Biodiversity Meets Neuroscience: From the Sequencing Ship (Ship-Seq) to Deciphering Parallel Evolution of Neural Systems in Omic’s Era

Abstract

The origins of neural systems and centralized brains are one of the major transitions in evolution. These events might occur more than once over 570–600 million years. The convergent evolution of neural circuits is evident from a diversity of unique adaptive strategies implemented by ctenophores, cnidarians, acoels, molluscs, and basal deuterostomes. But, further integration of biodiversity research and neuroscience is required to decipher critical events leading to development of complex integrative and cognitive functions. Here, we outline reference species and interdisciplinary approaches in reconstructing the evolution of nervous systems. In the “omic” era, it is now possible to establish fully functional genomics laboratories aboard of oceanic ships and perform sequencing and real-time analyses of data at any oceanic location (named here as Ship-Seq). In doing so, fragile, rare, cryptic, and planktonic organisms, or even entire marine ecosystems, are becoming accessible directly to experimental and physiological analyses by modern analytical tools. Thus, we are now in a position to take full advantages from countless “experiments” Nature performed for us in the course of 3.5 billion years of biological evolution. Together with progress in computational and comparative genomics, evolutionary neuroscience, proteomic and developmental biology, a new surprising picture is emerging that reveals many ways of how nervous systems evolved. As a result, this symposium provides a unique opportunity to revisit old questions about the origins of biological complexity.

Introduction

“There are reasons to believe that behavioral shifts have been involved in most evolutionary innovations, hence the saying ‘behavior is the pacemaker of evolution’. Any behavior that turns out to be of evolutionary significance is likely to be reinforced by selection of genetic determinant of such behavior known as the “Baldwin effect”” (Mayr 2001) (p. 137 the emphasis added). The term behavior is not restricted to a myriad of instant reactions controlled by nervous systems; it is a well-tuned systemic response of any organism, from a bacterium to an elephant, which allows rapid and flexible adaptations within changeable environments, whether being a search for food or an attempt at escape. Not surprisingly, deep evolutionary roots of neuronal functions were recognized in the 19th century (Darwin 1872, 1880; Jennings 1906), before the neuronal doctrine was established by Santiago Ramón y Cajal (Bullock et al. 2005; Guillery 2005, 2007; Baluska and Mancuso 2009; Baluska et al. 2009).

First evolutionary scenarios explaining the origin of neural organization were initiated by Haeckel’s students in the 1870s and 1880 (Kleinenberg 1872; Claus 1878; Hertwig and Hertwig 1878, 1879, 1880; Chun 1880; Hertwig 1880)—all of them took experimental and conceptual advantages from enigmatic, then and now, marine organisms such as ctenophores, medusas, sponges, and various marine larval forms. Earlier hypotheses were further developed during the 20th century by Mackie (1990), Pantin (1956), Parker (1919), Passano (1963), and many other pioneers of the field (reviewed by Moroz [2014]).

The conceptual framework of these approaches was based on three major cornerstones: (i) the predominance of electrical conductive properties and excitability of neurons as major features of neuronal organization; (ii) the view of sponges (Porifera) as the earliest evolving animal lineage; and (iii) a little-disputed postulate that such complex structures as a nervous system or a brain could evolve only once. The widespread utility of a few model organisms in the context of “evodevo”, and first molecular phylogenetic trees in the 1990s and 2000s apparently hardened these paradigms in reconstructions of the evolution of nervous systems.

Virtually unchallenged, and often passively accepted, hypotheses of a single neural origin became commonplace. Some textbooks’ summaries are still written within a framework of a simplified Ladder of Nature (Scala Naturae), which illustrates the neuronal evolution mostly as an incremental, steadily ascending road, leading to a gradual rise of neuronal complexity from diffuse nerve nets (as in Hydra), to the centralized cord-like nervous system in a common bilaterian ancestor, followed by concentration of neuronal elements or ganglia into a complex brain. Not surprisingly, such terms as “primitive”/“advanced” or “lower”/“higher” nervous systems can still be found in literature.

Indeed, the origins of neural systems and complex brains are one of the major transitions in evolution (Maynard Smith and Szathmary 1995; Calcott and Sterelny 2011). The emerging comparative genomic data seriously challenge the hypothesis of monophyletic origin of nervous systems (Moroz et al. 2014; Moroz 2015a). The revised metazoan phylogeny (Figs. 1 and 2) apparently “removes” Porifera from its approximately 150 year-old, the most-basal position in the animal tree of life (Ryan et al. 2013; Moroz et al. 2014; Borowiec et al. 2015; Halanych 2015; Whelan et al. 2015a, 2015b). Moreover, advances in neuroscience transform the original perception of neurons as primarily electrical and conducting circuits (the paradigm rooted in the discovery of animal electricity by Galvani and Volta) into integrative chemical/secretory/genomic views of a neuron (Fig. 3).

Fig. 1.

Emerging reference species for comparative and evolutionary neuroscience. These are illustrative examples of species representing lineages of animals critical for the analysis of neuronal origins and parallel evolution of neuronal centralization. The term “reference species” is preferable to “model organisms” in this context (for details see Striedter et al. 2014). The phylum Ctenophora is represented by the cydippid Pleurobrachia bachei and the lobate Bolinopsis infundibulum (Friday Harbor, USA); Porifera—the larval stages of the demosponge Amphimedon queenslandica (Heron Island, Australia); Cnidaria—the scyphozoan Aurelia aurita (Sweden); Xenoturbellida—Xenoturbella bocki (Sweden); Nemertodermatida—Meara stichopi (Sweden); Mollusca—the gastropods Conus gloriamaris (the glory of the sea cone, Philippines) and Hermissenda crassicornis as well as cephalopods (octopus from Indonesia); Echinodermata—unidentified crinoid from the Great Barrier Reef (Heron Island, Australia); and Hemichordata—Saccoglossus bromophemolosus (Padilla Bay, USA). Photographs by L. L. Moroz. (This figure is available in black and white in print and in color at Integrative and Comparative Biology online.)

Fig. 2.

Genealogy of the animal phylogeny. (A) From Ernst Haeckel, Geneological Tree of Humanity, 1891; http://psnt.net/blog/2011/11/16/. (B) Modern view of relationships among animal phyla reconstructed using phylogenomic approaches. Several nodes (e.g., within Ecdysozoa and Lophotrochozoa) are not yet resolved. Solid dots represent nodes with high bootstrap support combined from various phylogenomics assays (see text). The clade Xenacoelomorpha might also be placed at the base of Bilateria. This tree is combined with the presence or absence of a CNS or brain. Examples of parallel centralization of neural systems and secondarily loss of nervous systems (the lighting signs) in obligatory parasites are shown. Abbreviations: B, brains or a brain-type of neural organization; G, ganglionic nervous systems; D, diffuse nervous systems; “Stop sign”, illustrates three lineages with a primary absence of neural organization. Porifera and Placozoa do not have recognized neurons. Ctenophores apparently evolved neurons and, possibly, muscles, independently from other groups of animals. Although neuronal organization in Cnidaria and Ctenophora can be superficially presented as a nerve net or diffuse nervous system, many species have a prominent concentration of neuronal elements, and numerous and apparently autonomous networks governing surprisingly complex and well-coordinated behaviors. For example, there are well defined concentrations of neural elements associated with locomotory combs, the aboral organ in Ctenophora, and rhopalia in Cnidaria. Choanoflagellates are placed at the base of the tree as a sister group for Metazoa. See text for details. (This figure is available in black and white in print and in color at Integrative and Comparative Biology online.)

Fig. 3.

Analytical approaches that reveal molecular portraits of neurons. The photograph shows one of the neurons from the CNS of Aplysia californica with a giant nucleus and glial nuclei stained by DAPI. The diversity of existing and emerging microanalytical approaches and technologies as well as their relationships are shown by arrows. (This figure is available in black and white in print and in color at Integrative and Comparative Biology online.)

Novel data also suggest that a neuron is not a genetic category tracing its ancestry to a single cell lineage in the common metazoan ancestor (urmetazoan). In contrast, a generalized term of “neuron” is a functional category with many neuronal phenotypes and neuronal circuits evolving in parallel across different lineages. The origins of neurons or neurogenesis in the broad sense might occur more than once over 570–600 million years of biological evolution (Sakharov 1974; Moroz 2009, 2012; Northcutt 2010; Pennisi 2013) and a hypothetical Precambrian urmetazoan (ancestor of all metazoans) might not contain neurons, or even what Parker (1919) famously called, protoneurons.

This issue brought several experts to emphasize critical events leading to the development of complex integrative and cognitive functions within the animal kingdom. All authors presented highly debated topics related to neural and animal evolution at the annual Society for Integrative and Comparative Biology meeting in West Palm Beach (January 3–7, 2015). The symposium highlighted several reference species (Fig. 1) and interdisciplinary approaches in reconstruction of the phylogeny of animals (Whelan et al. 2015a) and the evolution of nervous systems: (i) from numerically simpler larval receptive fields (Nakanishi et al. 2015) to animals with elementary intelligence (Yoshida et al. 2015) and mechanisms of behavioral decision making (Gillette 2015); (ii) from epitranscriptomics (Moroz 2015b; Kohn et al. 2015) to epigenomics (Dabe et al. 2015) and new adaptive strategies inspired by comparative models (Gillette 2015; Luttrell et al. 2015; Paulay 2015; Satterlie 2015) as well as synapse evolution. Together with advantages in computational and comparative genomics, evolutionary neuroscience, and proteomic and developmental biology a new surprising picture has been emerging—the picture that reveals many ways of how nervous systems evolved. As a result, this symposium provided a unique opportunity to set the stage and revisit the old questions about the origins of biological complexity.

The unified theme of the symposium, and the future of experimental biology, is understanding of unprecedented biological diversity. This view, atypical for the majority of neuroscientists, facilitates usage of numerous “non-conventional” organisms (Fig. 1) and opens new frontiers in analysis of neuronal innovations from microevolution/speciation to macro-evolutionary scale. Here, I briefly emphasize the needs to marriage biodiversity and neuroscience. “Study nature, not books”, this motto of Louis Agassiz can be considered as a practical guide for a new generation of experimental biologists. His words “Go to nature; take the facts in your hands; look, and see for yourself”. “The book of nature is always open” is the unbiased research strategy to bring novel reference species (Striedter et al. 2014) to the laboratory or, if needed, to bring the modern physiological/genomic mobile laboratory to the sea.

Phylogenomics meets neuroscience

The timing of the origin of animals is unknown: the event apparently was triggered by several global glaciations (Snowball Earth) at the end of the Cryogenian period (850–635 million years ago or Mya) and the break-up of the supercontinent Rodinia during Ediacaran times (635–542 Mya) (Fedonkin et al. 2007). The first proposed metazoan fossils are dated about 600 Mya, but the famous Ediacaran biota (580–545) cannot be reliably traced to any extant animal lineages (Valentine 2004; Fedonkin et al. 2007; Erwin and Valentine 2013). The fossils with undisputed affinities to modern phyla date many deep branches of the animal tree to an evolutionary radiation in the early Cambrian—the “Cambrian Explosion”, with possible roots in the Ediacaran period (Erwin and Valentine 2013). The late Ediacaran (∼570–550 Mya) might also be the time when the first neural-like, or elementary integrative systems might have evolved to cost-efficiently coordinate ciliated locomotion, burrowing into sediments (trace fossils) and possible defensive injury-related responses in early urmetazoa (Moroz 2009; Walters and Moroz 2009). The Cambrian Explosion with its novel global ecology and the appearance of large predators could trigger arms races and the formation of skeletons. Combined with the raise of oxygen in the atmosphere, diverse Cambrian ecosystems could further drive the profound diversification of sensory, neural, and motor systems, leading to the establishing of major types of neuronal organizations by the end of Cambrian. It was a relatively short geological interval (10–30 million years) when the radiation of key animal body plans and neural/integrative systems occurred. The Cambrian Explosion led to more than 500 million years of parallel evolution of established neural architectures in dozens of animal lineages. Zoologists are still debating the exact number of monophyletic animal superclades, but the existence of about 35 distinct animal phyla is widely recognized (Kozloff 1990; Nielsen 2012; Dunn et al. 2014).

Rephrasing the famous motto of biology (Dobzhansky 1973), we can say “Nothing in Neuroscience make sense, except in the light of circuits & behavior”. Unfortunately, representatives of only five animal phyla (marked as bold text in Fig. 2B) are systematically studied in terms of their functional neuronal circuits, interneuronal signaling, and behaviors. In addition to vertebrates (Chordata), reasonably well-characterized neural circuits controlling stereotyped and learned behaviors are identified in several species of crustaceans (e.g., stomatogastric ganglia in crabs and lobsters) and in some insects, in a few gastropod molluscs (e.g., Aplysia, Lymnaea, Pleurobranchaea, Tritonia—see Gillette in this issue), in the nematode Caenorhabditis elegans, and in one annelid (the medicinal leech [Kristan et al. 2005; Calabrese et al. 2011; Mullins et al. 2011]). Less is known about cnidarians (see the review in this issue by Satterlie [2015]). The neural circuits within representatives of other 25+ animal phyla are still a terra incognita for neuroscientists. This is reflected in a number of conflicting hypotheses about neuronal origin(s) and evolution. Thus, the three most critical bottlenecks today are (i) the lack of molecular/genomic data about the cellular/physiological organization of neural circuits, (ii) paucity of knowledge about neurogenesis from the majority of phyla, and (iii) lack of unbiased classification of neurons, including the use of single-cell RNA-seq and epigenomic profiling.

It took more than one decade, when advantages of expressed sequence tags (established by JC Venter’s team in 1992–1993) (Adams et al. 1992, 1993a, 1993b; McCombie et al. 1992) started to be introduced into studies of biodiversity and evo-devo. However, even the first deep transcriptome sequencing (RNA-seq) from representatives of a dozen phyla uncovered unexpected relationships across Metazoa (Halanych 2004; Bourlat et al. 2006; Moroz et al. 2006; Kocot et al. 2011; Philippe et al. 2011; Smith et al. 2011; Dunn et al. 2014; Whelan et al. 2015b). Figure 2 summarizes the current view of the animal tree of life combined with the presence of diffuse versus centralized neural systems. Although some nodes remain unresolved (and most RNA-seq data were not obtained from neural tissues), this tree challenges previous paradigms about both animal phylogeny and neuronal evolution.

First the morphologically and behaviorally complex ctenophores (with true neurons, muscles, and mesoderm) are now placed as the lineage sister to all Metazoa. In contrast, the nerveless Porifera and Placozoa, morphologically simpler animals without muscles and mesoderm, are now “upgraded” to the second and third sequential splits in the branching of the animal tree of life (Moroz et al. 2014; Whelan et al. 2015a, 2015b). Cnidarians hold their position as the sister group to all bilaterians. Cnidarian might also have cryptic bilaterian symmetry as evidenced from the expression of several genes controlling polarity in animals; also, they shared many developmental toolkits and even chromosome synteny with bilaterians (Putnam et al. 2007).

The novel animal-tree topology, combined with the unique molecular/genomic organization of ctenophores, suggests that neurons as well as muscles and mesoderm are evolved independently in Ctenophora and the bilaterian/cnidarian clade, from the last common metazoan ancestor without a nervous system (Moroz 2014, 2015a; Moroz et al. 2014). The sponges and placozoans are morphologically (but not molecularly!) simpler animals, but it is highly unlikely that these lineages of free-living organisms secondarily lost their nervous systems as suggested (Rokas 2013). Two presentations at this symposium discuss these alternative hypotheses (Moroz 2015b; A. B. Kohn and L. L. Moroz, submitted for publication).

The discovery of majority, canonical, body-patterning genes cell adhesion and signal molecules both in many basal metazoans and in unicellular eukaryotes initially surprised researchers. However, Mikhailov et al. (2009) revisited earlier evolutionary hypotheses and presented a scenario of the origin of animals that is supported by genomic data. They suggested that the observed temporal differentiation, i.e., the presence of multiple types of cells in unicellular eukaryotes with complex life cycles, had been transformed into spatial differentiation in colonial organisms and multicellular animals—the event requiring relatively small changes in their overall gene complement.

The presence of apparently “neuronal-like” and “synaptic-like” genes cannot be considered as the evidence for the presence of neural systems in the last common ancestor for two major reasons: (i) these genes are not truly pan-neuronal or pan-synaptic and they are widely expressed outside of neural systems, e.g., see discussion by A. B. Kohn and L. L. Moroz (submitted for publication) and Norekian and Moroz (2015), and (ii) most of these genes are components of the ancestral eukaryotic machinery for secretion/exocytosis, cell polarity, excitability, and reception. Not surprisingly, such machinery can be independently recruited for more spatially and temporally localized neuronal signaling in a variety of animal lineages including ctenophores and bilaterians (Moroz 2015b). Toward this goal, at the symposium Nakanishi et al. (2015) presented novel data about the molecular dissection of complex sensory, secretory, and cell-type specification in sponges.

Second, parallel centralization events might have occurred more than once, suggesting that the last common bilaterian ancestor had no central nervous systems (CNS) as implicated previously (Moroz 2009). It is also unlikely that the last common ancestor of deuterostomes had a CNS (Lowe et al. 2003, 2015). Furthermore, we can now consider the existence of more than 10 independent centralization events in the evolution of the nervous system (Moroz 2012). Additional cases were also discussed at the meeting. For example, various adaptive strategies led to the formation of rhopalia and concentration of the nerve net in Cnidaria (Satterlie, 2015) or centralization of neural-like cords in basal deuterostomes (Luttrell et al. 2015) and Acoela/Xenacoelomorpha (Perea-Atienza et al. 2015).

Interestingly, even in the single phylum Mollusca, we can reconstruct at least five independent events of centralization of the nervous system (Moroz 2009), including the lineage leading to modern cephalopods. These “primates of the sea” (Akimushkin 1963) have achieved the level of neuronal complexity comparable to that of rodents and birds (see also Yoshida et al., 2015). The number of neurons in the Octopus nervous system can reach approximately 500,000,000 (Hochner 2013). There are three major neural innovations in cephalopods: (1) the vertical lobe (the memory center and an analog of the hippocampus), (2) the optic lobes forming enormous visual centers, and (3) arm cords with more than 300,000,000 neurons controlling very complex mechanics of the arms and suckers—far exceeding in its performance any known robotic system. In contrast, it appears that Phoronida, Brachiopoda (Temereva and Tsitrin 2014, 2015), and most of the basal metazoans such as Xenoturbellida and Nemertodermatida, might still “preserve” uncentralized nervous systems from a common animal ancestor.

Third, there is a secondary loss of nervous systems in three obligatory parasitic lineages of animals: (1) Myxozoa—parasitic cnidarians (Smothers et al 1994; Kent et al. 2001; Jimenez-Guri et al. 2007; Yang et al. 2014; Foox and Siddall 2015); (2) Dicyemida or Rhombozoa (parasites of cephalopods) (Czaker 2006, 2011; Ogino et al. 2011); and (3) Orthonectida (parasites of echinoderms) (Sliusarev 2008). The last two lineages (known as Mesozoa) apparently belong to the Spiralia/Lophotrochozoa clade (Hanelt et al. 1996; Pawlowski et al. 1996; Aleshin et al. 1999; Suzuki et al. 2010; Dunn et al. 2014). It is very likely that both Dicyemida and Orthonectida secondarily lost many characters as a result of adaptations to obligatory parasitic lifestyles within their hosts (Aleshin and Petrov 2002). In parasitic rhizocephalan crustaceans, there is also the loss of nervous systems in adults but canonical neural organization is present in developmental and larval stages (Høeg 1987). However, there is no reported case of a nervous system being lost in any free-living metazoan lineages.

On the other hand, we are still far away from a comprehensive view and understanding of existing diversities among various classes of neuronal organization, simply because we have barely touched the tip of the iceberg of animal biodiversity.

Scope of marine biodiversity is largely unknown

Life, eukaryotes, plants, and animals all arose in the sea, and the oceans hold most of the deep diversity of life, from prokaryotes to virtually unknown diversity of unicellular and colonial eukaryotes to invertebrate animals. Before the famous Challenger oceanographic expedition (1872–1876) cataloged more than 4000 unknown species, some scientists thought that no life existed below 600 m. Today, there are approximately 230,000 known species of marine invertebrates, and although marine life is substantially underdescribed (Appeltans et al. 2012), it is nevertheless much more taxonomically diverse than life on land. Thus, surveying marine life is the most effective way of addressing questions about animal and neuronal evolution. Yet, the oceans remain remarkably underexplored. Recent estimates suggest that up to 91% of all marine species have never been even named, including about 2 million species of animals (Mora et al. 2011). These estimates have been supported by the most recent genomic surveys, suggesting that even in the well-studied costal marine ecosystems of the USA (Virginia and Florida) only approximately 11% of invertebrate species could be matched to reference barcodes in public databases (Leray and Knowlton 2015). Comparable estimates were obtained using 18S ribosomal sampling of the eukaryotic plankton biodiversity during the circumglobal Tara oceanic expedition in 2009–2013 (de Vargas et al. 2015).

There are just three illustrative examples of how little we know about oceanic biodiversity (highlighted in the presentation of G. Paulay at this symposium). Prochlorococcus, the most abundant autotroph on Earth that produces approximately 50% of global O₂, was only discovered in 1986. The four most recently described animal phyla were all discovered near the cost of Scandinavia, one of the most depauperate parts of the ocean, thereby demonstrating that effort rather than diversity is limiting our knowledge of the sea. The enigmatic Dendrogramma, possibly representing an entirely new superclade of basal metazoans (in addition to five others known), was discovered in 2014 from the Sydney area (Just et al. 2014). Again it came from the reasonably well-sampled ecosystem.

The challenge and cost of working at sea and in remote areas, in contrast to focusing on the more visible and accessible terrestrial biosphere, have curtailed our understanding of the deep diversity, history, and functional biology of life. Most importantly, for marine biologists and neuroscientists in particular, is a possibility to perform experimental tests on living marine animals in their natural habitats. Today, such research is only possible at coastal marine stations and laboratories. However, many marine organisms are pelagic, seasonal, rare, or too fragile to transfer to land bases for physiological and molecular analyses. This challenge can be resolved using mobile laboratories for functional genomics placed on board oceanic ships. Some pioneering work has already been initiated using this approach.

Sequencing at sea (Ship-Seq) as the first step to closing the gap between biodiversity and neuroscience

Thousands of investigators have studied sea creatures at remote locations for years, often finding their work stalled by specimens that declined so quickly after collection that they were deteriorated by the time they reached laboratories. Such complex field logistics prevented performance of so-needed real-time physiological measurements and sophisticated molecular analyses. Occasionally, highly valuable organisms were lost in transition—and the delay in examination and research is often frustrating. A perfect solution for many comparative genomic studies, and neurobiological research in particular, is to take the modern, ideally genomic, laboratory to the sea. The idea is simple but the execution is complicated.

Craig Venter was the first person who used the advantages of modern genomics to study marine biodiversity. During the initial survey of the Sargasso Sea (Venter et al. 2004) and world oceans (Rusch et al. 2007; Yooseph et al. 2007), Venter and his team focused on the collection of viral and microbial samples that were then shipped to land-based laboratories for their processing and sequencing. Similar approach was used by the team of Tara Oceans during their circumglobal expedition (2009–2013) focusing on the plankton biodiversity. Again, as it was done by the Venter team, planktonic viral, bacteria, and small eukaryotic samples from diverse oceanic locations were preserved for later sequencing on land; and primarily ribosomal RNAs were used to estimate a scope of biodiversity within the photic zone (Brum et al. 2015; de Vargas et al. 2015; Lima-Mendez et al. 2015; Sunagawa et al. 2015; Villar et al. 2015).

The development of compact and powerful sequencers opens a new era in marine biology and comparative neuroscience. As an example of the proof-of-concept reported at the symposium, we designed a laboratory for physiological genomics onboard an oceanic 141-foot yacht Copasetic turned into a research vessel (Fig. 4). Specifically, we used a regular shipping container and converted it into a mobile laboratory equipped with microscopes and instrumentation, thereby allowing us to perform all steps from dissections of animals and single-cells, isolation of RNA and DNA, construction of next-generation sequencing libraries, sequencing, and bioinformatic analysis on site and in real-time. To deal with the ship’s movement and potential changes in humidity and temperature, we used Ion Torrent semiconductor sequencer (Life Technologies, now ThermoFisher). We performed two trips from Fort Lauderdale, Florida to the Bahamas and Florida Keys in January–February and April 2014, testing both the equipment and the sequencing pipeline. As a result, within 10 days, we processed samples from more than two hundreds of planktonic and benthic invertebrates, including echinoderm, molluscan, and hemichordate larvae as well as Gulf Stream ctenophores, and performed 22 sequencing cycles onboard. The sequencer server was linked by satellite to the University of Florida high-performance supercomputer (HiPerGator) for online assembly and annotation of the obtained data with a rapid feedback to the mobile laboratory (Ship-Seq). The work allowed us to collect and sample rare tornaria larvae and extremely fragile ctenophores as well as perform experiments on the regeneration of ctenophores (L. Moroz, A. Kohn, E. Dabe, R. Sanford, G. Winter, and G. Paulay, manuscripts in preparation).

Fig. 4.

The sequencing aboard the oceanic ship Copasetic or Ship-Seq in January–April 2014. In the mobile genome laboratory on board the ship sequencing data can be uploaded and analyzed in real time using satellite connections to a land-based supercomputer (see text for details). (This figure is available in black and white in print and in color at Integrative and Comparative Biology online.)

Of course, these are initial pilot studies, but they open novel opportunities both to study biodiversity and to integrate it with real-time physiological genomics at any remote location around the globe. Advances in separation technologies and single-cell high-throughput sequencing (Klein et al. 2015; Macosko et al. 2015), combined with access to a broad diversity of organisms, are the emerging frontiers in evolutionary neuroscience.

It is not difficult to foresee in the nearest future, sampling hundreds of reference species, their nervous systems, and even individual neurons in selected neural circuits, from representatives of all extant lineages of animals. The long-term goal would be unbiased classification of neurons (neurosystematics) and, ultimately, the reconstruction of the genealogy of neurons across phyla. It would be possible to reevaluate various hypotheses of neuronal evolution and decipher novel principles or constraints in the organization of neural circuits and behaviors, including examples of parallel and convergent evolution.

Studying different organisms and their adaptations to different environments, especially examples of convergent evolution, is important because it could reveal a variety of evolutionary solutions to perform similar cellular and systemic functions, like learning and memory mechanisms. Indeed, “Nature” has already performed countless numbers of experiments for us. We might consider them as perfectly tuned, and validated by natural selection, knock-out experiments accessible for experimental analysis. There are animals such as closely related aplysiid species with different capabilities for learning (Wright 1998; Erixon et al. 1999; Marinesco et al. 2003; Hoover et al. 2006). It is also known that different neurons learn differently, but also different neurons age differently (Moroz and Kohn 2010), and they might have already discovered “Fountains of Youth”. Some animals can efficiently heal wounds and regenerate integrative centers such as aboral organs in ctenophores (Mnemiopsis and Bolinopsis); some don’t (Pleurobrachia and Beroe).

So “Nature” provides us with at least 2 millions of enormously efficient experiments already carried out on marine animals. In fact, life on this planet is one big experiment, 3.5 billion years in the making. The advantage of functional (neuro)genomics is that we can obtain the genetic blueprint of virtually any creature in the sea to study these experiments.

There is no time to lose. Human activity dramatically changes pristine ecosystems around the globe, potentially triggering the sixth major extinction. Recent estimates suggest that, at average, one species is lost every 6 h. Thus, combining both research and private enterprises, it must be a “Race to save the species”.

Conclusion

Ernst Mayer, one of the founders of the modern evolutionary synthesis, made this comments about “behavior as the pacemaker of evolution” at the time when the first draft of the human genome became available (Lander et al. 2001; Venter et al. 2001)—the event launching a new scientific revolution in the 21st century. Animals display an enormous variety of forms of neuronal organization and behaviors—most of them still unknown and under-investigated. In combination with integrated “omic” approaches, such biodiversity offers unique opportunities to reconstruct the origins of biological innovations, including the genomic bases of origin and parallel evolution of neural circuits and brains. However, the study of “non-model” animals is nearly 10–20 years behind that of so-called model organisms. Novel technologies, as represented by single-cell, next-generation sequencing, direct sequencing onboard at any oceanic locations (Ship-Seq) complemented by super-resolution imaging, ultrasensitive single-cell metabolomics, and proteomics, are opening new frontiers to unbiased profile cell lineages in all extant metazoan phyla within a decade of focused research. Thus, the existing gaps in our knowledge about the organization of animals can be filled at an unprecedented rate (Bracken-Grissom et al. 2014; Striedter et al. 2014).

Funding

This work was supported by the National Science Foundation [grant numbers NSF-0744649, IOS-1457162, NSF CNS-0821622, and IOS 1457162]; National Institute of Health [grant numbers 1R01GM097502, R01MH097062]; National Aeronautics and Space Administration (NASA) [grant number NNX13AJ31G]; and McKnight Brain Research and University of Florida Opportunity Funds. The Society for Integrative and Comparative Biology provided assistance for attending the symposium at which this article was presented.