Oncogene (1997) 14, 2159 ± 2166
 1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00
Identi®cation of XLerk, an Eph family ligand regulated during mesoderm
induction and neurogenesis in Xenopus laevis
Teri L Jones1, Irina Karavanova2, Lisa Chong1, Ren-Ping Zhou3 and Ira O Daar1
1
Laboratory of Leukocyte Biology, 2Laboratory of Comparative Carcinogenesis; National Cancer Institute-Frederick Cancer
Research and Development Center, Frederick, Maryland 21702; 3Laboratory for Cancer Research, Rutgers University, Piscataway,
New Jersey 08855, USA
We have isolated and characterized the ®rst Xenopus
transmembrane Eph ligand, XLerk (Xenopus Ligand for
Eph Receptor Tyrosine Kinases). While this ligand has
72% identity with the closest mammalian family
member, Lerk-2, it is the cytoplasmic domain of this
molecule that is the most conserved domain with 95%
identity. XLerk exists as a maternally expressed mRNA,
however, expression of transcripts and protein increase
during gastrulation and again in the late swimming
tadpole stage. In the adult, XLerk is expressed at low
levels in most adult tissues with increased levels observed
in the kidney, oocytes, ovary and testis. While low levels
of XLerk expression are observed in the adult brain, in
situ hybridization analysis demonstrates prominent
expression in the developing olfactory system, retina,
hindbrain, cranial ganglia, and somites. Furthermore, we
have shown that XLerk transcripts are signi®cantly
elevated during mesoderm induction caused by activin
and FGF, but not during noggin-induced neuralization.
These results suggest a role for XLerk in the developing
mesenchymal and nervous tissue.
Keywords: Eph ligand; xenopus development; nervous
system
Introduction
Receptor tyrosine kinases (RTKs) phosphorylate
cellular substrates that lead to a multitude of
responses including the regulation of multiple kinase
cascades, gene expression, and ultimately cell division
or di€erentiation. The Eph (Erythropoietin Producing
Hepatocellular) family of RTKs is amongst the newest
and largest families to emerge, numbering at approximately 13 independent members (Pandey et al., 1995;
Tuzi and Gullick, 1994). These transmembrane RTKs
de®ne a new protein tyrosine kinase subclass on the
basis of structural di€erences that include a single
cysteine rich domain and two ®bronectin type III
repeats found in the extracellular domain. The original
Eph receptor cDNA was isolated from a human
hepatoma library and overexpression of Eph transcripts has been observed in some carcinomas of the
colon, lung, liver and in some melanomas (Easty et al.,
1995; Hirai et al., 1987; Maru et al., 1988).
The Eph receptors exhibit temporally and spatially
restricted patterns of expression during embryogenesis,
Correspondence: IO Daar
Received 12 November 1996; revised 4 February 1997; accepted
4 February 1997
and several Eph-related receptors are expressed in
structures of the developing nervous system (Brambilla
et al., 1995; Pandey et al., 1995; Tuzi and Gullick,
1994). Eph receptors have also been shown to play a
role in axon fasciculation and guidance (TessierLavigne and Goodman, 1996). One such study by
Henkemeyer et al. (1996) shows a central role for Nuk
in commissural axon path®nding demonstrated in mice
which lack Nuk or express an aberrant Nuk product.
Recently, Eph family members have been isolated
from Xenopus. Pagliaccio (XSek-1) is expressed in the
marginal zone and neural crest cells during development (Winning and Sargent, 1994). Ectopic expression
of dominant negative XSek-1 (Pag) during Xenopus
development disrupts segmental restriction from
rhombomeres three to ®ve (Xu et al., 1995). Moreover, expression of an EGF (extracellular domain)/
Pagliaccio (intracellular kinase domain) chimera during
early embryogenesis caused embryonic blastomeres to
disaggregate by the late blastula stage, implicating a
role for this receptor in cell adhesion or cell motility
(Winning et al., 1996). G50, a Xenopus Eck-related
receptor, is expressed in a subpopulation of neural
crest cells that migrates to the second hyoid arch
(Brandli and Kirschner, 1995). Finally, Xek (Xenopus
elk-like kinase), an Eph receptor of the elk-related
subclass, was identi®ed and shown to be a maternal
message that is expressed throughout the central
nervous system of tailbud stage embryos (Jones et al.,
1995).
Recent studies have identi®ed several ligands that
can bind to and/or activate members of the Eph family
of receptor tyrosine kinases. These ligands are all
membrane bound with the exception of one Xenopus
B61-related protein termed XElf-a' (Weinstein et al.,
1996). This is the only soluble form of an Eph ligand
described to date. A majority of the members, B61,
Elf-1, Ehk-L (Lerk3), Lerk4 and AL-1(RAGS) are
anchored by a glycosyl phosphatidylinositol (GPI) tail
(Bartley et al., 1994; Cheng and Flanagan, 1994; Davis
et al., 1994; Drescher et al., 1995; Kozlosky et al.,
1995; Shao et al., 1995; Winslow et al., 1995).
Although these ligands are able to bind to various
Eph receptor family members, they share only about
50% homology. Based upon their binding anity, the
GPI-linked ligands have been placed in a subgroup
having greater anity for Eck-related receptors, while
transmembrane ligands preferentially bind the Elkrelated receptors (Gale et al., 1996a). Lerk2 (Cek5-L),
HTK-L (Elf-2) and Elk-L3 are type I transmembrane
proteins which share very highly conserved C-terminal
domains (Beckmann et al., 1994; Bennett et al., 1995;
Bergemann et al., 1995; Gale et al., 1996b; Shao et al.,
XLerk expression during development
TL Jones et al
2160
1994). While membrane anchorage has been shown to
be necessary for the activation of receptors by these
ligands (Bennett et al., 1995; Bergemann et al., 1995;
Brambilla et al., 1995; Davis et al., 1994; Shao et al.,
1994, 1995), a soluble form of Lerk-2 ligand has been
reported to stimulate autophosphorylation activity of
the Hek-2 receptor (Bohme et al., 1996).
Functional roles for the receptor-ligand interactions
are beginning to be determined. Recent ®ndings
suggest that AL-1 and its cognate receptor participate
in axon bundling (Cheng et al., 1995; Drescher et al.,
1995; Winslow et al., 1995). Additionally, Cek5 and
Qek5 have been found to display complementary
gradients with their proposed ligands in the retinotectal system, suggesting a role in retinal neuron
guidance (Holash and Pasquale, 1995; Kenny et al.,
1995). Furthermore, Elf-1, has been demonstrated to
cause repulsion of axons (Gao et al., 1996; Nakamoto
et al., 1996).
We report the isolation of the ®rst Xenopus
transmembrane Eph ligand, XLerk, that shares a high
degree of homology with the mammalian Lerk-2
product. We de®ne the sequence of the full-length
cDNA that reveals a striking degree of evolutionary
conservation in the cytoplasmic domain when compared with the extracellular receptor binding domain.
The adult and embryonic patterns of expression are
described and suggest a role for this ligand in neuronal
and mesoderm derived tissues.
Results
Cloning and sequence analysis of XLerk
To isolate transmembrane proteins of the subclass
reported to be the ligands of the Elk-related receptors
(Gale et al., 1996), the entire 1.1 kb coding fragment
from murine Lerk-2 (Shao et al., 1994) was used to
screen a Xenopus stage 24 cDNA library. A partial
cDNA of 2.4 kb was isolated, and sequence analysis
determined that this isolate contained only the 3' half
of the coding region. Rescreening of the Xenopus stage
24 cDNA library with a 1.5 kb EcoRI fragment
obtained from the initial cDNA resulted in the
isolation of a 4.0 kb cDNA. Comparative databank
analysis revealed homology with the Lerk-2 sequence
(Bartley et al., 1994; Beckman et al., 1994; Fletcher et
al., 1994). This cDNA sequence was named XLerk
(Xenopus Ligand for Eph Receptor Tyrosine Kinases),
and shown to contain an open reading frame of 329
amino acids (Figure 1). The methionine codon is
consistent with a translation initiation site (Kozak,
1992) and is followed by a signal peptide sequence.
The XLerk sequence contains four cysteine residues
that are well conserved among the Eph family of
ligands, in addition to two potential N-linked
glycosylation sites. The residues between amino acids
221 and 248 are hydrophobic and encode a putative
transmembrane domain (Figure 1). The intracellular
Figure 1 Nucleotide and predicted amino acid sequences of the XLerk cDNA. The putative signal peptide and transmembrane region are
underlined. The conserved cysteine residues are circled. The potential N-glycosylation sites are boxed
XLerk expression during development
TL Jones et al
domain consists of 80 amino acids that is followed by
2.75 kb of 3' untranslated region lacking a polyadenylation site.
Figure 2 shows an amino acid comparison of XLerk
with two other transmembrane proteins, rat Lerk-2 and
murine HTK ligand. XLerk shares close homology
with rat Lerk-2 (72% identity) and less with the murine
HTK ligand (58%). While there is a very signi®cant
degree of homology, it is still dicult to proclaim
XLerk as the frog homolog of Lerk-2 with absolute
certainty. The two regions of greatest diversity between
rLerk-2 and XLerk include the signal peptide region
and the extracellular sequence adjacent to the
transmembrane domain known as the spacer region
(Pandey et al., 1995). Strikingly, the intracellular
domain is 95% conserved in these two evolutionarily
diverged species, suggesting an important function for
this domain.
Expression of XLerk during Xenopus developmental
stages and in adult tissues
Unfertilized Egg
Examination of XLerk mRNA in Xenopus embryonic
stages was performed by the RNase protection assay.
A labeled fragment of XFGFr RNA was used as an
internal loading control as previously described by
Musci et al. (1990). XLerk RNA is observed in
unfertilized eggs and thus exists as a maternal
transcript (Figure 3). Further analysis of RNA from
subsequent stages reveals continual expression of
XLerk. After the mid-blastula transition (stage 9),
there is a decrease in XLerk expression followed by
increased expression during late gastrulation. Further
increases in mRNA abundance are observed during the
swimming tadpole stage (stage 28) through stage 37 of
development.
To determine the temporal pro®le of XLerk protein
expression during development, we examined extracts
from unfertilized eggs and embryonic stages (stages 6 ±
41). Western blot analysis was performed on these
extracts using Lerk2 antibody directed against a
peptide representing the C-terminal 18 amino acids
(Santa Cruz Biotechnology). This antibody does not
cross react with Lerk 1, 3, 4 or Elf-1. However, we
have determined that this antibody will recognize
XLerk protein expressed in both rabbit reticulocyte
lysates and in Xenopus oocytes using Western blot and
immunoprecipitation analysis (data not shown). In
Figure 4, the putative XLerk protein is detected by late
gastrulation (stage 13) and expression increases until
stage 41 of embryonic development. While we cannot
exclude the possibility that the recognized protein may
be a cross reacting species, the pattern of expression is
6
Stages
9 11 13 16 18 20 24 28 30 37 41
51 Kd —
➝
36 Kd —
29 Kd —
Figure 2 Comparison of the amino acid sequences of XLerk, rat
Lerk-2 and murine HTK. Sequences were aligned using the
Genetics Computer Group program Best®t. Residues that are
identical are shaded and gaps are shown by periods
XLERT
▲
EF1α
Testis
Stomach
Spleen
Ovary
Oocyte
Muscle
Lung
Liver
Kidney
Intestine
Heart
26 28 30 37
Brain
18 24
tRNA
16
▲
▲
FGFR
10 12
▲
XLERK
7
Probe
Egg
Stages
Figure 4 Western blot analysis of the XLerk protein in Xenopus
embryonic stages. The putative XLerk protein is detectable by
stage 13 and increases in expression until stage 41. Three embryos
from each indicated stage were lysed and analysed by Western
blot analysis with a Lerk-2 antibody, followed by an HRPcoupled secondary antibody and developed using ECL. Molecular
weight markers are designated. An arrow indicates the putative
XLerk protein
Figure 3 RNase protection analysis of XLerk transcripts in
Xenopus embryonic stages. Following arti®cial fertilization,
embryos were staged and collected according to Nieukwoop and
Faber (1967). Each lane represents an assay performed on 20 mg
of total mRNA isolated from the indicated stages of embryos.
The results are representative of three experiments. The protected
fragments for XLerk and FGFr are indicated
Figure 5 RNase protection analysis of XLerk expression in adult
Xenopus tissues. Total mRNA (20 mg per lane) was isolated from
the various tissues and assayed. The results are representative of
three experiments. Protected fragments for XLerk and the EF-1a
control are indicated
2161
XLerk expression during development
TL Jones et al
2162
Figure 6 In situ hybridization of XLerk mRNA expression. Sagittal and coronal serial sections of whole embryos from stages 24,
33 and 41 were hybridized with 35S-labeled XLerk antisense and sense RNA probes, then processed for emulsion autoradiography.
(a,b) Stage 24 sagittal sections showing XLerk expression in the olfactory placode (op) and in the olfactory bulb (ob). Weaker
hybridization is observed along the midbrain and spinal cord with stronger hybridization throughout the hindbrain (hb). Expression
is also observed in the developing somites (s), cranial ganglia (cg) and nerves, and heart anlage (ha); (a) bright®eld (b) dark®eld.
(c,d) Stage 33 embryo showing a magni®cation of XLerk expression in the otic vesicle (ot) as indicated by the short open arrow and
hybridization of the geniculate ganglion (gg) just anterior to the otic vesicle as indicated by the long arrow; (c) bright®eld (d)
dark®eld. (e,f) In the coronal sections of stage 41, XLerk expression is observed in the di€erentiating olfactory epithelium (oe), and
the hindbrain-midbrain (hm); (e) bright®eld (f) dark®eld. (h) Stage 41 sagittal section shows punctate expression of XLerk in the gut
and the pronephros as indicated by the arrows. (g) sense probe showing no hybridization
XLerk expression during development
TL Jones et al
Regulation of XLerk during mesoderm induction by
activin and FGF
In Xenopus, the ectoderm will pursue an epidermal
lineage autonomously, while interaction with dorsal
mesoderm is required for the induction of the neural
cell fates (Doniach, 1995; Kessler and Melton, 1994).
Mesoderm is induced in the marginal zone of the frog
blastula through a process that involves a complex
interplay between multiple inductive signals (e.g. FGF
or activin). Exogenous treatment of animal cap
explants with FGF and activin can e€ectively mimic
the action of mesoderm induction (Smith and Harland,
1991). Neural inducing activity has been observed in
Noggin
β-gal
b
FGF
a
Activin
To examine the temporal and spatial localization of
XLerk RNA, the pattern of expression in tissue
sections from embryonic stages 24, 33 and 41 were
examined by in situ hybridization. 35S-labeled antisense
and sense RNA probes were hybridized with sagittal
and coronal sections. In the swimming tadpole stage
(stage 24, Figure 6a and b), XLerk expression was
observed in the olfactory placode and in the adjacent
neural tissue, the olfactory bulb. Weaker hybridization
was detected along the midbrain and spinal cord with
stronger hybridization observed throughout the
hindbrain. Expression of XLerk RNA was also
observed in the developing somites and heart anlage
(Figure 6a and b). Figure 6c and d shows a
magni®cation of XLerk expression in the otic vesicle
of a stage 33 embryo. In addition, there is
hybridization in the geniculate ganglion just anterior
to the otic vesicle. Signi®cant hybridization of XLerk
in the cranial ganglia and nerves is also observed
(Figure 6a ± d). In the coronal sections of embryonic
stage 41, XLerk expression was apparent in the
di€erentiating olfactory epithelium, the forebrain and
hindbrain (Figure 6e and f). XLerk expression was
also prominently observed in the neural retina, with
restricted hybridization at the midbrain/hindbrain
border. In sagittal sections of stage 41 embryo,
XLerk RNA was expressed in the branchial arches
and cranial facial nerves (data not shown). Very
speci®c punctate expression of XLerk can be seen in
the gut and the pronephros, suggestive of innervations
by the myenteric plexus (Figure 6h). These results are
consistent with the expression patterns of murine
Lerk-2 and Elf-2.
Noggin
In situ hybridization analysis of XLerk expression in
Xenopus embryonic stages
animal caps treated with noggin or follistatin (Lamb et
al., 1993; Hemmati-Brivanlou and Melton, 1994; Smith
and Harland, 1992).
To investigate the pathway underlying the regulation
of XLerk transcription, we examined mesoderm and
neural induction in animal cap explants from blastula
stage embryos (stage 8). Animal caps injected with bgalactosidase (b-gal) control mRNA were cultured in
the presence of activin, FGF or alone until midgastrulation (stage 11) (Figure 7a). Animal caps were
also prepared from mid-blastula stage embryos that
had been injected with noggin mRNA (0.125 ng) and
were cultured until mid-gastrulation (Figure 7a). As a
control for noggin-induced neuralization, animal caps
from b-gal or noggin RNA injected embryos were
cultured to the early tailbud stage (stage 26) (Figure
7b). We then determined whether the levels of XLerk
transcripts were altered during neural and/or mesoderm induction.
In response to mesoderm inducers activin and FGF,
the level of XLerk RNA increases signi®cantly by
mid-gastrulation as determined by RT ± PCR analysis
(Figure 7a). To verify the induction of dorsal
mesoderm by activin, mRNA encoding Xenopus
chordin, a gene induced by activin function was
assayed (Sasai et al., 1994; Figure 7a). In addition,
we detected the mRNA expression of Xbra (brachyury), a nuclear factor that is induced in response to
the ventral mesoderm inducer, FGF. XLerk expression
did not increase in response to noggin expression in
the mid-gastrula or the early tailbud (stage 26)
embryo (Figure 7a and b). The induction of neural
crest adhesion molecule (N-CAM) con®rms that the
animal cap has di€erentiated towards a neural fate in
response to noggin (Figure 7b). These results provide
evidence for an increase in XLerk mRNA abundance
during mesoderm induction.
β-gal
very consistent with the observed XLerk RNA
transcript levels (Figure 3).
XLerk mRNA levels were also analysed in Xenopus
adult tissues. A labeled probe of EF-1a RNA was used
as an internal loading control as previously described
by Krieg et al. (1989). XLerk RNA is expressed in
most adult tissues with an increased level observed in
the kidney, oocytes, ovary and testis (Figure 5). XLerk
RNA was also expressed at much lower levels in the
brain, heart, spleen and stomach (Figure 5). While
XLerk mRNA is not abundant in the adult brain
(Figure 5), these ®ndings are similar to that observed
with human and rat Lerk-2 (Beckmann et al., 1994;
Fletcher et al., 1994).
Chordin —
Xbra —
— N-CAM
XLerk —
— XLerk
EF1α —
— EF1α
Figure 7 Increased XLerk expression during mesoderm induction. The blastula stage animal caps were cultured in bu€er until
sibling controls reached mid-gastrulation (a) or tailbud stages (b).
Animal caps from embryos injected with b-galactosidase control
RNA and incubated in the presence of activin (50 ng/ml), bFGF
(100 ng/ml), or left untreated (b-gal), were assayed along with
caps from embryos injected with noggin RNA (0.125 ng). At midgastrulation, XLerk RNA expression was assayed by RT ± PCR
(a). The induction of chordin and Xbra was assayed to verify the
activity of each mesodermal inducer. In the tailbud embryo (stage
25), XLerk and N-CAM transcripts were assayed in animal caps
from embryos injected with b-galactosidase control RNA or
noggin RNA (0.125 ng). (b) As a control, EF-1a gene expression
was determined and an equivalent amount of RNA was used for
each sample
2163
XLerk expression during development
TL Jones et al
2164
Discussion
We report the isolation of XLerk, the ®rst Xenopus
cDNA encoding a transmembrane ligand for the Eph
receptor tyrosine kinase family. XLerk is a protein
consisting of a 222 amino acid extracellular domain, a
hydrophobic transmembrane region and an 80 amino
acid intracellular domain. While there is approximately
72% overall amino acid identity between the rat Lerk-2
protein and XLerk, the intracellular domain is extremely
well conserved (95%). This conservation is of interest
because if the sole function of these ligands is to activate
the receptor, it seems unlikely that the intracellular
domain would be the most conserved region. This
evolutionary conservation strongly suggests that the
cytoplasmic domain has an important function.
One intriguing possibility is that these transmembrane ligands are themselves signaling molecules. Upon
cell-cell contact, the ligand activates the tyrosine kinase
activity of the receptors residing on the surface of
another cell, while sending an unknown signal within
its own host cell. Support for this hypothesis comes
from the recent report by Henkemeyer et al. (1996), in
which a complete gene ablation for the Nuk receptor
caused abberant commissural axon guidance in the
mouse brain. However, expression of the receptor with
only the kinase domain ablated had a normal anterior
commissure, suggesting that the ligand may be a
signaling molecule. In addition, Holland et al. (1996)
has recently demonstrated that a clustered ectodomain
of the Nuk receptor can induce the phosphorylation of
the Elk ligand on tyrosine. There are ®ve potential
tyrosine phosphorylation sites in the cytoplasmic
domain, but none have been mapped yet.
While adult expression of XLerk mRNA in the
brain is low, it has rather prominent expression in the
embryonic brain. This is quite consistent with the
pattern of neural expression demonstrated for Lerk-2
in the mouse (Carpenter et al., 1995). A similar
temporal pattern of mouse Elk mRNA expression
was demonstrated, but with signi®cant di€erences in
spatial expression as was the case for the Xenopus elklike kinase, Xek (Jones et al., 1995). While the Xek
transcripts are dramatically induced at neurulation,
XLerk mRNA and protein levels increase at late
gastrulation and then again in the late swimming
tadpole stage (stage 28). The increases in XLerk
transcription and protein production occur at times
of massive cell movement which permit inductive
events of later development including neural induction.
In this report we provide direct evidence for the
mesodermal induction of XLerk transcripts. In
Xenopus, mesoderm induction involves several signaling molecules from the underlying endoderm including
FGF, activin and BMPs (Kessler and Melton, 1994).
The inductive abilities of these signaling molecules have
been studied using cultured explants from the
prospective ectoderm of the animal pole (animal cap).
This prospective ectoderm is capable of responding to
these inductive signals in culture, and will di€erentiate
towards an epidermal lineage in the absence of the
in¯uence of dorsal mesoderm.
We demonstrate using animal cap explants and
RT ± PCR analysis, that the mesodermal inducers,
activin and FGF, cause an increase in the level of
XLerk RNA by mid-gastrulation, while noggin, a
neural inducer, does not a€ect XLerk mRNA levels
(Figure 7). This suggests a possible role for XLerk
during mesoderm patterning and movement that sets
the stage for later developmental events.
At later developmental stages, XLerk has some
common regions of expression with adhesion molecules
that are likely implicated in morphoregulatory processes.
For example, there are overlapping areas of expression
between XLerk and Fak with regard to somitic and
neural expression (somites, hindbrain and cranial ganglia) (Hens and DeSimone, 1995). The formation of
somites, a segmented mesoderm where XLerk mRNA is
expressed, is very important for embryonic patterning, as
it gives rise to connective tissue, axial skeletal tissue and
muscles. Concurrent with motor neuron innervations to
the somites (at stage 22), in which motor axons invade
the myotomes, the expression of focal adhesion
molecules including Fak, tanabin, talin, paxillin and
integrins may also regulate somite formation (Hens and
DeSimone, 1995; Hemmati-Brivanlou et al., 1992;
Krotoski and Bronner-Fraser, 1990; Lefcort et al.,
1992; Ransom et al., 1993; Tomaselli and Reichardt,
1989). In addition, both N-Cadherin, a neural cell
adhesion molecule, and XLerk are expressed in the
olfactory placode, hindbrain, otic vesicle and heart
anlage (Simonneau et al., 1992).
XLerk mRNA expression is found in the telencephalon, ventral diencephlon, and as a stripe along the
dorsal cells of the spinal cord (Figure 6f) that are
believed to give rise to sensory neurons. XLerk is
highly expressed in sensory structures derived from
placodes (olfactory placode and otic vesicle) and in the
retina, a sensory structure derived from the neural
tube. In later stage embryos (stage 41), there is strong
expression in the nasal epithelium and the mesenchyme
of the developing otic vesicles, suggesting involvement
in morphogenic movements. The prominent expression
of XLerk in the olfactory bulb has also been observed
for Lerk-2 in the mouse (Bouillet et al., 1995;
Carpenter et al., 1995). Also of interest is the punctate
staining of XLerk surrounding the gut which is very
similar to the staining of mouse Hox 2.1 and suggestive
of expression in parasympathatic ganglia of the
myenteric plexus (Holland and Hogan, 1988). Taken
together, these data are consistent with a proposed role
for XLerk in neuronal development, such as axon
guidance, and morphoregulatory processes.
Materials and methods
Animals
Xenopus laevis were obtained from Xenopus I (Ann Arbor,
Michigan) and maintained in the animal facility at NCI ±
FCRDC. Embryos were obtained by in vitro fertilization
(Newport and Kirschner, 1982) and staged accordingly to
Nieuwkoop and Faber (1976).
Library screening
A stage 24 Xenopus embryo cDNA library (lgtll) (gift of
Igor Dawid, NIH) was screened using the entire 1.1 kb
coding region of murine Lerk2 (Shao et al., 1994) as a
probe. Individual clones from each primary positive were
plaque puri®ed. Phage DNA was recovered using a
Lambda DNA puri®cation kit (Stratagene) and subcloned
into the EcoRI Site of pBluescript SK+ (Stratagene).
XLerk expression during development
TL Jones et al
Sequencing of independent clones on both strands was
performed by the dideoxynucleotide chain termination
method as recommended by the manufacturer (USB).
Because these independent clones did not contain the 5'
end of XLerk, the above library was rescreened with a
probe generated from the partial open reading frame of
XLerk. Clones containing the additional 5' end were found
and con®rmed by further sequencing.
RNase protection assay
Total RNA was extracted from whole Xenopus embryos or
adult tissues using Rnazol B (Tel systems) as described by
the manufacturer. Brie¯y, embryos or tissues were
homogenized in Rnazol B with a te¯on homogenizer.
Following the addition of chloroform, the RNA was
precipitated with one volume of isopropanol, then treated
with 4 mM LiCl and reprecipitated at 7208C overnight.
Quantitation of total RNA was performed by spectrophotometry and ethidium bromide staining.
For antisense strand RNA synthesis, a 475 bp fragment of
XLerk that spans nucleotides 811 to 1286 was subcloned into
pBluescript SK+, linearized with BamHI and radiolabeled
with [a-32P]UTP (3000 Ci mmol71; Amersham) as recommended by the manufacturer (Riboprobe Gemini II core
system [Promega]). For unlabeled sense strand RNA
synthesis, the plasmid was linearized with HindIII and
transcribed using T3 polymerase. Antisense probes of the
Xenopus EF1-a and XFGF receptor were generated and used
as internal loading controls for adult tissues and embryo
stages, respectively.
Hybridization was performed using 20 mg of total RNA
with XLerk and EF1-a or XFGFR radiolabeled probes as
recommended by the manufacturer (HybSpeed RPA [Ambion
Inc.]). Samples were electrophoresed on a 6% polyacrylamide
sequencing gel and autoradiographed on Kodak XAR-5 ®lm
at 7708C.
In situ hybridization
For in situ hybridization, antisense or sense probes were
transcribed using BamHI or HindIII linearized plasmids,
T7 or T3 polymerase, and [35S]UTP as described above.
Embryos from various stages were collected and ®xed in
4% formaldehyde/PBS for 1 ± 2 h at 48C and processed for
paran embedding. The paran blocks were serially
sectioned at 6 mm. In situ hybridization was performed as
previously described by Jones et al. (1995).
Western blot
A total of three embroys from each of the various stages
were homogenized in lysis bu€er (137 m M NaCl, 20 mM
Tris (pH 8.0), 2 mM EDTA, 1% NP-40) containing 2 mM
phenylmethylsulfonyl ¯uoride (PMSF), aprotinin (0.15 U/
ml), 0.5 mM sodium vanadate and 20 mM leupeptin.
Insoluble material was removed by centrifugation at
14 000 g for 10 min at 48C. Lysates were resolved on a
10% SDS ± PAGE gel and transferred to Immobilon-P
membrane (Millipore). The membrane was blocked in 5%
powdered milk in 16TBS containing 0.2% Tween-20
(TBST) for 2 h, washed three times in TBST and
incubated overnight at 48C with antibody against LERK2
(C-18, Santa Cruz Biotechnology) at a dilution of 1 : 500.
The membrane was washed three times with TBST,
incubated 30 min with HRP-coupled secondary antibody
(Boehringer-Mannheim Corp.) at a 1 : 5000 dilution,
washed for 40 min with four changes of TBST and
developed using enhanced chemiluminescence (Amersham).
Embryo injections and animal cap assay
Capped mRNA was synthesized according to the manufacturer (mMessage machine [Ambion Inc.]). At the 2-cell
stage, each blastomere was injected with either 0.125 ng of
noggin capped mRNA or 10 ng of capped mRNA encoding
b-galactosidase. Animal caps were dissected at stage 8.5
and cultured at 238C in 67% Leibovitz's L-15 medium
(Gibco-BRL), 7 mM Tris-HCl (pH 7.5), and 1 mg/ml of
gentamycin, with or without 50 ng/ml of Activin (kindly
provided by Dr HF Kung, NCI, Frederick, MD) or
100 ng/ml of bFGF (Peprotech, Rocky Hill, NJ). The
animal caps were harvested at stage 11 and stage 26 as
determined by uninjected embryos.
RT ± PCR assay
RNA extraction from the animal cap explants was performed
essentially as described by the manufacturer (Triazol; GibcoBRL). To generate ®rst strand cDNA, RNA from two
animal cap equivalents was incubated with reverse transcriptase using oligo dT as primer according to the
manufacturer (Gibco-BRL). One-tenth of the reverse
transcription product was used in each PCR reaction with
[a-32P]dCTP (3000 Ci mmol71; Amersham) added. The
primers N-CAM, EF-1a and Xbra were designed for PCR
as described by Hemmati-Brivanlou and Melton (1994). The
primers for XLerk and Chordin were designed as follows:
XLerk
upstream,
5'-GCAGCACAGTCATAGATCCCAATG-3' and downstream, 5'-CAGGGTTTTGTACAGGACCATTCC-3'; Chordin (Sasai et al., 1994): upstream, 5'TTAGAGAGGAGAGCAACTCGGGCAAT-3' and downstream, 5'-CCTGTTGCGAAACTCTACAGA-3'. The conditions for PCR were as follows: 948C for 1 min, 578C for
1 min and 728C for 2 min for 24 cycles. The ®nal step was
728C for 10 min. Deviation from these conditions were as
follows: EF-1a and XLerk were done for a total of 19 and 34
cycles, respectively. The PCR products were loaded on a 6%
polyacrylamide gel and autoradiographed on Kodak XAR-5
®lm at 7708C for various exposure times.
Note added in proof
After acceptance of this manuscript, the nomenclature for
Eph ligands and receptors was agreed upon. Xlerk will be
referred to as x-ephrin-B1, while Xek will be termed XEphB1.
Acknowledgements
The authors thank Gretchen White and Frank Ruscetti for
critical reading of the manuscript, Nancy Papalopulu and
Robert Grainger for helpful discussions and Ren-He Xu
and Jaebong Kim for PCR primers and reagents. The
GenBank accession number for XLerk is U31427.
References
Bartley TD, Hunt RW, Welcher AA, Boyle WJ, Parker VP,
Lindberg RA, Lu HS, Colombero AM, Elliot RL, Guthrie
BA, Holst PL, Skrine JD, Toso RJ, Zhang M, Fernandez
E, Trail G, Varnum B, Yarden Y, Hunter T and Fox GM.
(1994). Nature, 368, 558 ± 560.
Beckmann MP, Cerretti DP, Baum P, Vanden Bos T, James
L, Farrah T, Kozlosky C, Hollingsworth T, Shilling H,
Maraskovsky E, Fletcher FA, Lhotak V, Pawson T and
Lyman S. (1994). EMBO J., 13, 3757 ± 3762.
2165
XLerk expression during development
TL Jones et al
2166
Bennett BD, Zeigler FC, Gu Q, Fendly B, Goddard AD,
Gillett N and Matthews W. (1995). Proc. Natl. Acad. Sci.
USA, 92, 1866 ± 1870.
Bergemann AD, Cheng H-J, Brambilla R, Klein R and
Flanagan JG. (1995). Mol. Cell. Biol., 15, 4921 ± 4929.
Bohme B, Vanden Bos T, Cerretti DP, Park LS, Holtrich U,
Rubsamen-Waigmann H and Strebhardt K. (1996). J.
Biol. Chem., 271, 24747 ± 24752.
Bouillet P, Oulad-Abdelghani M, Vicaire S, Garnier JM,
Schuhbaur B, Dolle P and Chambon P. (1995). Dev. Biol.,
170, 420 ± 433.
Brambilla R, Schnapp A, Casagranda F, Labrador JP,
Bergemann AD, Flanagan JG, Pasquale EB and Klein R.
(1995). EMBO J., 14, 3116 ± 3126.
Brandli AW and Kirschner MW. (1995). Dev. Dynamics, 203,
119 ± 140.
Carpenter MK, Shilling H, Vanden Bos T, Beckmann MP,
Cerretti DP, Kott JN, Westrum LE, Davison BL and
Fletcher FA. (1995). J. Neuro. Res., 42, 199 ± 206.
Cheng H and Flanagan J. (1994). Cell, 79, 157 ± 168.
Cheng H-J, Nakamoto M, Bergemann AD and Flanagan JG.
(1995). Cell, 82, 371 ± 381.
Davis S, Gale NW, Aldrich TH, Maisonpierre PC, Lhotak V,
Pawson T, Goldfarb M and Yancopoulos GD. (1994).
Science, 266, 816 ± 819.
Doniach T. (1995). Cell, 83, 1067 ± 1070.
Drescher U, Kremoser C, Handwerker C, Loschinger J,
Noda M and Bonhoe€er F. (1995). Cell, 82, 359 ± 370.
Easty DJ, Guthrie BA, Maung K, Farr CJ, Lindberg RA,
Toso RJ, Herlyn M and Bennett DC. (1995). Cancer Res.,
55, 2528 ± 2532.
Fletcher FA, Carpenter MK, Shilling H, Baum P, Ziegler SF,
Gimpel S, Hollingsworth T, Vanden Bos T, James L,
Hjerrild K, Davison B, Lyman SD and Beckmann MP.
(1994). Oncogene, 9, 3241 ± 3247.
Gale NW, Holland SJ, Valenzuela DM, Flenniken A, Pan L,
Ryan TE, Henkemeyer M, Strebhardt K, Hirai H,
Wilkinson DG, Pawson T, Davis S and Yancopoulos
GD. (1996a). Neuron, 17, 9 ± 19.
Gale NW, Flenniken A, Compton DC, Jenkins N, Copeland
NG, Gilbert DJ, Davis S, Wilkinson DG and Yancopoulos GD. (1996b). Oncogene, 13, 1343 ± 1352.
Gao P-P, Zhang J-H, Yokoyama M, Racey B, Dreyfus CF,
Black IB and Zhou R. (1996). Proc. Natl. Acad. Science,
93, 11161 ± 11166.
Hemmati-Brivanlou A, Mann RW and Harland RM. (1992).
Neuron, 9, 417 ± 428.
Hemmati-Brivanlou A and Melton DA. (1994). Cell, 77,
273 ± 281.
Henkemeyer M, Orioli D, Henderson JT, Saxton TM, Roder
J, Pawson T and Klein R. (1996). Cell, 86, 35 ± 46.
Hens MD and Desimone DW. (1995). Dev. Biol., 170, 274 ±
288.
Hirai H, Maru Y, Hagiwara K, Nishida J and Takaku F.
(1987). Science, 238, 1717 ± 1720.
Holash JA and Pasquale EB. (1995). Dev. Biol., 172, 683 ±
693.
Holland PWH and Hogan BLM. (1988). Development, 102,
150 ± 169.
Holland SJ, Gale NW, Mbamalu G, Yancopoulos G,
Henkemeyer M and Pawson T. (1996). Nature, 383,
722 ± 725.
Jones TL, Karavanova I, Maeno M, Ong RC, Kung H-F and
Daar IO. (1995). Oncogene, 10, 1111 ± 1117.
Kenny D, Bronner-Fraser M and Marcelle C. (1995). Dev.
Biol., 172, 708 ± 716.
Kessler DS and Melton D. (1994). Science, 266, 596 ± 604.
Kozak M. (1992). Ann. Rev. Cell Biol., 8, 197 ± 225.
Kozlosky CJ, Maraskovsky E, McGrew JT, Vanden Bos T,
Teppe M, Lyman SD, Srinivasan S, Fletcher FA, Gayle
RB III, Cerretti DP and Beckmann MP. (1995). Oncogene,
10, 299 ± 306.
Krieg P, Varnum S, Wormington M and Melton DA. (1989).
Dev. Biol., 133, 93 ± 100.
Krotoski D and Bronner-Fraser M. (1990). J. Exp. Zool.,
253, 139 ± 150.
Lamb TM, Knecht AK, Smith WC, Yancopolous GD and
Harland RM. (1993). Science, 262, 713 ± 718.
Lefcort F, Venstrom K, McDonald JA and Reichardt LF.
(1992). Development, 116, 767 ± 782.
Maru Y, Hirai H, Yoshida MC and Takaku F. (1988). Mol.
Cell. Biol., 8, 3770 ± 3788.
Musci TJ, Amaya E and Kirschner MW. (1990). Proc. Natl.
Acad. Sci. USA, 87, 8365 ± 8369.
Nakamoto M, Cheng H-J, Friedman GC, McLaughlin T,
Hansen MJ, Yoon CH, O'Leary D and Flanagan JG.
(1996). Cell, 86, 755 ± 766.
Newport T and Kirschner M. (1982). Cell, 30, 675 ± 686.
Nieuwkoop PD and Faber J. (1976). Normal Table of
Xenopus laevis. Daudin: North Holland, Amsterdam,
pp. 252.
Pandey A, Lindberg RA and Dixit VM. (1995). Current Biol.,
5, 986 ± 989.
Ransom DG, Hens MD and Desimone DW. (1993). Dev.
Biol., 160, 265 ± 275.
Sasai Y, Steinbiesser H, Geissert D, Gont LK and
DeRobertis EM. (1994). Cell, 79, 779 ± 790.
Shao H, Lou L, Pandey A, Pasquale E and Dixit VM. (1994).
J. Biol. Chem., 269, 26606 ± 26609.
Shao H, Lou L, Pandey A, Verderame MF, Siever DA and
Dixit V. (1995). J. Biol. Chem., 270, 3467 ± 3470.
Simonneau L, Broders F and Thiery J-P. (1992). Dev. Dyn.,
194, 247 ± 260.
Smith WC and Harland RM. (1991). Cell, 67, 753 ± 765.
Smith WC and Harland RM. (1992). Cell, 70, 829 ± 840.
Tessier-Lavigne M and Goodman CS. (1996). Science, 274,
1123 ± 1131.
Tomaselli KJ and Reichardt LF. (1989). The assembly of the
nervous system. Landmesser (ed) AR Liss, New York, pp.
81 ± 108.
Tuzi NL and Gullick WJ. (1994). Br. J. Cancer, 69, 417 ± 421.
Weinstein DC, Rahman SM, Ruiz JC and HemmatiBrivanlou A. (1996). Mech. Dev., 57, 133 ± 144.
Winning RS and Sargent TD. (1994). Mech. Dev., 46, 219 ±
229.
Winning RS, Scales JB and Sargent TD. (1996). Dev. Biol.,
179, 309 ± 319.
Winslow JW, Moran P, Valverde J, Shih A, Yuan JQ, Wong
SC, Tsai SP, Goddard A, Henzel WJ, Hefti F, Beck KD
and Carass W. (1995). Neuron, 14, 973 ± 981.
Xu Q, Alldus G, Holder N and Wilkinson DG. (1995).
Development, 121, 4005 ± 4016.