Integrative and Comparative Biology
Integrative and Comparative Biology, volume 52, number 6, pp. 829–834
doi:10.1093/icb/ics117
Society for Integrative and Comparative Biology
SYMPOSIUM
Ptychoderid Hemichordate Neurulation without a Notochord
Shawn Luttrell,*,† Charlotte Konikoff,*,‡ Alana Byrne,*,† Barbara Bengtsson* and Billie J.
Swalla1,*,†,‡
*
Department of Biology, University of Washington, Seattle, WA 98195, USA; †Friday Harbor Laboratories, University of
Washington, Friday Harbor, WA 98250, USA; ‡BEACON, Evolution in Action Science Technology Center, Michigan State
University, East Lansing, MI 48824, USA
From the symposium ‘‘Evo-Devo Rides the Genomics Express’’ presented at the annual meeting of the Society for
Integrative and Comparative Biology, January 3–7, 2012 at Charleston, South Carolina.
1
E-mail: [email protected]
Synopsis Enteropneust hemichordates share several characteristics with chordates, such as a Hox-specified anterior–
posterior axis, pharyngeal gill slits, a dorsal central nervous system (CNS), and a juvenile postanal tail. Ptychoderid
hemichordates, such as the indirect-developer Ptychodera flava, have feeding larvae and a remarkable capacity to regenerate their CNS. We compared neurulation of ptychoderid hemichordates and chordates using histological analyses, and
found many similarities in CNS development. In ptychoderid hemichordates, which lack a notochord, the proboscis
skeleton develops from endoderm after neurulation. The position of the proboscis skeleton directly under the nerve cord
suggests that it serves a structural role similar to the notochord of chordates. These results suggest that either the CNS
preceded evolution of the notochord or that the notochord has been lost in hemichordates. The evolution of the
notochord remains ambiguous, but it may have evolved from endoderm, not mesoderm.
Introduction
The origin of the chordates has intrigued biologists
for over 150 years (Gee 1996; Cameron et al. 2000;
Gerhart et al. 2005; Swalla and Xavier-Neto 2008).
Molecular phylogenies of metazoans constructed
over the past 15 years have shown that the lophophorates are not basal deuterostomes, but are, if
fact, within the clade of Lophotrochozoa (Halanych
2004). Re-examination of the long-held view of a
colonial, stalked deuterostome ancestor (Romer
1967) in light of molecular deuterostome phylogenies
resulted in a new hypothesis of a benthic, vermiform
deuterostome ancestor with gill slits, from which
chordates evolved (Cameron et al. 2000; Gerhart
et al. 2005; Brown et al. 2008; Swalla and Smith
2008). Enteropneust hemichordates share many key
features with chordates including gill slits, a postanal
tail in the larvae of direct-developing species, and an
anterior–posterior axis specified by a conserved set of
developmental genes (Lowe et al. 2003; Brown et al.
2008; Swalla and Smith 2008; Gillis et al. 2012).
In contrast to hemichordates and other invertebrates, chordates have been reported to have an
inversion of Bone Morphogenetic Protein (BMP) expression along the dorsal-ventral axis (Lowe et al.
2006). Over-expression of BMP in hemichordates
has been shown to extend gill slits circumferentially,
while under-expression extends the mouth (Lowe
et al. 2006). BMP antagonists such as chordin are
expressed ventrally early in hemichordates and dorsally in chordates, linking changes in early BMP expression to orientation of the gill slits (Lowe et al.
2006). This evidence, along with a reversed expression of nodal, suggests that inversion of the
dorsal-ventral axis occurred during chordate evolution (Duboc et al. 2005; Gerhart et al. 2005; Lowe
et al. 2006). One of the current hypotheses is that the
mouth of a worm-like deuterostome ancestor rotated
dorsally, resulting in the current chordate body plan
(Lowe et al. 2006). This hypothesis assumes that no
central nervous system (CNS) existed in the
worm-like deuterostome ancestor, only a diffuse
Advanced Access publication September 9, 2012
ß The Author 2012. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved.
For permissions please email: [email protected].
830
nerve net (Holland 2003; Lowe et al. 2006).
However, if BMP expression at the time of neurulation is compared between hemichordates and chordates, BMP expression is found in a dorsal strip down
the CNS midline during chick development (Selleck
et al. 2003; Martı́ 2000), exactly as has been reported
for hemichordates (Lowe et al. 2006). This alternative view suggests that the molecular mechanism of
defining dorsal during CNS development has been
conserved between hemichordates and chordates.
Hemichordates are the only deuterostomes that
have a tripartite adult body plan. These three regions, from anterior to posterior, are the proboscis,
collar, and trunk (Fig. 1). The collar region contains
the CNS, while the trunk region contains the gill
slits, gill bars, gonads, hepatic sacs, and intestine.
Saccoglossus kowalevskii has been shown to have an
ectodermal nerve net, with anterioposterior specification due to conserved transcription factors that are
expressed in the vertebrate CNS in an anterior to
posterior manner (Lowe et al. 2003). This has been
referred to as a ‘‘skin brain’’ (Holland 2003).
However, there is also a CNS in the dorsal region
of the collar in hemichordates (Brown et al. 2008;
Nomaksteinsky et al. 2009; Burke 2011). Thomas
Hunt Morgan initially described the process of neurulation in Balanoglossus tornaria in the late 1800s,
discussing homology with chordate neurulation
(Morgan 1894). Libbie Hyman (1959) also described
the CNS and the cartilaginous skeleton in enteropneust hemichordates, suggesting the gill bars and
proboscis skeleton developed from the endoderm.
Recent experiments showed that the gill bars do develop from endoderm in hemichordates, making an
unusual acellular cartilage (Smith et al. 2003; Rychel
and Swalla 2007; Gillis et al. 2012). Furthermore,
morphological and cytological evidence in developing embryos and juvenile worms suggests that the
collar cord is a crucial, centralized element in the
nervous system of S. kowalevskii, and that neurulation shares many key similarities with chordates
(Kaul and Stach 2010) and has specialized cell
types (Brown and others 2008; Nomaksteinsky
et al. 2009; Burke 2011).
Neuronal and pan-neuronal expression data in juvenile and adult worms (Nomaksteinsky et al. 2009)
shows the diffuse nerve net is a transitory phase in
juvenile enteropneusts, with dorsal and ventral
strands of localized neurons ultimately moving anteriorly and dorsally in the collar region. The hemichordate CNS may represent a transitional form
between protostomes and deuterostomes, as they
have both dorsal and ventral strands of neurons
before migration to the collar region occurs
S. Luttrell et al.
(Nomaksteinsky et al. 2009). Interestingly, a dorsally
migrating strand has also been found in amphioxus
(Benito-Gutiérrez et al. 2005), but its origin remains
unclear. CNS-related gene expression in enteropneust
hemichordates has also suggested the presence of
precursory ANR, ZLI, and IsO vertebrate-like signaling centers in these animals (Pani et al. 2012).
Formation of the central nervous
system in Ptychodera flava precedes
formation of the proboscis skeleton
during metamorphosis
Recent molecular developmental work has shown
that hemichordates and lancelets contain gill bars
made from endoderm, not mesoderm (Rychel and
Swalla 2007; Gillis et al. 2012). During those studies
we obtained evidence that the proboscis skeleton is
also formed from the endoderm. In cross-sections of
adult ptychoderid worms, the proboscis skeleton is
ventral to the CNS, serving a similar function of
stability as the notochord does in chordates (Brown
et al. 2008). Therefore, we sectioned metamorphosed
Ptychodera flava in order to determine whether the
proboscis skeleton or CNS forms first in the neck
region of metamorphosing larvae. Surprisingly, the
CNS develops first, rolling up over the dorsal
blood vessel in metamorphosing ptychoderid hemichordates (Fig. 2A–D). In contrast, when the CNS
first develops the proboscis skeleton is just beginning
to be formed from folds of endoderm, seen as thin
slivers of Alcian-blue-positive matrix in the middle
neck region (Fig. 2C). The posterior forks of the
proboscis skeleton develop first, then mineralization
of the endoderm proceeds in a posterior to anterior
manner in the neck region as metamorphosis continues. When these juvenile sections are compared to
adult sections of the neck region (Fig. 2F–H), we
find that the CNS has a strong cartilaginous and
mineralized structure just underneath it, the proboscis skeleton, that likely lends structural support similar to the notochord in chordates. Note that the
adult worm sections are approximately four times
larger than the metamorphosing juvenile, but are
shown here at the same size to directly compare
morphological structures.
Formation of the larval neural tube in
hemichordates shares similarities with
neurulation in vertebrates
Development of the larval ptychoderid nervous
system follows a remarkably similar pattern to that
of chordate neural development. Figure 3 compares
neural tube formation in metamorphosed larval
Hemichordate neurulation
831
Fig. 1 The tripartite adult body plan of the enteropneust P. flava. From anterior to posterior, enteropneust hemichordates possess
proboscis, collar, and trunk regions. The internal stomochord and proboscis skeleton are also shown. Lines a–d correspond to the P.
flava cross sections shown in Fig. 2. The posterior trunk region contains the genital wings and hepatic sacs.
Fig. 2 Neurulation in P. flava in the absence of a notochord. Illustration depicts approximate cross-section position relative to the
whole animal (Fig. 1). (A,E) correspond with line a, (B,F) with line b, (C,G) with line c, and (D,H) with line d. All sections were stained
with Milligan’s trichrome; blue indicates collagen and the proboscis skeleton. A–D are cross sections of metamorphosing larvae.
(A) The stomochord is formed in the proboscis while the neural tube is invaginating at the posterior neck region (D) in a metamorphosing larval worm. (B) The neural tube is rolled up and dropping into the interior of the neck region, anterior to posterior. The
dorsal vessel is seen clearly as a closed circle beneath the neural tube. The thickened epithelial at the bottom is the endoderm.
(C) Buds of the developing proboscis skeleton visible at the posterior side of the collar region. The proboscis skeleton is being
elaborated by the endoderm. Adult cross sections (E–H) shown anterior to posterior are approximately equivalent to the corresponding larval stage. (E) Fully developed stomochord above heart and kidney complex in the proboscis. (F) Neural tube above dorsal
vessel and fully developed proboscis skeleton, stained bright blue. This is in the anterior collar region. (G) Proboscis skeleton is forked
at the end, showing either side in cross-section. (H) The neural tube and dorsal vessel posterior to the proboscis skeleton in the neck
region. In metamorphosing cross sections (A–D) scale bars are 100 m. In adult sections (E–F) scale bars are 200 m.
hemichordates to vertebrate neurulation. The P. flava
larva sectioned in Fig. 3B is shown intact in Fig. 3A.
The collar ectoderm invaginates and rolls up,
forming the CNS in an anterior-to-posterior
manner (Fig. 3B). However, hemichordates lack a
notochord, which in chordates serves a structural
and signaling role in development (Gerhart et al.
2005; Brown et al. 2008). Instead of the notochord,
a dorsal blood vessel is ventral to the CNS,
suggesting there may be signaling molecules emanating from the dorsal vessel that induce neurulation.
The hemichordate stomochord has been suggested to
be the notochord homolog, primarily because of the
vacuolated appearance of the cells (Morgan 1894;
Balser and Ruppert 1990). However, the stomochord
develops in the proboscis, not in the collar region
(Fig. 2A), so is unlikely to serve a signaling role in
neurulation during CNS development. Instead,
832
S. Luttrell et al.
Fig. 3 Neurulation in the absence (P. flava) and presence (chick) of a notochord. (A) Photograph of a P. flava larva, 2 weeks after
collection from the plankton. The trunk region is at the top and proceeds anteriorly to the proboscis at the bottom. (B) Cross-sections
of a 2-week P. flava larva with posterior collar cross-sections at the top and proceeding anteriorly through the collar region. The CNS
forms by invagination of ectoderm and occurs dorsal to the dorsal vessel. (C) Illustration of neurulation in a developing chick embryo.
Posterior is at the top and the CNS forms by invagination of ectoderm and occurs dorsal to the notochord. (D) Illustration of a chick
embryo at stage 10 of embryonic development.
developmental signaling may be emanating from
the dorsal vessel and/or endoderm, which expresses
sonic hedgehog (Pani et al. 2012). The dorsal vessel
is located directly under the CNS and is dorsal to the
endoderm in both chordates and hemichordates
(Figs. 2B–D; 2F–H and 3). Neurulation occurs dorsally, suggesting that neural tube development may
be homologous in chordates and hemichordates.
Development of the larval ptychoderid nervous
system begins in the anterior region of the collar
and proceeds posteriorly through the collar. Similar
to the process in vertebrates, ectoderm invaginates
and rolls up forming the CNS (Fig. 3B). The dorsal
blood vessel is ventral to the CNS, suggesting there
may be signaling molecules emanating from the
dorsal vessel that induce neurulation. This process is
compared with formation of the neural tube in vertebrates, shown in Fig. 3. The developmental series is
similar, except that in hemichordates the neural tube
sinks within the collar region, maintaining a connection with the dorsal midline. We are investigating
whether this connection consists of neural or connective tissue. The other major difference is that the CNS
of hemichordates is found only in the collar region,
while in chordates the CNS is found along the entire
dorsal midline (Kaul and Stach 2010).
A summary of CNS development in chordates is
compared to that in hemichordates in Fig. 4. The
neural tubes of vertebrates and hemichordates both
develop dorsal to the dorsal vessel; however, vertebrates develop a notochord situated between the
neural tube and dorsal vessel. In vertebrates, neural
crest cells migrate ventromedially from the dorsal
region of the neural tube through the rostral half
of each somite and coalesce into ganglia beside the
dorsal vessel (Rickmann et al. 1985; Erickson et al.
1989; Schneider et al. 1999; Young et al. 2004). The
vertebrate dorsal vessel is ventral to the notochord
and dorsal to the endoderm. The dorsal vessel
secretes BMP molecules that signal neural crest cells
to differentiate into sympathetic neurons (Rickmann
et al. 1985; Schneider et al. 1999; Young et al. 2004),
while the notochord induces CNS formation. Due to
the absence of the notochord in hemichordates, the
dorsal vessel may control neurulation, but further
experiments will be necessary to test this hypothesis.
Conclusions
In light of developmental similarities between the
CNS in chordates and hemichordates, we propose
that the common ancestor of the deuterostomes had
a CNS that formed by the developmental in-pocketing
of ectoderm. Taken together with previous studies,
our results further suggest that neurulation and patterning in enteropneust hemichordates shares key
833
Hemichordate neurulation
Fig. 4 Diagrams of cross sections of a hemichordate and a vertebrate. In both pictures dorsal is to the top and ventral is to the
bottom. In the diagram on the left, the metamorphosing hemichordate neural tube (green) develops above the dorsal vessel (red). The
endoderm (yellow) is directly ventral to the dorsal vessel. In the illustration of the developing chick embryo on the right, the vertebrate
neural tube (green) develops above the notochord (pink). Somites (red) surround the neural tube and notochord. The dorsal vessel
(red) is ventral to the notochord and dorsal to the endoderm (yellow). Blue represents nonneural ectoderm in both images.
similarities and differences with chordates (Brown et
al. 2008; Kaul and Stach 2010; Pani et al. 2012). In the
chordate ancestor, the tripartite larva was lost, allowing direct development, losing the separate larval coeloms, and allowing the subsequent fusion of the neck
and pharyngeal regions. Our results also suggest that
the ancestral deuterostome cartilage developed from
endoderm (Rychel and Swalla 2007). The proboscis
skeleton may play additional roles, as it projects
more anteriorly to the mouth than a chordate
notochord.
Future studies aimed at exploring neural gene
expression and discovering the origins of various inductive signals will elucidate how conserved these
patterning principals are among deuterostomes and
among Bilateria as a whole. Here, we show that the
stomochord develops in the proboscis, not in the
collar. The stomochord is believed to be derived
from the anterior endoderm in the collar region,
which may then be pushed anteriorly into the proboscis as development progresses (Lowe et al. 2006;
Nomaksteinsky et al. 2009; Kaul and Stach 2010).
Thus, neuralizing signal from the stomochord
cannot be completely ruled out, because early stomochord precursors could send signals earlier in development while they are still in the collar region.
Looking beyond the deuterostome ancestor, studies in Platynereis dumerilii, a polychaete, strongly
suggest the ancestral bilaterians also had a body
plan that included a centralized nervous system
(Denes et al. 2007). Molecular and cellular data
have revealed that the neurectoderm of P. dumerilli
is patterned into vertebrate-like longitudinal precursor domains, and furthermore that neuronal differentiation occurs from mediolateral domains
strikingly similar to those found in vertebrates
(Denes et al. 2007). As with many other bilaterian
organisms, Denes et al. as well as Mizutani et al.
(2006) and Mizutani and Bier (2008) found that
BMP signaling in the polychaete plays a crucial role
in patterning the developing nervous system.
In summary, the deuterostome ancestor was likely
to be a vermiform animal with an anterior mouth
and posterior anus, with gill slits and a well-formed
nervous system. Indeed, it is possible that the deuterostome ancestor was very chordate-like, and that
losses in the ambulacraria resulted in the evolution
of radically different body plans. The loss of the
notochord would have preceded other changes in
body plan.
Funding
Both Shawn Luttrell and Alana Byrne were funded as
Mary Gates Scholars and as Hughes Undergraduate
Researchers to pursue research at the University of
Washington Friday Harbor Laboratories in the
Genomics 2010 Research Apprenticeship. This
834
research was reported at the 2010 Society for
Integrative and Comparative Biology meetings in
Seattle, Washington, and Shawn Luttrell thanks the
Department of Biology at the University of Washington for the Casey Travel Award she received to
attend the meetings. Part of this research was
funded by National Science Foundation #DEB0816892 to BJ Swalla. Dr Charlotte Konikoff is supported by the National Science Foundation under
Cooperative
Agreement
No.
DBI-0939454
(BEACON). Any opinions, findings, and conclusions
or recommendations expressed in this material are
those of the authors and do not necessarily reflect
the views of the National Science Foundation.
References
Balser E, Ruppert E. 1990. Structure, ultrastructure and function of the preoral heart-kidney in Saccoglossus kowalevskii
(Hemichordata, Enteropneusts) including new data on the
stomochord. Acta Zoologica 71:235–49.
Benito-Gutiérrez E, Nake C, Llovera M, Comella J, GarciaFernàndez J. 2005. The single AmphiTrk receptor highlights
increased complexity of neurotrophin signalling in vertebrates and suggests an early role in developing sensory
neuroepidermal cells. Development 132:2191–202.
Brown FD, Prendergast A, Swalla BJ. 2008. Man is but a
worm: Chordate origins. Genesis 46:605–13.
Burke R. 2011. Deuterostome neuroanatomy and the body
plan paradox. Evol Dev 13:110–5.
Cameron CB, Garey JR, Swalla BJ. 2000. Evolution of the
chordate body plan: new insights from phylogenetic analyses of deuterostome phyla. Proc Natl Acad Sci USA
97:4469–74.
Denes AS, Jékely G, Steinmetz PR, Raible F, Snyman H,
Prud’homme B, Ferrier DE, Balavoine G, Arendt D. 2007.
Molecular architecture of annelid nerve cord supports
common origin of nervous system centralization in bilateria. Cell 129:277–88.
Duboc V, Röttinger E, Lapraz F, Besnardeau L, Lepage T.
2005. Left-right asymmetry in the sea urchin embryo is
regulated by nodal signaling on the right side. Dev Cell
9:147–58.
Erickson C, Loring J, Lester S. 1989. Migratory pathways of
HNK-1 immunoreactive neural crest cells in the rat
embryo. Dev Biol 134:112–8.
Gee H. 1996. Before the backbone: views on the origin of the
vertebrates. London: Chapman & Hall.
Gerhart J, Lowe C, Kirschner M. 2005. Hemichordates and
the origin of chordates. Curr Opin Genet Dev 15:461–7.
Gillis J, Fritzenwanker J, Lowe C. 2012. A stem-deuterostome
origin of the vertebrate pharyngeal transcriptional network.
Proc Roy Soc B Biol Sci 279:237–46.
Halanych K. 2004. The new view of animal phylogeny. Annu
Rev Ecol Evol Syst 35:229–56.
Holland N. 2003. Early central nervous system evolution: an
era of skin brains? Nat Rev Neurosci 4:617–27.
S. Luttrell et al.
Hyman L. 1959. Smaller coelomate groups. The invertebrates.
New York: McGraw-Hill. p. 783.
Kaul S, Stach T. 2010. Ontogeny of the collar cord: neurulation in the hemichordate Saccoglossus kowalevskii. J
Morphol 271:1240–59.
Lowe C, Terasaki M, Wu M, Freeman R, Runft L, Kwan K,
Haigo S, Aronowicz J, Lander E, Gruber C, et al. 2006.
Dorsoventral patterning in hemichordates: insights into
early chordate evolution. PLoS Biol 4:1603–19.
Lowe CJ, Wu M, Salic A, Evans L, Lander E, StangeThomann N, Gruber CE, Gerhart J, Kirschner M. 2003.
Anteroposterior patterning in hemichordates and the origins of the chordate nervous system. Cell 113:853–65.
Martı́ E. 2000. Expression of chick BMP-1/Tolloid during
patterning of the neural tube and somites. Mech Dev
91:415–9.
Mizutani CM, Bier E. 2008. EvoD/Vo: the origins of BMP
signalling in the neuroectoderm. Nat Rev Genet 9:663–77.
Mizutani CM, Meyer N, Roelink H, Bier E. 2006.
Threshold-dependent BMP-mediated repression: a model
for a conserved mechanism that patterns the neuroectoderm. PLoS Biol 4:e313.
Morgan T. 1894. The development of Balanoglossus. J
Morphol 9:1–86.
Nomaksteinsky M, Rottinger E, Dufour H, Chettouh Z,
Lowe C, Martindale M, Brunet J. 2009. Centralization of
the deuterostome nervous system predates chordates. Curr
Biol 19:1264–9.
Pani A, Mullarkey E, Aronowicz J, Assimacopoulos S,
Grove E, Lowe C. 2012. Ancient deuterostome origins of
vertebrate brain signaling centres. Nature 483:289–94.
Rickmann M, Fawcett J, Keynes R. 1985. The migration of
neural crest cells and the growth of motor axons through
the rostral half of the chick somite. J Embryol Exp Morphol
90:437–55.
Romer A. 1967. Major steps in vertebrate evolution. Science
158:1629–37.
Rychel A, Swalla B. 2007. Development and evolution of
chordate cartilage. J Exp Zool Part B Mol Dev Evol
308B:325–35.
Schneider C, Wicht H, Enderich J, Wegner M, Rohrer H.
1999. Bone morphogenetic proteins are required in vivo
for the generation of sympathetic neurons. Neuron
24:861–70.
Selleck M, Garcı́a-Castro M, Artinger K, Bronner-Fraser M.
2003. Effects of Shh and Noggin on neural crest formation
demonstrate that BMP is required in the neural tube but
not ectoderm. Development 125:4919–30.
Smith S, Douglas R, Burke da Silva K, Swalla B. 2003.
Morphological and molecular identification of enteropneust
worms (Hemichordata: Harrimanidae) in the Pacific
Northwest. Can J Zool 81:133–41.
Swalla BJ, Smith AB. 2008. Deciphering deuterostome
phylogeny: molecular, morphological and palaeontological
perspectives. Philos Trans R Soc Lond B Biol Sci
363:1557–68.
Swalla BJ, Xavier-Neto J. 2008. Chordate origins and evolution. Genesis 46:575–9.
Young HM, Anderson RB, Anderson CR. 2004. Guidance cues
involved in the development of the peripheral autonomic
nervous system. Auton Neurosci 112:1–14.