Molecular genetic insights into deuterostome evolution from the direct-developing hemichordate Saccoglossus kowalevskii

Progress in developmental biology, phyloge- nomics, and palaeontology over the past 5 years has made major contributions to a long-enduring problem in comparative biology: the early origins of the deuterostome phyla. A detailed character- ization of the early development of the enter- opneust hemichordate Saccoglossus kowalevskii has revealed close developmental genetic simi- larities between hemichordates and chordates during early body plan formation. The two phyla share close transcriptional and signalling lig- and expression patterns during the early devel- opment of the anteroposterior and dorsoventral axes, despite large morphological disparity between the body plans. These genetic networks have been proposed to play conserved roles in patterning centralized nervous systems in meta- zoans, and probably play conserved roles in pat- terning the diffusely organized basiepithelial nerve net of the hemichordates. Developmental genetic data are providing a unique insight into early deuterostome evolution, revealing a com- plexity of genetic regulation previously attrib- uted only to vertebrates. Although these data allow for key insights into the development of early deuterostomes, their utility for recon- structing ancestral morphologies is less certain; morphological, palaeontological and molecu- lar data sets should all be considered carefully when speculating about ancestral deuterostome features.
10.1 Introduction

The deuterostome phyla form one of the major animal lineages of the bilaterians (Hyman, 1940; Brusca and Brusca, 1990). The evolutionary history of this group has been the subject of debate for over a century (Gee, 1996). The composition of deuteros- tomes has been in a state of flux since the advent of molecular systematics, making attempts to recon- struct the early history of the group very difficult. However, the bilaterian phyla belonging within the deuterostomes are now largely known (Turbeville et al., 1994; Halanych, 1995; Halanych et al., 1995; Bromham and Degnan, 1999; Cameron et al., 2000; Bourlat et al., 2003, 2006; Dunn et al., 2008). The following four phyla make up the deuterostomes: chordates, hemichordates, echinoderms, and xeno- turbellids.
Despite increased confidence in the relationships between the major deuterostome phyla, our under- standing of early deuterostome body plan evo- lution remains quite murky. There are two main factors that contribute to this uncertainty: a poor fossil record (Swalla and Smith, 2008) and a large morphological disparity between the body plans of the four phyla. Both of these factors make the reconstruction of ancestral features of early deuter- ostomes particularly challenging. This chapter will focus on the molecular genetic data from hemi- chordates that facilitate more direct comparisons with the chordate body plan and provide novel insights into the genetic networks that must have


93

94 AN I M AL EV O L UTI O N



been present in the common ancestor of all deuter- ostome phyla. First, I begin with a general introduc- tion to the deuterostome phyla and the challenges associated with reconstructing early deuterostome evolution. Second, I summarize the molecular gen- etic information, most recently generated from enteropneusts, involved in the anteroposterior and dorsoventral patterning of hemichordates. Finally, I will discuss what sort of insights can be gained from molecular genetic data sets and their utility for testing both general axial or organizational homologies and more traditional morphological homologies.


10.2 Problems in the reconstruction of ancestral deuterostome
characters

One of the most significant barriers to under- standing the evolution of early deuterostome evo- lution has been the difficulty of making direct comparisons between the adult body plans of the four deuterostome phyla: chordates, echinoderms, hemichordates, and xenoturbellids. There are few uncontested deuterostome synapomorphies and a poor early deuterostome fossil record, making attempts to reconstruct an ancestral deuterostome body plan difficult. Echinoderms typify the diffi- culties of body plan comparison across the deu- terostome phyla; the adult body is perhaps the most radical morphological departure of any of the bilaterian groups (Lowe and Wray, 1997). Extant species have lost ancestral bilateral symmetry and have become pentaradially symmetric as adults, while maintaining a bilaterally symmetric larva. There are two novel mesodermally derived struc- tures that are key components of their unusual body plan: the calcitic endoskeleton and water vas- cular system (Hyman, 1940). Their nervous system is largely diffuse and organized as a basiepithelial nerve net, with some evidence of integrative abilities in the radial nerves (Bullock and Horridge, 1965). Even gross axial comparisons between extant echi- noderms and other deuterostomes are problematic (Lowe and Wray, 1997) and it is not clear whether valid comparisons can be made to the anteroposter- ior and dorsoventral axes of the bilaterian groups. However, early stem-group fossil echinoderms give
some clues to their early bilaterian origins. These
fossils show evidence of gill slits (Dominguez et al.,
2002) and possibly even a muscular stalk (Swalla
and Smith, 2008). There are a few molecular gen-
etic studies that give insights into these questions
(Peterson et al., 2000; Morris et al., 2004; Morris and
Byrne, 2005; Hara et al., 2006), but many more data
are required before any strong comparative conclu-
sions can be drawn.
The most recent addition to the deuterostomes,
xenoturbellids (Bourlat et al., 2003, 2006; Dunn
et al., 2008) are morphologically rather unremark-
able: they have a ventral mouth, a blind gut, and
little in the way of external morphological features.
They do share the general organizational features
of the hemichordate nervous system (Pedersen and
Pedersen, 1986), but little else currently described
in their anatomy could be referred to as a deutero-
stome synapomorphy. There are still very few pub-
lished studies of their biology, though preliminary
developmental studies suggest that embryos are
brooded (Israelsson and Budd, 2005). However, it
is difficult to make any strong comparative conclu-
sions based on current data and further study is
needed, particularly in characterizing the develop-
ment of this animal.
Hemichordates are perhaps the most promis-
ing of the non-chordate deuterostome groups for
addressing issues of both early deuterostome evo-
lution and the evolution of the chordate body plan
(Cameron et al., 2000; Tagawa et al., 2001; Lowe et al.,
2003). The phylum is divided into two classes: the
enteropneust worms and the pterobranchs. Both
groups possess a similar tripartite body organ-
ization, but are characterized by distinct feeding
mechanisms. The pterobranchs are often small,
colonial animals, and feed using a lophophore—a
ciliated extension of the mesosome (Halanych,
1995). Enteropneusts are larger solitary animals,
and use their highly muscular, ciliated proboscis
(prosome) for direct particle ingestion and filter
feeding (Cameron, 2002). I will focus exclusively on
the body plan of enteropneusts, as there are cur-
rently few data on the body patterning of ptero-
branchs (Sato and Holland, 2008). The most recent
molecular phylogenies describe two main groups
of enteropneusts: the Harrimaniidae in one lin-
eage and the Ptychoderidae and Spengelidae on

D E V E LO P M EN T A L B I O L O GY O F SA C C OG L O S S U S 95



the other (Cameron et al., 2000). These two lineages have major life-history differences: harrimaniids are all direct developers, whereas the spengelids and ptychoderids are indirect developers with feeding larvae, which often spend many months in the plankton before metamorphosing into juven- iles (Lowe et al., 2004).
Phylogenetic relationships of the various hemi- chordate groups remain poorly resolved, and this area is in need of further research. Pterobranchs have traditionally been considered as basally branching hemichordates, based largely on the proposed homology of its lophophore with that in other lophophorate groups. However, reclassificat- ion of the lophophorates as protostomes reveals that the structural similarities of lophophores are due to convergence rather than homology (Halanych et al. 1995). Further molecular phylo- genetic studies have proposed that pterobranchs are perhaps nested within the enteropneusts (Cameron et al., 2000; Winchell et al., 2002), but this is weakly supported by current data sets. Clearly this issue should be revisited with broader phylo- genetic sampling.
Figure 10.1 outlines some of the main anatom- ical features of enteropneusts and shows a photo- micrograph of a juvenile worm of the harrimaniid Saccoglossus kowalevskii. The tripartite body plan is divided into an anterior prosome or proboscis, a mesosome or collar, and a metasome or trunk. The proboscis is muscular, ciliated, and highly inner- vated with sensory neurons (Bullock, 1945; Knight- Jones, 1952), and its primary functions are digging and feeding. The mouth opens up on the ventral side and marks the boundary between the probos- cis and the collar. In the most anterior region of the trunk, dorsolateral gill slits perforate the ecto- derm. The gill slits can be very numerous and con- tinue to be added as the animal grows (Bateson,
1885; Hyman, 1940). At the very far posterior end of the metasome, a ventral extension, sucker, or tail, extends ventrally from the anus and is used for locomotion by the post-hatching juvenile worm. A post-anal extension is present only in the juvenile of the harrimaniids, but not in other enteropneust groups, and is lost in adult worms. Given the uncer- tainty over the relationships of the major groups within the hemichordates, the possible homology





Prosome/proboscis Mesosome/collar






Mouth



Gill slits
Metasome/
trunk




Post-anal “tail” Anus
Figure 10.1 Organization of the adult body plan of enteropneusts. Light micrograph of a juvenile worm of the harrimaniid enteropneust Saccoglossus kowalevskii at day 13 of development. All major body regions (prosome, mesosome, and metasome) are well developed and several gill slits are perforated in the anterior metasome. The juvenile post-anal tail is still present, but is eventually lost in adult animals.

96 AN I M AL EV O L UTI O N



of this extension to the pterobranch stalk, and even more controversially, the chordate post-anal tail, remains unresolved (Cameron et al., 2000; Swalla and Smith, 2008).
The earliest descriptions of hemichordate anat- omy by Bateson (1884, 1885, 1886) and Morgan (1891, 1894) resulted in various hypotheses of mor- phological homologies between hemichordates and chordates (Bateson, 1886; Morgan, 1891; Nübler- Jung and Arendt, 1996). Most of these classical hypotheses are largely unsupported by both mor- phological and molecular data sets (Peterson et al.,
1999; Nieuwenhuys, 2002; Long et al., 2003; Ruppert,
2005; Lowe et al., 2006). However, only gill slits, as
primary ciliated pouches, may represent an ances-
tral feature of the deuterostomes (Ogasawara et al.,
1999; Okai et al., 2000; Tagawa et al., 2001; Cameron,
2002; Rychel et al., 2006; Rychel and Swalla, 2007)
and possibly the post-anal tail (Lowe et al., 2003). A
more in-depth discussion of potential hemichord-
ate and chordate morphological homologies is
found in Ruppert (2005). Morphological homology
aside, there has been little consensus over how
to compare the body plans of hemichordates and
chordates, even at a basic axial level.


10.3 The potential of molecular genetic data for providing insights into deuterostome evolution

Although establishing robust morphological hom- ologies between deuterostome groups is problem- atic, progress in developmental biology over the past 20 years, mainly from studies of arthropods and chordates, has allowed unprecedented axial comparisons between distantly related groups (Gerhart and Kirschner, 1997; Carroll, 2005). However, representation of the deuterostome adult body plans in broad metazoan comparative stud- ies has been dominated by chordate developmen- tal biology. There is an impressive literature on early embryonic and larval patterning in echino- derms, but only a handful of studies on the devel- opment of the adult body plan (Lowe and Wray,
1997; Ferkowicz and Raff, 2001; Lowe et al., 2002; Sly et al., 2002; Morris et al., 2004; Morris and Byrne,
2005; Hara et al., 2006). Recent studies in hemichor-
dates have revealed a novel way to compare the
adult body plans of chordates and hemichordates
(Okai et al., 2000; Harada et al., 2002; Taguchi et al.,
2002; Lowe et al., 2003, 2006). I will introduce these
comparative data sets and their role in compar-
ing anteroposterior and dorsoventral patterning
of bilaterian groups. I will then review the current
developmental genetic work from hemichordates
and how this impacts upon our understanding of
early deuterostome evolution.
Although it is now widely accepted that many of
the developmental regulatory cascades controlling
anteroposterior and dorsoventral axial patterning
during nervous system development are probably
homologous as regulatory modules (Gerhart and
Kirschner, 1997; Carroll, 2005; Davidson, 2006),
the extent to which the these data are effective for
testing hypotheses of morphological homology
remains controversial (Arendt and Nübler-Jung,
1996; De Robertis and Sasai, 1996; Lowe et al., 2003,
2006; Lichtneckert and Reichert, 2005; Denes et al.,
2007). Most studies have focused on similarities
in nervous system patterning along both dorso-
ventral and anteroposterior axes (Hirth et al., 2003;
Acampora et al., 2005; Lichtneckert and Reichert,
2005). Classical morphological comparisons gen-
erally converged on the hypothesis that the cen-
tral nervous systems of arthropods and chordates
evolved independently and that early bilaterian ner-
vous systems were generally quite simple (Holland,
2003). More recent interpretations of the molecular
genetic data, based on model systems with central
nervous systems, lead to quite different conclusions
and propose a protostome–deuterostome ancestor
with a complex, centralized nervous system with a
regionalized brain. The homologous suites of genes
involved in the patterning of the central nervous
systems of model systems have very similar spatial
domains during development (Arendt and Nübler-
Jung, 1994, 1996; Finkelstein and Boncinelli, 1994;
Sharman and Brand, 1998; Hirth et al., 2003). Along
the anteroposterior axis, the Hox genes are involved
in patterning the nerve cords of both arthropods
and chordates. The boundary of the trunk and the
rest of the anterior nervous system is marked by
the homeobox gene gbx or unplugged (Hirth et al.,
2003; Castro et al., 2006). Gbx is expressed at the
boundary between Hox genes and otx and marks a
morphological transition in the organization of the

D E V E LO P M EN T A L B I O L O GY O F SA C C OG L O S S U S 97



nervous system. Other anteriorly localized homeo- box genes such as orthodenticle (otx), pax6, distalless (dlx), emx, and retinal homeobox (rx) play conserved roles in brain patterning and exhibit similar rela- tive spatial localization during the development of the central nervous system (Lowe et al., 2003).
On the dorsoventral axis, early patterning events of the ectoderm are similarly conservative between arthropods, vertebrates, and annelids (Arendt and Nübler-Jung, 1994, 1996; De Robertis and Sasai,
1996; Holley and Ferguson, 1997; Cornell and Ohlen,
2000; Denes et al., 2007). The secreted factor chor-
din/short gastrulation is released dorsally in verte-
brates and ventrally in arthropods, protecting the
ectoderm from the neural-inhibiting effects of the
transforming growth factor (TGF)-E ligand Bmp,
which is expressed ventrally in vertebrates and
dorsally in flies. The interaction of these secreted
ligands results in the formation of the central ner-
vous system on the dorsal side in vertebrates and
the ventral side in arthropods and annelids. These
data have revived the venerable dorsoventral axis
inversion hypothesis of Dohrn (1875), which pro-
posed that the dorsoventral organization of chor-
dates is best explained by a complete body axis
inversion in the lineage leading to chordates.
Further similarities have been revealed in later
dorsoventral pattering of the neurectoderm of all
three groups, but most closely between the anne-
lid Platynereis dummerilii and vertebrates (Denes
et al., 2007). The similarities in relative expres-
sion domains and essential functions along both
the dorsoventral and anteroposterior axes during
central nervous system patterning have led to the
proposal that a tripartite brain is ancestral for bilat-
erians, implying that the protostome/deuterostome
ancestor and early deuterostomes were character-
ized by a complex, centralized nervous system.
Within a purely phylogenetic framework, based
on current molecular phylogenies, the outgroups
to the bilaterians are acoel flatworms and cnidar-
ians, both of which are characterized by a nerve net
(Holland, 2003). Within the bilaterians, the basie-
pithelial nerve net is quite common. Proposing a
protostome/deuterostome ancestor with a complex
central nervous system and brain implies that this
organization has been lost multiple times during
the evolution of the bilaterian phyla.
Until recently, the vast majority of data gener- ated in developmental biology have been from terrestrial model systems that are characterized by central nervous systems. This bias has begun to be addressed by a broad description of the gen- etic information involved in patterning the hemi- chordate enteropneust body plan, and its diffuse basiepithelial nerve net (Lowe et al., 2003, 2006). The organization of the nerve net is based around a broad distribution of cell bodies throughout the ectoderm. Despite the general diffuse organization of the nervous system, there is a significant dorso- ventral and anteroposterior polarity in its structure and organization, particularly in the dorsoventral dimension (Bullock, 1945; Knight-Jones, 1952). A mat of axons spreads out along the basement mem- brane, which is thickened in certain areas of the ectoderm—at the base of the proboscis, along the anterodorsal region of the body in the mesosome, and in both the dorsal and ventral midlines of the metasome. In the proboscis ectoderm, there is a dense concentration of nerve cells that have been proposed to be primarily sensory (Bullock, 1945; Knight-Jones, 1952) and are particularly thick at the base of the proboscis (Brambell and Cole, 1939).
Probably the most well known aspect of hemi- chordate anatomy is the mid-dorsal region of the dorsal cord, or collar cord, which is internalized into a hollow tube of epithelium in some species within the Ptychoderidae and in one species of the Spengelidae. However, in the Harrimaniids, there is no contiguous hollow tube, but scattered blind lacunae (Nieuwenhuys, 2002; Ruppert,
2005). This structure has been widely compared to the dorsal cord of chordates, not only due to the superficial similarities of the hollow nerve cord, but also because in some species, the collar cord forms by a process that resembles chordate neu- rulation (Morgan 1891). However, the similarities have generally been over-emphasized as it seems to be more of a conducting tract than a process- ing centre (Ruppert, 2005) as evidenced by both ultrastructural (Dilly et al., 1970) and physiological data (Pickens, 1970; Cameron and Mackie, 1996). Another striking feature of the dorsal cord is the presence of giant axons in some species. The cell bodies project their axons across the midline and continue posteriorly within the collar cord. It is

98 AN I M AL EV O L UTI O N



not known where the axons finally project; Bullock (1945) proposed that they innervate the ventrolat- eral muscles of the trunk and suspected that their primary function is to elicit a rapid contraction of these muscles. However, several groups do not possess giant axons and yet are still able to elicit a rapid retreat, so the role of the giant axons remains uncertain (Pickens, 1973).
In the metasome, the third body region, the most prominent features are the ventral and dor- sal nerve cords, which are both thickenings of the nerve plexus. The dorsal cord is contiguous with the collar cord and projects down the entire length of the metasome. The ventral cord is comparatively much thicker, with more associated cell bodies, but both cords are largely described as through axon tracts. However, at least one study describes the ventral cord as having some integrative func- tion (Pickens, 1970). It seems to play a role in rapid retreat of the animals following anterior stimula- tion (Knight-Jones, 1952; Bullock and Horridge,
1965; Pickens, 1973).


10.4 Anteroposterior patterning in hemichordates

Molecular genetic patterning information in hemi- chordates has the potential to address two major areas of comparative interest. First, these data could be another means to compare deuterostome body plans, giving insights into early deuterostome evolution. Second, hemichordates are representa- tive of the first basiepithelial nervous system to be characterized molecularly and allow insights into whether the complex networks of regulatory genes involved in patterning complex central nervous systems play similar roles in less complex, more diffusely organized, nervous systems. These ques- tions have been the focus of several papers over the past 10 years investigating the roles of body pat- terning genes in hemichordates. The first suite of papers focused on Ptychodera flava, an indirect-de- veloping species with a ciliated feeding larva and an extended planktonic larval period: most of these initial studies focused on the establishment of the larval body plan (Dohrn, 1875; Peterson et al., 1999; Harada et al., 2000, 2002; Okai et al., 2000; Tagawa et al., 2001; Taguchi et al., 2002). More recently, we
have developed a direct-developing species, the harrimaniid enteropneust S. kowalevskii, to investi- gate more directly the pattering of the adult rather than larval body plan of hemichordates.
In the first of two papers on body patterning, we investigated anteroposterior patterning (Lowe et al., 2003) by examining the expression of ortho- logues of 22 transcription factors that have con- served roles in the patterning of the brain and spinal cord of vertebrates along the anteroposterior axis. At least 14 of these genes also play conserved roles in the patterning of the central nervous sys- tem of the fruit fly Drosophila melanogaster. The
22 transcription factors can be divided into three broad expression and functional domains during the development of the vertebrate brain and cen- tral nervous system: (1) expressed and involved in forebrain development, (2) expressed during mid- brain development, and (3) expressed and func- tionally involved in the patterning of the hindbrain and spinal cord. In the first category a group of six transcription factors—six3, brain factor 1(bf-1), distal- less (dlx), nk2–1, ventral anterior homeobox (vax), and retinal homeobox (rx)—were all expressed in simi- lar domains during the early development of the embryo and juvenile (Figure 10.2). Their expression was restricted, for the most part, to the develop- ing proboscis ectoderm, the most anterior region of ectoderm. Unlike vertebrates and panarthropods, this expression is not restricted to the dorsal or ventral side, but rather forms rings encircling the entire dorsoventral aspect of the animal reflecting the inherent diffuse organization of the basiepithe- lial nerve net.
Vertebrate orthologues of the second group of genes are expressed with the posterior limit of expression marking the midbrain, and sometime hindbrain of vertebrates, including pax6, tailless (tll), barH, emx, orthopedia (otp), dorsal brain homeo- box (dbx), lim1/5, iroquois (irx), orthodenticle (otx), and engrailed (en). Similar to vertebrates, this group of genes is expressed in a more posterior position along the anteroposterior axis of the developing hemichordate embryo, in the posterior proboscis, collar, and anterior trunk. Of these genes, en is particularly interesting as it forms a sharp single ring of expression in the ectoderm of the anterior metasome over the forming first gill slit. En is a

D E V E LO P M EN T A L B I O L O GY O F SA C C OG L O S S U S 99


Forebrain Midbrain Hindbrain/spinal cord Tail















vax, nk2-1, rx, dlx, bf-1, otp
otx, pax6, emx barH, dbx, otp irx, lim1/5, en

gbx, hox 1, 2, 3,
4, 5, 6/7, 9/10 hox 11/13 a, b, c














Prosome/proboscis Mesosome/collar Metasome Tail

Figure 10.2 Summary of similarities between the enteropneust hemichordate Saccoglossus kowalevskii and vertebrates in the ectodermal expression of conserved transcriptional developmental regulatory genes. The upper panel represents an idealized vertebrate, and the bottom panel represents a juvenile hemichordate. The various shades of grey represent similarities in gene expression between the two groups.




critical gene in the formation of the vertebrate isth- mus. This then makes a compelling case for inves- tigating other genes involved in the formation of the midbrain/hindbrain division of the vertebrate brain and how much of this signalling regula- tory cassette is conserved, as most studies of basal chordates have suggested that ectodermal signal- ling centres evolved in association with complex vertebrate neural anatomy (Canestro et al., 2005).
The last group of genes includes gbx and Hox genes. In vertebrates, the regulatory interaction between otx and gbx is involved in positioning the isthmus along the anteroposterior axis, with gbx expressed posterior to otx (Rhinn et al., 2005). In the cephalochordate amphioxus, gbx is also
expressed in a mutually exclusive domain to otx in the central nervous system, suggesting a con- served interaction between the two genes (Castro et al., 2006). However, in ascidians gbx is absent from the genome. In S. kowalevskii, we observe a departure from chordates in that otx and gbx expression overlap extensively at all stages of development examined, suggesting that they do not share the same mutual antagonism as found in vertebrates.
A total of 11 Hox genes have now been cloned from S. kowalevskii (Aronowicz and Lowe, 2006). A study of the relative order of Hox genes in the gen- ome has now been completed and will be reported elsewhere (Gerhart et al., work in preparation).

100 AN I M AL EV O L UTI O N



Hox expression domains follow predictable nested domains with the most anterior Hox genes expressed in the most anterior regions of the meta- some ectoderm, and the most posterior members in the most posterior domains. At the stages that were examined, expression of many of the genes was tightly grouped, with little evidence of difference in anterior expression limits. Expression has not been examined for all genes at late developmental stages when the trunk begins to elongate and become fur- ther regionalized. Perhaps the anterior expression limits of Hox genes become more markedly differ- entiated in later stages. Posterior Hox family mem- bers were examined at later stages when the ventral post-anal tail was developing, and the expression of these posterior members was restricted to the post-anal tail in juveniles, which is similar to the expression of their orthologues in the dorsal post- anal tail of vertebrates. These data would support the proposed homology of the chordate and enter- opneust post-anal tails, although it is certainly also possible that independently evolved posterior extensions are likely to express posterior Hox genes already expressed in the posterior ectoderm.
A summary of the data from the development of the anteroposterior axis of the hemichordates is illustrated in Figure 10.2. These data clearly dem- onstrate similar relative expression of transcrip- tion factors with critical roles in anteroposterior patterning between chordates and hemichordates. Although most comparative studies and specula- tions on the nervous system of hemichordates have focused on the dorsal and ventral axon tracts as potential homologues of chordate central neural structures, the results from this study suggest, as was proposed by Bullock in 1945, that the appro- priate comparison is with the entire net. The cords are probably local thickenings of the nerve plexus rather than integrative centres (Dilly et al., 1970). The conclusions one can draw from these data are more complicated—particularly, whether these data can help to reconstruct early morphological evolution of deuterostomes. First, it is most parsi- monious to conclude that the similarities in the rela- tive expression domains of multiple genes between hemichordates and chordates are due to homology of a gene regulatory network. It is highly unlikely that all the similarities of gene expression along
the anteroposterior axes of both groups are a result of co-option of individual genes into convergently similar domains. However, the organizational difference in the nervous systems of both groups suggests that, despite this regulatory conservation, the evolutionary possibilities of the downstream morphologies have not been constrained. The ner- vous system, in particular, demonstrates this point effectively: the development of both the central nervous system of chordates and the basiepithelial nerve net of hemichordates is probably regulated by this conserved regulatory map (although it is important to note that this was not directly tested in Lowe et al., 2003). Clearly, the forebrain of ver- tebrates and the proboscis of hemichordates are not homologous structures. This suite of genes is not a reliable marker of morphological homology between groups.
By considering these data alone, we can specu- late that the deuterostome ancestor, and also the protostome/deuterostome ancestor, may have been characterized by a completely diffuse or fully cen- tralized nervous system, and all possible inter- mediates. Reconstructing ancestral morphologies from gene expression data can be problematic, even with such large expression data sets. These data, however, do give a unique insight into the antero- posterior patterning of the deuterostome ancestor, revealing a degree of transcriptional complex- ity previously attributed to the complex nervous system of vertebrates. Finally, the nervous system of hemichordates has been described as barely more complex than the cnidarian nervous system (Bullock, 1945; Bullock and Horridge, 1965) and yet there is an exquisite level of transcriptional pat- terning in the ectoderm. This may suggest a level of neural diversity currently not recognized in this group. Perhaps the complexity of the basiepithelial net of the hemichordates has been underestimated and would benefit from a modern approach to describing the neural diversity? Detailed physio- logical and molecular studies would be required to address this hypothesis.


10.5 Dorsoventral patterning

Hemichordates have a distinctive and marked dorsoventral axis. The mouth opens on the ventral

D E V E LO P M EN T A L B I O L O GY O F SA C C OG L O S S U S 101



side by convention, and the most obvious dorsal markers are the paired dorsolateral gill slits. The stomochord is an anterodorsal projection from the gut supporting the axial complex or heart and kid- ney complex. As previously discussed, the nerve net also exhibits marked dorsoventral polarity in the distribution of dorsal and ventral cords, and the presence of giant axons in the dorsal cords. The TGF-E signalling ligand, Bmp, and one of its antagonists, chordin, are involved in establishing the dorsoventral developmental axis in arthropods and chordates. This molecular axis has recently been investigated in S. kowalevskii (Lowe et al., 2006). Hemichordates occupy a key position for investigat- ing the evolution of this developmental pathway in dorsoventral patterning of the bilaterians (Nübler- Jung and Arendt, 1996; Lowe et al., 2006).
The most striking feature of bmp and chordin between vertebrates and arthropods is that their relative expression is inverted dorsoventrally with respect to each other (Arendt and Nübler- Jung, 1994, 1996; De Robertis and Sasai, 1996). In hemichordates, bmp2/4 is expressed along the dorsal midline throughout all stages of devel- opment, along with all the members of the Bmp synexpression group (Niehrs and Pollet, 1999; Karaulanov et al., 2004). At early developmental stages, chordin is expressed ventrally and very broadly on the opposite side to bmp2/4, almost up to the dorsal midline, but is increasingly restricted to the ventral side as development progresses. There are many genes that exhibit marked dorsoven- trally restricted expression domains along either dorsal or ventral midlines in ectoderm (tbx2/3, dlx, olig, netrin, pitx, poxN, lim3, admp, sim), endoderm (mnx, admp, sim, nk2.3/2.5), and mesoderm (mox/ gax). These data reveal a molecular dorsoventral asymmetry that perhaps underlies the morpho- logical asymmetry along this axis. Although in hemichordates the expression of chordin and bmp, in relation to the dorsoventral axis, is similar to protostomes, the early developmental action of Bmp and chordin does not result in segregation of a central nervous system from the general ecto- derm: there is no central nervous system, but a dif- fuse and broadly distributed nerve net.
What is the early role of Bmp in an animal with- out a non-neural ectoderm? Over-expression and
knockdown analyses addressed this question, and two main conclusions were presented; first, over- expression of Bmp did not result in the repression of neural fates; second, bmp plays a central and critical role in dorsoventral patterning (Lowe et al.
2006). In embryos incubated with recombinant vertebrate Bmp4 protein, endogenous hemichord- ate bmp2/4 expression was activated throughout the ectoderm, rather than localized along the dor- sal midline as in normal embryos. These treated embryos do not perforate a mouth, and with high levels of Bmp protein do not perforate gill slits. Additionally, in the endoderm, dorsolateral endo- dermal pouches, precursors to the gill slits, do not form, and the entire endoderm projects into the protocoel rather than a thin dorsal projection that would normally develop into the stomochord. Knockdown or diminished expression of bmp2/4 by injection of short-interfering RNAs (siRNAs) resulted in a complementary phenotype, particu- larly in relation to the mouth, which normally per- forates on the ventral side. In injected embryos, the mouth develops circumferentially, and eventually results in the detachment of the entire prosome.
The morphological interpretation of Bmp modu- lation experiments suggests that over-expression of Bmp ‘dorsalizes’ embryos, and knockdown of Bmp ‘ventralizes’ embryos. This was confirmed by further molecular analysis: markers of the dor- sal midline, in both the ectoderm and endoderm, expanded into circumferential rings in Bmp lig- and-treated embryos, suggesting that in normal embryos they are activated by Bmp signalling on the dorsal midline. Some of the same dorsal mark- ers failed to activate expression following siRNA injection, adding further support for a role of Bmp in patterning dorsal cell fates. Further experimen- tal evidence suggested that Bmp is involved in restricting the expression of ventrally expressed genes to the ventral midline, as Bmp ligand-treated embryos failed to express ventral markers, and these same markers expand to the dorsal side in siRNA-injected embryos.
The major differences between the hemichordates and vertebrates are summarized by two major cri- teria. First, in the disposition of the Bmp/chordin axis, which is inverted (Figure 10.3): hemichordates more closely resemble the protostomes with chordin

102 AN I M AL EV O L UTI O N



Vertebrate
D

N
Hemichordate
D
G

Bmp
Annelid
D

Bmp
Drosophila
D






Bmp
G
M M M M


Chordin/sog Bmp/dpp Endoderm



Figure 10.3 Expression of bmp/chordin in bilaterians. The diagrams represent idealized cross-sections through embryos, with dorsal (D) oriented up and ventral down. In vertebrates, the source of chordin is largely the notochord, here a grey circle marked with ‘N’, dorsal to the gut. The mouth in all panels is represented by M, and in hemichordates and vertebrates G represents the position of the gill slits. In annelids, the ventral, light grey colour represents the predicted domain of chordin expression, as this has yet to be published. The black symbol under the vertebrate model represents a potential dorsoventral axis inversion on the lineage leading to chordates.




expressed ventrally and bmp dorsally. Second, the mouth opens on the side of the embryo expressing chordin in hemichordates and other protostomes, but in the bmp domain in vertebrates. Functional experiments in hemichordates suggest that the Bmp/chordin axis is fundamental for the develop- ment of many components of the dorsoventral axis, and particularly important for the formation of the mouth. Lastly, although Bmp is directly involved in repressing neural fates in the developing epidermis of vertebrates, it plays no role in repressing neural cell fate in hemichordates, as neural markers are not down-regulated following Bmp4 treatment of embryos. This is also not surprising based on pre- vious descriptions of the distribution of neural cell bodies, which are present along the dorsal midline where Bmp is normally expressed.
How can the differences between hemichord- ates and vertebrates be explained, and do these data give any critical insights into early deuter- ostome evolution? One way to explain the data is partially to accept the basic model of inversion as proposed by Dohrn (1875). However, the modifi- cation to the model is that a hypothetical ances- tor was not necessarily characterized by a central
nervous system: the data from S. kowalevskii dem- onstrate that a dorsoventrally distributed Bmp/ chordin axis, although fundamentally involved in dorsoventral patterning, is not always linked to the formation of a central nervous system. Therefore, issues of inversion and centralization can be uncoupled and considered separately. It is formally possible that inversion of an animal with a diffuse nervous system gave rise to the chord- ates, and centralization happened secondarily. Following inversion, the definitive chordate mouth must have either migrated from the dorsal side or a new mouth formed de novo. The new mouth of ver- tebrates seems to have a novel relationship to the Bmp/chordin axis as it forms in a region of Bmp expression, which in hemichordates inhibits the formation of the mouth.


10.6 Life-history considerations

There has been a diverse range of hypotheses to explain the early evolution of our phylum. Some of the most influential of these can be roughly divided into two kinds; ones that derive the chordate body plan from a larval life-history

D E V E LO P M EN T A L B I O L O GY O F SA C C OG L O S S U S 103



stage, and others from the adult life-history stage (Gee, 1996). The data I have presented from the direct-developing hemichordate exhibit extensive similarities with the adult body plan of verte- brates. I would argue that the extensive molecu- lar similarities in patterning between adult body plans would argue strongly for adult life-history origins of chordates. However, there remain many supporters of larval origins of the chordate body plan (Nielsen, 1999; Tagawa et al., 2000, 2001; Poustka et al., 2004).
Walter Garstang proposed his Auricularian hypothesis over a century ago (Garstang, 1894,
1928), yet it remains as one of the most compelling of the plethora of hypotheses presented to explain the origins of the chordate body plan. The hypoth- esis derives the chordate dorsal nerve cord from a larval life-history stage by a dorsal migration and fusion of the ciliated bands, hypothesized to resem- ble that of many extant enteropneust and echino- derm species. Given the extensive similarities in the anteroposterior and dorsoventral patterning of many bilaterian groups, if chordates are derived from a larval life-history stage then we may expect to see some evidence of conserved regulatory net- works during larval development, typical in the adult chordate body plan. There are some limited similarities, but so far the evidence is far from con- vincing. Here I review some of the molecular data for larval development of both echinoderms and hemichordates.
Hox genes, as discussed previously, are perhaps the best known of the body-patterning genes for their unique chromosomal cluster organization and how it relates to its collinear expression along the anteroposterior axis of bilaterian developing embryos. The expression of particular Hox genes conveys spatial information to cells along the developing anteroposterior axis. These homeo- box genes have been used broadly as a compara- tive tool to investigate similarities in body plans between distantly related groups. The broad con- sensus from a wide range of bilaterians is that the Hox complex has played a central role in the evolution of the bilaterian anteroposterior axis (Krumlauf et al., 1993; Lumsden and Krumlauf,
1996; Pearson et al., 2005). If a larval life-history stage was involved in establishing the chordate
body plan, then we may expect to see some molecular evidence for a role of the Hox complex in ectodermal regionalization during the forma- tion of the larval body plan. Within larval species of echinoderms and hemichordates, there is so far only expression information for Hox genes in echinoids (Arenas-Mena et al., 2000). Interestingly there is only a small subset of the 13 Hox genes cloned from echinoderms expressed during the patterning of the larval body plan. This subset of Hox genes is not expressed in a coordinated and collinear fashion, but in a lineage-specific fashion. The first sign of colinear expression of the Hox cluster is later, during larval develop- ment as the adult radial echinoderm body plan is beginning to develop. This has also been found in lecithotrophic larvae of crinoids (Hara et al.,
2006) and direct-developing sea urchins (Morris and Byrne, 2005). Only the posterior cluster mem- bers have so far been examined, but their expres- sion is detected in the posterior coeloms in larval species, much later in development.
Other homeobox genes with potentially con- served roles in anteroposterior patterning between S. kowalevskii and chordates similarly present little evidence of a conserved role in larval anteropos- terior regionalization. For example, otx is a marker of the adult anterior nervous system during devel- opment of adult nervous systems (Finkelstein and Boncinelli, 1994; Acampora et al., 2005). The expres- sion of this gene has been examined quite exten- sively throughout different echinoderm groups and in Ptychodera flava, an indirect-developing species of hemichordate (Harada et al., 2000). However, otx expression is not restricted to any particular region of the ciliated band in any of these larval spe- cies. Distalless, another conserved anterior neural marker, shows quite varied expression domains in different echinoderm larvae (Lowe et al., 2002), but there is no evidence for a conserved role in larval anteroposterior patterning. Some authors have argued that there currently insufficient data to reject outright the paedomorphosis hypothesis proposed by Garstang (Poustka et al., 2004) and show examples of gene expression domains that are consistent with chordate larval origins. Certainly broader comparisons should be carried out in the roles of body patterning genes during the early

104 AN I M AL EV O L UTI O N



development of marine invertebrates with complex life histories. However, as Haag points out in his discussion of this point (Haag, 2005, 2006), whether or not an ancestral deuterostome was characterized by a feeding primary larva or not, it is important to consider carefully what exactly is being com- pared across groups, and the adult echinoderm body plan, however derived it may be, is probably the most relevant for body plan comparisons to the chordates. The data from Lowe et al. (2003, 2006) certainly support the hypothesis of adult origins of the chordate body plan. Echinoderm develop- ment has focused almost exclusively on the larva, and very few studies have been carried out on the development of the adult. Ultimately, a detailed comparison between direct- and indirect-devel- oping echinoderms and hemichordates will be necessary to test between competing hypotheses of chordate origins.


10.7 Conclusions

The molecular genetic body patterning data pre- sented in this chapter reveal some critical insights into the body plan of the deuterostome ancestor, and a unique way to compare the adult body plans of hemichordates and chordates. The detailed simi- larities in the transcriptional and signalling net- works are not likely to be a result of recruitment
of individual genes into convergently similar expression topologies. These exquisite similar- ities are almost certainly a result of homology. However, what we can most confidently recon- struct is an ancestral gene network rather than ancestral morphologies. Most of the gene networks discussed have been used comparatively to inves- tigate the nature of ancestral nervous systems, and yet hemichordates are a good example of how hom- ologous gene regulatory networks can be deployed to regulate the development of nervous systems with fundamental differences in their basic organ- ization. While gene networks are conserved over large evolutionary timescales, the broad range of morphologies that they regulate has not been con- strained by the higher-level regulatory control. Tight regulatory conservation is the foundation of both the highly complex vertebrate central ner- vous system and the basiepithelial nerve net of the hemichordates. Although these genetic networks have potential for testing hypotheses of morpho- logical homology, their reliability as informative characters is questionable given the range of neural morphologies regulated by this network. Caution should be exercised when reconstructing ances- tral neuroanatomies based on these data. Broader sampling and incorporation of fossil data sets will all be required for a more rigorous assessment of ancestral features of early deuterostomes.