Development 122, 2367-2374 (1996)
Printed in Great Britain © The Company of Biologists Limited 1996
Late emigrating neural crest cells migrate specifically to the exit points of
cranial branchiomotor nerves
Christiane Niederländer and Andrew Lumsden
Department of Developmental Neurobiology, UMDS Guy’s Hospital, London SE1 9RT, UK
Morphological segmentation of the avian hindbrain into
rhombomeres is also reflected by the emergent organisation of branchiomotor nerves. In each case, the motor
neurons of these nerves lie in two adjacent rhombomeres
(e.g. of the Vth nerve in r2 and r3, VIIth in r4 and r5 etc.),
and their outgrowing axons emerge into the periphery
through defined exit points in rhombomeres r2, r4 and r6,
respectively. Sensory axons of the cranial ganglia also enter
the neuroepithelium at the same points. Motor axon
outgrowth through experimentally rotated rhombomeres
has suggested that a chemoattractive mechanism, involving
the exit points, may form a component of their guidance.
Yet so far, nothing is known about the establishment of the
exit points or the identity of the cells that form them. In
this study, we describe a group of late emigrating cranial
neural crest cells which populate specifically the prospective exit points. Using chimaeras in which premigratory
chick neural crest had been replaced orthotopically by
quail cells, a population of neural crest was found to leave
the cranial neural tube from about stage 10+ onwards and
to migrate directly to the prospective exit points. These
cells define the exit points by stage 12+, long before either
motor or sensory axons have grown through them. The
entire neural crest population of exit point cells expresses
the recently described cell adhesion molecule c-cad7.
Further, heterotopic grafting experiments show that
midbrain and spinal cord crest, grafted at late stages in
place of r4 crest, share the same migratory behaviour to
the facial nerve exit points and express the same markers
as cells contributed by the native r4 crest. It was not
possible to generate new exit points in odd numbered
rhombomeres simply by experimentally increasing their
(normally insignificant) amount of crest production.
Initiation of the exit point region probably lies, therefore,
in the neuroepithelium.
neurons that make up the trigeminal nerve lie in rhombomere
(r) 2 and r3, while the facial neurons are located in r4 and r5
and the glossopharyngeal neurons in r6 and r7. Outgrowing
branchiomotor axons grow dorsally away from the floor plate,
remain for some distance within the neuroepithelium and exit
into the periphery through defined points in the alar plate of
r2, r4 and r6. Hence, axons from odd numbered rhombomeres
must turn rostrally to reach their exit points in the adjacent
rhombomere. In growing towards and through these exit
points, motor axons are grouped to form bundles which emerge
into the periphery. The branchiomotor nerve exit points are
also the entry points for ingrowing sensory axons.
It has been shown that the floor plate acts as a chemorepellent for outgrowing branchiomotor axons of the hindbrain
(Guthrie and Pini, 1995). Floor plate chemorepulsion has also
been shown to act on trochlear motor axons and in this case
could be attributed to netrin-1 (Colamarino and TessierLavigne, 1995). Other studies suggest that, once in the
periphery, outgrowing motor axons have their next intermediate target in the sensory ganglia, lying adjacent to the rhombomere that contains their exit point (Moody and Heaton,
1983c; Heaton and Wayne, 1986). Thus, it appears from a
number of studies that the outgrowth of motor axons in the head
and in the spinal cord, from their cell bodies to their final targets
The mechanisms that guide outgrowing motor axons from their
origin in the neuroepithelium into the periphery and then to
their final target are only partly understood. Axons of spinal
motor neurons of the trunk, for example, grow away from the
floor plate and exit the neural tube in a continuous band along
its anteroposterior axis, but at a specific dorsoventral level.
Once in the periphery, axons form nerve bundles due to the
instructive action of the adjacent segmented mesoderm: a
chemorepulsive signal in the posterior halves of the somites
directs spinal motor axons to grow toward the anterior half
somites and thus to form the spinal motor nerve bundles
(Keynes and Stern, 1984). Furthermore, it has been shown
recently that the determination of motor neurons in the trunk
to grow towards specific groups of muscles is correlated with
the expression of combinations of LIM-homeodomain transcription factors in subgroups of spinal motor neurons
(Tsuchida et al., 1994).
In the head, the segmentation of the neuroepithelium of the
hindbrain (Vaage, 1969; Lumsden, 1990) coordinates the
development of the cranial motor nerves. In particular the branchiomotor nerves reflect this rhombomeric organisation of the
hindbrain (Lumsden and Keynes, 1989). Cell bodies of motor
Key words: rhombomeres, neural crest, branchiomotor neurons,
quail-chick chimaeras, fate map, Krox-20, c-cadherin7
2368 C. Niederländer and A. Lumsden
in the periphery, involves their stepwise guidance by chemorepulsive and chemoattractive landmarks along their way.
The pattern of branchiomotor axon outgrowth to the nerve
exit points, which results in nerve formation, suggests that the
exit points represent one of the chemoattractive intermediate
targets guiding motor axons. Cell surface molecules in the neuroepithelium do not seem to be required to guide motor axons
to the exit points. This has been shown by experiments in
which odd-numbered rhombomeres were inverted in their
anteroposterior orientation: the outgrowing motor axons still
turn towards their anterior lying exit points (Guthrie and
To date, the morphology and cellular composition of the exit
point region is unknown and exit points only become distinguishable once sensory and motor axons have traversed
through them. It is likely that the establishment and maintenance of the exit point region involves a number of different
cells. The expression of genes which mark distinct cell groups
at the exit points suggests the existence of specified cells at
We have been interested in the role of neural crest cells in
establishing exit points for motor axons, focusing on the branchiomotor neurons of the hindbrain. Previous studies have
shown that the amount of neural crest produced along the
dorsal aspect of the neural tube varies greatly between different
rhombomeres (Lumsden et al., 1991; Sechrist et al., 1993).
Rhombomeres 2, 4 and 6 which contain the exit points,
produce (together with r1) the majority of neural crest cells in
the hindbrain, whereas r3 and r5 are massively depleted of crest
cells by apoptosis (Graham et al., 1993, 1994). Neural crest in
the hindbrain migrates in three streams lying adjacent r2, r4
and r6, whose prominence relates, at least in part, to the crest
depletion of r3 and r5, and along a number of dorsoventral
pathways, one of which leads crest cells along the side of the
basal lamina of the neural tube (Detwiler, 1937; Noden, 1988).
The predominant neural crest production of the rhombomeres
which contain the branchial nerve exit points and the pathway
of some of the emigrating crest cells along the side of the
neural tube prompted us to examine the role of neural crest
cells in determining the exit point region.
This study reveals a subpopulation of neural crest cells, emigrating from the neural tube of the hindbrain only shortly
before crest emigration ceases, which targets predominantly
the prospective exit point region. We have analysed the time
course of emigration in this cell population and some aspects
of its action at the target region using quail-chick chimaeras.
The results show that neural crest cells, leaving the hindbrain
in r2 from stage 10+ onward and in r4 from stage 11 onward,
migrate specifically to the prospective branchiomotor exit
point regions, which they reach at or before stage 12+. These
cells therefore arrive at exit points considerably in advance of
either sensory or motor axons, which do not reach the exit
points until at least stage 15 (Newgreen and Erickson, 1986;
Moody and Heaton 1983a; Covell and Noden, 1989). In a
parallel approach, DiI iontophoresis was used to label small
groups of premigratory crest cells which were also found to
migrate to the exit points. These crest cells express c-cad7 and
Krox-20 (r2 and r4 only). The cells do not become part of the
sensory ganglia, rather they attach closely to the surface of the
neural tube, frequently bulging into the neuroepithelium
through perforations in the basal lamina.
Heterotopic grafts of midbrain or spinal cord neural crest
into the hindbrain region show that crest cells from all axial
levels can contribute to the exit points. These cells appear to
be targeted to this region by a property of the neuroepithelium,
which may reside in its basal lamina. The signal retaining crest
cells at the exit point region seems to be universally recognised
by neural crest cells, making it likely that the time point of emigration from the neural tube or the time of maturation of the
exit point targets at the neuroepithelium surface play a role in
determining whether neural crest cells accumulate at the exit
MATERIALS AND METHODS
Neural crest transplantation experiments
Fertile hens’ eggs were obtained from a mixed flock (Poyndon Farm,
Enfield) and incubated in a forced draft incubator at 38°C to stage 1011. Quail eggs were obtained from Rosedean Farm, Cambridgeshire
and incubated to the same stages. Hens’ eggs were prepared for
grafting as described before (Simon et al., 1995). A narrow region of
the dorsal neural tube at the hindbrain level was removed using
needles flame-sharpened from 100 µm pure tungsten wire. This region
was replaced by the equivalent dorsal piece of neural tube from a stage
matched quail embryo. For fate mapping, the grafted region was
orthotopic. Further experiments used heterotopic grafts from the
spinal cord and midbrain, transplanted into r4, or grafts of r4 transplanted into the r3 position. Eggs were resealed with electrical tape
and incubated for a further 24-48 hours.
Hens’ eggs were incubated to stage 10-11 and prepared as for grafting.
They were then placed under an epifluorescence microscope and
viewed at 20× magnification using an ultra long working distance
objective. An aluminosilicate glass electrode micropipette was tipfilled with DiI and backfilled with 1 M KCl solution. The electrode
was held in a micromanipulator and its tip positioned over the dorsal
neural tube of the chick embryo. DiI was released from the electrode
for a few seconds by closing an electric circuit through the electrode
and the egg albumen with a 9 Volt battery. Extracellular release of
dye was thus spread onto the membrane of a small group of cells. The
success of the injection was monitored by a brief exposure to fluorescence. Eggs were resealed and incubated for further 24 hours. For
sectioning, embryos were embedded in 20% gelatine and fixed in 4%
paraformaldehyde; they were then cut at 50 µm on a vibratome and
viewed on a BioRad MRC 600 confocal microscope.
Fixed embryos were stained with the QCPN antibody (developed by
Carlson) using an indirect immunoperoxidase method (Lumsden and
Keynes, 1989; Guthrie and Lumsden, 1992). Stained embryos were
viewed as whole mounts or embedded for wax, cryo or ultrathin sectioning. Light microscopic sections were stained with toluidine blue.
The QCPN monoclonal antibody was obtained from the Developmental Studies Hybridoma Bank of the Johns Hopkins University
School of Medicine, Baltimore.
Embryos were fixed for 4 hours at 4°C in 2.5% glutaraldehyde in 0.1
M phosphate buffer (pH 7.3). After washing overnight in phosphate
buffer, embryos were osmicated in 1% aqueous osmium tetroxide and
dehydrated in an ascending methanol series. After embedding in Epon
(TAAB), semithin and ultrathin sections were cut. Some sections were
then stained again using lead citrate and uranyl acetate. Semithin
sections (1 µm) were stained with toluidine blue and viewed under
the light microscope. Ultrathin sections were viewed at 75 kV in a
Hitachi H700 transmission electron microscope.
Cranial nerve exit point cells 2369
Fig. 1. In situ hybridisation with probes that are expressed at the
branchiomotor exit point region. (A) Dorsal view of a stage (st.) 16
whole-mount embryo probed with c-cad7. The branchiomotor nerve
exit points and the motor column inside the neural tube are labelled.
(B) Transverse section through facial exit point region of a st. 16
embryo probed with c-cad7. (C) Coronal section through st. 15
hindbrain of Krox-20 probed embryo. Anterior is at the top. The
staining adjacent to r4, the unstained rhombomere, marks the facial
exit point. (picture courtesy of Dr Isobel Heyman and Dr Cairine
Logan). (D) Section through the facial exit point of a Krox-20 in situ
hybridised st. 17 embryo. Scale bar: A, 100 µm; B-D, 50 µm.
Fig. 2. Diagram of the different grafting procedures. (A) Isotopic and
isochronic dorsal tube grafts consisting of either r2-r4 or r4-r6 and
iontophoretic DiI injection into the same region. (B) Heterotopic
grafts replacing chick r3 dorsal tube with quail r4 dorsal tube.
(C) Heterotopic grafts replacing r4 dorsal tube with midbrain or
spinal cord premigratory crest.
Probes for in situ hybridisation
The c-cad7 probe was a kind gift from Dr S. Nakagawa and Dr
Masatoshi Takeichi. The region used for the in situ hybridisation
studies was the 450 bp PCR-fragment previously described
(Nakagawa and Takeichi, 1995). The Krox-20 probe was a 600 bp
Pst1-Apa1 fragment from the mouse gene, kindly provided by Dr
David Wilkinson (Wilkinson, 1992).
Whole-mount in situ hybridisation
Digoxigenin-labelled riboprobes (Boehringer, UK) were used to
detect gene transcripts. For Krox-20 in situs the protocol used was
that of Dr David Wilkinson (Wilkinson, 1992) with the omission of
the proteinase K step. For c-cad7, the protocol used was that of
Domingos Henrique and David Ish-Horowicz (Henrique et al., 1995)
with the omission of the proteinase K step. The protocol is as follows.
Embryos were fixed in 4% paraformaldehyde, washed twice in PBT
(PBS with 0.1% Tween-20), dehydrated by washing in 50%
methanol/PBT and 100% methanol and then rehydrated by passing
them through a methanol/PBT series. Embryos were then washed
twice with in PBT, once in hybridisation buffer/PBT (1:1) and then
preincubated in hybridisation buffer for 1 hour at 70°C. Embryos were
transferred into hybridisation buffer containing 1 µg/ml of DIGlabelled riboprobe. The hybridisation buffer consisted of 50%
formamide, 1.3× SSC, 5 mM EDTA, 50 µg/ml yeast RNA, 0.2%
Tween-20, 0.5% CHAPS and 100 µg/ml heparin. After hybridisation,
the embryos were washed twice at 70°C in hybridisation buffer
followed by one wash at the same temperature in MABT/hybridisation buffer (1:1). MABT consisted of 100 mM maleic acid, 150 mM
NaCl and 1% Tween-20, final pH 7.5. The embryos were then washed
in MABT followed by washing in MABT containing 2% Boehringer
Blocking reagent (Boehringer, UK) and 20% goat serum. The
embryos were incubated overnight in the same solution containing
alkaline phosphatase coupled anti-DIG antibody (Boehringer, UK) at
a 1:2000 dilution. After incubation with the antibody, the embryos
were washed in MABT over the next day. At the end of the day the
Fig. 3. Labelling of late emigrating neural crest in the hindbrain in
quail-chick chimaeras stained with the QCPN antibody, in wholemount views (A,B) and viewed confocally in transverse sections
after iontophoretic DiI injection (C,D). (A) Dorsal view of hindbrain
of a st. 14 embryo with r2-r4 grafted at st. 11−. Quail crest cells can
be seen at the trigeminal and facial exit points. (B) Side view of a st.
14+ embryo with r4-6 grafted at st. 11+; anterior is to the right,
dorsal is up. Quail crest cells are at the facial, glossopharyngeal and
vagal exit points. ot, otocyst. (C) Trigeminal exit points in st. 15
embryo after injection into r2 at st. 11. (D) Facial exit point in st. 15
embryo after injection at st. 11. Scale bar: A-C, 100 µm; D, 50 µm.
embryos were washed in NTMT (100 mM Tris-HCl pH 9.5, 100 mM
NaCl, 50 mM MgCl2, 1% Tween-20) twice for 10 minutes. The colour
reaction was developed in NTMT containing 5-bromo-4-chloro-3indolyl phosphate. Quail-chick chimaeras were then immunohistochemically stained using the QCPN antibody (see above). For sectioning, embryos were embedded in OCT compound, frozen and cut
at 13 µm in a cryostat.
Specific gene expression marks the exit point sites
Motor nerve exit points become readily distinguishable once
2370 C. Niederländer and A. Lumsden
axons have passed into and out of the neural tube. The
expression of certain genes seems to relate to future nerve exit
The chick homologue of cadherin-7, c-cad7, a calciumdependent cell adhesion molecule, has been cloned recently
(Nakagawa and Takeichi, 1995). The expression pattern is
associated with cranial nerve development, and Fig. 1A,B
shows its expression at prospective branchiomotor nerve exit
points. C-cad7 is first expressed by neural crest cells leaving
the midbrain at stage 10. Subsequently, between stage 10-11,
neural crest cells of the hindbrain start expressing c-cad7 at the
time of their emigration out of the neural tube. The c-cad7positive signal initially can be seen in a short stream away from
the hindbrain but between stage 12 and 13 the signal becomes
confined to the exit points of developing cranial nerves. C-cad7
is also expressed in the trunk by crest cells at the dorsal and
ventral root. Apart from neural crest, c-cad7 expression is
found throughout the motor column in the neural tube
(Nakagawa and Takeichi, 1995).
Krox-20 is a transcription factor of the zinc finger family
whose transcripts are restricted to r3 and r5 of the hindbrain
(Wilkinson et al., 1989). In addition, there is a distinct domain
of Krox-20 expression in the neural crest at sites which have
been identified as ‘boundary caps’ adjacent to r2 and r4,
described both in mouse (Wilkinson et al., 1989) and in chick
(C. Logan, unpublished data). In chick, this expression can be
seen from about stage 14 onward; transverse sections of wholemount in situs probed for Krox-20 transcripts (Fig. 1C,D) show
this ‘boundary cap’ at the same site as c-cad7 expression in the
These expression patterns suggest that the prospective exit
point region is populated by specialised cells before the arrival
of motor axons.
A late emigrating population of cranial neural crest
cells specifically targets nerve exit points
We were interested in determining the role of neural crest in
exit point formation. Because a universal and unique marker
for neural crest cells has not been described, we marked neural
crest cells by making quail-chick chimaeras. By replacing a
very narrow region of chick dorsal neural tube with the same
region of isochronic quail embryos (Fig. 2A), labelled crest
cells were produced whilst ensuring that as little as possible of
the remaining neural tube cells were labelled. Quail cells were
visualised immunohistochemically using an antibody (QCPN)
directed against a perinuclear epitope in quail cells.
At around stage 10+ dorsal neural tube from r2-4 or r4-6,
that is premigratory neural crest, was replaced by the same
region of quail embryos. Because crest emigration in the
hindbrain starts at around stage 9 and lasts for about 24 hours
(Newgreen and Erickson, 1986; Lumsden et al., 1991) this
procedure led to the labelling of neural crest cells that emigrate
from the neural tube comparatively late. These crest cells
ended up predominantly at the branchiomotor nerve exit points
(Fig. 3A,B). Crest cells from r2-4 were found at the trigeminal (r2) and facial (r4) exit points (Fig. 3A) and crest cells from
r4-6 migrated to the facial, glossopharyngeal (r6) and eventually the vagal (r7) nerve exit points (Fig. 3B).
To confirm these results, premigratory neural crest cells
were iontophoretically labelled with DiI (Fig. 2A). As in the
quail-chick chimaeras, we found neural crest cells migrating
specifically to the branchiomotor nerve exit points (Fig. 3C,D).
The observed migration behaviour in the quail-chick chimaeras
is therefore not due to the changed environment or the grafting
procedure: the same migration pattern to the exit points is
found in embryos where neural crest was labelled using the
much less invasive method of DiI iontophoresis.
Crest cells could be seen at the exit points as early as stage
12+, a stage at which neither sensory nor motor axons have yet
grown through the exit points (Fig. 4A shows a section through
a stage 13− embryo). This is, however, only the earliest time
point included in this study, since it is difficult to combine the
grafting procedure with a shorter incubation time which would
not allow sufficient healing of the graft. The fact that the exit
point region is already populated by many crest cells suggests
that the first time these cells can be found at prospective exit
points is even earlier. This view is also supported by the
expression pattern of c-cad7.
In embryos ranging from stage 16 to 21, the neural crest cells
stayed at the exit points, remaining separate from the proximal
part of the ganglia (Fig. 4B). At these later stages the crest cells
were seen to bulge deeply into the side of the neural tube (Fig.
4C,D), suggesting that the basal lamina of the neural tube is
no longer present there.
Some of the quail-chick chimaeras were embedded for high
resolution light microscopy and electron microscopy, to
analyse in greater detail the interaction of crest cells with the
basal surface of the neural tube. The quail cells were seen to
come close to the neural tube and at the electron microscopic
level to bulge into it (Fig. 5). The basal lamina of the neural
tube looked ruffled at the exit point region in contrast to
elsewhere. Fig. 5C,D shows labelled quail-cells in non-counterstained sections coming close to the neural tube. On a
parallel section to that shown in Fig. 5D, uranyl acetate and
lead citrate staining was applied to visualise the structure of
the basal lamina at the exit points (Fig. 5E,F,G). We found that
at the site where the labelled cell appears to bulge into the
neural tube the typical double layered structure of the basal
lamina is absent (compare Fig. 5F with 5G) and crest cell and
neuroepithelium are in direct contact.
The specificity of the prospective exit points is
determined before the arrival of neural crest cells
To determine whether the generation of exit points is
dependent on the amount of crest cells generated at a given
rhombomeric level, or on the specific anteroposterior identity
of the crest cells, heterotopic grafts were made. Quail dorsal
r4 was transplanted in place of dorsal r3, thereby creating a
continually high emigration level of crest cells from r2, r3/r4′
and r4 (Fig. 2B); r3 usually gives rise to a greatly reduced
number of neural crest cells (Graham et al., 1993; Sechrist et
al., 1993; Birgbauer and Fraser, 1994). In these embryos, the
r4 quail cells, coming from the axial level of r3 are seen to
migrate both anterior and posterior to the trigeminal and facial
nerve exit points, respectively, showing no specificity for their
native exit point, the facial, but also not generating an additional exit point in r3, the level of their emigration from the
neural tube (Fig. 6A). This suggests that the initial cues for exit
point formation lie in the neuroepithelium itself.
Heterotopic grafting experiments were also performed
which included premigratory neural crest of midbrain or spinal
cord level grafted in place of dorsal r4 (Fig. 2C). These neural
Cranial nerve exit point cells 2371
crest cells do not normally form exit points for mixed
sensory/motor nerves, since the latter form only at hindbrain
level. However, the quail-chick chimaeras resulting from these
experiments are indistinguishable from chimaeras with orthotopic grafts (Fig. 6B,C). This further strengthens the possibility that the initial clues for exit point formation lie within the
neuroepithelium, and are independent of both the amount of
available crest and its specific anteroposterior level of origin.
Gene expression and the late neural crest cells at
the exit points are congruent
Our results suggested that the late emigrating neural crest cells
we describe are the same cells that express Krox-20 and c-cad7
at the exit points. We have confirmed this by performing
double labelling using in situ hybridisation on quail-chick
chimaeras. We find the expression site of c-cad7 and the target
site of late emigrating neural crest cells are coextensive (Fig.
7A). The expression pattern of Krox-20 makes it very likely
that late emigrating neural crest cells at the exit point of the
trigeminal and facial nerves also express Krox-20. Moreover,
double labelling with c-cad7 in the heterotopically grafted
embryos where midbrain or spinal cord crest was grafted into
r4, shows that these cells express the appropriate marker for
their new site (Fig. 7B).
This study is the first to describe a specific cell population that
delineates the future site of the branchiomotor nerve exit point
prior to its penetration by axons from inside or outside the
The differentiation of motor neurons and the subsequent
outgrowth of motor axons has been reported to occur from
stage 13 onwards. The bulk of early motor axons passes
through the exit points around stage 15, even though the
earliest axons might reach the exit points as early as around
stage 13-14 (Noden, 1980; Moody and Heaton, 1983a,b,c;
Covell and Noden, 1989; Guthrie and Lumsden, 1992). The
population of late emigrating crest cells described here
however, were found at the prospective exit point at least from
stage 12+ onwards and very likely even earlier, judged by the
amount of crest cells found at the exit points at the earliest
stage examined. This subpopulation of crest cells is the earliest
marker of the region where the exit point will form.
In situ hybridisation showed that this cell population is also
distinguished by its specific expression of c-cad7 and Krox-20.
(Nakagawa and Takeichi, 1995; Wilkinson et al., 1989). Krox20 appears to be expressed only by cells at the trigeminal and
facial exit points. C-cad7 mirrors the behaviour of the exit
point crest cell population even more closely. C-cad7 expressing cells can be seen to emigrate from the neural tube of the
hindbrain between stage 10-11. This matches the results
obtained in our grafting experiments.
The neural crest cells adhere very closely to the neural tube
and remain separate from the nearby proximal pole of the
cranial sensory ganglion. At later stages, the cells are frequently seen to bulge into the neural tube and electron microscopic analysis has confirmed that the basal lamina of the
neural tube is absent at these sites.
Although this population of crest cells is the earliest marker
of the region where the hindbrain exit point will form, the
demarcation of presumptive motor axon exit sites in the spinal
neural tube by migratory cells has been observed before. Lunn
et al. (1987) describe a population of ventral neural tube cells
that appear to breach the basal lamina from the inside of the
neural tube and coalesce at the future ventral root before the
arrival of motor axons.
The neural crest cells we describe here are candidates for the
secretion of molecules to guide motor axons over their intraepithelial course and maybe also to attract incoming sensory
axons to the exit point (Guthrie and Lumsden, 1992). Removal
of the exit point cells and subsequent analysis of motor axon
pathfinding is an obvious test of this possible role. Although
the regenerative capacity of neural crest at the hindbrain level
(Scherson et al., 1993) rules out simple ablation experiments,
we are undertaking experiments that combine extensive
ablation of the neural crest together with the prevention of
contact between the residual neural tube and overlying surface
ectoderm, contact with which appears to be required for neural
crest regeneration (Liem et al., 1995).
Other possible candidates for the guidance of outgrowing
branchiomotor axons include the specialised cells of the neuroepithelium which presumably reside at the site of the exit
point and the ingrowing sensory axons from the cranial
ganglion that use the motor neuron exit point as their entry
point into the CNS. Equally, motor axons reaching the exit
point before sensory axons could serve as guiding targets for
ingrowing sensory axons. The first sensory and motor axons
appear to enter and exit the brain at around the same time,
between stage 13-14 in the chick; however, the timing of the
ingrowth of sensory axons versus outgrowth of motor axons is
not known with precision. Studies aimed at determining
whether the sensory or motor axons reach the exit points first
have been inconclusive (Noden, 1980; Moody and Heaton,
1983a-d; Covell and Noden, 1989).
Late emigrating neural crest cells are the first manifestation
of the prospective exit point region, but heterotopic grafts show
that these crest cells themselves do not determinate the site of
the exit point. The reduced amount of crest cells and the
absence of exit points in the odd numbered rhombomeres
initially raised the possibility that the high level of crest cell
generation in even numbered rhombomeres would elicit exit
point formation in the same rhombomere, since the migration
pathway of some neural crest cells extends ventrally alongside
the neural tube. By transplanting dorsal r4 in place of dorsal
r3, however, the amount of neural crest cell generation was
experimentally raised at the r3 level and yet these embryos still
failed to form exit points in r3. Moreover, r4 cells that at their
normal site of origin are seen to migrate to the facial exit point
in that rhombomere, migrate to both the trigeminal and facial
exit points when grafted to r3, showing no preference for their
The r4 into r3 grafting experiment also points out another
feature of the exit point region: whatever signal from the neural
tube prompts late emigrating neural crest cells to migrate there,
it is recognised by crest cells from different axial levels in the
hindbrain and it is the same for all exit points. Neural crest
cells seem to target the closest exit point to their migration
path. Surprisingly, crest cells originally from axial levels that
do not have mixed motor-sensory nerves, i.e. crest cells from
the midbrain or spinal cord levels, are targeted to the branchial
2372 C. Niederländer and A. Lumsden
Fig. 4. Transverse sections through the
trigeminal exit point at different stages.
(A) St. 13−. (B) St. 17, the exit point cells
do not spread out into the proximal part of
the ganglion. g, ganglion. (C) St. 19; neural
crest cells are closely attached to the neural
tube, giving the impression of bulging into
it. (D) Higher power view of the boxed
area in C. Scale bar: A-C, 50 µm; D, 5 µm.
Fig. 5. Transverse sections through quail-chick chimaeras (A,C) 1 mm semithin (A) and ultrathin (C) section through facial exit points of st. 16
chimaeras grafted at st. 10+. (B,D) 1 mm semithin (B) and ultrathin section (D) through trigeminal exit point of st. 14 embryo grafted at st. 11−.
(C-G) Electron micrographs. E-F are from a parallel section to D, but additionally stained with lead citrate and uranyl acetate to enhance the
contrast. The cells seen in D is marked with an arrow in E. (F,G) High power views of the cell’s membrane adjacent to the neural tube (sites
marked with arrowheads in E). In F the basal lamina on the side of the neural tube is clearly visible, whereas further along (G) where the
labelled cell seems to indent into the neural tube, the basal lamina is absent. The interface between the two cells in G is directly under the basal
lamina seen in F. nt, neural tube. Scale bar: A, 100 µm; B, 10 µm; C, D, 5 µm; E, 2.5 µm.
Cranial nerve exit point cells 2373
Fig. 6. Quail-chick chimaeras with heterotopic dorsal neural tube
transplants. (A) Dorsal whole-mount view of a st. 16 embryo where
the r3 tissue had been replaced by quail r4 tissue, the transplanted r4
crest cells have migrated anterior to the trigeminal exit points and
posterior to their native facial exit points. (B) Dorsal whole-mount
view of a st. 14 embryo where r4 had been replaced by midbrain
tissue. (C) Transverse section through a st. 14 embryo where r4
dorsal tube had been replaced by spinal cord premigratory crest.
Scale bar: A,B, 200 µm; C, 100 µm.
nerve exit points exactly as normal when heterotopically
grafted. These ectopic cells also express the appropriate
markers for their new environment. It is therefore likely that
proximity to the prospective exit points and the time of emigration from the neural tube are the factors which determine
the migration target of these cells.
Our experiments thus show that the determinant of the exit
point site must lie in the neural epithelium itself. Anteroposterior patterning, conferred for example by Hox genes and
Krox-20 expression (Wilkinson, et al., 1989; Krumlauf,
1994), establishing odd and even numbered rhombomeres,
together with dorsoventral patterning, conferred for example
by the expression of Pax genes (Goulding et al., 1993; Simon
et al., 1995; Mansouri et al., 1994 for review) could result in
specific surface properties of the basal lamina of the neural
tube, retaining crest cells at the site of the exit points. Even
numbered rhombomeres, in which the dorsoventral orientation has been inverted, generate their exit point in the former
basal plate, now in alar plate position, at the correct dorsoventral location (Simon et al., 1995). In this respect, it is interesting to note that the exit point neural crest population
expresses the cell adhesion molecule c-cad7. Cell adhesion
involving cadherins has mainly been described as homophilic
and c-cad7 expression could simply lead to the cohesion or
coalescing of exit point cells, since there is no c-cad7
expression at the corresponding sites in the neural tube.
However, heterophilic binding amongst different cadherins
and probably also binding of cadherins to other cells surface
proteins has recently been shown to be biologically important
(Redies and Takeichi, 1993; Williams et al., 1994). Possibly
an as yet unknown member of the cadherin family is
expressed in the neural tube at the exit point site that acts in
retaining c-cad7 expressing neural crest cells there. Also the
expression of c-cad7 in exit points and then later in the motor
neurons of the motor column which have exit points as their
Fig. 7. Quail-chick chimaeras, double stained using c-cad7 in situ
hybridisation and immunohistochemistry with the QCPN antibody.
(A) Transverse section through the facial exit point of a
homotopically grafted quail-chick chimaera showing the neural crest
cells at the site of c-cad7 expression. (B) Section through facial exit
point of a heterotopically grafted quail-chick chimaera where spinal
cord crest had been transplanted into r4. The ectopic spinal cord crest
cells express the appropriate marker, c-cad7 for their new
environment. Scale bar, 50 µm.
intermediate target, might constitute a recognition system
with a function in axon guidance to the intermediate target of
the exit points.
We have focused our study on the exit/entry points of mixed
motor-sensory nerves in the hindbrain and have identified a
population of crest cells defining this region. These cells are
specific for the region in which axons cross from the CNS into
the periphery and vice versa. However, they are not necessarily restricted only to the branchiomotor nerve exit points. ccad7, for example, is associated with several cranial motor
nerve exit points, like the oculomotor and abducens nerve. Our
preliminary evidence suggest that the abducens exit points,
which are rootlets on the ventral side of the neural tube in
rhombomere r5 and r6, are also defined by late emigrating
neural crest cells.
c-cad7 is also associated with dorsal root ganglia and spinal
motor nerves, as is Krox-20 (Wilkinson et al., 1989; SchneiderMaunoury et al., 1993). However, Krox-20 expression in the
trunk is associated with Schwann cell development (Topilko et
al., 1994). Whether similar mechanisms are involved in the
generation of nerve exit/entry points in the head and the spinal
cord remains to be investigated.
We wish to thank Dr Cairine Logan and Dr Isobel Heyman for a
picture of the Krox-20 in situ hybridisation and the use of unpublished
data. We thank Dr Anthony Graham and Dr Ian McKay for the careful
reading of the manuscript and for helpful discussion. The research was
supported by the Medical Research Council, the Wellcome Trust and
the Howard Hughes Medical Institute, of which A. L. is an International Research Scholar. C. N. received support from the Gottlieb
Daimler- and Carl Benz-Stiftung and the European Community.
Birgbauer, E. and Fraser, S. E. (1994). Violation of cell lineage restriction
compartments in the chick hindbrain. Development 120, 1347-1356.
Colamarino, S. A. and Tessier-Lavigne, M. (1995). The axonal
chemoattractant Netrin-1 is also a chemorepellent for trochlear motor axons.
Cell 81, 621-629.
Covell, D. A. and Noden, D. M. (1989). Embryonic development of the chick
primary trigeminal sensory-motor complex. J. Comp. Neurol. 286, 488-503.
Detwiler, S. R. (1937). Observations upon the migration of neural crest cells,
and upon the development of the spinal ganglia and vertebral arches in
Amblystoma. Am. J. Anat. 61, 63-94.
Goulding, M. D., Lumsden, A. and Gruss, P. (1993). Signals from the
2374 C. Niederländer and A. Lumsden
notochord and floor plate regulate the region-specific expression of two Pax
genes in the developing spinal cord. Development 117, 1001-1016.
Graham, A., Heyman, I. and Lumsden, A. (1993). Even-numbered
rhombomeres control the apoptotic elimination of neural crest cells from
odd-numbered rhombomeres in the chick hindbrain. Development, 119, 233245.
Graham, A., Francis-West, P., Brickell, P. and Lumsden, A. (1994). The
signalling molecule BMP-4 mediates apoptosis in the rhombencephalic
neural crest. Nature 372, 684-686.
Guthrie, S. and Lumsden, A. (1992). Motor neuron pathfinding following
rhombomere reversal in the chick embryo hindbrain. Development 114, 663673.
Guthrie, S. and Pini, A. (1995). Chemorepulsion of developing motor axons
by the floor plate. Neuron 14, 1117-1130.
Heaton, M. B. and Wayne, D. B. (1986). Specific responsiveness of chick
trigeminal motor nucleus explants to target-conditioned media. J. Comp.
Neurol. 243, 381 – 387.
Henrique, D., Adam, J., Myat, A., Chitnis, A., Lewis, J. and Ish-Horowicz,
D. (1995). Expression of a Delta homologue in the chick. Nature 375, 787790.
Keynes, R. J. and Stern, C. D. (1984). Segmentation in the vertebrate nervous
system. Nature 310, 786-789.
Krumlauf (1994). Hox genes in vertebrate development. Cell 78, 191-201.
Liem, K.F., Tremml, G., Roelink, H. and Jessell, T.M. (1995). Dorsal
differentiation of neural plate cells induced by BMP-mediated signals from
epidermal ectoderm. Cell 82, 969-979.
Lumsden, A. and Keynes, R. (1989). Segmental patterns of neuronal
development in the chick hindbrain. Nature 337, 424-428.
Lumsden, A. (1990). The cellular basis of segmentation in the developing
hindbrain. Trends Neurosci. 13, 329-335.
Lumsden, A., Sprawson, N. and Graham, A. (1991). Segmental origin and
migration of neural crest cells in the hindbrain region of the chick embryo.
Development 113, 1281-1291.
Lunn, E. R., Scourfield, J., Keynes, R. J. and Stern, C. D. (1987). The neural
tube origin of ventral root sheath cells in the chick embryo. Development
Mansouri, A., Stoykova, A. and Gruss, P. (1994). Pax genes in development.
[Review]. J. Cell Science-Suppl. 18, 35-42.
Moody, S. A. and Heaton, M. B. (1983a). Developmental relationships
between trigeminal ganglia and trigeminal motoneurons in chick embryos. I.
Ganglion development is necessary for motoneuron migration. J. Comp.
Neurol. 213, 327 – 343.
Moody, S. A. and Heaton, M. B. (1983b). Developmental relationship
between trigeminal ganglia and trigeminal motoneurons in chick embryos. II.
Ganglion axon ingrowth guides motoneuron migration. J. Comp. Neurol.
213, 344 – 349.
Moody, S. A. and Heaton, M. B. (1983c). Developmental relationships
between trigeminal ganglia and trigeminal motoneurons in chick embryos.
III. Ganglion perikarya direct motor axon growth in the peripherie. J. Comp.
Neurol. 213, 350 – 364.
Moody, S. A. and Heaton, M. B. (1983d). Ultrastructural observations of the
migration and early development of trigeminal motoneurons in chick
embryos. J. Comp. Neurol. 216, 20 – 35.
Nakagawa, S. and Takeichi, M. (1995). Neural crest specific cell-cell
adhesion by sequential and sub population specific expression of novel
cadherins. Development 121, 1321-1332.
Newgreen, D. F. and Erickson, C. A. (1986). The migration of neural crest
cells. Int. Rev. Cytol. 103, 89-145.
Noden, D. M. (1980). Somatotopic organisation of the embryonic chick
trigeminal neuron. J. Comp. Neurol. 190, 429-444.
Noden, D. M. (1988). Interactions and fates of avian craniofacial mesenchyme.
Development Supplement 103, 121-140.
Redies, C. and Takeichi, M. (1993). N- and R-cadherin expression in the optic
nerve of the chicken embryo. Glia 8, 161-171.
Schneider-Maunoury, S., Topilko, P., Seitanidou, T., Levi, G., CohenTannoudji, M., Pournin, S., Babinet, C. and Charnay, P. (1993).
Disruption of Krox-20 results in alteration of rhombomeres 3 and 5 in the
developing hindbrain. Cell 75, 1199-1214.
Scherson, T., Serbedzija, G., Fraser, S. and Bronner-Fraser, M. (1993).
Regulative capacity of the cranial neural tube to form neural crest.
Development 118, 1049-1061.
Sechrist, J., Serbedzija, G. N., Scherson, T., Fraser, S. E. and BronnerFraser, M. (1993). Segmental migration of the hindbrain neural crest does
not arise from its segmental generation. Development 118, 691-703.
Simon, H., Hornbruch, A. and Lumsden, A. (1995). Independent assignment
of antero-posterior and dorso-ventral positional values in the developing
chick hindbrain. Curr. Biol. 5, 205-214.
Topilko, P., Schneider-Maunouri, S., Levi, G., Baron-Van Evercooren, A.,
Ben Younes Chennnoufi, A., Seltanidou, T., Babinet, C. and Charnay, P.
(1994). Krox20 controls myelination in the peripheral nervous system.
Nature 371, 796-799.
Tsuchida, T., Ensini, M., Morton, S. B., Baldassare, M., Edlund, T., Jessell,
T. M. and Pfaff, S. L. (1994). Topographic organisation of embryonic motor
neurons defined by expression of LIM homeobox genes. Cell 79, 957-970.
Vaage, S. (1969). The segmentation of the primitive neural tube in chick
embryos (Gallus domesticus). Adv. Anat. Embryol. Cell Biol. 41, 1-88.
Wilkinson, D. G. (1992). Whole Mount In-situ hybridisation of vertebrate
embryos. IRL Press
Wilkinson, D. G., Bhatt, S., Chavrier, P., Bravo, R. and Charnay, P. (1989).
Segment-specific expression of a zinc-finger gene in the developing nervous
system of the mouse. Nature 337, 461-464.
Williams, E. J., Furness, J., Walsh, F. S. and Doherty, P. (1994). Activation
of the FGF receptor underlies neurite outgrowth stimulated by L1, N-CAM,
and N-Cadherin. Neuron 13, 583-594.
(Accepted 29 April 1996)