Echinoderm development and evolution in the post-genomic era

Developmental Biology (xxxx) xxxx–xxxx

Contents lists available at ScienceDirect

Developmental Biology journal homepage: www.elsevier.com/locate/developmentalbiology

Review article

Echinoderm development and evolution in the post-genomic era ⁎

Gregory A. Cary, Veronica F. Hinman

Department of Biological Sciences, Carnegie Mellon University, Mellon Institute, 4400 Fifth Ave, Pittsburgh, PA 15213, United States

A R T I C L E I N F O

A BS T RAC T

Keywords: Echinoderm GRN Evolution Evo-Devo Strongylocentrotus purpuratus Patiria miniata Sea urchin

The highly recognizable animals within the phylum Echinodermata encompass an enormous disparity of adult and larval body plans. The extensive knowledge of sea urchin development has culminated in the description of the exquisitely detailed gene regulatory network (GRN) that governs the specification of various embryonic territories. This information provides a unique opportunity for comparative studies in other echinoderm taxa to understand the evolution and developmental mechanisms underlying body plan change. This review focuses on recent work that has utilized new genomic resources and systems-level experiments to address questions of evolutionary developmental biology. In particular, we synthesize the results of several recent studies from various echinoderm classes that have explored the development and evolution of the larval skeleton, which is a major feature that distinguishes the two predominant larval subtypes within the Phylum. We specifically examine the ways in which GRNs can evolve, either through cis regulatory and/or protein-level changes in transcription factors. We also examine recent work comparing evolution across shorter time scales that occur within and between species of sea urchin, and highlight the kinds of questions that can be addressed by these comparisons. The advent of new genomic and transcriptomic datasets in additional species from all classes of echinoderm will continue to empower the use of these taxa for evolutionary developmental studies.

1. Introduction The sea urchin has been a potent model system for developmental biologists and biochemists for over a century, producing key insights into fundamental processes such as fertilization (Santella et al., 2012), cell cycle control (Yanagida, 2014; Dorée and Hunt, 2002), embryonic patterning (Molina et al., 2013; Annunziata et al., 2014; Angerer et al., 2011), including the regulative nature of early development (Angerer and Angerer, 1999), and the complex character of cis-regulatory control sequences (Yuh et al., 2004, 2001; Ransick and Davidson, 2006; Davidson, 1999). A key contribution of sea urchin research over the last several decades has been to describe early development using a hierarchical network of regulatory genes. Such gene regulatory networks (GRNs) explain the regulatory interactions that control successive stages of specification and differentiation (Davidson et al., 2002a, 2002b; Oliveri et al., 2008; Andrikou et al., 2015; Saunders and McClay, 2014). The GRN governing sea urchin embryogenesis is the most complete such network described to date, and portions of the network have reached a level of completeness allowing the generation of a computational boolean model in which in silico perturbations can predict known experimental outcomes (Peter et al., 2012). The characteristics that make sea urchins an attractive model to developmental biologists - namely the ease of acquiring large quantities

⁎

of eggs and synchronized developing embryos, visual transparency, and ease of manipulation - are shared by many other species in the phylum Echinodermata. Given the experimental accessibility of these species and the relative strength of the sea urchin model, recent work has leveraged the intricate knowledge of sea urchin developmental regulatory interactions as a basis for evolutionary comparisons within this phylum. This review focuses, in particular, on how recent genomic data has enabled systems-level inquiries from these species which have lead the way in our understanding of evolution of GRN for development. 2. Echinoderms: a rich disparity of body plans Echinoderms belong within the grouping of deuterostome animals, which only includes two other phyla; Chordata and Hemichordata. The echinoderms and hemichordates are further grouped together as the Ambulacraria (Fig. 1). There is a rich fossil record of echinoderms, which informs our current understanding of the evolution of this phylum. The earliest echinoderms are thought to have emerged in the Cambrian around 530-524 MYA (Smith, 1988). Crinoids, both stalked (sea lilies) and unstalked (feather stars), most likely diverged from the other echinoderm classes between 485 and 515 MYA (Rouse et al., 2013), however there are few living examples, and these tend to be found in deep water making them problematic for detailed functional

Corresponding author. E-mail address: [email protected] (V.F. Hinman).

http://dx.doi.org/10.1016/j.ydbio.2017.02.003 Received 12 September 2016; Received in revised form 4 February 2017; Accepted 6 February 2017 0012-1606/ © 2017 Elsevier Inc. All rights reserved.

Please cite this article as: Cary, G.A., Developmental Biology (2017), http://dx.doi.org/10.1016/j.ydbio.2017.02.003

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

Fig. 1. Phylogenetic relationships and example adult and larval morphologies within the Ambulacraria. Asterozoan topology, the consensus view of relationships within the echinodermata, is highlighted. Branch lengths are not drawn to scale. The images presented do not, in all cases, correspond to the example species cited. Photo credits: Adult Euechinoidea and Cidaroid are © Ann Cutting, Caltech; Holothuroidea is © Richard Ling/www.rling.com; Asteroidea is © Jerry Kirkhart, Los Osos, CA; Ophiuroidea is © Hans Hillewaert; Crinoidea is © NOAA Okeanos Explorer Program, INDEX-SATAL 2010; and Hemichordata is © Moorea Biocode / calphotos.berkeley.edu 4444 4444 0513 0997. Cidaroidea larval image is adapted from Bennett et al. (2012), all other whole (SEM) images of echinoderm and tornaria larvae are © T.C. Lacalli and T.H.J. Gilmour (University of Saskatchewan n. d.).

cucumbers form dipleurula like larval form (called a bipinnaria in sea stars and auricularia in sea cucumbers) (Nakano et al., 2003; Byrne et al., 2007), and while the larva of the sea lily is non-feeding, it nonetheless also forms a dipleurula larva (Nakano et al., 2003; Fig. 1). This larva-type is characterized by having two loops of ciliary bands that transverse the ectoderm. The similarity of this larval type to the tornaria larvae of hemichordates suggests that this larval form is basal among the Echinoderms, and possibly also Ambulacraria and hence more broadly the entire deuterostome clade (Cannon et al., 2014; Cameron et al., 2000). Sea urchins and brittle stars, by contrast, have a pluteus larva (echinopluteus in sea urchins and ophiopluteus in brittle stars) that have a single ciliary band around the oral ectoderm and a large, dominant larval skeleton. It is the larval skeleton that gives the plutei their striking armed phenotype. As sea urchins and brittle stars are not sister taxa, the origin of these larval types is difficult to resolve, but for now the most parsimonious explanation is that the plutei evolved independently from dipleurula larval forms within the lineages leading to the Echiniodea and Ophiuroidea (Morino et al., 2016). This disparity comprises a rich natural source of large scale body plan changes that permit investigation of deep-time divergence of dramatic body plan evolution. This, coupled with the extraordinary analyses of GRNs in sea urchins, provides an unparalleled potential to understand how GRNs have evolved for such developmental change. Additionally, smaller scale comparisons made within populations or

studies. The other four classes of echinoderms form a clear grouping known as the Eleutherozoa, which recent phylogenomics suggests separated into the four classes within a 5 Myr window around 480 MYA (Pisani et al., 2012; Telford et al., 2014). This rapid, ancient radiation has made it difficult to establish the relationship between the classes. Recently however, extensive genomic information has lead to the congruence of the grouping sea stars and brittle stars into one clade, called the Asterozoa and sea urchins and sea cucumbers to another, termed the Echinozoa (Telford et al., 2014; Reich et al., 2015). Within the echinoids, there are two broad taxa of sea urchin, the Cidariodea (pencil urchins, e.g. Eucidaris tribolidea), and Euechinoidea which comprises the thin spined sea urchins represented by well known model species (e.g. S. purpuratus, Lytechinus variegatus, Paracentrotus lividus) as well as the lesser studied sand dollars (e.g. Peronella japonica). It is apparent, even to someone with just a cursory knowledge of these animals, that there is an extraordinary body plan diversity among the classes of adult echinoderms. In contrast, however there has been little intraclass deviation in body plans in the almost 500 million years since their origin. All echinoderms also develop through a larval stage, which ancestrally was a feeding, planktotrophic larva, but has repeatedly and independently evolved in close sister taxa to non-feeding lecithotrophic forms (Puritz et al., 2012; Raff and Byrne, 2006). The feeding larval forms are also highly disparate between the classes. Sea stars and sea 2

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

echinoderms has shown that, for the most part, these genes are expressed in partially overlapping anterior-posterior domains including in S. purpuratus (e.g. hox7 and hox11/13b) (Howard-Ashby et al., 2006b), the sea cucumber A. japonicus (Kikuchi et al., 2015), and larval sea lily M. rotundus (Hara et al., 2006). One notable exception was found in the direct developing sand dollar P. japonica, where hox genes were found to be expressed in radial domains around the juvenile mouth, but linearly within the somatocoel (Tsuchimoto and Yamaguchi, 2014). This is reminiscent of observations made in S. purpuratus, where hox genes were found in a linear arrangement in the larval somatocoel that then asymmetrically changes as the adult rudiment matures (Arenas-Mena et al., 2000). Ultimately, pentameral symmetry is likely not a result of a phylumspecific hox cluster rearrangement as intact hox clusters have recently been described in the genome of the crown-of-thorns sea star A. planci (Baughman et al., 2014; Fig. 2). Furthermore, in both A. planci and the bat star P. miniata (Annunziata et al., 2013) the ParaHox genes are also found to be clustered in the genome, which contrasts to the lack of genomic arrangement of the S. purpuratus orthologs. These results refute the hypothesis that the evolution of the echinoderm pentameral body plan was associated with hox gene translocation, but do suggest significant rearrangement has occurred sometime following the split between Asterozoa and Echinozoa. Thus, the accumulation of genome data from species of Echinodermata and Hemichordata have asserted the phylogenetic grouping of these organisms within the deuterostomes, while highlighting specific gene families that have expanded or contracted with respect to other phyla (e.g. toll-like receptors in the sea urchin genome (Rast et al., 2006)). The interpretation of conserved synteny and syntenic breaks is more ambiguous; microsynteny, in particular, may indicate the presence of noncoding regulatory sequences in the proximity of genes within the syntenic block, however this has yet to be demonstrated in this phylum. Nonetheless, the availability of these genomes has greatly facilitated comparative EvoDevo research in Echinoderms. As we discuss more below, genome-level comparisons of gene expression patterns and cis regulatory control have enabled more directed comparisons of GRN topology and the dynamics of evolutionary change. These comparisons have been made across many scales; from closely related species to different classes of echinoderms, and even across phyla. These studies have compared the ways in which nodes (the genes) themselves evolve, the ways in which the edges changes (i.e. the regulation of one gene by another), how the topology of subcircuits can change, and the dynamics of how the GRN might change. These works collectively serve to underscore the inherent variability gene expression dynamics, and have hinted at a key feature of GRNs being robustness to this variability.

between closely related species with minimal morphological differences inform the capacity for developmental programs to overcome inherent variability in gene expression or cis-regulatory sequence to achieve very similar developmental outcomes. GRN comparisons are by their very nature enabled by high throughput, genome-based approaches, which therefore requires at least some level of assembled genome resource. We next therefore present a current examination of the efforts to accumulate assembled genome data from these species and follow on our review to examine how genome resources have facilitated recent comparative GRN studies. 3. Evolution of echinoderm genomes In the last decade numerous Echinoderm genomes have been sequenced, assembled and annotated to varying degrees of completeness (Sea Urchin Genome Sequencing Consortium et al., 2006; Cameron et al., 2015; Long et al., 2016; Baughman et al., 2014). The first analyses of these genomes revealed the greater homology between S. purpuratus and vertebrate gene sets, than between either taxon and protostomes, in keeping with the phylogenetic position of echinoderms among the Deuterostoma (Materna et al., 2006; Simakov et al., 2015). Furthermore, the protein domains identified among the genes present underscored similarities to vertebrate genomes; for example, the differences in gene content between deuterostome phyla were primarily expansions and contractions of gene families rather than innovations or loss of gene classes (Sea Urchin Genome Sequencing Consortium et al., 2006). This is especially important given the observation that transcription factor gene families are present and highly utilized during early development (Howard-Ashby et al., 2006b, 2006a; Tu et al., 2006). Recent descriptions of two hemichordate genomes (Simakov et al., 2015) provides the basis for useful inter-phylum genome comparisons within the Ambulacraria. This work again supports the phylogenetic grouping of echinoderms and hemichordates among the deuterostomes, across a variety of phylogenetic measures (protein similarity, intron structure, and coding indels). These genomes also highlight the extent of synteny among deuterostomes. For example, the “pharyngeal cluster”, which includes nkx2.1, nkx2.2, pax1/9, foxA, slc25A21 and mipol1 is present in both of the hemichordate genomes, all vertebrates, and in Acanthaster planci, the crown-of-thorns sea star (Simakov et al., 2015). Also, the tight genetic linkage between univin and bmp2/ 4 described in S. purpuratus (Range et al., 2007) is also observed in these hemichordate genomes. While the micro-synteny identified among these genomes may be a result of low rates of genomic rearrangement, it also highlights the potential for conserved regulatory interactions wherein noncoding sequences that regulate a gene are located proximal to neighboring genes (Irimia et al., 2012). Therefore, broken syntenic relationships in one clade relative to the others may indicate alterations to the regulatory architecture operating around those genes. One early revelation from the efforts to sequence the S. purpuratus genome was the odd arrangement of genes within the hox cluster (Cameron et al., 2006; Fig. 2). Genomic order and transcriptional orientation of hox cluster genes in the hemichordate genomes is largely conserved with chordates (Freeman et al., 2012; Fig. 2). Genes within the hox cluster are broadly utilized within bilaterians for developmental anterior-posterior axis patterning (Finnerty, 2003) with the remarkable feature of spatio-temporal colinearity of expression relative to genomic position (Kmita and Duboule, 2003). Early expression studies in sea urchins suggested that homeobox gene expression patterns did not appear to obey the gene expression profile rules emerging from other species, and therefore might account for evolution of the derived radial body plan (Lowe and Wray, 1997; Long and Byrne, 2001). The observation that the topology of this cluster is broken in the sea urchin genome was a potential explanation of the evolution of this feature. However, additional studies of hox gene expression patterns among

4. Fine-scale evolution within the Echinoidea: variability of gene expression and morphological consequences While the greater focus of comparative studies have been of the large scale comparisons across classes, there is an increasing emphasis on examinations of smaller scale, within-class comparisons. These fine scale comparisons of morphological etiology range from the variation within a single species to larger variations related to drastic changes in larval ecological strategies across species. These studies have highlighted the importance of variation within populations, and species level gene expression and how this might relate to morphological, genetic and/or environmental changes. Garfield et al. (2013) studied population-level variation in expression of genes within the well-studied primary mesenchyme cell (PMC) GRN in S. purpuratus. They found that expression of regulatory genes is variable within populations and that this variability could be correlated with changes to larval skeleton morphology by parent-oforigin effects. By overlaying expression variation with GRN topology they were able to show that the early GRN buffers this variation, while 3

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

Fig. 2. Hox cluster organisation across the invertebrate deuterostomes. The topology and organisation of identified Hox clusters for two echinoderms (S. purpuratus and A. planci), one hemichordate (S. kowalevskii), and one chordate (B. floridae) are depicted. Genes and intergenic distance is not drawn to scale. For each species, the order and orientation of genes within the cluster are indicated as is the relative distribution of anterior hox genes (purple), medial hox genes (blue), and posterior hox genes (red). Photo credits: S. purpuratus © Ann Cutting, Caltech; A. planci © user Rore bzh, Wikimedia Commons; S. kowalevskii © user Necrophorus on de.wikipedia; B. floridae © user Parent Géry, Wikimedia Commons.

specification and differentiation. These studies highlight the surprising degree to which gene expression levels can change while still preserving GRN topology and morphology. In each case, conservation of regulatory gene expression onset and kinetics, rather than absolute expression level, seems to be the important feature involved in the maintenance of morphological characteristics. Dramatic shifts in GRN structure and morphology between closely related species can however occur, perhaps due to strong selective pressure. Understanding this contrast will require further investigations into how GRN topologies can buffer changes to permit, or restrict, morphological changes.

variation associated with morphological differences in the skeleton are largely the result of gene expression changes occurring at the termini of the network. Several studies have also compared gene expression profiles between species of echinoids. Examination of developmental gene expression for Paracentrotus lividus revealed that expression levels are variable between individuals, but the timing of gene expression onset is highly consistent (Gildor and Ben-Tabou de-Leon, 2015). The P. lividus time-course data were then compared to the developmental time-course of S. purpuratus. Although ancestors of these two species diverged roughly 40 MYA, their developmental morphologies remain very similar. This comparison identified high levels of conservation in the expression profiles for genes within several GRN subcircuits. This then further highlights the robustness of the developmental GRN to genetic and environmental differences generated in the 40 million years of evolutionary history that separates these species. Drastic differences have however been observed at short evolutionary divergence times that correspond to alterations in larval life history strategies. Developmental transcriptomes of two species of Heliocidaris, one lecithotrophic (H. erythrogramma) and one planktotrophic (H. tuberculata), were compared to the planktotrophic outgroup L. variegatus (Israel et al., 2016). The most recent common ancestor of the Heliocidaris and Lytechinus taxa existed 35–45 MYA whereas the two Heliocidaris species shared a common ancestor about 5 MYA. In those intervening 5 Myr, however, an ancestor of H. erythrogramma shifted from developing into a feeding larva to a lecithotrophic mode where the larva depends on egg lipids and proteins to fulfill its energy requirements prior to metamorphosis. Expression profiles were more similar between the two planktotrophic species than the lecithotrophic species. As with the comparison between S. purpuratus and P. lividus, the expression of GRN components is strongly conserved between the two planktotrophic species, L. variegatus and H. tuberculata. In stark contrast the expression of GRN components in H. erythrogramma were quite different, and GRN-specific differences were found to be more pronounced than the expression differences across the transcriptome as a whole. The changes observed corroborated what is known regarding the altered embryogenesis of this species, including capturing the altered dynamics of larval skeleton

5. Evolution of the echinoderm skeletogenic subcircuit: cooption and innovation underlie novelty Some of the most significant comparison of GRNs among echinoderms have focused on the origins and evolution of the larval skeleton, in part because the cellular and gene regulatory controls that drive skeletogenesis in sea urchins are extremely well known. Before exploring the evolution of this structure, we start here with a brief description of the well-documented GRN for skeletogenesis in euechinoids (primarily in S. purpuratus and L. variegatus), which is provided in Fig. 3. Specification of the skeletogenic lineage begins in early embryogenesis as the asymmetric fourth and fifth cleavages that give rise to the large micromeres at the vegetal pole. The large micromeres then undergo two to three additional rounds of cell division, depending on the species, prior to ingressing into the blastocoel at the start of gastrulation at which point they are considered primary mesenchyme cells (Wu et al., 2007). These PMCs then migrate within the blastocoel in response to ectodermal signals, form a syncytial network, and begin to excrete the skeletal matrix biominerals. The GRN interactions controlling each of these transitions have been well described and reviewed in details elsewhere (Oliveri et al., 2008; McIntyre et al., 2014; Ettensohn, 2009; Rafiq et al., 2012; Wilt and Ettensohn, 2007). The transcription factor (TF) pmar1 is expressed solely in the micromeres coincident with their formation at the fourth cleavage and functions to repress the more ubiquitous hesC expression from the large micromeres (Oliveri et al., 2008; Revilla-i-Domingo 4

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

Fig. 3. Gene regulatory network (GRN) for specification of mesoderm and skeletogenic (skel) cells in different echinoderm classes. Images of representative larval morphologies are included for each class (see Fig. 1 legend for image credits). The GRN for euechinoid primary mesenchyme cell (PMC) specification is the most well established this set. The four other GRNs represent hypothetical networks based on the published data reported (described in the text) and should be viewed as a reasonable starting points to test network connectivity. In all cases, dashed edges are the speculative interactions based on wiring demonstrated in the euechinoid network. For the euechinoid GRN, the double negative gate sub-circuit underpinning PMC specification is indicated in the blue box. Genes highlighted by the orange box (skel) lie at the terminus of the network and represent the suite of effector genes responsible for cell behaviors involved in skeletogenesis. The nodes highlighted in yellow are genes commonly associated with larval and juvenile skeletogenesis centers across multiple echinoderms. This network is continuously updated and this version is adapted from the current model, hosted at the Davidson Lab website (sugp.caltech.edu/endomes).

the Euechinoids are the two major groups of Echinoids, and last shared a common ancestor around 268 million years ago. Since that time these lineages have diverged with respect to adult and juvenile skeletal plate morphology (Gao et al., 2015), and larval skeletal developmental regulatory origins (Yamazaki et al., 2014; Erkenbrack and Davidson, 2015). Cell populations involved in larval cidaroid skeletogenesis are notably different from those in the euechinoidea; in Euechinoids, the large micromere derived PMCs ingresses into the blastocoel invagination and form a syncytial network that hardens to form skeletal rods. By contrast, in Cidaroida, the cells that form the syncytium do not ingress into the blastocoel until mid-invagination, delaminating from the archenteron along with other mesodermally fated cell populations (Wray and McClay, 1988). A direct ortholog of the negative regulator pmar1 has not been detected in the cidaroid (E. tribuloides) genome or transcriptome assemblies (Erkenbrack and Davidson, 2015). There are also differences in regulatory connections between hesC and alx1, tbr, and ets1 in cidaroid compared to euechinoid urchins, which indicates there must be an alternative mechanism for specification of these cells during development (Yamazaki et al., 2014; Erkenbrack and Davidson, 2015). These differences may therefore account for the relatively delayed emergence of these cells in development in the pencil urchins. The comparison of these regulatory sub-circuits has enabled the reconstruction of the putative ancestral echinoid skeletal GRN (Erkenbrack et al., 2016). Analysis of larval skeleton formation in the Ophiuroids, or brittle stars, presents a fascinating opportunity to understand the GRN basis for convergent evolution. This is because the Echinoids and Ophiuroids both have a strikingly similar larval plutei form that, due to the phylogenetic relationship between these two classes (Fig. 1), seem likely to have independently evolved for 475 Myr (Pisani et al., 2012). In the brittle star Amphiura filiformis, like in S. purpuratus, and in contrast with E. tribuloides, skeletogenic mesenchyme ingresses into the blastocoel prior to invagination. Many gene orthologs of the sea urchin PMC GRN (i.e. alx, jun, p19, p58a, p58b) are expressed in this ophiuroid lineage (Dylus et al., 2016). There is also evidence that vegf

et al., 2007). hesC is known to repress several positive regulators of the PMC GRN (i.e. alx1, tbr, ets1) (Oliveri et al., 2008; Damle and Davidson, 2011). Thus exclusion of hesC expression from the large micromere population initiates the PMC GRN and establishes the skeletogenic lineage. Thus the activation of the PMC GRN requires the spatially restricted repression of a globally acting repressor. This double repressive circuit has therefore been termed the double negative gate (DNG, Fig. 3, blue box). Mid-level acting TFs then regulate the expression of an additional layer of regulatory genes (e.g. delta, foxb, dri), which in turn will activate the effector genes needed for the many functions of this lineage, including the formation of the biomineral (e.g. sm50, sm30, msp) (Fig. 3, orange box). While echinoderm larval skeletons are likely to be an evolutionarily acquired novelty, all adult echinoderms develop some form of biomineralized skeleton. Therefore, a reasonable and revelatory question to ask is whether the larval skeleton is co-opted from the adult GRN. Whole mount in situ analyses in S. purpuratus revealed that terminal differentiation genes, as well as components of the embryonic regulatory network from the PMC GRN, are also expressed in post-metamorphic, juvenile skeleton forming centers (Gao and Davidson, 2008). However, many of the earlier expressed genes, including those associated with the DNG, and micromere-specific signaling functions (delta signal under control of the DNG) are not expressed in the juvenile skeletogenic domains (Gao and Davidson, 2008). Many of the orthologs of these skeletogenic GRN genes are also expressed during juvenile skeletogenesis in P. miniata (ets1, alx1, hex) (Gao and Davidson, 2008) and E. tribuloides (vegfR, alx1, sm37) (Gao et al., 2015). These results suggest therefore that there are a core set of regulatory genes involved in establishing biomineralization in postmetamorphic, juvenile echinoderms that are also part of the embryonic GRN in sea urchins, consistent with the hypothesis that the PMC GRN has been in part coopted from an ancient, plesiomorphic GRN for adult skeletogenesis. Recent works have also been directed at resolving the GRN for PMC specification in the Cidaroids, or pencil urchins. Cidaroids along with

5

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

eventual downstream function have changed dramatically. Recent work within sea urchins also shows that similar circuits may also be observed across even greater evolutionary distances. Martik and McClay (2015) identified a subcircuit of the genes pax6, six3, six1/2, eya, and dach1 that is needed to direct the precise homing mechanism of late mesodermal progenitors in the sea urchin. They noted that this circuitry is striking similar to that used to specify vertebrate retinal cells. In this case, also, both upstream and downstream functions are dissimilar in these divergent taxa, but subcircuit topology is conserved. These examples clearly demonstrate that GRNs can function and evolve as discrete modules of subcircuits, and that similarity at one level in the GRN hierarchy does not require similarity at other levels. These cases studies on the of PMC GRN in echinoderms illuminate fundamental mechanisms by which GRNs can evolve and how change leads to the evolution of body plans. These studies have shown, (i) that GRNs can evolve as modules, (ii) that similar subcircuit modules can be observed over immense evolutionary distances, and (iii) that conservation of circuitry does not necessitate similar morphological outcomes, but perhaps implies shared developmental trajectories. They also suggest that (iv) co-option of these modules may be a prevalent mechanism for change, and that (v) independent co-option of circuits may explain convergence.

and vegfR are expressed in mesenchyme and associated epidermis in brittle stars, similar to the expression of the sea urchin orthologs (Morino et al., 2012), and that genes involved in sea urchin pluteus arm development (i.e. fgfA, pax2/5/8, pea3, otp, wnt5, and tet) are common to brittle star pluteus arm development (Morino et al., 2016). This extensive similarity therefore indicates, that if indeed independently co-opted from adult programs, similar circuitry has been coopted in these different lineages. This points to constraints of the ways that GRNs can evolve through co-option. However, as with the previous comparisons to the embryonic euechinoid skeletogenic GRNs, there are also some drastic differences in the brittle star network. The double negative gate, responsible for PMC specification in euechinoids, is not functional, or if so, does not use pmar1 or hesC in A. filiformis. The most closely related pmar ortholog in A. filliformis, pplx, has similar expression patterns to its sea urchin counterpart, however it possesses no repressive activity or skeletogenic potential when ectopically expressed in S. purpuratus. Furthermore, hesC is co-expressed with genes it is known to repress in S. purpuratus (i.e. delta, ets1/2, tbr), indicating hesC must not dominantly repress these genes in A. filiformis. Additionally, the interlocking loop of S. purpuratus (i.e. hex activates erg, which activates tgif) is inverted (i.e. tgif activates erg, which activates hex) in A. filiformis (Dylus et al., 2016). Thus, there are components of the skeletogenic GRN that are in common between these classes, but there are also significant network rearrangements that point to alternative mechanisms for making and patterning these morphologically similar structures. While sea star bipinnaria larvae do not form any skeletal elements, the sea cucumber auricularia larvae do form small biomineralized spicules. P. parvimensis Alx1, the ortholog of a key component of skeletogenic GRNs in other echinoderms, is associated with a cryptic skeletogenic lineage in the sea cucumber and its knock-down leads to the loss of the larval spicule (McCauley et al., 2012). Although sea cucumbers do not form micromeres, four cells expressing alx1 first appear in the vegetal plate prior to gastrulation. They ingress at the onset of invagination and form a dorsally localized ring within the blastocoel in a pattern that is highly reminiscent of pattern seen in sea urchin larval skeletogenesis. As in the sea urchin, Pp-alx1 is coexpressed with genes in the skeletogenesis GRN (e.g. ets1, erg, and tbr) prior to gastrulation. However, in contrast to the sea urchin where alx1 expression is restricted to the large micromeres, Pp-alx1 is also expressed with genes found more broadly in the endomesoderm (e.g. gata4/5/6). During ingression only a subset of skeletogenesis genes remain co-expressed (ets1 and erg). Thus at all stages of development there are dramatic differences in the regulatory state of sea urchin and sea cucumber skeletogenic lineage (McCauley et al., 2012). The sea star bipinnaria larva does not make any type of skeleton until metamorphosis. Unexpectedly, however, a comparison of GRN circuitry between the sea urchin and sea star demonstrated that both embryos use a similar motif, i.e. that regulatory interactions between erg, hex, and tgif are similarly, although not identically, engaged in a positive cross regulation to specify their respective mesodermal lineages (McCauley et al., 2010). However, this circuitry is not activated by a double negative gate in sea stars. While HesC can act as a repressor in the ectoderm, it does not function in this capacity in the sea star mesoderm, and hence is coexpressed with e.g. ets1, erg and tbr (McCauley et al., 2010), as with orthologs in the brittle stars and sea cucumbers. Instead, the sea star establishes this circuitry through a nuclear β-catenin gradient that exists in the vegetal pole of these embryos, in which high doses specify mesoderm and lower doses initiate endoderm circuits (McCauley et al., 2015). Nuclear β-catenin is required to specify all EM territories in the sea urchin (Logan et al., 1999; Weitzel et al., 2004; Lhomond et al., 2012), but as yet no dose response has been noted. The putative conservation of this subcircuit therefore suggests that these deeply divergent echinoderms nonetheless use highly similar mechanisms to specify mesoderm state, but that both the mechanism of the activation of this circuit as well as its

6. Evolution of GRNs occurs through changes to both cis regulatory sequences and transcription factor function Echinoderms have been models for unraveling the complex mechanisms of cis regulatory control for over three decades. The GRN models, genomic BAC libraries, and more recent BAC recombineering technologies have served to further strengthen these investigations. Population-level comparisons of CRM sequences have revealed high levels of polymorphism that may alter nominally functional aspects of the regulatory sequence. For example, examination of intra-species polymorphisms in a known CRM involved in regulating the expression of endo16 (Balhoff and Wray, 2005; Garfield et al., 2012) revealed low levels of sequence conservation, high rates of sequence indels, and even alterations within functionally characterized TF binding sites (TFBS). While the cis-regulatory regions were less polymorphic than intronic regions, they contained more polymorphisms, in particular insertions and deletions, than coding regions. Similar patterns of low conservation within CRM were also observed across congeners that diverged 3– 9 MYA (Balhoff and Wray, 2005; Kober and Bernardi, 2013). This, in theory, conflicts with the observation that sites of interspecific noncoding sequence conservation are indicative of function (Yuh et al., 2002). Measurement of interspecific polymorphisms occurring within 6 different CRM sequences between S. purpuratus and S. franciscanus, congeners which diverged approximately 18 MYA, also found similarly high levels of polymorphisms in these CRM. However, large indels ( > 20 bp) were mostly absent from these regions (Cameron et al., 2005). As with the population-level analyses, changes (i.e. SNPs and small indels) were observed to occur between TFBS. However, a careful cataloging of 46 TFBS identified in 8 different CRMs, revealed that the majority were unchanged with respect to the set of TF inputs and relatively invariant in function between S. purpuratus and L. variegatus. Some TFBS did change relative position within the examined CRM and/or showed evolutionary gain or loss. However, despite the differences detected, all CRM function equivalently in both species by BAC reporter assays (Cameron and Davidson, 2009). This is similar to the observation from comparisons of sea urchin and sea star CRM that, while individual TFBS are not conserved, the inputs to a given target and clusters of TFBS with the CRM are conserved even at these extremely large evolutionary distances (~500 Mya) (Hinman et al., 2007). While sequence conservation, even individual TFBS, may not be an ideal signature of functional CRM, there are two general rules that can be derived from these comparisons; (i) TF inputs are maintained, even 6

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

This reveals the potential for TF to evolve over these large interclass or even inter-phylum distances. The developmental consequences of this change are as yet unknown. However, genome-wide binding of Tbr in both S. purpuratus and P. miniata using ChIP-seq analyses show that both primary and secondary motif use occurs in CRMs of genes at all levels of the GRN in both organisms. This implies that compensatory features exist within CRMs in vivo that stabilize binding of SpTbr to the low affinity sites. Furthermore, there is very little overlap in orthologous gene regulation across sea urchins and sea stars (Cary et al., 2017). More datasets of these kinds are required to unravel the relative contributions of CRM sequence changes and TF functional changes to the evolution of GRNs. Almost certainly this will involve the consideration of local network features and how these may permit or hinder various kinds of evolutionary change within each GRN.

when the relative organisation of the TFBS are not, and (ii) TFs that operate in heterodimeric pairs might maintain TFBS for both TFs in proximity. Advances in chromatin assay and CRM screening will greatly facilitate the investigation of CRM evolution. For example, integration of conserved sequence, TFBS signatures, along with chromatin accessibility assays (e.g. ATAC-seq data available on Echinobase.org) may help further inform us about the evolutionary signatures of active functional CRM. Finally, recently developed sensitive, multiplexed assays of CRM detection (Tulin et al., 2016) and function (Nam and Davidson, 2012; Nam et al., 2010) will increase the throughput of these CRM analyses. More recent work has indicated that the different classes of Echinoderms may represent the ideal evolutionary distances to study evolutionary changes in the function of the TFs themselves. A great deal of information has been gathered over the past decades about the evolution of CRMs, and in particular the idea that CRMs are more amenable to change, as mutations arising in these are less likely to be detrimental. By contrast TFs have many targets across the genome, are typically reused in various developmental contexts, and are therefore highly pleiotropic and hence more constrained. This contrast has largely borne out by experimental evidence of the past 30 years. However, it is becoming increasing evident that TF themselves also evolve altered functions (Cheatle Jarvela and Hinman, 2015). One recent example is the T-box transcription factor Tbrain. Within the sea urchin, tbrain (SpTbr) is expressed in the highly-specialized mesodermal skeletogenic micromeres where it acts as a top-level regulator of the skeletogenic GRN. However in asteroids, holothurians, and most other deuterostomes, tbrain is more broadly expressed throughout the endomesoderm and ectoderm. Sea star Tbrain (PmTbr) is necessary for specification of cell types within all three germ layers of developing P. miniata (Hinman and Davidson, 2007; Cheatle Jarvela et al., 2014). Protein-binding microarrays were used to establish DNA sequence binding preferences, and showed that while PmTbr and SpTbr proteins, and indeed their mouse ortholog Eomesodermin, bound a highly similar primary motif with high affinity, the sea star PmTbr also bound a secondary motif with lower affinity (Fig. 4; Cheatle Jarvela et al., 2014). Intriguingly, mouse Eomesodermin, also had a preference for a different secondary motif.

7. Conclusion In the post-genomic era, echinoderm developmental model systems present a unique opportunity to interrogate the mechanisms of morphological evolution. Indeed echinoderms, and in particular the sea urchin, have been powerful models for developmental biology for over a century. The GRN for early development provides an unmatched framework for understanding morphogenetic processes, in particular, skeletogenesis. The remarkable diversity in adult and larval body plans provides the context to address numerous questions concerning the nature and mechanisms of GRN evolution and how changes to regulatory interactions drive modifications to developmental outcomes. This experimental framework has yielded several surprising lessons. First that there is an immense level of variability that is permissible, not only in the noncoding sequences that comprise CRM but also in absolute expression levels, without drastically affecting the overall developmental outcome. While it is clear that the GRN is robust to such variability, it remains unclear what aspects of the regulatory network buffer against these changes and also the ways in which changes can break through, leading to adaptive alterations to developmental programs. It appears that one mode in which GRN topology may be altered is through co-option of network motifs from other contexts, such as we discuss for the module that drives skeletogenesis

Fig. 4. Tbrain orthologs have altered preferences for secondary sites. The primary and secondary site motifs for the Tbrain orthologs from sea star (PmTbr), sea urchin (SpTbr) and mouse (MmEomes) are shown. The motifs for sea star and sea urchin proteins were identified by protein binding microarray analyses (Cheatle Jarvela et al., 2014). There was no detected secondary site preference for the sea urchin protein. The motifs of the mouse ortholog, Eomesodermin, were retrieved from UniProbe database (Hume et al., 2015); accession: UP00068).

7

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

in various echinoderm larvae and adults. However, the field is currently poised to address other mechanisms through which regulatory networks evolve, including through changes to both cis and trans components. Along with the accumulation of additional genomic datasets from this phylum, the utilization of ATAC-seq and ChIP-seq to define chromatin features of interest, and leveraging the recombinant BAC libraries and nanotag CRM interrogation methodologies will facilitate further comparisons. In particular, these tools will enhance the ability to detect signatures of cis-regulatory sequence through the identification of microsynteny, patches of sequence conservation, chromatin features, and clusters of TFBS that may suggest functionality. This will also permit the exploration of the ways in which transcription factor modularity permits regulatory network evolution and how pervasive this recently described phenomenon is across the GRN. This phylum of animals is situated to continue to make significant contributions to our understanding of how both regulatory networks and the morphologies they encode evolve.

Mol. Biol. Evol. 31 (10), 2672–2688. Cheatle Jarvela, A.M., Hinman, V.F., 2015. Evolution of transcription factor function as a mechanism for changing metazoan developmental gene regulatory networks. Evodevo 6 (1), 3. Damle, S., Davidson, E.H., 2011. Precise cis-regulatory control of spatial and temporal expression of the alx-1 gene in the skeletogenic lineage of s. purpuratus. Dev. Biol. 357 (2), 505–517. Davidson, E.H., 1999. A view from the genome: spatial control of transcription in sea urchin development. Curr. Opin. Genet. Dev. 9 (5), 530–541. Davidson, E.H., Rast, J.P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C.-H., Minokawa, T., Amore, G., Hinman, V., Arenas-Mena, Cesar, Otim, O., Brown, C.T., Livi, C.B., Lee, P.Y., Revilla, R., Rust, A.G., Pan, Z. jun, Schilstra, M.J., Clarke, P.J.C., Arnone, M.I., Bolouri, H., 2002a. A genomic regulatory network for development. Science 295 (5560), 1669–1678. Davidson, E.H., Rast, J.P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C.-H., Minokawa, T., Amore, G., Hinman, V., Arenas-Mena, C.ésar, Otim, O., Brown, C.T., Livi, C.B., Lee, P.Y., Revilla, R., Schilstra, M.J., Clarke, P.J.C., Rust, A.G., Pan, Z., Arnone, M.I., Bolouri, H., 2002b. A provisional regulatory gene network for specification of endomesoderm in the sea urchin embryo. Dev. Biol. 246 (1), 162–190. Dorée, M., Hunt, T., 2002. From Cdc2 to Cdk1: when did the cell cycle kinase join its cyclin partner? J. Cell Sci. 115 (Pt 12), 2461–2464. Dylus, D.V., Czarkwiani, A., Stångberg, J., Ortega-Martinez, O., Dupont, S., Oliveri, P., 2016. Large-scale gene expression study in the ophiuroid Amphiura filiformis provides insights into evolution of gene regulatory networks. Evodevo 7, 2. Erkenbrack, E.M., Ako-Asare, K., Miller, E., Tekelenburg, S., Thompson, J.R., Romano, L., 2016. Ancestral state reconstruction by comparative analysis of a GRN kernel operating in echinoderms. Dev. Genes Evol. 226 (1), 37–45. Erkenbrack, E.M., Davidson, E.H., 2015. Evolutionary rewiring of gene regulatory network linkages at divergence of the echinoid subclasses. Proc. Natl. Acad. Sci. USA 112 (30), E4075–E4084. Ettensohn, C.A., 2009. Lessons from a gene regulatory network: echinoderm skeletogenesis provides insights into evolution, plasticity and morphogenesis. Development 136 (1), 11–21. Finnerty, J.R., 2003. The origins of axial patterning in the metazoa: how old is bilateral symmetry? Int. J. Dev. Biol. 47 (7–8), 523–529. Freeman, R., Ikuta, T., Wu, M., Koyanagi, R., Kawashima, T., Tagawa, K., Humphreys, T., Fang, G.-C., Fujiyama, A., Saiga, H., Lowe, C., Worley, K., Jenkins, J., Schmutz, J., Kirschner, M., Rokhsar, D., Satoh, N., Gerhart, J., 2012. Identical genomic organization of two hemichordate hox clusters. Curr. Biol. 22 (21), 2053–2058. Gao, F., Davidson, E.H., 2008. Transfer of a large gene regulatory apparatus to a new developmental address in echinoid evolution. Proc. Natl. Acad. Sci. USA 105 (16), 6091–6096. Gao, F., Thompson, J.R., Petsios, E., Erkenbrack, E., Moats, R.A., Bottjer, D.J., Davidson, E.H., 2015. Juvenile skeletogenesis in anciently diverged sea urchin clades. Dev. Biol. 400 (1), 148–158. Garfield, D., Haygood, R., Nielsen, W.J., Wray, G.A., 2012. Population genetics of cisregulatory sequences that operate during embryonic development in the sea urchin Strongylocentrotus purpuratus. Evol. Dev. 14 (2), 152–167. Garfield, D.A., Runcie, D.E., Babbitt, C.C., Haywood, R., Nielsen, W.J., Wray, G.A., 2013. The impact of gene expression variation on the robustness and evolvability of a developmental gene regulatory network. PLoS Biol. 11 (10). http://dx.doi.org/ 10.1371/journal.pbio.1001696, (e1001696) (PMID: 24204211). Gildor, T., Ben-Tabou de-Leon, S., 2015. Comparative study of regulatory circuits in two sea urchin species reveals tight control of timing and high conservation of expression dynamics. PLoS Genet. 11 (7), e1005435. Hara, Y., Yamaguchi, M., Akasaka, K., Nakano, H., Nonaka, M., Amemiya, S., 2006. Expression patterns of Hox genes in larvae of the sea lily Metacrinus rotundus. Dev. Genes Evol. 216 (12), 797–809. Hinman, V.F., Davidson, E.H., 2007. Evolutionary plasticity of developmental gene regulatory network architecture. Proc. Natl. Acad. Sci. USA 104 (49), 19404–19409. Hinman, V.F., Nguyen, A., Davidson, E.H., 2007. Caught in the evolutionary act: precise cis-regulatory basis of difference in the organization of gene networks of sea stars and sea urchins. Dev. Biol. 312 (2), 584–595. Howard-Ashby, M., Materna, S.C., Brown, C.T., Chen, L., Cameron, R.A., Davidson, E.H., 2006a. Gene families encoding transcription factors expressed in early development of Strongylocentrotus purpuratus. Dev. Biol. 300 (1), 90–107. Howard-Ashby, M., Materna, S.C., Brown, C.T., Chen, L., Cameron, R.A., Davidson, E.H., 2006b. Identification and characterization of homeobox transcription factor genes in Strongylocentrotus purpuratus, and their expression in embryonic development. Dev. Biol. 300 (1), 74–89. Hume, M.A., Barrera, L.A., Gisselbrecht, S.S., Bulyk, M.L., 2015. UniPROBE, update 2015: new tools and content for the online database of protein-binding microarray data on protein-DNA interactions. Nucleic Acids Res. 43, D117–D122. Irimia, M., Tena, J.J., Alexis, M.S., Fernandez-Miñan, A., Maeso, I., Bogdanovic, O., de la Calle-Mustienes, E., Roy, S.W., Gómez-Skarmeta, J.L., Fraser, H.B., 2012. Extensive conservation of ancient microsynteny across metazoans due to cis-regulatory constraints. Genome Res. 22 (12), 2356–2367. Israel, J.W., Martik, M.L., Byrne, M., Raff, E.C., Raff, R.A., McClay, D.R., Wray, G.A., 2016. Comparative developmental transcriptomics reveals rewiring of a highly conserved gene regulatory network during a major life history switch in the sea urchin Genus Heliocidaris. PLoS Biol. 14 (3), e1002391. Kikuchi, M., Omori, A., Kurokawa, D., Akasaka, K., 2015. Patterning of anteroposterior body axis displayed in the expression of Hox genes in sea cucumber Apostichopus japonicus. Dev. Genes Evol. 225 (5), 275–286. Kmita, M., Duboule, D., 2003. Organizing axes in time and space; 25 years of colinear tinkering. Science 301 (5631), 331–333.

Acknowledgments The authors thank the reviewers for providing constructive feedback during the preparation of this manuscript. This work was supported by grants from the National Science Foundation, NSF IOS 0844948 and IOS-1557431. References Andrikou, C., Pai, C.-Y., Su, Y.-H., Arnone, M.I., 2015. Logics and properties of a genetic regulatory program that drives embryonic muscle development in an echinoderm. eLife 4. Angerer, L.M., Angerer, R.C., 1999. Regulative development of the sea urchin embryo: signalling cascades and morphogen gradients. Semin. Cell Dev. Biol. 10 (3), 327–334. Angerer, L.M., Yaguchi, S., Angerer, R.C., Burke, R.D., 2011. The evolution of nervous system patterning: insights from sea urchin development. Development 138 (17), 3613–3623. Annunziata, R., Martinez, P., Arnone, M.I., 2013. Intact cluster and chordate-like expression of ParaHox genes in a sea star. BMC Biol. 11, 68. Annunziata, R., Perillo, M., Andrikou, C., Cole, A.G., Martinez, P., Arnone, M.I., 2014. Pattern and process during sea urchin gut morphogenesis: the regulatory landscape. Genesis 52 (3), 251–268. Arenas-Mena, C., Cameron, A.R., Davidson, E.H., 2000. Spatial expression of Hox cluster genes in the ontogeny of a sea urchin. Development 127 (21), 4631–4643. Balhoff, J.P., Wray, G.A., 2005. Evolutionary analysis of the well characterized endo16 promoter reveals substantial variation within functional sites. Proc. Natl. Acad. Sci. USA 102 (24), 8591–8596. Baughman, K.W., McDougall, C., Cummins, S.F., Hall, M., Degnan, B.M., Satoh, N., Shoguchi, E., 2014. Genomic organization of Hox and ParaHox clusters in the echinoderm, Acanthaster planci. Genesis 52 (12), 952–958. Bennett, K.C., Young, C.M., Emlet, R.B., 2012. Larval development and metamorphosis of the deep-sea cidaroid urchin Cidaris blakei. Biol. Bull. 222 (2), 105–117. Byrne, M., Nakajima, Y., Chee, F.C., Burke, R.D., 2007. Apical organs in echinoderm larvae: insights into larval evolution in the Ambulacraria. Evol. Dev. 9 (5), 432–445. Cameron, C.B., Garey, J.R., Swalla, B.J., 2000. Evolution of the chordate body plan: new insights from phylogenetic analyses of deuterostome phyla. Proc. Natl. Acad. Sci. USA 97 (9), 4469–4474. Cameron, R.A., Chow, S.H., Berney, K., Chiu, T.-Y., Yuan, Q.-A., Krämer, A., Helguero, A., Ransick, A., Yun, M., Davidson, E.H., 2005. An evolutionary constraint: strongly disfavored class of change in DNA sequence during divergence of cis-regulatory modules. Proc. Natl. Acad. Sci. USA 102 (33), 11769–11774. Cameron, R.A., Davidson, E.H., 2009. Flexibility of transcription factor target site position in conserved cis-regulatory modules. Dev. Biol. 336 (1), 122–135. Cameron, R.A., Kudtarkar, P., Gordon, S.M., Worley, K.C., Gibbs, R.A., 2015. Do echinoderm genomes measure up? Mar. Genom. 22, 1–9. Cameron, R.A., Rowen, L., Nesbitt, R., Bloom, S., Rast, J.P., Berney, K., Arenas-Mena, C., Martinez, P., Lucas, S., Richardson, P.M., Davidson, E.H., Peterson, K.J., Hood, L., 2006. Unusual gene order and organization of the sea urchin hox cluster. J. Exp. Zool. B Mol. Dev. Evol. 306 (1), 45–58. Cannon, J.T., Kocot, K.M., Waits, D.S., Weese, D.A., Swalla, B.J., Santos, S.R., Halanych, K.M., 2014. Phylogenomic resolution of the hemichordate and echinoderm clade. Curr. Biol. 24 (23), 2827–2832. Cary, G., Chaetle-Jarvela, A., Francolini, R., Hinman, V., 2017. Genome-wide use of high and low affinity Tbrain transcription factor binding sites during echinoderm development. Proc. Natl. Acad. Sci. USA. Cheatle Jarvela, A.M., Brubaker, L., Vedenko, A., Gupta, A., Armitage, B.A., Bulyk, M.L., Hinman, V.F., 2014. Modular evolution of DNA-binding preference of a Tbrain transcription factor provides a mechanism for modifying gene regulatory networks.

8

Developmental Biology (xxxx) xxxx–xxxx

G.A. Cary, V.F. Hinman

into the immune system of the sea urchin. Science 314 (5801), 952–956. Reich, A., Dunn, C., Akasaka, K., Wessel, G., 2015. Phylogenomic analyses of Echinodermata support the sister groups of Asterozoa and Echinozoa. PLoS One 10 (3), e0119627. Revilla-i-Domingo, R., Oliveri, P., Davidson, E.H., 2007. A missing link in the sea urchin embryo gene regulatory network: hesC and the double-negative specification of micromeres. Proc. Natl. Acad. Sci. USA 104 (30), 12383–12388. Rouse, G.W., Jermiin, L.S., Wilson, N.G., Eeckhaut, I., Lanterbecq, D., Oji, T., Young, C.M., Browning, T., Cisternas, P., Helgen, L.E., Stuckey, M., Messing, C.G., 2013. Fixed, free, and fixed: the fickle phylogeny of extant Crinoidea (Echinodermata) and their Permian-Triassic origin. Mol. Phylogenet. Evol. 66 (1), 161–181. Santella, L., Vasilev, F., Chun, J.T., 2012. Fertilization in echinoderms. Biochem. Biophys. Res. Commun. 425 (3), 588–594. Saunders, L.R., McClay, D.R., 2014. Sub-circuits of a gene regulatory network control a developmental epithelial-mesenchymal transition. Development 141 (7), 1503–1513. Sea Urchin Genome Sequencing Consortium, Sodergren, E., Weinstock, G.M., Davidson, E.H., Cameron, R.A., Gibbs, R.A., Angerer, R.C., Angerer, L.M., Arnone, M.I., Burgess, D.R., Burke, R.D., Coffman, J.A., Dean, M., Elphick, M.R., Ettensohn, C.A., Foltz, K.R., Hamdoun, A., Hynes, R.O., Klein, W.H., Marzluff, W., Wright, R., 2006. The genome of the sea urchin Strongylocentrotus purpuratus. Science 314 (5801), 941–952. Simakov, O., Kawashima, T., Marlétaz, F., Jenkins, J., Koyanagi, R., Mitros, T., Hisata, K., Bredeson, J., Shoguchi, E., Gyoja, F., Yue, J.-X., Chen, Y.-C., Freeman, R.M., Sasaki, A., Hikosaka-Katayama, T., Sato, A., Fujie, M., Baughman, K.W., Levine, J., Gonzalez, P., Gerhart, J., 2015. Hemichordate genomes and deuterostome origins. Nature 527 (7579), 459–465. Smith, A.B., 1988. Fossil evidence for the relationships of extant echinoderm classes and their times of divergence. In: Paul, C.R.C., Smith, A.B. (Eds.), Echinoderm Phylogeny and Evolutionary Biology. Clarendon Press, Oxford, 85–97. Telford, M.J., Lowe, C.J., Cameron, C.B., Ortega-Martinez, O., Aronowicz, J., Oliveri, P., Copley, R.R., 2014. Phylogenomic analysis of echinoderm class relationships supports Asterozoa. Proc. Biol. Sci. 281 (1786). Tsuchimoto, J., Yamaguchi, M., 2014. Hox expression in the direct-type developing sand dollar Peronella japonica. Dev. Dyn. 243 (8), 1020–1029. Tulin, S., Barsi, J.C., Bocconcelli, C., Smith, J., 2016. Genome-wide identification of enhancer elements. Int. J. Dev. Biol. 60 (4–6), 141–150. Tu, Q., Brown, C.T., Davidson, E.H., Oliveri, P., 2006. Sea urchin Forkhead gene family: phylogeny and embryonic expression. Dev. Biol. 300 (1), 49–62. University of Saskatchewan University Archives & Special Collections, T.C. Lacalli fonds [Online]. Available at: 〈http://scaa.sk.ca/gallery/lacalli〉 (Accessed 11 September 2016). Weitzel, H., Illies, M., Byrum, C., Xu, R., et al., 2004. Differential stability of β-catenin along the animal-vegetal axis of the sea urchin embryo mediated by dishevelled. Available at: 〈http://dev.biologists.org/content/131/12/2947.short〉. Wilt, F.H., Ettensohn, C.A., 2007. The Morphogenesis and Biomineralization of the Sea Urchin Larval Skeleton. In: Buerlein, E. (Ed.), Handbook of Biomineralization. Wiley-VCH Verlag GmbH, Weinheim, Germany, 182–210. Wray, G.A., McClay, D.R., 1988. The origin of spicule-forming cells in a “primitive” sea urchin (Eucidaris tribuloides) which appears to lack primary mesenchyme cells. Development 103 (2), 305–315. Wu, S.-Y., Ferkowicz, M., McClay, D.R., 2007. Ingression of primary mesenchyme cells of the sea urchin embryo: a precisely timed epithelial mesenchymal transition. Birth Defects Res. C Embryo Today 81 (4), 241–252. Yamazaki, A., Kidachi, Y., Yamaguchi, M., Minokawa, T., 2014. Larval mesenchyme cell specification in the primitive echinoid occurs independently of the double-negative gate. Development 141 (13), 2669–2679. Yanagida, M., 2014. The role of model organisms in the history of mitosis research. Cold Spring Harb. Perspect. Biol. 6 (9), a015768. Yuh, C.H., Bolouri, H., Davidson, E.H., 2001. Cis-regulatory logic in the endo16 gene: switching from a specification to a differentiation mode of control. Development 128 (5), 617–629. Yuh, C.-H., Brown, C.T., Livi, C.B., Rowen, L., Clarke, P.J.C., Davidson, E.H., 2002. Patchy interspecific sequence similarities efficiently identify positive cis-regulatory elements in the sea urchin. Dev. Biol. 246 (1), 148–161. Yuh, C.-H., Dorman, E.R., Howard, M.L., Davidson, E.H., 2004. An otx cis-regulatory module: a key node in the sea urchin endomesoderm gene regulatory network. Dev. Biol. 269 (2), 536–551.

Kober, K.M., Bernardi, G., 2013. Phylogenomics of strongylocentrotid sea urchins. BMC Evol. Biol. 13, 88. Lhomond, G., McClay, D.R., Gache, C., Croce, J.C., 2012. Frizzled1/2/7 signaling directs β-catenin nuclearisation and initiates endoderm specification in macromeres during sea urchin embryogenesis. Development 139 (4), 816–825. Logan, C.Y., Miller, J.R., Ferkowicz, M.J., McClay, D.R., 1999. Nuclear beta-catenin is required to specify vegetal cell fates in the sea urchin embryo. Development 126 (2), 345–357. Long, K.A., Nossa, C.W., Sewell, M.A., Putnam, N.H., Ryan, J.F., 2016. Low coverage sequencing of three echinoderm genomes: the brittle star Ophionereis fasciata, the sea star Patiriella regularis, and the sea cucumber Australostichopus mollis. Gigascience 5, 20. Long, S., Byrne, M., 2001. Evolution of the echinoderm Hox gene cluster. Evol. Dev. 3 (5), 302–311. Lowe, C.J., Wray, G.A., 1997. Radical alterations in the roles of homeobox genes during echinoderm evolution. Nature 389 (6652), 718–721. Martik, M.L., McClay, D.R., 2015. Deployment of a retinal determination gene network drives directed cell migration in the sea urchin embryo. eLife 4. Materna, S.C., Berney, K., Cameron, R.A., 2006. The S. purpuratus genome: a comparative perspective. Dev. Biol. 300 (1), 485–495. McCauley, B.S., Akyar, E., Saad, H.R., Hinman, V.F., 2015. Dose-dependent nuclear βcatenin response segregates endomesoderm along the sea star primary axis. Development 142 (1), 207–217. McCauley, B.S., Weideman, E.P., Hinman, V.F., 2010. A conserved gene regulatory network subcircuit drives different developmental fates in the vegetal pole of highly divergent echinoderm embryos. Dev. Biol. 340 (2), 200–208. McCauley, B.S., Wright, E.P., Exner, C., Kitazawa, C., Hinman, V.F., 2012. Development of an embryonic skeletogenic mesenchyme lineage in a sea cucumber reveals the trajectory of change for the evolution of novel structures in echinoderms. Evodevo 3 (1), 17. McIntyre, D.C., Lyons, D.C., Martik, M., McClay, D.R., 2014. Branching out: origins of the sea urchin larval skeleton in development and evolution. Genesis 52 (3), 173–185. Molina, M.D., de Crozé, N., Haillot, E., Lepage, T., 2013. Nodal: master and commander of the dorsal-ventral and left-right axes in the sea urchin embryo. Curr. Opin. Genet. Dev. 23 (4), 445–453. Morino, Y., Koga, H., Tachibana, K., Shoguchi, E., Kiyomoto, M., Wada, H., 2012. Heterochronic activation of VEGF signaling and the evolution of the skeleton in echinoderm pluteus larvae. Evol. Dev. 14 (5), 428–436. Morino, Y., Koga, H., Wada, H., 2016. The conserved genetic background for pluteus arm development in brittle stars and sea urchin. Evol. Dev. 18 (2), 89–95. Nakano, H., Hibino, T., Oji, T., Hara, Y., Amemiya, S., 2003. Larval stages of a living sea lily (stalked crinoid echinoderm). Nature 421 (6919), 158–160. Nam, J., Davidson, E.H., 2012. Barcoded DNA-tag reporters for multiplex cis-regulatory analysis. PLoS One 7 (4), e35934. Nam, J., Dong, P., Tarpine, R., Istrail, S., Davidson, E.H., 2010. Functional cis-regulatory genomics for systems biology. Proc. Natl. Acad. Sci. USA 107 (8), 3930–3935. Oliveri, P., Tu, Q., Davidson, E.H., 2008. Global regulatory logic for specification of an embryonic cell lineage. Proc. Natl. Acad. Sci. USA 105 (16), 5955–5962. Peter, I.S., Faure, E., Davidson, E.H., 2012. Predictive computation of genomic logic processing functions in embryonic development. Proc. Natl. Acad. Sci. USA 109 (41), 16434–16442. Pisani, D., Feuda, R., Peterson, K.J., Smith, A.B., 2012. Resolving phylogenetic signal from noise when divergence is rapid: a new look at the old problem of echinoderm class relationships. Mol. Phylogenet. Evol. 62 (1), 27–34. Puritz, J.B., Keever, C.C., Addison, J.A., Byrne, M., Hart, M.W., Grosberg, R.K., Toonen, R.J., 2012. Extraordinarily rapid life-history divergence between Cryptasterina sea star species. Proc. Biol. Sci. 279 (1744), 3914–3922. Raff, R.A., Byrne, M., 2006. The active evolutionary lives of echinoderm larvae. Heredity 97 (3), 244–252. Rafiq, K., Cheers, M.S., Ettensohn, C.A., 2012. The genomic regulatory control of skeletal morphogenesis in the sea urchin. Development 139 (3), 579–590. Range, R., Lapraz, F., Quirin, M., Marro, S., Besnardeau, L., Lepage, T., 2007. Cisregulatory analysis of nodal and maternal control of dorsal-ventral axis formation by Univin, a TGF-beta related to Vg1. Development 134 (20), 3649–3664. Ransick, A., Davidson, E.H., 2006. cis-regulatory processing of Notch signaling input to the sea urchin glial cells missing gene during mesoderm specification. Dev. Biol. 297 (2), 587–602. Rast, J.P., Smith, L.C., Loza-Coll, M., Hibino, T., Litman, G.W., 2006. Genomic insights

9