Current Biology, Vol. 13, 1571–1582, September 16, 2003, 2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/j.cub.2003.08.030
Eph/Ephrin Signaling Regulates the Mesenchymal-
to-Epithelial Transition of the Paraxial Mesoderm
during Somite Morphogenesis
mation of these segments into epithelial somites sepa-
rated by intersomitic furrows. Major progress has been
the establishment of a segmental pattern prior to furrow
formation (reviewed in [1, 2]). Few studies, however,
have addressed the molecular mechanisms that drive
the morphogenetic processes occurring during somite
Prior to somite formation, cells within each segment
of the rostral presomitic mesoderm (PSM) acquire either
anterior or posterior character. Anterior-posterior polar-
ity within the presumptive somite is evident as segmen-
tal patterns of gene expression regulated by periodic
activation of Notch signaling [3, 4]. Boundaries of gene
expression between cells with anterior identity and cells
with posterior identity are subsequently translated into
morphological furrows between somitic cells that un-
dergo the mesenchymal-to-epithelial transition. Genes
segmentally expressed in the rostral PSM are therefore
cesses leading to somite formation.
The Eph family of receptor tyrosine kinases and their
Ephrin ligands show Notch signaling-dependent seg-
mental expression in the rostral PSM [5, 6]. Eph proteins
are membrane-spanning receptor proteins involved in
intercellular signaling in many morphogenetic processes
during embryonic development (reviewed in [7–9]).
Ephrin ligands for these receptors are classified into
two groups according to their association to the cell
membrane. Class A Ephrins are tethered to the mem-
brane by a GPI linkage and preferentially bind EphA
bind EphB receptors) are transmembrane ligands with
an intracellular domain. EphA4 is the only receptor that
binds both classes of ligands . A notable feature of
this pathway is that bidirectional signaling cascades are
triggered upon Eph/Ephrin interaction (reviewed in ).
signaling, whereas signaling downstream of Ephrins is
called reverse signaling. Molecules that directly or indi-
tors are in most cases regulators of the cytoskeleton
and cell adhesion (reviewed in ) as well as other
transmembrane receptors [13, 14]. One frequent conse-
populations , and roles in cell sorting and boundary
formation in the zebrafish hindbrain have been demon-
strated [16,17]. Ephreceptors andEphrins aretherefore
excellent candidates for driving the morphogenetic
events associated with somite epithelialization.
We have previously shown that disruption to Eph/
Ephrin signaling results in a lack or aberrant formation
of intersomitic furrows . In this paper, we have used
the mutant fused somites (fss) as an in vivo system to
study the role of Eph/Ephrin signaling during somite
morphogenesis. fss encodes Tbx24, a T box transcrip-
tion factor involved in maturation of the PSM . fss
Arantza Barrios,* Richard J. Poole, Lindsey Durbin,
Caroline Brennan,1Nigel Holder,2
and Stephen W. Wilson*
Department of Anatomy and Developmental Biology
University College London
London WC1E 6BT
Background: During somitogenesis, segmental pat-
terns of gene activity provide the instructions by which
mesenchymal cells epithelialize and form somites. Vari-
ous members of the Eph family of transmembrane re-
ceptor tyrosine kinases and their Ephrin ligands are ex-
pressed in a segmental pattern in the rostral presomitic
mesoderm. This pattern establishes a receptor/ligand
interface at each site of somite furrow formation. In the
fused somites (fss/tbx24) mutant, lack of intersomitic
boundaries and epithelial somites is accompanied by a
lack of Eph receptor/Ephrin signaling interfaces. These
observations suggest a role for Eph/Ephrin signaling in
the regulation of somite epithelialization.
in the paraxial mesoderm of fss mutants rescues most
aspects of somite morphogenesis. First, restoration of
distinct boundaries. Second, activation of EphA4 leads to
the cell-autonomous acquisition of a columnar morphol-
ogy and apical redistribution of ?-catenin, aspects of
epithelialization characteristic of cells at somite bound-
aries. Third, activation of EphA4 leads to nonautonomous
acquisition of columnar morphology and polarized relo-
calization of the centrosome and nucleus in cells on the
omous aspects of epithelialization may involve interplay
of EphA4 with other intercellular signaling molecules.
Conclusions: Our results demonstrate that Eph/Ephrin
signaling is an important component of the molecular
mechanisms driving somite morphogenesis. We pro-
pose a new role for Eph receptors and Ephrins as inter-
cellular signaling molecules that establish cell polarity
during mesenchymal-to-epithelial transition of the par-
Somites are the most obvious manifestation of verte-
brate embryonic metamerism. Somitogenesis involves
the specification of groups of cells within the paraxial
mesoderm as segments and the subsequent transfor-
*Correspondence: email@example.com (A.B.), firstname.lastname@example.org
1Present address: Queen Mary College, University of London, Lon-
don E1 4NS, United Kingdom.
2This paper is dedicated to the memory of Nigel Holder.
Figure 1. Cells Undergo Mesenchymal-to-Epithelial Transition at Somite Boundaries
(A–L) Dorsal views of the left-sided paraxial mesoderm of embryos labeled with Bodipy ceramide (which reveals cell morphology; [A], [E], and
[I]) or with Bodipy 505-515 (which reveals nuclear position, [C], [G], and [K]) or immunostained for ?-catenin ([B], [F], and [J]) or for ?-tubulin
(which labels centrosomes) and stained with phalloidin (which labels actin) ([D], [H], and [L]). Anterior is oriented toward the top. (A–D) Cells
at somite boundaries in wild-type embryos. The arrowheads point to the intersomitic boundary. (E–H) Cells in the presomitic mesoderm (PSM)
of wild-type embryos. The arrows point to epithelial adaxial cells in which centrosomes are apically localized (H), as also seen in epithelial
cells at somite boundaries (D). Centrosomes are randomly positioned in other PSM cells. (I–L) Cells in the somitic mesoderm of fss?/?embryos.
n, notochord; ac, adaxial cells. The scale bars represent 10 ?m.
[6, 20]. We show that in fss?/?embryos, disruption of
the Eph/Ephrin signaling interface is accompanied by a
failure of paraxial mesoderm cells to undergo the mes-
mation. Using a genetic mosaic approach, we demon-
strate that restoration of unidirectional or bidirectional
Eph/Ephrin signaling in the paraxial mesoderm of fss?/?
ically distinct boundaries. Furthermore, many aspects
aries induced in fss?/?embryos. For instance, activation
calization of ?-catenin. Conversely, apical localization
of the centrosome and basally directed relocalization of
the nucleus is a cell-nonautonomous consequence of
EphA4 activation. Nuclear relocalization is dependent
ing; these observations suggest the involvement of a
parallel pathway activated by EphA4 signaling. Alto-
gether, these results reveal a pivotal role for Eph recep-
tors and Ephrins as effectors of somite morphogenesis.
cell movement during this process in fish , (see
Movie 1 in the Supplemental Data available with this
article online). However, the boundary cells do undergo
changes associated with transformation from a mesen-
chymal to an epithelial morphology. Epithelial morphol-
ogy is revealed by the acquisition of a columnar shape
(Figure 1A), accumulation of molecules associated with
adhesion complexes, such as ?-catenin, at the apical
pole of the cells (Figure 1B), basally directed relocaliza-
tion of cell nuclei toward the somite boundary (Figure
1C), and apical relocalization of centrosomes (Figure
1D). These changes are initiated contemporaneously
with, and not prior to, intersomitic boundary formation
(see Figure S1 in the Supplemental Data). Unlike cells
at the boundaries, those within the core of the somite
remain mesenchymal like cells of the PSM.
undergo some reorganization but somite boundaries do
not form. Cells fail to epithelialize at segment borders
and remain mesenchymal (Figure 1I), ?-catenin appears
localized homogenously throughout the cell membrane
1K), and centrosomes are distributed randomly within
the cytoplasm (Figure 1L). The morphology of the cells
in the somitic mesoderm of fss mutants resembles the
mesenchymal morphology of cells in the core of the
PSM of wild-type embryos (Figures 1E–1H). The PSM
of wild-type zebrafish embryos also contains cells with
epithelial morphology at sites where the paraxial meso-
derm borders with the notochord, neural and surface
ectoderm, and lateral plate (Figure S1). In fss?/?em-
bryos, equivalent cells also display epithelial morphol-
Cells of the Paraxial Mesoderm Fail to Undergo
in fss Mutant Embryos
During somite formation, PSM cells positioned at either
side of the prospective intersomitic boundary align to
form palisade-like structures along the intersomitic fur-
row (Figure 1A). Unlike in chick , there is virtually no
Eph Signaling Regulates Somite Epithelialization
Figure 2. Expression of Eph Family Members in the Paraxial Mesoderm of Wild-Type and fss?/?Embryos
Dorsal views of the paraxial mesoderm of 8-somite-stage wild-type and 10-somite-stage fss?/?embryos and schematics with anterior oriented
toward the top. The arrowheads indicate the position of the most recently formed intersomitic boundary.
(A and B) Living wild-type and fss?/?embryos labeled with Bodipy ceramide. In the wild-type embryo, the positions of the last two somites
formed (SII, SI), the forming somite (S0), and the two presumptive somites in the PSM (S-I, S-II) are indicated. The arrowhead points to the
(C–J) Expression of (C and D) ephA4, (E and F) ephrin-B2b, (G and H) ephrin-B2a, and (I and J) ephrin-A1 in wild-type (top row) and fss mutant
(bottom row) embryos in the region of the paraxial mesoderm shown in (A) and (B).
(C?–J?, K, and L) Schematics summarizing expression of Eph family members in the paraxial mesoderm of wild-type and fss?/?embryos.
nt, neural tube; n, notochord. The scale bars represent 30 ?m.
ogy (data not shown and ), indicating that the intra-
cellular machinery required for cellular epithelialization
is functional in fss?/?cells.
pression of ephrin-B2a is detected throughout the so-
mite-II/somite II region (Figure 2H). Segmental expres-
sion of ephrin-A1 is also lost in fss?/?embryos, and, in
the somite-II/somite II region, transcripts are only de-
tected in the medial region of the paraxial mesoderm
(Figure 2J). Analysis of the expression patterns of these
Eph family members demonstrates the absence of a
ligand-receptor interface in the PSM of fss mutants; this
lack of interface suggests that Eph/Ephrin signaling is
disrupted in the region where somite boundaries should
be forming (Figure 2L).
Eph/Ephrin Signaling Is Disrupted in the Paraxial
Mesoderm of fss?/?Embryos
In the rostral PSM of wild-type embryos, somite bound-
ary formation is preceded by the segmental expression
of EphA4, two Ephrins (Ephrin-B2a and Ephrin-A1) that
bind this receptor , and Ephrin-B2b, which is also
likely to bind EphA4 . ephA4 and ephrin-B2b are
expressed in two or three stripes in the PSM  such
that by the stage that cells are in somite 0 (the somite
being formed), expression is restricted to one or two
rows of cells in the most anterior region of the segment
adjacent to the forming somite boundary (Figures 2C
sion within presumptive somites, and the highest expres-
sion is in posterior cells adjacent to the forming boundary
(Figures 2G and 2I). Each new intersomitic furrow, there-
fore, forms at the interface between posterior cells in
somite 0, expressing high levels of ephrin-B2a and
levels of ephA4 and ephrin-B2b (Figure 2K).
Segmental expression of these Eph family members
is abolished in fss mutants. ephA4 and ephrin-B2b ex-
pression in cells with anterior identity is absent in the
paraxial mesoderm (Figures 2D and 2F), whereas ex-
Restoration of the Eph/Ephrin Signaling Interface
Rescues the Formation of Morphologically
Distinct Boundaries in the Paraxial
Mesoderm of fss?/?Embryos
To test the hypothesis that loss of ephA4 expression in
in fss?/?embryos, we designed a series of experiments
to restore the Eph/Ephrin interface in the paraxial meso-
derm of fss?/?mutants. Wild-type donor cells express-
ing an EphA4-eGFP fusion protein  were transplanted
to the prospective paraxial mesoderm of fss?/?host
embryos. When clusters of wild-type cells expressing
exogenous EphA4 were present in the paraxial meso-
derm of fss?/?embryos, ectopic morphologically dis-
tinct boundaries were visible at the interface between
EphA4-expressing donor cells and Ephrin-expressing
Figure 3. Eph/Ephrin Signaling Restores Morphologically Distinct Boundaries in fss?/?Embryos
(A–E and A?–E?) (A–E) DIC and fluorescence overlays and (A?–E?) DIC images of the paraxial mesoderm of fss?/?hosts into which wild-type
(wt) or fss?/?cells expressing various GFP-tagged reagents (green labeling in [A], [B], [D], and [E]) or containing rhodamine dextran (RD, red
labeling in [C]) have been transplanted. Reagents are indicated at the bottom of the panels. (C?) ephA4 expression (blue) is absent from the
transplanted cells. The arrowheads point to morphologically distinct boundaries formed at the interface between donor and host cells. nt,
fss?/?host cells in 84% (n ? 57) of the cases (Figure
3A, Table 1). As irregular and weak boundaries do occa-
fss?/?embryos, we assayed the frequency with which
aries. When fss?/?cells were transplanted to the pro-
spective paraxial mesoderm of fss?/?host embryos, we
found only 14% of the clones aligned with an endoge-
nous boundary (Table 1).
In order to address the possibility that factors other
than exogenous EphA4 mediate the rescue of bound-
aries when wild-type cells are transplanted to fss?/?
hosts, wild-type donor cells expressing only GFP were
transplanted into fss?/?hosts. Wild-type cells did not
restore boundary formation, and clones coincided with
the rare endogenous boundaries in the somitic meso-
derm of fss?/?hosts at a frequency no greater than
we had observed for control transplants of fss?/?cells
(Figure 3B, Table 1). This was surprising since we ex-
pected that cell-autonomous activity of wild-type Fss
might promote the endogenous expression of anterior
markers including ephA4 in wild-type cells transplanted
to fss?/?hosts. However, although wild-type cells do
cell-autonomously express various anterior segmental
markers when transplanted into fss?/?hosts (see Figure
S2 in the Supplemental Data), there is no detectable
ephA4 expression (Figure 3C). Together, these results
indicate that wild-type cells are neither able to rescue
an Eph/Ephrin signaling interface nor formation of mor-
phologically distinct boundaries when transplanted into
Fss regulates the expression of many genes in addi-
tion to ephA4, and so we tested whether exogenous
EphA4 requires the activity of other Fss-dependent pro-
teins to rescue the formation of morphologically distinct
boundaries. When clusters of EphA4-expressing fss?/?
cells were present in the somitic mesoderm of fss?/?
hosts, ectopic, morphologically distinct boundaries
formed at the interface between donor and host cell
populations in 87% (n ? 23) of the cases (Figure 3D,
be induced in fss?/?embryos solely by the restoration
of an Eph/Ephrin signaling interface.
Our favored interpretation of these results is that Eph/
Ephrin signaling functions as one of the final steps in
boundary formation. An alternative possibility is that
genes that function in the anterior region of the forming
somite and that it is these other factors that mediate
boundary formation. However, this is unlikely, as wild-
type cells with anterior character (but lacking ephA4
expression) do not induce boundaries (Figure 3B), ex-
pression ofseveral anterior markers appearsto be unaf-
fected by EphA4 activity, and exogenous EphA4 can
still promote boundary formation when Notch signaling
is disrupted (Figure S2). These observations suggest
that Eph/Ephrin signaling directly mediates boundary
formation downstream of the acquisition of “anterior”
identity and does not induce boundaries by changing
the fate of presomitic cells.
Table 1. Rescue of Boundary Formation in Mosaic Experiments
fss → fss
wt → fss
wt ? EphA4 → fss
fss ? EphA4 → fss
wt ? dnEphA4 → fss
wt ? EphA4 → fss ? dnEphrin-B2a
n is the total number of fss host embryos analyzed. Data are pooled
from several independent experiments. wt, wild-type; fss, fused
Eph Signaling Regulates Somite Epithelialization
Transplanting EphA4-expressing cells into fss?/?em-
bryos is likely to activate signaling downstream of both
receptor-expressing donor cells and Ephrin-expressing
host cells. To test whether signaling downstream of
Ephrins is sufficient to restore the formation of morpho-
logically distinct boundaries, we activated Ephrins with
that lacks the intracellular tyrosine kinase domain (see
the Experimental Procedures). Truncated membrane
bound Eph receptors are able to bind and promote clus-
tering andactivation of their counterpartligands in adja-
cent cells [23, 24]; however, the lack of a tyrosine kinase
domain renders the receptor incapable of transducing
an intracellular signal. When wild-type cells expressing
truncated EphA4 receptor were transplanted into the
somitic mesoderm of fss?/?embryos, ectopic bound-
aries were rescued at the interface between donor and
3E, Table 1). This result suggests that activation of
Ephrin reverse signaling in cells on one side of the na-
distinct boundaries in the paraxial mesoderm of fss?/?
Boundaries Established by the Restoration of EphA4/
Ephrin Signaling in fss Embryos Mature Normally
during Muscle Differentiation
A major consequence of somite boundary formation is
the alignment of muscle fiber attachment sites at the
intersomitic boundary (Figure 4A). To assess if morpho-
of Eph/Ephrin signaling in fss?/?mutants resemble wild-
nization in fss?/?embryos with and without transplants
of EphA4-expressing cells. In fss?/?embryos, there is
no precise alignment of muscle fibers. Even when rare,
irregular, and aberrantly shaped endogenous bound-
aries are present, muscle fibers still frequently cross the
boundaries (Figure 4B).
The morphologically distinct boundaries that form in
fss?/?embryos following unidirectional or bidirectional
are capable of organizing muscle fibers. Both host and
donor muscle fiber attachment sites are aligned at the
induced boundaries, and muscle fibers do not extend
across the boundary into adjacent cell populations.
Muscle fibers in EphA4-expressing clusters of wild-type
cells (Figure 4C) appear more compact and better
aligned than those in EphA4-expressing fss?/?cell clus-
ters (Figure 4D) or in truncated EphA4-expressing wild-
type cell clusters (Figure 4E). This suggests that both
signaling downstream of the receptor and other Fss-
dependent factors contribute to the proper morphogen-
esis and differentiation of the clones (see below and the
Figure 4. Boundaries Restored by Eph/Ephrin Signaling Are Main-
tained during Muscle Differentiation
(A–E) Lateral views of (A) wild-type and (B–E) fss?/?somitic muscles
immunostained for myosin (green). Anterior is oriented toward the
left. (C), (D), and (E) show transplanted (C and E) wild-type or (D)
fss?/?cells (red) expressing (C and D) full-length or (E) truncated,
distinct furrows. The arrows point at fss?/?muscle fibers that span
Apical Distribution of ?-Catenin and Acquisition
of Columnar Morphology Are Downstream
Consequences of EphA4 Signaling
In addition to the formation of a morphologically distinct
boundary, somite morphogenesis leads to epithelializa-
in fss?/?mutants (Figure 1). To investigate whether Eph/
Ephrin signaling is involved in mediating this mesenchy-
mal-to-epithelial transition, we analyzed whether cells
along boundaries induced by Eph/Ephrin signaling in
fss?/?embryos became epithelialized.
In 70% (n ? 20) of the cases in which morphologically
Figure 5. Eph/Ephrin Signaling Rescues Epithelialization of Cells at Morphologically Distinct Boundaries
(A–L) (A, E, and I) Confocal images showing Bodipy ceramide-labeled somitic mesoderm of fss?/?host embryos (green) containing rhodamine
dextran-labeled donor cells (red). (B, F, and J) Confocal images of ?-catenin immunolocalization (green) in transplanted rhodamine dextran-
distinct boundaries form (visible with DIC optics, not shown). In (B) and (J), ?-catenin is reduced on the basal surfaces of these cells. (C, G,
and K) Confocal images showing Bodipy 505-515-labeled somitic mesoderm of fss?/?host embryos (green) containing rhodamine dextran-
labeled donor cells (red). The white arrowhead points to nuclei localized at the basal pole of host fss?/?cells, adjacent to the boundaries
created between donor and host cells. (D, H, and L) Confocal images showing phalloidin-labeled somitic mesoderm of fss?/?host embryos
(red) containing CFP-labeled donor cells (blue) and immunostained for ?-tubulin (green). The white arrows point to centrosomes. These are
localized at the apical pole of host fss?/?cells adjacent to the boundary created between donor and host cells in (D) but are randomly
positioned in (H) and (L). Donor cells are wild-type cells expressing (A–D) full-length or (E–H) truncated, dominant-negative EphA4 or fss?/?
cells expressing (I–L) full-length EphA4. n, notochord; ac, adaxial cells. The scale bars represent 10 ?m.
distinct boundaries were established at the interface
between ephA4-expressing, wild-type donor cells and
ephrin-expressing fss?/?host cells, the donor cells at
Procedures) morphology (Figure 5A, Table 2). In 80%
(n ? 15) of the cases, donor cells were judged to show
increased levels of ?-catenin at the apical pole of the
cell and reduced levels basally (Figure 5B, Table 2). To
investigate if this epithelialization is dependent upon
EphA4 interacting with other Fss-dependent factors, we
analyzed whether epithelialization of fss?/?cells occurs
following activation of EphA4. In 70% (n ? 19) of the
to Ephrin-expressing fss?/?host cells (Figure 5J, Table
2). In seven of nine cases, donor fss?/?cells acquired
a columnar morphology (Figure 5I, Table 2).
Although these results suggest that EphA4 signaling
Table 2. Rescue of Epithelialization in Mosaic Experiments
Donor Cells Host Cells
Columnar Morphology ?
Wt ? EphA4 → fss
Wt ? dnEphA4 → fss
Fss ? EphA4 → fss
n is the total number of fss host embryos analyzed. Data are pooled from several independent experiments. wt, wild-type; fss, fused somites.
80% (n ? 15)
7% (n ? 14)
70% (n ? 19)
70% (n ? 20)
8% (n ? 12)
77% (n ? 9)
52% (n ? 49)
7% (n ? 29)
13% (n ? 23)
50% (n ? 14)
0% (n ? 6)
11% (n ? 9)
Eph Signaling Regulates Somite Epithelialization
mediates epithelialization, an alternative possibility is
that Eph signaling establishes a boundary and that as
a consequence of boundary formation, other EphA4-
independent events lead to epithelialization. To address
if this may be the case, we transplanted wild-type cells
boundaries but cannot transduce intracellular signaling
downstream of EphA4. Although truncated EphA4 in-
duces boundary formation, cells do not adopt a colum-
nar morphology and ?-catenin remains throughout the
membrane, including at the basal pole adjacent to the
boundary (Figures 5E and 5F, Table 2). These results
indicate that cell shape changes and ?-catenin relocal-
ization are dependent on cell-autonomous activation of
EphA4 signaling and are not secondary consequences
of boundary formation.
Next, we attempted to more directly address whether
reverse signaling downstream of Ephrins is required for
epithelialization. To do this, we overexpressed an intra-
throughout the somitic mesoderm of fss?/?mutant hosts.
Truncated Ephrin-B2a is able to bind and activate EphA4
likely to suppress endogenous Ephrin-B reverse signaling
[17, 18] (and see the Experimental Procedures). When
clones of wild-type donor cells expressing EphA4 were
fss?/?hosts, the acquisition of columnar morphology
and nuclear relocalization still occurred in host cells
in 57% of the cases in which morphologically distinct
boundaries formed (n ? 22; see Figure S3A in the Sup-
plemental Data). This frequency is not significantly dif-
ferent from that seen in experiments in which reverse
Ephrin signaling was unperturbed. Together, these re-
sults demonstrate that elongation and polarization of
fss?/?host cells are cell-nonautonomous effects down-
stream of EphA4 signaling for which Ephrin reverse sig-
naling may not be essential.
pathway that becomes activated by EphA4 signaling
in the receptor-bearing cells and that signals back to
adjacent fss?/?cells to promote cell elongation and nu-
clear migration. To investigate whether this parallel
pathway is fully functional in fss?/?mutants, we trans-
planted EphA4-expressing fss?/?cells into fss?/?hosts
fss?/?host cells undergo elongation and relocalization
of cellular organelles in response to EphA4-expressing
seen when the donor cells lacked Fss activity (Figures
5I–5L, Table 2). These results indicate that the factor/s
is not present or is functionally compromised in fss?/?
Epithelialization of Host fss?/?Cells Is a Cell-
Nonautonomous Effect of EphA4 Signaling
Dependent on Fss
When wild-type cells expressing exogenous EphA4
were present in the somitic mesoderm of fss?/?hosts,
tion. In 52% of these cell groups (44% of the total num-
ber of clones), the nuclei of Ephrin-expressing fss?/?
host cells became localized at the basal pole, toward
the boundary (Figure 5C, Table 2). Basal nuclear relocal-
ization in fss?/?host cells was accompanied by the ac-
quisition of a columnar morphology and apical relocal-
ization of the centrosome (Figures 5A and 5D, Table 2).
Epithelialization of host fss cells is, therefore, a cell-
nonautonomous effect of EphA4 activity. Despite acquir-
ing these features of epithelialization, Ephrin-expressing
host cells showed no evidence of apical ?-catenin relo-
calization (not shown). Similarly, relocalization of the
nucleus and the centrosome was not observed in donor
EphA4-expressing cells (not shown). Together, these
results indicate that while epithelial morphology is res-
cued on both sides of the Eph/Ephrin-induced bound-
aries, more subtle aspects of cell polarity may remain
Given that Ephrins can signal intracellularly, a cell-
nonautonomous effect of the receptor suggests a role for
Ephrin reverse signaling in epithelialization. We therefore
performed several sets of experiments to elucidate the
requirement for Ephrin signaling in cell elongation and
cell polarity in fss?/?host cells at boundaries induced
by transplantation of truncated EphA4-expressing wild-
type donor cells. Truncated EphA4 is able to activate
Ephrin reverse signaling in adjacent cells but is unable
to signal intracellularly. Although 93% (n ? 29) of the
clones formed morphologically distinct boundaries, no
obvious rescue of cell elongation, basal nuclear relocal-
ization, or apical relocalization of the centrosome was
observed in the adjacent Ephrin-expressing fss?/?host
cells (Figures 5E, 5G, and 5H, Table 2). This result indi-
cates that the reverse signaling induced by truncated
EphA4 is not sufficient to promote cell epithelialization.
As full-length, but not truncated, EphA4 can nonautono-
mously restore epithelial morphology in adjacent cells,
this result also implies that intracellular signaling down-
stream of EphA4 is important in this event.
During somite morphogenesis, a furrow of de-adhesion
creates a boundary between the populations of paraxial
mesodermal cells that will form adjacent somites. Cells
on both sides of the forming somite boundary undergo
a mesenchymal-to-epithelial transition that involves
changes in cell shape, cell adhesive interactions, and
we have presented several lines of evidence that Eph/
Ephrin signaling has key roles in boundary formation
and somite morphogenesis. First, in embryos that lack
somites, restoration of Eph/Ephrin signaling interfaces
rescues the formation and subsequent maturation of
morphologically distinct boundaries. Second, activation
of EphA4 signaling both cell-autonomously and -nonau-
tonomously rescues various aspects of somite bound-
ary cell epithelialization.
An Eph/Ephrin Interface Is Required for Morphological
Segmentation of the Paraxial Mesoderm
We have previously demonstrated that disrupting Eph/
Ephrin signaling in the paraxial mesoderm of wild-type
zebrafish embryos disturbs the formation of somites.
for which Eph/Ephrin signaling is required through anal-
ysis of the fss mutant, which lacks organized somites.
be disrupted in fss mutants, our data show that loss of
somite boundaries is most likely due to the absence of
Eph/Ephrin signaling interfaces. The absence of ephA4
interfaces between EphA4-expressing and EphrinB2a/
EphrinA1-expressing cells that normally occur between
ration of Eph/Ephrin signaling in the paraxial mesoderm
by the apposition of either wild-type or fss?/?cells ex-
pressing EphA4 with Ephrin-expressing fss?/?host cells
results in the rescue of morphologically distinct bound-
aries. In the case of fss?/?cells expressing EphA4, this
rescue occurs without restoring the expression of other
Conversely, apposition of wild-type cells, which lack
detectable ephA4 expression, with fss?/?cells does not
result in boundary formation, despite the fact that wild-
type cells do express anterior segmental markers other
than ephA4. These results demonstrate that the pres-
ence of EphA4 is sufficient to restore boundaries to the
paraxial mesodermof fss?/?embryos. Furthermore,they
imply that Eph/Ephrin signaling mediates the final step
sition of anterior or posterior segmental character.
An intracellularly truncated form of EphA4 can also
restore boundaries in fss?/?embryos. This suggests that
signaling downstream of the receptor is not essential
and that Ephrin reverse signaling is sufficient to induce
the formation of a physical furrow between adjacent
cell populations. In contrast, bidirectional Eph/Ephrin
signaling is required to restrict cell intermingling and
boundary formation in blastomere intermixing assays
. However, in these assays, cells normally intermin-
gle extensively, whereas in the rostral PSM, there is
hardly any cell movement and cell mixing does not oc-
cur. Therefore, it may be that in situations in which cell
movements are limited, unidirectional Eph/Ephrin sig-
naling is sufficient to induce boundary furrow formation.
Similarly, unidirectional Eph/Ephrin signaling is suffi-
cient to restrict cell movement within hindbrain rhom-
In addition to initiating boundary formation, restora-
tion of Eph/Ephrin signaling leads to the appropriate
alignment of muscle fiber attachment sites at these
boundaries, even in situations in which the initial epithe-
lialization of boundary cells fails to occur. Therefore,
epithelialization is not an absolute prerequisite for the
maturation of somite boundaries. However, it may be
important for the correct organization and compaction
of muscle fibers since these features were more com-
pletely rescued in situations where boundary cells had
containing adherens junctions . Both wild-type and
fss?/?cells expressing exogenous EphA4 acquire a co-
lumnar morphology and show apical localization of
?-catenin when transplanted into the paraxial meso-
derm of fss?/?mutants. These results suggest that these
tion of EphA4 signaling. In support of this interpretation,
expression of intracellularly truncated EphA4 does not
lead to the acquisition of a columnar morphology or
?-catenin relocalization, despite the fact that it is able
to induce the formation of morphologically distinct
boundaries. Therefore, epithelialization is not simply a
consequence of the formation of boundaries between
paraxial mesodermal cells. A similar conclusion has
been reached from analysis of mice lacking function of
the transcription factor Paraxis ; in these mice, the
paraxial mesoderm exhibits intersomitic furrows, but
cells at the boundaries fail to become epithelial. To-
gether, these results suggest that boundary furrow for-
mation isan earlyevent inthe partitioningof theparaxial
within the cells adjacent to the boundary require addi-
Further features of the epithelialization of somite cells
are the basally directed relocalization of the nucleus and
the apical relocalization of the centrosome. Perhaps sur-
prisingly, relocalization of the cellular organelles was un-
coupled from some other aspects of epithelialization and
occurred in host fss?/?mutant cells only when they were
confronted with EphA4-expressing wild-type cells. There-
fore, these aspects of epithelialization are a cell-nonau-
tonomous consequence of EphA4 signaling. Cell-non-
autonomous effects of Eph receptors are generally
thought to be independent of the intracellular domain
of the receptor and are usually attributed to reverse
signaling downstream of Ephrins expressed in adjacent
cells (see  for a review). However, nuclear relocaliza-
tion in fss?/?host cells requires the intracellular domain
of EphA4 and may be independent of Ephrin reverse
signaling. One possible explanation of these observa-
tions is that a parallel pathway is activated downstream
of EphA4 signaling in the receptor-bearing cells that
triggers nuclear relocalization in adjacent cells. Since
Eph receptors are known to interact with other trans-
membrane receptors during synapse formation, cranio-
facial development, and neural connectivity (reviewed
in ), it is plausible that cross-talk is also occurring
during somite epithelialization. Nevertheless, a role for
Ephrin reverse signaling in epithelialization cannot be
ruled out since the intracellular domain of the receptor
in adjacent cells.
Several lines of evidence led us to investigate the
nent of a pathway cooperating with EphA4 in promoting
epithelialization (Figure S3). However, Papc (in the pres-
ence of EphA4) is neither sufficient in fss?/?cells nor
likelyto berequired inwild-type cellsto promoteepithe-
lialization of adjacent fss?/?cells (Figure S3). Alternative
candidates for proteins that interact with Eph receptors
to promote epithelialization includethe integrin family of
molecules are the classical ligands for integrins, mem-
EphA4 Signaling Leads to Epithelialization
of Boundary Cells
During epithelialization, cells acquire a columnar mor-
phology and form apically positioned cadherin/catenin-
Eph Signaling Regulates Somite Epithelialization
Figure 6. Summary Model of Roles for Eph/Ephrin Signaling during Somite Epithelialization
The red circles or rectangles represent donor wild-type or fss?/?cells during (circles) and after (rectangles) epithelialization. The white circles
or rectangles represent host fss?/?cells. Although Ephrins (orange) are represented as transmembrane molecules, GPI-linked Ephrins may
also play a role. The cofactor (yellow) interacting with EphA4 in donor cells is represented as a transmembrane molecule; however, it could
be a secreted molecule or other form of protein.
(A1 and B1) Donor cells expressing EphA4 (blue) and host cells expressing interacting Ephrins (orange) are in contact within the paraxial
mesoderm. This allows for Eph/Ephrin binding and activation. Both cells present mesenchymal morphology. ?-catenin (green) is homogenously
distributed throughout the cell membrane, the nucleus is localized in the center of the cell body, and the centrosome (red) is localized randomly
within the cytoplasm.
(A2 and B2) Activation of Ephrin reverse signaling leads to localized de-adhesion and boundary formation between both cells. Activation of
EphA4 signaling leads to relocalization of ?-catenin toward the apical pole of the cell, cell elongation, and boundary formation. (A2) In wild-
type cells, EphA4 signaling also leads to the activation of factor X (yellow) in donor cells and its interaction with factor Y (brown) in adjacent
host cells. Signaling downstream of factor Y results in the basal relocalization of the nucleus toward the boundary, the apical relocalization
of the centrosome, and the acquisition of a columnar morphology. (B2) In fss?/?donor cells, where factor X is not present or has compromised
function, signaling downstream of factor Y in adjacent fss?/?host cells does not occur and they remain mesenchymal, despite boundary
brane bound ligands also exist (see  for a review).
In mouse embryos lacking ?5-integrin, the paraxial
mesoderm segments, but epithelial somites fail to form
, suggesting a role for integrin signaling in somite
formation. Furthermore, roles for Eph receptor signaling
in other situations [30–36]. At present, little is under-
stood about the biochemical interactions between Eph
receptors/Ephrins and integrins, but it is clearly an area
in need of further exploration.
mental Data); this removal of protein presumably termi-
nates Eph/Ephrin signaling. Signaling downstream of
EphA4subsequently leads to the acquisition of a colum-
nar morphology and apical accumulation of ?-catenin
in a cell-autonomous manner. Activation of EphA4 sig-
naling also leads to the activation of a parallel pathway
that promotes the acquisition of columnar morphology
and polarized relocalization of the nucleus and the cen-
trosome in adjacent cells (Figure 6A). In fss?/?cells,
lialization in the adjacent cell population (Figure 6B).
EphA4 promotes different aspects of epithelialization
occurs in receptor-bearing cells, whereas relocalization
of the nucleus and the centrosome occur in ligand-
expressing cells. However, in wild-type embryos, cells
on both sides of the somite boundary undergo all as-
pects of epithelialization. One possible explanation of
why cells expressing exogenous EphA4 do not undergo
relocalization of the subcellular organelles may be that
overexpression of the receptor leads to a gain-of-func-
tion phenotype and not to the phenotype that is a result
Eph Receptors and Ephrins as Effectors
of Somite Morphogenesis
We present a model to explain our results and to predict
how Eph/Ephrin signaling regulates boundary formation
and epithelialization (Figure 6). In the PSM of fss?/?chi-
maeras, cells expressing exogenous EphA4 are in close
proximity to cells expressing endogenous Ephrin ligands,
with the consequence that receptors and ligands bind,
cluster, and become activated. Events downstream of
the Eph receptor and Ephrin activation lead to local
de-adhesion and boundary formation. Coincident with
furrow formation, EphA4 protein is removed from the cell
surface facing the boundary (see Figure S4 in the Supple-
lecular Probes) staining was performed after antibody staining. Em-
of 1:40 (in 2% PBT) from the commercial stock.
of normal activation of EphA4. Indeed, ectopic EphA4
catalytic activity in Xenopus embryos leads to the loss
of cell polarity in early blastula cells . In our experi-
ments, however, activation of EphA4 in wild-type or
fss?/?cells does lead to the acquisition of columnar
morphology and the polarized distribution of ?-catenin,
type somite boundaries. Therefore, our data are consis-
tent with the possibilitythat exogenous EphA4 recapitu-
lates the normal activity of the receptor and that nuclear
relocalization is a cell-nonautonomous effect of EphA4
To extrapolate our results to a wild-type situation
where all aspects of epithelialization occur in cells on
both sides of the intersomitic boundary, we need to
postulate that both cell populations express receptor
and ligand. Although a receptor expressed in the poste-
rior domain of the forming somite has not yet been
identified in zebrafish, several Eph proteins that could
fulfill this role are known in other species [38–41]. Based
on these observations, we think it highly likely that the
posterior regionof the forming somiteexpresses an Eph
receptor yet to be identified in fish. This receptor could
then mediate ?-catenin relocalization in posterior cells
and relocalization of cellular organelles in anterior cells
on the other side of the forming somite boundary.
Cloning, Synthesis, and Injection of mRNA
As previously described , the EphA4eGFP fusion construct was
made by inserting full-length Xenopus EphA4 into the pEGFP-N1
vector (Clontech) upstream of, and in frame with, the GFP coding
sequence. The coding sequence for the fusion protein was subse-
quently cloned into the pCS2 vector for in vitro transcription. The
same procedure was used for truncated EphA4eGFP and truncated
EphrinB2a. EphA4 was truncated after amino acid 602 by using PCR
primers (5? primer, 5?-TCAGATCTGCCACCATGGCTGGGATTGTA-3?;
3? primer, 5?-ATCCCGGGATTCATAAGTAAATGGGTC-3?). Ephrin-
B2a was truncated after amino acid 251 by using PCR primers
(5? primer, 5?-TACCGCGGACCATGGGCGACTCT-3?; 3? primer, 5?-
GTGGATCCCGTCGTCGATACTTCAGGAG-3?) . Capped mRNA
was synthesized as previously described , and 400 pl was in-
jected into embryos at the 1- or 2-cell stage. The concentrations at
which the different mRNAs were injected were 300–400 ng/?l for
full-length EphA4eGFP and 200 ng/?l for truncated EphA4eGFP
and truncated Ephrin-B2eGFP. At these concentrations, truncated
formation in wild-type embryos (data not shown).
Mosaic experiments were performed as previously described .
Donor embryos were injected with mRNA encoding GFP fusion pro-
teins or a mixture of RNA encoding GFP or CFP and the mRNA
of interest. Donor cells were selected for transplantation by green
fluorescence to ensure that they were expressing the injected
mRNAs. In some experiments, donors were coinjected with fixable
rhodamine dextran (Molecular Probes) to allow visualization of the
transplanted cells in host embryos stained with green dyes.
The results presented in this paper indicate that Eph/
Ephrin signaling in the rostral PSM is an important com-
ponent of the molecular machinery that drives somite
morphogenesis. Restoration of EphA4/Ephrin signaling
in the paraxial mesoderm of fss?/?mutants is sufficient
to rescue the formation of morphologically distinct
boundaries. Moreover, activation of EphA4 signaling re-
sults in the mesenchymal-to-epithelial transition in the
morphology ofboundary cellsand therebyrecapitulates
most aspects of somite morphogenesis. However, res-
tionally require the activity of pathways acting in parallel
to Eph/Ephrin signaling.
Visualization of Cell Morphology in Living Embryos
Living embryos at stages between 8 and 12 somites were incubated
ing the protocol of Cooper et al. . Embryos were mounted in
agarose for visualization under confocal microscopy by using an
Cells were considered of columnar morphology when the height/
width ratio was between 2.3 and 3, compared to an average height/
width ratio of 1.3 for mesenchymal cells.
Supplemental Data including one movie and four additional figures
with discussion are available at http://www.current-biology.com/
Maintenance of Fish
Breeding zebrafish (Danio rerio) were maintained at 28?C on a 14
hr light/10 hr dark cycle. Wild-type and mutant embryos were col-
lected by natural spawning and were staged according to Kimmel
et al. . Mutant embryos were generated from fish carrying the
te314a or ti1 alleles of fused somites .
We thank Patrick Blader, in whose laboratory some of this work
was done, Stephen Nurrish for his patience and support, Simon
Hughes for providing the myosin antibody, Sharon Amacher for
providing the Papc constructs, Ajay Chitinis for the XDelSTU con-
struct, and Julie Cooke for comments on the manuscript. We would
like to thank members of our group for discussions on this work,
Daniel Ciantar for his help with confocal microscopy, Carole Wilson
and her colleagues for care of the fish, and many colleagues in
the community for provision of various reagents. This work was
supported by grants from the Wellcome Trust and Biotechnology
and Biological Sciences Research Council to S.W.W. and N.H.
S.W.W. is a Wellcome Trust Senior Research Fellow.
In Situ Hybridization and Immunocytochemistry
the protocol of Thisse et al. . For muscle labeling, antibody
staining was performed essentially as previously described , by
using pan-myosin antibody 1025 (a gift from Simon Hughes) at 1:2.5
dilution. For ?-catenin and ?-tubulin immunostaining, embryos were
were permeabilized by acetone treatment at ?20?C for 5 min and
were then washed in PBT (0.8% Triton X-100 in PBS). For ?-catenin,
embryos were permeabilized by methanol treatment at ?20?C and
were then washed in PDT (PBT ? 1% DMSO). Monoclonal anti-
?-catenin antibody (BD Transduction Laboratories) was used at
1:500 dilution.Monoclonal anti-?-tubulinantibody (Sigma)was used
at 1:200 dilution. Alexa fluor 488 anti-mouse IgG (Molecular Probes)
was used as a secondary antibody at 1:200 dilution. Phalloidin (Mo-
Received: March 14, 2003
Revised: July 2, 2003
Accepted: July 25, 2003
Published: September 16, 2003
Eph Signaling Regulates Somite Epithelialization
mite boundary formation revealed by time-lapse analysis. Sci-
ence 298, 991–995.
22. Wood, A., and Thorogood, P. (1994). Patterns of cell behaviour
tebrate embryos. Dev. Dyn. 201, 151–167.
23. Davis, S., Gale, N., Aldrich, T., Maisonpierre, P., Lhotak, V.,
Pawson, T., Goldfarb, M., and Yancopoulos, G. (1994). Ligands
for EPH-related receptor tyrosine kinases that require mem-
brane attachment or clustering for activity. Science 266,
sion of truncated Sek-1 receptor tyrosine kinase disrupts the
segmental restriction of gene expression in the Xenopus and
zebrafish hindbrain. Development 121, 4005–4016.
25. Tepass, U., Godt, D., and Winklbauer, R. (2002). Cell sorting in
animal development: signalling and adhesive mechanisms in
the formation of tissue boundaries. Curr. Opin. Genet. Dev. 12,
26. Burgess, R., Rawls, A., Brown, D., Bradley, A., and Olson, E.
(1996). Requirement of the paraxis gene for somite formation
and musculoskeletal patterning. Nature 384, 570–573.
27. Kullander, K., and Klein, R. (2002). Mechanisms and functions
28. Bokel, C., and Brown, N. (2002). Integrins in development. Mov-
ing on, responding to, and sticking to the extracellular matrix.
Dev. Cell 3, 311–321.
29. Yang, J.T., Bader, B.L., Kreidberg, J.A., Ullman-Cullere, M.,
Trevithick, J.E., and Hynes, R.O. (1999). Overlapping and inde-
pendent functions of fibronectin receptor integrins in early me-
sodermal development. Dev. Biol. 215, 264–277.
30. Zou, J.X., Wang, B., Kalo, M.S., Zisch, A.H., Pasquale, E.B., and
Ruoslahti, E. (1999). An Eph receptor regulates integrin activity
through R-Ras. Proc. Natl. Acad. Sci. USA 96, 13813–13818.
31. Huynh-Do, U., Stein, E., Lane, A.A., Liu, H., Cerretti, D.P., and
Daniel, T.O. (1999). Surface densities of ephrin-B1 determine
EphB1-coupled activationof cellattachmentthrough alphavbeta3
and alpha5beta1 integrins. EMBO J. 18, 2165–2173.
32. Becker, E., Huynh-Do, U., Holland, S., Pawson, T., Daniel, T.O.,
and Skolnik, E.Y. (2000). Nck-interacting Ste20 kinase couples
Eph receptors to c-Jun N-terminal kinase and integrin activa-
tion. Mol. Cell. Biol. 20, 1537–1545.
33. Miao, H., Burnett, E., Kinch, M., Simon, E., and Wang, B. (2000).
Activation of EphA2 kinase suppresses integrin function and
causes focal-adhesion-kinase dephosphorylation. Nat. Cell
Biol. 2, 62–69.
34. Genda, T., Sakamoto, M., Ichida, T., Asakura, H., and Hirohashi,
S. (2000). Loss of cell-cell contact is induced by integrin-medi-
static hepatocellular carcinoma cells. Lab. Invest. 80, 387–394.
35. Davy, A., and Robbins, S.M. (2000). Ephrin-A5 modulates cell
adhesion and morphology in an integrin-dependent manner.
EMBO J. 19, 5396–5405.
36. Zou, J.X., Liu, Y., Pasquale, E.B., and Ruoslahti, E. (2002). Acti-
vated SRC oncogene phosphorylates R-ras and suppresses
integrin activity. J. Biol. Chem. 277, 1824–1827.
37. Winning, R.S., Wyman, T.L., and Walker, G.K. (2001). EphA4
activity causes cell shape change and a loss of cell polarity in
Xenopus laevis embryos. Differentiation 68, 126–132.
38. Kilpatrick, T., Brown, A., Lai, C., Gassman, M., Goulding, M.,
and Lemke, G. (1996).Expression of the Tyro4/Mek4/Cek4 gene
specifically marks a subset of embryonic motor neurons and
their muscle targets. Mol. Cell. Neurosci. 7, 62–74.
39. Ellis, J., Liu, Q., Breitman, M., Jenkins, N.A., Gilbert, D.J., Cope-
land, N.G., Tempest, H.V., Warren, S., Muir, E., Schilling, H., et
al. (1995). Embryo brain kinase: a novel gene of the eph/elk
receptor tyrosine kinase family. Mech. Dev. 52, 319–341.
40. Soans, C., Holash, J.A., Pavlova, Y., and Pasquale, E.B. (1996).
Developmental expression and distinctive phosphorylation of
the Eph-related receptor tyrosine kinase, Cek9. J. Cell Biol. 135,
41. Araujo, M., and Nieto, M.A. (1997). The expression of chick
EphA7 during segmentation of the central and peripheral ner-
vous system. Mech. Dev. 68, 173–177.
42. Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B., and
1. Maroto, M., and Pourquie, O. (2001). A molecular clock involved
in somite segmentation. Curr. Top. Dev. Biol. 51, 221–248.
2. Saga, Y., and Takeda, H. (2001). The making of the somite:
molecular events in vertebrate segmentation. Nat. Rev. Genet.
3. Dale, J., Maroto, M., Dequeant, M., Malapert, P., McGrew, M.,
and Pourquie, O. (2003). Periodic Notch inhibition by lunatic
fringe underlies the chick segmentation clock. Nature 421,
tor and wavefront that create the zebrafish somite. Semin. Cell
Dev. Biol. 13, 481–488.
5. Barrantes, I.B., Elia, A.J., Wunsch, K., De Angelis, M.H., Mak,
T.W., Rossant, J., Conlon, R.A., Gossler, A., and de la Pompa,
J.L. (1999). Interaction between Notch signalling and Lunatic
fringe during somite boundary formation in the mouse. Curr.
Biol. 9, 470–480.
B., Brennan, C., Green, A., Wilson, S., and Holder, N. (2000).
Anteroposterior patterning is required within segments for so-
mite boundary formation in developing zebrafish. Development
7. Holder, N., and Klein, R. (1999). Eph receptors and ephrins:
effectors of morphogenesis. Development 126, 2033–2044.
8. Palmer, A., and Klein, R. (2003). Multiple roles of ephrins in
morphogenesis, neuronal networking, and brain function.
Genes Dev. 17, 1429–1450.
9. Murai, K.K., and Pasquale, E.B. (2003). ‘Eph’ective signaling:
forward, reverse and crosstalk. J. Cell Sci. 116, 2823–2832.
T., Henkemeyer, M., Strebhardt, K., Hirai, H., Wilkinson, D., et
al. (1996). Eph receptors and ligands comprise two major speci-
embryogenesis. Neuron 17, 9–19.
11. Cowan, C.A., and Henkemeyer, M. (2002). Ephrins in reverse,
park and drive. Trends Cell Biol. 12, 339–346.
12. Kalo, M.S., and Pasquale, E.B. (1999). Signal transfer by Eph
receptors. Cell Tissue Res. 298, 1–9.
13. Halford, M.M., Armes, J., Buchert, M., Meskenaite, V., Grail, D.,
Hibbs, M.L., Wilks, A.F., Farlie, P.G., Newgreen, D.F., Hovens,
C.M., et al. (2000). Ryk-deficient mice exhibit craniofacial de-
fects associated with perturbed Eph receptor crosstalk. Nat.
Genet. 25, 414–418.
14. Dalva, M.B.,Takasu, M.A.,Lin, M.Z.,Shamah, S.M.,Hu, L.,Gale,
N.W., and Greenberg, M.E. (2000). EphB receptors interact with
NMDA receptors and regulate excitatory synapse formation.
Cell 103, 945–956.
15. Mellitzer, G., Xu, Q., and Wilkinson, D.G. (1999). Eph receptors
and ephrins restrict cell intermingling and communication. Na-
ture 400, 77–81.
16. Xu, Q., Mellitzer, G., Robinson, V., and Wilkinson, D.G. (1999).
In vivo cell sorting in complementary segmental domains medi-
ated by Eph receptors and ephrins. Nature 399, 267–271.
17. Cooke, J., Moens, C., Roth, L., Durbin, L., Shiomi, K., Brennan,
C., Kimmel, C., Wilson, S., and Holder, N. (2001). Eph signalling
functions downstream of Val to regulate cell sorting and bound-
ary formation in the caudal hindbrain. Development 128,
18. Durbin, L., Brennan, C., Shiomi, K., Cooke, J., Barrios, A., Shan-
mugalingam, S., Guthrie, B., Lindberb, R., and Holder, N. (1998).
Eph signalling is required for segmentation and differentiation
of the somites. Genes Dev. 12, 3096–3109.
19. Nikaido, M., Kawakami, A., Sawada, A., Furutani-Seiki, M.,
Takeda, H., and Araki, K. (2002). Tbx24, encoding a T-box pro-
tein, is mutated in the zebrafish somite-segmentation mutant
fused somites. Nat. Genet. 31, 195–199.
20. van Eeden, F., Granato, M., Schach, U., Brand, M., Furutani-
Seiki, M., Haffter, P., Hammerschmidt, M., Heisenberg, C.-P.,
Jiang, Y.-J., Kane, D., et al. (1996). Mutations affecting somite
formation and patterning in the zebrafish, Danio rerio. Develop-
ment 123, 153–164.
21. Kulesa, P.M., and Fraser, S.E. (2002). Cell dynamics during so-
Current Biology Download full-text
Schilling, T.F. (1995). Stages of embryonic development of the
zebrafish. Dev. Dyn. 203, 253–310.
43. Thisse, C., Thisse, B., Schilling, T.F., and Postlethwait, J.H.
(1993). Structure of the zebrafish snail1 gene and its expression
44. Blagden, C.S., Currie, P.D., Ingham, P.W., and Hughes, S.M.
(1997). Notochord induction of zebrafish slow muscle mediated
by Sonic hedgehog. Genes Dev. 11, 2163–2175.
45. Cooper, M., D’Amico, L., and Henry, C. (1999). Confocal micro-
scopic analysis of morphogenetic movements. Methods Cell
Biol. 59, 179–204.