Larval single-cell maps point to a shared ancestral sensory toolkit
Single-cell transcriptomics in marine larvae is now pushing evolutionary biology toward a finer question: not when nervous systems appeared, but how the molecular parts were assembled before true neurons emerged. In a comparative analysis of larval cell states from the demosponge Amphimedon queenslandica and four cnidarian species—Nematostella vectensis, Astroides calycularis, Pocillopora acuta and Clytia hemisphaerica—investigators used cross-species mapping to compare the transcriptional architecture of sensory, secretory and neural-associated programs across early-diverging animals.
The biological question is sharpened by the larval stage itself. Sponge and cnidarian larvae are not passive developmental intermediates; they must detect environmental cues, interpret them, and coordinate swimming, dispersal and settlement. Those behaviors depend on specialized sensing and signaling systems, which makes larvae a particularly revealing window into the evolutionary history of cell identity. In this study, larval cell populations were compared across species with Self-Assembling Manifold mapping, a cross-species alignment approach that can place cells from different organisms into shared transcriptional space when underlying gene programs overlap.
Cnidarian larval cell identities remain broadly conserved across species
Across the cnidarian datasets, the mapping recovered broad conservation of major cell-type identities. Neural, secretory, cnidocyte, epithelial, gastrodermal and apical populations clustered with recognizable correspondence between species, suggesting that the larval cellular framework of cnidarians is built on a deeply conserved set of transcriptional programs. This kind of conservation matters because it indicates that distinct species can still retain comparable cell-state logic even after substantial evolutionary divergence.
The patterns also fit what developmental biologists expect from larvae that must perform similar tasks in different ecological settings. Sensory detection, environmental integration and coordinated behavior are not optional features in free-swimming larvae; they are survival functions. The transcriptomic data suggest that these functions are supported by cell identities that remain legible across cnidarian lineages, despite differences in body plan and life history.
Sponge larval cells share transcriptional features with neural and progenitor-like states
The comparison becomes more provocative when sponge larvae are added to the analysis. Extending the same cross-species mapping to A. queenslandica revealed transcriptional similarities between several sponge larval cell populations and cnidarian neural, sensory and progenitor-associated cell types. Within the sponge larva, these populations were not isolated as a single proto-neural compartment. They remained integrated within broader epithelial and sensory cell identities, implying that the relevant molecular toolkit was distributed rather than centralized.
That distribution is central to the evolutionary argument. Instead of a lone ancestral cell type already functioning as a neuron in waiting, the data support a model in which the ingredients for neural function were spread across multiple ancestral cell states. Such a configuration would allow later lineages to reassemble, specialize and concentrate those components into dedicated neural and sensory-secretory cells. In other words, nervous systems may have emerged less by invention than by reorganization.
Neural-associated gene modules were split across multiple sponge cell populations
The transcriptomic signal did not point to a single sponge cell type carrying all the hallmarks of a neural precursor. Neural-associated gene modules were distributed across distinct epithelial and sensory-like sponge larval cell populations. Those modules included conserved families linked to photoreception, neural fate specification and postsynaptic scaffolding. In a vertebrate or cnidarian nervous system, such genes would usually be interpreted in the context of neural signaling, synaptic organization or sensory transduction. In the sponge larva, they appear in a more dispersed and developmentally flexible arrangement.
That split architecture is biologically meaningful. Photoreceptive components can support light-sensitive behavior even in the absence of neurons. Fate-specification genes can prepare cells for developmental divergence without committing them to a mature neural lineage. Postsynaptic scaffolding proteins, meanwhile, can be repurposed as structural and signaling elements before canonical synapses exist. The overall picture is one of modularity: molecular parts that later became hallmarks of nervous systems were already active in non-neural cells, where they likely served sensory or signaling roles in a simpler behavioral network.
A pre-neural common ancestor may have carried the raw materials for neural evolution
The strongest inference from this work is evolutionary rather than anatomical. The findings suggest that molecular components later incorporated into neural systems were already deployed across diverse cell types in a common ancestor of sponges and cnidarians. If so, the origin of neurons did not require assembling an entirely new molecular repertoire from scratch. It required selecting, concentrating and reorganizing older sensory-secretory programs into more specialized cellular architectures.
That model also helps explain why comparisons between early-diverging animal lineages can be so informative. Sponges lack neurons, yet their larvae still express gene modules that in other animals are tightly associated with nervous systems. Cnidarians possess neurons, but those neurons may represent an elaboration of a more ancient toolkit that was first used in mixed sensory and secretory contexts. The evolutionary path therefore looks incremental, with repeated reuse of conserved components rather than a single abrupt origin event.
The study’s framing is especially useful because it bridges molecular data and cell-type evolution without forcing present-day cell identities into overly rigid categories. What looks like a neuron-specific gene in one lineage may have had broader ancestral functions in another. What appears to be a sensory-like sponge cell may preserve a transitional state between generalized epithelial physiology and specialized neural circuitry. Those possibilities are exactly where comparative single-cell biology earns its value.
Cross-species transcriptomics narrows the gap between sponges and neurons
By integrating larval datasets from five species, the analysis supports an evolutionary scenario in which extant neural and sensory-secretory cell types arose through diversification and reorganization of ancestral molecular programs distributed across multiple cell states. The work does not claim that sponge larvae already possess neurons. It does something more precise: it shows that the molecular groundwork for nervous system evolution was already present in non-neural contexts and was not confined to a single proto-neural cell type.
A practical limitation follows from the evidence itself. Transcriptomic similarity cannot by itself establish ancestral cell function, nor can it prove that a shared gene module performed the same task in every lineage. Functional validation, developmental perturbation and broader taxon sampling will be needed before the evolutionary sequence can be tightened further. Even so, the data provide a strong comparative framework for thinking about how sensory and neural programs were assembled from older cellular parts, and they place larvae—rather than adult body plans—at the center of that history.