/. Embryol. exp. Morph. Vol. 57, pp. 71-78, 1980
Printed in Great Britain © Company of Biologists Limited 1980
7\
Is there a ventral neural ridge in chick embryos?
Implications for the origin of adenohypophyseal and
other APUD cells
By N. B. LEVY, ANN ANDREW 1 , B. B. RAWDON
AND BEVERLEY KRAMER
From the Department of Anatomy, University of the Witwatersrand
Medical School, Johannesburg
SUMMARY
Two- to ten-somite chick embryos were studied in order to ascertain whether, as has been
proposed, there exists a 'ventral neural ridge' which gives rise to the hypophyseal (Rathke's)
pouch. Serial sections and stereo-microscopy were used.
The neural ridges arch around the rostral end of the embryo onto the ventral surface of the
head, but no evidence was found for their extension to form a 'ventral neural ridge' reaching
the stomodaeum: in fact a considerable expanse of non-thickened surface ectoderm was
seen to separate the ventral portions of the neural ridges from the stomodaeum. The thickening of neural ectoderm which does appear on the ventral surface of the head results from
apposition and fusion of the opposite neural ridges flanking the neural plate and thus the tip
of the anterior neuropore - the classically accepted mode of closure of the neuropore.
These findings are in accord with the generally accepted concept of the origin of the hypophyseal pouch rather than with its derivation from a 'ventral neural ridge'.
No sign of neural crest formation was encountered ventrally; this observation excludes the
possibility that endocrine cells of the APUD series could originate from neural crest in this
region.
INTRODUCTION
The idea of a' ventral neural ridge' which gives rise to the hypophyseal (Rathke's)
pouch and the floor of the diencephalon in chick embryos has been formulated
by Takor Takor & Pearse (1975). Based on the examination of serial sections
of White Leghorn embryos, their claim is that thickening of the ventral ectoderm around the tip of the anterior neuropore, beginning at the 4-somite stage,
extends the neural ridges which flank the neural plate, to form a 'ventral neural
ridge'. This, it is maintained, reaches the stomodaeum by the 7-somite stage.
It is reported to become concave ventrally, forming a ventral neural sulcus, the
floor of which, cranial to the optic chiasma, is in contact with the floor of
the prosencephalon. Here cell lysis brings the prosocoel into continuity with the
1
Author's address (for reprints): Department of Anatomy, Medical School, Hospital
Street, Johannesburg 2001, South Africa.
72
N. B. LEVY AND OTHERS
amniotic cavity ventrally. Subsequent closure of the sulcus results in incorporation of part of the 'ventral ridge' into the floor of the diencephalon. The caudal
portion of the ' ventral neural ridge' between the optic chiasma and the stomodaeum is considered to give rise to the hypophyseal pouch and hence to the
adenohypophysis. These authors also describe cells freed from the 'ventral
neural ridge' migrating into the surrounding mesenchyme, and imply that these
are neural crest cells.
Takor Takor & Pearse's study opened up the possiblity that adenohypophyseal cell types and perhaps other endocrine APUD (Amine Precursor Uptake
and Decarboxylation) cells might arise from the 'ventral neural ridge' itself or
neural crest cells emigrating from it. This is important in relation to the proposal
that all APUD cells are of neurectodermal origin (Pearse, Polak & Bussolati,
1972; Pearse, 1977).
We have therefore undertaken a comparable study using serial sections and
examination of unfixed heads under a stereomicroscope, to reinvestigate the
matter.
MATERIALS AND METHODS
Eggs of the Black Australorp breed of domestic fowl were incubated at
37-5 °C for 31-36 h. The eggs were opened into chick Ringer's solution and the
embryos removed by the paper ring method of Low (1967). Embryos at 2- to
10-somite stages (stages 7 to 10 of Hamburger & Hamilton, 1951) were pinned
out in wax dishes containing Ringer's solution which was gradually replaced by
Bouin's fixative. The embryos were embedded in paraplast, serially sectioned
transversely or sagitally at 5/tm and stained with periodic acid-Schiff and
Table 1. Numbers of embryos studied
Intact heads
A
Scanning
election
microscopy
Stage
Serially
sectioned
2-3 somites
4 somites
5 somites
6 somites
7 somites
8 somites
9 somites
10 somites
3
3
3
3
7
2
1
2
1
1
1
2
2
2
2
2
5
6
7
3
2
24
7
29
Totals
60
Stereomicroscopy
Ventral neural ridge in chick embryos?
73
haematoxylin. Transverse sections of three other embryos of the same breed,
stained with Azure A, were also available for study.
Unfixed heads of embryos were studied and photographed in Ringer's solution under a Wild photomacroscope at the dark-field setting. Some heads were
examined by scanning electron microscopy but as these showed no more than
the unfixed heads, the former procedure was discontinued. Study of unfixed
heads has the advantage that no fixation artifacts are introduced.
The numbers of embryos prepared by the different methods and available for
study are given in Table 1.
RESULTS
Observations made on serially sectioned embryos, and on intact heads by
stereomicroscopy are integrated in the following description.
At the 2-, 3- and 4-somite stages the neural plate in the head is beginning to
roll up. The plate arches around the rostral end of the embryo onto the ventral
surface. The neural ridges flanking the plate are fairly prominent except around
its ventral tip where they are continuous with each other across the midline.
Here they blend smoothly into the ventral surface ectoderm.
The neural ridges are raised higher above the surface ectoderm at the 5-somite
stage, and the plate extends somewhat further ventrally (Fig. 1) but the tip
still does not have a well-demarcated caudal border.
Because the anterior neuropore is continuous around the rostral end of the
embryo onto the ventral surface, transverse sections may be obtained that show
both the dorsal and the ventral ends of the neuropore (Fig. 7).
A little later (5- to 6-somite stage), protruberant neural ridges have approached
each other rostrally so that they border a slit-like anterior neuropore. The
caudal end of the neuropore may be wider (Fig. 3) as would be expected when
the rounded end of a plate, flanked by a continuous ridge, rolls up. A subsidiary
pore of this description may be present until the 7-somite stage (Fig. 4). In
other specimens of 5- to 7- and even 8-somite embryos the neural ridges are
closely apposed to one another in the midline (Figs, 2, 5). A groove is then
present on the surface between them (Figs. 8-10), and this groove may continue onto the ventral surface ectoderm immediately caudal to the tip of the
neural plate. This is probably a result of the mechanical stresses involved in
approximation of the ridges. Around the tip, the neural ridges are seen to be
continuous with each other in serial sections, but there is no marked caudal
border. Here, the thick neural ectoderm is reflected into the caudally tapering
ventral ectoderm. The appearance of this transition in a longitudinal section
(Fig. 11) is identical to that seen where the dorsal portions of the ridges merge
with surface ectoderm laterally (for example, Fig. 8).
A completely closed anterior neuropore was first encountered at the 8-somite
stage (Fig. 6). Serial sections show definite fusion of the ridges caudally with
74
N. B. LEVY AND OTHERS
Ventral neural ridge in chick embryos?
75
cellular continuity across the midline. Further rostrally the ridges are only in
apposition. At this stage the ventral groove is not obliterated by fusion of the
ridges.
By the 9- and 10-somite stages, the ventral ends of the neural ridges have
fused leaving a groove which is usually no longer flanked by marked protruberances.
In no specimen at any stage studied is any sign of neural crest formation seen
in association with the ventral ends of the neural ridges (Figs. 7-11). The
appearance which would be expected is seen dorsally in Fig. 10.
The prochordal plate is evident from the earliest stage studied as a mass of
cells (Figs. 11, 12) without regular lateral boundaries. It is confluent with the
neural ectoderm, the rostral end of the foregut and the notochord.
As early as the 5-somite stage a very slight depression may foreshadow the
formation of the stomodaeum and delimit the oral plate. This depression appears
in one or two of the 6- and 7-somite-stage embryos as well (Fig. 12). By the
8-somite stage the depression is well-marked.
In 7-somite embryos the infundibulum takes the form of a V-shaped depression,
a diverticulum (Fig. 12) or a proliferating mass. The latter appearance is also
seen at the 8-somite stage: tubular or bud-like outgrowths are forming at this
time and in 9-somite embryos. No indication of the formation of the hypophyseal pouch is present.
The optic chiasma is not identifiable at the stages studied.
DISCUSSION
No evidence was seen for extension of the neural ridges beyond the tip of the
neural plate as proposed by Takor Takor & Pearse. We found that ventrally
the neural plate merges into the surface ectoderm in exactly the same manner as
the dorsal neural ridges blend into the ectoderm lateral to them. A mid-line
thickening of neural ectoderm does develop on the ventral surface of the head,
and it does bear a longitudinal groove. However, the thickening is formed by the
apposition of the opposite neural ridges which flank the tip of the neural plate and
therefore the anterior neuropore. As the ridges meet and fuse a median groove is
FIGURES
1-6
Ventral views of unfixed heads of chick embryos to show the ventral portions of
the neural ridges and closure of the anterior neuropore. The light area caudal to the
tip of the neural plate is attributable to a deep structure, probably the prochordal
plate, x 60.
Fig. 1. 5-somite stage.
Figs. 2, 3. 6-somite stage.
Figs. 4, 5. 7-somite stage.
Fig. 6. 8-somite stage.
76
N. B. LEVY AND OTHERS
\
Ventral neural ridge in chick embryos?
11
apparent between them. The groove may temporarily end in a caudal pit, but
apart from this, closure of the anterior neuropore proceeds from the ventral
tip - as well as from the dorsal extremity. This finding is in keeping with
the classical description of neuropore closure given by Romanoff (1960, p. 217).
Furthermore there was no sign of lysis between the neural ridge tissue around
the median groove on the one hand and the floor of the neural tube on the
other.
Probably Takor Takor & Pearse did not realise that the tip of the neural
plate, and therefore the anterior neuropore, extend onto the ventral surface of
the head. Several of their sections which they interpret as successive stages formation of the neural sulcus (their Fig. 8), union of the 'ventral neural ridge*
and the neural tube (their Fig. 9), lysis in this zone of contact (their Fig. 10)
and consequent communication of the prosocoel and the amniotic cavity (their
Fig. 11), appear to be from the same 6-somite-stage embryo. As such they
reflect apposition of the neural ridges and the resultant groove in the ventral
midline (their Fig. 8); more rostrally, flexure in the floor of the neural tube
curving around the rostral end of the embryo (their Figs. 9, 10); and more
rostrally still, the prosocoel open both dorsally and ventrally through the curve
of the anterior neuropore. Sections of a single embryo, comparable to those of
Takor Takor & Pearse, are shown in our Figs. 7-9.
The latter authors maintain that the hypophyseal pouch arises from a proliferation of the caudal part of the 'ventral neural ridge' between the optic
chiasma and the stomodaeum; this is held to bulge into the space between the
diencephalon and Seesel's pouch. The optic chiasma is not identifiable in our
embryos, but the stomodaeum is definable at least by the 8-somite stage. The
FIGURES 7-12
Sections through the heads of embryos orientated with the dorsal surface uppermost.
Figs. 7-9. Transverse sections through the rostral end of a 5-somite embryo at progressively more caudal levels, x 150.
Fig. 7. This section is cut through the anterior neuropore both dorsally and ventrally
(arrowheads).
Figs. 8, 9. Approximation and fusion of the neural ridges on the ventral surface
(arrowheads).
Fig. 10. Transverse section through the rostral end of a 7-somite embryo showing
neural crest (NC) formation dorsally and fusing neural ridges ventrally (arrowhead).
xl50.
Fig. 11. Sagittal section through the head of a 7-somite embryo showing the
anterior neuropore (NP), foregut (G) and prochordal plate (P). Note that ventrally
the neural ectoderm is continuous with caudally tapering surface ectoderm (arrow).
xl50.
Fig. 12. Parasagittal section (slightly oblique) through the head of a 7-somite
embryo showing closure of the dorsal end of the anterior neuropore (arrowhead),
infundibulum (I), foregut (G), prochordal plate (P) and stomodaeum (S). x 150.
6
EMB
57
78
N. B. LEVY AND OTHERS
ventral portions of the neural ridges are separated from the stomodaeum by a
considerable expanse of non-thickened surface ectoderm, and there is no pouch
present here. Nor is a hypophyseal pouch formed yet in the classical position
in our embryos-it would not be expected until the 14- or 15-somite stage
(Hammond, 1974). We could, however, detect the infundibulum at the 7somite stage, which is a little earlier than reported by Hammond. Also Silver
& Gould (1977) have found the hypophyseal pouch to be separated by a wide
gap from the anterior neuropore, in harmony with the generally accepted
concept of the origin of the pouch.
Our particular interest in Takor Takor & Pearse's report relates to their claim
that neural crest cells disperse from a 'ventral neural ridge' into the surrounding mesoderm. This would then constitute a possible so-far unrecognized source
for APUD cells of disputed origin, such as the endocrine cells of the gastrointestinal tract. The fact that we could find no 'ventral neural ridge' as described by these authors, and no sign at all of neural crest formation ventrally
has allayed any idea that gut endocrine cells could arise from this source.
The authors wish to thank the Senate Research Committee of the University of the
Witwatersrand, Johannesburg for financial support. We are grateful to Professor P. V.
Tobias for his interest and encouragement and to Mrs S. Brandes and Mr W. Tadiello for
skilled technical assistance.
REFERENCES
V. & HAMILTON, H. L. (1951). A series of normal stages in the development
of the chick embryo. / . Morph. 88, 49-92.
HAMMOND, W. S. (1974). Early hypophysial development in the chick embryo. Am. J. Anat.
141, 303-316.
Low, F. N. (1967). Developing boundary membranes in the chick embryo. Anat. Rec. 159,
231-238.
PEARSE, A. G. E. (1977). The APUD concept and its implications: related endocrine peptides
in brain, intestine, pituitary, placenta and anuran cutaneous glands. Med. Biol. 55,115-125.
PEARSE, A. G. E., POLAK, J. M. & BUSSOLATI, G. (1972). The neural crest origin of gastrointestinal and pancreatic endocrine polypeptide cells and their distinction by sequential
immunofluorescence. Folia Histochem. Cytochem, Krakow 10, 115-120.
ROMANOFF, A. L. (1960). The Avian Embryo. Structural and Functional Development. New
York: MacMillan.
SILVER, P. H. S. & GOULD, R. P. (1977). Rathke's pouch and the anterior neuropore of the
chick embryo (Gallus domesticus). Proc. Anat. Soc. Great Britain 36, 29 (Abstract).
TAKOR TAKOR, T. & PEARSE, A. G. E. (1975). Neuroectodermal origin of avian hypothalamohypophyseal complex: the role of the ventral neural ridge. / . Embryol. exp. Morph. 34,
311-325.
HAMBURGER,
{Received 14 August 1979, revised 15 November 1979)