JOURNAL OF EXPERIMENTAL ZOOLOGY (MOL DEV EVOL) 308B (2007)
Primitive Versus Derived Traits
in the Developmental Program of the
Vertebrate Head: Views From Cyclostome
Developmental Studies
SHIGERU KURATANI! AND KINYA G. OTA
Evolutionary Morphology Research Group, Center for Developmental Biology,
RIKEN, Kobe, Japan
ABSTRACT
Evolution can be viewed as a series of changes in the developmental program along
the phylogenetic tree. To better understand the early evolution of the vertebrate skull, we can use
the embryos of the cyclostome species as models. By comparing the cyclostome developmental
patterns with those of gnathostomes, it becomes possible to distinguish the primitive and derived
parts of the developmental program as taxon-specific traits. These traits are often recognizable as
developmental constraints that define taxa by biasing the developmental trajectories within a certain
limited range, resulting in morphological homologies in adults. These developmental constraints are
distributed on the phylogenetic tree like the morphological character states of adult animals and are
associated with specific regions of the tree. From this perspective, we emphasize the importance of
considering gene expression and embryonic anatomy as the mechanistic bases that can result in
homologous or nonhomologous morphological patterns at later developmental stages. Taking the
acquisition of the jaw and trabecula cranii as examples, we demonstrate that a set of embryonic
features can be coupled or decoupled during evolution and development. When they are coupled,
they exert an ancestral developmental constraint that results in homologous morphological patterns,
and when they are decoupled, the ancestral constraints tend to be abandoned, generating a new body
plan. The heterotopy behind the specification of the oral domain is an example of decoupling, based
on shifted tissue interactions. We also stress the importance of ‘‘developmental burden’’ in
determining the sequential order of changes through evolution. J. Exp. Zool. (Mol. Dev. Evol.) 308B,
2007.
r 2007 Wiley-Liss, Inc.
How to cite this article: Kuratani S, Ota KG. 2007. Primitive versus derived traits in the
developmental program of the vertebrate head: views from cyclostome developmental
studies. J. Exp. Zool. (Mol. Dev. Evol.) 308B:[page range].
Identification of the morphological ground plan
of the vertebrate skull, and the mechanical basis
for its morphological variation, used to be primarily a question of transcendental anatomy
(reviewed by De Beer, ’37; Kuratani, 2005a).
Goethe (1790), for example, called the ground
pattern of animal morphology the Urtyp, the
idealistic archetype (more anatomically formulated and conceptual than the phylotypic stage;
see below) from which can be derived all the
various features of an animal species through
metamorphosis. Goethe (1790), as well as Oken
(1807), proposed that the vertebrate skull is
primarily composed of several vertebrae, which
have changed their shapes and functions depending on their positions along the body axis. This
r 2007 WILEY-LISS, INC.
notion survived in comparative embryology until
the 20th century, and remains a question for
evolutionary developmental biology in terms of
the presence of metamerical developmental units
in the embryonic vertebrate head (reviewed by
Kuratani, 2003, 2005a).
In comparative morphology and embryology,
which were originally rooted in the work of
Goethe, metameric organization and positional
!Correspondence to: Shigeru Kuratani, Group Director, Laboratory
for Evolutionary Morphology, Center for Developmental Biology,
RIKEN, 2-2-3 Minatojima-minami, Chuo, Kobe, Hyogo 650-0047,
Japan. E-mail: saizo@cdb.riken.jp
Received 19 April 2007; Revised 2 July 2007; Accepted 2 July 2007
Published online in Wiley InterScience (www.interscience.wiley.
com). DOI: 10.1002/jez.b.21190
2
S. KURATANI AND K.G. OTA
value-based transformation have therefore been
the two major key concepts, both of which are now
central issues in developmental biology in molecular terms (see Holland, ’88; McGinnis and
Krumlauf, ’92; Slack et al., ’93). The latter
includes the expression domains of homologous
genes in various embryos. For example, the shared
patterns of Hox gene expression in bilaterians is
called a ‘‘zootype’’, which is regarded as a
synapomorphy of the metazoans at the level of
the developmental program (Slack et al., ’93).
Needless to say, such a ‘‘ground type’’ perceived in
a certain animal group, such as the vertebrates,
cannot be assumed to be shared by all the animal
phyla beyond that group, but any body plan must
have had its origin in the course of metazoan
evolution. Therefore, in the contemporary context,
comparative morphology must inevitably consider
the origin of the archetype per se, once it is
identified by comparative studies.
The archetype can be rephrased today in more
evolutionarily and developmentally accurate
terms as the ancestral developmental program
putatively possessed by the common ancestor of
the animals under consideration. Additional
changes were introduced into this primitive
developmental program, resulting in the various
modifications of the ancestral form into diverse
morphological types in the vertebrates. This
review is intended to relate the findings of
comparative developmental biology to the hypothetical evolutionary sequence by considering
the phylogenetic relationships of animal lineages,
and discussing the evolutionary and ontogenetic
behavior of the developmental constraints.
tree. Thus, the analyses are reversed: the traits
shared by two animal lineages can be regarded as
already established in the most recent common
ancestor of both, unless the traits resulted from
parallel evolution (Fig. 1). Conversely, traits found
in only one of the two groups may represent a
synapomorphic (newly added) trait that defines
that group, or else a plesiomorphic (primitive and
ancestral) feature that has been lost from the
other group. Therefore, it is a prerequisite for this
analysis to determine whether a trait is comparatively new or old within the evolutionary time
course, and thus to discern the polarity of the
observed developmental changes. For example, the
absence of vertebrae observed in the hagfish may
represent either the primitive, prevertebrate state
or a secondary loss, the derived state. A so-called
out-group comparison must be made to clarify
this, but an ideal out-group may not always be
available. When no out-group is available, developmental biology may provide a clue because, even
after the trait in question is lost, a part of the
genetic cascade may still be present. In other
cases, the loss of certain structures can be
attributed to the loss of competence of some
PHYLOGENY AND COMPARATIVE
EMBRYOLOGY
When comparing the characteristic developmental patterns and processes of the embryos of
different taxa, it is crucial to choose a reliable
phylogenetic tree, reconstructed ideally with no
information on comparative morphological, embryological, or developmental traits, to circumvent
circular arguments. The trees obtained by recent
molecular phylogenetic analyses are adequate in
this regard (see Narita and Kuratani, 2005). The
purpose of this comparative analysis is not to
establish the phylogenetic relationships between
different animal lineages, but to determine the
sequence of changes introduced into the developmental program of these animals so that this
sequence is consistent with the most acceptable
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
Fig. 1. A simplified phylogenetic tree showing the relationships between vertebrate groups. In this scheme, the
vertebrates consist of the cyclostomes and gnathostomes.
Apomorphic changes in the developmental programs (A to C)
are plotted onto the tree based on the distribution patterns of
these traits in the animal groups (developmental traits A to
C). For example, the vertebrates are characterized by the
possession of developmental trait A, which is missing in the
out-group. Because this trait can be observed in both
gnathostomes and cyclostomes, it is expected to have occurred
in at least the most recent common ancestor of all the
vertebrates.
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
growth factors that rely on the specific molecular
function of, for example, a receptor of those
ligands (reviewed by Hall, ’98).
Alternatively, the complexity of the structure,
organ, or developmental cascade can represent the
derived condition relative to their simpler counterparts. It is equally possible that the less complicated state of the character represents a secondary
simplification. It is not essential to deal with as
many characters as possible. It is more important
to understand the sequence of changes in terms
of developmental and morphological patterns
(or even in terms of body plans) by visualizing
the hypothetical embryonic anatomy as an integrated entity. One trait cannot always evolve
independently, without affecting or being influenced by other structures or developmental
events. This issue will be revisited below in the
context of ‘‘developmental burdens’’ and the ‘‘key
innovations’’ involved in the integrity of nasal and
oral evolution in the early vertebrates.
The phylogenetic position of the hagfish and the
naming of the animal groups are most problematic
for our understanding of the early history of the
vertebrates when we use the phylogenetic strategy
described above. The hagfish has often been placed
basal to the rest of the vertebrates, as an outgroup of the latter, mainly based on a number of
primitive morphological, physiological, and
embryological characters (see Yalden, ’85; Forey
and Janvier, ’93; Forey, ’95; Janvier, ’96; Braun,
’98). Conversely, molecular analyses of a number
of genes have almost unanimously supported the
monophyly of the cyclostomes (Stock and Whitt,
’92; Mallatt and Sullivan, ’98; Furlong and
Holland, 2002; Takezaki et al., 2003; reviewed by
Ota and Kuratani, 2007, in press; the shared
molecular mechanism that generates variable
lymphocyte receptors in the cyclostomes, as
discovered by Pancer et al., 2004, 2005, may
represent a plesiomorphy). Although we support
the cyclostome monophyly in this review, the
ultimate resolution of the dilemma described
above requires a thorough embryological investigation of the hagfish at the molecular level. A
comparison of the lamprey and hagfish should be
very important in shedding light on the putatively
primitive state of the vertebrate developmental
program, especially in the context of olfactory–
hypophysial patterning (see below).
If we recognize the cyclostomes as a monophyletic group, can we call the other branch the
‘‘gnathostomes’’? Although the entire group of the
jawed vertebrates we see today and the common
3
ancestor of the cyclostomes are thought to have
been generated at the point of this branching, it is
not equivalent to the point of jaw acquisition (Fig.
2). Instead, based on this node-dependent definition of the cyclostomes and gnathostomes, the
gnathostomes should include a number of fossil
agnathans as the stem group of the gnathostomes.
In this context, newly acquired developmental
changes in the ‘‘early gnathostomes’’ may not
necessarily explain the acquisition of the jaw
(Janvier, 2001). Many of the gnathostome-specific
developmental traits may have been shared by
most fossil agnathans, but not by the cyclostomes.
In an alternative classification, the jawless vertebrates can only be defined as a paraphyletic group
containing the cyclostomes and other fossil agnathans if we depend on the synapomorphy, ‘‘the
presence of a jaw’’, to define the group Gnathostomata (Fig. 2). However, the apparently primitive
features of the cyclostome developmental patterns
may have already been lost in the most recent
ancestor of the jawed gnathostomes. Those primitive or derived features of the developmental
patterns of gnathostome model animals that are
defined by comparison with agnathan animals
may always represent synapomorphies more specific to the crown taxa, such as birds or mammals.
Even if such an overstatement can be avoided by
comparing more gnathostome taxa, the primitive
characters that are possessed only by lampreys
and/or hagfish may still represent cyclostomespecific traits, in the absence of proper out-groups.
In the following discussion, these inherent problems associated with the evaluation of developmental characters are addressed as they relate to
each specific issue of vertebrate head evolution.
PRIMARY FEATURES OF THE
VERTEBRATE HEAD—MESODERMAL
SEGMENTATION
Classically, the central topic in vertebrate skull
morphology used to be its segmental archetype,
the idea that the skull is composed of segments or
compartments equivalent to the somites in the
trunk, as mentioned above. At least in comparative embryology, this idea depended largely on the
finding of ‘‘head cavities’’ in shark embryos
(Balfour, 1878; van Wijhe, 1882; reviewed by
Goodrich, ’30; De Beer, ’37; Jarvik, ’80; Jefferies,
’86; Kuratani, 2003, 2005a; for the influence of
this discovery on vertebrate comparative embryology, see Gee, ’96; Fig. 3). There are three (or
sometimes four) pairs of mesodermal epithelial
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
4
S. KURATANI AND K.G. OTA
Fig. 2. An elaborated version of the phylogenetic tree in Fig. 1. Fossil animals are included. In this tree, the cyclostomes and
their sister group, the gnathostomes, are defined by a hypothetical ‘‘node’’ that splits the two lineages. By this node-dependent
definition, the gnathostomes include the advanced jawless vertebrates as the stem group. The apomorphic character, ‘‘the biting
jaw’’, was acquired after the split. The animals basal to the node represent the nonvertebrate animals (out-groups), if the
vertebrates include both the cyclostomes and gnathostomes. If the gnathostomes are defined by the possession of the jaw, the
jawless vertebrates can now be regarded as paraphyletic agnathans on this tree. Comparison of the developmental programs of
living animals indicates that the former method of grouping is more appropriate.
coeloms in the embryonic shark head, and they
appear to give rise to myoblasts that differentiate
into the extrinsic eye muscles (Neal, ’18a). Interestingly, these coeloms arise in a pattern that
reflects the number and arrangement of the
cranial nerves that innervate the muscles (Fig.
3B). The rostralmost premandibular cavity differentiates into four muscles innervated by the
oculomotor nerve; the second or mandibular
cavity becomes the inferior oblique muscle,
innervated by the trochlear nerve; and the hyoid
cavity, the lateral rectus, is innervated by the
abducens (Fig. 3A and B). Of course, the terminology relating to the cavities reflects the idea that
the metamerism of the cavities is synchronized
with that of the pharyngeal arches.
The muscle–nerve relationships along the anteroposterior axis are reminiscent of the segmental
organization between the myotomes and spinal
motor nerves, as typically encountered in the
anatomical pattern of the trunk. This has driven
embryologists to look for a segmental plan of head
development in various vertebrate embryos (Neal,
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
1896; reviewed by Neal, ’18b). The segmental view
of vertebrate head development basically shows
the head as a structure equivalent to an array of
trunk segments, which are primarily derived from
somites. This idea was so influential that nonelasmobranch embryos were described as if they
were modified shark embryos (see Koltzoff, ’01;
Jarvik, ’80). The search for head segments thus
resulted in a conclusive scheme, presented by
Goodrich (’10, ’30), who drew the generalized
vertebrate embryonic head as a modified trunk,
based on the shark embryonic pattern (Fig. 3).
Obviously, the somitomeric segmental scheme
more closely resembles the acraniate pattern than
the actual vertebrate head (reviewed by Kuratani,
2003). Nevertheless, to determine how similar the
head mesoderm and somites are is still an
important task, because this also relates to the
evolution of the vertebrae in the basal vertebrates.
In cyclostomes, in which the centrum does not
develop, the posterior part of the cranium chondrifies into the chordal cranium near the notochord (parachordals; Holmgren, ’46; Johnels, ’48;
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
see below), just as the sclerotomal cells do in the
gnathostome trunk. Therefore, the hagfish may
not represent the ancestral ‘‘prevertebrate’’ condition, but may simply have lost their vertebrae
5
secondarily. In this context, the differentiation of
the somites into epithelial dermomyotomes and
mesenchymal components has also been observed
in the hagfish (Kupffer, ’00; also see Ota et al.,
2007; and unpublished observation by Ota and
Kuratani; also see Fig. 4).
Another line of evidence supporting cephalic
mesodermal segmentation comes from a more
recent observation made with scanning electron
microscopy. By observing whole-mount embryonic
heads with this technique, an array of incomplete
segments was found in the cephalic mesoderm,
which were designated ‘‘cephalic somitomeres’’
(Meier, ’79; Fig. 3E). Although somitomeres have
Fig. 3. Segmental theories of the vertebrate head. (A)
Head cavities of the shark embryo. The head cavities are
mesodermal coeloms and the shark pharyngula generally
exhibits three pairs of cavities called, from rostral to caudal,
the premandibular, mandibular, and hyoid cavities. The
posterior two are confluent with the coelom in two rostral
pharyngeal arches (mandibular and hyoid). Two more head
cavities are assumed to have been suppressed caudal to the
hyoid arches by the enlarged otic capsule. Thus, the shark
embryo is believed to possess somitomeric compartments in
the postoptic region, the segments equivalent to the somites in
the trunk. This metamerism has been equated to that in the
pharyngeal arches that have the same names as the cavities.
(B) Goodrich’s scheme of the vertebrate head, based on the
embryo of the shark, which has four pairs of head cavities.
According to Goodrich, the rostral-most one, Platt’s vesicle,
represents the mesodermal coelom of the remnant premandibular arch, with which the premandibular cavity used to be
associated. (C) Mesodermal components illustrated in (B),
clearly showing that Goodrich regarded the vertebrate head to
be a modified somitomeric structure. (D) Head cavities are
distributed in vertebrate clades as a synapomorphy of the
gnathostomes. White circles represent epithelial head cavities,
which are missing in lampreys (and also in hagfishes).
Modified from Kuratani (2005a). (E) Fate mapping of the
chicken head mesoderm by Noden (’88), based on a stage 10
embryo. Here, the somitomeres are used only as a reference to
locate the pieces of head mesoderm that Noden excised as
grafts. Note that the head mesoderm at this stage is
anteroposteriorly well specified (as head cavities in elasmobranchs), and differentiate into extrinsic eye muscles in the
order that these muscles are innervated by the cranial nerves
(nerves III, IV, and VI). However, this does not support the
presence of segments in the head mesoderm. ba1, first
branchial (third pharyngeal) arch; hbm, hypobranchial muscles; hm, hyoid mesoderm; hya, hyoid arch; hyc, hyoid cavity;
ma, mandibular arch; mm, mandibular mesoderm; mnc,
mandibular cavity; moc (III), extrinsic ocular muscles innervated by nerve III; mos (IV), superior oblique muscle
innervated by nerve IV; mrl, lateral rectus muscle innervated
by nerve VI; myo, myotomes; mV, mandibular arch muscles
innervated by the trigeminal nerve; mVII, hyoid arch muscles
innervated by nerve VII; ot, otic vesicle; otc, otic capsule; pl,
Platt’s vesicle; pm, premandibular mesoderm; pmc, premandibular cavity; III–XII, cranial nerves.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
6
S. KURATANI AND K.G. OTA
been described in the head mesoderm of various
gnathostome embryos (reviewed by Jacobson, ’88),
there are no equivalent bulges in the cephalic
mesoderm of the lamprey species, Lethenteron
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
japonicum, observed with the same technique
(Kuratani et al., ’99). Under the influence of
‘‘elasmobranch worship’’ (reviewed by Gee, ’96;
Fig. 3A and B), some classical histological
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
observations have detected clearly segmented mesodermal blocks (Koltzoff, ’01; Veit, ’39; Damas, ’44).
Cephalic somitomeres and head cavities should
not be confused: somitomeres are, if they do exist,
mesenchymal masses of cells that are only incompletely separated from each other, whereas
head cavities are histologically very overt epithelial structures that show a typical configuration in
the late pharyngula (Jacobson, ’88). More importantly, they differ in number: head cavities appear
as three pairs, whereas there are seven somitomeres (Fig. 3A and E). Therefore, these two kinds
of segments cannot be identical developmental
units if a pair of two successive somitomeres
corresponds with one head cavity, as suggested
by Jacobson (’88). It is also possible that if
head cavities do not represent the primary
segments of the vertebrate head, somitomeres
do, or vice versa.
To ascertain the developmental nature of the
cephalic somitomeres, a computer-based analysis
was performed to measure the size of the head
mesoderm (Freund et al., ’96). However, this did
not detect any metameric bulge in the mesoderm.
The expression of mesodermal segmentation-related genes has also been analyzed recently in
early chicken embryos. Some regulatory genes
exhibit oscillating pulses of regulation, with each
cycle corresponding to the generation of one
somite (reviewed by Maroto and Pourquie, 2001).
Gene expression analyses in early chicken embryos detected only two pulses of cyclic gene
expression: one in the prechordal plate region,
and the other in the rest of the entire head
mesoderm (Jouve et al., 2002). This is apparently
consistent with the view that the only mesodermal
segment in the lamprey head is the premandibular
mesoderm, which is derived directly from the
prechordal plate (Kuratani et al., ’99). At least in
the lamprey, most of the head mesoderm shows no
sign of primary segmentation. It is only secondarily ‘‘regionalized’’ by nonmesodermal embryonic
structures, such as the eyes, pharyngeal pouches,
7
and otocysts. As these structures grow, the head
mesoderm becomes divided into domains, to which
particular names such as the ‘‘mandibular mesoderm’’ and the ‘‘hyoid mesoderm’’ can be applied
(Fig. 3). Importantly, real segmentation and
secondary regionalization of the head mesoderm
cannot be distinguished merely by histological
observation (see Koltzoff, ’01 and Kuratani
et al., ’99).
Do the head cavities then represent the primary
segments of the vertebrate head mesoderm? So far,
there have been no reports of the presence of overt
head cavities (with the typical epithelial configuration) in lamprey embryos (Kuratani et al., ’99; also
see Sewertzoff, ’16), although the segmental and
enterocoelic nature of the mandibular mesoderm
has been implied (Holland et al., ’97; also see
Koltzoff, ’01; Veit, ’39; Damas, ’44), i.e., the
lamprey and gnathostome mandibular mesoderm
commonly expresses the En homologue, as do the
rostral enterocoelic somites of amphioxus. The
situation is probably the same in another cyclostome species, the hagfish, because no epithelial
coelomic structures have been reported in the
limited literature dealing with the embryos of
Bdellostoma stouti, or in our own embryonic
specimens of Eptatretus burgeri (data not shown;
also see Fig. 4).
In purely morphological terms, the full set
(three pairs) of typical head cavities is only
encountered in elasmobranch embryos (Balfour,
1878; for another pair of rostral cavities, called
Platt’s vesicles, see Goodrich, ’30; Jarvik, ’80;
Jefferies, ’86; Kuratani, 2003, 2005a; Fig. 3B).
Some of the so-called chondrosteans possess
premandibular and mandibular cavities, but not
hyoid cavities (De Beer, ’24; Kuratani et al., 2000),
and a number of amniote embryos, including the
human embryo, have been described as having
only a premandibular cavity (van Wijhe, 1883;
Adelmann, ’26; Wedin, ’49, ’53a, b; Fraser, ’15;
Gilbert, ’47, ’53, ’54, ’57; Fig. 3D). What seems
clear from these observations is that, unlike the
Fig. 4. Hagfish and their embryos. (A and B) The adult hagfish, Eptatretus burgeri, studied in our laboratory. (A) Slime
secreted by the hagfish in captivity. (C–E) A 7.0-mm embryo at the stage corresponding to the late neurula to early pharyngula.
Note in (C) that the somites are restricted to the level posterior to the otic pit (ot). (D) Enlargement of the head region showing
some anatomical structures. (E) The head region of the same embryo before treatment with methyl benzoate, showing the
surface structure. (F–K) Transverse sections of the same embryo at the level of (F) the optic vesicle (ov), (G) hypothalamus (ht),
(H) mandibular arch, (I) rostral hindbrain (hb), (J) otic placode (ot), and (K) posterior hindbrain, showing the ectodermal
placodes (arrowheads). The placodes are continuous, forming a horseshoe pattern, similar to the hypothetical pan-placodal
domain (see Fig. 3c). Arrow in (I) indicates the tendency of the putative crest cells to form an anlage for the trigeminal nerve
root. coe, coelom; fb, forebrain; lp, lens placode; oc, oral cavity; ph, pharynx; po, postotic placode; TC, mandibular arch
ectomesenchyme; trg, trigeminal placode. Bars 5 100 mm.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
8
S. KURATANI AND K.G. OTA
apparently primitive and primary preconceptions
we tend to have about head cavities, they seem to
have arisen in certain basal gnathostome as a
synapomorphy, and have thereafter tended to
disappear, from a posterior to anterior direction,
in the embryos of the crown groups (see Kuratani,
2003, 2005a; Fig. 3D).
As noted above, the significance of elasmobranch
head cavities can be interpreted in opposite
directions depending on whether we see the
cavities as ancestral traits or derived traits.
Similarly, if we consider the vertebrate head to
be a derived trait, it can be described as a
nonsegmental rostral part of the body axis, in
terms of the mesodermal developmental pattern.
However, it might once have had an array of
primary mesodermal segments in the ancestral
condition, somewhat similar to that still seen in
amphioxus. Therefore, the search for mesodermal
segments in the vertebrate head, as schematized
by Goodrich (’30), is in some ways a search for a
putative plesiomorphy in the developmental program of the vertebrates. Conversely, to stress the
absence of mesodermal segments would imply the
derived nature of the vertebrate head, a feature
that is not shared by the hypothetical ancestor.
Thus, the neural crest cells in the vertebrate head
are closely related to the pharyngeal arches and
rhombomeres, and not to any possible mesodermal
segments, because the crest cells are distributed in
a metameric pattern only in the trunk, where the
segmented crest cell streams correspond to the
somites (Fig. 3B).
The vertebrate head, seen as a vertebratespecific trait, lacks the typical somites and
somite-derived segmental pattern. It is not clear
whether cephalic somitomeres or yet unknown
gene expression patterns can possibly demonstrate vestigial ancestral segments, and these have
yet to be proven. The gnathostome head cavities
appear to have been overemphasized. They are
at best a synapomorphy of the gnathostomes
(Fig. 3D), and an elasmobranch-like condition
cannot represent the vertebrate ground plan in
the phylogenetic context, even if it served as an
embryological archetype in the transcendental
background. However, if head cavities represent
a pan-vertebrate trait, we have to assume that
they have been secondarily lost in the cyclostome
lineage. This may be plausible in the hagfish, in
which the eyes are a degenerative condition.
However, from the anatomy of the lamprey and
many nonelasmobranch gnathostomes that share
a similar (but not identical) set of extrinsic eye
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
muscles (Marinelli and Strenger, ’54), we can
safely say that the head cavities represent neither
fundamental nor essential developmental components in the basic patterning of the vertebrate
head.
The most curious question in the context of head
segmentation seems to remain unanswered. We do
not know if the ancestor of the vertebrates
possessed mesodermal segments that extended to
the rostral tip of the anteroposterior axis. It is
undeniable that until recently comparative morphologists have been somewhat biased in imagining the ancestor of the vertebrates to be like
amphioxus, possessing a segmented mesoderm to
the rostral tip of the body axis (as is the tentative
assumption on which the present paper is based).
This would cause them to think that the vertebrate ancestor lost some head somites. Molecular
phylogeny, in contrast, has suggested that tunicates, rather than amphioxus, are the closest sister
group to the vertebrates (Delsuc et al., 2006). To
close this section, it is worth mentioning that,
although the tunicates are likely to have independently experienced substantial modifications in
their developmental program and body plan, they
resemble the vertebrates in that their typical
trunk-like morphology is restricted to their tail
region, just as the true somites of vertebrates are
restricted to the postotic level (even in lamprey
and hagfish embryos; see Dean, 1899; Kuratani
et al., ’99; Ota et al., 2007).
NEURAL CREST, PLACODES,
AND THE HOX CODE
Since Gans and Northcutt (’83), the neural crest
and placodes have been recognized as the developmental components most essential for the
patterning of the vertebrate head, and as possibly
the most fundamental synapomorphy defining this
group.
Induced on both sides of the neural plate, the
neural crest delaminates to yield migratory cells,
called the ‘‘neural crest cells’’, which provide
extensive ectomesenchyme (mesenchyme derived
from the ectoderm) and mainly occupy the ventral
portion of the embryonic head, extending posteriorly to form the ‘‘neck’’ region in gnathostomes
(reviewed by Noden, ’88; and by Kuratani, ’97; for
the posterior limit of the cephalic crest cell
distribution, see Matsuoka et al., 2005). Therefore,
the vertebrate pharyngeal arch contains neural
crest-derived mesenchyme as well as the mesodermal mesenchyme that forms the core of each
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
arch and that eventually differentiates into the
pharyngeal arch muscles (see Kimmel et al., 2001).
In the head region, the crest-derived ectomesenchyme differentiates into various types of cells
and tissues, including peripheral sensory neurons
and supporting cells in the cranial ganglia,
endocrinal cells, connective tissues for the cranial
muscles and vessels, skeletal tissues, including
both bone and cartilage as well as pigment cells
(reviewed by Le Douarin, ’82; Le Douarin and
Kalcheim, ’99). In particular, the skeletal differentiation of the neural crest characterizes the
vertebrate head. However, these data largely
depend on a few experimental model animals,
and the skeletogenic potency of the crest in other
vertebrate taxa has yet to be identified with proper
labeling experiments, especially in the trunk
region.
In the history of comparative embryology, the
presence and developmental roles of the cephalic
neural crest have been neglected in the interpretation of the basic morphological plan of the
head, although these cells are recognized by
experimental embryologists to be the source of
the craniofacial and pharyngeal arch skeletons
(Platt, 1893; reviewed by Hall and Hörstadius, ’88;
Hanken and Hall, ’93; Hall, ’99). However, both
immunohistochemical staining and recent embryological observations based on in situ hybridization, which show regulatory gene expression
patterns, have confirmed that the cephalic crest
cells form three distinct cell populations (reviewed
by Kuratani, ’97, 2005a; Horigome et al., ’99).
These are distributed in a semisegmental pattern
in the head, connecting an even-numbered rhombomere with each pharyngeal arch (Fig. 5;
Kuratani and Eichele, ’93; Lumsden, 2004). Thus,
the segmental developmental program is more
plausible in the vertebrate head than in the
patterning of the central nervous system and
branchiomeric skeletomuscular system (Fig. 5).
Hox gene expression patterns are also shared by
the rhombomeres and crest cells at the same axial
level, implying that these embryonic cell populations are specified commonly as segmental units,
by the same set of gene transcripts, along the
anteroposterior axis (Hox code; Hunt et al., ’91;
Fig. 5).
The basic topographical patterns of crest cell
distribution are similar in lamprey and gnathostome embryos, as are the expression patterns of
some of the Hox genes expressed in the rostral
part of the head (Kuratani et al., ’97b; Horigome
et al., ’99; Meulemeans and Bronner-Fraser, 2002;
9
Fig. 5. Rhombomeres, neural crest cells, and placodes.
(A) Hindbrain of the mouse embryo with regulatory
gene expression patterns, branchiomeric motoneurons, and
primordial cranial nerve roots. In gnathostomes, the boundaries of the Hox gene expression domains and the origins
of the motoneurons correspond to those of the rhombomeres
(r1–7), the metamerical neurepithelial compartments of
the hindbrain. b1–3, pharyngeal arches; V–X, positions of
the branchiomeric nerve roots. Modified from Kuratani
(2004b). (B) As shown in the chicken embryo, the segmental
pattern of the cranial nerve roots originates in the selective
adhesion of the crest cells to even-numbered rhombomeres.
In contrast, the migration of trunk crest cells is divided
by the posterior halves of the somites (ps), which leads
to the somitomeric segmental pattern of the crest-derived
dorsal root ganglia (sp). sc, spinal cord. Modified from
Kuratani (2004b). (C) Placodes are specified in a horseshoe-like pattern along the outer edge of the neural plate
of the early vertebrate embryo. This ectodermal domain
is defined by the combined expression of some regulatory
genes, rostral and lateral to the neural crest domain.
ebp, epibranchial placode; lp, lens placode; hyp, hypophysial placode; np, nasal placode. Modified from Schlosser
(2006).
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
10
S. KURATANI AND K.G. OTA
McCauley and Bronner-Fraser, 2003; Takio et al.,
2004, 2007; but also see Cohn, 2002). As in most
nonelasmobranch gnathostome embryos, the
lamprey cranial nerve roots also form on evennumbered rhombomeres, reflecting the selective
adhesion of the neural crest cells, the precursors of
the cranial nerves, to those rhombomeres
(Kuratani and Eichele, ’93; Kuratani et al., ’97b;
Horigome et al., ’99; Fig. 5A and B).
Although the morphology of the cephalic ectomesenchyme and the Hox code patterns remain to
be determined in the hagfish, the basic embryonic
patterns, as described above, appear to have been
shared by the common ancestor of the cyclostomes
and gnathostomes, an animal close to the ancestor
of the entire vertebrates (Takio et al., 2004;
summarized in Figs 8 and 9). Simultaneously, this
pattern appears to have been a developmental
constraint, imposing a surprisingly conservative
morphological pattern on the cranial nerves and
their innervation patterns, as well as on the basic
morphology of the pharyngeal arch derivatives. It
is even possible to establish the morphological
homology of these anatomical elements, at least to
some extent, among all vertebrate species (Holmgren, ’46; Wicht and Northcutt, ’95; Kuratani et al.,
’97b; Kuratani, 2005a; regarding the rostral part
of the head, including the jaw, see below). Here,
we deduce that the developmental constraints on
the vertebrates are the developmental mechanical
basis of this homology. These constraints are also
pattern generative for the distribution of these
cephalic crest cells, setting up the basic topographical relationships of the cells, which will later
serve as the basis for the local tissue interactions
required for the further differentiation of the
embryonic head (Kuratani, 2003).
Naturally, the origin of the neural crest must
have predated the establishment of the vertebratelike head, as discussed above. In this context,
neural crest-like cells have recently been found in
tunicate larvae; these cells share similar sets of
gene expressions with true neural crest cells and a
similar capacity to differentiate into melanocytes
(Jeffery et al., 2004). Consistent with this, we have
recently confirmed the development of delaminating crest cells in the hagfish embryo (Ota et al.,
2007), in contrast to the previous observation that
the hagfish crest develops as an epithelial growth
(Conel, ’42). The hagfish crest also appears to be
specified by part of the highly conserved molecular
cascade that functions in the neuroectodermal
specification of other vertebrates, and involves
Pax6, Pax3/7, SoxE, and Sox9. The hagfish and
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
lampreys seem to consistently form a monophyletic group, the cyclostomes, and the origin of the
neural crest should be sought in the nonvertebrate
chordates (summarized in Figs 8 and 9).
The ectodermal placodes are the source of the
sensory ganglia, sensory organs, and possibly also
the lateral line in the hagfish embryo. Although
the lateral line is missing in the adult lamprey,
Wicht and Northcutt (’95) have described an
independent series of placodes (the dorsolateral
placodes) as a possible remnant of the lateral line
primordia in the hagfish embryo (also see Neumayer, ’38; Holmgren, ’46). A horseshoe-shaped
continuous epidermal thickening was observed in
one of our hagfish embryo specimens, corresponding to the early pharyngula, and a Sox9-positive
otic placode arose as part of the latter (Ota and
Kuratani, 2007). This pattern of placode development is reminiscent of the ‘‘pan-placodal domain’’
proposed by Schlosser (2005, 2006; Fig. 5C). Such
a pattern of placodal development is very hard to
observe in the lamprey embryo, because the
epidermal cells contain large yolk granules in the
early pharyngular stages.
In addition to a delaminating neural crest, the
common vertebrate ancestor seems to have possessed a placode (summarized in Figs 8 and 9).
The otic placode was already differentiated in this
ancestor, although it is assumed to have been a
specialized lateral line placode. No animal has yet
been found that displays an intermediate state
that would support the lateral line origin of the
otic placode.
ORIGIN OF THE JAW
The Hox genes described above are generally
classified according to their relative positions in
the cluster they form, into 13 (or 14) paralogue
groups (PGs). The Hox code generally refers to the
organized nested expression patterns of these
genes along the anteroposterior axis, which is
commonly observed in the embryos of all the
metazoa. Genes with smaller PG numbers are
expressed more rostrally and those with larger PG
numbers more caudally.
Among the vertebrate Hox genes, the PG2 genes
are expressed in the ectomesenchyme of the
second and more posterior pharyngeal arches,
and the PG3 genes in the third and more posterior
arches. Curiously, the mandibular arch never
expresses any of the Hox genes and the mandibular arch in gnathostome model animals appears
to be specified by this Hox-code default state (Rijli
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
et al., ’93; Gendron-Maguire et al., ’93; Couly
et al., ’98). These expression patterns, or the
rostral part of the cephalic Hox code, are commonly observed in lampreys and gnathostomes
(Takio et al., 2004, 2007).
The mandibular arch is also identifiable in the
lamprey based on its innervation by the trigeminal
nerve as well as on its generalized embryonic
morphology (Kuratani et al., 2001; Takio et al.,
2004). Again, this pattern of Hox expression is
very likely to have been established in the
vertebrate ancestor (summarized in Figs 8 and
9). However, in the lamprey, no Hox genes are
expressed specifically in the arch 4 ectomesenchyme or extending posteriorly, but a PG4 gene
expresses a gradient of transcripts that becomes
most intense in the caudalmost pharynx. This is
quite unusual for vertebrate Hox genes. In the
lamprey, multiple Hox genes are also expressed in
the caudalmost pharyngeal pouch, which has not
been observed in any gnathostome species. Thus,
the organized nested pattern of the Hox code is
specific to the gnathostomes, and the cyclostomes
seem to have evolved their own derived pattern of
Hox regulation (Takio et al., 2004).
The part of the Hox code shared by the
vertebrates is consistent with the general rule of
pharyngeal arch differentiation. That is, only the
mandibular and hyoid arches differentiate into
their own specific morphologies, namely the jaw
and jaw supporting apparatus, whereas the more
posterior arches (postotic arches) almost always
have a monotonous morphology, forming the
respiratory skeletal complex. This situation is
repeated in the ammocoete larvae of the lamprey,
indicating that this triple morphological specification is probably shared by all vertebrate species. In
other words, no vertebrates have been found
displaying the unspecified condition of the branchiomeric units, as illustrated in textbooks, in
which all the arches are in an undifferentiated
state. The mandibular arch may have always
differed from the rest of the arches, being involved
in the formation of the oral apparatus.
To form a jaw, the gnathostome-specific trait,
the mandibular arch must be divided into dorsal
and ventral halves, and each half must be specified
in its own way. Although the Hox genes function
in specification along the anteroposterior axis,
another group of homeobox genes, the Dlx genes,
exhibit nested expression patterns along the
dorsoventral axis of each arch (Depew et al.,
2002). However, from a preliminary report of
Petromyzon marinus (Neidert et al., 2001), the Dlx
11
code, unlike the Hox code, does not seem to be
present in the lamprey. Therefore, the completion
of Hox–Dlx codes, the Cartesian grid of homeobox
gene expression in the pharyngeal ectomesenchyme, appears to be at best a gnathostome
synapomorphy (summarized in Figs 8 and 9).
Again, the condition in the jawless stem gnathostomes remains unknown. Is it possible to determine whether it dates back to a point before the
acquisition of the jaw, or after it? Would the jaws
of basal gnathostomes, in which the upper and
lower jaws look alike, reflect the pre-Dlx-code state
of our ancestor (reviewed by Koentges and
Matsuoka, 2002)?
The gnathostome jaw is specified by the Hox
code default state of the mandibular arch ectomesenchyme. In fact, overexpression of the PG2
genes in the mandibular arch partially transforms
the arch into the identity of the hyoid, and the
deletion of Hoxa2 expression in the hyoid arch
results in the partial transformation of the hyoid
arch into the identity of the mandibular arch (Rijli
et al., ’93; Gendron-Maguire et al., ’93; Pasqualetti
et al., 2000; Grammatopoulos et al., 2000; Hunter
and Prince, 2002). The Hox default state of the
lamprey mandibular arch implies that the jaw
evolved via the differentiation of the pharyngeal
arch, which had already been specified as the
mandibular arch, and by additional changes in the
developmental program, which may not have
involved the Hox code per se (but also see Cohn,
2002). Conversely, comparative embryology has
shown that the position and expansion of the oral
area differ between the lamprey and the gnathostomes. Thus, the upper and lower lips of ammocoete larvae are not homologous to the upper and
lower jaws. The lamprey upper lip differentiates
from a premandibular component, unlike the
upper jaw, which develops as part of the mandibular arch (Kuratani et al., 2001; Shigetani et al.,
2002).
Curiously, however, the genes involved in the
establishment of the proximal and distal polarity
of the oral apparatus are expressed in similar
patterns in the gnathostomes and lamprey, as if
the upper and lower lips were homologous to the
upper and lower jaws, respectively (Shigetani
et al., 2002). The expression of homologous genes
is uncoupled from the morphological homology.
These are two growth factor-encoding genes, Fgf8
and Bmp2/4, the cognates of which are expressed
in the proximal and distal epidermis of the oral
region, respectively. Their target genes, Dlx1 and
Msx, are also ectomesenchymally expressed with
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
12
S. KURATANI AND K.G. OTA
the same polarity in both animals. Therefore, the
cascade of genes was probably established in the
agnathan ancestor, and regulation of the genes
encoding growth factors was secondarily shifted
posteriorly in the lineage of the gnathostomes
(summarized in Figs 8 and 9). As a result of the
heterotopic shift in this signaling system during
evolution, the morphological homology of the oral
apparatus was lost in the transition to the jawed
state of the vertebrates. Thus, the jaw can be seen
as a true novelty because it does not have an exact
homologue in its ancestor.
TRABECULA IN VERTEBRATES
The heterotopic origin of jaw evolution may
possibly explain the appearance of the gnathostome-specific trabecula, the precursor of the interorbital and nasal septa in the gnathostomes
(reviewed by Goodrich, ’30; De Beer, ’37; Kuratani
et al., ’97a; Kuratani, 2005a, b; Figs 6 and 7). Two
major sections can be distinguished in the early
gnathostome neurocranium: the posterior part,
generally recognized as a pair of parachordal
cartilages derived from the mesoderm; and the
rostral part, derived from the neural crest-derived
ectomesenchyme (De Beer, ’37; Couly et al., ’93;
Fig. 6A). The parachordal and associated structures
that comprise the posterior neurocranium are
collectively called the ‘‘chordal cranium’’, indicating
that this portion arises at the level of the notochord,
which is required for the chondrification of the
mesodermal cells, just as the notochord is necessary
for the sclerotomal differentiation in the somites
(Couly et al., ’93; Fig. 6). The more rostrally located
neurocranial portion is called the ‘‘prechordal
cranium’’ and, as its name implies, is located
rostral to the notochordal tip (Couly et al., ’93).
In the gnathostomes, the trabecula is the major
precursor of this prechordal cranium (see De Beer,
’37; Fig. 6C). It thus appears to be derived from
the prechordal ectomesenchyme, which is also
called the ‘‘premandibular ectomesenchyme’’
when it is viewed as the rostral part of the
extensive trigeminal crest cells, which also includes the ectomesenchyme of the mandibular
arch (Shigetani et al., 2000; reviewed by Kuratani,
2004a). The term ‘‘premandibular’’ tends to imply
the classical morphological concept of ‘‘branchiomerism’’ or the repetition of the pharyngeal
arches, which assumes the presence of one or
more pharyngeal arches in front of the mandibular arch (Sewertzoff, ’11; Goodrich, ’30;
Portmann, ’76; Jarvik, ’80; reviewed by Kuratani
et al., ’97a). However, current evolutionary embryology should not support such a hypothesis
without reference to the developmental pattern of
the endodermal pouches in actual embryos. Therefore, if the premandibular and mandibular portions of the trigeminal crest cells form the lamprey
oral region, no trabecular cartilage could differentiate in this animal.
However, the trabecula has been described in
the lamprey (Sewertzoff, ’16; Neumayer, ’38;
Johnels, ’48; reviewed by Holmgren, ’46; also see
Koltzoff, ’01; De Beer, ’37; this cartilage in the
lamprey embryo has been given various names by
each author; Fig. 7), and some authors have even
described it as having a neural crest origin
(Damas, ’44). However, the morphology of this
cartilage (the initial anlage found at the level of
the mandibular artery and lateral to the notochord) and the results of preliminary labeling
studies suggest that it is most likely to be of
mesodermal origin, probably corresponding to
the anteriorly extended parachordal cartilage
Fig. 6. Neural crest-derived prechordal cranium. In the gnathostome skull, the rostral half of the so-called neurocranium
(rostral to the level of the hypophysis, colored light) is derived from the neural crest, unlike the posterior moiety derived from
the mesoderm. The crest-derived neurocranium does not require the presence of the notochord to chondrify. (A) The origin of
the chicken neurocranium. The transverse dotted line indicates the boundary of the crest- and mesoderm-derived parts of the
neurocranium. The former is called the ‘‘prechordal neurocranium’’ and the latter the ‘‘chordal neurocranium’’. Modified from
Couly et al. (’93). Note that the chordal cranium is coextensive with the notochord (n). (B) Developmental composition of the
shark cranium by extrapolating the chicken data in (A). (C) Early configuration of the gnathostome neurocranium and its
origins. The neurocranium consists of the crest-derived rostral part, which primarily consists of the trabecula and nasal capsule
(and also the polar cartilage, which can be seen as the posteriormost element of the trabecula), and the mesoderm-derived
chordal part, which consists of the orbital, parachordal, and occipital cartilages. Except for the occipital cartilage, which has a
segmental anlage derived from the rostral-most group of somites, most of the chordal cranium is unsegmented head mesoderm.
Note that the boundary between the prechordal and chordal crania corresponds to the position of the hypophysis (hyp). e, eye;
fb, forebrain; fh, hypophysial foramen; ios, interorbital septum derived from the trabecula; mb, midbrain; n, notochord: nas,
nasal capsule; ncr, neurocranium; np, nasal placode; oc, orbital cartilage; occ, occipital cartilage; pcc, parachordal cartilage; tr,
trabecula (of gnathostomes); vcr, viscerocranium; III–VI, nerves innervating the extrinsic ocular muscles.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
(Sewertzoff, ’16; Neumayer, ’38; Johnels, ’48;
Kuratani et al., 2004; Fig. 7C and F).
A similar morphology is found in the hagfish, as
is a similar type of confusion. In the most detailed
description by Holmgren (’46) of the Myxine
13
embryo, the trabecular cartilages are illustrated
as a pair of rods located lateral to the notochord,
again too posterior for the prechordal cranium.
Rostral to these rods, a transverse cartilage
commissure unites the mid parts of the trabeculae
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
14
S. KURATANI AND K.G. OTA
at a level just behind the hypophysis (Fig. 7F). The
pair of cartilaginous rods further extends rostrally
to reach the nasal region. The rostral bar thus
resembles the gnathostome trabecula, but the
posthypophysial portion of the rod more likely
represents the parachordal of the gnathostomes.
This longitudinal cartilage of the hagfish, as a
whole, probably represents a composite of the
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
parachordal and the true trabecula, as seen in the
chondrocranium of the early gnathostome, but the
anteroposterior level of fusion has been misunderstood. The rostral bar (true trabecula) does not
seem to exist in the lamprey, but apparently
differentiates into the mesenchymal component
of the upper lip. At any rate, the rostral trabecula
of the hagfish embryo lies in the ‘‘palate’’ of this
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
animal (the rostral septum that separates the
nasopharyngeal duct and the oral cavity), and this
trabecula occupies the same topographical position as the upper lip of the ammocoete larva of the
lamprey (reviewed by Kuratani et al., ’97a, 2001).
MONORHINY–DIPLORHINY TRANSITION
AND THE PITUITARY: DEVELOPMENTAL
BURDEN AND KEY INNOVATION
The mesodermal origin of the cyclostome
trabecula is consistent with the heterotopy theory
of the origin of the gnathostome jaw described
above. In this theory, the true trabecula in
gnathostomes, the neural crest-derived prechordal
cranial element, can only arise from an embryo in
which the premandibular component of the
ectomesenchyme is freed from the formation of
the upper lip, as found in lamprey larvae (Fig. 8).
However, the heterotopic caudal shift in tissue
interactions (for oral patterning), as noted above,
alone does not seem to automatically lead to the
appearance of the trabecula, because this cartilage
requires a space in which to chondrify along the
medial sagittal line between the nasal placodes on
both sides. This is not present in the lamprey
(Kuratani et al., 2001; Kuratani, 2005b). The
lamprey has only a single anlage along the midline
for the nasal epithelium, and the posterior portion
of this placode also differentiates into the pituitary. Because of this common anlage, called the
‘‘nasohypophysial plate’’, the premandibular
ectomesenchyme in the lamprey cannot simply
grow rostromedially to occupy the space, in a
manner similar to that of the gnathostome
trabecula. Therefore, we can easily imagine that
the transition from monorhiny (the single-nostril
state) to diplorhiny (paired-nostril state; see
15
Janvier, ’96) was a prerequisite for the appearance
of the trabecula, which is very likely to have
predated the heterotopy described above (summarized in Figs. 8 and 9).
Given that the monorhinous state of the
lamprey represents the plesiomorphic condition,
the transition toward diplorhiny would probably
have happened long after the split between the
cyclostomes and gnathostomes, which is estimated
to have been more than 500 mya (Figs. 8 and 9;
Kuraku and Kuratani, 2006). This assumption is
consistent with the fact that there are several
fossil osteostracan groups (stem gnathostomes)
showing two olfactory organs (diplorhinous, at
least internally). A situation such as is seen in the
galeaspids, in which the two olfactory organs open
into one external nostril, implies that the transition from monorhiny to diplorhiny may not have
been a simple event, but that it rather proceeded
in multiple steps of morphogenetic changes, as did
the origin of the trabecula. In this context, the
embryonic origin of the trabecular cartilage is not
well understood, even in the crown gnathostomes.
This cartilage may arise not only from the
postoptic premandibular crest cells, but the
preoptic crest cells may also participate in its
formation, which has not been discussed above.
The basic morphological distribution pattern of
the premandibular crest cells per se is apparently
shared by the lamprey and the gnathostomes
(Horigome et al., ’99; Kuratani et al., 2001).
The nasohypophysial plate in the lamprey is
recognized as the common anlage for the nasal
epithelium and the adenohypophysis (reviewed by
Kuratani et al., 2001; Uchida et al., 2003). In the
hagfish, a similar median placode also occurs, and
its posterior fold has been explained as the
primordium of the adenohypophysis, as observed
Fig. 7. Cranial development in the cyclostomes. Cartilage elements are either shaded or shown in specific colors when the
embryonic origin of the cartilage is inferred (see ‘‘E’’ for reference). (A–C) Neurocranial development of the lamprey. (A and B).
Lateral and ventral views, respectively, of the neurocranium in an early ammocoete larva. The cartilage rod, called the trabecula
of the lamprey (trl), represents an anteriorly extended parachordal. (C) In an early embryo of the lamprey, the anlage of the
trabecula appears lateral to the notochord at the level of the first aortic arch. Thus, the trabecula is initially chordal in its
anteroposterior level, and its location is similar to those of mesodermal elements. Based on Johnels (’48). (D–F) Hagfish. (D) The
chondrocranium of the hagfish by Cole (’05). (E and F) Embryonic chondrocranium seen from lateral (E) and dorsal (F) views,
as depicted by Holmgren (’46). In the hagfish embryo, there is a pair of longitudinal cartilage bars, called the ‘‘trabecula’’ (trh).
From the position of the notochordal rostral tip, as well as that of the hypophysis, it seems most likely that only the rostral part
of the hagfish trabecula is equivalent to the gnathostome trabecula (tr). The caudal part corresponds to the gnathostome
parachordal cartilage (pc). In hagfish embryology, only the caudal nodule lateral to the notochord is called the parachordal. In
(E), the cartilages are colored according to the morphological evaluation used by Holmgren (’46) for arches 1–3; refer to the
Discussion of this paper for the distinction between parachordal and trabecular cartilages. e, eye; hf, hypophysial foramen; n,
notochord; oc, otic capsule; ot, otocyst; pc, real parachordals; pch, parachordals of the hagfish; pcl, parachordals of the lamprey
(posterior portion of the parachordals); tr, real trabecula (homologous to the cartilage of the same name in the gnathostomes);
trh, trabecula of the hagfish; trl, trabecula of the lamprey (anteriorly extended parachordal).
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
16
S. KURATANI AND K.G. OTA
Fig. 8. Sequence of changes in the embryonic and morphological patterns evolving toward the gnathostomes. The sequence
of changes, from the perspective of developmental burden, is plotted onto the phylogenetic tree. Each set of traits represents
specific developmental constraints that define the taxonomic groups.
in lamprey development (Kupffer, ’00). However,
in the development of E. burgeri, the adenohypophysis does not differentiate from this posterior
fold, but from a certain anlage that is located more
posteriorly, in proximity to the notochordal rostral
tip and the ventral diencephalon (unpublished
observation by Ota and Kuratani). Thus, the
nasohypophysial plate may not be homologous
even in the lamprey and hagfish. In this context,
the adenohypophysis of the hagfish has been
inferred to arise from an endodermal component,
not from the rostral ectoderm (Gorbman and
Tamarin, ’85). It will require more observations
and experiments on hagfish embryos to resolve
this issue. The topographical expression domains
of signaling molecules, such as fibroblast growth
factors, bone morphogenetic proteins, and Hedgehog, are not conserved, at least between lamprey
and gnathostome development, whereas the embryonic distribution of transcription factors is
always associated with identical cell types (TTF1 as a ventral diencephalic marker, pitx1 as a
rostral ectodermal marker, and so forth). It seems
very likely that the evolution of the regulation of
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
signaling molecules is involved in heterotopic
evolution, especially regarding the factors that
change the sites of tissue interactions.
CONCLUSIONS AND PERSPECTIVES
Developmental constraint has been defined as
the bias that limits the direction of evolution
(Maynard-Smith et al., ’85). Pattern-generative
constraints are a subtype of developmental constraints, in which the bias imposed on the
ontogenetic trajectory by a certain fixed developmental pattern at specific and critical embryonic
stages (such as the appearance of somites or
rhombomeres) establishes stereotypical topographical relationships between tissues or cells.
Pattern-generative constraints can even establish
the highly sophisticated and conserved gene
expression networks or their expression domains
that tend to result in fixed patterns of phenotypes.
It is in these patterns of phenotypes that comparative morphologists have perceived morphological homologies at the anatomical level.
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
Comparative embryology has identified such
morphological identities at the level of the primordia or cell populations, which lead directly to
the adult anatomy. Thus, developmental constraints are the bases of morphological homologies, and even possibly the developmental bases
for taxonomy. They are specifically associated with
a certain animal group, which is defined by a
limited variation in its morphological pattern (due
to the specific developmental constraint), and
largely overlap the synapomorphy that defines
the same group. From this perspective, evolutionary developmental biological studies are more or
less aimed at determining the sequence of changes
in these developmental constraints or their developmental mechanical backgrounds, which underlie the evolution of new patterns along the
phylogenetic tree. They focus at various levels,
such as the molecular and cellular biological, and
genomic levels, to understand the dynamic roles of
developmental patterns and processes as the
targets of natural selection.
From this developmental view of phenotypic
evolution, novelty can be defined as a change in
the ancestral constraints, not simply a modifica-
17
tion of the same basic pattern. This can be seen in
the evolution of the bat wing, which is anatomically identical to the foreleg of other mammals.
Heterotopy, defined as a shift in developmental
place, provides an opportunity to generate
completely new patterns of development simply
by changing the place of the tissue interaction,
based on the same set of regulatory genes as
are involved in jaw evolution. The concept of
heterotopy was initially coined by Haeckel to
describe, for example, instances in which identical
structures or cells, such as germ cells, are derived
from different germ layers in different animals
(Haeckel, 1874; also see De Beer, ’58). As can
easily be imagined, if the developmental constraints are strong enough to maintain conserved
developmental trajectories, including cell lineages,
von Baer’s ‘‘germ layer theory’’ (in which homologous organs are always derived from the same
germ layer) cannot be refuted (Baer, 1828; see
De Beer, ’58).
In evolutionary developmental studies of the
cyclostomes, there remains a curious question to
be resolved in relation to the validity of the
germ layer theory, namely, the evolutionary and
Fig. 9. Sequence of changes in the developmental programs evolving toward the gnathostomes. The traits plotted on the
previous tree (Fig. 8) can be rephrased as changes in developmental programs, such as gene duplications, changes in molecular
cascades, or changes in gene regulation. Some of the suggested changes are plotted on this tree so that they can be correlated
with the embryonic patterns of the previous tree.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
18
S. KURATANI AND K.G. OTA
developmental origin of the vertebrate pituitary
mentioned above. In the hagfish, this organ
appears to arise from the endoderm, not from
the ectoderm as in the lamprey and gnathostomes
(Gorbman, ’83, ’85). From the limited number of
specimens and illustrations available today, this
organ does appear to arise from the endoderm.
However, it is possible that this reflects a
complicated reorganization of the oropharyngeal
membrane specific to this animal, and that the
hagfish pituitary also arises from the ectoderm, as
in the rest of the vertebrates (see Jefferies, ’86).
Consistent with this, the younger embryo of the
hagfish appears to possess a single median
placode, which resembles the nasohypophysial
plate of the lamprey (Kupffer, 1899; Gorbman,
’83). Even if this is so, we must consider the fact
that Hatschek’s pit in amphioxus, the putative
homologue of the vertebrate pituitary, undoubtedly arises from the endoderm (Hatschek, 1881;
see Uchida et al., 2003).
Finally, developmental constraints can be
coupled not only in an anatomically integrative
manner in the embryo, mainly due to cell–cell
interactions, but also during the time courses of
both evolution and development. As an example,
we have shown that the separation of the
agnathan-type nasohypophysial plate into the
pituitary and a pair of olfactory placodes, as seen
in gnathostome embryos (monorhiny to diplorhiny
transition), was a prerequisite for jaw and trabecular evolution (Figs. 8 and 9). These structures
are not only coupled as integrated units, but
one trait must precede the other to achieve
the integrity of the whole. At the evolutionary
time point at which diplorhiny first arose,
it might have simply been one of the possible
variations. However, once the ectomesenchyme
that filled the space between the double
nostril formed an adaptive trabecula, the
diplorhinous state could no longer arbitrarily
disappear, because of the ‘‘developmental burden’’
(Riedl, ’78) imposed by the later developmental
phenomenon. We now realize that the concept
of developmental burden is the ontogenetic key
to determining the phylogenetic sequence
of developmental events. Thus, evolutionary developmental biological studies may have now
attained scientific access to the formulation of
transcendental ideas (or tendencies) in classical
comparative embryology, collectively called ‘‘recapitulation’’, with our molecular embryological,
developmental biological, and phylogenetic
strategies.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
LITERATURE CITED
Adelmann HB. 1926. The development of the premandibular
head cavities and the relations of the anterior end of the
notochord in the chick and robin. J Morphol Physiol
42:371–439.
Baer KE von. 1828. Entwicklungsgeschichte der Thiere:
Beobachtung und Reflexion. Königsberg: Born Träger.
Balfour FM. 1878. The development of the elasmobranchial
fishes. J Anat Physiol 11:405–706.
Braun CB. 1998. Schreicher organs: a new craniate chemosensory modality in hagfishes. J Comp Neurol 392:135–163.
Cole FJ. 1905. A monograph on the general morphology of the
Myxinoid fishes based on the study of Myxine. Part I. The
anatomy of the skeleton. Trans R Soc Edinb 41:749–788.
Cohn MJ. 2002. Evolutionary biology: lamprey Hox genes and
the origin of jaws. Nature 416:386–387.
Conel JL. 1942. The origin of the neural crest. J Comp Neurol
76:191–215.
Couly GF, Coltey PM, Le Douarin NM. 1993. The triple origin
of skull in higher vertebrates: a study in quail-chick
chimeras. Development 117:409–429.
Couly G, Grapin-Botton A, Coltey P, Ruhin B, Le Douarin
NM. 1998. Determination of the identity of the derivatives
of the cephalic neural crest: incompatibility between Hox
gene expression and lower jaw development. Development
125:3445–3459.
Damas H. 1944. Recherches sur la développement de
Lampetra fluviatilis L. Contribution à l’ étude de lq
céphalogenèse des vertébrés. Arch Biol 55:1–284.
Dean B. 1899. On the embryology of Bdellostoma stouti.
A genera account of myxinoid development from the egg and
segmentation to hatching. Festschrift zum 70ten Geburststag Carl von Kupffer 220–276.
De Beer GR. 1924. The prootic somites of Heterodontus and of
Amia. Q J microsc Sci 68:17–38.
De Beer GR. 1937. The development of vertebrate skull.
London: Oxford University Press.
De Beer GR. 1958. Embryos and ancestors. Oxford: Oxford
University Press.
Delsuc F, Brinkmann H, Chourrout D, Philippe H. 2006.
Tunicates and not cephalochordates are the closest living
relatives of vertebrates. Nature 439:965–968.
Depew MJ, Lufkin T, Rubenstein JL. 2002. Specification of
jaw subdivisions by Dlx genes. Science 298:381–385.
Forey P, Janvier P. 1993. Agnathans and the origin of jawed
vertebrates. Nature 361:129–134.
Forey PL. 1995. Agnathans ercent and fossil, and the origin of
jawed vertebrates. Rev Fish Biol Fish 5:267–303.
Fraser EA. 1915. The head cavities and development of the eye
muscles in Trichosurus vulpecula, with notes on some other
marsupials. Proc Zool Soc Lond 22:299–346.
Freund R, Dörfler D, Popp W, Wachtler F. 1996. The
metameric pattern of the head mesoderm-does it exist?
Anat Embryol 193:73–80.
Furlong RF, Holland PW. 2002. Bayesian phylogenetic
analysis supports monophyly of ambulacraria and of
cyclostomes. Zool Sci 19:593–599.
Gans C, Northcutt RG. 1983. Neural crest and the orgin of
vertebrates: a new head. Science 220:268–274.
Gee H. 1996. Before the backbone: views on the origin of the
vertebrates. London: Chapman and Hall.
Gendron-Maguire M, Mallo M, Zhang M, Gridley T. 1993.
Hoxa-2 mutant mice exhibit homeotic transformation of
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
skeletal elements derived from cranial neural crest. Cell
75:1317–1331.
Gilbert PW. 1947. The origin and development of the head
cavities in the human embryo. J Morphol 90:149–188.
Gilbert PW. 1953. The premandibular head cavities of the
opossum, Didelphys virginiana. Anat Rec 115:392–393.
Gilbert PW. 1954. The premandibular head cavities in the
opossum, Didelphys virginiana. J Morphol 95:47–75.
Gilbert PW. 1957. The origin and development of the human
extrinsic ocular muscles. Cont Embryol 36:59–78.
Goethe JW. 1790. Das Schädelgrüt aus sechs Wirbelknochen
aufgebaut. Zur Naturwissenschaft überhaupt, besonders
zur Morphologie. II 2 (cited in Gaupp 1898).
Goodrich ES. 1910. On the segmentation of the occipital
region of the head in the Batrachia Urodela. Proc Zool Soc
Lond 1910:101–121.
Goodrich ES. 1930. Studies on the structure and development
of vertebrates. London: Macmillan Press.
Gorbman A. 1983. Early development of the hagfish pituitary
gland: Evidence for the endodermal origin of the adenohypophysis. Am Zool 23:639–654.
Gorbman A, Tamarin A. 1985. Early development of oral,
olfactory and adenohypophyseal structures of agnathans
and its evolutionary implications. In: Foreman RE, Gorbman A, Dodd JM, Olsson R, editors. Evolutionary Biology of
Primitive Fishes. New York: Plenum. 165–185 p.
Grammatopoulos GA, Bell E, Toole L, Lumsden A, Tucker AS.
2000. Homeotic transformation of branchial arch identity
after Hoxa2 overexpression. Development 127:5355–5365.
Haeckel E. 1874. Die Gastrea-Theorie, die phylogenetische
Klassification des Tierreiches und Homologie der Keimblätter. Jena Z Naturwiss 8:1–55.
Hall BK. 1998. Evolutionary developmental biology. 2nd ed.
London: Chapman & Hall.
Hall BK. 1999. The neural crest in developmental and
evolution. New York: Springer.
Hall BK, Hörstadius S. 1988. The neural crest. Oxford: Oxford
University Press.
Hanken J, Hall BK. (eds.) 1993. The skull vols. Chicago, USA:
University of Chicago Press. 1–3.
Hatschek B. 1881. Studien über die Entwicklung des Amphioxus. Arb Zool Inst Univ Wien 4:1–88.
Holland LZ, Kene M, Williams NA, Holland ND. 1997.
Sequence and embryonic expression of the amphioxus
engrailed gene (AmphiEn): the metameric pattern of
transcription resembles that of its segment-polarity homolog in Drosophila. Development 124:1723–1732.
Holland PW. 1988. Homeobox genes and the vertebrate head.
Development 103(suppl):17–24.
Holmgren N. 1946. On two embryos of Myxine glutinosa. Act
Zool 27:1–90.
Horigome N, Myojin M, Ueki T, Hirano S, Aizawa S, Kuratani
S. 1999. Development of cephalic neural crest cells in
embryos of Lampetra japonica, with special reference to
the evolution of the jaw. Dev Biol 207:287–308.
Hunt P, Gulisano M, Cook M, Sham MH, Faiella A, Wilkinson
D, Boncinelli E, Krumlauf R. 1991. A distinct Hox code for
the branchial region of the vertebrate head. Nature
353:861–864.
Hunter MP, Prince VE. 2002. Zebrafish Hox paralogue group
2 genes function redundantly as selector genes to pattern
the second pharyngeal arch. Dev Biol 247:367–389.
Jacobson AG. 1988. Somitomeres: mesodermal segments of
vertebrate embryos. Development 104(suppl):209–220.
19
Janvier P. 1996. Early vertebrates. Oxford: Clarendon Press.
Janvier P. 2001. Ostracoderms and the shaping of the
gnathostome characters. In: Ahlberg PE, editor. Major
events in early vertebrate evolution: paleontology. Phylogeny, genetics and development. London and New York:
Taylor & Francis, pp 172–186.
Jarvik E. 1980. Basic structure and evolution of veretebrates,
vol 2. New York: Academic press.
Jefferies RPS. 1986. The ancestry of the vertebrates. London:
The British Museum (Natural History).
Jeffery WR, Strickler AG, Yamamoto Y. 2004. Migratory
neural crest-like cells form body pigmentation in a urochordate embryo. Nature 431:696–699.
Johnels AG. 1948. On the development and morphology
of the skeleton of the head of Petromyzon. Acta Zool 29:
140–279.
Jouve C, Iimura T, Pourquie O. 2002. Onset of the segmentation clock in the chick embryo: evidence for oscillations in
the somite precursors in the primitive streak. Development
129:1107–1117.
Kimmel CB, Miller CT, Keynes RJ. 2001. Neural crest
patterning and the evolution of the jaw. J Anat 199:105–120.
Koentges G, Matsuoka T. 2002. Jaws of the fates. Science 298:
371–372.
Koltzoff NK. 1901. Entwicklungsgeschichte des Kopfes von
Petromyzon planeri. Bull Soc Nat Moscou 15:259–289.
Kupffer C von. 1899. Zur Kopfentwicklung von Bdellostoma.
Sitzungsber. Ges Morphol Physiol 15:21–35.
Kupffer C von. 1900. Studien zur vergleichenden Entwicklungsgeschichte des Kopfes der Kranioten. 4 Heft: Zur
Kopfentwicklung von Bdellostoma. München und Leipzig:
Verlag von J. F. Lehmann. 86 p.
Kuraku S, Kuratani S. 2006. Timescale for cyclostome
evolution inferred with a phylogenetic diagnosis of hagfish
and lamprey cDNA sequences. Zool Sci 23:1053–1064.
Kuratani S. 1997. Distribution of postotic crest cells in the
chick embryo defines the trunk/head interface: embryological interpretation of crest cell distribution and evolution of
the vertebrate head. Anat Embryol 195:1–13.
Kuratani S. 2003. Evolutionary developmental biology and
vertebrate head segmentation: a perspective from developmental constraint. Theory Biosci 122:230–251.
Kuratani S. 2004a. Evolution of the vertebrate jaw: comparative embryology reveals the developmental factors behind
the evolutionary novelty. J Anat 205:335–347.
Kuratani S. 2004b. Evolutionary morphology: bauplan and
embryonic development of vertebrates. Tokyo: Tokyo University Press (in Japanese).
Kuratani S. 2005a. Craniofacial development and the evolution of the vertebrates: the old problems on a new
background. Zool Sci 22:1–19.
Kuratani S. 2005b. Developmental studies of the lamprey and
hierarchical evolutionary steps towards the acquisition of
the jaw. J Anat 207:489–499.
Kuratani S, Matsuo I, Aizawa S. 1997a. Developmental
patterning and evolution of the mammalian viscerocranium:
Genetic insights into comparative morphology. Dev Dyn
209:139–155.
Kuratani S, Ueki T, Aizawa S, Hirano S. 1997b. Peripheral
development of cranial nerves in a cyclostome, Lampetra
japonica: morphological distribution of nerve branches and
the vertebrate body plan. J Comp Neurol 384:483–500.
Kuratani S, Horigome N, Hirano S. 1999. Developmental
morphology of the head mesoderm and reevaluation of
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
20
S. KURATANI AND K.G. OTA
segmental theories of the vertebrate head: evidence from
embryos of an agnathan vertebrate, Lampetra japonica. Dev
Biol 210:381–400.
Kuratani S, Nobusada Y, Saito H, Shigetani, Y. 2000.
Morphological development of the cranial nerves and
mesodermal head cavities in sturgeon embryos from early
pharyngula to mid-larval stages. Zool Sci 17:911–933.
Kuratani S, Nobusada Y, Horigome N, Shigetani Y. 2001.
Embryology of the lamprey and evolution of the vertebrate jaw: insights from molecular and developmental
perspectives. Phil Trans R Soc Lond B Biol Sci 356:
1615–1632.
Kuratani S, Murakami Y, Nobusada Y, Kusakabe R, Hirano S.
2004. Developmental fate of the mandibular mesoderm in
the lamprey, Lethenteron japonicum: comparative morphology and development of the gnathostome jaw with special
reference to the nature of the trabecula cranii. J Exp Zool B
(Mol Dev Evol) 302:458–468.
Kuratani SC, Eichele G. 1993. Rhombomere transplantation
repatterns the segmental organization of cranial nerves and
reveals autonomous expression of a homeodomain protein.
Development 117:105–117.
Le Douarin N. 1982. The neural crest. Cambridge: Cambridge
University Press.
Le Douarin N, Kalcheim C. 1999. The neural crest. Cambridge: Cambridge University Press.
Lumsden A. 2004. Segmentation and compartition in the early
avian hindbrain. Mech Dev 121:1081–1088.
Mallatt J, Sullivan J. 1998. 28S and 18S rDNA sequences
support the monophyly of lampreys and hagfishes. Mol Biol
Evol 15:1706–1718.
Marinelli W, Strenger A. 1954. Vergleichende Anatomie und
Morphologie der Wirbeltiere. Wien: Franz Deuticke.
Maroto M, Pourquie O. 2001. A molecular clock involved in
somite segmentation. Curr Top Dev Biol 51:221–248.
Matsuoka T, Ahlberg PE, Kessaris N, Iannarelli P, Dennehy
U, Richardson W, McMahon AP, Koentges G. 2005. Neural
crest origins of the neck and shoulder. Nature 436:
347–355.
Maynard-Smith J, Burian R, Kauffman S, Alberch P, Campbell J, Goodwin B, Lande R, Raup D, Wolpert L. 1985.
Developmental constraints and evolution. Q Rev Biol 60:
265–287.
McCauley D, Bronner-Fraser M. 2003. Neural crest contributions to the lamprey head. Development 130:2317–2327.
McGinnis W, Krumlauf R. 1992. Homeobox genes and axial
patterning. Cell 68:283–302.
Meier S. 1979. Development of the chick mesoblast. Formation
of the embryonic axis and establishment of the metameric
pattern. Dev Biol 73:25–45.
Meulemeans D, Bronner-Fraser M. 2002. Amphioxus and
lamprey AP-2 genes: implications for neural crest evolution
and migration patterns. Development 129:4953–4962.
Narita Y, Kuratani S. 2005. Evolution of the vertebral
formulae in mammals: a perspective on developmental
constraints. J Exp Zool B Mol Dev Evol 304:91–106.
Neal HV. 1896. A summary of studies on the segmentation of
the nervous system in Squalus acanthias. Anat Anz 12:
377–391.
Neal HV. 1918a. History of eye muscles. J Morphol 30:
433–453.
Neal HV. 1918b. Neuromeres and metameres. J Morphol 31:
293–315.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b
Neidert AH, Virupannavar V, Hooker GW, Langeland JA.
2001. Lamprey Dlx genes and early vertebrate evolution.
Proc Natl Acad Sci USA 98:1665–1670.
Neumayer L. 1938. Die Entwicklung des Kopfskelettes von
Bdellostoma. St. L. Arch Ital Anat Embryol 40(suppl) (cited
by Holmgren, 1946).
Noden DM. 1988. Interactions and fates of avian craniofacial
mesenchyme. Development 103(suppl):121–140.
Oken L. 1807. Über die Bedeutung der Schädelknochen.
Bamberg: Göbhardt.
Ota KG, Kuratani S. 2007. Cyclostome embryology and early
evolutionary history of vertebrates. Integ Comp Biol (in press).
Ota KG, Kuraku S, Kuratani S. 2007. Hagfish embryolgy with
reference to the evolution of the neural crest. Nature
446:672–675.
Pancer Z, Amemiya CT, Ehrhardt GR, Ceitlin J, Gartland GL,
Cooper MD. 2004. Somatic diversification of variable
lymphocyte receptors in the agnathan sea lamprey. Nature
430:174–180.
Pancer Z, Saha NR, Kasamatsu J, Suzuki T, Amemiya CT,
Kasahara M, Cooper MD. 2005. Variable lymphocyte receptors in hagfish. Proc Natl Acad Sci USA 102:9224–9229.
Pasqualetti M, Ori M, Nardi I, Rijli FM. 2000. Ectopic Hoxa2
induction after neural crest migration results in homeosis of
jaw elements in Xenopus. Development 127:5367–5378.
Platt JB. 1893. Ectodermic origin of the cartilage of the head.
Anat Anz 8:506–509.
Portmann A. 1976. Einführung in die vergleichende Morphologie der Wirbeltiere. Basel: Schwabe & Co.
Riedl R. 1978. Order in living organisms. New York: Wiley.
Rijli FM, Mark M, Lakkaraju S, Dierich A, Dolle P,
Chambon P. 1993. A homeotic transformation is
generated in the rostral branchial region of the head by
disruption of Hoxa-2, which acts as a selector gene. Cell 75:
1333–1349.
Schlosser G. 2005. Evolutionary origins of vertebrate placodes:
insights from developmental studies and from comparisons
with other deuterostomes. J Exp Zool B (Mol Dev Evol) 304:
347–399.
Schlosser G. 2006. Induction and specification of cranial
placodes. Dev Biol 294:303–351.
Sewertzoff AN. 1911. Die Kiemenbogennerven der Fische.
Anat Anz 38:487–495.
Sewertzoff AN. 1916. Etudes sur l’évolution des vertebrés
inférieurs. I. Morphologie du squelette et de la musculature
de la tête des cyclostomes. Arch russes d’Anat d’Hist et
d’Embryol I (cited by Holmgren, 1946).
Shigetani Y, Nobusada Y, Kuratani S. 2000. Ectodermallyderived FGF8 defines the maxillomandibular region in the
early chick embryo: epithelial–mesenchymal interactions in
the specification of the craniofacial ectomesenchyme. Dev
Biol 228:73–85.
Shigetani Y, Sugahara F, Kawakami Y, Murakami Y, Hirano
S, Kuratani S. 2002. Heterotopic shift of epithelialmesenchymal interactions for vertebrate jaw evolution.
Science 296:1316–1319.
Slack JM, Holland PW, Graham CF. 1993. The zootype and
the phylotypic stage. Nature 361:490–492.
Stock DW, Whitt GS. 1992. Evidence from 18S ribosomal RNA
sequences that lampreys and hagfishes form a natural
group. Science 257:787–789.
Takezaki N, Figueroa F, Zaleska-Rutczynska Z, Klein J. 2003.
Molecular phylogeny of early vertebrates: monophyly of the
CYCLOSTOME EMBRYOLOGY AND THE VERTEBRATE HEAD
agnathans as revealed by sequences of 35 genes. Mol Biol
Evol 20:287–292.
Takio Y, Pasqualetti M, Kuraku S, Hirano S, Rijli FM,
Kuratani S. 2004. Lamprey Hox genes and the evolution of
jaws. Nat OnLine 429:1 p following 262. http://www.nature.
com/cgi-taf/DynaPage.taf?file 5 /nature/journal/v429/n6989/
full/nature02616_fs.html.
Takio Y, Kuraku S, Kusakabe R, Murakami Y, Pasqualetti M,
Rijli FM, Narita Y, Kuratani S, Kusakabe R. 2007. Hox gene
expression patterns in Lethenteron japonicum embryos
insights into the evolution of the vertebrate Hox code. Dev
Biol. in press.
Uchida K, Murakami Y, Kuraku S, Hirano S, Kuratani S.
2003. Development of the adenohypophysis in the
lamprey: evolution of the epigenetic patterning programs
in organogenesis. J Exp Zool (Mol Dev Evol) 300B:
32–47.
Veit O. 1939. Beiträge zur Kenntnis des Kopfes der
Wirbeltiere. III: Beobachtungen zur Frühentwicklung
21
des Kopfes von Petromyzon planeri. Morphol Jb 84:
86–107.
Wedin B. 1949. The development of the head cavities in
Alligator mississippiensis Daud. Lunds Univ Arssikr NF
avs 2 45:1–32.
Wedin B. 1953a. The development of the head cavities in
Ardea cinerea L. Act Anat 17:240–252.
Wedin B. 1953b. The development of the eye muscles in Ardea
cinerea L. Act Anat 18:38–48.
Wicht H, Northcutt RG. 1995. Ontogeny of the head of the
Pacific hagfish (Eptatretus stouti, Myxinoidea): development
of the lateral line system. Phil Trans R Soc Lond B 349:
119–134.
van Wijhe JW. 1882. Über die Mesodermsegmente und die
Entwicklung der Nerven des Selachierkopfes. Groningen:
Ver Akad Wiss Amsterdam. p 1–50.
van Wijhe JW. 1883. Zool Anz 6:657 (cited in Wedin, ’53).
Yalden DW. 1985. Feeding mechanisms as evidence for
cyclosome monophyly. Zool J Linn Soc 84:291–300.
J. Exp. Zool. (Mol. Dev. Evol.) DOI 10.1002/jez.b