Fig 1.
Identification and characterization of neuropeptide precursors and final products in Hydra vulgaris.
Amino acid sequences of 16 precursors identified encoding 64 final peptide-candidates via dibasic cleaving sites. Red shows an identified signal sequence detected by signalP(25), green a predicted final peptide, yellow identified cleavage sites. Underlined sequences are peptides verified experimentally using MS [27]. Identified versions are sometimes shorter or longer than found ones. Full list of transcriptome and genome position in S1 Table. Neuropeptide family assignments are indicated in cyan.
Fig 2.
Identification and interspecies comparison of Hydra´s GPCRs.
Cluster analysis of major Class A GPCRs from cnidarian, bilaterian, and placozoan species, illustrating sequence similarity relationships. Each node represents a GPCR sequence, with edges indicating similarity-based connections. Different species groups GPCRs are color-coded. Orange highlights GPCRs sequences that have been deorphanized and shown experimentally to bind neuropeptides in Nematostella vectensis. These were used as a guidance to determine GPCR from Hydra vulgaris that putatively can bind neuropeptides (highlighted in green). Edges are colored by intensity. Fruchterman Reingold algorithm used for clustering with repulsion set to 40 and attraction between nodes set to 10.
Fig 3.
Predicted interactions between identified peptides and GPCRs in Hydra.
Binding prediction results for identified peptides and GPCRs, with top 10% highest scores highlighted by black circles. Colors represent precursors from which the respective peptides are synthesized. Prediction scores integrate results from AlphaFold and DeepTMHMM to assess expected binding potential of the receptor-neuropeptide combination within the predicted membrane location (see Methods). Note that several different receptors can bind each neuropeptide. Further, several of receptors are found to bind more than one neuropeptide and have high scores even across families, reinforcing the idea that the neuropeptide network is flexible and relies on highly distributed set of interactions.
Fig 4.
Specific expression of predicted neuropeptide and GPCR genes in Hydra´s neuronal subtypes.
Transcriptomic profiles of predicted neuropeptide and neuropeptide-binding GPCR genes across neuronal subtypes in Hydra. X axis categorizes different transcriptomic neuron types found in Hydra and on Y axis detected Neuropeptide (left) or GPCR (right) transcripts. Note that there are only 16 neuropeptide transcripts - these encode the final 65 detected peptides. The color shows the average expression across cells in those categories. The unique expression profiles for different neuronal subtypes suggest highly specific communication pathways within the organism. The cladograms show similarity between cell types based on receptor and or prepropeptide expression.
Fig 5.
Dense neuropeptide-GPCR network.
Network derived from predicted neuropeptide-GPCR interaction, showing hub nodes within and between cell types. (A) Cellular matrix, where rows denote sending neurons, while columns denote receiving neurons. Communication strength is color-coded, with red indicating high intensity and blue indicating low intensity. Strength is calculated based on number and expression levels of GPCR-neuropeptide combinations. Note widespread high density of interactions. (B) Spatial distribution of 11 neuronal subtypes along the Hydra´s body (adapted from [21]. (C) Neuronal cell type-level network showing strongly connected ectodermal communication hub, with clustering based on connection similarity. (D) Topology of neuronal interactions via neuropeptides/GPCRs, illustrating a densely connected network with multiple communication pathways between cell types. Node placement is determined by directed force algorithms based on connections. Endodermal and ectodermal neurons are colored blue and pink, respectively. (E) Simplified version of D with strongest 30% of connections, highlighting ectodermal hub nodes.
Fig 6.
Connectivity between endoderm and ectoderm nerve nets.
A. Heatmaps showing directed connectivity strengths between endoderm and ectoderm nodes. Left: Connectivity from endoderm to ectoderm nodes. Right: Connectivity from ectoderm to endoderm nodes. Color scales represent connection strength, with higher values indicating stronger connectivity. Each row represents a sending node group, and each column represents a receiving node group. Color bars indicate the range of connectivity strengths. B. Rich neuropeptide connectivity between endoderm and ectoderm. Left: Connections from Endoderm to Ectoderm nodes. Right: Connections from Ectoderm to Endoderm nodes. Edges are colored according to connection strength (see color bars), with node labels indicating node groups (e.g., en1, ec5). Node sizes are proportional to their overall connectivity strength. A variety of connections between endoderm and ectoderm are shown, with different cell types exhibiting specific subfunctions highlighted by the reduced figure. Note that en1 is a receiving hub for ectodermal signals, while en2 sends strong signals towards the ectoderm.
Fig 7.
Hydra´s neuropeptide network can implement attractor states.
A. Activity of one of the network nodes (en) across different initial starting points demonstrating a dynamical split into two different final states depending on the initial input towards the network. B. Representative trajectories of the recurrent dynamical system starting from different initial conditions, projected into the first three principal components of the network’s state space. Each trajectory is shown in a different color. Despite diverse starting points, trajectories converge toward two distinct terminal regions, reflecting the multistable nature of the system. Axes correspond to the reduced state-space dimensions (x₁ = PC1, x₂ = PC2, x₃ = PC3). C. Shows a pseudo-energy measure for the final steady states reached from a range of initial conditions. The resulting landscape is visualized in the reduced state space defined by the first two principal components (PC1 and PC2). Brighter regions indicate lower relative pseudo-energy (more stable attractor basins). The landscape reveals two distinct basins of attraction, consistent with the presence of multiple stable equilibrium states supported by the dense and recurrent architecture of the neuropeptide signaling network. Axes correspond to the reduced state-space dimensions (x₁ = PC1, x₂ = PC2).