7. Embryol. exp. Morph. 85, 111-119 (1985)
Printed in Great Britain © The Company of Biologists Limited 1985
111
The kinetic behaviour of the cranial neural
epithelium during neurulation in the rat
FIONA TUCKETT AND GILLIAN M. MORRISS-KAY
Department of Human Anatomy, University of Oxford, South Parks Road, Oxford
OX13QX, U.K.
SUMMARY
The kinetic behaviour of the cranial neuroepithelial cells of rat embryos during neurulation is
described..Serial transverse sections of 4-, 8-, 12- and 16-somite-stage embryos show that
differential mitosis does not play a part in the mechanisms responsible for effecting cranial
neural tube closure. A constant cell number is found in the midbrain/hindbrain neural
epithelium during all four stages; the mitotic spindle axes are orientated parallel to the long axis
of the embryo, so that increase in cell number occurs in this direction only.
Growth is only expressed by an expansion in the volume of the forebrain, which projects
rostral to the notochordal tip. [3H]thymidine studies (using an in vitro culture technique) show
no significant variation in the cell cycle time between the forebrain and the midbrain/anterior
hindbrain neural epithelium. It is suggested that the neural epithelium is afluidstructure whose
overall shape is strictly controlled while the cells within it flow towards and into the rapidly
expanding forebrain.
INTRODUCTION
Growth kinetics of the rodent embryo have previously been studied during the
pre-implantation period of development, the immediate postimplantation period,
and the period after cranial neural tube closure. During cleavage of the mouse
blastocyst the length of the cell cycle time is approximately 20 h, and this is halved
by the time implantation occurs (Graham, 1973). Prior to primitive streak
formation the cell cycle time continues to shorten to around 9h; during primitive
streak formation itself, there is a period of rapid cell proliferation in which the
average cell cycle time is reduced to 5 h; analysis of the proliferation rates within
the epiblast revealed that 10 % of the epiblast had a very short generation time of
between 2 and 3h, the remaining tissue having a cell cycle time of 6-5 h (Snow,
1976,1977).
During the period immediately after neural tube closure (10th day of gestation
in the mouse), the cell cycle time was found to be 8-5 h in the thoracic neural tube
(Kauffman, 1966, 1968), and in the mesencephalon (Wilson, 1974), whereas
Key words: Mitosis, neural tube, cranial neuroepithelial cells, cell numbers, rat embryo.
112
F. TUCKETT AND G. M. MORRISS-KAY
Hoshino, Matsuwaza & Murakami (1973) found that the telencephalon had a
shorter cell cycle time of only 7 h. These studies suggest that during the late somitic
period of development of the mouse, differences exist in the generation time of
cells in different regions of the neural tube.
Cell kinetic studies have not previously covered the period between the
formation of the primitive streak and closure of the cranial neural tube, i.e. during
early somitogenesis. The aim of this study was to observe cell proliferation and
growth within the neural epithelium during the period of cranial neurulation by
both light and scanning electron microscopy, to discover whether morphogenesis
of the neural plate during neurulation could be correlated with proliferative
changes within the epithelium itself. Embryos of four different somite stages (4,8,
12 and 16) were used. At the 4-somite stage, convex neural folds are beginning to
rise up out of the plane of the neural plate; by the 8-somite stage the forebrain has
expanded greatly in size and the neural folds have begun to fuse in the
myelencephalon as an upward continuation of spinal neural tube closure; in the 12somite-stage embryo the forebrain folds have apposed and fusion has occurred at
the forebrain/midbrain boundary, while the hindbrain folds have continued to
fuse leaving a midbrain neuropore between the two regions of fusion; by the 16somite stage, closure of the cranial neural folds is complete.
MATERIALS AND METHODS
Wistar strain rat embryos were explanted in Tyrode's saline on either day 9 or day 10 of
pregnancy (day of positive vaginal smear = day 0). In those embryos which were subsequently to
be cultured only Reichert's membrane was opened; in all other embryos the extraembryonic
membranes were removed before fixation.
Whole embryo culture
Embryos were cultured at a temperature of 38°C in 60 ml Pyrex glass bottles containing 2-5 ml
immediately centrifuged, heat-inactivated rat serum (Steele & New, 1974), 2-5 ml Tyrode's
saline and lOjul penicillin/streptomycin (5000 i.u. ml" 1 and 5000jug ml" 1 ). The bottles were
gassed with 5 % CO2:5 % O2:90 % N2 (New, Coppola & Cockroft, 1916a,b) prior to sealing. The
bottles were continuously rotated at 30r.p.m.
Determination of cell cycle times
The embryos were treated in culture continuously with [3H]thymidine for varying periods of
time; 5 juCi/ml [3H]thymidine (specific activity 2 Ci/mmol; Amersham International) was added
to each culture bottle (five embryos/bottle). Table 1 explains the experimental arrangement for
the addition of the [3H]thymidine so that the total culture period for all embryos was 10 h.
After the culture period the labelled embryos were thoroughly washed before fixation in
Bouin's aqueous fluid. The embryos were dehydrated through graded alcohols and embedded in
paraffin wax. Serial transverse sections of 5 fjim thickness were obtained and mounted on clean
glass slides. The wax sections were hydrated and washed thoroughly in tap water before
transferring to a darkroom. Under a safe light the slides were coated with Ilford nuclear
emulsion K5 and left to expose in the dark at 4°C for 2 weeks, after which they were developed
with Kodak D-19 developer, andfixedwith 30 % sodium thiosulphate (hypo). The sections were
subsequently stained with cresyl violet, dehydrated and mounted in DPX.
Under the light microscope the percentage of neuroepithelial cells in the forebrain, midbrain
and hindbrain regions with labelled nuclei was determined and the cell cycle time was calculated
Kinetic behaviour of the cranial neural epithelium
113
3
to be the time taken for all the cells capable of dividing to incorporate [ H]thymidine i.e. the
time taken to reach a plateau in the labelling index when it was plotted against the time period
after the addition of [ H]thymidine. The difference between the level of the plateau in labelling
index and an index of 100% was believed to be associated with a non-dividing pool of
neuroepithelial cells, although the size of the non-mitotic pool was not determined.
Distribution and orientation of mitotic cells determined by light microscopy (LM)
After removal of the membranes, embryos were fixed in 2-5% cacodylate-buffered
glutaraldehyde (0-1 M, pH 7-2 with 2mM-CaCl2), postfixed with cacodylate-buffered osmium
tetroxide, washed and dehydrated through graded alcohols, and embedded in Spurr resin. Serial
sections of 1 jum thickness were obtained, mounted on glass slides and stained with either equal
parts of 1 % methylene blue and 1 % azure II in 1 % borax, or 1 % toluidine blue in 1 % borax.
The resolution obtained with plastic semithin sections was much greater than with the thicker
wax sections and enabled an easy examination of the mitotic apparatus at relatively low
magnifications.
The transverse sections of neural epithelium were subdivided into three regions for
observation of the distribution of mitotic cells. The most lateral part of the neural epithelium
was termed the lateral region; around the neural groove was termed the basal region, and the
remainder the intermediate region. The orientation of the mitotic spindle apparatus was
determined from examination of both transverse and sagittal sections.
Scanning electron microscopy
After fixation in glutaraldehyde the embryos were washed, and sagittally bisected with a
cataract knife. The halved embryos were dehydrated through a graded series of acetones,
critical-point dried with a Polaron Critical-Point Drying Apparatus E3000, mounted on
aluminium stubs with double-sided tape, and coated with gold using Polaron E5100 series II
Cool Sputter Coater before viewing with the Jeol scanning electron microscope JSM-T20. From
the micrographs tracings were made of the forebrain and the midbrain/rostral hindbrain regions
(Fig. 1) and the area occupied by these regions was determined; the sulcus in the lateral neural
epithelium was taken as the forebrain/midbrain boundary.
RESULTS
Cell cycle times within the neural epithelium
The cell cycle times calculated from the continuous labelling of cells with
[3H]thymidine are presented in Table 2; five embryos were used for each of the
somite stages investigated, and the data were cumulated before determination of
Table 1. Experimental procedure for the continuous pulse method for labelling cells
Culture bottle
number
1 (control)
2
3
4
5
6
7
8
Hours of culture in
control medium
before addition of
[3H]thymidine
10
0
1
2
3
4
5
6
Hours of culture with
[3H]thymidine
0
10
9
8
7
6
5
4
114
F . TUCKETT AND G. M. MORRISS-KAY
the generation time. A two-way analysis of variance was performed which
revealed that there was no significant variation in the cell cycle times between
either the three brain regions or the four different somite stages (P> 0-2 in both
!^
. . II
fe'
Fig. 1. Scanning electron micrographs of the cranial neural folds in the four different
somite-stages studied; (A) 4-somite-stage embryo, (B) 8-somite-stage embryo, (C) 12somite-stage embryo, and (D) 16-somite-stage embryo. The scale bar represents
50jum. The boundaries taken for the measurement of surface area in Table 6 are
indicated by the lines a-a' and b-b'. The forebrain lies rostral to a-a'; b-b' overlies
the pre-otic sulcus.
The pre-otic sulcus is proposed as a barrier preventing cell movement between the
regions of hindbrain rostral and caudal to it. The rostrad-pointing arrows illustrate the
hypothesis that cells generated within the neural epithelium rostral to the pre-otic
sulcus move towards and into the rapidly expanding forebrain.
Kinetic behaviour of the cranial neural epithelium
115
instances), even though the calculated cell cycle time was nearly one hour greater
for the 12-somite-stage embryos.
Distribution and orientation of mitotic cells
In transverse sections, there was no concentration of mitotic cells within the
lateral, intermediate or basal regions at any of the four stages examined. This
would suggest that there is no active growth, at the lateral edges or elsewhere,
which might contribute towards the mechanism of neural tube closure. Further
evidence to support this was the finding that the number of cells in each transverse
section remained constant in the midbrain and hindbrain regions during all four
somite stages observed (Table 3).
Table 2. Cell cycle times calculated by the continuous pulse method
Somite number
Forebrain
Midbrain
Hindbrain
4
8
12
16
5-9±0-lh*
6-l±0-lh
6-9 ± 0-1 h
6-l±0-lh
6-3 ± 0-5 h
6-l±0-lh
6-9 ± 0-1 h
6-l±0-lh
5-9±0-lh
6-l±0-lh
7-0±0-lh
6-l±0-lh
* standard deviation.
From the two-way analysis of variance no significant difference exists between the calculated cell
cycle times.
Fbrain region = 0'009, m = 2, X\2 = 6, P> 0-2.
Fsomite stage = 0-458, 1^ = 3, n 2 = 6, P> 0-2.
For each cell cycle determination, counts were made on ten sections within each brain region
and data were cumulated from five different embryos.
Table 3. Mean number of cells/transverse section in the midbrain/ hindbrain
Embryo
0)
(ii)
(iii)
(iv)
(v)
(vi)
(vii)
(viii)
(ix)
(x)
Mean
4
124 ± 1 *
124 ± 2
125 ± 2
124 ± 1
124 ± 1
127 ± 2
124 ± 2
125 ± 2
124 ± 1
124 ± 2
125
Somite stage
12
8
124 ± 2
124 ± 2
125 ± 3
126 ± 3
125 ± 1
124 ± 1
125 ± 1
125 ± 2
125 ± 1
125 ± 1
125
124 ± 2
126 ± 2
126 ± 2
125 ± 1
124 ± 2
124 ± 1
124 ± 1
124 ± 1
125 ± 1
125 ± 1
125
16
126 ± 2
124 ± 1
125 ± 1
127 ± 3
125 ± 1
125 ± 1
125 ± 1
125 ± 1
124 ± 1
125 ± 1
125
* standard deviation.
From the analysis of variance no significant difference was found between the cell numbers in
transverse sections.
(F = 0-669, ni = 3, n2 = 27, P>0-2).
Cell counts were obtained from sections at 5 jum intervals along the total length of the neural
tube within the midbrain and hindbrain regions.
116
F. TUCKETT AND G. M. MORRISS-KAY
The majority of metaphase cells were orientated so that cell division would
result in an elongation of the neural epithelium along the long axis of the embryo.
Table 4 gives details of the orientation of the metaphase spindle axes in embryos
from 4- and 8-somite-stage embryos and the open region of 12-somite-stage
embryos. Table 5 gives details of the orientation of the metaphase spindle axes in
embryos from the closed region of 12-somite-stage embryos and from the whole
cranial neural tube in 16-somite-stage embryos. Of the small number of cells not
dividing along the length of the embryo, the majority were orientated with their
spindle axes parallel to the luminal border in the transverse plane. Since there is no
significant increase in cell number per transverse section (Table 3), the increase
which results from transversely orientated division is most likely to represent a
numerically controlled replacement of cells lost from the neural epithelium by
either cell death or migration of neural crest cells, although cell movement either
rostrally or caudally could also be involved. It was very rare for a cell to have its
spindle axis perpendicular to the luminal border: this orientation was observed in
less than 1 % of all neuroepithelial cells in sections where neural tube closure had
not yet taken place. After neural tube closure the percentage of metaphase cells
with this orientation had increased but was still less than 5 %; this slightly higher
Table 4. Orientation of the metaphase spindle axis in the open neural folds. Percentage
of cells with their spindle axis along the long axis of the embryo
Embryo
Lateral
region
Intermediate
region
Basal
region
Mean
4-Somite
(i)
(ii)
(iii)
(iv)
(v)
Mean
86
83
88
92
88
87
97
98
97
95
96
97
89
98
96
100
100
97
91
93
94
96
95
94
8-Somite
(i)
(ii)
(iii)
(iv)
(v)
Mean
82
90
81
87
91
86
98
94
99
96
97
97
100
100
100
94
95
98
93
95
93
92
94
94
88
94
97
96
91
93
95
100
96
94
100
97
94
97
93
93
92
94
12-Somite (open region)
100
(i)
98
(ii)
85
(iii)
(iv)
88
85
(v)
Mean
91
Kinetic behaviour of the cranial neural epithelium
117
count reflects the onset of cell migration away from the epithelium as primary
neuroblasts.
Scanning electron microscopy
Fig. 1 shows the areas measured. Measurement of the surface area (Table 6)
gives only an approximate indication of change in overall size of the neural
epithelium, since changes in thickness (cell height) also occur during this period
(Morriss-Kay, 1981); no allowance was made for the concavity of the developing
optic sulcus, so the observed increase in forebrain surface area is a slight
underestimate of the actual increase. The measurements show an increase in
forebrain surface area of 20-fold during the period between the 4- and 8-somite
stages but only a slight increase (1-3 times) between the 8- and 16-somite stages.
There was no significant difference in the midbrain/rostral hindbrain surface area
during the period of cranial neurulation (x2 = 0-303 with 2 degrees of freedom,
0-2>P>0-l).
Table 5. Orientation of the metaphase spindle axis in the closed neural tube.
Percentage of cells with their spindle axis parallel to the luminal border along the long
axis of the embryo
Lateral
Embryo
region
12-Somite (closed region)
Intermediate
region
Basal
region
Mean
92
81
79
(ii)
(hi)
(iv)
87
88
83
84
84
89
95
76
87
88
79
(v)
71
Mean
85
79
84
82
70
83
73
82
(i)
78
16-Somite
(i)
00
(iii)
(iv)
(v)
Mean
74
86
74
78
78
76
77
76
75
77
77
77
87
83
80
82
74
72
89
74
81
81
75
78
Table 6. Areas of the brain regions measured from scanning electron micrographs
/Ml 2 X 10 4
MB + RHB
/an 2 x 104
FB
Somite number
FB + MB + RHB
4
8
12
16
1-14
2-47
2-64
3-05
1-07
1-075
1-078
1-221
0-07
1-395
1-562
1-829
/mi2 x 104
118
F. TUCKETT AND G. M. MORRISS-KAY
DISCUSSION
This study has shown that during the period of cranial neurulation, the cranial
neural epithelium is a highly mitotically active tissue, and that the cell cycle times
vary little between different regions of the developing brain or between different
somite stages. The regional differences observed in the cell cycle times in late
somitic embryos (Hoshino et al 1973; Kauffman, 1966, 1968; Wilson, 1974) are
most likely associated with the cellular differentiation occurring within the neural
tube at that time, e.g. development of the motor columns (Kauffman, 1968) and
cranial nerve ganglia (Altman & Bayer, 1982). The lack of regional or temporal
differences in cell cycle times observed in the present study could be associated
with the pre-differentiation and pre-specialization state of the neural epithelium
during the period of cranial neurulation.
Measurement of the surface area and cell number per transverse section of the
midbrain plus rostral hindbrain regions showed that there was no growth here
during neurulation. It was reported previously that this region does not grow in
length along the ventral midline (pre-otic sulcus to tip of notochord), nor in crosssectional area, at this time (Morriss-Kay, 1981). Jacobson & Tarn (1982) found an
increase in length of the ventral midline neural epithelium when the caudal
reference point was extended to the level of the first somite. Measurements on
their illustrations suggest that there is no discrepancy between the two sets of
results, since the common measurement (pre-otic sulcus to tip of notochord) is
constant in their SEMs, while the part caudal to it (pre-otic sulcus to
metencephalic-myelencephalic junction) increases.
In contrast to the constant size and cell number of the midbrain/rostral
hindbrain region, the forebrain expands rapidly during cranial neurulation.
Comparison of cell cycle time calculations with volumetric growth shows that
intrinsic cell division within the forebrain neural epithelium is insufficient to
account for forebrain expansion, while intrinsic cell division within the
midbrain/rostral hindbrain regions produces large numbers of daughter cells in a
region of negligible growth. These results suggest that cells generated in the
midbrain/rostral hindbrain regions migrate rostrally to populate and augment the
expanding forebrain. There is no morphological evidence of a specific migration
pathway to explain this phenomenon. We propose therefore that the whole
neuroepithelial sheet of cells moves rostrad passively due to the high rate of cell
division within an area of controlled cell number, the rostrocaudal orientation of
the mitotic spindles, and the presence of a barrier to caudad movement. This
hypothesis also requires that morphogenesis of the neural epithelium is controlled
within this fluid, forward-flowing sheet, and that the line of cells which form the
forebrain/midbrain boundary is a continuously changing population (Fig. 1). The
nature of the barrier to caudad migration is unknown; the pre-otic sulcus is a
candidate since it forms a clear morphological discontinuity across the whole width
of the neural epithelium.
Kinetic behaviour of the cranial neural epithelium
119
Experiments are at present in progress to test this hypothesis by injecting
radioactively labelled cells into the neural epithelium, followed by a period of in
vitro development.
We wish to thank Mr M. Barker for technical assistance. Fiona Tuckett was supported by an
M.R.C. Research Studentship.
REFERENCES
J. & BAYER, S. (1982). Development of the cranial nerve ganglia and related nuclei in
the rat. Adv. Anat. Embryol. Cell Biol. 74.
GRAHAM, C. F. (1973). The cell cycle during mammalian development. In The Cell Cycle in
Development and Differentiation (1st Symposium Br. Soc. devl Biol.). London: Cambridge
University Press.
HOSHINO, K., MATSUWAZA,T. &MURAKAMI, O. (1973). Characteristics of the cell cycle of matrix
cells in the mouse embryo during histogenesis of telencephalon. Expl Cell Res. 77, 89-94.
JACOBSON , A. G. & TAM, P. P. L. (1982). Cephalic neurulation in the mouse embryo analysed
by SEM and morphometry. Anat. Rec. 203, 375-396.
KAUFFMAN, S. L. (1966). An autoradiographic study of the generation cycle in the ten-day mouse
embryo neural tube. Expl Cell Res. 42, 67-73.
KAUFFMAN , S. L. (1968). Lengthening of the generation cycle during embryonic differentiation of
the mouse neural tube. Expl Cell Res. 49, 420-424.
MORRISS-KAY, G. M. (1981). Growth and development of pattern in the cranial neural epithelium
of rat embryos during neurulation. /. Embryol. exp. Morph. 65, (Suppl.) 225-241.
NEW, D. A. T., COPPOLA, P. T. & COCKROFT, D. L. (1976a). Improved development of headfold
rat embryos in culture resulting from low oxygen and modifications of the culture serum.
J. Reprod. Fert. 48, 219-222.
NEW, D. A. T., COPPOLA, P. T. & COCKROFT, D. L. (1976b). Comparison of growth in vitro and in
vivo of post-implantation rat embryos. /. Embryol. exp. Morph. 36, 133-144.
SNOW, M. H. L. (1976). Embryo growth during the immediate postimplantation period. Ciba
Foundation Symp. 40,53-66.
SNOW, M. H. L. (1977). Gastrulation in the mouse: Growth and regionalization of the epiblast.
/. Embryol. exp. Morph. 42, 293-303.
STEELE, C. E. & NEW, D. A. T. (1974). Serum variants causing the formation of double hearts
and other abnormalities in explanted rat embryos. /. Embryol. exp. Morph. 31, 707-719.
WILSON, D. B. (1974). Proliferation in the neural tube of the splotch (Sp) mutant mouse.
J. comp. Neurol. 154, 249-256.
ALTMAN,
(Accepted 23 July 1984)