Download Cell movements driving neuruiation in avian embryos

157
Development Supplement 2, 1991, 157-168
Printed in Great Britain © The Company of Biologists Limited 1991
Cell movements driving neuruiation in avian embryos
GARY C. SCHOENWOLF
Department of Anatomy,
University of Utah, School of Medicine, Salt Lake City, Utah 84132, USA
Summary
Neuruiation, formation of the neural tube, a crucial
event of early embryogenesis, is believed to be driven by
the coordination of a number of diverse morphogenetic
cell behaviors. Such behaviors include changes in cell
number (division, death), cell shape and size (wedging,
palisading and spreading), cell position (rearrangement
or intercalation) and cell-cell and cell-matrix associations (including inductive interactions). The focus of
this essay is on epiblast cell movements and their role in
shaping and bending of the neural plate. Neuruiation is a
multifactorial process requiring both intrinsic (within
the neural plate) and extrinsic (outside the neural plate)
forces. The origin and movements of three populations
of epiblast cells have been studied in avian embryos by
constructing quail/chick transplantation chimeras and
by labeling cells in situ with identifiable, heritable
markers. MHP (median hinge-point neurepithelial) cells
originate principally from a midline epiblast area rostral
to and overlapping Hensen's node. In addition, a few
caudal MHP cells originate from paranodal epiblast
areas. MHP cells stream down the length of the midline
neuraxis in the wake of the regressing Hensen's node.
This streaming occurs as a result of cell division
(presumably oriented so that daughter cells are placed
into the longitudinal plane rather than into the
transverse plane) and rearrangement (intercalation),
resulting in a narrowing of the width of the MHP region
with a concomitant increase in its length. L (lateral
neurepithelial) cells originate from paired epiblast areas
flanking the rostral portion of the primitive streak, and
they stream down the length of the lateral neuraxis
concomitant with regression of Hensen's node. They do
so both by oriented cell division and by intercalation. SE
(surface epithelial) cells originate from the epiblast of the
area pellucida, as far lateral as near the area pellucidaarea opaca border. From this area they stream medially,
toward the forming lateral margins of the neural plate.
Collectively, movements of the three populations of
epiblast cells generate the convergent-extension movements characteristic of the epiblast during neuruiation.
Heterotopic grafting has been used to assess the
relationship between cell position and cell fate and to
determine whether transplanted heterotopic cells can
adopt the behaviors typical of the new site. For example,
SE cells can replace L cells, changing their fate and
adopting L-cell behavior. Similarly, prospective MHP
and L cells both can change their fate and adopt the
behavior of SE cells. L cells, when placed into
prospective MHP-cell territory, move out of this
territory by intermingling with adjacent host L cells.
Likewise, prospective MHP cells placed into L-cell
territory, move out of this territory by intermingling
with host MHP cells. Collectively, these results suggest
that cell fate is determined principally by the ultimate
position of cells, and that adjacent, different cell
populations are restricted from intermingling with one
another. How positional information is specified, the
nature of restriction of intermingling and the guidance
cues used for cell navigation during streaming remain to
be elucidated.
Introduction
tail bud, into an epithelial, longitudinal cord, the
medullary cord, which subsequently cavitates, thereby
establishing a lumen and forming the caudal neural
tube. Why the embryo builds its neural tube in two
strikingly different manners is unknown, but the fact
that it does so underscores the importance of using
restraint when generalizing among species and among
developing systems. The only way to determine with
certainty how one species constructs its rudiments, is to
carefully catalog and analyze the events underlying the
formation of each rudiment in that organism. Generalization does provide insight and allows one to make
Formation of the neural tube, the process of neuruiation, occurs in two phases in avian and mammalian
embryos. During the first phase or primary neuruiation,
the flat, ectodermal neural plate rolls up into a
cylindrical neural tube. This phase of neuruiation is the
one that has received the most attention and, consequently, it is the one that is best understood. Secondary
neuruiation, the second of the two phases, begins when
primary neuruiation is nearing completion. It involves
the aggregation of mesenchymal cells, derived from the
Key words: cell division, cell rearrangement, cell shape,
ectoderm, epiblast, neural plate, neural tube,
neurepithelium, surface epithelium.
158
G. C. Schoenwolf
predictions, but insight and predictions do not necessarily provide facts.
I will restrict my discussion in this essay to primary
neurulation, a process that can be subdivided into four
distinct but spatially and temporally overlapping stages
(Fig. 1): (1) formation of the neural plate, (2) shaping
of the neural plate, (3) bending of the neural plate and
(4) closure of the neural groove. My discussion will be
further restricted to just two of these stages, shaping
and bending of the neural plate, and will be based
principally on information obtained, largely during the
last five years in my laboratory, from chick and quail
embryos. Recent studies have shown that shaping and
bending of the neural plate are driven by certain
morphogenetic cell behaviors, which occur both within
the neural plate as well as within the surrounding
tissues. A major task, which is not yet complete, has'
been to catalog the relevant behaviors and to assess
their respective roles in neurulation. This catalog
reveals that neurulation is not an all-or-none event, but
rather requires an interacting series of diverse cell
behaviors, which are precisely coordinated in both time
and space. Such behaviors include changes in cell
number (division and death), cell shape and size
(wedging, palisading and spreading), cell position
(rearrangement or intercalation) and cell-cell and
cell-matrix associations (including inductive interactions). The mechanisms underlying these obligatory cell
behaviors, and the manner in which they are coordinated, are largely unknown and remain as challenges
for future studies. Rather than discussing all cell
behaviors involved in neurulation, my focus in this brief
essay will be on one particular behavior, cell movement, which results in a change in the cell's position
within the embryo. Neurulation is discussed more
broadly in several recent reviews (Gordon, 1985;
Jacobson et al. 1986; Martins-Green, 1988; Copp et al.
1990; Schoenwolf and Smith, 1990a,6; Schoenwolf,
1991).
Shaping and bending of the neural plate are
multifactorial processes
Shaping and bending of the neural plate are multifactorial processes involving both intrinsic (i.e. within the
neural plate) and extrinsic (i.e. outside the neural plate)
forces acting in concert. Although the initiation of
shaping precedes bending, with the onset of bending,
these two processes become contemporaneous (e.g.
Schoenwolf, 1982, 1983).
Shaping involves intrinsic forces
Shaping of the neural plate includes three characteristic
events (Fig. 1A,B; Burnside and Jacobson, 1968;
Jacobson and Gordon, 1976; Morriss-Kay, 1981; Jacobson and Tarn, 1982; Schoenwolf, 1985; Schoenwolf and
Powers, 1987; Schoenwolf and Alvarez, 1989): (1)
apicobasal thickening as most neurepithelial cells
composing the pseudostratified, columnar epithelium
of the flat neural plate increase their height; (2)
transverse narrowing: (a) as neurepithelial cells increase in height, consequently decreasing their diameter; and (b) as neurepithelial cells rearrange, intercalating such that the number of cells spanning the
transverse dimension of the neural plate decreases,
while the number spanning the longitudinal dimension
increases; and (3) longitudinal lengthening: (a) as
neurepithelial cells intercalate, moving to decrease the
width of the neural plate and to increase its length; and
(b) as neurepithelial cells divide, with most daughter
cells positioned so as to increase the length of the neural
plate rather than its width. These processes still occur
when the neural plate is isolated from lateral tissues
(Schoenwolf, 1988), caudal tissues (Schoenwolf et al.
1989a) and underlying tissues (Alvarez and Schoenwolf,
1991a), suggesting that requisite forces for shaping of
the neural plate originate exclusively (or at least
principally) within the neural plate; that is, these forces
are intrinsic to the neural plate.
Bending involves intrinsic and extrinsic forces
Bending of the neural plate includes two characteristic
events (Figs 1B-D, 2; Schroeder, 1970; Morriss-Kay,
1981; Jacobson and Tarn, 1982; Schoenwolf, 1982, 1983;
Schoenwolf and Desmond, 1984): (1) furrowing; and
(2) folding. Furrowing of the neural plate, the
formation of shallow, gutterlike depressions, which
extend down the length of the neural plate, is driven (at
least for the median hinge-point region; see below) by
intrinsic forces and it is associated with the formation of
localized regions termed hinge points. Three hinge
points form during neurulation, a single median and
paired dorsolateral; the paired regions of neurepithelium between the median and dorsolateral hinge points
on each side are called the L (lateral) regions. The
median hinge point (MHP) forms at the midbrain
through spinal cord levels and consists of the midline
neurepithelium and underlying notochord (actually, it
could be argued that the MHP also forms at the
forebrain level, where it consists of the midline
neurepithelium and underlying prechordal plate mesoderm; my main reason for excluding the forebrain level
is that this region, unlike the rest of the neuraxis, does
not form a floor plate, a structure seemingly derived
from MHP cells; Schoenwolf et al. 1989/?; Fraser et al.
1990). It is defined as the midline area where the neural
plate becomes anchored to the notochord and longitudinal furrowing of the neurepithelium occurs. This
furrowing occurs autonomously within the MHP (i.e. it
does not require the presence of forces from more
lateral tissues; Schoenwolf, 1988), and it occurs as a
result of a change in neurepithelial cell shape from
columnlike to wedgelike (i.e. cell wedging) with an
accompanying decrease in neurepithelial cell height
(Schroeder, 1970; Brun and Garson, 1983; Schoenwolf
and Franks, 1984; Schoenwolf, 1985). Presumably, such
changes in cell shape and size generate intrinsic forces,
which deform the neural plate locally and establish the
longitudinal furrow. Much less is known about the
paired dorsolateral hinge points (DLHPs). These form
throughout most brain levels, as well as the most caudal
Cell movements driving neundation
159
A
(A*
flbr
1 1 H» •
i
Fig. 1. Light micrographs of living chick blastoderms on the surface of the yolk viewed dorsally through the vitelline
membranes. Formation of the neural plate occurred earlier in development, so that a flat neural plate (diagonal lines) is
already present in A. (A) Flat neural plate stage. (B) Initial neural groove stage; note that considerable shaping of the
neural plate has occurred and that bending of the neural plate (arrow) has been initiated. (C) Incipient neural tube stage:
note that the neural groove has closed at the future midbrain level (arrow). (D) Definitive neural tube stage: note that the
neural tube has formed throughout most of the length of the neuraxis. ps. primitive streak: nt. neural tube. Adapted (with
permission) from Schoenwolf and Watterson (1989). Bar=200/im (A); 400/im (B-D).
160
G. C. Schoenwolf
Fig. 2. Scanning electron micrographs of transverse slices of chick blastoderms during shaping and bending of the neural
plate. Neurepithelial cells of the MHP and DLHPs have been outlined. (A-C) Low power overviews; (D,E) enlargements
of the MHP region showing an early and a late stage (E is an enlargement of C); (F) enlargement of a DLHP region,
n, notochord; nf, neural fold; ng, neural groove; np, neural plate; ps, primitive streak; se, surface epithelium; arrow (in F),
initial furrowing of the DLHP. Bars=100,i<m (A-C); lOjum (D-F).
Cell movements driving neurulation
primary neural tube level (i.e. the region of the caudal
neuropore, a level derived from the sinus rhomboidalis
portion of the ectodermal neural plate) and are
composed, on each side, of the dorsolateral neurepithelium and the adjacent surface epithelium; that is, the
two layers constituting the paired bilaminar neural
folds. Whether furrowing of the DLHPs occurs
autonomously within these regions and whether the
associated changes that occur in the shape of its
neurepithelial cells (i.e. wedging and increase in
neurepithelial cell height) are induced by surface
epithelium have not been studied.
Folding of the neural plate, the uplifting of the neural
folds with the consequent formation of the neural
groove, is centered around the three hinge points and
requires (at least for neural fold elevation; see below)
extrinsic forces (i.e. forces generated outside the neural
plate). As extrinsic forces act upon the neural plate,
furrowing of the neurepithelium within the hinge points
facilitates subsequent neural plate folding, and anchorage of the neurepithelium to the underlying tissues (as
well as apicobasal polarization of the neurepithelium)
assists in preventing the neural plate from being
everted; thus, 'invagination' of the neural plate occurs
rather than evagination, establishing the neural groove.
Extrinsic forces cause the neural folds first to rotate
dorsally (i.e. to undergo elevation), with their axis oi
rotation centered on the MHP, and then medially (i.e.
to undergo convergence), with the axis of rotation on
each side centered on the corresponding DLHP. Thus,
in the hinge-point model (reviewed by Schoenwolf and
Smith, 1990b), there are 'hinge pins' (formed by
localized anchorage and furrowing at the hinge points)
and 'hinge plates' (formed by the paired L regions of
the neural plate for the MHP and by each neural fold
for the DLHPs), which are acted upon by forces arising
outside the neural plate.
Historically, bending of the neural plate has been
viewed as being driven exclusively by intrinsic forces
(reviewed by Schoenwolf and Smith, 1990a). This idea
arose with the classical experiments of Wilhelm Roux
(in 1885 and 1895; translated by Weiss, 1939), who
reported that the neural plate 'when isolated could
bend into a tube without external assistance.' However,
experiments in my laboratory (unpublished) have
shown that although pieces of avian epiblast do rapidly
roll up when isolated completely in a fluid medium
(Fig. 3A-C), they consistently roll up inside out (i.e.
the epithelium curls so that its original apical side, as
indicated by the presence of mitotic figures, lines the
outside of the vesicle, whereas its original basal side,
lines the inside of the vesicle; Fig. 3D). Hence, the fact
that isolated neural plates roll up in a direction directly
opposite to that occurring during normal neurulation
(also see experiments by Burnside, 1972; Jacobson,
1981; Vanroelen et al. 1982; Stern et al. 1985), suggests
not that neurulation is driven exclusively by intrinsic
forces (as originally concluded from Roux's experiment), but that such rolling up is an artifact of the
isolation and culturing of epithelial sheets.
The role of extrinsic forces in folding of the neural
161
Fig. 3. Light micrographs of isolated plugs of epiblast.
(A-C) Living plugs viewed from their basal (ventral) side
over a 2 h period (lettered in order of increasing time)
following their isolation; note the direction of curling.
(D) Section through an isolated plug; note the direction of
curling. Arrow, mitotic figure at the apical side of the plug.
Bar=120f<m (A-C); 30jxm (D).
162
G. C. Schoenwolf
plate has been assessed in two general ways: by ablation
experiments, and by the analysis of the behaviors of
cells in non-neurepithelial tissues. Ablation experiments involving enzymatic (or otherwise) depletion of
various components of the extracellular matrix, or
microsurgical removal of tissues lateral to the neural
plate, resulted in a delay, severe inhibition or complete
failure of bending of the neural plate (Morriss-Kay and
Crutch, 1982; Morriss-Kay etal. 1986; Morriss-Kay and
Tuckett, 1989; Anderson and Meier, 1982; Schoenwolf
and Fisher, 1983; Schoenwolf, 1988; Smits-van Prooije
etal. 1986; Tuckett and Morriss-Kay, 1989). In contrast,
removal of tissue caudal to or beneath the neural plate
had little or no effect on this process (Schoenwolf et al.
1989«; Alvarez and Schoenwolf, 1991a). Collectively,
these results suggest that tissues (i.e. cells and
associated extracellular matrix) lateral to the neural
plate cause folding of the neural plate. What might the
relevant tissues be? Three tissues form lateral to the
neuraxis: (1) the surface epithelium, continuous
medially with the neural plate; (2) the mesoderm,
consisting principally of lateral plate; and (3) the
endoderm, forming the fore, mid- and hindgut regions.
Removal of the endoderm and mesoderm, but not the
ectoderm, lateral to the neural plate does not prevent
folding of the neural plate, implicating the only
remaining tissue, the surface epithelium and its
associated extracellular matrix, in this process (Alvarez
and Schoenwolf, 1991a). What are the morphogenetic
cell behaviors that occur within the surface epithelium
to generate extrinsic neurulation forces? Much work
remains to be done. Nevertheless, our initial study of
the surface epithelium indicates that at least three cell
behaviors are involved (Schoenwolf and Alvarez,
1991): cell division, cell rearrangement and change in
cell shape from columnar to squamous (i.e. cell
spreading). Spreading of the epidermis (i.e. surface
epithelium) toward the dorsal midline also occurs
4A
_
during neurulation in amphibians (Jacobson, 1962;
Jacobson and Jacobson, 1973; Brun and Garson, 1983).
Cell movements within the epiblast during
neurulation
The movements of three types of epiblast cells, MHP
cells (i.e. neurepithelial cells within the MHP), L cells
(i.e. lateral neurepithelium cells between the MHP and
DLHPs) and surface epithelial (SE) cells (i.e. ectodermal cells lateral to the neural plate) have been studied
during shaping and bending of the neural plate. Two
experimental paradigms have been used to analyze the
direction and amount of such movement in each cell
population (Figs 4, 5): (1) construction of quail/chick
transplantation chimeras (Schoenwolf et al. 19896;
Schoenwolf and Alvarez, 1989, 1991; Alvarez and
Schoenwolf, 1991b); and (2) microinjection of a
fluorescent-histochemical marker (rhodamine-conjugated horseradish peroxidase) into the epiblast,
through which single epiblast cells or small groups of
epiblast cells are labeled (Schoenwolf and Sheard, 1989,
1990). In essence, the two approaches provided the
same information, so their results will be combined in
the following discussion.
MHP cells
Shortly after formation of the neural plate, prospective
MHP cells are located in three areas of the epiblast. The
first two of these areas contribute almost all the MHP
cells; the last area's contribution is relatively minor.
Essentially all the MHP cells of the future brain levels
are situated in a single, midline area located just rostral
to Hensen's node. From this area cells stream caudally,
in synchrony with the regression of Hensen's node, to
form the MHP cells of the midbrain and hindbrain
levels, as well as contributing some of the MHP cells of
the spinal cord level (see below). Cells from the
B
Fig. 4. Sections of a quail/chick chimera 24 h after transplantation (A; arrows indicate quail cells in the surface epithelium
of the neural fold) and of a chick embryo (B), whose flat neural plate was injected 24h earlier with rhodamine-conjugated
horseradish peroxidase (arrows indicate labeled cells in the neural tube). Bar=20,um (A); 15jum (B).
Cell movements driving neurulation
163
Fig. 5. Combined fluorescent and bright-field micrographs of
the dorsal surface of chick blastoderms injected with
rhodamine-conjugated horseradish peroxidase. Two cases are
shown (A,B> one case; C,D, another case). Note that
injected areas of the epiblast (encircled) move medially and
caudally during shaping and bending of the neural plate.
A,B, Oh and 6.5h after injection, respectively, n, notochord;
ps, primitive streak; hn, Hensen's node. C,D, Oh and 6.5h
after injection, respectively. Bar=200,um.
prenodal area also move rostrally to contribute to the
floor of the forebrain and the ventrolateral aspects of
the optic vesicles. Hensen's node is a second area that
contributes MHP cells. Although it is unclear exactly
where such prospective MHP cells arise in the node,
because the node is capped by epiblast, which is directly
continuous with the surrounding perinodal epiblast, it
seems likely that the cap is the source of these cells.
Hensen's node contributes MHP cells to the spinal cord
level, supplementing those derived from the prenodal
area. Recent experiments suggest that the paranodal
areas immediately flanking Hensen's node collectively
constitute a third area normally providing MHP cells;
however, the paranodal areas provide only a few MHP
cells and they do so only to the caudal levels of the
primary neural tube (Schoenwolf and Alvarez, 1991).
Interference with the contribution made by any one of
the three sources of MHP cells, results in a compensa-
164
G. C. Schoenwolf
tory increase in the number of cells contributed by the
remaining sources (Smith and Schoenwolf, 1991;
Schoenwolf et al. unpublished data). Hence, each of the
three sources of MHP cells apparently gauges the
number of such cells required at each rostrocaudal level
throughout the process of MHP formation; it then
inserts the appropriate number of cells into the forming
MHP region to fill in any gaps that would arise
otherwise. Seemingly, such a system would work only if
prospective MHP cells are induced rather than preprogrammed (which they are; see below), and it would
increase the assurance that a normal MHP region, an
important region for organizing neurons within the
developing neural tube (Placzek et al. 1990a; Hirano et
al. 1991), would form.
Prospective MHP cells exhibit at least four types of
morphogenetic behavior during shaping and bending of
the neural plate: cell division, change in cell shape and
size, cell-cell inductive interactions and cell rearrangement. The cell-cycle time of MHP cells is about 12 h
(e.g. at brain levels), approximately 65% longer than
that of cells of the flat neural plate prior to formation of
MHP and L regions (Smith and Schoenwolf, 1987).
Furthermore, most (i.e. approximately 70%) MHP
cells are wedgeshaped, in contrast to cells of the fiat
neural plate, most (i.e. approximately 70%) of which
are spindleshaped, and MHP cells have a decreased
height that is about 25 % less than that of the cells of the
flat neural plate (Schoenwolf and Franks, 1984;
Schoenwolf, 1985). These changes in cell shape and
size, relative to that of neurepithelial cells of the flat
neural plate, arise as a result of cell-cell inductive
interactions with the notochord (van Straaten et al.
1988; Smith and Schoenwolf, 1989; Placzek et al.
1990fo). Finally, MHP cells undergo extensive cell
rearrangement during neurulation, namely, lateromedial intercalation (Schoenwolf and Alvarez, 1989).
During such intercalation, cells move medially within
the width of the neural plate, and as a consequence of
this rearrangement, the number of cells arrayed within
the longitudinal plane of the neural plate is increased,
while the number of cells spanning its transverse
dimension is decreased. Such intercalation results in
convergent-extension of the neural plate, a movement
defined as a narrowing of the width of the neural plate
with a concomitant increase in its length. Convergentextension has perhaps been best documented in
echinoderm, amphibian and fish embryos, where it is
believed to be one of the principal motors driving
gastrulation (Keller, 1980; Keller et al. 1985; Keller and
Trinkhaus, 1987; Ettensohn, 1985; Hardin, 1989; Warga
and Kimmel, 1990).
L cells
L cells apparently have a simpler origin than do MHP
cells. Shortly after formation of the neural plate,
prospective L cells are located in paired, bandlike
epiblast areas flanking Hensen's node and the rostral
portion of the primitive streak (hence, they overlap the
areas that contribute a few MHP cells to the caudal
primary neural tube). From these bilateral areas cells
stream caudally, in synchrony with the regression of
Hensen's node, to form the paired L regions of the
midbrain through spinal cord levels. Cells from the
bilateral areas also move rostrally to contribute to the
roof of the forebrain and the dorsolateral aspects of the
optic vesicles. The exact longitudinal extent of the areas
alongside the primitive streak that contribute to the L
regions of the neural plate is a matter of controversy;
estimates of the area's length range from about 500 ,um
(Spratt, 1952) to something over 1 mm (Schoenwolf and
Sheard, 1990).
Prospective L cells exhibit at least three types of
morphogenetic behavior during shaping and bending of
the neural plate: cell division, change in cell shape and
size and cell rearrangement. The cell-cycle time of L
cells is about 8h (e.g. at brain levels), identical to that
of cells of the flat neural plate prior to formation of
MHP and L regions (Smith and Schoenwolf, 1987).
Furthermore, most (i.e. approximately 70%) L cells
are spindleshaped, like the cells of the flat neural plate,
but L cells have an increased height that is about 30 %
greater than that of the cells of the flat neural plate
(Schoenwolf and Franks, 1984; Schoenwolf, 1985).
Finally, L cells undergo extensive rearrangement
during neurulation as a result of lateromedial intercalation, and through this process they generate a
convergent-extension movement (Schoenwolf and
Alvarez, 1989).
SE cells
Shortly after formation of the neural plate, prospective
SE cells are located in an ovalshaped ring (with its long
axis oriented rostrocaudally) whose outer edge lies near
the area pellucida-area opaca border of the epiblast.
Cells within this ring form both extraembryonic and
intraembryonic SE cells. Extraembryonic SE cells
remain within the peripheral portion of the area
pellucida, and with formation of the body folds (i.e. the
head, lateral and tail body folds), these cells become
distinguishable from the intraembryonic SE cells. They
remain peripheral to the body folds and contribute to
the extraembryonic membranes (i.e. amnion and
chorion). Extraembryonic SE cells exhibit at least two
types of morphogenetic behavior during shaping and
bending of the neural plate (Smith and Schoenwolf,
1987; Schoenwolf and Alvarez, 1991): cell division, and
change in cell shape and size. Extraembryonic SE cells
undergo about two divisions (i.e. the cell-cycle time of
SE cells is about 10 h, or in other words, it is shorter
than that of MHP cells but longer than that of L cells).
In addition, extraembryonic SE cells change their shape
and size, converting from cuboidal to squamous and
thereby decreasing their height by about 60%. As a
consequence of cell division and change in cell shape
and size, grafts of extraembryonic SE cells exhibit
radial expansion (as considered en face in surface view),
increasing their surface area extensively.
Intraembryonic SE cells become incorporated into
the embryo proper with formation of the body folds,
and they contribute to the outer epithelial 'tube' (i.e.
the 'skin') of the typical vertebrate tube-within-a-tube
Cell movements driving neurulation
body plan. Intraembryonic SE cells, like extraembryonic SE cells, exhibit at least two types of morphogenetic behavior (Smith and Schoenwolf, 1987; Schoenwolf
and Alvarez, 1991). These cells undergo about two
divisions during shaping and bending of the neural
plate, but they essentially maintain their original shape
and size. Additionally, intraembryonic SE cells
undergo extensive rearrangement, namely, lateromedial intercalation, which generates a convergentextension movement.
How are cell fates determined?
There are two basic scenarios to account for the
specification of cell fate during neurulation. Either cells
could be predetermined for a particular developmental
fate prior to initiating their movement (i.e. cells already
could be specified to be: neurepithelium or surface
epithelium; MHP neurepithelial cells or L neurepithelial cells; or extraembryonic or intraembryonic
surface epithelial cells), or they could be undetermined
for a particular development fate until after their
movement is completed (i.e. cell fate could be specified
by local cell-cell inductive interactions near the final
area of cell residence). Recent experiments, based
either on (1) addition or subtraction of notochords or
notochordal precursor cells, or on (2) heterotopic grafts
of quail epiblast plugs to host chick embryos, provide
strong evidence that the latter scenario is the principal
one used. In the first type of experiment, it was shown
that in the absence of the notochord, MHP cells fail to
form, and in the presence of ectopic notochords,
supernumerary MHPs develop (van Straaten etal. 1988;
Smith and Schoenwolf, 1989; Placzek et al. 1990ft).
Thus, instructive, inductive interactions are required
for MHP formation. In the second type of experiment,
it was shown that prospective L cells, isolated in MHPcell territory, form MHP cells when they come to
overlie the notochord, and prospective MHP cells,
isolated in L-cell territory, form L cells when they are
removed from the influence of the notochord (Alvarez
and Schoenwolf, 19916). This experiment provides
further evidence of the induction of MHP cells by the
notochord and also provides evidence of the induction
of L cells (namely, because L cells differ from cells of
the flat neural plate, and when the latter are transplanted, they form L cells; therefore, this difference
between cells of the flat neural plate and L cells depends
upon the cell's location, not its area of origin). Whether
this induction occurs 'vertically' (i.e. a signal originating
from mesodermal and/or endodermal tissues underlying L cells) or 'horizontally' (i.e. an 'edgewise' signal
originating from adjacent L cells) remains to be
elucidated. Further experiments of the second type
have shown that SE and neurepithelial cells also arise as
a result of local inductive interactions (Schoenwolf and
Alvarez, 1991). For example, prospective SE cells
placed into L-cell territory form typical L cells and,
conversely, prospective L cells placed into surface
epithelium form surface epithelium (alternatively, in
this latter case the surface epithelium could be merely a
default pathway). Collectively, this series of exper-
165
iments demonstrates that cells of avian embryos possess
a tremendous ability to adapt to changes in their local
environment, presumably accounting (at least in large
part) for their capability to regulate and reconstitute
regions of the neurepithelium deleted experimentally
(e.g. Alvarez and Schoenwolf, 1991ft).
How are patterns of cell movement established?
The patterns of cell movement during neurulation are
complex and the question of how they are established is
far from answered. Here, I will consider two points.
First, individual cell movement is highly influenced by
movement of like cells in a surrounding cell stream.
Thus, for example, when plugs of prospective L cells
are transplanted to L-cell territory with the correct
apical-basal orientation, but with incorrect rostralcaudal and medial-lateral orientations, their pattern of
rearrangement still corresponds to that of the host
(Alvarez and Schoenwolf, 1991ft). This result shows
that cells do not go 'upstream' (i.e. against a 'current' of
streaming cells) even when 'disoriented' in their normal
territory and surrounded by like cells. Second, patterns
of cell movement for some types of cells are shown to be
labile and for other types of cells are shown to be fixed,
when cells are placed in atypical locations (Alvarez and
Schoenwolf, 1991ft; Schoenwolf and Alvarez, 1991).
For example, prospective L cells (or prospective MHP
cells) placed into surface epithelium adopt the rearrangement pattern of surrounding host SE cells, and,
conversely, prospective SE cells placed into L-cell
territory adopt the rearrangement pattern of surrounding host L cells. By contrast, prospective MHP cells
placed into L-cell territory, prospective L cells placed
into MHP-cell territory and prospective SE cells placed
into MHP-cell territory cannot adopt the rearrangement pattern typical of their new site. Instead, these
cells intermix (when the opportunity exists) with
adjacent host cells of the like type, preferentially
intercalating with them. How and when such cell
rearrangement preference is established is unknown.
Hypothesis: inhibition of cell rearrangement results in
neural tube defects
According to the hinge-point model, discussed above,
bending of the neural plate is driven by both intrinsic
forces (i.e. generated by cell wedging, which leads to
furrowing of the neural plate) and extrinsic forces
(generated by lateral tissues, presumably surface
epithelium and its associated matrix, which leads to
folding of the neural plate; that is, to the elevation and
convergence of the neural folds). Cell rearrangement
during shaping of the neural plate facilitates subsequent
folding of the neural plate by decreasing the width of
the neural plate and by increasing its length, thereby
allowing extrinsic forces to lift the neural folds toward
the dorsal midline where they fuse to establish a closed
neural tube. An hypothesis that emerges from the work
discussed above is that inhibition of neurepithelial cell
rearrangement results in neural tube defects; namely,
that such inhibition would result in a neural plate with a
wider (mediolateral) and shorter (rostrocaudal) con-
166
G. C. Schoenwolf
figuration than is normal, and consequently, that such
an ungainly neural plate would serve as an impediment,
opposing extrinsic forces in their endeavor to elevate
the neural folds. Unpublished evidence from my
laboratory suggests that in many cases, embryos with
anencephaly have apparently wider and shorter neural
tubes than do embryos undergoing normal neurulation
(Fig. 6). Moreover, recall that SE cells, like neurepithelial cells, also undergo rearrangement, and that this
process is implicated in providing driving forces for
bending of the neural plate. Consequently, inhibition of
cell rearrangement within the surface epithelium might
decrease the overall magnitude of the extrinsic forces,
thereby resulting in neural tube defects. The possibility
that some neural tube defects are generated by
decreased cell rearrangement is an important consideration worthy of future studies, and it emphasizes the
value of careful analysis of the cell behaviors underlying
normal morphogenesis. Undoubtedly, neural tube
defects arise in a multiplicity of ways, and it is not
unreasonable to suspect that inhibition of any of the
fundamental morphogenetic cell behaviors would cause
neurulation to go awry. Thus, we must remember the
complexity of the system when we consider the highly
relevant but difficult-to-answer question underlying
birth defects: What went wrong?
Concluding remarks
Analysis of cell behaviors during morphogenesis reveals
both the native beauty of the embryo as well as its
immense complexity. In view of such complexity, how
do we proceed in our analysis? The approach we have
chosen in my laboratory is based on the assumption that
the behaviors of individual cells and groups of cells
drive morphogenesis. Consequently, my laboratory has
focused on cataloging normal cell behaviors and
exploring how such behaviors are regulated. Many
questions about mechanisms of regulation still remain.
The embryo is 'a difficult nut to crack," but with the
proper "nutcrackers" the job can be done.
I wish to thank Jennifer Parsons and Fahima Rahman for
excellent assistance, and Dr I. Santiago Alvarez for providing
Figs 3D and 4A, unpublished micrographs from his studies in
my laboratory. The original research described herein from
my laboratory was funded by grants principally from the
National Institutes of Health.
References
ALVAREZ. I. S. AND SCHOENWOIT. G. C. (199hj)- Expansion of
surface epithelium provides the major extrinsic force for bending
of the neural plate. J. exp. Zooi, in press.
ALVAREZ. I. S. AND SCHOENWOLF. G. C. (1991/;). Patterns of
neurepithelial cell rearrangement during avian neurulation are
established independently of notochordal inductive interactions.
Deri Biol. 143. 78-42.
ANDKRSON. C. B. AND MEIER. S. (1482). Effect of hyaluronidase
treatment on the distribution of cranial neural crest cells in the
chick embryo. J. exp. Zool. 221. 329-335.
Fig. 6. Light (A) and scanning electron (B) micrographs of
the dorsal surface of chick blastoderms exhibiting
spontaneous neural tube defects (anencephaly). Arrow,
closed neural tube. The MHP region is outlined in B.
Bar=200,<mi (A); 50;/m (B).
BRUN. R. B. AND GARSON. J. A. (1983). Neurulation in the
Mexican salamander {Ambvstoma mcxieanum): A drug study
and cell shape analysis of the epidermis and the neural plate. J.
Embryol. exp. Morph. 74. 275-295.
BURNSIDE. B. (1972). Experimental induction of microfilament
formation and contraction. J. Cell Biol. 55. 33a.
Cell movements driving neurulation 167
BURNSIDE, M. B. AND JACOBSON, A. G. (1968). Analysis of
morphogenetic movements in the neural plate of the newt
Taricha torosa. Devi Biol. 18, 537-552.
COPP, A. J., BROOK, R. A., ESTIBEIRO, J. P., SHUM, A. S. W. AND
COCKROFT, D. L. (1990). The embryonic development of
mammalian neural tube defects. Prog. Neurobiol. 35, 363-403.
ETTENSOHN, C. A. (1985). Mechanisms of epithelial invagination.
Quart. Rev. Biol. 60, 289-307.
FRASER, S., KEYNES, R. AND LUMSDEN, A. (1990). Segmentation in
the chick embryo hindbrain is defined by cell lineage
restrictions. Nature 344, 431-435.
GORDON, R. (1985). A review of the theories of vertebrate
neurulation and their relationship to the mechanics of neural
tube birth defects. J. Embryol. exp. Morph. 89 Suppl., 229-255.
HARDIN, J. (1989). Local shifts in position and polarized motility
drive cell rearrangement during sea urchin gastrulation. Devi
Biol. 136, 430-445.
HIRANO, S., FUSE, S. AND SOHAL, G. S. (1991). The effect of the
DODD, J. (1990b). Mesodermal control of neural cell identity:
Induction of floor plate-specific properties by the notochord.
Science 250, 985-988.
SCHOENWOLF, G. C. (1982). On the morphogenesis of the early
rudiments of the developing central nervous system. Scanning
Electron Microsc. 1982/1, 289-308.
SCHOENWOLF, G. C. (1983). The chick epiblast: A model for
examining epithelial morphogenesis. Scanning Electron Microsc.
1983/HI, 1371-1385.
SCHOENWOLF, G. C. (1985). Shaping and bending of the avian
neuroepithelium: Morphometric analyses. Devi Biol. 109,
127-139.
SCHOENWOLF, G. C. (1988). Microsurgical analyses of avian
neurulation: Separation of medial and lateral tissues. J. comp.
Neurol. 216, 498-507.
SCHOENWOLF, G. C. (1991). Neurepithelial cell behavior during
avian neurulation. In Cell-Cell Interactions in Early Development
(ed. J. Gerhart), New York: Alan R. Liss, Inc. (in press).
floor plate on pattern and polarity in the developing central
nervous system. Science 251, 310-313.
JACOBSON, A. G. (1981). Morphogenesis of the neural plate and
tube. In Morphogenesis and Pattern Formation (eds. T. G.
Connelly, L. L. Brinkley and B. M. Carlson), pp. 233-263. New
York: Raven Press.
JACOBSON, A. G. AND GORDON, R. (1976). Changes in the shape of
the developing vertebrate nervous system analyzed
experimentally, mathematically and by computer simulation.
J. exp. Zool. 197, 191-246.
SCHOENWOLF, G. C. AND ALVAREZ, I. S. (1989). Roles of
JACOBSON, A. G., OSTER, G. F., ODELL, G. M. AND CHENG, L. Y.
SCHOENWOLF, G. C. AND DESMOND, M. E. (1984). Descriptive
(1986). Neurulation and the cortical tractor model for epithelial
folding. J. Embryol. exp. Morph. 96, 19-49.
JACOBSON, A. G. AND TAM, P. P. L. (1982). Cephalic neurulation
in the mouse embryo analyzed by SEM and morphometry.
Anat. Rec. 203, 375-396.
JACOBSON, C.-O. (1962). Cell migration in the neural plate and the
process of neurulation in the axolotl larva. Zool. Bidrag
(Uppsala) 35, 433-449.
JACOBSON, C.-O. AND JACOBSON, A. (1973). Studies on
morphogenetic movements during neural tube closure in
amphibia. Zoon 1, 17-21.
KELLER, R. E. (1980). The cellular basis of epiboly: an SEM study
of deep cell rearrangement during gastrulation in Xenopus
laevis. J. Embryol. exp. Morph. 60, 201-234.
KELLER, R. E., DANILCHIK, M., GIMLICH, R. AND SHIH, J. (1985).
The function and mechanism of convergent extension during
gastrulation of Xenopus laevis. J. Embryol. exp. Morph. 89
Suppl., 185-209.
KELLER, R. E. AND TRINKAUS, J. P. (1987). Rearrangement of
enveloping layer cells without disruption of the epithelial
permeability barrier as a factor in Fundulus epiboly. Devi Biol.
120, 12-24.
MARTINS-GREEN, M. (1988). Origin of the dorsal surface of the
neural tube by progressive delamination of epidermal ectoderm
and neuroepithelium: Implications for neurulation and neural
tube defects. Development 103, 687-706.
MORRISS-KAY, G. M. (1981). Growth and development of pattern
in the cranial neural epithelium of rat embryos during
neurulation. J. Embryol. exp. Morph. Suppl. 65, 225-241.
MORRISS-KAY, G. M. AND CRUTCH, B. (1982). Culture of rat
embryos with /J-D-xyloside: Evidence of a role for proteoglycans
in neurulation. J. Anat. 134, 491-506.
MORRISS-KAY, G. M. AND TUCKETT, F. (1989).
Immunohistochemical localisation of chondroitin sulphate
proteoglycans and the effects of chondroitinase ABC in 9- to 11day rat embryos. Development 106, 787-798.
MORRISS-KAY, G. M., TUCKETT, F. AND SOLURSH, M. (1986). The
effects of Streptomyces hyaluronidase on tissue organization and
cell cycle time in rat embryos. J. Embryol. exp. Morphol. 98,
59-70.
PLACZEK, M., TESSIER-LAVIGNE, M., JESSELL, T. AND DODD, J.
(1990a). Orientation of commissural axons in vitro in response
to a floor-plate-derived chemoattractant. Development 110,
19-30.
PLACZEK, M., TESSIER-LAVIGNE, M., YAMADA, T., JESSELL, T. AND
neuroepithelial cell rearrangement and division in shaping of the
avian neural plate. Development 106, 427-439.
SCHOENWOLF, G. C. AND ALVAREZ, I. S. (1991). Specification of
neurepithelium and surface epithelium in avian transplantation
chimeras. Development 112, 713-722.
SCHOENWOLF, G. C , BORTIER, H. AND VAKAET, L. (19896). Fate
mapping the avian neural plate with quail/chick chimeras:
Origin of prospective median wedge cells. J. exp. Zool. 249,
271-278.
studies of occlusion and reopening of the spinal canal of the
early chick embryo. Anat. Rec. 209, 251-263.
SCHOENWOLF, G. C , EVERAERT, S., BORTIER, H. AND VAKAET, L.
(1989a). Neural plate- and neural tube-forming potential of
isolated epiblast areas in avian embryos. Anat. Embryol. 179,
541-549.
SCHOENWOLF, G. C. AND FISHER, M. (1983). Analysis of the effects
of Streptomyces hyaluronidase on formation of the neural tube.
J. Embryol. exp. Morph. 73, 1-15.
SCHOENWOLF, G. C. AND FRANKS, M. V. (1984). Quantitative
analyses of changes in cell shapes during bending of the avian
neural plate. Devi Biol. 105, 257-272.
SCHOENWOLF, G. C. AND POWERS, M. L. (1987). Shaping of the
chick neuroepithelium during primary and secondary
neurulation: Role of cell elongation. Anat. Rec. 218, 182-195.
SCHOENWOLF, G. C. AND SHEARD, P. (1989). Shaping and bending
of the avian neural plate as analysed with a fluorescenthistochemical marker. Development 105, 17-25.
SCHOENWOLF, G. C. AND SHEARD, P. (1990). Fate mapping the
avian epiblast with focal injections of a fluorescent-histochemical
marker: Ectodermal derivatives. J. exp. Zool. 255, 323-339.
SCHOENWOLF, G. C. AND SMITH, J. L. (1990a). Mechanisms of
neurulation: Traditional viewpoint and recent advances.
Development 109, 243-270.
SCHOENWOLF, G. C. AND SMITH, J. L. (1990b). Epithelial cell
wedging: A fundamental cell behavior contributing to hinge
point formation during epithelial morphogenesis. In Control of
Morphogenesis by Specific Cell Behaviors (ed. R. E. Keller and
D. Fristrom). London: W.B. Saunders Co., Vol. 1, 325-334.
SCHOENWOLF, G. C.'AND WATTERSON, R. L. (1989).
Laboratory
Studies of Chick, Pig, and Frog Embryos: Guide and Atlas of
Vertebrate Embryology. 6th ed. New York: Macmillan
Publishing Company.
SCHROEDER, T. E. (1970). Neurulation in Xenopus laevis. An
analysis and model based upon light and electron microscopy. J.
Embryol. exp. Morph. 23, 427-462.
SMITH, J. L. AND SCHOENWOLF, G. C. (1987). Cell cycle and
neuroepithelial cell shape during bending of the chick neural
plate. Anat. Rec. 218, 196-206.
SMITH, J. L. AND SCHOENWOLF, G. C. (1989). Notochordal
induction of cell wedging in the chick neural plate and its role in
neural tube formation. J. exp. Zool. 250, 49-62.
SMITH, J. L. AND SCHOENWOLF, G. C. (1991). Further evidence of
extrinsic forces in bending of the neural plate. J. comp. Neurol.
307, 225-236.
168
G. C. Schoenwolf
SMITS-VAN PROOIJE, A., POELMANN, R., DUBBELDAM, J., MENTINK,
M. AND VERMEU-KEERS, C. (1986). The formation of the neural
tube in rat embryos, cultured in vitro, studied with teratogens.
Ada Histochem. 32, 41-45.
SPRATT, N. T., JR (1952). Localization of the prospective neural
plate in the early chick blastoderm. J. exp. Zool. 120, 109-130.
STERN, C. D . , MANNING, S. AND GILLESPIE, J. I. (1985). Fluid
transport across the epiblast of the early chick embryo. J.
Embryol. exp. Morph. 88, 365-384.
TUCKETT, F. AND MORRISS-KAY, G. M. (1989). Heparitinase
treatment of rat embryos during cranial neurulation. Anat.
Embryol. 180, 393-400.
VANROELEN, C , VERPLANKEN, P. AND VAKAET, L. C. A. (1982).
The effects of partial hypoblast removal on the cell morphology
of the epiblast in the chick blastoderm. J. Embryol. exp. Morph.
70, 189-196.
VAN STRAATEN, H. W. M., HEKKING, J. W. M., WIERTZ-HOESSELS,
E. J. L. M., THORS, F. AND DRUKKER, J. (1988). Effect of the
notochord on the differentiation of a floor plate area in the
neural tube of the chick embryo. Anat. Embryol. Ill, 317-324.
WARGA, R. M. AND KIMMEL, C. B. (1990). Cell movements during
epiboly and gastrulation in zebrafish. Development 108, 569-580.
WEISS, P. (1939). Principles of Development. New York: Henry
Holt and Company.