Download Mesoderm induction in Xenopus laevis:responding

609
Development 104, 609-618 (1988)
Printed in Great Britain © The Company of Biologists Limited 1988
Mesoderm induction in Xenopus laevis: responding cells must be in
contact for mesoderm formation but suppression of epidermal
differentiation can occur in single cells
K. SYMES, M. YAQOOB and J. C. SMITH
Laboratory of Embryogenesis, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK
Summary
When Xenopus embryos are cultured in calcium- and
magnesium-free medium (CMFM), the blastomeres
lose adhesion but continue dividing to form a loose
heap of cells. If divalent cations are restored at the
early gastrula stage the cells re-adhere and eventually
form muscle (a mesodermal cell type) as well as
epidermis. If, however, the cells are dispersed during
culture in CMFM, muscle does not form following
reaggregation although epidermis does. This suggests
that culturing blastomeres in a heap allows the transmission of mesoderm-induction signals from cell to cell
while dispersion effectively dilutes the signal.
In this paper, we have attempted to substitute for
cell proximity by culturing dispersed blastomeres in
XTC mesoderm-inducing factor (MIF). We find that
dispersed cells do not respond to XTC-MIF by forming mesodermal cell types after reaggregation, but the
factor does inhibit epidermal differentiation. One
interpretation of this observation is that an early stage
in mesoderm induction is the suppression of epidermal
differentiation and that formation of mesoderm may
require contact-mediated signals that are produced in
response to XTC-MIF.
We have gone on to study the suppression of
epidermal differentiation in more detail. We find that
this is a dose-dependent phenomenon that can occur in
single cells in the absence of cell division. Animal pole
blastomeres become more difficult to divert from
epidermal differentiation at later stages of development and by stage 12 they are 'determined' to this
fate. Fibroblast growth factor (FGF) also suppresses
epidermal differentiation in isolated animal pole
blastomeres and transforming growth factor-/fr acts
synergistically with FGF in doing so.
Introduction
only the ectoderm and endoderm markers were
detected: muscle-specific actin genes were not
expressed.
These results suggest that culture of cells in a heap
allows signals to pass from cell to cell which result in
the specification of mesoderm. Such signals are likely
to be identical to those transmitted in the more
familiar demonstration of mesoderm induction, in
which animal and vegetal pole regions of the blastula
are juxtaposed. This results in the formation of
mesoderm by the animal pole component of the
combination (Nieuwkoop, 1969, 1973; Sudarwati &
Nieuwkoop, 1971; Dale et al. 1985; Gurdon et al.
1985; Smith et al. 1985).
Recently, soluble factors that induce isolated ani-
The clearest evidence that the mesoderm of Xenopus
laevis arises through cell-cell interactions comes from
work by Gurdon et al. (1984) and Sargent et al.
(1986). In these experiments, Xenopus embryos were
disaggregated into single cells by culture in calciumand magnesium-free medium (CMFM) from the fertilized egg stage to the beginning of gastrulation. At
this point divalent cations were returned to the
medium and the cells allowed to reaggregate. If the
cells were kept in a heap during culture in CMFM,
they subsequently expressed gene products characteristic of all three germ layers: ectoderm, mesoderm
and endoderm. If, however, the cells were dispersed,
Key words: mesoderm induction, Xenopus laevis,
epidermal differentiation, MIF, fibroblast growth factor
(FGF), transforming growth factor-/?!.
610
K. Symes, M. Yaqoob and J. C. Smith
mal pole ectoderm to form mesoderm have been
discovered. One of these is a protein derived from the
Xenopus XTC cell line (Smith, 1987). This XTC
mesoderm-inducing factor (MIF) has recently been
purified and shown to be a protein of apparent
relative molecular mass (Mr) 23500. On reduction
XTC-MIF loses its activity and polyacrylamide gel
electrophoresis reveals one or two bands of approximately 15000Mr (Smith et al. 1988). XTC-MIF is
active at picomolar concentrations and may be related to TGF-£ (Smith et al. 1988; see below). A
variety of embryological experiments suggest that it is
similar to a natural mesoderm inducer (Symes &
Smith, 1987; Cooke et al. 1987). Another group of
MIFs consists of the heparin-binding growth factors,
such as fibroblast growth factor (FGF) (Slack et al.
1987). These tend to induce ventral mesoderm cell
types rather than dorsal (Slack etal. 1987), although if
transforming growth factor-/31 (TGF-/S1) is present,
differentiation of muscle, a dorsal cell type, is
enhanced (Kimelman & Kirschner, 1987). TGF-j81
alone has no mesoderm-inducing activity (Slack et al.
1987; Kimelman & Kirschner, 1987), but the closely
related TGF-/J2 does induce muscle from isolated
Xenopus animal pole regions (Rosa et al. 1988a).
If, as suggested above, mesoderm is formed after
culture of disaggregated blastomeres in a heap because their close proximity allows them to pass MIFs
to each other, we should expect that culture of
dispersed cells in exogenous MIF would have the
same effect. In this paper, we test this prediction. Our
results indicate that culture of dispersed cells in XTCMIF cannot substitute for cell proximity, because the
cells do not go on to form notochord, muscle or
blood. However, the factor is not without effect, for it
suppresses epidermal differentiation both in whole
embryos that are reaggregated after dispersion and in
dispersed animal pole cells. We suggest that an early
stage in mesoderm induction is the suppression of
epidermal differentiation in the same way that suppression of epidermis formation is an early stage in
neural induction (Jamrich et al. 1987). Formation of
mesoderm may require further contact-mediated signals that are produced in response to XTC-MIF.
Materials and methods
Embryos
Embryos of Xenopus laevis were obtained by artificial
fertilization as described by Smith & Slack (1983). They
were chemically dejellied using 2 % cysteine hydrochloride
(pH 7-8-8-1), washed, and transferred to Petri dishes
coated with 1% Noble Agar and containing 10% normal
amphibian medium (NAM: Slack, 1984). The embryos
were staged according to Nieuwkoop & Faber (1967).
Mesoderm-inducing factors
Serum-free conditioned medium from the XTC cell line
(Pudney et al. 1973) was prepared as described by Smith
(1987) except that the cells were grown on glass roller
bottles. For most experiments, XTC-MEF was partially
purified from heated conditioned medium by DEAESepharose chromatography followed by Phenyl Sepharose
chromatography (see Cooke et al. 1987). In these experiments, mesoderm-inducing activity was quantified as
described by Cooke et al. (1987) and Godsave et al. (1988).
We define one unit of mesoderm-inducing activity as the
minimum quantity that must be present in 1 ml medium for
induction to occur. The sample of partially purified XTCMIF used for these experiments contained approximately
7-7X103 units mg"1 protein. In some experiments, electrophoretically homogeneous XTC-MIF with a specific activity
of SxlC^unitsmg"1 was used (Smith et al. 1988).
Recombinant basic fibroblast growth factor (bFGF) was
supplied by BCL. A sample purified from bovine brain was
the generous gift of Dr J. Heath. Transforming growth
factor-^1 (TGF-/51) was kindly donated by Drs M. Sporn
and H. Moses. TGF-/31 alone had no inducing activity, but
it enhanced the effect of bFGF as described by Kimelman &
Kirschner (1987).
Embryo culture and disaggregation
Embryos to be disaggregated were transferred to agarcoated 35 mm Petri dishes containing calcium- and
magnesium-free medium (CMFM) prepared according to
Sargent et al. (1986). The vitelline membranes of these
embryos were removed between the 1- and 32-cell stages
and the cells were either left undisturbed to divide and form
a compact heap, or were gently separated and dispersed by
passage of medium from a Pasteur pipette. The cells of
dispersed cultures were kept apart by repeating this action
every 30min.
When control embryos reached stage 10-10i, the
beginning of gastrulation, calcium chloride and magnesium
chloride were added to the cultures at final concentrations
of 1 ITIM. The cells of dispersed cultures were swirled to the
centre of the Petri dishes and they and the cells already in
heaps were allowed to adhere. In some experiments,
adhesion was facilitated by transferring the cells to the wells
of Nunc microwell plates. The wells had previously
received 50 jd of molten 1 % agarose, which sets to form a
shallow cup. The cells roll down the walls of the cup to form
a clump. The resulting 'embryos' were maintained at 18°C
for 3 days before being fixed.
In another series of experiments, blastula animal pole
regions were disaggregated into single cells and dispersed.
These cells were maintained in CMFM, in the presence or
absence of inducing factors, until control embryos reached
stage 19 (late neurula stage). They were then processed for
immunofluorescence as described below. In some cases,
animal cap cells were disaggregated and dispersed by
culture in CMFM containing lOjtgml"1 cytochalasin D
(Aldrich) (see Jones & Woodland, 1986). These cells were
cultured overnight until controls reached stage 19, when
they too were processed for immunofluorescence.
Mesoderm induction in Xenopus
611
Immunofluorescence analysis
Whole embryos or reaggregated groups of cells that were to
be analysed by immunofluorescence were fixed in 2%
trichloroacetic acid at 4°C for one hour to overnight. They
were dehydrated in ethanol and embedded in polyethylene
glycol 400 distearate (PEDS: Koch Chemicals) plus 1%
cetyl alcohol (Koch-Light) at 40cC (Dreyer et al. 1983).
Sections were cut at 10 ^m and brought to PBS-A through
an acetone series. The sections were analysed by indirect
immunofluorescence exactly as described by Dale et al.
(1985).
Single cell suspensions were dried onto microscope slides
at 25°C (F. M. Watt, personal communication). The dried
cells were fixed and permeabilized by immersion in methanol at -20°C for 5min and then analysed by indirect
immunofluorescence as described by Dale et al. (1985).
Antibodies
Four antibodies were used in this study. The monoclonal
antibody 12/101 (Kintner & Brockes, 1984) recognizes
Xenopus somite muscle. MZ15 is a monoclonal antibody
that recognizes keratan sulphate (Zanetti et al. 1985). It
reacts specifically with Xenopus chordocytes and particularly with the notochord sheath. From stage 27 onwards, it
also interacts with the ventral portion of the ear vesicle
lumen (Slack etal. 1985; Smith & Watt, 1985). RD35/3a is a
monoclonal antibody that, at stages 17 to 25, recognizes
exclusively the epidermis of Xenopus embryos (Godsave
etal. 1986). Finally, a rabbit antiserum was raised against
Xenopus globins (Cooke & Smith, 1987).
Results
Blastomeres cultured in a heap eventually form large
amounts of mesoderm: dispersed blastomeres do not
In preliminary experiments, we confirmed the
observations of Sargent et al. (1986), except that the
results were analysed by immunofluorescence of
specimens fixed after 3 days' culture, rather than by
cDNA probes. Xenopus embryos were disaggregated
between the 1- and 32-cell stages and cultured in a
heap until the early-to-mid gastrula stage (stage
10-10i). Divalent cations were then restored to the
culture and the reaggregated cells were fixed 3 days
later. Notochord was identified in at least 14 out of 34
cases (Fig. 1A,B) and muscle in 29 (Fig. 1C,D).
Differentiation of blood was not observed in the 6
cases that were examined. The mesodermal cell types
were at least partly surrounded by epidermis
(Fig. 1E,F).
When cells were dispersed during morula and
blastula stages, rather than being cultured in a heap,
mesoderm was formed less frequently and in smaller
amounts. Muscle, the most abundant mesodermal
cell type, was used to monitor mesoderm formation.
Notochord (which is easily recognized in histological
sections) was never observed and blood was not
Fig. 1. Mesodermal cell types are formed following
culture of Xenopus embryos in a heap in CMFM.
Disaggregated embryos were allowed to reaggregate at
stage 10 by addition of calcium and magnesium chlorides
to 1 mM each. The reaggregated embryos were fixed at
control stage 40 and processed for immunofluorescence
using antibodies recognizing notochord (A,B), muscle
(C,D) or epidermis (E,F). A, C and E are stained with
4',6-diamidino-2-phenylindole-dihydrochloride (DAPI) to
show nuclei. B, D and F show tissue-specific
fluorescence. Scale bar in B is lOO^m, and also applies to
A. Scale bar in F is 200/im, and also applies to C, D and
E.
detected in the 5 cases examined with the anti-globin
antibody. Least muscle was observed if embryos were
reaggregated at the crescent-shaped blastopore stage
(stage 10i), or if they were disaggregated before the
8-cell stage and reaggregated at the early gastrula
stage (stage 10-10J). In some cases, no muscle cells at
all could be identified with the 12/101 monoclonal
antibody (Fig. 2A,B) although epidermis was formed
(Fig. 2C,D); in others, there were scattered groups of
weakly fluorescent muscle cells (Fig. 2E,F). Significant amounts of muscle frequently formed, however,
if embryos were disaggregated later than the 16-cell
stage or reaggregated prior to stage 1(H (Fig. 2G,H).
These conclusions are based on a total of 104 cases
obtained in 9 separate experiments. The results differ
slightly from those of Sargent et al. (1986), in that a
mesodermal cell type can arise from Xenopus embryo
cells after a period of dispersion during blastula stages
followed by reaggregation at the early gastrula stage.
However, the difference is probably due to minor
inconsistencies in interpreting the stage series of
Nieuwkoop & Faber (1967), since in our hands
612
K. Symes, M. Yaqoob and J. C. Smith
derm-inducing signals from cell to cell, while dispersion effectively dilutes the signal and prevents induction from occurring. In an attempt to substitute for
cell proximity, we therefore cultured dispersed
blastomeres in various concentrations of partially
purified XTC-MIF. At stage 10-10i divalent cations
were returned to the media, the cells were reaggregated, and they were cultured for 3 days. They were
then fixed and examined as before.
In 5 experiments, involving a total of 35 reaggregated embryos, we never observed significant mesoderm formation. Small amounts of muscle were
observed in 4 cases, but no more than was formed
after dispersion in the absence of inducing factor (see
above). Blood and notochord were never detected.
Partially purified XTC-MIF was used at concentrations of 260-1690 ngmP 1 (2-0-13-0 units ml" 1 ),
which almost invariably induces muscle from intact
animal pole regions (see Cooke et al. 1987; Smith et
al. 1988). Fig. 3A,B shows a typical negative result.
Fig. 2. Less mesoderm is formed if cells in CMFM are
dispersed during blastula and early gastrula stages.
Dispersed cells were reaggregated at the stage indicated
and cultured until control stage 40. They were then fixed
and sectioned and reacted with antibodies specific for
muscle (A,B,E,F,G,H) or epidermis (C,D). Panels A, E
and G show DAPI staining of panels B, F and H,
respectively. When dispersed blastomeres are
reaggregated at stage 10i muscle rarely forms (A,B), but
large amounts of epidermis differentiate (C,D).
Occasionally, isolated muscle cells are seen after
reaggregation at stage 10i (E,F). Large amounts of
muscle are formed if embryos are reaggregated at stage 9
(G,H). Scale bar in (B) is 200/im, and applies to all
panels.
virtually no muscle is formed if reaggregation is
carried out at what we call stage 10i, the crescentshaped blastopore stage.
We conclude, in agreement with Sargent et al.
(1986), that culture of blastomeres in a heap during
blastula and early gastrula stages permits mesoderm
formation, whereas dispersion of cells during these
stages does not.
Mesoderm-inducing factors cannot substitute for cell
proximity and induce mesoderm
Like Sargent et al. (1986), we reasoned that close
proximity of blastomeres allows the transfer of meso-
Mesoderm-inducing factors suppress epidermal
differentiation
Although XTC-MIF did not induce mesoderm from
blastomeres that had been dispersed between the
morula and gastrula stages, it was not without effect.
At all MIF concentrations, epidermis differentiation
was greatly reduced and usually completely inhibited
(compare Fig. 3C,D with Fig. 2C,D). Sargent et al.
(1986) have correlated such a depression in cytokeratin gene expression with mesoderm induction, pointing out that the induction process diverts cells away
from their original epidermal specification. The suppression of epidermal differentiation observed in
response to XTC-MIF may therefore represent an
early step in mesoderm induction that can occur in
single cells in calcium-free medium; further contactdependent steps may be required for mesodermal
differentiation. We discuss the possible nature of such
steps later; the next sequence of experiments concentrates on the suppression of epidermis formation by
mesoderm-inducing factors.
Single-cell analysis of suppression of epidermis
differentiation
The suppression of epidermal differentiation shown
in Fig. 3C,D was observed after 3 days in culture.
This is many hours and several cell divisions after the
initial specification of epidermis and expression of
epidermal markers (Slack, 1985; Jones & Woodland,
1986) and it is possible that events during this period
within the large cell mass (see Gurdon, 1987) reverse
initial epidermal differentiation. To investigate this,
epidermal differentiation was studied using an early
marker, in single cells, and in the absence of cell
division.
Mesoderm induction in Xenopus
613
Table 1. Proportion of cells expressing the
epidermal marker RD35/3a after disaggregation at
different stages
Proportion of cells expressing
RD35/3a (%)*
Stage of
disaggregation
7
8
9
9i
10
Experiment 1
Experiment 2
a
20-8
35-0
27-5
32-9
30-6
54
77
52
01
* A minimum of 500 cells was counted for each point.
lOO-i
9080-
Fig. 3. XTC-MIF does not substitute for cell proximity
and induce muscle from dispersed Xenopus blastomeres,
but it does suppress epidermal differentiation. Xenopus
embryos were cultured in CMFM from fertilization
onwards and blastomeres were kept dispersed by gentle
jets of medium from a Pasteur pipette. XTC-MIF was
added at control stage 1{ and the cells were reaggregated
at control stage 10i. The reaggregated embryos were
allowed to develop until control stage 40 when they were
fixed, sectioned and reacted with antibodies specific for
muscle (A,B) or epidermis (C,D). Panels A and C show
DAPI staining of panels B and D, respectively. Neither
muscle nor epidermis was formed. Panels E and F show
positive controls in which muscle- and epidermis-specific
antibodies, respectively, were reacted with sections of
embryos that had been cultured in CMFM in their
vitelline membranes before reaggregation. Scale bar in F
is 200/xm and applies to all panels.
In preliminary experiments, animal pole regions
were dissected from embryos at stages 7, 8,9,94 or 10
and disaggregated into single cells by culture in
CMFM. They were cultured in CMFM overnight at
18°C until controls reached stage 19, when they were
processed for indirect immunofluorescence using the
monoclonal antibody RD35/3a. The results from two
such experiments are shown in Table 1. The proportions of animal cap cells that expressed RD35/3a
varied from 20 % (Table 1) to almost 90 % (see
Fig. 4), depending on the egg batch. However, unlike
Jones & Woodland (1986), who have performed
similar experiments with a different epidermal
marker, we saw little effect of the stage at which the
cells were disaggregated on the proportion of positive
cells. This reflects the fact that Jones & Woodland's
marker, 2F7.C7, requires cell-cell contact for ex-
^
70-
'%
50-
\ \ \
\
\
a
•1 4 ° cd
1 30"
\
\
\
2010-
\
06-5
13 26
65 130 260 650 1300
Partially purified XTC-MIF (ngmP 1 )
Fig. 4. Epidermal differentiation of single animal pole
cells is inhibited by XTC-MIF in a dose-dependent and
stage-dependent manner. Animal cap cells from stage 8i
( • ) or stage 10i (O) were disaggregated and dispersed by
culture in CMFM. They were incubated in the indicated
concentrations of partially purified XTC-MIF
(7-7X10 3 unitsmg" 1 ) and processed for
immunofluorescence with RD35/3a after 16 h, when
controls were at stage 19. Notice that higher
concentrations of XTC-MIF are required to suppress
epidermis differentiation in stage 10i blastomeres than in
stage 8i blastomeres.
pression; by contrast, the keratins are expressed
autonomously (Sargent et al. 1986; see Smith et al.
1987). In view of the great variability between different egg batches, each subsequent experiment was
carried out with animal pole regions derived from a
single egg batch.
The next experiments investigated whether mesoderm-inducing factors can suppress epidermal differentiation in single cells. Animal pole explants were
disaggregated into single cell suspensions and cul-
614
K. Symes, M. Yaqoob and J. C. Smith
100
~l
90%
£
7060-
1 508. 40c
3
3020-
I 10*
0J
2
5
10
FGF(ngmP')
20
Fig. 5. Fibroblast growth factor inhibits epidermal
differentiation in dispersed animal pole blastomeres in a
dose-dependent fashion. TGF-j31 acts synergistically with
FGF. Stage-8 animal cap cells were disaggregated and
dispersed by culture in CMFM. They were incubated in
the indicated concentrations of FGF, with (O) or without
(•), 34ngmr 1 TGF-^l. Notice that TGF-/51 acts
synergistically with FGF in suppressing epidermal
differentiation but that it has no activity alone.
tured in CMFM in the presence of different concentrations of XTC-MIF or bFGF. They were processed
for immunofluorescence when controls reached stage
19. Both XTC-MIF (partially-purified and electrophoretically homogeneous) and FGF reduced the
proportions of cells expressing RD35/3a in a dosedependent manner (Figs 4-6), and the effective concentrations of inducing factors were similar to those
that caused mesoderm induction. Furthermore,
although TGF-/31 alone did not suppress epidermal
differentiation, it acted synergistically with FGF in
doing so (Fig. 5).
A similar suppression of epidermis formation could
be observed in single cells treated with cytochalasin
D, which inhibits cell division. In the absence of
inducing factors such cells express the RD35/3a
antigen (Fig. 6E,F) but expression is suppressed by
XTC-MIF (Fig. 6G,H).
Although XTC-MIF and bFGF suppressed epidermis formation in dispersed animal pole cells cultured
in CMFM, use of the monoclonal antibody 12/101
demonstrated that these factors never induced muscle
(data not shown).
Determination of epidermal differentiation
The proportion of cells expressing the RD35/3a
antigen is not greatly affected by the stage at which
they are disaggregated, but their response to mesoderm-inducing factors is. Cells disaggregated from
later-staged embryos require higher concentrations of
XTC-MIF to suppress epidermis formation. In one
experiment (Fig. 4), cells disaggregated at stage 8i
required 47ngml~ (0-4 units ml" 1 ) of partially purified XTC-MIF to halve the proportion of cells
expressing
the
RD35/3a
antigen,
whereas
lOOngml" 1 (0-8 units ml" 1 ) was needed for cells dis-
Fig. 6. Appearances of single cells cultured in the
absence (A,B,E,F) or presence (C,D,G,H) of XTC-MIF
and in CMFM (A-D) or CMFM containing lO^gml"1
cytochalasin D (E-H). The cells were reacted with the
antibody RD35/3a, which recognizes an epidermal
cytokeratin. Panels B, D, F and H show DAPI staining of
panels A, C, E and G, respectively. In the absence of
XTC-MIF most of the cells form epidermis (A,B), even
in the presence of cytochalasin (E,F). In the presence of
inducing activity, epidermal differentiation is suppressed
(C,D), even in the presence of cytochalasin (G,H). Scale
bar in H is 50 /xm and applies to all frames.
aggregated at stage 10£. Similar results have been
observed in two different experiments.
Beyond stage 10i, animal pole cells become even
more difficult to divert from epidermal differentiation
Mesoderm induction in Xenopus
a?
lOO-i
90-
Table 2. Induction by XTC-MIF does not suppress
protein synthesis in disaggregated animal pole cells
I 80° 70> 60•a
50-
8.
•S
S
U
403020100J
615
Incorporation of [35S]methiomne
(ctsmin"' ^g;"' protein)
10
I
11
ll|
Stage
Treatment
12
Fig. 7. Almost all animal pole blastomeres are committed
to epidermal differentiation by stage 12. Animal pole
blastomeres were obtained from embryos at the indicated
stages and dispersed by incubation in CMFM. They were
cultured in the absence (open bars) or presence (solid
bars) of 2-6/igml"1 partially-purified XTC-MIF
(20 units ml"'). At stage 10 XTC-MIF suppressed
epidermal differentiation in all the cells, but at later
stages increasing numbers of cells become refractory to
the factor.
and, by stage 12, almost all are determined to this fate
(Fig. 7). This is also the stage at which intact animal
pole regions are no longer able to respond to XTCMIF (Symes & Smith, in preparation), providing
further evidence that suppression of epidermis differentiation is an early step in mesoderm induction.
Suppression of epidermal differentiation is not a
toxic effect of inducing factors
The results described above show that the properties
of XTC-MIF and bFGF in suppressing epidermal
differentiation are correlated with their abilities to
induce mesoderm. Nevertheless, it is important to
demonstrate that the inhibition of epidermal differentiation is a specific effect of MIFs and not due to
toxicity. Thus, we find that the rate of cell division in
cultures maintained in MIFs was similar to that in
controls (compare cell sizes in Fig. 6A,B with those in
Fig. 6C,D) and the rate of incorporation of
[35S]methionine into trichloroacetic-acid-precipitable
material was actually enhanced (Table 2). The rate of
incorporation of [3H]uridine into TCA-precipitable
material appeared to be slightly inhibited by XTCMIF, but this was due to a reduction in uptake of the
labelled precursor rather than a real decrease in RNA
synthesis (Table 3). By contrast, bFGF caused an
increase in incorporation of [3H]uridine, but there
was an equivalent increase in uptake of the precursor.
Epidermal growth factor, which does not suppress
epidermal differentiation (data not shown), also
increased incorporation of [3H]uridine. Taken
together, these results suggest that the effect of
mesoderm-inducing factors in inhibiting epidermal
differentiation is not due to toxicity. In support of
this, we also find that animal pole cells that have been
Control
XTC-MTF (20 units/ml)
Without
cytochalasin
With
cytochalasin
1110
1490
590
790
Disaggregated cells were cultured in CMFM containing
10/igml" 1 cytochalasin D in the presence or absence of
2-6/igmi"1 (20units ml" 1 ) partially purified XTC-MIF.
[35S]methionine was added to 40/jCiml"1 and the cells were
incubated for 2h. At the end of this period the cells were
homogenized. One aliquot of the homogenate was used to
determine TCA-precipitable counts and the other was used for
protein determination by the method of Sargent (1987).
Table 3. The effects of XTC-MIF, FGF and EGF
on incorporation of [3H]uridine into disaggregated
animal pole cells
Incorporation of [3H]uridine into
TCA-insoluble material
Treatment
Control
XTC-MIF (20 units/ml)
FGF(100ng/ml)
EGF(100ng/ml)
ctsmin"' [ig~l
protein
ctsmin '/
total ctsmin"1
(xlO3)
598 ± 175
387 ± 33
931 ± 252
953 ± 291
30-8 ±3-2
29-8 ±1-1
33-7 ±5-0
38-8 ±0-9
Disaggregated animal pole cells were cultured in CMFM with
the indicated additions and 60/iCiml" 1 [3H]uridine. When
control embryos reached stage 13, the cells were washed three
times in CMFM and homogenized. One aliquot of the
homogenate was used to determine TCA-precipitable counts,
one to determine total counts and one was used for protein
determination by the method of Sargent (1987). Each
measurement was performed in triplicate and the values shown
are means ± standard errors.
induced by 3h contact with vegetal pole regions do
not express RD35/3a antigens after overnight culture
in CMFM (data not shown).
Discussion
The main conclusion to be drawn from this paper is
that mesoderm induction involves at least two steps.
The first can occur in single cells and results in the
suppression of epidermal differentiation. The second
requires cell-cell contact during induction and results
in mesoderm formation.
Mesoderm induction and the suppression of
epidermal differentiation
Jones & Woodland- (1986) have emphasized that the
616
K. Symes, M. Yaqoob and J. C. Smith
early Xenopus embryo consists of two cell types:
vegetal pole presumptive endoderm and animal pole
presumptive ectoderm. Mesoderm arises through an
inductive interaction in which vegetal pole tissue acts
on adjacent animal pole cells (reviewed by Smith,
1988). The early 'specification' (Slack, 1983) of the
entire animal hemisphere is therefore to form epidermis, and mesoderm induction must involve inhibition
of this mode of differentiation as well as the activation of mesoderm-specific genes. Our observations
suggest that these two processes occur by distinct
mechanisms, because inhibition of epidermal differentiation in response to mesoderm-inducing factors
can occur in a population of dispersed cells, whereas
if muscle is to differentiate, responding cells must be
in close proximity.
The inhibition of a certain mode of cell differentiation may be a general mechanism in development.
The next inductive interaction to which Xenopus
ectoderm is exposed is neural induction (Spemann &
Mangold, 1924; Smith & Slack, 1983; Gimlich &
Cooke, 1983; Kintner & Melton, 1987), in which
ectoderm overlying dorsal mesoderm is induced to
form nervous system. Here, Jamrich et al. (1987) have
shown that transcription of epidermis-specific cytokeratin genes commences in presumptive neural
tissue only to be terminated in response to neural
induction. At the same time, or perhaps subsequently, neural-specific gene transcription is activated (Kintner & Melton, 1987).
The suppression of epidermal differentiation in
response to MIFs satisfies two criteria defined by
Gurdon (1987), who proposed that the molecular
analysis of induction would be simplified if the
phenomenon could be studied in single cells, in the
absence of cell division. The Xenopus epidermal
cytokeratin genes have been cloned (Jonas et al.
1985), so it should be possible to use this system to
study their regulation.
Mesoderm formation requires cell contact during
induction
Our results indicate that responding cells must be in
contact during treatment with XTC-MIF if mesoderm
is to form. This suggests that animal pole blastomeres
produce additional, short-range, inductive signals in
response to the initial stimulus. Our prediction from
this is that if protein synthesis is inhibited during
induction of intact animal pole regions, and subsequently allowed to recover, mesoderm should not
form. This experiment has been carried out by Cascio
& Gurdon (1987), using animal-vegetal combinations, and they find that muscle-specific gene activation is indeed inhibited. However, all genes whose
activation occurs during gastrulation are inhibited,
whether or not their transcription depends on induc-
tion. Further evidence is therefore required for the
existence of additional mesoderm induction signals.
The idea that secondary cell interactions are
required for mesoderm formation after treatment
with inducing factors was first put forward by Minuth
& Grunz (1980), who proposed that the 'vegetalizing
factor' of Born et al. (1972) acts by inducing some of
the responding ectoderm cells to become endoderm,
and that these cells than induce mesoderm from
previously uninduced ectoderm. The conclusion that
the first action of the vegetalizing factor is to induce
endoderm was based on the observation that when
gastrula animal cap regions of Triturus alpestris are
exposed for 6h to a pellet of vegetalizing factor,
disaggregated for 20 h, and then reaggregated, they
differentiate as liver or intestine; that is, as endodermal structures.
If the situation in Xenopus is like that in the slowerdeveloping Triturus, then the suppression of epidermal differentiation seen in this paper would represent
a switch to endodermal differentiation. Vegetal pole
blastomeres of Xenopus differentiate poorly in culture and there is no reliable early marker of endodermal differentiation. The clone DG42 described as
endodermal by Sargent et al. (1986) is now known to
display a more complicated expression pattern (Rosa
et al. 1988b). We therefore intend to investigate this
question using the single cell transfer technique of
Heasman et al. (1984) and Wylie et al. (1987).
However, one argument that Xenopus animal pole
blastomeres are not specified as endoderm is that the
inducing activity of Xenopus vegetal pole regions is
lost at stage 10 or soon afterwards (Nakamura et al.
1970; Dale et al. 1985; Gurdon et al. 1985; Jones &
Woodland, 1987) while Xenopus animal pole regions
are still responsive to XTC-MIF at this stage and even
later (Symes & Smith, in preparation). If XTC-MIF
were to act by converting some ectoderm cells to
endoderm, even if this transformation was instantaneous, there would barely be enough signalling
time for the subsequent mesoderm induction to
occur. An alternative suggestion is that the shortrange inductive signal we propose is that which is also
responsible for homeogenetic induction, which has
recently been observed in Xenopus (Cooke et al.
1987) and is known to occur in other amphibian
species (Kaneda, 1981; Kurihara & Sasaki, 1981;
Kaneda & Suzuki, 1983).
We are grateful to Drs J. Heath, H. Moses and M. Sporn
for gifts of growth factors, to Dr J. Brockes for 12/101, to
Dr F. Watt for MZ15, and to Prof. C. Wylie for RD35/3a.
We also thank Jonathan Cooke for his comments.
References
BORN, J., GEITHE, H. P., TIEDEMANN, H., TIEDEMANN, H.
Mesoderm induction in Xenopus
& KOCHER-BECKER, U. (1972). Isolation of a
vegetalizing inducing factor. Z. Physiol. Chem. 353,
1075-1084.
CASCIO, S. & GURDON, J. B. (1987). The initiation of new
gene transcription during Xenopus gastrulation requires
immediately preceding protein synthesis. Development
100, 297-305.
COOKE, J. & SMITH, J. C. (1987). The midblastula cell
cycle transition and the character of mesoderm in u.v.induced nonaxial Xenopus development. Development
99, 197-210.
COOKE, J., SMITH, J. C , SMITH, E. J. & YAQOOB, M .
(1987). The organization of mesodermal pattern in
Xenopus laevis: experiments using a Xenopus
mesoderm-inducing factor. Development 101, 893-908.
DALE, L., SMITH, J. C. & SLACK, J. M. W. (1985).
Mesoderm induction in Xenopus laevis: a quantitative
study using a cell lineage label and tissue-specific
antibodies. J. Embryol. exp. Morph. 89, 289-312.
DREYER, C , WANG, W. H., WEDLICH, D. & HAUSEN, P.
(1983). Oocyte nuclear proteins in the development of
Xenopus. In Current Problems in Germ Cell
Differentiation (ed. A. McLaren & C. C. Wylie), pp.
329-352. Cambridge & London: Cambridge University
Press.
GIMUCH, R. L. & COOKE, J. (1983). Cell lineage and the
induction of second nervous systems in amphibian
development. Nature, Lond. 306, 471-473.
GODSAVE, S. F., ANDERTON, B. H. & WYLIE, C. C.
(1986). The appearance and distribution of
intermediate filament proteins during differentiation of
the central nervous system, skin and notochord of
Xenopus laevis. J. Embryol. exp. Morph. 97, 201-223.
GODSAVE, S. F., ISAACS, H. V. & SLACK, J. M. W.
(1988). Mesoderm-inducing factors: a small class of
molecules. Development 102, 555-566.
GURDON, J. B. (1987). Embryonic induction-molecular
prospects. Development 99, 285-306.
GURDON, J. B., BRENNAN, S., FAIRMAN, S. & MOHUN, T.
J. (1984). Transcription of muscle-specific actin genes
in early Xenopus development: nuclear transplantation
and cell dissociation. Cell 38, 691-700.
GURDON, J. B., FAIRMAN, S., MOHUN, T. J. & BRENNAN,
S. (1985). The activation of muscle-specific actin genes
in Xenopus development by an induction between
animal and vegetal cells of a blastula. Cell 41, 913-922.
HEASMAN, J., WYLIE, C. C , HAUSEN, P. & SMITH, J. C.
(1984). Fates and states of determination of single
vegetal pole blastomeres of Xenopus laevis. Cell 37,
185-194.
JAMRICH, M., SARGENT, T. D. & DAWID, I. B. (1987).
Cell-type-specific expression of epidermal cytokeratin
genes during gastrulation of Xenopus laevis. Genes &
Development 1, 124-132.
JONAS, E., SARGENT, T. D. & DAWID, I. B. (1985).
Epidermal keratin gene expressed in embryos of
Xenopus laevis. Proc. natn. Acad. Sci., U.S.A. 82,
5413-5417.
JONES, E. A. & WOODLAND, H. R. (1986). Development
of the ectoderm in Xenopus: Tissue specification and
the role of cell association and division. Cell 44,
617
345-355.
E. A. & WOODLAND, H. R. (1987). The
development of animal cap cells in Xenopus: a measure
of the start of animal cap competence to form
mesoderm. Development 101, 557-563.
KANEDA, T. (1981). Studies on the formation and state of
determination of the trunk organizer in the newt
Cynops pyrrhogaster. III. Tangential induction in the
dorsal marginal zone. Dev. Growth Diff. 23, 553-564.
JONES,
KANEDA, T. & SUZUKI, A. S. (1983). Studies on the
formation and state of determination of the trunk
organizer in the newt, Cynops pyrrhogaster. IV. The
association of the neural-inducing activity with the
mesodermization of the trunk organizer. Wilhelm
Roux's Arch, devl Biol. 192, 8-12.
KIMELMAN, D. & KIRSCHNER, M. (1987). Synergistic
induction of mesoderm by FGF and TGF-/3 and the
identification of FGF in the early Xenopus embryo.
Cell 51, 869-877.
KINTNER, C. R. & BROCKES, J. P. (1984). Monoclonal
antibodies identify blastemal cells derived from
dedifferentiating muscle in newt limb regeneration.
Nature, Lond. 308, 67-69.
KINTNER, C. R. & MELTON, D. A. (1987). Expression of
Xenopus N-CAM RNA in ectoderm is an early
response to neural induction. Development 99,
311-325.
KURIHARA, K. & SASAKI, N. (1981). Transmission of
homoiogenetic induction in presumptive ectoderm of
newt embryo. Dev. Growth Diff. 23, 361-369.
MINUTH, M. & GRUNZ, H. (1980). The formation of
mesodermal derivatives after induction with
vegetalizing factor depends upon secondary cell
interactions. Cell Diff. 9, 229-238.
NAKAMURA, O., TAKASAKI, H. & ISHIHARA, M. (1970).
Formation of the organizer from combinations of
presumptive ectoderm and endoderm. I. Proc. Japan
Acad. 47, 313-318.
NIEUWKOOP, P. D. (1969). The formation of mesoderm in
Urodelean amphibians. I. Induction by the endoderm.
Wilhelm Roux's Arch. EntwMech. Org. 162, 341-373.
NIEUWKOOP, P. D. (1973). The "organization centre" of
the amphibian embryo, its origin, spatial organization
and morphogenetic action. Adv. Morphogen. 10, 1-39.
NIEUWKOOP, P. & FABEK, J. eds (1967). Normal Table of
Xenopus laevis (Daudin). 2nd ed. Amsterdam: NorthHolland.
PUDNEY, M., VARMA, M. G. R. & LEAKE, C. J. (1973).
Establishment of a cell line (XTC-2) from the South
African clawed toad, Xenopus laevis. Experientia 29,
466-467.
ROSA, F., ROBERTS, A. B., DANIELPOUR, D., DART, L. L.,
SPORN, M. B. & DAWID, I. B. (1988a). Mesoderm
induction in amphibia: the role of TGF-/32-like factors.
Science 239, 783-785.
ROSA, F., SARGENT, T. D., REBBERT, M. L., MICHAELS,
G. S., JAMRICH, M., GRUNZ, H., JONAS, E., WINKLES,
J. A. & DAWID, I. B. (1988fc). Accumulation and decay
of DG42 gene products follow a gradient pattern
during .Xenopws embryogenesis. Devi Biol. 129,114-123.
SARGENT, M. G. (1987). Fiftyfold amplification of the
618
K. Symes, M. Yaqoob and J. C. Smith
Lowry protein assay. Anal. Biochem. 163, 476-481.
T. D., JAMRICH, M. & DAWID, I. B. (1986).
Cell interactions and the control of gene activity during
early development of Xenopus laevis. Devi Biol. 114,
238-246.
SLACK, J. M. W. (1983). From Egg to Embryo.
Cambridge and London: Cambridge University Press.
SLACK, J. M. W. (1984). Regional biosynthetic markers in
the early amphibian embryo. J. Embryol. exp. Morph.
80, 289-319.
SLACK, J. M. W. (1985). Peanut lectin receptors in the
early amphibian embryo: Regional markers for the
study of embryonic induction. Cell 41, 237-247.
SARGENT,
SLACK, J. M. W., CLEINE, J. H. & SMITH, J. C. (1985).
Regional specificity of glycoconjugates in Xenopus and
axolotl embryos. J. Embryol. exp. Morph. 89
Supplement 137-153.
SLACK, J. M. W., DARLINGTON, B. G., HEATH, J. K. &
GODSAVE, S. F. (1987). Mesoderm induction in early
Xenopus embryos by heparin-binding growth factors.
Nature, Lond. 326, 197-200.
SMITH, J. C. (1987). A mesoderm-inducing factor is
produced by a Xenopus cell line. Development 99,
3-14.
SMITH, J. C. (1988). Cellular interactions in establishment
of regional patterns of cell fate during development. In
Developmental Biology: A Comprehensive Synthesis,
vol. 5 (ed. L. Browder), pp. 79-125. New York &
London: Plenum Press.
SMITH, J. C , DALE, L. & SLACK, J. M. W. (1985). Cell
lineage labels and region-specific markers in the
analysis of inductive interactions. /. Embryol. exp.
Morph. 89 Supplement 317-331.
SMITH, J. C. & SLACK, J. M. W. (1983). Dorsalization and
neural induction: properties of the organizer in
Xenopus laevis. J. Embryol. exp. Morph. 78, 299-317.
SMITH, J. C , SYMES, K., HEASMAN, J., SNAPE, A. &
WYLIE, C. C. (1987). The Xenopus animal pole
blastomere. BioEssays 7, 229-234.
J. C. & WATT, F. M. (1985). Biochemical
specificity of Xenopus notochord. Differentiation 29,
109-115.
SMITH,
SMITH, J. C , YAQOOB, M. & SYMES, K. (1988).
Purification, partial characterization and biological
effects of the XTC mesoderm-inducing factor.
Development 103, 591-600.
SPEMANN, H. & MANGOLD, H. (1924). Uber Induktion
von Embryonenanlagen durch Implantation artfremder
Organisatoren. Wilhelm Roux's Arch EntwMech. Org.
100, 599-638.
SUDARWATI, S. & NIEUWKOOP, P. D. (1971). Mesoderm
formation in the Anuran Xenopus laevis (Daudin).
Wilhelm Roux's Arch. EntwMech. Org. 166, 189-204.
SYMES, K. & SMITH, J. C. (1987). Gastrulation
movements provide an early marker of mesoderm
induction in Xenopus laevis. Development 101,
339-349.
WYLIE, C. C , SNAPE, A., HEASMAN, J. & SMITH, J. C.
(1987). Vegetal pole cells and commitment to form
endoderm in Xenopus laevis. Devi Biol. 119, 496-502.
ZANETTI, M., RATCLIFFE, A. & WATT, F. M. (1985). Two
subpopulations of differentiated keratinocytes
identified with a monoclonal antibody to keratan
sulphate. J. Cell Biol. 101, 53-59.
(Accepted 20 September 1988)