/ . Embryo/, exp. Morph. Vol. 33, 4, pp. 957-967, 1975
Printed in Great Britain
957
Xenopus laevis cement gland as an experimental
model for embryonic differentiation
I. In vitro stimulation of differentiation by ammonium chloride
By J. J. PICARD 1
From the Laboratoire cTEmbryologie, Louvain-la-Neuve
SUMMARY
Ectoblastic cells explanted from the animal pole of young Xenopus laevis gastrulae have
been cultured in vitro. When these cells were cultured for five days in standard salt solutions
they formed atypical epidermis. When they were first submitted for 6 h to Holtfreter solution
containing ammonium chloride and then transferred for five days in standard Barth's solution
they underwent differentiation into typical cement gland tissue.
The optimum concentration of ammonium chloride was 10mM. Below and above this
concentration the resulting cement glands had a smaller volume.
The optimum duration for the initial stimulation with 10 mM ammonium chloride in
Holtfreter solution was 6 h. Shorter stimulation times produced only small cement glands.
Longer initial incubations in ammonium chloride resulted in progressive dissociation of the
explants.
To obtain the best differentiation into cement gland it appeared that the pH of the
ammonium chloride solution should be between 7-5 and 7-7.
When the stimulation is performed under these optimum conditions the cement gland
measured after five days of culture accounts for 80-90 % of the explanted tissue. This means
that all or almost all of the competent superficial layer of the ectoblastic cells underwent
differentiation into cement gland. No other differentiated tissue was observed in the explants.
The cement gland is a very simple organ containing only one single cell type. The gland
obtained under the described in vitro conditions is therefore proposed as an experimental
model for biochemical studies on early embryonic differentiation.
INTRODUCTION
Many experimental models, suitable for biochemical studies on cell differentiation, have been described. However, very few in vitro systems homogeneous
with respect to cell population are available for the study of early embryonic
differentiation.
In the following series we will describe a new experimental model, i.e. the
cement gland of Xenopus laevis (Daudin). Differentiation of this gland may be
induced in vitro from early gastrula cells by using a simple starting stimulus,
ammonium chloride. Under defined conditions, most if not all of the stimulated
cells undergo differentiation, and the cement gland is the only tissue to develop
1
Author's address: Laboratoire d'Embryologie, Sciences 12, 1348 Louvain-la-Neuve,
Belgium.
958
J. J. PICARD
from the superficial layer of the ectoderm. The gland is very simple, containing
only a single cell type. Differentiation is well characterized both morphologically
and biochemically.
The discovery of the ability of ammonium chloride to stimulate the differentiation of the cement gland was the result of experiments designed to test the
hypothesis that neural induction might be mediated by intracellular ionic
modifications in the receptive cells. In these experiments undifferentiated ectoblastic cells of the gastrula were incubated in media containing various
concentrations of the chloride salts of alkaline metals as well as of the related
ammonium. None of these tissues produced neural structures, but the explants
incubated in ammonium chloride underwent an almost complete differentiation
into cement gland.
The following series of articles will describe the experimental conditions for
the optimal differentiation of the gland and will show that this model may be
used for biochemical studies on early embryonic differentiation.
MATERIAL AND METHODS
Adult male and female Xenopus were purchased from the South African
Snake Farm, Fish Hoek, South Africa. They were maintained in tap water at
room temperature and fed twice a week with diced horse liver.
Mating was induced by injecting the male and the female with 100 i.u. chorionic gonadotropin (Pregnyl, N.V. Organon-OSS, Netherland). On the next day
the male and the female received further 250 i.u. and 300 i.u. doses respectively.
The eggs were shed and fertilized overnight. The stages of development are
described according to Nieuwkoop & Faber (1967).
About 500 embryos at stages 8 and 9 were chemically dejellied according to
Dawid (1965). They were washed ten times in sterile Barth's solution (Barth &
Barth, 1959) containing 500 i.u. of penicillin and 500 /ig of streptomycin per ml.
Finally, they were rinsed with the same solution containing 100 i.u. penicillin
and 100 /tg streptomycin per ml and evenly distributed among four Petri dishes.
All the following operations were performed in a culture room maintained
at 18 °C±0-2. Fifteen stage-10 embryos were collected in a small Petri dish,
their vitelline membranes were removed with sharpened watchmaker forceps
and the required explants were dissected with tungsten needles.
The explanted region is depicted in Fig. 1. A cup was dissected containing
the major part of the ectoblast and extending from about 20° below the animal
pole dorsally to about 70° below the animal pole ventrally. The 15 explants
were briefly rinsed in the test solution and transferred to a small vessel containing
2 ml of the same solution. After incubation for the required time in the test
solution, the explants were washed in Barth's saline, containing 100 i.u. penicillin
and 100 /tg streptomycin, and cultured for 5 days in the same solution. About
150 explants can be processed within a day according to this method.
In vitro studies on Xenopus cement gland. I
959
P.I A.
20°
Fig. 1. Diagram of a median section through a stage-10 gastrula indicating the position of the explanted ectoblastic cup. The superficial layer is finely dotted. P.A.,
Animal pole; P. V., vegetal pole; Blp., dorsal lip of the blastopore; Blc, blastocoelic
cavity.
The explants were fixed in Bouin solution and serial paraffin sections were
stained according to Cleveland & Wolfe (Romeis, 1948). The volumes of the
individual tissues were measured in each explant by the method of Hennig &
Meyer-Arendi (1963).
RESULTS
(1) Explants cultured in standard salt solutions
Stage-10 ectoblast cells were incubated for 6 h in one of the following standard
solutions, (a) Barth solution (pH 8) (Barth & Barth, 1959); (b) normal Holtfreter
solution (pH 7-8) (Holtfreter, 1931); (c) Holtfreter solution modified to contain
10 mM NaHCO 3 (pH 8-2). After this initial incubation, they were all cultured
for 5 days in Barth's solution.
In all three instances, the explants became mainly or completely epidermal.
The disorderly mass of this atypical epidermis has often been described (see
Holtfreter, 1945). In a few explants (4 cases in 50), small cement glands could
be observed (Fig. 2). The volume of these glands accounted for only 0-01 %
to 5 % of the corresponding explant. This amount is negligible with regard to
the total mass of the 50 cultured explants (< 0-1 %). No other differentiated
tissue was observed.
960
J. J. PICARD
50/an
Fig. 2. Histological section of an explant cultured for 6 h in Holtfreter solution and
5 days in Barth's solution. The whole explant is composed of undifferentiated tissue
except for a small cement gland containing one single cell (arrow). Erythrosine,
orange G, anilin (x 225).
(2) Influence of ammonium chloride concentration in the stimulation medium
Explants were incubated for 6 h in the modified Holtfreter solution containing increasing concentrations of ammonium chloride. The pH of these
solutions was 7-6. For ammonium chloride concentrations above 20 mM, the
solutions were first neutralized with the required amount of NaOH 0-1 JV and
the pH was raised to 7-6 by 10 mM NaHCO 3 . After this stimulation, the explants
were transferred to Barth saline for 5 days. They were then processed for
histological examination and the volumes of the tissues were measured.
The results of this series of experiments are given in Fig. 3. On average, the
total volume of the explants remained approximately constant over the range
of concentrations used during the stimulation period. The slope is not significantly different from zero at the 5 % level (Fig. 3 A).
The mean volume of the cement gland in each explant (Fig. 3 B) increased
rapidly when the stimulation occurred at concentrations which rose from 0 to
10 mM ammonium chloride. Above the latter concentration the observed volume
was progressively less important.
The percentage of the total explanted tissue occupied by the cement gland is
given in Fig. 3 C. The curve has the same pattern as in Fig. 3 B, but the highest
percentage was obtained at 5 mM ammonium chloride. At this concentration
about 90 % of the explanted tissue was cement gland (Fig. 4).
In vitro studies on Xenopus cement gland. I
3U
961
B
40 _
T
J>39
30 -
f\
T/
12f Tl3\
20
10
10.
20
30
40
-h
50
Ammonium chloride (ITIM)
^^419
l
|
I
i
i
?
10
20
30
40
50
60
Ammonium chloride (ITIM)
E
2 0 -
10 -
10
20
30
40
50
Ammonium chloride (ITIM)
10
20
30
40
50
60
Ammonium chloride (ITIM)
Fig. 3. Influence of ammonium chloride concentration in the stimulation medium.
The incubation time was 6 h. The volumes were measured after 5 days of culture in
Barth's solution. The ordinate indicates the average volume of several explants
whose number is given. Vertical bars are the confidence limits at 95 %. The curves
are hand-drawn, except for A where it was drawn by the method of least squares.
For each explant the total volume is the sum of the volume of the cement gland and
of the undifferentiated tissue. A, total volume of the explanted tissue; B, volume
of the cement gland; C, relative volume of the cement gland; D, volume of the undifferentiated tissue. The filled and hollow circles and triangles refer to different
egg-batches.
The gland was the only tissue to differentiate in these explants. The remaining
tissue was atypical epidermis. The amount of this undifferentiated tissue fell
rapidly when the concentration was raised from 0 to 5 mM ammonium chloride.
At higher concentrations, the undifferentiated tissue increased progressively
(Fig. 3D).
(3) Influence of the duration of stimulation by ammonium chloride
Explants were first stimulated during a definite period with 10 mM ammonium chloride in the modified Holtfreter solution (pH 7-6). They were then
962
J. J. PICARD
200 fim
Fig. 4. Histological section of an explant stimulated for 6 h in Holtfreter solution containing 10 mM ammonium chloride and cultured for 5 days in Barth's solution. The
tissue is almost completely cement gland covered by a single layer offlatcells (arrow).
Erythrosine, orange G, anilin (x 46).
cultured for 5 days in Barth's solution and the volumes of the observed tissues
were measured.
The mean total volume per explant remained constant when the duration of
stimulation was lower or equal to 7 h (Fig. 5 A). For longer stimulations, the
volume of the explants diminished rapidly, due to cell dissociation. The dissociated cells rounded up, did not divide nor differentiate and remained loaded
with yolk platelets. Dissociation of the explants mostly continued during the
first three days of culture. A 15-5 h stimulation left only two small explants out
of 15 remaining after 5 days of culture. Stimulations of 20 and 25 h produced
complete dissociation of all the explants.
The explants contained only small cement glands when the stimulation had
been shorter or equal to 4 h (Fig. 5B). The optimal stimulation is about 6 h.
Longer incubation times resulted in smaller volumes of the gland, due to
progressive dissociation of the explants.
The percentage of the explants occupied by the cement gland rose when the
duration of stimulation was increased. From 6 to 11 h of stimulation this
percentage remained at values of about 80 % (Fig. 5C).
No other differentiated tissue was observed in the explants. The amount of
undifferentiated tissue decreased as a function of the duration of the stimulation
In vitro studies on Xenopus cement gland. I
50 _
T .
40
[39
963
A
ill 17
30
A
20
\
\
"K
10
1
k
1
10
2/15
15
-----jl6
20
15.
25
5
Duration of the stimulation (h)
100
50
C
15
20
D
40
f\Y 16
-
10
Duration of the stimulation (h)
15
30
50 _
20 _
.15
\.39
M3
'As
I
5
i
10
10
I
I
i
15
20
25
Duration of the stimulation (h)
1
^15
i
i
i
5
i
i
i
2/15
1 1 4-4-1.1 ' * '
10
15
i
16
15
20
25
i i i i i i i i
Duration of the stimulation (h)
Fig. 5. Influence of the duration of stimulation by 10 mM ammonium chloride. The
general conditions and the symbols are the same as in Fig. 3. The four curves are handdrawn.
(Fig. 5D). This is the result of both differentiation of the cement gland and
the progressive dissociation of the explants.
(4) Influence of the pH of the stimulation medium
As will be described in a following article, phosphate buffers should be
avoided in the stimulation medium, because they depress the differentiation of
the cement gland. In the present experiments the pH was therefore controlled
by sodium bicarbonate.
Five series of explants were incubated for 6 h in Holtfreter solutions containing 10 mM ammonium chloride. The Holtfreter solutions were modified to
contain 0-7 mM, 2 mM, 12-6 mM and 30 mM sodium bicarbonate. The pH of
these solutions was 7, 7-4, 8 and 8, respectively. The fifth solution contained
2 mM sodium bicarbonate and 0-96 mM sodium hydroxide and had a pH of 8-5.
After this initial stimulation, the explants were cultured for 5 days and the
volumes of the observed tissues were measured. The results are given in Table 1.
6o
BIB
33
964
J. J. PICARD
Table 1. Influence of the pH of the stimulation medium
Incubation time was 6 h. The first three columns indicate the pH and the corresponding modifications of the standard Holtfreter solution containing 10 mM ammonium
chloride. The number of explants is given in the fourth column. The volumes are in
/tm3 with their standard deviation, S.D.
Volume 10 6 /tm 3 ±s. D.
Experimental conditions
pH
NaHCO 3
NaOH
n
70
0-7 mM
—
21
7-4
2 0 mM
—
16
80
12-6 mM
—
17
80
300 mM
—
19
8-5
2 0 mM
0-96 mM
19
Undifferentiated
tissue
Cement
gland
Total
tissue
49-5
±131
37-2
±16-2
16-5
±8-7
5-5
±7-8
19-5
±14-4
0-2
±2-5
3-7
±6-4
19-6
± 5-2
10-4
±4-7
170
±9-4
49-7
±130
40-9
±13-4
361
±10-5
159
Percentage
cement
gland
0-3
90
54-3
65-4
±8-2
36-5
±10-7
46-6
When the stimulation was performed at pH 7 only a negligible volume of
cement gland (0-17 x 106/*m3) was observed in the explants. This accounts for
only 0-3 % of the total volume of the corresponding explants. The amount of
the cement gland was still rather low when the stimulation solution had a pH
of 7-4 (3-7 x 106/*m3, accounting for 9-0 % of the total volume). At pH 8 the
cement gland had a mean volume of 19-6 x 106 /tm3 and occupied 54-3 % of
the total volume.
Statistical analysis shows that the volumes of the cement gland obtained at
pH 7-4 (0-025 < P < 0-010) and pH 8 (P < 0-001) are significantly higher than
those obtained at pH 7. The difference between the values observed at pH 7-4
and pH 8 is also statistically significant (P < 0-001).
When the concentration of sodium bicarbonate was 30 mM (pH 8), the mean
volume of the cement gland (10-4 x 106 /tm3) was about 50 % lower than when
the concentration was 12-6 mM (pH 8). The difference is statistically significant
(P < 0-001). However, under these conditions the percentage of the total tissue
occupied by the cement gland (65-4 %) was greater than the corresponding value
(54-3 %) of the explants incubated in the solution containing 12-6 mM sodium
bicarbonate.
When the stimulation was performed at pH 8-5 (NaHCO 3 2 mM and NaOH
0-96 mM) the volume of the cement gland was approximately equal to the
value obtained at pH 8 (NaHCO 3 12-6 mM).
Finally, it may be mentioned that the mean total volume of the explants, as
In vitro studies on Xenopus cement gland. 1
965
measured in the first four series after 5 days of culture, decreases linearly as
a function of the concentration of sodium bicarbonate.
No tissue other than the cement gland was observed to differentiate in these
series.
DISCUSSION
Explants cultured for 5 days in standard salt solutions contained only
negligible amounts (< 0-1 %) of cement gland tissue. The bulk of the explants
was atypical epidermis. Similar results have been previously observed with
explants from Rana fusca and Rana esculenta (Holtfreter, 1938), from Bufo
viridis and Bufo vutgaris (Raunich, 1941, 1942), from Rana temporaria, Rana
esculenta and Bufo viridis (Gulinati, 1964), from Rana esculenta (Delpino, 1932)
and Rana nigromaculata (Yamada, 1933, 1938). However, the percentage of
explants containing cement gland tissue seemed always higher in these reports
than in our results. This percentage was reported to be 50-9 % with Rana
esculenta (55 explants) and 86-8 % with Rana fusca (38 explants) (Holtfreter,
1938). These differences may be due to different potencies of the ectoblastic
cells in the genera Rana, Bufo and Xenopus or to in vivo differences in the
volume of the cement gland in the experimental species.
When explants from Xenopus laevis gastrulae are first incubated for 6 h in the
presence of ammonium chloride, a considerable amount of these explants will
differentiate into cement gland after 5 days of culture in a standard saline solution. The concentration of ammonium chloride in the stimulation medium which
produces the best differentiation is 10 mM. Under these conditions 80 to 90 %
of the total explanted tissue differentiated into cement gland.
This concentration may be considered as rather high. Physiological concentrations of ammonium chloride in human arterial (Muting et at. 1970) and venous
(Sinniah, Fulton & McCullough, 1970) blood are about 80 fi% per 100 ml
(5 x 10~2 mM). Blood concentrations of about 10 mM are highly toxic in mammals, especially for the nervous system. However, mammalian cells are much
more resistant under in vitro culture conditions, and concentrations near or above
10 mM have been used (Zimber & Topping, 1970; Yisek, Kolodny & Gross,
1972; Kloppick, Jacobasch & Rapoport, 1967). Adult amphibians are able to
survive with high concentrations of ammonium chloride in the blood. Adult
Xenopus have been reared for 28 days in water containing 50 mM ammonium
chloride without any apparent harm (Janssens, 1972). At the end of this incubation period ammonium chloride concentration in both liver and muscles was
50 mM. Moreover, whole embryos of Xenopus laevis are able to develop normally
from blastula to stage 40 in water containing 10 mM ammonium chloride
(unpublished results).
The optimal duration of stimulation was 6 h. Under these conditions 79 %
of the total explanted tissue differentiated into cement gland. Shorter incubation
times resulted in the appearance of only small cement glands. When the
60-2
966
J. J. PICARD
stimulation was longer than 6 h the explants underwent an increasing
dissociation and the cement gland tissue was correspondingly smaller.
The volume of the cement gland was highest when the pH of the ammonium
chloride solution was 8. However, the total volume of the explanted tissue
remaining after 5 days culture decreased linearly as a function of sodium bicarbonate concentration in the stimulation medium. Therefore high concentrations
of sodium bicarbonate will be avoided in subsequent experiments and a value
of 10 mM (pH 7-5 to 7-7) has been selected.
After optimal stimulation, the cement gland tissue accounts for about 80 %
of the total tissue. This means that all or almost all the competent cells underwent differentiation. Indeed, the explanted tissue contains not only the competent superficial layer but also the underlying layer which will never form
cement gland. Although it was not possible, after 5 days of culture, to distinguish the undifferentiated cells derived from both layers, it seems highly
probable that the undifferentiated tissue (20 %) derived from the underlying
layer.
The development of the cement gland under these in vitro conditions seems
to be an interesting experimental model for biochemical studies on embryonic
differentiation. In addition to the high proportion of cement gland, the explants
contain no other differentiated tissue. Moreover, the cement gland is a very
simple organ composed of only a single cell type.
The reason for the selective differentiation of the cement gland as well as the
mechanism of stimulation by ammonium chloride remains unknown. Some
insight will be presented in the following articles.
REFERENCES
L. G. & BARTH, L. J. (1959). Differentiation of cells of Rana pipiens gastrula in
unconditioned medium. J. Embryol. exp. Morph. 7, 210-222.
DAWID, I. B. (1965). Deoxyribonucleic acid in amphibian eggs. /. molec. Biol. 12, 581-599.
DELPINO, I. (1932). Richerche sperimentali sullo sviluppo degli organi adesivi in Rana
esculenta. Arch. zool. Hal. 17, 401-415.
GULINATI, A. M. (1964). Osservazione sul differenziamento dell'organo adesivo negli anfibi
anuri. Boll. Zool. 31, 1157-1164.
HENNIG, A. & MEYER-ARENDI, J. R. (1963). Microscopic volume determination and
probability. Lab. Invest. 12, 460-464.
HOLTFRETER, J. (1931). Uber die Aufzucht isolierter Teile des Amphibienkeimes. II. Ziichtung
von Keimen und Keimteilen in Salzlosung. Wilhelm Roux Arch. EntwMech. Org. 124,
404-466.
HOLTFRETER, J. (1938). Differenzierungspotenzen isolierter Teile der Anurengastrula.
Wilhelm Roux Arch. EntwMech. Org. 138, 658-738.
HOLTFRETER, J. (1945). Neuralization and epidermization of gastrula ectoderm. /. exp. Zool.
98, 161-209.
JANSSENS, P. A. (1972). The influence of ammonia on the transition to ureotelism in Xenopus
laevis. J. exp. Zool. 182, 357-366.
KLOPPICK, E., JACOBASCH, G. & RAPOPORT, S. (1967). Steigerung der Glycolyse durch den
Einfluss von Ammonionen auf die Phosphofructokinaseaktivitat. Acta biol. med. germ.
18, 37^2.
BARTH,
In vitro studies on Xenopus cement gland. I
967
D., HEENZE, J., MUMPER, M., SCHWARZ, M. & SCHMIDT, F. M. (1970). Methodology
and clinical significance of blood-ammonia estimation. Germ. Med. Monthly 15, 523-528.
NIEUWKOOP, P. D. & FABER, J. (1967). Normal Table of Xenopus laevis (Daudin), 2nd ed.
Amsterdam: North Holland Publishing Co.
RAUNICH, L. (1941). Risultati sperimentali su espianti ectodermici di Anfibi Anuri (Bufo
viridis). La Ricerca sclent. 12, 1284-1292.
RAUNLCH, L. (1942). Contributo alia conoscenza della determinazione dell'organo adesivo
degli Anfibi anuri {Bufo vulgaris). Monitore zool. ital. 53, 17-26.
ROMEIS, B. (1948). Mikroskopische Technik. Miinchen: Oldenburg.
SINNIAH, D., FULTON, T. T. & MCCULLOUGH, H. M. (1970). Venous blood ammonium levels
in control subjects and in patients with disorders of the liver. /. clin. Path. 23, 720-726.
VJSEK, W. J., KOLODNY, G. M. & GROSS, P. R. (1972). Ammonia effects in cultures of normal
and transformed 3T3 cells. /. cell Physiol. 80, 373-381.
YAMADA, T. (1933). liber die Determination der Haftdriisen bei Rana nicromaculata. J. Fac.
Sci. Tokyo Univ. Sect. IV, Zool. 3, 239-254.
YAMADA, T. (1938). Weitere Analyse der Determination der Haftdriise bei Rana nicromaculata, mit einigen Bemerkungen iiber die Induktion anderer Kopforgane. /. Fac. Sci. Tokyo
Univ. Sect. IV, Zool. 3, 133-163.
ZIMBER, A. & TOPPING, D. C. (1970). Metabolic effects of ammonia on ascites tumor cells
and mouse spleen cells in vitro. Fedn Proc. Fedn. Am. Socs exp. 19, 428.
MUTING,
(Received 21 October 1974)