TrendsAccumulating genomic data stronglysupport the position of Cnidaria asthe sister clade to Bilateria. The emer-gence of a simple nerve net togetherwith biological, structural and func-tional diversity within this taxonomicgroup make cnidarians highly informa-tive for comparative approaches.

Recently sequenced genomes andtranscriptomes provide insights intothe molecular complexity of cnidariannerve nets. The diversity of synapticproteins, small neurotransmitters, neu-ropeptides, and their processingmachinery and receptors, is compar-able with that of chordates.

Recent advances in imaging and genemanipulation techniques make cnidar-ians now amenable to functional ana-lysis addressing molecular, behavioraland evolutionary questions.

Accumulating evidences point to mul-tiple roles of the simple nervous sys-tems. Emerging evidence pointsto functions of nervous systemsbeyond simple sensory and motorcoordination.

1University of Kiel, Kiel, Germany2Ruper Boskovi�c Institute, Zagreb,Croatia3Catholic University of Croatia,Zagreb, Croatia4Institute of Physiology, RWTHAachen University, Germany5Centre for Organismal Studies,Heidelberg, Germany

ReviewBack to the Basics:Cnidarians Start to FireThomas C.G. Bosch,1,* Alexander Klimovich,1

Tomislav Domazet-Loso,2,3 Stefan Gründer,4

Thomas W. Holstein,5 Gáspár Jékely,6 David J. Miller,7

Andrea P. Murillo-Rincon,1 Fabian Rentzsch,8

Gemma S. Richards,8,9 Katja Schröder,1 Ulrich Technau,10 andRafael Yuste11,*

The nervous systems of cnidarians, pre-bilaterian animals that diverged close tothe base of the metazoan radiation, are structurally simple and thus have greatpotential to reveal fundamental principles of neural circuits. Unfortunately,cnidarians have thus far been relatively intractable to electrophysiologicaland genetic techniques and consequently have been largely passed over byneurobiologists. However, recent advances in molecular and imaging methodsare fueling a renaissance of interest in and research into cnidarians nervoussystems. Here, we review current knowledge on the nervous systems of cnidar-ian species and propose that researchers should seize this opportunity andundertake the study of members of this phylum as strategic experimentalsystems with great basic and translational relevance for neuroscience.

The Power of a Comparative Approach to Understand Neural CircuitsSince the time of Cajal, comparative approaches have been powerful tools in neuroscience [1].However, in contrast to Cajal and Sherrington's ‘neuron doctrine’, which established that theindividual neuron is the functional unit of the nervous system [2], modern neuroscience is nowfocused on understanding entire neural circuits, as they may have multicellular responsible foremergent functional properties [3]. With the advent of innovative methods, researchers expect torecord and manipulate entire neural circuits. This could generate a dynamic picture that willreveal, perhaps for the first time in depth, how complex neural circuits generate behavior andinternal functional states.

The immense complexity of the human brain, consisting of a hundred billion neurons of a yetunknown number of different types, with each neuron able to connect to tens of thousands ofother neurons, makes a holistic understanding of how the system works, or how neuronalcircuits work on a scale of the entire nervous system of an organism, extremely challenging.Therefore, there is a need to study alternative models with smaller and simpler nervous systems.

Examples of the strength of this comparative approach in neuroscience in the 20th centuryinclude the use of invertebrate models, such as the marine mollusc Aplysia californica, toelucidate mechanisms of neural function that specifically mediate habituation, sensitization,and forms of associative learning [4]. The large neurons of Aplysia allowed the detailedexamination of neuronal architecture, physiology, and control of behaviors at the level of singlewell-characterized cells and defined signaling pathways. Other classical examples of break-throughs that were made possible by using a comparative approach include the understanding

92 Trends in Neurosciences, February 2017, Vol. 40, No. 2 http://dx.doi.org/10.1016/j.tins.2016.11.005

© 2016 Elsevier Ltd. All rights reserved.

6Max Planck Institute forDevelopmental Biology, Tübingen,Germany7ARC Centre of Excellence for CoralReef Studies, Townsville, Australia8Sars International Centre for MarineMolecular Biology, University ofBergen, Norway9University of Queensland, Brisbane,Australia10University of Vienna, Wien, Austria11Neurotechnology Center, ColumbiaUniversity, New York, NY, USA

*Correspondence:tbosch@zoologie.uni-kiel.de(Thomas C.G. Bosch) andrafaelyuste@columbia.edu (R. Yuste).

of the ionic basis of the action potential (squid) [5], the discovery of adult neurogenesis (canary)[6], conditioned reflexes (dog) [7], and the earlier discovery of dendritic spines (chicken) [8]. Morerecently, significant advancements in temporal control of neuronal function through optoge-netics have been made thanks to the characterization of channel-rhodopsins in algae [9].

In contrast to the comparative tradition in neuroscience, the power of molecular genetics hasdriven the exploitation of a selected group of model organisms, particularly Caenorhabditiselegans, Drosophila melanogaster, Xenopus, zebrafish, and mice. Without doubt, these modelorganisms have revolutionized our understanding of biological processes. At the same time,particularly for neuroscience, the emphasis on small number model organisms has come at thecost of essentially ignoring the rich structural and functional diversity of nervous systems in theanimal kingdom. This situation arose due to the difficulty of applying genetic or molecularmethods to most species, thereby leaving them outside the reach of molecular neuroscience.

This outlook has changed dramatically in the last few years due to the introduction ofmolecular genetics that can be applied to a wide variety of species. In many of these cases,it was the Human Genome Project [10] that opened the way for systematically sequencinggenomes of representatives of every animal phylum. Every area of biology has been trans-formed by the development of sequencing technologies and transgenics. This, together withgene-editing techniques such as CRISPR/Cas9 [11], have led to a ‘democratization’ ofmolecular biology, with functional and genomic analyses now possible in a wide range ofspecies. A similar case can be made for the use of calcium imaging of neural circuits [12],which has enabled access to functional information from neurons that were previously toodifficult to record from with electrical methods. Taken together, these advances have led to arenewed interest in studying nervous system evolution, structure, and function, by using non-traditional model animals with simple nervous systems. Such investigations are now not onlypossible, but also appear necessary for understanding the structural determinants of behaviorin more complex animals.

In the following, we briefly review some of the basic features of the neurobiology of variouscnidarians and then illustrate some examples of how modern techniques are starting to yieldsignificant insights into their nervous systems. We end with a ‘call to arms’, pointing out theunique opportunities and potential benefits if we add basal metazoans to the neurosciencemenu.

The Earliest Nervous Systems Were Present in the Common Ancestor ofCnidaria and BilateriaIf one aims to understand how nervous systems function by identifying cardinal shared featuresof neurons and neural circuits via the comparative approach, it is essential to study the earliestevolutionary examples. Remarkably, nervous systems appeared very early in animal evolutionand were certainly in place prior to the origin of bilaterally symmetric animals – also known asBilateria (Figure 1). Of the lineages that diverged prior to the bilaterian radiation, nervous systemsare present only in Ctenophora and Cnidaria. Intriguingly, while cnidarian and bilaterian nervoussystems have many characteristics in common, those of Ctenophora appear to differ funda-mentally – for example, glutamate and neuropeptides may be the sole neurotransmitters used[13]. But, in addition to the controversial phylogenetic position of Ctenophora [14,15], a paucityof data and limited amenability to technical approaches currently make these animals difficultcandidates for comparative studies of neurobiology. Amore practical choice for such studies arecnidarians, a phylum of�11 000 aquatic animals that occupies a strongly supported position asthe sister clade of Bilateria (Figure 1). Cnidarians, which include jellyfishes and polyps and arecharacterized by their nematocytes (stinging cells), have long been utilized in laboratories asexperimental organisms to address diverse biological questions [16–20].

Trends in Neurosciences, February 2017, Vol. 40, No. 2 93

Staurozoa Cubozoa Scyphozoa HydrozoaAnthozoaNematostella∗

XeniaCarybdeaTripedalia∗

AureliaCassiopeaCyanea

AglanthaCly�a∗Hydra∗

Hydrac�nia

Porifera Cnidaria Lophotrochozoa Ecdysozoa Deuterostomia

Bilateria

Eukaryota

Eumetazoa

Protostomia

Bilateria

Ctenophora

Placozoa

Metazoa

(A)

(B)

Focus

Why cnidarians?

Why now?

Cnidarian nervous systems are nerve nets and offer great poten�alfor understanding the basic design principles of neural circuits

● Evolu�onary posi�on as sister group to Bilateria● Rela�vely simple and small nervous systems● Diversity of nervous system design● Diversity of behavioral pa�erns● High plas�city of nervous system and adult neurogenesis

● Genome-wide sequences available in broad range of species● Gene edi�ng has become recently available● New imaging techniques are being applied to these models● Tools to control nerve cells (e.g. optogene�cs) are being developed

Medusozoa

Figure 1.

(Figure legend continued on the bottom of the next page.)

Cnidarians as Model Organisms for Studies of Nervous Systems. (A) Schematic phylogenetic treeshowing the relationships of the five classes within the phylum Cnidaria. Species of cnidarians used for research on nerve

94 Trends in Neurosciences, February 2017, Vol. 40, No. 2

Interestingly, in stark contrast to the centralized nervous systems of bilaterians, cnidariannervous systems function as diffuse nerve nets that can possess varying degrees of regionalgrouping and specialization (Figure 2). This simple neural architecture offers great potential forunderstanding the basic design principles and evolutionary trajectories of nervous systemcircuitry.

In support of the value of analyzing the development, structure, and function of cnidarian nervoussystems, significant advances have been made not only in the application of experimentaltechnologies, but also in the number of cnidarian species that can be interrogated by suchmeans (Figure 1). In recent years, the classical freshwater polyp Hydra and colonial hydrozoanHydractinia echinata have been joined in the laboratory by a sea anemone, Nematostellavectensis and a jellyfish, Clytia hemispherica [21]. For these species, extensive genomic andtranscriptomic resources now exist, and transgenesis and gene-editing technologies are avail-able. Thus, a suite of pre-bilaterian nervous systems is now open to functional analyses toaddress questions that range from the molecular to the behavioral and evolutionary levels. Inaddition, an extensive set of genomic and transcriptomic databases is becoming available forseveral sponge species: Amphimedon queenslandica, Oscarella carmela, and Sycon ciliatum,as well as for a placozoan Trichoplax adhaerens [22–25]. Because sponges and placozoansrepresent an outgroup to cnidarians and bilaterians (Figure 1), and are phyla of metazoanslacking nervous systems, these data are valuable for comparative genomic approaches.

Most Nervous System Components Have Ancient OriginsAlthough phylostratigraphy [26] implies that vertebrate nervous systems have their origins in thecommon ancestor of Cnidaria and Bilateria (herein referred to as the Eumetazoa), individualmolecular components clearly predate this. For example, the processing enzyme for neuro-peptides, a dual function peptidylglycine/-amidating monooxygenase (PAM), is present in earlydiverging phyla such as sponges (Porifera) that lack a nervous system, supporting the idea that,within Metazoa, amidated peptides may have originally functioned in communication betweenepithelial cells [27]. Also, most of the structural components of the postsynaptic density (PSD)are present in the sponge Amphimedon [22,28], and some have clear homologs in choano-flagellates and other unicellular holozoans [29,30]. Other synaptic proteins, particularly thoseinvolved in vesicle exocytosis, are also represented in unicellular holozoans and homologs ofsome of the SNARE proteins (e.g., synaptobrevin2 and syntaxin1) and their interacting partners(e.g., tomosyn) have been even identified in fungi and plants [31].

Thus, remarkably, not only some individual neuronal components predate the appearance ofEumetazoa, but the cnidarian repertoire of proteins associated with nervous systems is essen-tially complete. In fact, homologs of some vertebrate synaptic proteins not present in the modelinvertebrates Drosophila and Caenorhabditis, such as Narp/Pentraxin, are present in cnidarians(though absent from sponges, placozoans and protists). An ancestral richness of cnidarians interms of synaptic proteins is therefore apparent [28,32]. Moreover, because orthologs ofsynaptic proteins may be present in organisms that lack nervous systems, their roles mustbe different [33], suggesting that the eumetazoan ancestor co-opted these to enable neuronalcommunication.

In fact, few synaptic proteins are unique to vertebrates; of the 74 protein components of themammalian nervous system listed by Burkhardt et al. [33], only three (SynCam, Piccolo, and

systems and referred to in this Review are listed. Complete genome sequences are available for several cnidarian species(marked with asterisk) [21]. (B) Schematic phylogenetic tree showing main branches of metazoan evolution and position ofCnidaria among the non-bilaterian Metazoa. As the phylogenetic position of comb jellies (Ctenophora) remains controversial[13–15], the branch leading to this group is represented by a dashed line.

Trends in Neurosciences, February 2017, Vol. 40, No. 2 95

Bidirec�onal synapses

Complex sense organsonr

sc

shc

Sensory cnidocytes

Local condensa�on Chemical heterogeneity

Complex behaviors

Nerve net

Plas�city and regenera�on

Figure 2. Distinctive Features of Cnidarian Nervous System. A diffuse nerve net is a basic design of cnidarian nervesystems. Nerve net of the sea anemone Nematostella vectensis revealed by expression of a mCherry protein under nerve-specific elav promotor. Adapted from [98]. Local condensation of the nerve net is observed in most cnidarians [69,111,112].For instance, in Hydra, high density of the JD1-positive sensory neurons is observed around the mouth opening (hypostome)of a polyp. Adapted from [93]. Chemical heterogeneity of the cnidarian nerve net contrasts with its apparent morphologicalsimplicity. A broad range of neurotransmitters present in cnidarian nerve cells are expressed by particular subsets of neurons.For instance, in Hydra RFamide (left) and GLWamide (right) are produced by two apparently nonoverlapping populations ofneurons. Adapted from [93]. Sensory cnidocytes represent a type of mechanosensory cell found exclusively in cnidarians.Mechanical stimulation of a resting cnidocyte (left) triggers rapid discharge of a nematocyst and release of a harpoon (right) tospear and paralyze prey. Adapted from [113]. Complex sense organs are found in cnidarians. A statocyst from the umbrellarmargin of the hydrozoan jellyfishAglantha contains a concretion (c) surrounded by sensory cells (sc) with sensory cilia (sh), thatare connected to the outer nerve ring (onr) of the umbrella. Declination of the concretion stimulates cilia, enabling the vestibularsense of the medusa. Adapted from [66]. Bidirectional synapses are common for cnidarians. An electron micrograph of asynapse between two axons in the scyphozoan Cyanea represents neurotubules (t) and small vesicles (bs) on both sides of asynaptic cleft (s). Adapted from [36]. Plasticity and regeneration are characteristic for nervous systems of cnidariansundergoing constant asexual proliferation. In vivo imaging of transgenic cells (here a neuron of Hydra, adapted from[114]) allows studying how this plasticity is accomplished by integrating stem-cell-derived migratory neuronal precursorcells into the nervous system. Complex behaviors, such as somersaulting in Hydra, emerge from the activity of simple nervenets; the circuit, cellular and molecular mechanisms behind remain poorly understood. Adapted from [92].

Bassoon) lack homologs in other phyla. A caveat here is that generalizations about early nervoussystem evolution are currently based on relatively sparse taxonomic sampling; at presentgenome assemblies are available for only a few non-bilaterian animals or holozoans. The patchydistribution of some gene families across Metazoa, perhaps resulting from stochastic gene loss

96 Trends in Neurosciences, February 2017, Vol. 40, No. 2

[34] implies that more extensive taxon sampling may eventually identify larger numbers ofsynaptic proteins in both cnidarians and other groups, mandating the revisiting of evolutionaryscenarios based on presently available data.

The Hidden Complexity of Cnidarian Nervous SystemsThe anatomical simplicity of the cnidarian nervous system masks some remarkable neurophys-iological specializations [35] including, for example, bidirectional chemical synapses (Figure 2)[36,37], signaling by a diversity of peptide-gated channels [38–40], the rapid discharge ofnematocysts (Figure 2) [41], and axons with two kinds of impulse propagation [35,42–45]. Inthe following section, we briefly touch on these mechanisms and other distinct neural character-istics to illustrate the functional sophistication of these primitive nervous systems and highlightfertile avenues for neurobiological research. These examples could also help put in perspectivestandard mechanisms of synaptic transmission and neuronal integration found in bilaterians.

Neurotransmitters and ReceptorsFunctional analysis of cnidarian nervous systems requires knowledge of the mode of synaptictransmission between individual neurons. This could reveal fundamental insights into basicdesign principles of excitatory and inhibitory transmission, and into how this design for com-munication is used within a neural circuit. While unidirectional synapses are the norm in bothcnidarians and bilaterians, bidirectional synapses can be found, for example, in the mammalianretina and olfactory system [46,47]. Bidirectional synapses in Cnidaria (Figure 2) were firstdescribed at the ultrastructural level, containing synaptic vesicles accumulated on both sides ofthe synapse [36]. Electrophysiological recordings revealed that such bidirectional synapses areexcitatory and nonpolarized, and conduct equally well in either direction [37].

Besides ultrastructural evidence for chemical synapses, ample histochemical, biochemical, andfunctional data have indicated the presence in Cnidaria of different small molecule neuro-transmitters, such as catecholamines, serotonin, acetylcholine, glutamate, and g-aminobutyricacid (GABA) [48,49]. Their precise functional role, however, remains largely undefined, mainlybecause receptors for these putative transmitters have not yet been identified using biochemicalmethods.

In Cnidaria, cloning and characterization of transmitter receptors have in the past been slow andmainly performed by researchers focused on individual gene families [38,39], thus leaving thereceptor repertoire of Cnidaria as mostly unknown territory. But sequencing the genomes ofdifferent cnidarian species, in particular those of the anthozoans Nematostella vectensis,Acropora, and Aiptasia and the hydrozoan Hydra magnipapillata (Figure 1) [50,51], has offeredthe possibility to quickly gather an overview on the entire complement of receptors in thesespecies. Moreover, the ability to perform in situ hybridization (ISH), allows a quick overview of theexpression of a specific receptor in the whole organism, with particular focus on whether a givenreceptor gene is expressed in nerve cells.

The metabotropic and ionotropic receptor repertoire of Nematostella has been inspected bygenomic analysis in some detail [52], revealing the presence of several metabotropic glutamateand GABA(B) receptors, as well as a large number of G-protein-coupled receptors (GPCRs) formonoamines and melatonin [52]. Strikingly, specific biogenic amine-synthesizing enzymes aresparse, serotonin is apparently lacking, and it has been concluded that aminergic-like trans-mitters unique to sea anemones act on these receptors [52]. Interestingly, again, many trans-mitter-related protein classes appeared closer to vertebrate than to invertebrate counterparts;for example, octopamine and its receptors, which are common in invertebrates, are missing inNematostella [52]. Also, while not present in Nematostella [52], two muscarinic acetylcholinereceptors have been identified in Hydra [53].

Trends in Neurosciences, February 2017, Vol. 40, No. 2 97

Fast synaptic transmission and synaptic plasticity in Bilateria relies on ion channel receptors. Thefunctional requirements of complex neuronal signaling have driven the evolution of various ionchannel gene families that differ with respect to ligands, ion selectivity, and kinetics (onset ofactivation and desensitization time course). In addition to metabotropic receptors, the Nem-atostella genome contains several genes for the major classes of ionotropic receptors [nicotinicacetylcholine receptors, glutamate receptors of the /-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA), kainate and N-methyl-D-aspartate (NMDA) type, GABA(A), and P2X];only an 5-hydroxytryptamine-3 receptor homolog is apparently lacking [52]. Thus, the diversity ofreceptor subtypes in cnidarians is, in principle, comparable to that of chordates and notablyincludes AMPA and NMDA receptors, which mediate synaptic plasticity in the vertebrate centralnervous system [54]. The presence of these receptors suggests that the cnidarian nerve net ischemically complex and uses several small molecule transmitters. It also suggests that cni-darians build their nervous systems using essentially the same building blocks present inbilaterians, so nervous system evolution may result more from the rearrangement of existingmolecular components than from the invention of novel ones. It should be emphasized,however, that the assignment of cnidarian genes to a receptor class was exclusively basedon homology to bilaterian receptors [52]. An unequivocal assignment still requires the detailedfunctional analysis of ligand specificity and, in the case of ionotropic receptors, ion selectivity. Offurther note, the diversity of ion channel receptors in Cnidaria is in stark contrast to the relativepaucity of ion channel families present in ctenophores, supporting views of independent neuralevolution in the Ctenophora [13].

In addition to small-molecule neurotransmitters, the cnidarian nervous system uses neuro-peptides extensively and perhaps predominately [52,55]. In Nematostella, the diversity ofneuropeptides is matched by >80 putative GPCRs for neuropeptides [52]. Strikingly, neuro-peptide signaling in Hydra is mediated not only by GPCRs, but also by ion channel receptors[38,40,56]. These channels are directly gated by RFamide neuropeptides, which are alsopresent in vertebrates and invertebrates [57]. The presence of RFamide neuropeptides in large,dense core vesicles at the neuromuscular junction [58] and of the ionotropic RFamide-receptors(iRFa-Rs) in epitheliomuscular cells together with pharmacological evidence suggests that theiRFa-Rs are involved in neuromuscular transmission [39,40]. Homologous channels also exist inNematostella [56] and in some Protostomia [59], suggesting that neuropeptides have a broadfunction in fast neuronal communication in cnidarians.

While in most cases it is not clear whether these transmitters and receptors identified incnidarians have a neuronal location and function, the large repertoire of GPCRs and ion channelreceptors suggests that the cnidarian nerve net, while structurally simple, is chemically complex.The cnidarian nerve net thus appears to be an interesting object to study how a complex toolboxof transmitters and receptors is used to enable the emergence of behaviors within a morpho-logically simple nerve net structure. This apparent paradox makes it possible that cnidarians usea ‘chemical connectome’ [60], that is, a sophisticated chemical agonist–receptor matchingspace that implements the selectivity necessary for specific behavioral patterns. From this view,a specific muscle contraction, for example, could be achieved not by precisely connecting amotor neuron with a particular muscle fiber, but by expressing a particular receptor combinationin that muscle fiber, while the neurotransmitter is widely distributed across all muscle fibers.Thus, instead of implementing computations in its physical wiring, the cnidarian nervous systemmay primarily operate using a ‘chemical wiring diagram’. We note that this view supports earlierideas proposed by Carl Pantin [61].

Specialized Cellular Composition of Cnidarian Nervous SystemsWhile the relatively morphological simplicity of their nerve nets is a basic feature of cnidariannervous systems (Figure 2), they have independently evolved several fascinating neural

98 Trends in Neurosciences, February 2017, Vol. 40, No. 2

structures and properties that provide unique opportunities to understand how more complexnervous systems can evolve.

For example, all cnidarians are endowed with specialized mechanosensory cells (the nema-tocyst/cnidocysts-containing nematocytes/cnidocytes, which give name to the phylum,Figure 2) that enable them to spear and paralyze prey with supersonic harpoons. Indeed,the discharge of nematocysts is one of the fastest processes in biology. This discharge istriggered by mechanical stimulation, and uses a fast-recoiling, silk-like elastic protein, cnidoin,that incorporates into the capsule of the nematocyst and stores kinetic energy [41]. There is astriking structural similarity between nematocysts and the ciliary–microvillar sensory appara-tus of mechanosensory cells in some cnidarians [62,63], suggesting that nematocystsand mechanosensory structures are evolutionarily related. The ciliated mechanosensory cellsin cnidarians are also similar to the hair cells of vertebrates, and may represent descendantsof an ancient metazoan sensory cell type [64]. A better understanding of their cell biologycould inform us further about the evolution and diversification of animal mechanosensorysystems.

Sense Organs and NavigationIn addition to cnidocytes, some cnidarians have elaborate sense organs and perform intricatebehaviors (Figure 2) that suggest advanced neural integration. For example, the statocysts inScyphozoa (Figure 1) are small tentacle-like organs that hang at the outer side of the bell of somemedusa and are involved in sensing gravity, a sense still poorly understood even in bilaterians(Figure 2). Statocysts contain a concretion in their distal part, and are surrounded by nonmotilemechanosensory cilia [65,66]. Movement deflects cilia, allowing the medusa to sense itsorientation in the water. These sense organs, therefore, are intriguing objects to study ancestralfeedback control mechanisms and may even be informative for understanding bilaterian ves-tibular sensing systems.

In terms of vision, an interesting specialization is found in the cubomedusae Tripedalia andCarybdea (Cubozoa, Figure 1), which have four sensory structures, the rhopalia, which containcamera-type eyes. These eyes enable the box jellyfish to avoid obstacles and to navigate in thewater based on terrestrial cues [67–69]. Rhopalia are integrated with the rest of the nervoussystem [70], and visual input modulates the pacemaker system that controls the contraction ofthe swimming bell [69,71]. The structure and function of such visual systems in non-bilateriansprovides further support to the view [72] that camera-type eyes have independently evolvedseveral times. Further investigations may moreover reveal how complex eyes can function in theabsence of a centralized nervous system.

Giant AxonsAnother fascinating specialization of cnidarians is found in some medusae, which are endowedwith larger axonal structures. The giant axon system of a hydrozoan Aglantha digitale (Figure 1)represents an advanced specialization that has been extensively studied due to its amenability toelectrophysiological recordings [42,45,73–78]. Interestingly, the giant axons in the jellyfishAglantha are reported to have two kinds of impulse propagation [79]. During slow swimming,low amplitude Ca2+ spikes drive weak muscle contractions, but during predator avoidance, thejellyfish switches to fast escape swimming that relies on strong contractions elicited by rapidlyconducted Na+-dependent action. The same eight giant motor axons thus can mediate eitherslow or fast swimming modes, synapsing on the contractile myoepithelial cells as part of anelaborate neural circuitry. This system arguably represents the best-understood circuitry in anycnidarian [80]. The circuits mediating locomotion, feeding, and tentacle contractions in Aglanthaare composed of at least 14 distinct neuron types with dedicated functions, including a relay, acarrier, and a pacemaker system [42,45,73,74]. This complexity and functional sophistication

Trends in Neurosciences, February 2017, Vol. 40, No. 2 99

stands in contrast to the default concept of a simple nerve net in cnidarians, and indicates thatAglantha has found ingenious ways to implement different behaviors using the same available‘hardware’.

Behavior in CnidariansOne of the greatest surprises in cnidarian neuroscience has been the realization that theirbehavioral repertoire is unexpectedly complex, given the apparent structural simplicity ofcnidarian nerve nets (Figure 2). Their behavioral sophistication is something that was alreadyappreciated in the 18th century, when Abraham Trembley first described somersaultinglocomotion in Hydra [81] (Figure 2). As a single footed polyp, Hydra cannot translate itsposition in the bottom of fresh water ponds and rivers without being carried away bythe current. As a solution, Hydra bends its body column, attaches its tentacles to thesubstrate, releases its foot, swings its body over to reattach its foot, and then releases thetentacles to become erect again. It is currently not at all understood how this complexbehavior, which indeed not all authors of this review can master, is achieved via a diffuse netof neurons.

Spontaneous body column contractions in Hydra is another coordinated behavior, which isbased on a subpopulation of nerve cells [44,82,83]. Beyond Hydra, spontaneous rhythmicpulsations of tentacles in soft corals (as Xenia, Anthozoa, Figure 1) [84] were first noted byLamarck nearly 200 years ago. Among medusae, such as Cassiopea (Scyphozoa, Figure 1),the spontaneous rhythmic pulsations of the medusa bell present a behavior pattern in whichthe influence of certain sense organs and the overall control of the nervous system have beeninvestigated [85,86]. Moreover, growth pulsations, that is, successive rhythmic extensions andretractions of the shoot and stolon growing tips, are thought to be essential for growth andmorphogenesis of colonial hydroid polyps (Hydrozoa, Figure 1) [87–89]. Another impressivebehavior among cnidarians is the ‘wedding dance’, or courting behavior exhibited by somecubozoan medusae [90]. In the sexually dimorphic cubomedusa Carybdea sivickisi (Cubozoa,Figure 1), sexually mature females produce conspicuous velar spots and males only courtfemales with these spots. During mating, the male attaches a tentacle to the female and bringstheir oral openings (manubriums) into direct contact. The male forms a spermatophore, whichis then transferred to one of the female tentacles. Following release by the male, the femaleinserts the spermatophore into her manubrium [91]. This behavior includes elements ofrecognition of the other sex and the assessment of sexual maturity. Because the conspicuousvelar spots of the females are important for mating, recognition is likely mediated at least partlyby visual cues.

Feeding behavior in jellyfish and polyps is another highly coordinated behavior, includingnematocyte discharge, tentacle flexing, and lip flaring to ingest prey [74,92,93]. In Hydra, forexample, this behavior can be elicited in the laboratory by micromolar concentrations of reducedglutathione, therefore allowing to study and manipulate behavior under controlled conditions.Feeding behavior of the polyps can be easily monitored using a dissecting microscope, andvideo-recorded to register and quantify sequence and timing of the events. The Hydra nervoussystem must implement a sustained behavioral plan, composed of many independent modulesthat are precisely arranged in time. This provides us with an example of an early-evolved fixedaction pattern.

Finally, some sea anemones show aggression behavior towards other species by stretching outtheir tentacles to their adversaries, firing the nematocysts, and ripping off tissues from them. Thisbehavior usually occurs in clonal species, in which sea anemone differentiate into five morpho-logically distinct types including warriors and non-fighting polyps, and the fighting behaviorinvolves communication between these types.

100 Trends in Neurosciences, February 2017, Vol. 40, No. 2

Plasticity of the Nerve NetCnidarians have a long history as experimental animals for regeneration and pattern formationbeginning with Abraham Trembley's bisection experiments in 1744 [81] and the classic‘developmental organizer’ transplant experiments of Ethel Browne in Thomas Hunt Morgan'slaboratory [94], both using Hydra. The spectacular ability to rebuild any missing body partincludes the generation of large numbers of new nerve cells that seamlessly connect with theexisting nervous system (Figure 2). In Hydra, this is accomplished by integrating stem-cell-derivedmigratory neuronal precursor cells into the nervous system [95,96]. Indeed, an increasein nerve cell density is the first change detected in cell distribution upon budding and regener-ation [97]. In 12–24 h, the local density of neurons may double in the amputation region oremerging bud.

Both normal and regeneration-induced neurogenesis implies fast and effective integration ofnew neurons into the nerve net, through rewiring and re-establishment of synaptic contacts. Thisremarkable neurogenic potential also characterizes the initial formation of the nervous systemduring embryogenesis. For example, in the sea anemone Nematostella vectensis, neurons aregenerated throughout the ectoderm as well as in the endoderm [98]. The molecular control ofneurogenesis is remarkably similar to that in vertebrates: Notch signaling controls the number ofneural progenitor cells; SoxB and bHLH transcription factor encoding genes regulate theirfurther differentiation; and Wnt signaling gradients are involved in the patterning of the nervoussystem [99–101]. The spatially broad neurogenic potential of cnidarians poses specific chal-lenges for the developmental control of neurogenesis, for example, how is the balance betweenneural and non-neural cell types regulated? Which mechanisms determine the number of neuralprogenitor cells? Given the molecular similarities, understanding these questions in cnidarianscan shed light on vertebrate neurogenesis. In addition, understanding in detail how the cnidarianneurogenic program can be promiscuously activated throughout both germ layers of the embryoand during regeneration has the potential to stimulate research to improve the generation ofneurons in pathological conditions in humans.

Modern Methods Reach CnidariaThere are several fascinating aspects of the cnidarian nervous system that merit further attention(see Outstanding Questions), from the molecular and subcellular level, through the cellular,neural circuit, and organ levels, and up to the behavioral level. Studying cnidarians permitsfundamental questions about the function and evolution of nervous systems to be addressed,such as: (1) how can relatively complex macro-behaviors such as swimming/feeding/reproduc-ing/somersaulting be coordinated by noncentralized neural networks; and (2) what are the corestructural and functional features of all metazoan neurons.

The relevance of cnidarians for understanding the evolution of nervous systems and informationprocessing in simple nerve nets has long been recognized, but now faces a reignited interest(Figure 1). Indeed, over the past decade, modern molecular biology technology has beenimplemented in several cnidarian species, allowing many of the questions that have beeninaccessible in the past to be finally addressed. Stable transgenesis in a cnidarian was firstachieved with Hydra [102] and is also now available in both Nematostella and Hydractinia[103,104] (Figure 2). The use of cell-type-specific promoters for reporter genes has alsoimproved the visualization of neuronal morphology and tracking of the developmental originof nerve cells [98,100]. Transgenic lines can also be used to isolate specific neuronal sub-populations by fluorescence-activated cell sorting and determination of their transcriptome andproteome profiles. The availability of stable transgenesis and neuron-specific promoters nowallows the conditional ablation of neurons in adult animals, opening the possibility to analyzethe regenerative capacity of cnidarians in new experimental paradigms. In addition to well-established transient knockdown strategies (double-stranded RNA, and morpholinos),

Trends in Neurosciences, February 2017, Vol. 40, No. 2 101

Outstanding QuestionsHow do simple nerve nets work?Reconstructing entire neural circuits,recording their dynamic activity mapscould lead to understanding organism-wide neural connectivity and how itgenerates different activity patterns.

What was the original function of thenervous system? Did it emerge to con-trol motility, generate internal functionalstates, monitor the environment, or toorchestrate multiple functions, asdevelopment, tissue homeostasis,and host–microbiome interactions?

How are constant neurogenesis, nervenet regeneration and plasticityachieved in adult asexually proliferatingcnidarians? What are the core princi-ples and the main molecular mecha-nisms of neurogenesis that stay activethroughout the life of a cnidarian?

What were the main trajectories of ner-vous systems evolution? Reconstruct-ing evolution of nervous systems wouldclarify the role of invention of novel

genome-editing technologies like TALENs (transcription activator-like effector nucleases) orCRISPR/Cas9 are now available in some cnidarian model systems [105,106].

The transgenic expression of pre- and postsynaptic marker proteins will significantly facilitate thedescription of neural connectivity and how it is established. Indeed, the first experiments usinggenetically encoded calcium indicators to monitor organism-wide neural activity are nowunderway [107] and in conjunction with the use of optogenetics tools (e.g., light-controlledion channels), these will be instrumental in understanding how noncentralized nervous systemsgenerate behavior and react in response to environmental cues. Finally, detailed maps of theentire nervous system in cnidarian models generated with the aforementioned technologies maybecome a valuable input for computational models, intended to simulate function of a nerve netand provide testable predictions.

Concluding Remarks and Future PerspectivesIn summary, we are at the beginning of what could constitute a revolution in the study of theneurobiology of cnidarians and other basal metazoans. The systematic use of large scaleimaging, optogenetics and molecular engineering methods could permit the elucidations ofbasic principles of neural circuits and the expansion of synthetic biology to metazoans. Thecomparative analysis of cnidarian and bilaterian nervous systems may allow the identification ofancient and therefore fundamental principles of nervous system structure and function (Box 1).The great molecular similarity between the cnidarian and bilaterian nervous systems makestenable the possibility that many of the basic circuit mechanisms are also conserved. At present,there are large research communities deciphering these principles in the nervous systems of

Box 1. Potential Promise of Cnidarian Neuroscience

Cnidaria offer a ‘rich playground’ to explore in relatively simple and accessible nervous systems many of the classicalquestions in neuroscience. Using calcium imaging of the complete pattern of neural activity, the neural basis of manybehaviors such as feeding, swimming, egestion, and somersaulting could be elucidated, given the small number ofneurons that some cnidarian have. These efforts could be aided by the reconstructions of their connectomes; a task thatappears feasible for a sparse nerve net, such as those present in many cnidarians. Similarly, complete access to suchsimple systems could also enable the rigorous modeling of neural circuits, with simulations that are constrainedexperimentally in terms of the number of neurons, connections between the neurons, and activity patterns.

The decoding of the activity patterns involved in behavior could reveal general principles of neural circuit function usedmore widely in evolution. For example, it is likely that neural circuits generate emergent states of activity, such asdynamical attractors [115]. These attractors are difficult to identify and manipulate in most laboratory species, due to thefact that functional recordings or manipulation of the activity are always incomplete. In the case of a cnidarian, however,one could measure the complete record of neural activity for every neuron for a long period of time. This could enable therigorous identification of a large portion of the dynamical space of the nerve net, and the systematic mapping of attractorsto behavioral or functional states. While it is impossible to predict how these dynamical landscapes will appear, it is likelythat insights acquired in this research could help guide similar research agenda in Bilateria. The initial results in the brain-wide imaging ofHydra vulgaris, which displays robust endogenous and sensory-drive dynamics [107], bodes well for thesuccess of such an experimental program.

Besides offering a comparative approach to classical questions, there is a deep evolutionary question that cnidariancould help answer; namely, the function of the first nervous systems. Indeed, the original function of the nervous systemstill remains mysterious [116]. While most ascribe the invention of nervous systems to the evolutionary need for motorcontrol, some animals can move in a coordinated fashion without a nervous system, such as sponges and nerve-freeHydra, where epitheliomuscular cells form excitable epithelia. Understanding in detail the neurobiology of the cnidarian,which are among the most primitive representatives of the first nervous systems in evolution, can shed light on this issue.

Cnidarians can also help explore noncanonical functions of nervous systems. Current opinion mostly perceives thenervous system as ameans of communication and information exchange between the central nervous system, the rest ofthe body and the environment. However, the spectrum of functions performed by nervous systems is broader.Noncanonical functions may include a role in regeneration, development of innervated tissues, and tissue homeostasis[117,118], as well as bidirectional communication with the commensal microbiota [119–124].

molecular elements and rearrange-ment of existing ones. It would revealancient and therefore fundamentalprinciples of nerve system's structureand function.

How is the complex toolbox of trans-mitters, receptors, and ion channelsused in cnidarians to enable emer-gence of complex behaviors with aseemingly simple nerve net structure?

102 Trends in Neurosciences, February 2017, Vol. 40, No. 2

Caenorhabditis and Drosophila. A large body of literature covers the molecular makeup and thewiring of their nervous systems and how specific behaviors emerge from these properties. In thiscontext, a small investment towards elucidating the cnidarian nervous systems could have amajor payoff. Although the standard model animals have been useful in the past and will nodoubt continue to deliver important insights in the future, one has to be aware that bothCaenorhabditis and Drosophila are highly diverged organisms characterized by high levels ofgene loss and sequence divergence, whereas at least some cnidarians have retained much ofthe ancestral genetic complexity of metazoans [108–110]. Reflecting this ancestral core,cnidarian genes are often more similar to vertebrate genes than vertebrate genes are to thoseof more diverged animals like Caenorhabditis or Drosophila [50,109]. Thus, in addition to thesimple design of the cnidarian nervous system, its comparative analysis might reveal new andunexpected basic features of nervous systems. It seems to us that, armedwith a new generationof methods, the time is ripe for a deep exploration of the neurobiology of cnidarians.

AcknowledgmentsThe work in the Bosch laboratory (T.B., A.K., A.M., K.S.) related to this review was supported in part by grants from the

Deutsche Forschungsgemeinschaft (DFG), the CRC 1182 (“Origin and Function of Metaorganisms”) and the Cluster of

Excellence “Inflammation at Interfaces”. Support by the Alexander von Humboldt foundation (A.K.) and Max Plank Institute

for Evolutionary Biology (A.M.) is gratefully acknowledged. Work in the laboratory of S.G. was supported by the DFG grant

GR1771/7-1. F.R. and G.R. are supported by the Sars Centre core budget and a Marie Curie Incoming Postdoctoral

Fellowship, respectively. D.J.M. gratefully acknowledges the support of the Australian Research Council, both directly

(DP1095343) and indirectly via the ARC Centre of Excellence for Coral Reef Studies (CE14100020). U.T. is funded by the

Austrian Science Fund (FWF P27353). T.D.-L. acknowledges Adris Foundation and City of Zagreb. R.Y. thanks the NEI

(DP1EY024503) and Rob Steele, Christophe Dupre, Shuting Han, and John Szymanski for thoughtful comments. This

material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No.

HR0011-17-C-0026.

References

1. Cajal, S.R. y (1967) The structure and connexions of neurons. In

Nobel Lectures, Physiology or Medicine, 1901–1921, pp. 220–253, Elsevier Publishing Company

2. Shepherd, G.M. (1991) Foundations of the Neuron Doctrine,Oxford University Press

3. Yuste, R. (2015) From the neuron doctrine to neural networks.Nat. Rev. Neurosci. 16, 487–497

4. Kandel, E. and Schwartz, J. (2013) Principles of Neural Science.(5th edn), McGraw-Hill Education

5. Hodgkin, A.L. and Huxley, A.F. (1952) The components of mem-brane conductance in the giant axon of Loligo. J. Physiol. 116,473–496

6. Nottebohm, F. et al. (1990) Song learning in birds: the relationbetween perception and production. Philos. Trans. R. Soc. Lon-don B Biol. Sci. 329, 115–124

7. Pavlov, I.P. (1951) Conditioned reflex. Feldsher Akush 10, 3–10

8. Cajal, S.R. y (1888) Estructura de los Centros Nerviosos de lasaves. Rev. Trim. Histol. Norm. Pat. 1, 1–10

9. Nagel, G. et al. (2003) Channelrhodopsin-2, a directly light-gatedcation-selective membrane channel. Proc. Natl. Acad. Sci. 100,13940–13945

10. Yager, T.D. et al. (1991) The Human Genome Project: creatingan infrastructure for biology and medicine. Trends Biochem.Sci. 16, 454

11. Doudna, J.A. and Charpentier, E. (2014) The new frontierof genome engineering with CRISPR-Cas9. Science 346,1258096

12. Yuste, R. and Katz, L.C. (1991) Control of postsynaptic Ca 2+influx in developing neocortex by excitatory and inhibitory neuro-transmitters. Neuron 6, 333–344

13. Moroz, L.L. et al. (2014) The ctenophore genome and the evolu-tionary origins of neural systems. Nature 510, 109–114

14. Jékely, G. et al. (2015) The phylogenetic position of ctenophoresand the origin(s) of nervous systems. Evodevo 6, 1

15. Pisani, D. and Liu, A.G. (2015) Animal evolution: only rocks canset the clock. Curr. Biol. 25, R1079–R1081

16. Technau, U. and Steele, R.E. (2011) Evolutionary crossroads indevelopmental biology: Cnidaria. Development 138, 1447–1458

17. Watanabe, H. et al. (2014) Nodal signalling determines biradialasymmetry in Hydra. Nature 515, 112–115

18. Kelava, I. et al. (2015) Evolution of eumetazoan nervous systems:insights from cnidarians. Phil. Trans. R. Soc. B 370, 20150065

19. Bosch, T.C.G. (2014) Rethinking the role of immunity: lessonsfrom Hydra. Trends Immunol. 35, 495–502

20. Bosch, T.C.G. et al. (2014) How do environmental factors influ-ence life cycles and development? An experimental frameworkfor early-diverging metazoans. BioEssays 36, 1185–1194

21. Technau, U. and Schwaiger, M. (2015) Recent advances ingenomics and transcriptomics of cnidarians. Mar. Genomics24, 131–138

22. Srivastava, M. et al. (2010) The Amphimedon queenslandicagenome and the evolution of animal complexity. Nature 466,720–726

23. Srivastava, M. et al. (2008) The Trichoplax genome and thenature of placozoans. Nature 454, 955–960

24. Fortunato, S.A.V. et al. (2015) Comparative analyses of develop-mental transcription factor repertoires in sponges reveal unex-pected complexity of the earliest animals. Mar. Genomics 24,121–129

25. Riesgo, A. et al. (2014) The analysis of eight transcriptomes fromall poriferan classes reveals surprising genetic complexity insponges. Mol. Biol. Evol. 31, 1102–1120

26. Šestak, M.S. et al. (2013) Phylostratigraphic profiles reveal adeep evolutionary history of the vertebrate head sensory sys-tems. Front. Zool. 10, 1

27. Attenborough, R.M.F. et al. (2012) A “neural” enzyme in non-bilaterian animals and algae: preneural origins for peptidylglycine/-amidating monooxygenase. Mol. Biol. Evol. 29, 3095–3109

Trends in Neurosciences, February 2017, Vol. 40, No. 2 103

28. Sakarya, O. et al. (2007) A post-synaptic scaffold at the origin ofthe animal kingdom. PLoS ONE 2, e506

29. Suga, H. et al. (2013) The Capsaspora genome reveals a com-plex unicellular prehistory of animals. Nat. Commun. 4, http://dx.doi.org/10.1038/ncomms3325

30. Burkhardt, P. (2015) The origin and evolution of synaptic proteins–choanoflagellates lead the way. J. Exp. Biol. 218, 506–514

31. Roshchina, V.V. (2016) New trends and perspectives in theevolution of neurotransmitters in microbial, plant, and animalcells. InMicrobial Endocrinology: Interkingdom Signaling in Infec-tious Disease and Health, pp. 25–77, Springer

32. Alié, A. and Manuel, M. (2010) The backbone of the post-syn-aptic density originated in a unicellular ancestor of choanoflagel-lates and metazoans. BMC Evol. Biol. 10, 1

33. Burkhardt, P. et al. (2014) Evolutionary insights into premetazoanfunctions of the neuronal protein homer. Mol. Biol. Evol. 31,2342–2355

34. Forêt, S. et al. (2010) New tricks with old genes: the geneticbases of novel cnidarian traits. Trends Genet. 26, 154–158

35. Meech, R.W. (2015) Electrogenesis in the lower Metazoa andimplications for neuronal integration. J. Exp. Biol. 218, 537–550

36. Horridge, G.A. and Mackay, B. (1962) Naked axons and sym-metrical synapses in coelenterates. J. Cell Sci. 3, 531–541

37. Anderson, P.A. (1985) Physiology of a bidirectional, excitatory,chemical synapse. J. Neurophysiol. 53, 821–835

38. Golubovic, A. et al. (2007) A peptide-gated ion channel from thefreshwater polyp Hydra. J. Biol. Chem. 282, 35098–35103

39. Dürrnagel, S. et al. (2010) Three homologous subunits form ahigh affinity peptide-gated ion channel in Hydra. J. Biol. Chem.285, 11958–11965

40. Assmann, M. et al. (2014) The comprehensive analysis of DEG/ENaC subunits in Hydra reveals a large variety of peptide-gatedchannels, potentially involved in neuromuscular transmission.BMC Biol. 12, 1

41. Beckmann, A. et al. (2015) A fast recoiling silk-like elastomerfacilitates nanosecond nematocyst discharge. BMC Biol. 13, 1

42. Mackie, G. and Meech, R. (1995) Central circuitry in the jellyfishAglantha. I: The relay system. J. Exp. Biol. 198, 2261–2270

43. Westfall, J.A. et al. (1980) Neuro-epitheliomuscular cell and neuro-neuronal gap junctions in Hydra. J. Neurocytol. 9, 725–732

44. Takaku, Y. et al. (2014) Innexin gap junctions in nerve cellscoordinate spontaneous contractile behavior in Hydra polyps.Sci. Rep. 4, 3573

45. Mackie, G. and Meech, R. (1995) Central circuitry in the jellyfishAglantha. II: The ring giant and carrier systems. J. Exp. Biol. 198,2271–2278

46. Shepherd, G.M. and Greer, C.A. (1990) Olfactory bulb. In Thesynaptic Organization of the Brain, (3rd edn), pp. 133–169,Oxford University Press

47. Sterling, P. (1990) Retina. In The Synaptic Organization of theBrain (Shepherd, G.M., ed.), pp. 170–213, Oxford UniversityPress

48. Kass-Simon, G. and Pierobon, P. (2007) Cnidarian chemicalneurotransmission, an updated overview. Comp. Biochem.Physiol. Part A Mol. Integr. Physiol. 146, 9–25

49. Pierobon, P. (2012) Coordinated modulation of cellular signalingthrough ligand-gated ion channels in Hydra vulgaris (Cnidaria,Hydrozoa). Int. J. Dev. Biol. 56, 551–565

50. Putnam, N.H. et al. (2007) Sea anemone genome reveals ances-tral eumetazoan gene repertoire and genomic organization. Sci-ence 317, 86–94

51. Chapman, J. et al. (2010) The dynamic genome of Hydra. Nature464, 592–596

52. Anctil, M. (2009) Chemical transmission in the sea anemoneNematostella vectensis: a genomic perspective. Comp. Bio-chem. Physiol. Part D Genomics Proteomics 4, 268–289

53. Collin, C. et al. (2013) Two types of muscarinic acetylcholinereceptors in Drosophila and other arthropods. Cell. Mol. Life Sci.70, 3231–3242

54. Ryan, T.J. and Grant, S.G.N. (2009) The origin and evolution ofsynapses. Nat. Rev. Neurosci. 10, 701–712

104 Trends in Neurosciences, February 2017, Vol. 40, No. 2

55. Grimmelikhuijzen, C.J.P. et al. (2004) Neuropeptides in cnidar-ians. In Cell Signalling in Prokaryotes and Lower Metazoa, pp.115–139, Springer

56. Gründer, S. and Assmann, M. (2015) Peptide-gated ion chan-nels and the simple nervous system of Hydra. J. Exp. Biol. 218,551–561

57. Jékely, G. (2013) Global view of the evolution and diversity ofmetazoan neuropeptide signaling. Proc. Natl. Acad. Sci. 110,8702–8707

58. Koizumi, O. et al. (1989) Ultrastructural localization of RFamide-like peptides in neuronal dense-cored vesicles in the peduncle ofHydra. J. Exp. Zool. 249, 17–22

59. Lingueglia, E. et al. (1995) Cloning of the amiloride-sensitiveFMRFamide peptide-gated sodium channel. Nature 378,730–733

60. Andrews, A.M. (2013) The BRAIN Initiative: toward a chemicalconnectome. ACS Chem. Neurosci. 4, 645

61. Pantin, C.F.A. (1956) The origin of the nervous system. Pubbl.Staz. Zool. Napoli 28, 171–181

62. Mattern, C.F.T. et al. (1965) Electron microscope observationson the structure and discharge of the stenotele of Hydra. J. CellBiol. 27, 621–638

63. Tardent, P. and Schmid, V. (1972) Ultrastructure of mechanor-eceptors of the polyp Coryne pintneri (Hydrozoa, Athecata). Exp.Cell Res. 72, 265–275

64. Arendt, D. et al. (2015) Gastric pouches and the mucociliary sole:setting the stage for nervous system evolution. Phil. Trans. R.Soc. B 370, 20150286

65. Horridge, G.A. (1969) Statocysts of medusae and evolution ofstereocilia. Tissue Cell 1, 341–353

66. Singla, C.L. (1983) Fine structure of the sensory receptors ofAglantha digitale (Hydromedusae: Trachylina). Cell Tissue Res.231, 415–425

67. Garm, A. et al. (2007) Visually guided obstacle avoidance in thebox jellyfish Tripedalia cystophora and Chiropsella bronzie. J.Exp. Biol. 210, 3616–3623

68. Garm, A. et al. (2011) Box jellyfish use terrestrial visual cues fornavigation. Curr. Biol. 21, 798–803

69. Satterlie, R.A. (2014) Multiple conducting systems in the cubo-medusa Carybdea marsupialis. Biol. Bull. 227, 274–284

70. Garm, A. et al. (2006) Rhopalia are integrated parts of the centralnervous system in box jellyfish. Cell Tissue Res. 325, 333–343

71. Garm, A. and Bielecki, J. (2008) Swim pacemakers in boxjellyfish are modulated by the visual input. J. Comp. Physiol.A 194, 641–651

72. Nilsson, D-E. (2013) Eye evolution and its functional basis. Vis.Neurosci. 30, 5–20

73. Mackie, G.O. and Meech, R.W. (2000) Central circuitry in thejellyfish Aglantha digitale. III. The rootlet and pacemaker systems.J. Exp. Biol. 203, 1797–1807

74. Mackie, G.O. et al. (2003) Central circuitry in the jellyfish Aglanthadigitale IV. Pathways coordinating feeding behaviour. J. Exp. Biol.206, 2487–2505

75. Mackie, G.O. et al. (1992) Giant axons and escape swimming inEuplokamis dunlapae (Ctenophora: Cydippida). Biol. Bull. 182,248–256

76. Mackie, G.O. (2004) Central neural circuitry in the jellyfish Aglan-tha. Neurosignals 13, 5–19

77. Roberts, A. and Mackie, G.O. (1980) The giant axon escapesystem of a hydrozoan medusa, Aglantha digitale. J. Exp. Biol.84, 303–318

78. Kerfoot, P.A. et al. (1985) Neuromuscular transmission in thejellyfish Aglantha digitale. J. Exp. Biol. 116, 1–25

79. Mackie, G.O. and Meech, R.W. (1985) Separate sodium andcalcium spikes in the same axon. Nature 313, 791–793

80. Katsuki, T. and Greenspan, R.J. (2013) Jellyfish nervous sys-tems. Curr. Biol. 23, R592–R594

81. Trembley, A. (1744) Mémoires pour servir à l’histoire d’un genrede polypes d’eau douce, à bras en forme de cornes. Mémoirespour servir à l’histoire d’un genre de polypes d’eau douce, à brasen forme de cornes, Jean and Herman Verbeek, (Leiden)

82. Anderson, P.A. (1990) Evolution of the First Nervous Systems,Plenum Press

83. Mackie, G.O. (1990) The elementary nervous system revisited.Am. Zool. 30, 907–920

84. Kremien, M. et al. (2013) Benefit of pulsation in soft corals. Proc.Natl. Acad. Sci. 110, 8978–8983

85. Hamlet, C. et al. (2011) A numerical study of the effects of bellpulsation dynamics and oral arms on the exchange currentsgenerated by the upside-down jellyfish Cassiopea xamachana.J. Exp. Biol. 214, 1911–1921

86. Santhanakrishnan, A. et al. (2012) Flow structure and transportcharacteristics of feeding and exchange currents generated byupside-down Cassiopea jellyfish. J. Exp. Biol. 215, 2369–2381

87. Labas, Y.A. et al. (1981) On pulsating growth in multicellularorganisms. Dokl. Akad. Nauk SSSR 257, 1247–1250

88. Beloussov, L.V. et al. (1993) Growth pulsations in hydroid polyps:kinematics, biological role and cytophysiology. InOscillations andMorphogenesis (Rensing, L., ed.), pp. 183–193, Marcel DekkerNew York

89. Kosevich, I.A. (2006) Mechanics of growth pulsations as thebasis of growth and morphogenesis in colonial hydroids. Russ.J. Dev. Biol. 37, 90–101

90. Werner, B. (1973) Spermatozeugmen und paarungsverhalten beiTripedalia cystophora (Cubomedusae). Mar. Biol. 18, 212–217

91. Lewis, C. and Long, T.A.F. (2005) Courtship and reproduction inCarybdea sivickisi (Cnidaria: Cubozoa). Mar. Biol. 147, 477–483

92. Wagner, G. (1905) Memoirs: on some movements and reactionsof Hydra. J. Cell Sci. 2, 585–622

93. Koizumi, O. (2016) Origin and evolution of the nervous systemconsidered from the diffuse nervous system of cnidarians. In TheCnidaria, Past, Present and Future, pp. 73–91, Springer

94. Browne, E.N. (1909) The production of new hydranths in hydraby the insertion of small grafts. J. Exp. Zool. 7, 1–23

95. Hager, G. and David, C.N. (1997) Pattern of differentiated nervecells in hydra is determined by precursor migration.Development124, 569–576

96. Technau, U. and Holstein, T.W. (1996) Phenotypic maturation ofneurons and continuous precursor migration in the formation ofthe peduncle nerve net in Hydra. Dev. Biol. 177, 599–615

97. Bode, H. et al. (1973) Quantitative analysis of cell types duringgrowth and morphogenesis in Hydra. Wilhelm Roux’Archiv fürEntwicklungsmechanik der Org. 171, 269–285

98. Nakanishi, N. et al. (2012) Nervous systems of the sea anemoneNematostella vectensis are generated by ectoderm and endo-derm and shaped by distinct mechanisms. Development 139,347–357

99. Layden, M.J. et al. (2012) Nematostella vectensis achaete-scutehomolog NvashA regulates embryonic ectodermal neurogenesisand represents an ancient component of the metazoan neuralspecification pathway. Development 139, 1013–1022

100. Richards, G.S. and Rentzsch, F. (2014) Transgenic analysis of aSoxB gene reveals neural progenitor cells in the cnidarian Nem-atostella vectensis. Development 141, 4681–4689

101. Watanabe, H. et al. (2014) Sequential actions of b-catenin andBmp pattern the oral nerve net in Nematostella vectensis. Nat.Commun. 5, http://dx.doi.org/10.1038/ncomms6536

102. Wittlieb, J. et al. (2006) Transgenic Hydra allow in vivo tracking ofindividual stem cells during morphogenesis. Proc. Natl. Acad.Sci. U.S.A. 103, 6208–6211

103. Renfer, E. et al. (2010) A muscle-specific transgenic reporter lineof the sea anemone, Nematostella vectensis. Proc. Natl. Acad.Sci. 107, 104–108

104. Künzel, T. et al. (2010) Migration and differentiation potential ofstem cells in the cnidarian Hydractinia analysed in eGFP-trans-genic animals and chimeras. Dev. Biol. 348, 120–129

105. Ikmi, A. et al. (2014) TALEN and CRISPR/Cas9-mediatedgenome editing in the early-branching metazoan Nematostellavectensis. Nat. Commun. 5, 5486

106. Kraus, Y. et al. (2016) Pre-bilaterian origin of the blastoporal axialorganizer. Nat. Commun. 7, http://dx.doi.org/10.1038/ncomms11694

107. Juliano, C.E. and Hobmayer, B. (2016) Meeting report on“Animal Evolution: New Perspectives From Early EmergingMetazoans”, Tutzing, September 14-17, 2015. BioEssays 38,216–219

108. Kortschak, R.D. et al. (2003) EST analysis of the cnidarianAcropora millepora reveals extensive gene loss and rapidsequence divergence in the model invertebrates. Curr. Biol.13, 2190–2195

109. Technau, U. et al. (2005) Maintenance of ancestral complexityand non-metazoan genes in two basal cnidarians. TRENDSGenet. 21, 633–639

110. Raible, F. and Arendt, D. (2004) Metazoan evolution:some animals are more equal than others. Curr. Biol. 14,R106–R108

111. Satterlie, R.A. and Eichinger, J.M. (2014) Organization of theectodermal nervous structures in jellyfish: scyphomedusae. Biol.Bull. 226, 29–40

112. Eichinger, J.M. and Satterlie, R.A. (2014) Organization of theectodermal nervous structures in medusae: cubomedusae. Biol.Bull. 226, 41–55

113. Holstein, T. (1981) The morphogenesis of nematocytes in Hydraand Forsklia: an ultrastructural study. J. Ultrastruct. Res. 75,276–290

114. Khalturin, K. et al. (2007) Transgenic stem cells inHydra reveal anearly evolutionary origin for key elements controlling self-renewaland differentiation. Dev. Biol. 309, 32–44

115. Hopfield, J.J. (1982) Neural networks and physical systems withemergent collective computational abilities. Proc. Natl. Acad. Sci.79, 2554–2558

116. Jekely, G. et al. (2015) An option space for early neural evolution.Phil. Trans. R. Soc. B 370, 20150181

117. Ivashkin, E. et al. (2014) A paradigm shift in neurobiology: periph-eral nerves deliver cellular material and control development.Zoology 117, 293–294

118. Adameyko, I. and Fried, K. (2016) The nervous system orches-trates and integrates craniofacial development: a review. Front.Physiol. 7, 49

119. Yano, J.M. et al. (2015) Indigenous bacteria from the gutmicrobiota regulate host serotonin biosynthesis. Cell 161,264–276

120. Williams, B.B. et al. (2014) Discovery and characterization of gutmicrobiota decarboxylases that can produce the neurotransmit-ter tryptamine. Cell Host Microbe 16, 495–503

121. Barrett, E. et al. (2012) g-Aminobutyric acid production by cul-turable bacteria from the human intestine. J. Appl. Microbiol. 113,411–417

122. Cryan, J.F. and Dinan, T.G. (2012) Mind-altering microorgan-isms: the impact of the gut microbiota on brain and behaviour.Nat. Rev. Neurosci. 13, 701–712

123. Forsythe, P. and Kunze, W.A. (2013) Voices from within: gutmicrobes and the CNS. Cell. Mol. life Sci. 70, 55–69

124. Mu, C. et al. (2016) Gutmicrobiota: the brain peacekeeper. Front.Microbiol. 7, 345

Trends in Neurosciences, February 2017, Vol. 40, No. 2 105