/ . Embryol. exo. Morph. Vol. 59, pp. 157-173, 1980
Printed in Great Britain © Company of Biologists Limited 1980
\ 57
Ectoderm and mesoderm interactions
in the limb bud of the chick embryo studied by
transfilter cultures: cartilage differentiation
and ultrastructural observations
By MADELEINE GUMPEL-PINOT1
From the Institut d'Embryologie experimentale, du CNRS et du
College de France, Nogent-sur-Mame
SUMMARY
The wing mesoderm of the chick embryo cultured in vitro without ectoderm is able to
differentiate into cartilage from stage 17 (Hamburger & Hamilton, 1951). But before this
stage the presence of ectoderm is necessary.
In transfilter cultures of wing-bud ectoderm and mesoderm, the mesodermal response as
measured by chondrogenesis was directly related to the pore size (0-2-1 /*m) of the filter.
Filters of 0-2/tm pore size and 10 /tm thickness gave no increase in chondrogenesis over that
of mesoderm cultures alone.
The lower face of filters on the upper face of which mesoderm or ectoderm had been cultured was observed by scanning electron microscopy. With ectoderm, no cell processes crossed
the filter. In contrast, with mesoderm, cell processes crossed the filter and this was also
related to pore size. A good correlation was observed between the mass and density of
processes crossing the filter and the mesodermal response.
It is concluded that induction of cartilage in limb mesoderm cannot be classified as a
'long-range transmission' system. It requires ectoderm and mesoderm to be separated by a
very narrow gap and this condition can be brought about in vitro by extension of mesodermal
processes through the filter close to the ectoderm.
The results are discussed in relation to a possible role of the basement membrane and
associated extracellular matrix in limb cartilage induction.
INTRODUCTION
Studies on chondrogenesis in the chick embryo show that the differentiation
of vertebral cartilage depends on the presence of neural tube and notochord
(review by Lash, 1968), while the otic capsule induces differentiation of the otic
cartilage (Benoit, 1960). In the developing limb, chondrogenesis requires the
influence of the limb ectoderm both in the chick (Gumpel-Pinot, 1972, 1973) and
in the mouse (Milaire & Mulnard, 1968; Luger, 1980). Thus, in these different
embryonic systems where mesenchyme becomes chondrogenic, a neighbouring
1
Author's Address: Institut d'Embryologie experimentale du CNRS et du College de
France, 49bis, Avenue de la Belle-Gabrielle, 94130-Nogent-sur-Marne, France.
II
EMB
59
158
M. GUMPEL-PINOT
tissue can be shown to be responsible for the induction. Since the result is
invariably cartilage differentiation, it could be expected that the inductive
process is the same, even though it is normally exercised by different tissues.
According to this hypothesis, each inductive tissue would be capable of initiating
cartilage differentiation in all three different types of chondrogenic mesenchyme.
However, vertebral chondrogenesis in vitro can be induced neither by the otic
vesicle (Benoit, 1960) nor by ectoderm (Holtzer, 1961; Luger, 1980). In addition,
while neural tube and notochord initiate otic chondrogenesis (Benoit, 1960),
these tissues fail to induce cartilage differentiation in the mesenchyme of the
chick limb bud (Gumpel-Pinot, 1972, 1973). In the mouse embryo, different
workers have reached different conclusions: Milaire & Mulnard (1968) report
that chondrogenesis in the limb bud is induced both by ectoderm and the neural
tube and notochord, while Luger (1980) claims that, as in the chick, neural tube
and notochord cannot induce limb chondrogenesis. Thus, as pointed out by
Luger, an unexpected specificity of cartilage induction exists at the level of
different organs.
The way in which the inductive influence is exercised has been studied particularly in the case of vertebral chondrogenesis. Here, it has been demonstrated
that extracellular matrix produced by the axial organs plays a determining role
in the inductive process (Strudel, 1971; Minor, 1973; Kosher Lash & Minor,
1973; Kosher & Church, 1975; Lash & Vasan, 1977; Lash, Belsky & Vasan,
1977).
Where chick limb chondrogenesis is concerned, culture experiments in which
ectoderm and mesoderm were separated by vitelline membrane (Gumpel-Pinot,
1973, and unpublished results) suggest that chondrogenesis takes place only
when ectodermal and mesodermal cells can establish 'contact' with each
other.
The aim of the present work is to clarify the precise conditions under which
ectoderm can induce cartilage differentiation in limb mesoderm, by setting up
transfilter associations of ectoderm and mesoderm.
MATERIALS AND METHODS
These experiments used White Leghorn chick embryos. The exact stage
reached at the time of excision of the wing primordia was specified by the somite
number and by Hamburger & Hamilton (1951) stages.
(i) Dissociation of the wing bud
Ectoderm was separated from mesoderm by means of trypsin digestion
(Zwilling, 1955). For exact details, see Gumpel-Pinot (1972, 1973).
Ectoderm and mesoderm interactions in chick limb bud
Control
159
Experiment
/
Nuclepore
poe
filter
\
Stage 17 = 2 wing buds
Stage 15-16 = 4 wing buds
Stage 14 = 6 wing buds
Fig. 1. Technique used in transfilter cultures. M, mesoderm; E, ectoderm.
(ii) Transfilter cultures
The culture medium was GIBCO H 21 medium supplemented with 12%
heat-inactivated foetal calf serum. Nuclepore filters (pore size 0-2-1 /«n, thickness 10/tm) were sterilized in 50% ethanol, washed three times in Tyrode's
solution and left in the culture medium at room temperature. Ectoderm was
attached to the lower surface of the filter with agar (half 1 % agar in Gey solution,
half culture medium). Mesoderm was placed on the opposite side (Fig. 1). The
control explants consist of mesoderm placed on the upper surface of the filter,
the lower surface being either left uncovered or covered by a film of agar. The
number of mesodermal limb 'cores' used for each explant was varied according
to the stage: at least one at stage 18, two at stage 17, four at stages 15-16 and
six at stage 14. For the scanning electron microscope study each explant consisted of 10-12 mesodermal or ectodermal components. The cultures were kept
at 38 °C in an humidified incubator in an atmosphere of 5 % CO2 in air.
(iii) Histological and cytological techniques
(a) 5 [im sections-light microscopy
After 5-7 days in culture, explants were fixed in Carnoy's solution and
embedded in wax. The sections were stained by haematoxylin-alcian blue and
observed by light microscopy to evaluate the histodifferentiation of the mesoderm.
(b) Semi-thin sections
After 2 days in culture, explants were fixed for 45 min in 2-5 % glutaraldehyde
in 0-1 M phosphate buffer (pH 7-3) at room temperature. They were washed in
three changes of the same buffer with 0-23 M saccharose and then post-fixed for
160
M. GUMPEL-PINOT
Table 1. Transfilter cultures (Nuclepore)
Pore size: 0 4 and 06 /tm. Results according to stage of explant.
Stage
(H and H)
St. 14
(21-23 somites)
St. 15-16
(24-28 somites)
St. 17
(29-32 somites)
St. 18
Total no. of
pairs of explants
Experiment
Differentiation of
cartilage
12
7
(60%)
17
(58%)
27
(81 %)
4
29
33
4
Control
Differentiation of
cartilage
0
2
(6%)
19
(57%)
4
1 h in 1 % osmic acid in the same buffer. Serial ethanolic dehydration was followed by embedding in Epon. The 1 /tm sections were stained with toluidineblue.
(c) Scanning electron microscopy
After 2 or 3 days in culture, explants formed by mesoderm or ectoderm
cultured on the upper face of filters were fixed for 45 min in 2 % glutaraldehyde
in 0-1 M cacodylate buffer (pH 7-3, at room temperature). They were washed in
0-2 M buffer and post-fixed for 1 h in 1 % osmic acid, then washed in distilled
water, dehydrated through an ethanol series followed by an ethanol-freon series
and critical-point dried. The explants were mounted on stubs so that the surface
of the filter opposite to the culture was visible, coated with gold and examined in
a Cameca MEB 07 electron microscope.
RESULTS
(i) Macroscopic observations
(a) Cartilage differentiation. Tables 1 and 2 summarize the results of cartilage
differentiation in the wing mesoderm. In Table 1, the results are given according
to the stage of the embryos from which the buds have been excised and concern
only explants cultured on Nuclepore filters with 0-4 and 0-6 jtim pore size. At
stage 17, the mesoderm of the wing cultured without ectoderm (controls) differentiated into cartilage in 57 % of the explants. This percentage was much
higher when ectoderm was cultured on the opposite surface of the filter (81 %).
Control explants were always much smaller than experimental ones, and cartilage, if it differentiated at all, was present in reduced amounts.
At stages 14-16, chondrogenesis was very rare in the controls (0-6%), but
ectoderm cultured on the opposite surface of the filter raised cartilage differentiation to 58 or 60 % of the explants. The growth of mesoderm in these explants
was considerably greater than in the controls (Fig. 2 a, b).
Ectoderm and mesoderm interactions in chick limb bud
161
In Table 2, the results are presented according to the pore size of the filters and
concern wing buds excised from embryos at stage 15-16. At this stage, as noted
above, differentiation of cartilage was very rare in the controls (Fig. 2c, Table 1).
Chondrogenesis was still rare (5 % of the explants) when ectoderm and mesoderm were separated by 0-2 ftm pore size filters. Cartilage differentiated in 44 %
of the explants when the pore size was 0-4 ftm i n 65 % of the explants when the
pore size was 0-6 ftm and in 80% of the explants when the pore size was 0-8
or 1 ftm.
Table 2. Transfilter cultures {Nuclepore)
Stage 15-16. Results according to the pore size of the filter.
Pore size
Total no. of
pairs of explants
02/mi
^
0-4 fim
0-6 fim
0-8orl/tm
9
20
15
Expsriment
Differentiation
of cartilage
J(5%)
4(44%)
13(65%)
12(80%)
Control
Differentiation
of cartilage
I
0
2
0
* In these 8 cases, ectoderm and mesoderm have moved relative to one another, so that
they are no longer opposite each other at the end of the culture period.
(b) Filter-exp/ant adhesivity. The ectoderm adhered weakly so that it was
difficult to avoid it separating from the filter before embedding. Similarly, the
mesoderm cultured without ectoderm separated easily from the filter when the
pore size was less than 0-6 ftm. When ectoderm and mesoderm were cultured on
opposite sides of the filter, the adhesion of mesoderm was related to the pore
size of the filter. When the pore size was 0-2 fim, ectoderm and mesoderm
frequently (8 out of 18 explants) moved relative to one another, so that they were
no longer opposite each other at the end of the culture period; in this case, both
tissues easily separated from the filter. In contrast, when the pore size reached
0-4 ftm and especially 0-6-1 ftm, the mesoderm was strongly adhesive to the filter.
(ii) Microscopic observations
(a) 5 fim sections. When separated by a filter, explants of ectoderm and mesoderm frequently contained several densely stained chondrogenic areas or a large
irregular cartilage mass (Figs. 2a, c and 3a). The control explants, which did not
show any chondrogenic differentiation, were small (Fig. 2 b), mainly composed
of loose undifferentiated mesenchyme (Fig. 2d). At stages 15-16 and even 17,
when areas of cartilage differentiated in the controls, these were small and
stained lightly with alcian blue (Fig. 3 b). In both control and experimental
explants, cartilage always tended to occupy a central position and was usually
separated from the filter by two or three sheets of non-chondrified cells.
162
M. GUMPEL-PINOT
1 mm
1mm
Ectoderm and mesoderm interactions in chick limb bud
163
Myocytes differentiated in both control and experimental explants, whether or
not cartilage was present (Fig. 3 b, c, d). Myocytes were observed in all the
explants of stage 17 and in most of those of stage 15-16 (42 out of 58) but were
never observed in the cultures of stage 14.
The periphery of the explants was occupied by undifferentiated mesenchyme
and pycnotic cells. The experimental explants were always more healthy than
the controls.
(b) Semi-thin sections. Explants formed by ectoderm and mesoderm from
stage-15 to stage-16 and stage-17 wing buds cultured on each side of Nuclepore
filters (pore size of 0-6 jim) were fixed after 2 days in culture. Semi-thin sections
of such explants showed the presence of abundant cytoplasmic processes in the
pores of the filter between ectoderm and mesoderm (Fig. 4 a, b). The mesoderm
adhered to the filter directly opposite the ectoderm, but tended to separate from
it at the periphery (Fig. 4 a).
(c) Scanning electron microscopy. The lower surface of Nuclepore filters was
examined after 2 or 3 days culture with stage-15 to stage-16 and stage-17 wingbud ectoderm or mesoderm on the upper surface. Cell outgrowths through the
filters were comparable for these different stages and the periods of culture, but
differed according to the pore size and tissue type.
Culture of mesoderm
When the pore size was 0-2 /tm, only a very few small rounded cell outgrowths
occurred on the lower surface of the filter (Fig. 5 a). Pore sizes of 0-4 and 0-6 /*m
allowed cell processes to clearly project out of the filter surface (Fig. 5 b, c).
These were often rounded, but sometimes they spread on the filter, forming very
long filopodia which occasionally exceeded 50 fim in length. When the pore size
of the filter was 0-8 or 1 /.vm (Fig. 6a) the density and volume of projecting cell
processes increased. Filopodia spread on the surface of the filter, bulged, joined
outgrowths from other pores and sometimes formed a very compact network at
the surface of the filter (Fig. 6 b). It should be emphasized that the cytoplasmic
processes inside the filter, when visible (Fig. 6b), were very thin, their diameter
generally being no more than 0-25 fim to 0-35 fim even when they occupied much
Fig. 2. (a) Stage 15-16. Experimental culture. Four mesoderm cores and eight
ectoderms were cultivated on opposite sides of a 0-6 /*m Nuclepore filter for 6 days.
Note the large quantity of cartilage, (b) Stage 15-16. Control culture. Four mesoderm
cores were cultured on the 0-6 fim filter for 6 days. The explant is small and failed to
form cartilage, (c) Stage 15. Explant made up of four mesoderm cores and eight ectoderms of the wing cultured on opposite sides of a 0-6 fim pore Nuclepore filter. The
explant contains several centrally positioned nodules of cartilage which do not
contact filter. The ectoderm became detached during fixation, (d) Stage 15. Explant
made up of four mesoderms cultured on thefilter,from which the explant has become
detached. There is no cartilage and many cells are pycnotic. c, cartilage.
164
'C
M. GUMPEL-PINOT
Ectoderm and mesoderm interactions in chick limb bud
165
larger pores; this cannot be due merely to a shrinkage artifact since the large
pores sometimes contained several filopodia. The density of the cell processes
was related to the pore size of the filter but even with the largest diameter, many
pores were apparently empty or at least not completely filled with cytoplasmic
processes.
When mesoderm at stages 15-16 or 17 has been cultured on filters, especially
of 0-6-1 ^m pore size, extracellular material accumulated between the cell
outgrowths. This formed an amorphous matrix on which granules were
visible (Fig. 6a). This phenomenon has been described and studied in detail
by Lash et al. (1977) in the case of somitic mesenchyme cultures in which
extracellular matrix produced by differentiating cells passes through the filter
and accumulates on the opposite side. By successive enzymic digestions, it has
been shown that this matrix was formed mainly of collagen and proteoglycans
(granules).
Culture of ectoderm
Whatever the pore size, cytoplasmic processes were never observed on the
lower surface of filters on the upper surface of which ectoderm had been cultured
for 2 or 3 days.
CONCLUSIONS AND DISCUSSION
(i) Transfer of the inductive signal
On the basis of the experiments just described, it is concluded that the transmission of the inductive signal from ectoderm to mesoderm cannot take place
at a distance, and needs conditions of 'contact' between the two tissues.
The problem of cellular interaction in development has been the subject of
numerous studies. Recently attention has been focused on the mechanism of
cellular communication, i.e. on the way the signal is transferred from one cell to
another. While there is little information concerning the nature of the signal,
observations are accumulating on the conditions necessary for its transmission.
Fig. 3. (a) Stage 17. Explant formed by two mesoderm cores cultured for 6 days with
four ectoderms on opposite sides of a 0-4 /im Nuclepore filter. The ectoderm has
become detached. The explant is rounded and almost entirely composed of cartilage.
(b) Stage 17. Control explant formed by two mesoderms cultured for 6 days on a
0-4 /im filter. Cartilage and differentiated myocytes are present, (c) Stages 15-16.
Explant formed by four mesoderms cultured with eight ectoderms for 7 days on
either side of a 0 4 /tm Nuclepore filter. The ectoderm and the filter have become
detached. There is no cartilage, but differentiated myocytes are present, (d) Stages
15-16. Control explant formed by four mesoderms cultured on a 0-6 /*m filter. The
explant contains no cartilage. High magnification of the myocytes. c, cartilage;
m, myocytes.
166
M. GUMPEL-PINOT
100 Hm
-M
Fig. 4. (a) Semi-thin section across an explant composed of four stage-15 to -16
mesoderm cores cultured with four ectoderms on opposite sides of a 0-6 /tm Nuclepore filter. Note that the mesoderm adheres to the filter opposite the ectoderm and
tends to detach at the periphery (arrowed), (b) The same explant. Enlargement
showing a part of the filter in the region separating the two tissues. Cellular processes
fill the pores of the filter. M, mesoderm; E, ectoderm.
These conditions vary according to the system being studied, and according to
Saxen (1977) two main types of system can be distinguished:
(1) Long-range transmission, where the cells are separated by a gap of the
order of 50-000 nm.
(2) Short-range transmission, where the cells are in direct contact, i.e. their
membranes are separated by a distance of the order of 5-10 nm. The present
author considers there is a third, intermediate group since, in certain systems, although the signals cannot be transmitted over a large gap, extracellular material
which may or may not be associated with the presence of a basement membrane
plays a direct short-range role in the interaction. Examples of this type are provided by the work of Meier & Hay (1975), Hay & Meier (1976), Hay (1977) on
corneal differentiation and of Thesleff (1977) on odontoblast differentiation.
Ectoderm and mesoderm interactions in chick limb bud
Fig. 5. SEM of the surface of the Nuclepore filters in regions where mesodermal
explants were positioned on the opposite surface, (a) 0-2 /tm pores, (b) 0-4 /«n pores,
(c) 0-6 /*m pores. Note the increasing density and volume of the mesodermal cell
processes which have crossed the filter.
167
168
M. GUMPEL- P I N O T
Fig. 6. SEM of the surface of Nuclepore filters in regions having mesodermal explants positioned on the opposite surface, (a) 0-8 /im pores. The mesodermal cell
processes are very dense. Note the presence of extracellular matrix (arrowed) on the
filter surface between the processes. (6) 0-8 /tm pores. The cell processes from several
pores forming a complex network on the filter surface.
Ectoderm and mesoderm interactions in chick limb bud
169
The infrastructure of the ectomesodermic space in the developing chick limb
has been studied between stages 10 and 26 (Jurand, 1965; Berczy, 1966; Ede,
Bellairs & Bancroft, 1974; Smith, Searls & Hiffer, 1975; Kaprio, 1977; GumpelPinot, unpublished observations). According to these studies a continuous
basement membrane is always found between ectoderm and mesoderm. Only
Ede et al. (1974) describe gaps in the basement membrane, and these are at a
late stage (24-26) and only under the apical ectodermal ridge, a structure whose
presence is not necessary for cartilage induction (Gumpel-Pinot, 1972, 1973).
Apart from this, the different accounts agree that at all the stages studied, cell
processes from the mesoderm establish contact with the basement membrane.
No direct contact between mesodermal and ectodermal cell membranes has ever
been described. Thus, in the normal relationship between limb mesoderm and
ectoderm, the basement membrane constitutes a permanent 'barrier'. Ultrastructural studies on the relationship when ectoderm and mesoderm are separated by Nuclepore filters should make it clear whether the basement membrane
is also a necessary intermediary for the transfer of the inductive signal. It is
worth emphasizing that a role for the basement membrane has recently been
demonstrated in differentiation of odontoblasts in the mouse embryo by Thesleff
(1977, 1978), Thesleff, Lehtonen & Saxen (1978) and Hurmerinta, Thesleff &
Saxen (1979). In this system these authors suggest that glycoproteins and
glycosaminoglycans in the basement membrane and on the surface of mesenchymal cells could have a role in the differentiation of odontoblasts. Changes
can be observed in the matrix proteins, fibronectin and collagen during the
differentiation of the tooth (Thesleff, Stenman, Vaheri & Timpl, 1979). Fibronectin, which is an important component of the basement membranes, seems to
increase in the basement membrane between epithelium and mesenchyme at the
time of odontoblast differentiation. Fibronectin could be involved in the attachment of the mesenchymal cells to the basement membrane.
Work by Milaire & Mulnaid (1968) demonstrated that in the mouse limb bud
the epiblast induced cartilage differentiation in the limb mesoderm. At that time
they set up transfilter cultures according to the Grobstein (1956) method, using
Millipore filters of type THWP (pore size 0-45 jam, thickness 25 jLom). Since the
epiblast effect took place across the filter, the authors concluded that the inductive factor was diffusible. But it has since been demonstrated that such filters
do not prevent cellular contacts (Nordling et al. 1971; Wartiovaara, Lehtonen,
Nordling & Saxen, 1972) and this conclusion should therefore be revised. At the
same time, the authors pointed out an interesting observations in relation to a
possible role of the basement membrane: some samples of trypsin which were
particularly active when used as dissociating agents, suppressed or reduced the
inductive properties of the epiblast. In these cases the authors observed the
disappearance or profound alteration in the basement membrane.
170
M. GUMPEL-PINOT
(ii) Differentiation of cartilage in in vitro culture and
the ectodermal role
In earlier work (Gumpel-Pinot, 1972, 1973) using a different culture method,
limb buds were grown on the semi-solid medium of Wolff & Haffen (1952) and
explants were wrapped in a fragment of vitelline membrane according to a
technique recommended by Wolff (1961). Under these culture conditions, when
mesoderm was cultured without ectoderm (controls), 1-9 % (3 out of 156) of the
stage-15 to stage-16 explants became chondrogenic, the proportion rising
to 13-3% of stage-17 explants (18 out of 135), to 39-2% of stage-18 explants
and finally to 79 % of stage-19 to stage-21 explants. In the culture conditions
described here, while the differentiation of cartilage remains rare at stages 15-16
(6 % of explants) it becomes dramatically more frequent at stage 17, since 57 %
of explants of this stage become chondrogenic. The comparison between the
results obtained in the first series of cultures and in the second one, in which
conditions seem to be more favourable for the phenotypic expression of cartilage,
highlights the difference between environmental factors acting once the mesoderm is capable of chondrogenesis without ectoderm (Gumpel-Pinot, 1972,1973)
and the specific ectodermal factor which is necessary to induce cartilage in
mesoderm incapable of autonomous chondrogenic differentiation.
At whatever stage of explantation, chondrogenic nodules most frequently
differentiated in the centre of the explant. If the prechondrocytes have to enter
into direct contact across the filter with the ectoderm in order to become chondrogenic, they must then move from the filter towards the explant centre. Such a
movement may also take place in vivo between the sub-ectodermal region and
the median part of the limb bud where the cartilaginous skeleton differentiates.
(iii) Attachment of explant cells to the filter: possible significance
The ectoderm, which does not put out cell processes capable of crossing the
filter, detaches from it very easily. The mesoderm from the control explants
adheres better to the filter when the pores are wider, i.e. when the density of the
processes which crossed the filter is greater. The attachment of explants to the
filter thus seems to be related to the density of cell processes anchoring tissue to
filter. Now, the mesoderm adhered better to the filter when ectoderm was present on the opposite side for all pore sizes. Semi-thin sections showed that
between the two tissues, pores of 0-6 jtim were filled with cellular material
although with the same diameter, scanning shows numbers of empty or incompletely filled pores when mesoderm alone was grown. This suggests that the
presence of ectoderm stimulates the outgrowth of mesoderm cell processes. Thus
the surface of 'contact' between ectoderm and mesoderm placed on either side
of the filter could be greater than suggested by the scanning observations.
Ectoderm and mesoderm interactions in chick limb bud
111
(iv) Myocyte differentiation
Myocyte differentiation is only observed in explants later than stages 15-16
(24-28 somites). Recent works show that the wing musculature will develop
from a cell line of somitic origin. The migration of this cell population towards
the somatopleure at the wing level begins at age 20 to 22 somites (stage 13-14)
(Christ, Jacob & Jacob, 1974, 1977; Chevallier, 1978). Premyocytes could
therefore be already present in the wing mesoderm explanted at stage 14. On the
other hand, it is noteworthy that in all cases where myocyte differentiation was
observed, a substantial number of cells were always present. It is thus possible
that a few isolated cells either cannot differentiate or escaped observation. So far,
absence of myocyte differentiation in stage-14 explants can be regarded as
support for the hypothesis of a non-somatopleural origin of the myocyte population. Furthermore, the differentiation of myocytes appears to be independent
both of the presence of ectoderm and of the formation of cartilage in the explant.
The author is most grateful to the 'Service Central de Microscopie Electronique de l'lnstitut Pasteur' in which the scanning observations have been carried out and especially to
Pr. L. G. Chevance.
Sincere thanks are expressed to J. R. Hinchliffe and F. D. Newgreen for their help during
the preparation of the manuscript and to M. Bontoux and L. Boule for their technical
assistance.
REFERENCES
J. A. A. (1960). Etude experimentale des facteurs de l'induction du cartilage otique
chez les embryons de poulet et de truite. Ann. Sci. nat. Zool. 12e Ser. 22, 323-385.
BERCZY, J. (1966). Zur Ultrastruktur des Extremitatenknospe. Z. Anat. Entwickl. Gesch. 125,
295-315.
CHEVALLIER, A. (1978). Etude de la migration des cellules somitiques dans le mesoderme
somatopleural de l'ebauche de l'aile. Wilhelm Roux's Arch, devl Biol. 184, 57-73.
CHRIST, B., JACOB, H. J. & JACOB, M. (1974). Uber den Ursprung der Flugelmuskulatur.
Experimentelle Untersuchungen mit Wachtel- und Huhnerembryonen. Experientia 30,
1446-1449.
CHRIST, B., JACOB, H. J. & JACOB, M. (1977). Experimental analysis of the origin of the wing
musculature in Avian embryos. Anat. Embryol. 150, 171-186.
EDE, D. A., BELLAIRS, R. & BANCROFT, M. (1974). A scanning microscope study of the early
limb bud in normal and talpid mutant chick embryos. / . Embryol. exp. Morph. 31,761-785.
GROBSTEIN, C. (1956). Transfilter induction of tubules in mouse metanephrogenic mesenchyme. Expl Cell Res. 10, 424-440.
GUMPEL-PINOT, M. (1972). Culture in vitro de l'ebauche de l'aile de l'embryon de, Poulet.
Role de l'ectoderme sur la chondrogenese. C.r. hebd. Seanc. Acad. Sci., Paris 274, 27862789.
GUMPEL-PINOT, M. (1973). Culture in vitro du bourgeon de l'aile de l'embryon de Poulet.
Differentiation du cartilage. Annee Biol. 12, 417-428.
HAMBURGER, V. & HAMILTON, H. L. (1951). A series of normal stages in the development of
the chick embryo. / . Morph. 88, 49-92.
HAY, E. D. (1977). Interaction between the cell surface and extracellular matrix in corneal
development. In Cell and Tissue Interactions (ed. J. W. Lash & M. M. Burger), pp.
115-137. New York: Raven Press.
HAY, E. D. & MEIER, S. (1976). Stimulation of corneal differentiation by interaction between
cell surface and extracellular matrix. II. Further studies on the nature and site of transfilter
•induction'. Devi Biol. 52, 141-157.
BENOIT,
172
M. GUMPEL-PINOT
H. (1961). Aspects of chondrogenesis and myogenesis. In Molecular and Cellular
Synthesis (ed. D. Rudnick), pp. 35-87. New York: Ronald Press.
HURMERINTA, K., THESLEFF, I. & SAXEN, L. (1979). Inhibition of tooth germ differentiation
in vitro by diazo-oxo-norleucine (DON). J. Embryol. exp. Morph. 50, 99-109.
JURAND, A. (1965). Ultrastructural aspects of early development of the fore-limb buds in the
chick and the mouse. Proc. R. Soc. B 162, 387-405.
KAPRIO, E. A. (1977). Ectodermal-mesenchymal interspace during the formation of the
chick leg bud. Wilhelm Roux's Arch, devl Biol. 182, 213-225.
KOSHER, R. A., LASH, J. W. & MINOR, R. R. (1973). Environmental enhancement of in vitro
chondrogenesis. IV. Stimulation of somite chondrogenesis by exogenous chondromucoproteins. Devi Biol. 35, 210-220.
KOSHER, R. A. & CHURCH, R. C. (1975). Stimulation of in vitro chondrogenesis by procollagen and collagen. Nature, Lond. 258, 327-330.
LASH, J. W. (1968). Somitic mesenchyme and its response to cartilage induction. In EpithelialMesenchymal Interactions (R. Fleischmajer and R. E. Billingham, eds), pp. 165-172.
Baltimore: Williams & Wilkins.
LASH, J. W. & VASAN, N. S. (1977). Tissue interactions and extracellular matrix components.
In Cell and Tissue Interactions (ed. J. W. Lash & M. Burger), pp. 101-113. New York:
Raven Press.
LASH, J. W., BELSKY, E. & VASAN, N. S. (1977). Stimulation of chondrogenic differentiation
with extracellular matrix components: an analysis using scanning electron microscopy.
In Cell Interactions in Differentiation (ed. M. Karkinen-Jaaskelainen, L. Saxen & L. Weiss),
pp. 263-272. Academic Press.
LUGER, O. (1980). Specificity of cartilage induction in somitic and limb bud mesoderm.
Wilhelm Roux" Arch devl Biol. (in Press).
MEIER, S. & HAY, E. D. (1975). Stimulation of corneal differentiation by interaction between
cell surface and extracellular matrix. I. Morphogenetic analysis of transfilter induction.
/. Cell Biol. 56,275-291.
MILAIRE, J. & MLLNARD, J. (1968). Le role de l'epiblaste dans la chondrogenese du bourgeon
de membre chez la souris. /. Embryol. exp. Morph. 20, 215-236.
MINOR, R. R. (1973). Somite chondrogenesis. /. Cell Biol. 56, 27-50.
NORDLING, S., EKBLOM, P., LEHTONEN, E., SAXEN, L. & WARTIOVAARA, J. (1977). Kidney
tubule induction: physical and chemical interference. In Cell Interactions in Differentiation
(eds. M. Karkinen-Jaaskelainen, E. Saxen & L. Weiss), pp. 249-261. London: Academic
Press.
SAXEN, L. (1977). Morphogenetic tissue interactions: an introduction In Cell Interactions in
Differentiation (ed. M. Karkinen-Jaaskelainen, L. Saxen & L. Weiss), pp. 145-151. London:
Academic Press.
SMITH, A. A., SEARLS, R. L. & HILFER, R. S. (1975). Differential accumulation of extracellular
materials beneath the ectoderm during development of the embryonic chick limb and
flank regions. Devl Biol. 46, 222-236.
STRUDEL, G. (1971). Materiel extracellulaire et chondrogenese vertebrate. C.r. hebd. Seanc.
Acad. Sci., Paris 272, 473-476.
THESLEFF, I. (1977). Tissue interactions in tooth development in vitro. In Cell Interactions in
Differentiation (ed. M. Karkinen-Jaaskelainen, L. Saxen & L. Weiss), pp. 191-207. London:
Academic Press.
THESLEFF, I. (1978). Role of the basement membrane in odontoblast differentiation. /. biol.
Buccale 6, 241-249.
THESLEFF, I., LEHTONEN, E. & SAXEN, L. (1978). Basement membrane formation in transfilter tooth culture and its relation to odontoblast differentiation. Differentiation 10, 71-79.
THESLEFF, I., STENMAN, S., VAHERI, A. & TIMPL, R. (1979). Changes in the matrix proteins,
fibronectin and collagen, during differentiation of mouse tooth germ. Devl Biol. 70, 116—
126.
WARTIOVAARA, J., LEHTONEN, E., NORDLING, S. & SAXEN, L. (1972). Do membrane filters
prevent cell contacts? Nature, Lond. 238, 407-408.
HOLTZER,
Ectoderm and mesoderm interactions in chick limb bud
173
(1961). Utilisation de la membrane vitelline de l'oeuf de Poule en culture organotypique. I. Technique et possibility. Devi Biol. 3, 767-786.
WOLFF, ET. & HAFFEN, K. (1952). Sur une methode de culture d'organes embryonnaires
in vitro. Texas Rep. Biol. Med. 10, 463-472.
ZWILLING, E. (1955). Ectoderm-mesoderm relationship in the development of the chick
embryo limb bud. /. exp. Zool. 128, 423-442.
WOLFF, ET.
{Received 2 October 1979, revised 1 March 1980)
EMB 59