/. Embryol. exp. Morph. Vol. 44, pp. 121-131, 1978
Printed in Great Britain © Company of Biologists Limited 1978
Ultrastructural localization
of non-specific alkaline phosphatase during cleavage
and blastocyst formation in the mouse
By JACQUES MULNARD 1 AND ROGER HUYGENS 1
From the Department of Anatomy and Embryology,
Free University of Brussels Medical School, Brussels, Belgium
SUMMARY
The localization of non-specific alkaline phosphatase activity during cleavage and blastocyst formation has been investigated in the mouse by electron microscopy. The activity is
detectable for the first time at the two-cell stage and is localized on the surface of the interblastomeric plasma membranes and on small cytoplasmic inclusions. It increases in the
following stages, predominantly on the interblastomeric membranes, the outside membranes
remaining devoid of reaction. From the four-cell stage on, small reactive grains are also
observed in the crystalloid plates of the cytoplasm. At the morula stage, the plasma membranes of the inner mass cells are entirely marked by the reaction whereas the trophoblastic
cells are polarized, with their inner surfaces positive and outside surfaces negative. At the
blastocyst stage the enzyme is gradually eliminated from the membranes bordering the
blastocoel and from the interblastomeric furrows of the trophoblast and primary endoderm. The significance of the differential localization of the enzyme is discussed, especially
in relation to the differentiation mechanisms of the trophoblast and inner cell mass.
INTRODUCTION
In a previous paper (Mulnard, 1974) the localization of non-specific alkaline
phosphatase in the early development of the rat and mouse were comparatively
reevaluated by means of an improved technique involving fixation by glutaraldehyde. The original method of Gomori-Takamatsu, applied to whole mounted
embryos, gave the following results in the case of the mouse: (1) the enzymic
activity appears at the four-cell stage and perhaps, but not conclusively, at
the two-cell stage, (2) at all cleavage stages it is mainly localized on the interblastomeric membranes and absent from the outside surfaces of the blastomeres. Activity was also found on small cytoplasmic inclusions, (3) at the
blastocyst stage, the reaction fades rapidly in the trophoblastic and primary
endodermal cells where it disappears from the interblastomeric membranes and
from the surface bordering the blastocoel; hence in the egg-cylinder-stage
1
Authors' address: Laboratoire d'Anatomie et d'Embryologie humaines, Faculte de
Medecine et de Pharmacie de PUniversite Libre de Bruxelles, Rue aux Laines, 97, B 1000
Bruxelles, Belgium.
122
J. MULNARD AND R. HUYGENS
embryo of the early 6th day the reaction appears to be restricted to the primary
ectoderm, (4) the extra-embryonic ectoderm is devoid of reaction from the
beginning of its formation.
Such a gross analysis was only a prelude to an ultrastructural study, the
results of which are presented in this paper.
MATERIALS AND METHODS
The animals were of the white (NMRL) or black (C57B1) randomly bred
strains. Ovulation was always spontaneous and the day of discovery of the
copulation plug was counted as the first of development. The embryos were
prefixed for 1 h at 2 °C in a 2 % solution of glutaraldehyde in 0-1 M cacodylate
buffer (pH 7-4) containing 8 % sucrose. They were then washed at 2 °C for
2-3 h in the buffer and incubated for 15 min at 37 °C. The incubation medium
was that of Mayahara, Hirano, Saito & Ogawa (1967) with Reynold's solution
as a source of Pb ions (Tris-HCl buffer 0-2 M, pH 8-5, 1-4 ml; Na/? glycerophosphate 0-1 M, 2 ml; Reynold's solution, 4 ml; 0-15% MgSO4, 2-6 ml). Sucrose
was added at the final concentration of 8 % and pH was adjusted to 9-2-9-4
by 0-1 N-HC1. Control embryos were incubated in the same mixture in which
glycerophosphate was replaced by the corresponding amount of buffer. After
incubation, the embryos were rinsed for a few minutes in cacodylate buffer and
immediately post-fixed for 45 min in a 2 % solution of OsO4 in 0-1 M cacodylate
buffer (pH 7-4) at 2 °C in the dark. After a short rinsing in the buffer, they were
dehydrated and embedded in Epon, without propylene oxide treatment.
Thin (1-2 fim) or ultrathin (0-03 jam) sections were cut on a LKB microtome.
The thin sections were mounted without staining for examination by phasecontrast. The ultrathin sections were directly examined with a M III Philips
electron microscope. In order to verify the good preservation of the structures,
some ultrathin sections were stained with uranyl acetate-lead citrate.
For each stage some embryos were treated after incubation and rinsing by
diluted (NH4)2S and mounted for in toto examination.
RESULTS
No reaction could be detected in the uncleaved egg, whether fertilized or not.
The first evidence of alkaline phosphatase activity was found at the two-cell stage
(Fig. 1 A): very small grains of lead precipitate are scattered on the adjacent
membranes of the two blastomeres and mainly on the surface of the microvilli
and on the walls of the thin cisternae which are already commonly present
in the interblastomeric space. A few grains of the same size are found in the
cytoplasm, either isolated or in small clusters, in the vicinity of the membrane.
At the four-cell stage (Fig. 1B) the localization is identical but the grains of
precipitate are more numerous and thicker than at stage II. Also some of the
Alkaline phosphatase in early mouse embryos
123
V .
<
•!<•.•
'• A
A
.
.
".
B
Fig. 1. Electron micrographs of the interblastomeric regions at the two-cell (A) and
four-cell (B) stages. Mayahara's method, uncontrasted. Small grains of lead phosphate precipitate on the plasma membranes, on the surface of the microvilli and on
the wall of the few intercellular cisternae. lntracytoplasmic grains of the same size,
isolated or in clusters, some of them in small crystalloid plates (CR). Obvious
increase of the positive structures from stage II to stage IV.
intracytoplasmic clusters are larger and exhibit a crystalloid structure. At the
two-cell as well as the four-cell stage the outside membrane of the blastomeres is
devoid of reaction. The nuclei are also entirely negative.
At the eight-cell stage (Fig. 2) the situation is the same as at stage IV, with a
continuous increase of the number and size of the reactive grains on the surface
of the inside membranes of the cells and in the cytoplasm where large clusters of
crystalloid appearance are found in increasing number (Fig. 2C). Higher
magnification (Fig. 2D) indicates that the large clusters correspond to the wellknown 'crystalloid plates' (see Calarco & Brown, 1969; Gianguzza & Mulnard,
124
J. MULNARD AND R. HUYGENS
.*.*.^»
,<•*
\
.. ^ — r * ^ ' ^
• ^h
_.
Alkaline phosphatase in early mouse embryos
125
1972) in which isolated or aggregated particles of lead phosphate precipitate
are scattered. Again at stage YID the outside membranes and nuclei are devoid
of reaction.
The overall distribution of the membranar enzymic activity in the morula and
young biastocyst of the 4th day (16-32 cells) is clearly demonstrated by phasecontrast examination of thin sections (Fig. 3A): the reaction is positive along
the interblastomeric furrows and absent from the outside membranes of the
trophoblast which now forms a continuous envelope around the inner cell mass.
At the electron-microscopical level (Fig. 3B) the interblastomeric plasma
membranes of trophoblastic as well as inner mass cells are seen to be coated
with a more or less regular row of large and dense grains of lead phosphate
precipitated on the outer surface of each membrane, as is clearly shown where
the two membranes are locally separated to form intercellular cisternae. One
can also see how abruptly the reaction stops at the junction between a peripheral
interblastomeric furrow and the outside membrane of the two adjacent trophoblastic cells (Fig. 3B, arrow). The membranar reaction is interrupted along the
few tight junctions which have developed between the blastomeres (Fig. 3C,
arrow). In contrast with the increased membranar activity, the number of
positive intracytoplasmic structures - isolated or aggregated granulations has somewhat decreased.
In the early 5th day blastocyst a rapid change occurs in the distribution of the
membranar alkaline phosphatase activity (Fig. 4): the reaction remains unchanged on the intercellular membranes of the primary ectoderm cells and
practically also on the inside membranes of the cells which immediately surround
them. The outside membranes of the trophoblastic cells remain negative.
But the reaction has considerably faded in the intercellular furrows of the
FIGURE 2
Distribution of the alkaline phosphatase activity at the eight-cell stage. Mayahara's
method, uncontrasted.
(A) An embryo treated by NH42S after incubation and photographed in toto: absence
of reaction on the outside membranes facing the zona pellucida. Heavy reaction on
the interblastomeric regions.
(B) Thin section examined under phase-contrast: positive reaction in the interblastomeric furrows and on a few cytoplasmic inclusions. No reaction on the outside membranes. The nucleoli have the same contrast in control sections of embryos
incubated without substrate.
(C) Uncontrasted electron micrograph of the central region of an eight-cell embryo:
increased number and size of the grains of lead precipitate on the membranes of
three adjacent blastomeres. In the cytoplasm, numerous grains of precipitate, either
isolated or aggregated in small clusters. Positive granular reaction on large crystalloid plates (CR).
(D) Higher magnification of the same preparation to demonstrate two crystalloid
plates containing numerous positive inclusions. Fibrous material (F) is also visible
but shows no reaction.
9
EMB
44
126
J. MULNARD AND R. HUYGENS
\c
IC
1 //m
,*,
':
\
•
•
1 //m
c
Alkaline phosphatase in early mouse embryos
111
trophoblast and of the primary endoderm, where only a few very small grains can
still be observed, and on the membranes, either trophoblastic or endodermal,
which border the blastocoel. In the vicinity of the cell surfaces and especially along
the endodermal wall, small cytoplasmic fragments, either free or still attached
to the blastomeres and covered with tiny grains of lead precipitate, are seen
floating in the blastocoelic fluid, giving the impression of a 'desquamation' of
the cell surface. As in the morula isolated or aggregated positive granules and
crystalloid plates are less frequently observed than at the eight-cell stage.
DISCUSSION
The electron-microscopical analysis has confirmed and given finer detail to
the main results obtained by the method of Gomori-Takamatsu applied to
whole embryos (Mulnard, 1974). Up to the morula stage non-specific alkaline
phosphatase activity is observed on the inside membranes of all blastomeres
and absent from their outside membranes. In the young blastocyst it disappears
gradually from the trophoblast and from the primary endoderm except where
these groups are in contact with the primary ectoderm cells.
Before any conclusion can be drawn from these observations, it is necessary
to consider the possibility of an artifact due to selective inactivation or elimination of the superficial enzyme by the fixative or to removal of its reaction
products. Selective and irreversible inactivation by glutaraldehyde is highly
improbable: considering the easy penetration of the fixative, any irreversible
action should affect the deep regions of the embryo as well as its surface, or at
the least result in an outside-inside gradient of inactivation. The clear-cut
limit between the positive interblastomeric reaction and the negative outside
surfaces of adjacent trophoblastic cells of the morula (see Fig. 3B) is not
explicable by an inactivation effect of the fixative. The hypothesis of a selective
FIGURE 3
Morula stage (16-32 cells). Mayahara's method, uncontrasted.
(A) Thin section photographed under phase-contrast: positive reaction on the inside
membranes of all blastomeres. No reaction on the outside membranes of the trophoblast. The nucleoli have the same contrast as in control embryos.
(B) Low magnification electron micrograph of the peripheral region of a 16-cell
morula: positive reaction on the adjacent plasma membranes of inner mass (IC)
and trophoblastic (T) cells. Clear-cut limit between the positive interblastomeric and
the negative outside membranes of two adjacent trophoblastic cells (arrow). The
isolated or aggregated reactive grains and the crystalloid plates are less numerous
than at stage VIII.
(C) Higher magnification of the central region of a similar morula showing the
more or less regular row of grains of lead phosphate precipitate and their situation
on the external aspect of the adjacent plasma membranes. Note the absence of
reaction along the tight junction uniting two of the blastomeres (arrow).
9-2
\
N3 d
*
•
• *
i
33 d
4-
SNH9AHH
H QNV
'£
831
Alkaline phosphatase
in early mouse embryos
129
removal of the reaction product by the post-fixation and embedding procedures
must also be discarded since control treatment of whole embryos by ammonium
sulfide immediately after incubation gives the same negative result on the
outside membranes (see Fig. 2A). Perhaps more difficult to eliminate is the
possibility of a 'scouring' action of the fixative on the free membranes with
subsequent removal of the membranar debris by washing, an action which
would not affect the better protected interblastomeric membranes and, to a
lesser degree, the membranes bordering the blastocoel where the debris would
not be entirely washed away (Fig. 4). Such an artifact is also improbable when
one considers (1) that at the blastocyst stage (Fig. 4) the same residual activity
is observed not only along the membranes bordering the blastocoel but also in
the interblastomeric furrows of the trophoblast and of the primary endoderm
where the membranes should be protected against any removal effect, whereas
the reaction remains unchanged between the cells of the primary ectoderm,
(2) that after identical fixation procedure the reaction is positive on the outside
membranes of the trophoblast for other membrane-bound phosphatases such
as 5'-nucleotidase (unpublished observations).
However, the differential localization of the alkaline phosphatase in the
preimplantation mammalian embryo remains a subject of controversy, in
spite of the increasing precision of the cyto-enzymological methods. In a recent
publication on membrane-bound phosphatases in the early development of the
mouse, Vorbrodt, Konwinski, Solter & Koprowski (1977) describe an entirely
different distribution of the alkaline phosphatase activity which is claimed to be
predominantly localized on the outside and adjacent trophoblastic membranes
of the blastocyst and only occasionally detected in the inner cell mass. Such a
radical discrepancy can perhaps be explained by an unusually long incubation
(60-120 min) at a relatively low pH (8-6 instead of 9-2-9-4) at which other
hydrolytic enzymes present on the outside membranes - such as 5'-nucleotidase
- are capable of cleaving glycerophosphate or CMP (see Goff & Milaire, 1974).
On the other hand, our results are in good agreement with the conclusions of
FIGURE 4
Blastocyst of the late 4th day having spontaneously shed its zona pellucida. Mayahara's method, uncontrasted.
(A) Thin section examined under phase-contrast: heavy positive reaction in the
interblastomeric furrows of the primary ectoderm (P.EC). No reaction detectable
with that technique in the trophoblast (T) or the primary endoderm (P. EN).
(B) Low magnification electron micrograph of a similar blastocyst showing the
intense positive reaction on the adjacent membranes of the primary ectoderm cells
(P. EC). Absence of reaction on the outside membranes of the trophoblastic cells (T).
Decreasing number and size of the positive grains in the interblastomeric furrows
and on the blastocoelic surface of the trophoblast and primary endoderm (P. EN).
Presence in the blastocoelic fluid of membranar debris covered with small positive
grains, some still attached to the inside surface of the primary endoderm cells.
130
J. MULNARD AND R. HUYGENS
Izquierdo and his co-workers, who found the alkaline phosphatase activity of
the mouse morula and blastocyst is definitely higher in the inner cell mass than
in the trophoblast (Izquierdo & Marticorena, 1975; Izquierdo & Ortiz, 1975;
Izquierdo 1977).
The accumulating evidence of a differential, temporal and spatial localization
of non-specific alkaline phosphatase points to the possibility of a relation of the
enzyme to the mechanisms which regulate the differentiation of the two fundamental groups of cells in the preimplantation embryo, inner cell mass and
trophoblast. Up to the young blastocyst stage the membrane phosphatase
activity is localized along the interblastomeric furrows, the free surfaces of the
blastomeres being entirely negative. The result is that as soon as a cell becomes
completely surrounded by others the enzymic activity extends to the entire
surface of its plasma membrane whereas the peripheral blastomeres remain
polarized, with their inside membranes positive and outside membranes devoid
of activity. It is thus conceivable that the alkaline phosphatase is implicated in
general surface modifications responsible for the differentiation of the two
groups of blastomeres. Such a mechanism would be in accordance with Tarkowski
& Wrobleska's conception (1967) of a differentiation depending on the central
or peripheral position of the blastomeres in the morula.
Also of special interest is the disappearance of the alkaline phosphatase
from the cell groups of the blastocyst which undergo early extra-embryonic
differentiation: trophoblast and primary endoderm. The ultrastructural analysis
has confirmed this finding of a previous in toto study (Mulnard, 1974): it shows a
considerable reduction of the size and number of the grains of lead phosphate
precipitated along the blastocoelic wall and in the interblastomeric furrows. It
also suggests that the enzyme is possibly eliminated by a sort of desquamation of
the cell membranes bordering the blastocoel and finally released with the
membrane debris in the blastocoelic fluid. This modification could explain the
difference observed by Calarco & Epstein (1973) in scanning electron microscopy
between the cell surfaces bordering the blastocoel and the outer cell surface
of the mouse blastocyst. However, it is not clear whether the cytoplasmic
debris covered with positive grains and floating in the blastocoelic fluid
represent a process of membrane elimination or the remnants of trophoblastic
fragments isolated at the wall of the blastocoel during its formation (unpublished
personal observations).
The functional significance of the non-specific alkaline phosphatase in the
early mammalian development has not been elucidated so far. In the mouse, the
predominant membranar localizations are remarkable because the final product
of the reaction is situated on the external aspect of the interblastomeric plasma
membranes (see Fig. 3), suggesting that the enzymic molecules are inserted in
the membrane with their active group turned outside. This is consistent with the
hypothesis that the alkaline phosphatase could be implicated, here as in various
adult tissues (small intestine, kidney tubules... see Fernley, 1971), in the active
Alkaline phosphatase in early mouse embryos
131
transport of small molecules across the plasma membrane. In addition to the
membrane reaction, more discrete sites of enzymic activity were observed in
the depth of the cytoplasm. They show a maximum development between
stages II and VIII and tend to be less frequent in the morula and blastocyst.
The granules found in the vicinity of the membranes, either isolated or in small
clusters, could be the cytochemical visualization of small Golgi vesicles representing the transportation sites of the enzyme to its functional membranar
localization. Of particular interest is the granular reaction found in the crystalloid plates. It was previously demonstrated that these structures are closely
associated with small elements of rough endoplasmic reticulum (Gianguzza &
Mulnard, 1972). The positive granules observed in the plates (see Fig. 2D)
would possibly correspond to sites of enzymic synthesis. Those interpretations
are however conjectural in the present state of our knowledge.
We are indebted to W. Zwijsen for technical assistance.
REFERENCES
P. & BROWN, E. (1969). An ultrastructural and cytological study of preimplantation development of the mouse. /. exp. Zool. Ill, 253-284.
CALARCO, P. & EPSTEIN, C. (1973). Cell surface changes during preimplantation development
in the mouse. Devi Biol. 32, 208-213.
FERNLEY, H. (1971). Mammalian alkaline phosphatases. In The Enzymes, vol. 4 (ed. Paul
D. Boyer), pp. 417-447. London, New York: Academic Press.
GIANGUZZA, M. & MULNARD, J. (1972). Quelques aspects ultrastructuraux de 1'activite
lysosomiale dans les premiers stades du developpement de la souris. Archs Biol. 83,
499-512.
GOFF, R. & MILAIRE, J. (1974). A comparative analysis of the histochemical dephosphorylation of ATP and AMP in cryostat and paraffin sections of early chick embryos. Acta
Histochem. 51, 220-254.
IZQUIERDO, L. (1977). Cleavage and differentiation. In Development in Mammals, vol. 2
(ed. Martin H. Johnson), pp. 99-1.18. Amsterdam: North-Holland.
IZQUIERDO, L. & MARTICORENA, P. (1975). Alkaline phosphatase in preimplantation mouse
embryos. Expl Cell Res. 92, 399-402.
IZQUIERDO, L. & ORTIZ, M. (1975). Differentiation in the mouse morula. Wilhelm Roux
Arch. EntwMech. Org. Ill, 61-14.
MAYAHARA, H., HIRANO, H., SAITO, T. & OGAWA, K. (1967). The new lead citrate method for
the ultracytochemical demonstration of activity of non-specific alkaline phosphatase
(orthophosphoric monoester phosphohydrolase). Histochemie 11, 88-96.
MULNARD, J. (1974). Les localisations de la phosphatase alcaline dans les phases precoces du
developpement embryonnaire du rat et de la souris: devaluation comparative. Bull. Acad.
r. Med. Belg. 129, 677-697.
TARKOWSKT, A. & WROBLESKA, J. (1967). Development of blastomeres of mouse eggs isolated
at the 8-cell stage. /. Embryol. exp. Morph. 18, 155-180.
VORBRODT, A., KONWINSKI, M. SOLTER, D. & KOPROWSKF, H. (1977). Ultrastructural cytochemistry of membrane-bound phosphatases in preimplantation mouse embryos. Devi
Biol. 55, .117-134.
{Received 3 August 1977)
CALARCO,