Download 3.5 MB

CHANGES IN DISTRIBUTION OF ALKALINE PHOSPHATASE DURING
EARLY IMPLANTATION AND DEVELOPMENT OF THE MOUSE
By M. S. R. SMITH*
[Manuscript received 11 August 1972]
Abstract
Changes in the uterus of the mouse during the oestrous cycle and early
development from 90 to 190 hr post coitum were examined to demonstrate variations
in the occurrence of alkaline phosphatase (ALP) activity.
Uterine ALP was shown to have distinct patterns at each stage of the oestrous
cycle. There was an increase in the uterine epithelium activity during the prooestrous and oestrous phases and an increase in the periglandular cells from prooestrus reaching maximum activity at metoestrus.
Embryonic ALP activity was confined to the inner cell mass destined to
become mainly ectoderm. At no stage examined was there any detectable ALP
activity in the invasive trophoblast.
During pregnancy ALP activity in the uterine stroma was associated with the
decidual reaction. Its first signs were observed near the inner cell mass and then
through the stroma with the initiation of decidualization. By the end of the
series ALP activity was observed in most of the uterine stroma.
The ectoplacental cone and trophoblast showed no detectable ALP activity
but by 170 hr post coitum strong ALP activity was observed on the border of the
ectoplacental cavity and in the labyrinthine zone of the developing placenta.
1. INTRODUCTION
The presence of alkaline phosphatase (ALP) in the uterus of the rat and mouse
has been described by Christie (1966), Finn and Hinchliffe (1964), and Finn and
McLaren (1967). Christie (1966) suggested that ALP was associated with glycogen
deposition during the period from the third to the ninth day of pregnancy. Finn and
Hinchliffe (1964) and Finn and McLaren (1967) suggested that the ALP reaction
was an index of decidua1ization. They also described some of the ALP reactions in
the embryo and uterus during the implantation period. Manning et aZ. (1969)
demonstrated that there was a clear-cut correspondence between the implantation
process and the rise in uterine ALP activity in the rat.
In general the implication of ALP in important cellular processes are as follows:
secretory function (Dempsey and Wislocki 1945; Moog 1946; Bradfield 1950),
phospholipid synthesis (Malone 1960), RNA synthesis (Gavasto and Pileri 1958), and
carbohydrate metabolism (Moog and Wenger 1952).
The histochemical changes in uterine ALP during the oestrous cycle and changes
in uterine and embryonic ALP in the mouse from the third to the eighth day of
pregnancy are described in this investigation; the general purpose being to elucidate
some of the uterine-embryonic relationships during early pregnancy.
* School of Anatomy, University of New South Wales, Kensington, N.S.W. 2033.
Aust. J.)iol. Sci., 1973, 26, 209-17
210
M. S. R. SMITH
II.
MATERIALS AND METHODS
Female mice of the CBA strain were mated with males of the A strain to give A x CBA
hybrid embryos. The mice were placed together at 1700 hr and checked for vaginal plugs at
midnight, giving an average time of mating of 2030 hr. The females were killed during the period
90-190 hr post coitum. The uteri were fixed in 10% formol calcium (4°C) for 18 hr and then
transferred to sucrose solution at 4°C. Implantation sites were examined from pregnant females
at the following times:
No. of females examined
Hours post coitum
15 10
90 96
12
100
14
110
12
115
11
16
130 145
12 14
170 190
Eighteen females were followed by vaginal smears through three complete oestrous cycles
(average 4-!- days per cycle). They were then killed at the requisite stages of the cycle: four at
dioestrus; four at pro-oestrus; four at oestrus; six at metoestrus. This material was fixed in
the same way as the implantation sites.
The cold-fixed material was sectioned at 8 p.m on a microtome cryostat. ALP was localized
by a slightly modified method of Burstone (1958). The incubation period ranged from 5 to 30
min, with 20 min giving the optimum result. The constituents of the incubation medium were
fast violet B (Sigma: F-2000), naphthol As-Bi phosphoric acid (sodium salt) (Sigma No. N-2250),
and Tris buffer at pH 8·8 and 9· O. Controls were employed at each stage by removing the garnet
salt from the incubation medium.
At each stage a small series of embryos was sectioned and incubated according to the
Gomori (1952) method for ALP but these results did not differ in any marked way from the
modified Burstone method except that the latter method gave more precise localization of the
enzyme.
III.
RESULTS
(a) Distribution of ALP in the Uterus during the Oestrous Cycle
The activity of ALP in the uterus during the oestrous cycle is shown in Table 1.
(i) Pro-oestrus
There was strong non-specific ALP activity at the lumen border and in the
supranuclear region in the uterine surface and glandular epithelial cells of 'both
uterine horns. No activity was observed in the stromal cells surrounding the uterine
lumen at any stage of the oestrous cycle, in contrast to that in cells immediately
adjacent to the glandular epithelium (Fig. 2). ALP activity in the region adjacent to
the blood vessels was reasonably strong, extending to the second layer of cells from
the vessel basement membrane. This activity did not appear to differ greatly between
stages of the cycle.
(ii) Oestrus
Activity at the lumen surface and in the supranuclear region of the surface
epithelial cells was similar to the pro-oestrus stage but the intensity was reduced.
The glandular epithelium exhibited no supranuclear activity and only slight activity
at the lumen surface. But the activity in stromal cells surrounding the glands increased
from the previous stage, extending at least two stromal cells from the gland basement
membrane.
(iii) Metoestrus
There was no ALP activity in uterine or glandular epithelia. As in other stages
of the oestrous cycle, there was no mucosal activity in the vicinity of the uterine
HISTOCHEMICAL OBSERVATIONS ON EARLY IMPLANTATION
211
lumen, but there was an increase in the amount of ALP activity in the mucosal areas
surrounding glandular epithelia. ALP activity in the cytoplasm of these stromal
cells was often observed up to four cell layers from the glandular basement membrane (Fig. 3).
(iv) Dioestrus
There was very little ALP activity in the uterus, the only marked activity being
observed in a few stromal cells immediately adjacent to the glandular epithelium
basement membrane.
TABLE
1
DISTRIBUTION OF ALKALINE PHOSPHATASE IN THE MOUSE UTERUS DURING THE
OESTROUS CYCLE
+ + + Strong activity; + + medium activity; + weak activity
Tissue
Uterine epithelium
Lumen surface
Supranuclear cytoplasm
Glandular epithelium
Lumen surface
Supranuclear cytoplasm
Mucosa
Uterine epithelium
Glands
No. of cells deep
Vascular system
Height (/Lm)
Uterine epithelium
Glandular epithelium
Pro-oestrus
Oestrus
+++
+++
++
++
+++
++
+
++
++
Metoestrus
+++
Dioestrus
+
Up to 1
Upt02
Upt04
++
++
41
20
43
19
36
24
14
++
++
17
>1
(b) Distribution of Alkaline Phosphatase during Early Pregnancy·
(i) 90 hr post coitum
In all experiments at this stage the embryo was lost from the implantation
cavity and information on embryonic ALP activity was unobtainable. Patterns of ALP
activity in the uterus basically resembled those of dioestrus with very little mucosal
activity. There was some moderate activity in the supranuclear region of the uterine
surface epithelium, in the region immediately mesometrial to the implantation cavity.
The cavity in the mucosal cells around the glands was similar to the dioestrouspro-oestrous pattern, and the perivascular pattern of activity resembled that found
during the oestrous cycle.
(ii) 96-105 hr post coitum
ALP activity in uterine surface epithelia mesometrial to the implantation site
was still present and there was a considerable increase in enzyme activity in the
stroma immediately adjacent to the surface epithelium near the embryonic inner cell
mass (Figs. 1 and 4). The area of activity appeared to spread into the antimeso-
212
M. S. R. SMITH
metrial region (Fig. 1). The ALP activity appeared to increase during this period,
and by approximately 105 hr post coitum a crescentic area of activity had been
formed. Stromal activity around the glands was similar to that in the pro-oestrousoestrous stages but it was apparent that it was concentrated in the antimesometrial
half of the uterus following the onset of the decidual reaction.
90 hr
110-115 hr
96 hr
130 hr
100 hr
190 hr
Fig. I.-Diagrammatic representation of the alkaline phosphatase reaction in the uterus and
embryo from 90 to 190 hr post coitum. The ALP activity is depicted by the shaded areas; strong
activity being fully shaded and weak activity being cross-hatched.
In the embryo the inner cell mass showed marked ALP activity, whereas in the
highly invasive trophoblast the histochemical methods employed in this study showed
no detectable activity.
The myometrium was found to have considerable ALP activity throughout all
stages examined.
(iii) 110-115 hr post coitum
Implantation had commenced by this stage and the invasive trophoblast, with
no detectable ALP activity, was prominent in the antimesometrial region. In the
HISTOCHEMICAL OBSERVATIONS ON EARLY IMPLANTATION
213
Fig. 2.-Pro-oestrus. This shows the alkaline phosphatase reaction in the supranuclear region of
the uterine epithelium, gland epithelium, and stroma surrounding the glands. 20 min incubation.
x 260.
Fig. 3.-Metoestrus. The strong alkaline phosphatase activity is obvious in the stromal cells
surrounding the glands. 25 min incubation. x 260.
Fig. 4.-96 hr post coitum. The inner cell mass activity is obvious. No detectable alkaline phosphatase is present in the trophoblast or uterine epithelium. The first uterine stromal activity is
adjacent to the inner cell mass. 20 min incubation. x 130.
Fig. 5.-115 hr post coitum. This shows the extent of the decidual reaction and the lack of
detectable activity around the embryo. 20 min incubation.
Fig. 6.-115 hr post coitum. The embryonic ectoderm has strong alkaline phosphatase activity.
The decidual cell reaction is obvious and the trophoblast has no detectable activity, and the
abembryonic and lateral stromal cell regions also have no detectable or reduced activity. 30 min
incubation.
Fig. 7.-130 hr post coitum. The decidual cell alkaline phosphatase activity is obvious and in the
embryo there is no activity in the extra-embryonic ectoderm. 20 min incubation.
214
M. S. R. SMITH
Fig. 8.-160 hr post coitum.
The only embryonic alkaline
phosphatase activity is shown
in the ectoderm. The antimesometrial decidual cell
activity is stronger than that
in the mesometrial region.
20 min incubation.
Fig. 9.-160 hr post coitum.
At this stage there are signs
of the commencement of
placental alkaline phosphatase activity in the labyrinthine region (couche p1asmodiale compact) (arrow).
20 min incubation.
Fig. 10.-190 hr post coitum.
A sagittal section of the
embryo showing the alkaline
phosphatase activity in the
primitive streak (arrow) and
the ectoderm. 20 min
incubation.
HISTOCHEMICAL OBSERVATIONS ON EARLY IMPLANTATION
215
lateral region of the implantation cavity stromal cell ALP activity was reduced when
compared with the rest of the stroma (Figs. 1, 5, and 6). The number of glands
appeared to be reduced in the antimesometrial region but there was still considerable
activity in mesometrial glands.
The pattern of embryonic ALP activity was similar to that at 100 hr post coitum.
(iv) 130 hr post coitum
The predominant feature at this stage was the strong ALP activity in stromal
cells, which exhibited marked decidual reaction. The activity was still mainly in the
mesometrial region (Figs. 1 and 7), and the antimesometrial area once again showed
no detectable ALP activity.
The major embryological feature was the appearance of the pro-amniotic
cavity. There was marked ALP activity in the embryonic ectoderm cells but no
detectable activity in the embryonic endoderm, extraembryonic ectoderm, or the
trophoblast (Fig. 7). In most samples, however, there was strong ALP activity on
the border of the pro-amniotic cavity.
(v) 145 hr post coitum
The decidual cell activity in the uterus became more widespread with ALP
activity moving from the antimesometrial region to the lower mesometrial region.
The only area apparently not transformed by the decidual reaction lay close to the
uterine circular musculature.
The distribution of embryonic ALP activity was similar to that at 120-130 hr,
with strong activity in the embryonic ectoderm but none in the endoderm, extraembryonic ectoderm, ectoplacental cone, trophoblast, and trophoblastic giant cells.
(vi) 160 hr post coitum
The decidual reaction was well developed and, although ALP actlVlty had
spread further into the mesometrial region of the uterus, it was strongest in the
antimesometrial region (Fig. 8). An area with no detectable ALP activity was still
present in the stroma adjacent to the embryonic mid trunk, although much reduced
from earlier stages.
In the embryo, ALP activity was very similar to that at 145 hr post coitum.
There was no detectable activity in the mesoderm even though it was in the primitive
streak. Placental activity appeared to increase but there was still no reaction product
in the ectoplacental cone. The major activity in the placental region was found around
the ectoplacental cavity and the labyrinthine region (couche plasmodiale compact)
(Fig. 9).
(vii) 190 hr post coitum
Uterine ALP activity was generally distributed throughout the stroma, though
weak near the circular muscle band.
In the embryo, evidence of activity was confined to the ectoderm with the
strongest reaction in the primitive streak and the tail-fold region (Figs. I and 10).
In the embryonic ectoderm it was strongest in the region furthest from the mesoderm.
The mesoderm and endoderm as at earlier stages had no detectable activity. Placental
ALP activity appeared to have increased from the 160-hr stage with the strongest
216
M. S. R. SMITH
activity concentrated in the blood island region, labyrinthine region, and ectoplacental
cavity. There was no detectable activity in the yolk sac endoderm, somatopleure, or
spanchnopleure, but there was slight activity in the allantois.
IV.
DISCUSSION
This study presents a chronology of the histochemical changes of ALP in the
mouse during the oestrous cycle and from the third to the eighth day of pregnancy.
The oestrous cycle can be divided into two phases: (1) anabolic (pro-oestrus and
oestrus); (2) catabolic (metoestrus and dioestrus (Bronson et aZ. 1966). During the anabolic phase there is synthesis and growth of the uterine and glandular epithelia. At this
period the uterus is under the influence of oestrogen and there is evidence that uterine
ALP in the mouse is oestrogen-dependent (Manning et aZ. 1969). In the catabolic
phase, and probably consequent on the degenerative process, there is no ALP activity
in the epithelia. Activity in stromal cells surrounding the uterine glands is cyclical,
being minimal at dioestrus and increasing considerably to a peak during metoestrus.
The increase was not correlated with gland secretion which has been shown to occur
at dioestrus and pro-oestrus (Smith 1970). Thus, although ALP activity in the stromal
glands has been shown to coincide with the period of oestrogen production, the
functional significance of the activity around the glands is still unclear.
At the stages of pregnancy examined there were a number of precise changes in
the localization of embryonic ALP. The earliest blastocyst obtained had marked
ALP activity in the inner cell mass but none detectable in the trophoblast. Throughout
the series there was no activity in the trophoblast, ectoplacental cone, extra-embryonic
ectoderm, mesoderm, or endoderm including the allantois and yolk sac. The only
tissue which displayed marked ALP activity at each stage was the ectoderm. During
placental development there was considerable activity in the labyrinthine region,
blood islands, and in the surface areas of the ectoplacental cavity. This result appears
to conflict with the results of Padykula (1958) who observed strong ALP activity in
the syncytiotrophoblast of the rat from the 13th day of pregnancy. The lack of ALP
activity in the invasive trophoblast and trophoblastic giant cells could be attributed
to one of the three processes advocated by Deane et aZ. (1962), viz. phagocytosis,
production of trophic hormone (luteotrophin), and production of steroid hormone.
On the other hand, Manning et aZ. (1969) suggested that the blastocyst must retain
its invasiveness and viability to induce and sustain the rise in ALP.
ALP activity in the uterine stroma was first observed immediately adjacent to
the developing inner cell mass. Some form of embryo-maternal interaction is thus
apparent, but its precise nature is still in doubt (Finn and McLaren 1967). The
response could be related to the presence of the lysosomal border on the leading edge
of the trophoblast at the embryo-maternal junction at c. 95 hr post coitum (Smith
and Wilson 1971). At later stages ALP activity spread to become prominent throughout the antimesometrial half of the uterine stroma, which confirms that it may be
regarded as an index of decidualization (Finn and Hinchliffe 1964). ALP activity was
absent from the stromal region adjacent to the abembryonic trophoblast and was
reduced around the lateral margins of the implantation cavity. The reasons for this
are obscure but may be related to the invasive properties of the trophoblast. High
HISTOCHEMICAL OBSERVATIONS ON EARLY IMPLANTATION
217
levels of ALP activity in the endometrium of the ewe and rabbit doe have been
implicated in the provision of maternal nutrients for the preimplantation conceptus
(Murdoch 1970). However, it is debatable whether such conclusions may also apply
to the mouse, since the results of the present study demonstrate a lack or reduction
of ALP activity in the region immediately surrounding the conceptus.
As development proceeds the decidual reaction spreads to involve the majority
of the uterine stromal tissue, which has been shown to give a positive reaction for
glycogen in the rat (Christie 1966). An association is suggested, therefore, between
ALP distribution and the decidual reaction.
The nature of the antimesometrial region of the stroma and the effect of
trophoblast on stromal cells in the vicinity of the implantation cavity is under further
investigation.
V. ACKNOWLEDGMENTS
The author is indebted to Dr. D. Pugh and Dr. R. N. Murdoch for advice and
criticism of this work. The technical assistance of Margaret McPherson is gratefully
acknowledged. Part of this work was completed while the author held a Wellcome
Trust Fellowship at Southampton University.
VI. REFERENCES
BRADFIELD, J. R. G. (1950).-Biol. Rev. 25,113.
BRONSON, F. H., DAGG, C. P., and SNELL, G. D. (1966).-Reproduction. In "Biology of the
Laboratory Mouse". (Ed. E. L. Green.) Ch. 11. p. 187. (McGraw-Hill Book Company:
New York.)
BURSTONE, M. S. (1958).-J. nath. Cancer Inst. 21, 523.
CHRISTIE, G. A. (1966).-J. Reprod. Fert. 12, 279.
DEANE, H. 'V., RUBIN, B. L., DRIKS, E. C., LOBEL, B. L., and LEIPSNER, G. (1962).-Endocrinology
70,402.
DEMPSEY, E. W., and WISLOCKI, G. B. (1945).-Am. J. Anat. 76, 277.
FINN, C. A., and HINCHLIFFE, J. R. (1964).-J. Reprod. Fert. 8, 331.
FINN, C. A., and McLAREN, A. (1967).-J. Reprod. Fert. 13, 259.
GAVASTO, F., and FILERI, A. (1958).-Cancer, N. Y. 11, 222.
GOMORI, G. (1952).-"Microscopic Histochemistry." (Chicago Univ. Press.)
MALONE, T. E. (1960).-Am. J. Anat. 106,41.
MANNING, J. P., STEINETZ, B. G., and GIANNINA, T. (1969).-Ann. N. Y. Acad. Sci. 166, 482.
MOOG, F. (1946).-Biol. Rev. 21, 41.
MoOG, F., and WENGER, E. L. (1952).-Am. J. Anat. 90, 339.
MURDOCH, R. N. (1971).-Aust. J. bioI. Sci. 24, 331.
PADYKULA, H. A. (1958).-J. Anat. 92,118.
SMITH, M. S. R. (1970).-J. Reprod. Fert. 22, 461.
SMITH, M. S. R., and WILSON, I. B. (1971).-J. Embryol. expo Morph. 25,165.