Deuterostomes comprise vertebrates, the related invertebrate chordates (tunicates and cephalochordates) and three other invertebrate taxa: hemichordates, echinoderms and Xenoturbella. The relationships between invertebrate and vertebrate deuterostomes are clearly important for understanding our own distant origins. Recent phylogenetic studies of chordate classes and a sea urchin have indicated that urochordates might be the closest invertebrate sister group of vertebrates, rather than cephalochordates, as traditionally believed. More remarkable is the suggestion that cephalochordates are closer to echinoderms than to vertebrates and urochordates, meaning that chordates are paraphyletic. To study the relationships among all deuterostome groups, we have assembled an alignment of more than 35,000 homologous amino acids, including new data from a hemichordate, starfish and Xenoturbella. We have also sequenced the mitochondrial genome of Xenoturbella. We support the clades Olfactores (urochordates and vertebrates) and Ambulacraria (hemichordates and echinoderms). Analyses using our new data, however, do not support a cephalochordate and echinoderm grouping and we conclude that chordates are monophyletic. Finally, nuclear and mitochondrial data place Xenoturbella as the sister group of the two ambulacrarian phyla. As such, Xenoturbella is shown to be an independent phylum, Xenoturbellida, bringing the number of living deuterostome phyla to four.

We have compared the dorsoventral development of hemichordates and chordates to deduce the organization of their common ancestor, and hence to identify the evolutionary modifications of the chordate body axis after the lineages split. In the hemichordate embryo, genes encoding bone morphogenetic proteins (Bmp) 2/4 and 5/8, as well as several genes for modulators of Bmp activity, are expressed in a thin stripe of ectoderm on one midline, historically called “dorsal.” On the opposite midline, the genes encoding Chordin and Anti-dorsalizing morphogenetic protein (Admp) are expressed. Thus, we find a Bmp-Chordin developmental axis preceding and underlying the anatomical dorsoventral axis of hemichordates, adding to the evidence from Drosophila and chordates that this axis may be at least as ancient as the first bilateral animals. Numerous genes encoding transcription factors and signaling ligands are expressed in the three germ layers of hemichordate embryos in distinct dorsoventral domains, such as pox neuro, pituitary homeobox, distalless, and tbx2/3 on the Bmp side and netrin, mnx, mox, and single-minded on the Chordin-Admp side. When we expose the embryo to excess Bmp protein, or when we deplete endogenous Bmp by small interfering RNA injections, these expression domains expand or contract, reflecting their activation or repression by Bmp, and the embryos develop as dorsalized or ventralized limit forms. Dorsoventral patterning is independent of anterior/posterior patterning, as in Drosophila but not chordates. Unlike both chordates and Drosophila, neural gene expression in hemichordates is not repressed by high Bmp levels, consistent with their development of a diffuse rather than centralized nervous system. We suggest that the common ancestor of hemichordates and chordates did not use its Bmp-Chordin axis to segregate epidermal and neural ectoderm but to pattern many other dorsoventral aspects of the germ layers, including neural cell fates within a diffuse nervous system. Accordingly, centralization was added in the chordate line by neural-epidermal segregation, mediated by the pre-existing Bmp-Chordin axis. Finally, since hemichordates develop the mouth on the non-Bmp side, like arthropods but opposite to chordates, the mouth and Bmp-Chordin axis may have rearranged in the chordate line, one relative to the other.

Neuroectodermal signalling centres induce and pattern many novel vertebrate brain structures but are absent, or divergent, in invertebrate chordates. This has led to the hypothesis that signalling centre genetic programs were first assembled in stem vertebrates, which potentially drove morphological innovations. However, this scenario presumes that extant cephalochordates accurately represent ancestral chordate characters, which has not been tested using close chordate outgroups. Here, we report that genetic programs homologous to three vertebrate signalling centres; the anterior neural ridge, zona limitans intrathalamica, and isthmic organizer are present in the hemichordate Saccoglossus kowalevskii. Fgf8/17/18, sfrp1/5, hh, and wnt1 are expressed in vertebrate-like arrangements in hemichordate ectoderm, and homologous genetic mechanisms regulate ectodermal patterning in both animals. We propose these genetic programs were components of an unexpectedly complex, ancient genetic regulatory scaffold for deuterostome body patterning that degenerated in amphioxus and ascidians, but was retained to pattern divergent structures in hemichordates and vertebrates.

While some aspects of the phylogeny of the five living echinoderm classes are clear, the position of the ophiuroids (brittlestars) relative to asteroids (starfish), echinoids (sea urchins) and holothurians (sea cucumbers) is controversial. Ophiuroids have a pluteus-type larva in common with echinoids giving some support to an ophiuroid/echinoid/holothurian clade named Cryptosyringida. Most molecular phylogenetic studies, however, support an ophiuroid/ asteroid clade (Asterozoa) implying either convergent evolution of the pluteus or reversals to an auricularia-type larva in asteroids and holothurians. A recent study of 10 genes from four of the five echinoderm classes used 'phylogenetic signal dissection' to separate alignment positions into subsets of (i) suboptimal, heterogeneously evolving sites (invariant plus rapidly changing) and (ii) the remaining optimal, homogeneously evolving sites. Along with most previous molecular phylogenetic studies, their set of heterogeneous sites, expected to be more prone to systematic error, support Asterozoa. The homogeneous sites, in contrast, support an ophiuroid/echinoid grouping, consistent with the cryptosyringid clade, leading them to posit homology of the ophiopluteus and echinopluteus. Our new dataset comprises 219 genes from all echinoderm classes; analyses using probabilistic Bayesian phylogenetic methods strongly support Asterozoa. The most reliable, slowly evolving quartile of genes also gives highest support for Asterozoa; this support diminishes in second and third quartiles and the fastest changing quartile places the ophiuroids close to the root. Using phylogenetic signal dissection, we find heterogenous sites support an unlikely grouping of Ophiuroidea þ Holothuria while homogeneous sites again strongly support Asterozoa. Our large and taxonomically complete dataset finds no support for the cryptosyringid hypothesis; in showing strong support for the Asterozoa, our preferred topology leaves the question of homology of pluteus larvae open.

The evolutionary origins of the chordate central nervous system remain uncertain. Conclusions drawn from classical morphological comparisons and from a broad range of metazoan phyla conflict with the new molecular genetic information from developmental model systems characterized by central nervous systems. This has led to debate as to the nature of the ancestral deuterostome nervous system. Hemichordates as basal deuterostomes occupy a phylogenetically critical position for addressing hypotheses on the evolution of the chordate nervous system. Characterizing the molecular basis of the development of their diffuse nervous system offers insights into the role of conserved body patterning genes in the evolution of specific neural anatomies. We present a description of hox gene expression during the development of the direct-developing hemichordate Saccoglossus kowalevskii. The nested ectodermal expression of these genes in a hemichordate suggest that they play a general patterning role in the anterior/posterior regionalization of the ectoderm of bilaterians rather than being uniquely associated with the development and evolution of central nervous systems.