ec5 (Hym176C, G016165), localized to the peduncle (arrow), and ec3A (G021930) and ec1D (G004200), localized to the basal disk (arrows). See Fig. S7 for whole animal images of ec4A, ec4B, and ec1C FISH experiments. Nuclei are counterstained with Hoechst (blue). All panels show standard FISH labeling except panel B, in which ec1E neurons required localization using hybridization chain reaction (HCR).

Figure 3.. Neuropeptide expression delineates distinct neural circuits in Hydra vulgaris.

(A, B) Schematics illustrate the spatial distribution of neuron subtypes along the Hydra body column. (C) The RP1 circuit comprises ec3A–C neurons, which express neuropeptides from the GLWa family. The RP2 circuit includes endodermal en1 neurons, which also express GLWa. (D) Immunostaining using an antibody against GLWa (green) reveals the RP1 circuit extending from the basal disk to the tentacles. (E, F) GLWa immunostaining in body column rings shows endodermal en1 neurons of the RP2 circuit. (G) The CB circuit includes ectodermal neuron subtypes ec1A–D and ec5, which express Hym176 neuropeptides. (H) GFP immunostaining (green) in the Hym176B reporter line marks CB circuit neurons along the body column. (I) The RFa-A neuropeptide is expressed in ec4 and ec5 neurons, the latter being part of the CB circuit. (J) Immunostaining with an antibody against RFa (green) reveals ec5 neurons in the peduncle and ec4 neurons near the oral end. In all immunostaining panels (D, F, H, J), co-labeling with PNab (magenta) highlights the full structure of the nervous system. Scale bars: 200 μm. The lower images in panel D and the upper images in panel H are reproduced from Keramidioti et al., 2024.

Figure 4.. Localization of neuron subtypes in the basal disk and hypostome.

(A–C) Double FISH using probes for G021930 (magenta, ec3A) and G004200 (green, ec1D) confirms that these neuron subtypes are distinct populations in the basal disk. Nuclei are counterstained with Hoechst (blue) (D–F) Immunofluorescence with antibodies against GLWa (green, ec3) and PNab (magenta, pan-neuronal) reveals GLWa-negative/PNab-positive neurons in the basal disk, consistent with ec1D identity. (G–I) Co-immunostaining in the Hym17B-GFP reporter line using anti-GFP (green) and anti-GLWa (magenta) shows the spatial relationship of ec1 and ec3 neurons in the hypostome. (J–L) Co-immunostaining in the Hym17B-GFP reporter line with anti-GFP (green) and anti-RFa (magenta) illustrates the positioning of ec1 and ec4B neurons in the hypostome.

Figure 5.. Trajectory reconstruction of neurogenesis reveals distinct developmental pathways for ectodermal and endodermal neurons.

(A) Differentiation trajectories for eleven Hydra neuron subtypes were reconstructed using URD from single-cell transcriptomics data (Farrell et al., 2018). Interstitial stem cells (ISCs) were set as the “root” or starting point, and each neuron subtype served as a “tip” or endpoint. The tree is colored by pseudotime (developmental progression), with earlier pseudotime at the top (blue) and later pseudotime at the bottom (red). The boxed region highlights areas detailed in panels C, H, and M. Asterisks indicate segments corresponding to early differentiation events from ISCs into ectodermal or endodermal neuronal progenitors. (B) Spline plot showing expression trends for five genes during ec3A differentiation. The x-axis represents pseudotime (earlier on the left, later on the right), and the y-axis represents scaled gene expression levels. Each point is the average expression for five cells. (C-U) Double FISH validates predicted transition states during ec3A differentiation. Expression states in FISH images are indicated by arrow types: closed arrows for cells expressing only the first gene, open arrows for cells expressing only the second, and double arrows for cells co-expressing both genes. (C-G) Validation of early ec3A transition states co-expressing bhlha15 and gata3. (C) Visualization of bhlha15 and gata3 expression on the URD trajectory. Co-expressing cells (orange) are indicated with arrowheads; black represents cells expressing neither gene. (D-G) Confocal microscopy of double FISH for early transitions: (D) Body region imaged, (E) bhlha15 expression (red), (F) gata3 expression (yellow), and (G) Overlay with Hoechst-stained nuclei (gray). (H-L) Validation of mid ec3A transition states co-expressing hym355 (magenta) and gata3 (yellow). (H) Visualization of hym355 and gata3 expression on the URD trajectory. Co-expressing cells (orange) are highlighted with arrowheads. (I-L) Confocal microscopy of double FISH for mid transitions: (I) Body region imaged, (J) hym355 expression (red), (K) gata3 expression (yellow), and (L) Overlay with Hoechst-stained nuclei (gray). (M-U) Validation of late ec3A transition states co-expressing hym355 (magenta) and ec3A marker G021930 (yellow). (M) Visualization of hym355 and G021930 expression on the URD trajectory. Co-expressing cells (orange) are marked with arrowheads. (N-U) Confocal microscopy of double FISH for late transitions: (N, R) Body region imaged, (O, S) hym355 expression (red), (P, T) G021930 expression (yellow), and (Q, U) Overlay with Hoechst-stained nuclei (gray). Scale bars: 50 μm. Pink dotted lines in microscopy images (G, L, Q, U) indicate the boundary between the body column and peduncle, determined by nuclei morphology.

Figure 6.. Characterization of the neuronal chromatin landscape using ATAC-seq.

(A-B) Neuron-enriched ATAC-seq libraries were generated from six samples: two from the Tg(tba1c:mNeonGreen)^cj1-gt transgenic line (green) and four from the Tg(actin1:GFP)^rs3-in transgenic line (pink) (Keramidioti et al., 2024). To assess neuronal peak enrichment, three previously published whole-animal ATAC-seq libraries (AEP1–3; Cazet et al., 2023) were used as a reference. Panels (A) and (B) show the average accessibility changes in gene-proximal peaks for Tg(actin1:GFP)^rs3-in and Tg(tba1c:mNeonGreen)^cj1-gt, respectively, compared to whole-animal data sets. Neuronal genes exhibit significant peak enrichment within 10,000 base pairs of the transcription start site relative to non-neuronal genes (p < 0.001, t-test). (C) ATAC-seq tracks for the alpha tubulin locus (tba1c, G019559) show peaks in the neuron-enriched libraries that are absent in the whole animal libraries. (D) UMAP visualization of the neuron-enriched scRNA-seq dataset shows that tba1c is ubiquitously expressed in neurons. (E) UMAP visualization of the whole-animal scRNA-seq dataset shows that tba1c expression is restricted to neurons (Cazet et al., 2023; Siebert et al., 2019). (F) ATAC-seq tracks for the wnt3 locus (G010730) show peaks in the whole-animal libraries that are absent in the neuron-enriched libraries. (G) UMAP visualization of the neuron-enriched scRNA-seq dataset shows no expression of wnt3 in neurons. (H) UMAP visualization of the whole animal scRNA-seq dataset shows that wnt3 is expressed in the oral ectodermal and endodermal epithelium (Cazet et al., 2023; Siebert et al., 2019).

Figure 7.. A putative gene regulatory network (GRN) underlying ec4 neuron specification.

(A) Spline plots showing transcription factor expression dynamics during ec2 and ec4 differentiation. (B) Sox2 (G009896) is broadly expressed in oral neuron subtypes ec3C, ec1B, ec2, and ec4. (C) Zic4 (G004456) is expressed in ec1B, ec2, and ec4. (D) Arx (G021458) is restricted to ec2 and ec4. (E) Co-expression trajectory tree of Gata3 (G022640, magenta) and Zic4 (G004456, green) show mutually exclusive expression. (F) Co-expression trajectory tree of Litaf (G007969, magenta) and Gsx1/2 (G023503, green) reveals subtype-specific expression in ec2 and ec4 neurons. (G) Co-expression trajectory tree of Gsx1/2 (magenta) and Noto (G003891, green) shows mutually exclusive expression in ec4. (H) ATAC-seq tracks at the Sox2 locus reveal open chromatin regions containing Tcf, Sox2, and Myb binding motifs. (I) ATAC-seq tracks at the Noto locus highlight peaks containing Tcf, Sox2, Myc, and Gsx2 binding motifs. (J) A proposed GRN model summarizing transcriptional regulators and regulatory relationships involved in ec4 neuron specification in Hydra vulgaris. Plots in this figure are made using the scRNA-seq portal (https://research.nhgri.nih.gov/HydraAEP/SingleCellBrowser/).