Special Reviews

Paleogenomics of Echinoderms

See allHide authors and affiliations

Science  10 Nov 2006:
Vol. 314, Issue 5801, pp. 956-960
DOI: 10.1126/science.1132310


Paleogenomics propels the meaning of genomic studies back through hundreds of millions of years of deep time. Now that the genome of the echinoid Strongylocentrotus purpuratus is sequenced, the operation of its genes can be interpreted in light of the well-understood echinoderm fossil record. Characters that first appear in Early Cambrian forms are still characteristic of echinoderms today. Key genes for one of these characters, the biomineralized tissue stereom, can be identified in the S. purpuratus genome and are likely to be the same genes that were involved with stereom formation in the earliest echinoderms some 520 million years ago.

Paleogenomics is the addition of the component of deep time to the field of genomics (1). Initial studies have concentrated on reconstructing regions of the ancestral mammalian genome (2) or sequencing preserved DNA of recently extinct organisms, such as the wooly mammoth (3). Although such studies present many exciting possibilities, the prospects for paleogenomics are much broader.

Genomics offers the opportunity of identifying genes that are responsible for the evolution of key shared characters of organisms, or synapomorphies, which are ultimately used to reconstruct the tree of life. Paleogenomics thus allows for both the geologic and genetic fossil records to shed light on the origin and subsequent evolution through time of key genes and the key synapomorphies that they encode.

Echinoderm Paleogenomics and the Stereom Skeleton

The initial appearance of biomineralized skeletal tissues in the fossil record (4), just before the beginning of the Cambrian ∼542 million years ago (Ma), coincides with the start of the rapid increase in the diversity of metazoans termed the Cambrian explosion (5). Later in the Early Cambrian, by 520 Ma, a variety of biomineralized skeletal structures had appeared. Among the most distinctive is a major echinoderm synapomorphy: the unique endoskeletal tissue called stereom.

Stereom is composed of calcite organized into a meshlike structure (Fig. 1, A to D), the pores of which in life are populated with dermal cells and fibers (6). Much is known about this structure from studies of representatives of the approximately 7000 species that constitute the five clades of living echinoderms (crinoids or sea lilies, ophiuroids or brittle stars, asteroids or sea stars, holothuroids or sea cucumbers, and echinoids or sea urchins), all of which produce stereom endo-skeletons (6). Stereom forms structural elements (Fig. 1) that can be embedded in soft tissues or may be fused together to form larger compound plates, generating the various types of echinoderm skeletons (6).

Fig. 1.

Stereom formation in modern and fossil echinoderms. (A) A spine of the modern sea star Asterias with soft tissue removed, constructed of meshlike stereom. (B) An occular plate of the modern sea star Asterias with soft tissue removed, constructed of meshlike stereom. (C) Median cross section of the stylocone from the Middle Cambrian stylophoran ?Ceratocystis, showing stereom construction. (D) Detail of stereom from the inner face in external view of the stylocone from the Middle Cambrian stylophoran ?Ceratocystis. (E) Larval spicule of S. purpuratus with biomineralized stereom (calcite) dissolved away, showing the distribution of spicule matrix proteins. [(A) and (B) are used with permission from C. Sumrall; (C) and (D) are reprinted by permission from Macmillan Publishers Ltd. (Nature) (11), copyright (2005); (E) is reprinted from (27), copyright 1983, with permission from Elsevier.] Scale bars in (A) to (C), 500 μm; scale bar in (D), 100 μm; magnification of (E) is ×3000.

Because the high-magnesium calcite of which stereom is constructed is stable, echinoderm skeletons are very durable during the process of fossil preservation, and this has led to an abundant and well-understood echinoderm fossil record (6, 7).

Echinoderm Phylogeny and the History of Stereom

Echinoderms, one of the three major phyla of the Deuterostomia, make their first appearance in the fossil record during the Early Cambrian, about 520 Ma, but the most primitive echinoderms are the stylophorans, a bizarre group first recorded in the fossil record in the Middle Cambrian (∼510 Ma) (8) (Figs. 2 and 3A). Stylophorans are recognized as echinoderms because they possess stereom (Fig. 1, C and D). Another major echinoderm synapomorphy is the water vascular system (8, 9), a closed circulatory system that uses ambient seawater to provide the hydraulic force necessary to extend the tube feet of living forms. The water vascular system first made its appearance in another group of Cambrian forms, the solutes (Figs. 2 and 3B). Both stylophorans and solutes are “stem-group” echinoderms (8) because they are more closely related to living echinoderms than they are to living hemichordates, the closest living relatives of modern echinoderms, but they are not descended from the last common ancestor of the living echinoderms (Fig. 2).

Fig. 2.

Evolutionary history of the major echinoderm groups. Deuterostomia consists of three major groups: the chordates, hemichordates, and echinoderms, all with fossil representatives in the Cambrian. Cambrian echinoderms are recognized by the possession of stereom, but the phylogenetically most basal groups (such as stylophorans) lack the water vascular system, are highly asymmetrical, and possess gill slits. Pentameral symmetry is seen in two major Early Cambrian lineages, the edrioasteroids and eocrinoids; a third Early Cambrian taxon, the helicoplacoids, have an unusual threefold symmetry thought to be derived from the ancestral pentameral arrangement (10). All stem-group echinoderm lineages became extinct by the Carboniferous (indicated with crosses). Crown-group echinoderms, indicated by the yellow circle, consist of the five major extant lineages in addition to numerous extinct lineages not shown. Most class-level crown groups first appear in the latest Paleozoic–early Mesozoic, including echinoids. The lineage leading to echinoids, and hence to S. purpuratus, is indicated in purple. Known stratigraphic ranges are shown with thick lines, and inferred range extensions are shown with thin lines.

Fig. 3.

Stem-group (A to E) and crown-group (F to H) echinoderms. (A) Thestylophoran Cothurnocystis bifida (Middle Cambrian, Utah, USA). The putative gill skeletons as viewed from the back side are indicated with an arrowhead. M is the putative mouth. The arrow indicates the posterior appendage. (B) The solute Coleicarpus sprinklei (Middle Cambrian, Utah, USA). The arrow indicates the posterior appendage, and the double arrow points to the single ambulacrum. (C) The helicoplacoid Helicoplacus (Early Cambrian, California, USA). The double arrow points to one of the ambulacral grooves. (D) The eocrinoid Gogia spiralis (Middle Cambrian, Utah, USA). The double arrow points to one of the five arms. (E) The edrioasteroid Edriophus bigsbyi (Ordovician, Ontario, Canada). It displays conspicuous pentameral symmetry; one of the arms is indicated by the double arrow. (F) The crinoid Dorycrinus mississippiensis (Mississippian, Indiana, USA). (G) The asteroid Furcaster palaeozoicus (Devonian, Budenbach, Germany). (H) The echinoid Bothriocidaris (Ordovician, Estonia) [reprinted with permission from A. B. Smith, from (30)]. Scale bar, 0.5 cm in (A); 0.75 cm in (B) to (E); 2 cm in (F); 1.3 cm in (G); and 0.15 cm in (H). Part of a penny is shown for scale in (C).

Other stem-group echinoderms that produced both a biomineralized stereom skeleton and plate morphologies indicating the presence of a water vascular system appear in the stratigraphic record from about 520 Ma (10). These include the helicoplacoids, eocrinoids, and edrioasteroids (Fig. 3, C to E). Pentameral symmetry of the adult body, a highly characteristic echinoderm synapomorphy of crown-group echinoderms, makes its initial appearance in the edrioasteroids and the eocrinoids (10).

Crown-group echinoderm fossils (Fig. 2) occur in the earliest Ordovician at about 485 Ma, in the form of primitive crinoids, another immobile filter-feeding group (11) (Figs. 2 and 3F). The remaining crown-group echinoderms have mobile life habits on and in the seafloor and as a group are termed eleutherozoans (asteroids, ophiuroids, holothuroids, and echinoids) (Fig. 3, G and H). Eleutherozoans with biomineralized stereom ossicles also first appear in the earliest Ordovician about 475 to 480 Ma (12), with the earliest occurrence of a fossil sea star (asteroid). Sea urchins (echinoids) do not occur in the fossil record until the Late Ordovician (∼450 Ma) (13) (Fig. 3H). The modern forms trace their roots back to the Late Permian, when the first cidaroid echinoids (“pencil-spined” sea urchins) appeared (14). Perhaps only two echinoid lineages survived the end-Permian mass extinction ∼252 Ma. Strongylocentrotus purpuratus, the modern sea urchin whose genome sequence is now available, is a regular euechinoid.

Echinoderm Biomineralization: Cell Biology and Genes

The process by which the biomineralized stereom skeleton is formed in echinoids is coming to light, through combined approaches of cell, molecular, and developmental biology (15). First, embryonic mesenchymal cells secrete the earliest portions of the skeleton to appear. The larval spicules (Fig. 1E) and additional independent sites in the larva are the starting points for the adult plates. The biomineral is composed of calcite (CaCO3) containing 5% MgCO3. It is secreted into an extracellular space, probably sequestered from the surrounding environment, initially as amorphous calcium carbonate, which then undergoes a regulated transition to the crystalline form (16).

Occluded within the calcite is an organic matrix of proteins that make up about 0.1% of the mass, and the birefringent optical properties of the skeletal elements result from the regular alignment of the crystals in the matrix (17). As shown in Fig. 1E (18), which portrays a skeletal element after demineralization, the triradiate physical form is a property of the matrix proteins, which originally were deposited with the biomineral. Additionally, there is an envelope of proteins around the mineral portion. Initial surveys indicate that a large number of separable proteins or protein derivatives is associated with the mineral (15). Something is known of the structure and deployment of seven of these proteins, and four have been studied in detail, namely SM50, SM30, SM37, and PM27 (15).

The multiplicity of the spicule matrix proteins is reflected by structural and functional variety within this sample, though all have a C-type lectin domain, which is a calcium-dependent carbohydrate binding motif. SM30 and SM37 are glycosylated. SM30 is known to occupy the occluded protein compartment, whereas SM50 and PM27 are found occluded and in the extracellular matrix around the spicule. SM50 contains an unusual proline- and glycine-rich repeat sequence similar to the pericardin repeat motif (19). Little is known, though, about the exact functions of these proteins except that interference with the expression of SM50 inhibits spicule formation in S. purpuratus embryos (20).

The sea urchin genome project revealed the seven known spicule matrix genes and eight new ones as well (21). Furthermore, the genome sequence provides the opportunity to observe the arrangement of the spicule matrix protein genes, and the results illuminate an aspect of their evolution. These genes occur in small clusters and thus are likely to be the related products of local gene family expansions. For example, four related SM30 genes were found to be arranged in tandem on a single assembly scaffold (21). In addition, the SM37 gene is closely linked to the SM50 gene (2123).

The S. purpuratus sequence also contains homologs of many of the signaling molecules and extracellular matrix proteins involved in vertebrate biomineralization (21). But in contrast to these, the sea urchin C-lectin spicule matrix proteins share little similarity with the well-characterized vertebrate skeletogenic proteins. Nor are they similar to any sequences present in current databases of expressed sequence tags (ESTs) from the hemichordate Saccoglossus kowalewsaki or the cephalochordate Branchiostoma floridae. With respect to any other known genome, the spicule matrix proteins of echinoids are encoded by a clade-specific set of genes. This may be true for echinoderms in general, but there is too little sequence data from other classes to make the conclusion definitive. The stereom structure is so similar among the classes that it would be remarkable if these proteins were not a character of the phylum.

The molecular and cell biology of stereom biomineralization in the sea urchin offers a fascinating glimpse into the genetic underpinnings of an echinoderm synapomorphy that arose in the Early Cambrian. A suite of identifiable unique genes (except that they have in common a domain encoding a calcium-dependent lectin) evolved to construct the unique biomineral structure of the stereom. The basic pattern of fenestration in the stereom (Fig. 1) is the property of a single differentiated cell type, defined by the expression of a battery of matrix genes, which first appeared in echinoderms at least 520 Ma.


Paleogenomics adds a genomic dimension to the paleontological description of synapomorphies. Stereom is an iconic synapomorphy for echinoderms, much as bone is for the vertebrates, and it was the first to arise in the divergence of echinoderms from the other major deuterostome lineages. Thus, we hypothesize that the evolution of the spicule matrix genes occurred after the Precambrian-Cambrian boundary (542 Ma), but before the time in the Early Cambrian when stereom-containing fossils first appear (∼520 Ma) (Fig. 2). The specific prediction is that a unique echinoderm synapomorphy, a definitive property not shared with phylogenetic sister groups, will be the genomic constituents of the calcite/stereom differentiation gene battery (Table 1). That is, echinoderms will in general share variants of the same biomineralization genes and use the same transcriptional regulatory controllers of these genes. This is of course open to experimental verification by comparative molecular analysis of biomineralization in modern forms of echinoderm. Were it found that the genetic repertoire used to produce stereom in the diverse echinoderm classes is indeed similar, then it would be indisputable that their stem-group ancestors used the same genetic apparatus.

Table 1.

Predicted features of the echinoderm biomineralization gene battery extrapolated from S. purpuratus.

Regulatory apparatus Multiple specific regulatory genes (such as Alx1, Ets1, Dr, and Hnf6) with feed-forward input into biomineralization genes (View inline)
Cellular biology Non—echinoderm-specific molecules used for secretion and motility (View inline)
Biomineralization genes Echinoderm-specific genes featuring glycine- and proline-rich repeats on the same protein with C-type lectin domains (View inline, View inline)

Paleogenomics is a knife that cuts two ways: We gain insights not only into the genes that built the structures of our fossils, but also into the evolutionary origin of the gene networks that operate in the construction of modern animal body parts. In this case, we propose that the specific stereom matrix gene battery (that is, the variety of structural functions encoded in its diverse proteins, plus its regulatory controls) must have been assembled as such in Early Cambrian time. It has remained a feature of echinoderm genomes ever since. Something is already known of the regulatory network apparatus controlling spicule matrix protein expression in the S. purpuratus embryo. The differentiation genes of the biomineralization gene battery of this embryo are together regulated by a specific small set of transcriptional control genes (24, 25).

Paleogenomic approaches can be extended to other clades for which there is both a sequenced genome and a well-preserved fossil record. The objective is the identification of clade-specific gene batteries that encode clade-specific features of the body plan. What emerges will add a time dimension to specific parts of the underlying gene regulatory networks. Thus we will be able to “age-date” portions of the functional genome to determine parts of genomes that are relatively young, in contrast to others that are extraordinarily old.

It is unusual in the consideration of body plan evolution to be able to cite a phylum-specific set of structural or differentiation genes, of which it can literally be said that their evolution underlay a phylum-specific morphological feature. The organization of body plans obviously depends in general on the regulatory control of the developmental process, which in turn depends at the genomic level on the organization of developmental gene regulatory networks. In general, therefore, the evolution of diversity of body plans depends on changes in the architecture of gene regulatory networks. But the regulatory genes constituting these networks, which encode transcription factors and intercellular signaling components, are notoriously not clade-specific: They are largely pan-bilaterian (2628), if not pan-metazoan. Similarly, the downstream differentiation genes that produce the proteins from which major components of the body plan are constructed are often not clade-specific either. For example, muscles, nervous systems, and hearts use many orthologous genes across the Bilateria. Clade-specific sets of genes that are often noted in animal genome sequences, such as genes of the immune system or smell receptor genes, are not very likely to produce signatures in the fossil record.

In temporal and historical aspects, as well as in architectural and functional terms, genetic systems for the control of development are internally inhomogeneous (27). Some subcircuits are very ancient and have changed little since their early evolution hundreds of millions of years ago; others are of more recent origin and have arisen in given evolutionary branches. This view is, of course, inconsistent with the microevolutionary presumption of temporal uniformity in evolutionary processes. The broad objectives of paleogenomics are convergent with those of “regulatory phylogenetics” [for example, gene regulatory network comparisons (29)], in that both result in the association of given genetic components with specific time-resolved evolutionary nodes. A distinction, as in the example described in this paper, is the direct relation between genes producing a structure and the fossil record, rather than the indirect relation between the regulatory genes and the body plan. It is satisfying to be able to apply genomic data directly to the origins of a character that arose over half a billion years ago and is found in animals present on Earth today.

References and Notes

View Abstract

Navigate This Article