вход по аккаунту


Ultrastructural evidence for loss of the trophoblastic layer in the chorioallantoic placenta of Australian Bandicoots MarsupialiaPeramelidae.

код для вставкиСкачать
Ultrastructural Evidence for Loss of the Trophoblastic
Layer in the Chorioallantoic Placenta of Australian
Bandicoots (Marsupial ia : Peramelidae)
Division of A n i m a l Physiology, CSIRO, I a n Clunies Ross A n i m a l Research
Laboratory-Prospect, P. 0. B o x 239, Blacktown, N . S . W . 21 48, Australia
and Laboratory of Electron Microscopy, Wellesley College, Wellesley, Massachusetts 02181, U.S.A. and Department of Zoology, T h e University of
British Columbia, Vancouver, B.C. V6T 1W5, Canada
In most marsupials, placentation involves only the yolk sac;
however. in the bandicoot family, Peramelidae, a functional chorioallantoic placentation develops in addition (Hill, 1895, 1897, 1900; Flynn, '22, '23). This
duality is viewed as having evolutionary significance because most eutheria
have both placentae. Furthermore, the bandicoot trophoblast was reported to
vanish from the chorioallantoic site in late gestation (Hill, 1897; Flynn, '23);
whereas, the eutherian trophoblast is identifiable throughout later pregnancy
and may act as an immunological barrier between maternal and fetal genotypes
(Kirby, '68). Thus we have re-examined this singular chorioallantoic placenta
of the bandicoot in plastic sections with light and electron microscopy.
A distinctive feature of bandicoot placentation is the transformation of the
uterine simple columnar luminal epithelium into a highly vascular lining composed almost entirely of discrete syncytial masses (homokaryons 1. Endometrial
blood vessels penetrate among the homokaryons to create a rich network of large
diameter capillaries at extremely superficial locations near the maternal surface.
In the chorioallantoic placenta ( 7 mm to 10-11 mm crown-rump embryos)
the microvillous surface of the maternal homokaryons interdigitates with the
microvillous border of the fetal trophoblast with desmosomal interaction. This
trophoblast consists of a single layer of tall columnar undifferentiated cells rich
in ribosomes-polysomes, poor in cytoplasmic membranes, and with large nuclei
that have distinct clumps of heterochromatin and conspicuous nucleoli. It is thus
remarkable that these undifferentiated cells disappear as a recognizable layer
later in gestation (12 mm crown-rump embryos j.
Flynn's hypothesis that the trophoblastic cells disappear by fusing with maternal syncytia gains support from the existence of two populations of nuclei
in the syncytial masses only at the chorioallantoic site. One population is comparable to that occurring in the homokaryons of the yolk sac placenta, i.e., pale
staining nuclei with little heterochromatin and small peripheral nucleoli. However, the other nuclei resemble those of the trophoblast cells. Since the trophoblastic cells before their disappearance as a layer possess properties associated
with potential for further diff'erentiation, the possibility of fusion between the
maternal homokaryons and the fetal trophoblastic cells to form heterokaryons
composed OI two genotypes merits further consideration.
The disappearance of the trophoblastic layer and the superficial positioning
of the maternal capillaries bring the maternal and fetal bloodstreams into closest
proximity near term (12 mm crown-rump embryo). The thinnest parts of the
barrier consist of delicate cytoplasmic extensions from the syncytial masses (that
may be maternal in origin or jointly maternal and fetal) and a layer of maternal
stroma intervening between the maternal and fetal endothelia. Thus the chorioallantoic placental barrier of the marsupial bandicoot is unlike any thus far described for eutherian mammals.
Received Mar. 22, ' 7 6 . Accepted May 7 , ' 7 6 .
1 Supported by USPHS Research Grant HD-01026 and HD-09209 (Padykula).
2 Supported by National Research Council uf Canada Grant A-3462 and U.B.C. Research Grants (Taylor).
3 This investigation was initiated during the tenure of a USPHS Special Rcsearch Fellowship HD-52725
awarded to Dr. Padykula.
4 This investigation was initiated during the tenure of a sabbatical leave from The University of British
Columbia given to Dr. Taylor.
ANAT. REC.,186: 357-386.
The occurrence of chorioallantoic placentation has been established in only one
family of marsupials, the Peramelidae or
bandicoots. This important discovery was
made by J. P. Hill in 1895 who described
in Isoodon obesulus (Perameles obesula)
a discoidal chorioallantoic placenta that
was fused to the endometrium. Later Hill
(1897) provided a careful histological description of this chorioallantoic placental
barrier, from which all subsequent interpretations have essentially been derived.
An unusual feature of this barrier is the
disappearance of the trophoblastic layer
near term.
Placentation in the bandicoots includes
both the yolk sac and allantois, as is true
for most eutherian mammals. The possibility of de nouo evolution of chorioallantoic
placentation in marsupials and eutheria
is generally advanced as conspicuous evidence of convergence in these two mammalian lineages. Despite the potential evolutionary significance, our knowledge of
this placental barrier is limited to the
studies of Hill (1895, 1897, 1900) and
Flynn (’22, ’ 2 3 ) . These classical descriptions are important but complex; moreover, the limitations of the preparative
methods as well as of the resolving power
of the light microscope become evident in
attempts to interpret the cellular boundaries in this unique placental barrier. Thus
we decided to reinvestigate the chorioallantoic placental barrier in Perameles
nasuta and Isoodon macrourus using semithin and ultrathin plastic sections for light
and electron microscopic study.
A distinctive feature of bandicoot placentation is the conversion of the simple
columnar luminal epithelium of the uterus
into one composed of highly vascularized
discrete syncytial masses. These individual
syncytial masses may be designated “homokaryons” in accordance with the tenninology used in studies of experimental cell
fusion in vitro (Harris, ’70). This specialized epithelium persists throughout gestation and becomes a prominent component
of both yolk sac and chorioallantoic placentae. Flynn (’22, ’ 2 3 ) believed that during late gestation the trophoblast (chorion)
fuses with the maternal syncytial luminal
epithelium at the chorioallantoic site. Although Hill (1897) initially described the
trophoblast as degenerating and thus disappearing, he later (Hill, ’32; Pearson, ’49)
espoused Flynn’s hypothesis of fusion, a s
the more likely possibility. If true, such a
fusion would involve union of cellular layers representing two different genotypes;
and this idea is not so startling during the
current era of in vitro experimentation
with cell fusion (e.g., Harris, ’70) and also
expanding knowledge of ultrastructural aspects of implantation (Enders, ’72; Schlafke and Enders, ’75).
The uniqueness of the chorioallantoic
barrier in the bandicoot is substantiated
and further defined i n the present report.
We have found that i n late gestation the
trophoblast disappears as a recognizable
layer and, as a consequence, in the chorioallantoic placental barrier the maternal
and fetal vessels are separated by highly
attenuated syncytial epithelium and maternal stroma. The occurrence of two
categories of nuclei only in the syncytial
masses (“heterokaryons”) at the chorioallantoic site may be interpreted as being
the result of fusion of cellular layers of
two different origins. A preliminary report
(Padykula and Taylor, ’74) has been published.
Table 1 indicates the stages of pregnancy in two genera of bandicoots, Perameles and Isoodon, that were collected i n
the area of Sydney, Australia, from September 1971 to early February 1972. Of
the 38 female bandicoots that were captured in wire live-traps, 12 were pregnant.
The condition of pregnancy could not be
recognized with certainty from external
anatomical features; this limited our attempts to improve the sequence of developmental stages. This limitation may also be
related to the distinctive feature of marsupial reproduction that gestation usually
occurs within the estrous cycle (Sharman,
’69) and that the non-pregnant luteal
phase closely resembles that of pregnancy
(Tyndale-Biscoe, ’73). Hughes (’62) stated
that the estrous cycle of P. nasuta averages
26 days and that the length of gestation
falls between 8 and 12 days. Stodart (’66)
succeeded in breeding P . nasuta in captivity, and reported that gestation lasted 12.5
days. More recently Tyndale-Biscoe (’74)
indicated that in P. nasuta the free blasto- Isoodon 55, Perameles 307 and Isoodon 2.
cyst stage lasts seven to eight days and that The methods of specimen preparation for
embryogenesis (and thus placentogenesis) these three specimens are reported fully
occurs during the last four days. Less is below. Subsequent reports will consider the
known about I. macrourus except that the ultrastructure of the maternal syncytium
gestation period is one to four hours less and the yolk sac placenta.
than 12.5 days, and is the shortest mamOur first specimens were preserved by
malian gestation recorded (Lyne, '74).
immersion fixation, and this proved inI n the absence of timed pregnancies, adequate for careful isolation of various
crown-rump (CR) measurements of the regions of the two placentae that is needed
embryo-fetus were used to sequence pla- particularly for ultrastructural analysis.
cental stages. This measurement was made Midway in the collecting period we
on freshly isolated embryos when immer- switched to perfusion of fixative via the
sion fixation was used (except for the em- abdominal aorta, and our best material
bryos of P. nusuta 305 which were fixed for ultrastructural analysis was thus obin glutaraldehyde before measurement). tained (table 1 ). Furthermore, the chorioIn the three animals that were perfused allantoic placenta is a relatively small diswith fixative, this measurement is derived coidal region, and can be missed if the
delicate extra-embryonic membranes are
from fixed embryos.
This report is in general derived from allowed to collapse. According to Hill
analysis of all of the specimens in table 1 (1897, 1900), a n 8.00-8.75 mm CR embut it is based primarily on the following bryo ( P . obesula j had a chorioallantoic
animals: Perameles 305, Isoodon 55, Pera- placental region that measured 9 X 12
meles 307, and Isoodon 2. Because the goal mm, and the 12.5 mm CR embryo was asof this report is the cytological organiza- sociated with a chorioallantoic region that
tion of the chorioallantoic placenta, i t is was 6 X 8 m m in length and width. Rederived primarily from our latest stages: cently Hughes ('74 j illustrated the gross
features of a full term fetus of P . nasuta
that had a CR length of 13.8 m m and a
chorioallantoic placenta that meaBandicoot deuelopmental stages obtained
sured 11.5m m X 10 mm.
Peramelidae: Perameles nasuta and
Two bandicoots (PerameEes 355 and IsoIsoodon macrourus
odon 5 5 ) were perfused with aldehydic fixBlastocysts
atives as follows. The anesthetic, Inactin
(Promota, West Germany), was injected
Perameles 306
1.7m m
intraperitoneally at a dose of 80-150 mg/
Perameles 316
2.1 mm
kg body weight. A cannula (18 gauge
Perameles 323
5.0-6.2 mm
needle) was inserted into the abdominal
Not known
aorta just above the left renal artery. The
cannula was fastened by ligatures or by a
( C R ) Lengths
* Perameles 355
tissue adhesive (IBC-2 Ethicon from Ethi* Perameles 362
5.0 mm
con, Inc., Somerville, New Jersey). The inPerameles 305
5.0 mm
* Isoodon
6.5 mm
ferior vena cava was slit to allow outflow,
7.0 mm
as the first perfusion with heparinized saIsoodon
7.4 mm
line (2,500 IU in 100 m l j commenced a t
Perameles 307
a pressure of 140 m m Hg. The blood was
10-11 mm
12.0 mm
washed out, and then the first fixative
(1 % paraformaldehyde and 1.25% glutarL 2 Blastocyst stage occupies seven to eight days and
embrvonenesis the last four days i n P e r a m e k s (Tynaldehyde in 0.067 M cacodylate buffer, pH
dale-Bis%oe, '74).
7.2 with calcium chloride added 125 mg
3Apposition of the allantois to the chorion occurs
at 6.6 m m CR in P. gunnii (Flynn, '23).
CaCl, anhydrous to 50 ml fixative]) was
4 This animal was in captivity for 10.75 days before
introduced. This was followed immediately
5 The CR length of P. nasuta at birth h a s been reby perfusion with a more concentrated fixported to he 14 m m (Hill, 1897), 12.8 (Lyne, '64), and
13.8 mm (Hughes, '74).
ative ( 4 % paraformaldehyde and 5% glu*These specmens were fixed by aldehydic perfutaraldehyde buffered as above). About 200
ml of each perfusate was used. The animal
was then left i n the refrigerator overnight,
and the specimens of the female reproductive tract were isolated on the next morning. The isolated specimens were placed
into cold 0.1 M cacodylate buffer for 24
hours, and then osmicated in 1% OsO, in
0.14 M veronal acetate buffer for two
hours. Some of this methodology was derived from the procedures of Karnovsky
('65), Reese and Karnovsky ('67), and
Peters et al. ( ' 6 8 ) . The specimens fixed
for electron microscopy were dehydrated
in a graded series of ethanols and embedded in Araldite.
The placentae and endometria of Isoodon 2 and Perarneles 307 were fixed by
immersion as follows: 6.25% glutaraldehyde in 0.067 M cacodylate buffer with
0.002 M CaClz at pH 7.4 or 1.5% glutaraldehyde in 0.22 M collidine buffer (pH
7.4) followed by osmication in 1% OsOl
buffered in veronal acetate. Also some
samples of Isoodon 2 were fixed only in
1% OsO, in 0.14 M veronal acetate buffer.
Semi-thin plastic sections (1 pm) were
stained with toluidine blue. Ultrathin sections were doubly stained with lead hydroxide and uranyl acetate, and then studied in a Siemens Elmiskop IA. The large
size of the homokaryons and heterokaryons necessitated precise correlation of photomicrographs and electron micrographs,
a s in figures 9a-d a s well a s the construction of montages from low power (2,0004,000 X j electron micrographs, such as
that in figure 9b.
The uterine luminal epithelium forms a
conspicuous and persisting component of
the placenta of the bandicoot. During the
blastocyst stage the endometrial lining is
a regular simple columnar epithelium composed mainly of non-ciliated cells with occasional interspersed ciliated cells. It is
closely underlain by a rich network of thinwalled blood vessels of large caliber that
provide the basis for the vascular component of the maternal placenta. As in the
North American opossum (Padykula and
Taylor, '761, the endometrium during gestation is richly glandular and stromal
ground substance is abundant, but there
are relatively few stromal cells and little
fibrous material.
During the neural fold stage (Perumeles
nasuta 305, 6.5 m m CR) the superficial
endometrial blood vessels penetrate among
the columnar epithelial cells carrying with
them a thin stromal investment. At the
same time, discrete syncytial masses appear as components of the luminal epithelium; these multinucleate cells appear to
arise through fusion of individual columnar epithelial cells. The nuclei of these
multinucleate cells look alike (plate 1, figs.
la--c) and hence these units may be called
homokaryons according to the terminology
used to describe fusion occurring among
like cells in vitro (Harris, '70). The homokaryons are held together in epithelial
fashion by junctional complexes; occasional ciliated cells are located between
them. During this period of neural fold
formation (Flynn, '23) the allantois makes
contact with the chorion.
Two placental membranes are formed
through fetal-maternal interaction, the yolk
sac placenta and the chorioallantoic placenta (text fig, 1 ) . Placental samples were
routinely taken at three sites (see text fig.
1) : (1 j the chorioallantoic placenta, ( 2 j
the vascular yolk sac placenta (visceral
splanchnopleure), and ( 3 ) non-vascular
yolk sac placenta (bilaminar omphalopleure). Figures la-c and 2a,b demonstrate histological-cytological similarities
as well as regional differences at these placental sites at two stages of gestation ( I .
~ U C T O U ~ U55,
7 m m CR and P . nusutu
307, 10-1 1 mm CR j. An anatomical similarity with evident functional significance
is the location of large diameter, thinwalled maternal capillaries in a highly superficial position among the homokaryons
of the non-vascular yolk sac (fig. l a ) , the
vascular (visceral j yolk sac (figs. l b , 2a),
and the chorioallantoic placenta (figs, l c ,
2b). This vascular penetration causes a
scalloped appearance at the basal surface
of the epithelium, as vascularized prongs
of stroma separate syncytial masses. The
cytoplasm of the homokaryons is attenuated over the most superficially placed maternal capillaries at all placental sites.
The light microscopic organization of
the three placental sites at the 7 m m CR
stage is illustrated i n figures la-c (fixation
by perfusion). In the non-vascular yolk sac
(bilaminar omphalopleure) (fig. l a ) , the
fetal component shows cytological signs of
Chorro allantorC
,Vun iiusrular
Text Fig. 1 Diagrammatic representation of
the placental relationships of the bandicoots
(Peramelidae) (Modified slightly from Amoroso,
'52). In this marsupial family, a chorioallantoic
placenta is formed in addition to the yolk sac
placenta. The location of the sinus terminalis
marks the boundary of the vascular and nonvascular regions of the yolk sac placenta. Tissue
samples were regularly taken from: (1) the chorioallantoic placenta, ( 2 ) vascular yolk sac placenta,
and ( 3 ) non-vascular yolk sac placenta.
transport activity : ( 1) the endodermal epithelium has large intercellular dilations
and ( 2 ) long microvilli cover the free surfaces of the dome-shaped trophoblastic
cells. The vacuolation of the supranuclear
trophoblastic cytoplasm may also reflect
transport functions. The large nuclei of
the trophoblastic cells contain conspicuous
central nucleoli and evenly distributed heterochromatin granules. Remnants of the
shell membrane occur in the uterine cavity
located between the microvillous trophoblastic surface and the maternal multinucleate luminal epithelium (fig. la). Here,
as in the maternal homokaryons of the
vascuIar yolk sac (fig. l b ) and the nonvascularized chorioallantoic site (fig. l c ) ,
the nuclei are all alike i n being palestaining and in having peripherally placed
large nucleoli. The pale-staining of these
nuclei originates from the relative lack
of heterochromatin and a preponderance
36 1
of euchromatin. The cytoplasm of the homokaryons is rich i n fine granules that
represent primarily a n abundance of mitochondria, a s determined from electron micrographs. The elongate dark nuclei that
seem to be within the maternal epithelium
are actually of either endothelial or perivascular origin.
The vascular yolk sac (fig. I b ) differs
from that of the non-vascular portion in
two major respects besides the presence of
vitelline vessels. It has shorter, smaller
trophoblastic cells, and greater attenuation
of the cytoplasm of the homokaryons over
the most superficial maternal capillaries.
At the presumed chorioalIantoic site at
this stage (fig. l c ) fetal blood vessels
are absent, and tall irregularly columnar
trophoblastic cells are closely apposed to
the maternal endometrial surface. The
trophoblastic cells here represent a distinct
variant in comparison to those associated
with the non-vascular (fig. l a ) and vascular (fig. l b ) yolk sac. They differ in overall
cell shape, nuclear structure, and the a p
pearance of the cytoplasm.
At a later developmental stage, 10-11
m m CR, distinct differences exist between
the structure of the vascular yolk sac (fig.
2 a ) and that of the now vascularized chorioallantoic placenta (fig. 2b). The trophoblastic cells of the vascular yolk sac are
greatly attenuated whereas those of the
chorioallantois are irregularly tall columnar with intercellular dilations. An important difference has emerged i n the syncytial cells of the chorioallantoic site; two
categories of nuclei exist in multinucleate
masses and thus it is appropriate to designate them a s heterokaryons (Harris, '70).
Pale staining nuclei, comparable to those
occurring in the syncytia associated with
the vascular yolk sac, are present; in addition; however, there are nuclei with distinct clumps of heterochromatin and large
nucleoli that are more centrally placed
(compare figs. 2a, 2b). The more heterochromatic nuclei tend to be larger than the
pale nuclei; also they resemble closely the
nuclei of the trophoblastic cells. Compare
figure 2b with l c to see that, before vascularization, the multinucleate cells of the
chorioallantoic placenta are homokaryons
whereas after vascularization they are heterokaryons. A striking difference is also
evident in the cytoplasm of the multinu-
cleate cells a t these two placental sites. In
those associated with the visceral yolk sac,
the mitochondria are aggregated as a large
central pool (fig. 2a), whereas a t the chorioallantoic site, there is no evident mitochondrial aggregation (fig. 2b).
Our observations on this stage (10-11
m m CR) of placental differentiation agree
with those of Hill (1897) and Flynn ('23)
in regard to regional differences in the
state of the trophoblastic cells. Those 10cated peripherally i n the placental d ~ s c
tend to be more regular in form and arrangement than those more centrally positioned.
A remarkable histological change occurs
in the chorioallantoic barrier during the
period of embryonic growth from 10-12
mm CR length. The conspicuous trophoblastic layer of the 10 mm CR stage (fig.
2b) has disappeared by the 12 m m CR
stage (fig. 3 ) . As a result, the fetal allantoic and maternal blood vessels are placed
into much closer proximity. The two categories of nuclei persist in the syncytial
masses after the disappearance of the
trophoblastic layer. Furthermore, the heterochromatic nuclei are much larger than
the pale nuclei at the 10-11 m m CR stage
(fig. 2b). The ultrastructure of the two nuclear forms at the 12 mm CR stage is illustrated in figures 4a,b.
It is significant that, as late as the 10
m m CR stage, the trophoblastic cells of the
chorioallantoic placenta have ultrastruct u r d features that are usually interpreted
as being characteristic of undifferentiated
cells. Their cytoplasm is rich i n ribosomespolysomes, and the membrane systems are
relatively sparse (figs. 5, 8 ) . A special feature of their cytoplasm is the occurrence of
clusters of aggregated dense material (fig.
7). The maternal pole of the trophoblast is
differentiated into a microvillous border
that interdigitates closely with comparable
projections from the maternal syncytium
(figs. 6, 7). Syncytial masses and ciliated
cells on the maternal side interact with the
trophoblastic surface (fig. 6 ) , and desmosoma1 associations are formed between the
fetal and maternal surfaces. At the fetal
pole of the trophoblastic cells, however,
the surface is quite smooth and closely
apposed to the allantoic endothelium
with minimal intervening extracellular substance (fig. 8).
The undifferentiated state of the chorioallantoic trophoblast (figs. 5, 8 ) before its
disappearance as a layer is a prominent
feature. Although some ultrastructural
manifestations of transport activity were
observed in these trophoblastic cells in Isoodoia 55 and Perameles 307, they were
meager (i.e., relatively few coated vesicles,
vacuoles, and structures resembling multivesicular bodies). However, the trophoblastic cells, associated with the vascular and
non-vascular region of the yolk sac placenta in these specimens, possess specializations at the cell surface and in the membrane systems that are characteristic of
absorptive cells (Padykula and Taylor, unpublished observations). This regional differential in trophoblastic specialization
may reflect differing functional activities.
In the specimen, I . macroums 2, 12 m m
CR, i t has been possible to delineate the
cellular layers of the chorioallantoic placenta near term. Our evidence indicates
that in the thinnest part of the barrier the
fetal allantoic blood vessels are separated
from the maternal endometrial vessels by
highly attenuated cytoplasmic extensions
from the syncytial masses (heterokaryons),
and by a layer of maternal stroma. These
relationships are illustrated by four micrographs of a thin region of the chorioallantoic placenta which is shown at progressively higher magnification and resolution.
In figure Sa at three points (arrows), the
continuity of the attenuated layers of
cytoplasm with the body of the syncytial
masses can be seen. Figure 9b identifies
the fetal and maternal endothelia, and the
epithelial and stromal layers that intervene
between the two bloodstreams. Figure 9c
allows visualization of the components of
the maternal stroma, particularly the collagen fibrils and profiles of stromal cells.
Although the ultrastructural preservation
of this specimen in figure 9c is suboptimal,
it allows recognition of the components of
the chorioallantoic placental barrier near
term as follows starting from the maternal
blood : ( 1 ) maternal endothelium; ( 2 )
basal lamina of the maternal endothelium;
( 3 ) maternal stroma; ( 4 ) syncytial masses; and (5) fetal endothelium (with in-
distinct basal lamina). Finally figure 9d
defines the intimate interaction of the pleomorphic distal surface of the syncytium
with the basal surface of the allantoic endothelium.
To set a n appropriate background for
discussion, it should be indicated that current knowledge on chorioallantoic placentation in the bandicoots rests on relatively
few specimens that are beyond the blastocyst stage. Hill (1897, 1900) provided four
specimens: (Stage B, 7 mm; Stage C, 7
m m ; Stage D (1897), 8.75 mm; Stage D
(1900), 12.5 m m ) . Flynn ('23) added two
specimens (Stage 1, 6.1 m m and Stage 2,
6.6 m m ) , and reexamined some of Hill's
slides. Our collection (table 1) adds eight
new placental specimens and this brings
the total number examined and reported
to only 1 4 placentae.
I n both of Flynn's specimens the placenta was infiltrated by polymorphonuclear
leucocytes. Moors ('74) uses the presence
of these leucocytes in Flynn's material to
support the hypothesis that a maternal
homograft reaction occurs in marsupials
once the shell membrane degenerates, and
the fetal and maternal tissues juxtapose.
He interprets this reaction as the mechanism which imposes a time limit on marsupial intrauterine development.
This summary emphasizes the need for
devising a method for obtaining timed
pregnancies in the wild animal under laboratory conditions as has been done for
the opossum, Didelphis (New and Mizell,
'72; Renfree, '74) and the tammar wallaby,
Macropus (Renfree and Tyndale-Biscoe,
'73a). This is particularly critical in the
bandicoots since chorioallantoic placentation proceeds extremely rapidly in late gestation,
The entire differentiation of the chorioallantoic placenta must take place within
the last two (plus) days of gestation. For
example, Isoodon 55, CR 7 mm (early placental stage) had been isolated from males
in captivity for 10.75 days; hence a maximum of 1.75 days remained for radical
placental changes, such as here described,
and for close to doubling of fetal length
(table 1).
The possibility of fusion between fetal
and maternal cells in t h e
bandicoot placenta
We have been able to confirm most of
Hill's elegant 1897 histological description
of bandicoot placentation. Our plastic sections, of course, allow more definition of
intercellular relationships and intracellular
differentiation. The disappearance of the
trophoblastic layer a t the chorioallantoic
site during late gestation was first reported
by Hill (1897) and confirmed by Flynn
('23) as well as by the observations reported here. Cellular analysis of the mode
of removal of the trophoblastic layer must
await further work involving a close sequence of specimens from 10-14 mm CR.
Following an original suggestion of Hubrecht ('Og), Flynn ('23) presented the
stimulating hypothesis that the chorioallantoic trophoblast fused with the maternal syncytium and disappeared through
this merger. Although Hill (1897) had reported disappearance through degeneration, his drawing did not illustrate this;
eventually he accepted Flynn's hypothesis
(Hill, '32; Pearson, '49). Our evidence obtained from comparison of the 10 and 12
m m specimens favors the possibility of
fusion on the basis of the following two
pieces of new evidence as well as from correlations with the literature.
First, the maternal syncytial masses of
the chorioallantoic placenta have two categories of nuclei (heterokaryons) whereas
those associated with the yolk sac placenta
have only one of these nuclear types (homokaryons) (compare figs. 2a and 2b).
Hill (1897) did not mention this, although
some of his drawings show two nuclear
populations in the allantoic syncytia. Flynn
('23) emphasized the difference between
the nuclear structure of the trophoblastic
cells and syncytial masses, but did not report the condition evident i n the present
figure 2b which shows an intact columnar
trophoblastic layer and subjacent syncytial
masses already with two populations of nuclei. A s a n alternate interpretation to fusion, i t is reasonable to consider that the
presence of the chorioallantoic membranes
may have a n inductive effect on the maternal syncytia which stimulates RNA synthesis in some of the nuclei (e.g., fig. 4b)
and in turn new protein synthesis. How- allantoic (fig. 2b) syncytia may be a gross
ever, the dramatic disappearance of the cellular reflection of a change i n metabolic
trophoblast as a layer makes Flynn’s ( ’ 2 3 ) activity in the heterokaryons. In other
interpretation, that the two nuclear popu- words, the fetal genome might significantly
lations reflect the product of cellular fu- alter the metabolic and transport activity
sion, quite attractive. The large hetero- of the maternal syncytia for the specific
chromatic nuclei of the syncytia resemble needs of the extremely rapid final fetal difclosely those of individual trophoblastic ferentiation. The matter of immunologic
cells. Some nuclei of the heterokaryons in incompatibility would be lessened since the
the 12 mm CR stage are exceptionally fused complex exists only two to three
large, and this suggests that nuclear fu- days in. utero and could be eliminated postsion may occur (Harris, ”70). A second natally by phagocytes migrating from the
strong reason for favoring fusion is de- endometrial stroma. (The chorioallantoic
rived from our ultrastructural observations placenta is non-deciduate [Hill, 18971.)
which characterize the trophoblastic cells Harris (’70) concludes that the cells of
a s being undifferentiated, ribosome-rich different vertebrate species are immunocells as late as the 10-11 m m embryo logically compatible with each other after
stage. This ultrastructural image indicates in vitro fusion into a single hybrid entity.
considerable potential for future differen- Because the cytological evidence presented
tiation and thus seems incompatible with here strongly favors the hypothesis that
the possibility of degeneration. Further- cell fusion occurs between fetal and mamore, study of Hill’s (1897) drawings re- ternal cells of different genome, i t is worth
veals no sign of placental degeneration, citing Harris’ (’70) conclusion directly:
except in the day 1 postpartum prepara“It thus appears that in the cells of vertetions (Stage E ) . Nevertheless, i n the abbrates there are, in general, no intmcellular
mechanisms for the recognition of incomsence of direct proof to the contrary, it
patibility similar to those responsible for the
should be kept in mind that trophoblastic
recognition and destruction of tissue or ordegeneration might occur rapidly in develgan grafts exchanged between different inopmental stages not yet studied.
dividuals. Not only do the cytoplasms of
these different cells fuse amicably together,
The function of the bandicoot allantoic
but their nuclei dso; and after nuclear futrophoblast may be related to attachment,
sion has taken place the composite cell cara s suggested by Hill ( 1897), page 404:
ries out its functions in a perfectly integrated
“The role of the ectoderm is apparently merely
that of attaching the embryo to the previously prepared maternal syncytium. Once the
allantoic capillaries have spread out on its
inner surface, it degenerates and disappears
in order to allow of closer proximity between
the foetal and maternal capillaries, and thus
takes no part in the constitution of the functional placenta.”
The trophoblast may be attached by microvillous interdigitation, as in figure 7, which
includes desmosomal complexes (as in fig.
6 ) shared by fetal and maternal cellular
surfaces. Actual fusion of the fetal trophoblast with the maternal homokaryons
would effect the firmest of physical attachments. Furthermore, as in various in vitro
experiments on cell fusion (Harris, ’ 7 0 ) ,
the metabolic expression of maternal syncytial masses might be influenced considerably through fusion with the fetal trophoblast. The difference in the mitochondria1
distribution of the yolk sac (fig. 2a) and
way, and may, in some cases, undergo vigorous and indefinite multiplication.”
The possibility of fusion gains some impressive support from current knowledge
of the progestational behavior of the uterine luminal epithelium in the rabbit. Enders and Schlafke (’71) demonstrated
ultrastructurally that, during early implantation in the rabbit, knobs of syncytial
trophoblast containing many fetal nuclei
fuse with individual luminal epithelial
cells. After this union with the trophoblast,
the uterine cell retains its junctional associations with adjacent cells. Enders (’72)
has stated that the uterine cytoplasm is
“rapidly converted into ‘trophoblast’ cytoplasm’’ (p, 319). Further differentiation
involves expansion of the trophoblastic peg
apparently without incorporation of more
luminal epithelial cells. Several investigators (Larsen, ’61; Steer, ’71; Enders and
Schlafke, ’71 ) have reported the existence
of two types of nuclei within syncytial
masses during later implantation in the
rabbit. Hence in this eutherian event there
is considerable potential similarity with
the cytological picture in the marsupial
chorioallantoic placenta. However, in the
rabbit the embryonic-maternal fusion represents a transient stage in that “the
resultant cytoplasmic mass rather than
being an invasive tissue is believed to degenerate” (Enders, ’72, p. 319). In the
bandicoot the postulated fusion would occur much later in gestation and would create a persisting continuous placental layer,
Additional similarity between the bandicoot and rabbit resides in the well-known
fact that the simple columnar luminal epithelium is converted into multinucleated
masses during pseudopregnancy and pregnancy (Larsen, ’62). The multinucleate
cells arise through an initial phase of mitosis that is followed by cell fusion. Recently Davies and Hoffman (’75) reported
that these two events are accompanied by
nuclear differentiation, i.e., during the proliferative phase the nuclei have prominent
nucleoli and many pores whereas before
fusion both these features diminish. In
their micrographs the nuclei of the resultant “multinucleate cells” appear to be all
alike. This homogeneity in the absence of
embryonic tissue enhances the possibility
that the heterogeneity in nuclear populations in the bandicoot chorioallantoic placenta may represent fusion of different
cell types.
The structure of the chorioallantoic
placental barrier near term
The chorioallantoic placental barrier
near term, as shown in figures 9a-d, involves apposition of the fetal allantoic capillaries to an intact syncytial layer, that is,
in part or entirely, of maternal origin.
Thus, in relation to Grosser’s histological
scheme (’09) for placental classification,
this barrier might be designated epithelioendothelial in its final organization. HOWever, such designation should be withheld
until direct evidence allows a decision to
be made about whether or not the fetal
trophoblast has merged with the original
maternal syncytium.
Our correlated light and electron micro-
scopic observations (figs. 9a-d) establish
that in the thin regions of the chorioallantic placental barrier the fetal and maternal
capillaries are separated by two layers :
(1) a thin layer of syncytial cytoplasm,
and ( 2 ) maternal stroma. These photoand electron micrographs are partly in
agreement with Hill’s ( 1897) description
which is quoted here (pp. 414, 415):
“With this irregularly ridged and highly vascular surface the allantoic capillaries are in
most intimate contact; so close, indeed, is the
attachment that the walls of the capillaries
appear as if united with the syncytial cytoplasm. . . . it will be noted that the foetal
and maternal blood-streams are separated
from each other only by the thickness of two
endothelial walls, with at most the addition
of a thin layer of syncytial protoplasm.”
Hill’s (1897) drawing (his fig. 17) illustrates the syncytial layer that intervenes
between the two blood streams; however,
the stromal layer was not resolved. From
the above description, i t is understandable
that this barrier was designated as probably “endothelial-endothelial” in organization (Amoroso, ’52, ’55).
Since eutherian chorioallantoic placentae are endocrine organs, the possibility of
such a function should be considered for
this marsupial counterpart. Recent evidence indicates that the placenta of the
quokka ( S e t o n i x ) contains some enzymes
involved in the synthesis of progesterone
(Bradshaw et al., ’75). From the cytological evidence currently available on the
bandicoot trophoblast, e.g., the paucity of
membrane systems, there is no obvious
visible signal that steroid or protein molecules are being synthesized. However, the
membrane systems of the endoplasmic reticulum and Golgi complex in the heterokaryons are elaborately developed (Padykula and Taylor, unpublished observations),
and thus the secretory and absorptive potential is considerable. If the syncytium of
late gestation is the resultant of fused
trophoblast and maternal luminal epithelium, the known ability of trophoblast
(from eutherian studies) to synthesize
hormones might be expressed through interaction of the two metabolic cellular systems. This entire issue will have to await
direct study.
The structural uniqueness of the loss of
trophoblast in the bandicoot is reinforced
when considered alongside Wynn’s (’73)
recent review of eutherian placental structure :
“Finally it is significant that, in every placenta examined by electron microscopy, at
least one layer of trophoblast persists essentially throughout gestation. The conclusion
that immunologic protection of the placental
homograft is provided, at least in part, by the
trophoblast and the pericellular sialomucins
on its plasma membranes therefore seems
reasonable.” (p. 273.)
Thus the absence of a trophoblastic layer
near term endows the bandicoot chorioallantoic placenta with an ontogenetic
uniqueness that may be phylogenetically
significant in marsupial evolution,
We express our gratitude to the Ian
Clunies Ross Animal Research Laboratory
in the Division of Animal Physiology of the
Commonwealth Scientific and Industrial
Research Organization in Australia for providing us most generously with excellent
facilities and professional assistance. Our
sponsor at this Laboratory was Dr. A. Gordon Lyne whose impressive knowledge of
natural history of the bandicoot facilitated
this investigation. The method for perfusion of the bandicoots was devised and
executed under the expert guidance of Dr.
Brian D. Stacy. Mr. Robert Gemmell assisted in all aspects of the fixation and
processing of the specimens. His enthusiastic informed participation is most gratefully acknowledged. Special care and handling of captive bandicoots was skillfully
done by Mrs. Robyn Smith and her technical assistance i n preparing the animals for
laboratory investigation is also greatly appreciated. We are grateful for the aid given
by Mr. David E. Hollis and Mr. Chris Williams in locating and trapping bandicoots
used in this study. At the Wellesley laboratory, Ann G. Campbell and Ann W. Hobbs
prepared the semi-thin and ultrathin sections; Richard V. T. Steams and Tim Burke
prepared the photographic illustrations.
Amoroso, E. C. 1952 Placentation. In: Marshall’s
Physiology of Reproduction. Vol. 2. Third ed.
A. S . Parkes, ed. Longmans Green & Co., London, pp. 127-311.
1955 The Comparative Anatomy and
Histology of the Placental Barrier. Josiah
Macy, Jr. Foundation. First Conference on Gestation, New York.
Bradshaw, S. D., I. R. McDonald, R. Hiihnel and
H. Heller 1975 Synthesis of progesterone by
the placenta of a marsupial. J. Endocr., 65:
Davies, J., and L. H. Hoffman 1975 Studies on
the progestational endometrium of the rabbit.
11. Electron microscopy, day 0 to day 13 of
gonadotrophin-induced pseudopregnancy. Am. J.
Anat., 142: 335-366.
Enders, A. C. 1972 Mechanisms of implantation of the blastocyst. In: Biology of Reproduction. J. E. Velardo and B. A. Xasprow, eds. I11
Pan American Congress of Anatomy, New
Orleans, pp. 313-333.
Enders, A. C., and S. J. Schlafke 1971 Pcnetration of the uterine epithelium during implantation in the rabbit. Am. J. Anat., 132: 219-240.
Flynn, T. T. 1922 The phylogenetic significance
of the marsupial allantoplacenta. Proc. Linn.
SOC.,N.S.W., 47: 541-544.
1923 The yolk-sac and allantoic placenta i n Perameles. Quart. J. Micr. Sci., 67:
123-1 82.
Grosser, 0. 1909 Vergleichende Anatomie und
Entwicklungsgeschichte der Eihaute und der
Placenta. Wien-Leipzig, W. Braumuller.
Harris, H. 1970 Cell Fusion. The Dunham Lectures. Harvard University Press, Cambridge.
Hill, J. P. 1895 Preliminary note on the occurrence of a placental connection i n Perameles
obeszda, and on the foetal membranes of certain Macropods. Proc. Linn. SOC.,N.S.W., 10:
1897 The placentation of Perameles.
Quart. J. Micr. Sci., 40: 385-446.
1900 Contributions to the embryology
of the Marsupialia. 2. O n a further stage of
placentation of Perameles. 3. On the foetal
membrane of Mucropus purma. Quart. J. Micr.
Sci., 43: 1-22.
1932 I1 Crooniari Lecture. The developmental history of the primates. Phil. Trans. Roy.
SOC.(London), Series B1,221: 45-178.
Hubrecht, A . A. W. 1909 Early ontogenetic
phenomena in mammals and their bearing o n
our interpretation of the phylogeny of the vertebrates. Quart. J. Micr. Sci., 53: 1-181.
Hughes, R. L. 1962 Role of the corpus luteum
i n marsupial reproduction. Nature, 1 9 4 : 890891.
1974 Morphological studies on implantation i n marsupials. J . Reprod. Fert., 39: 173186.
Karnovsky, M. J. 1965 A formaldehyde-glutaraldehyde fixative of high osmolality for use
in electron microscopy. J. Cell Biol., 27: 137A.
Kirby, D.R.S. 1968 Immunological aspects of
pregnancy. In: Advances in Reproductive Physiology. Vol. 3. A. McLaren, ed. pp. 33-79.
Larsen, J. F. 1961 Electron nlicroscopy of the
implantation site i n the rabbit. Am. J. Anat.,
109: 319-334.
1962 Electron microscopy of the uterine epithelium in the rabbit. J. Cell Biol., 14:
Lyne, A. G. 1964 Observations on the breeding
and growth of the marsupial Perameles nasuta
Geoffroy, with notes on other bandicoots. Aust.
J. ZOO^., 12: 322-339.
1974 Gestation period and birth in the
marsupial Isoodon macrourus. Aust. J. Zool.,
22: 303-309.
Moors, P. J. 1974 The foeto-maternal relationship and its significance in marsupial reproduction: a unifying hypothesis. J. Aust. Mammal
SOC., 1: 263-266.
New, D. A. T., and M. Mizell 1972 Opossum
fetuses grown in culture. Science, 175: 533.
Padykula, H. A., and J. M. Taylor 1974 Cytological observations on marsupial placentation:
The Australian bandicoots (Perameles and
Isoodon). Anat. Rec., 178: 434 (abstract).
1976 Cellular mechanisms involved i n
cyclic stromal renewal of the uterus. I. The
opossum, Didelphis virginiana. Anat. Rec., 184:
Pearson, J. 1949 Placentation of the Marsupialia. Proc. Linn. SOC.Lond., 161: 1-9.
Peters, A. F., C. C. Proskauer and I. R. KaisermanAbramof 1968 The small pyramidal neuron
of the rat cerebral cortex. The axon hillock and
initial segment. J. Cell Biol., 39: 604-619.
Reese, T., and M. J. Karnovsky 1967 Fine struc-
tural localization of blood-brain barrier to exogenous peroxidase. J. Cell Biol., 34: 207-217.
Renfree, M. B. 1974 Ovariectomy during gestation in the American opossum, Didelphis marsupialis virginiana. J. Reprod. Fert., 39: 127-130.
Renfree, M. B., and C. H. Tyndale-Biscoe 1973
Intrauterine development after diapause in the
marsupial Macropus eugenii. Dev. B i d , 32:
Schlafke, S., and A. C. Enders 1975 Cellular
basis of interaction between trophoblast and
uterus at implantation. Biol. Reprod., 12: 41-65.
Sharman, G. B. 1969 Reproductive physiology
of marsupials. Science, 167: 1221-1228.
Steer, H. W. 1971 Implantation of the rabbit
blastocyst: The adhesive phase of implantation.
J. Anat., 109: 215-228.
Stodart, E. 1966 Management and behaviour
of breeding groups of the marsupial Perameles
nasuta Geoffroy in captivity. Aust. J. Zool.,
14: 611-623.
Tyndale-Biscoe, C. H. 1973 Life of Marsupials.
Edward Arnold, Ltd., London.
1974 Reproduction in marsupials. J.
Aust. Mammal SOC.,1: 175-180.
Wynn, R. M. 1973 Fine structure of the placenta. In: Handbook of Physiology. Vol. 11. Section 7, Endocrinology. Female reproductive system, part 2. R. 0. Greep, E. B. Astwood and
S. R. Geiger, eds. American Physiological SOC.,
Washington, D.C., pp. 261-276.
Figs. la-c.
Yolk sac placenta and chorioallantoic placenta, lsoodon mucrourus 5 5 ;
7 mm CR. Semi-thin plastic sections, toluidine blue, x 400.
Bilaminar omphalopleure of the yolk sac placenta.
The non-vascularizcd yolk sac (bilaininar omphalopleure) occurs at the top,
and consists of a n endodermal layer ( E ) characterized by large intcrcellular spaces
and a columnar layer of trophoblastic cells (T) that have distinct microvillous
borders and vacuolated supranuclear cytoplasm. Remnants of the shell membrane
(sm) occur i n the compartment between the fetal trophoblast ( T ) and maternal
syncytial ( S ) luminal epithelium. At this time the maternal epithelium is composed of individual multinucleate masses (homokaryons). Note the extremely superficial penetration of the large diameter thin-walled endometrial vessel (mv)
and the attenuation of the uterine epithelium at such sites (arrows). Note also
that the nuclei of the homokaryons are structurally alike; differences in nuclear
diameter are caused by plane of section. CT, endometrial stroma.
Visceral layer (vascular splanchnopleure) of the yolk sac placenta.
A vitelline vessel ( f v ) occurs at the top (the mesenchyme and endoderm are
not included here) and contains two nucleated blood cells. The trophoblastic layer
( T ) is thinner than that i n the non-vascular yolk sac (fig. l a ) and that at the
chorioallantoic site (fig. l c ) . The fetal-maternal junction is marked by asterisks.
The maternal thin-walled blood vessel (mv) appears closely applied to the trophoblastic layer because of the extreme attenuation of the cytoplasm of the syncytial
masses of the luminal epithelium. Note the homogeneity of the nuclei of the maternal homokaryons ( S ) . The granules i n the cytoplasm of the homokaryons represent mitochondria primarily. CT, endometrial stroma. Compare with the latter
developmental stage shown in figure 2.
Chorioallantoic placental region without fetal vascularization.
The fetal placental components in the upper field consist of inesothelial cells
( m ) and an irregularly columnar trophoblastic layer ( T ) . Some trophoblastic cells
protrude slightly into the maternal tissue (arrows). The nuclear structure of these
trophoblastic cells diIPers from the trophoblast i n the two regions of the yolk sac
placenta (figs. l a , and b). The fetal-maternal junction is marked by an asterisk
( * ) . Compare also the chorioallantoic trophoblastic structure with that of the later
developmental stage shown in figure 2 . The syncytial masses ( S ) are large, are
homokaryons, and possess numerous mitochondria, that are seen here as fine cytoplasmic granules. The superficial location of maternal vessels (mv) that ramify
among the homokaryons is evident.
Helen A. Padykula and J. Mary Taylor
3 70
Visceral layer of the yolk sac placenta, Perumeles nasutu 307; 10-11 mm CR. Semithin plastic section, toluidine blue. x 400.
The large syncytial masses ( S ) or homokaryons that comprise the luminal
epithelium of the uterus contain many pale nuclei ( n ) and central aggregations of
mitochondria ( m ) . Maternal blood vessels ( m v ) have penetrated between syncytial masses to acquire an extremely superficial position. The trophoblastic surface
of the visceral yolk sac has separated from the uterine surface leaving a n artifactual space (x). Fetal blood vessels (fv) occur i n the mesenchyme between the
attenuated trophoblast ( T ) and the visceral endoderm ( E ) . CT, endometrial stroma.
Compare with figure 2b.
Chorioallantoic placenta, Perumeles nasutu 307; 10-1 1 mm CR. Semi-thin plastic
section, toluidine blue. x 400.
The syncytial masses ( S ) are distinctly separated by stromal septa that place
the maternal blood vessels (mv) into a most superficial location. Note the two
types of nuclei in the syncytial masses, pale ones ( n l ) with peripheral nucleoli
and distinctly heterochromatic ones (nz) with large nucleoli and more heterochromatin. The more heterochromatic nuclei tend to be larger. Thus these syncytial masses are heterokaryons in comparison with those of the yolk sac placenta
(fig. 2a). At this placental site, the trophoblast cells ( T ) are conspicuous and
tall. Compare the nuclear structure of the trophoblastic cells with that of the
larger, heterochromatic nuclei of the heterokaryons. Fetal blood vessels ( f v ) occur
i n the mesenchyme beneath the trophoblast. Compare with figure 2a. CT, endometrial shoma.
S 12 mm CR. Semi-thin plastic secChorioallantoic placenta, Isoodon ~ U C T O I L ~ U2;
tion, toluidine blue. x 480.
As in the earlier stage of chorioallantoic placentation shown in figure 2b,
the syncytial masses ( S ) are heterokaryons with two distinct types of nuclei,
small pale ones ( n l ) and large, more heterochromatic ones ( n 2 ) . Also i n the
heterokaryons the heterochromatic nuclei appear to be larger than the heterochromatic ones at the earlier stage (fig. 2b). The striking change that occurs near
term is the disappearance of the trophoblast (compare with fig. 2b). As a result,
the maternal (mv) and fetal (fv) blood vessels, are placed in close apposition.
CT, endometrial stroma.
Helen A. Padykula and J. Mary Taylor
4a, b
3 72
Two nuclear types i n the syncytial masses at the chorioallantoic
site. Isoodon macrourus 2; 12 mm CR. 4a, x 7,000; 4b, x 13.500.
The pale nuclei of the syncytia shown in figure 4a have two major
properties, the peripheral position of the small nucleoli (Nc) and
the paucity of heterochromatin. The other nuclear population (fig.
4b) is characterized by opposite qualities, large nucleoli (Nc) more
central i n position and conspicuous heterochromatin ( H ) .
Helen A. Padykula and J. Mary Taylor
Ultrastructure of the nuclear region of the columnar trophoblastic
cells of the chorioallantoic placenta. lsoodon mucTourus 5 5 ; 7 mm CR.
X 9,500.
Note the large intercellular spaces ( I S ) and the loose associations
of the lateral surfaces of the trophoblastic cells. This specimen was
fixed by perfusion, and this procedure may have influenced the size
of the extracellular compartment. The hophoblastic nuclei possess
large nucleoli (Nc) and distinct clumps of heterochromatin ( H ) . The
predominant cytoplasmic features are a n abundance of ribosomal
material and a paucity of membranes. Cisternae of rough endoplasmic
reticulum (arrows) occur singly. m, mitochondria.
Helen A. Padykula and J. Mary Taylor
3 75
Ultrastructure of the chorioallantoic placental barrier at the fetomaternal junction. lsoodon nzacrourus 55; 7 mm CR. X 11,500.
The trophoblastic cells (TR j are interrelated with the syncytial
masses ( S ) by microvillous (mv::”) processes from each surface that
may form desmosomal ( D ) associations. The interweaving of microvilli is complex and irregular. The trophoblastic cytoplasm is rich i n
free polysomes and poor in membranes. Cytological signs of transport
activity i n the trophoblast are meager, e.g., coated vesicle (cv) and
vacuole ( v ) with content. At this location, a maternal ciliated cell
(C) is linked to a syncytial mass by the usual epithelial intercellular
junctions (arrows). The surface area of the syncytial masses i s increased by proximal ( * ) as well as distal ( ‘ k + j microvillous projections. The proximal microvilli rest on a n epithelial basal lamina (bl);
however, i n this field the separateness of the epithelial basal lamina
from that of the basal lamina of the maternal endothelium ( M E ) is
not as evident as in figure 7. MV, maternal vessel; m, mitochondria;
bl, basal lamina.
Helen A. Padykula and J. Mary Taylor
Ultrastructure of the chorioallantoic placental barrier at the maternal
pole. lsoodon ~ U C ~ O U T U5S5 ; 7 mm CR. x 33,200.
At the fetal-maternal junction, microvilli (mv) of fetal and maternal origin interdigitate. The fetal trophoblast ( T R ) is rich i n r h o somal-polysomal material; cytoplasmic membrane systems are relatively sparse (arrows); and clustered densities ( d ) are characteristic.
The lateral surfaces of two syncytial masses ( S ) are closely apposed
( " ) and possess specialized junctions (arrowhead) at the apical surfaces. A thin stromal layer (CT) intervenes between the basal laminae
( b l ) of the syncytial epithelium and the endothelium ( M E ) of the
maternal capillaries. MV, maternal vessel; m, mitochondria.
Helen A. Padykula and J. Mary Taylor
3 79
Ultrastructure of the trophoblast in the chorioallantoic placenta at
the fetal pole. Perameles nasuta 307; 10-11 m m CR. x 37,500.
The fetal surfaces of the columnar trophoblast cells ( T R ) (fig. 2b)
abut o n the allantoic endothelium ( A E ) without much intervening
extracellular material ( " ) . The lateral surfaces of 2 trophoblast cells
are loosely associated, and there are wide intercellular spaces ( I S ) .
Polysomes predominate in this region of the trophoblastic cytoplasm.
fv, fetal allantoic vessel.
Helen A. Padykula and J. Mary Taylor
38 1
Chorioallantoic placental barrier near term. Isoodon macrourus 2; 12 mm CR.
Semi-thin plastic section, toluidine blue. x 800.
The structural relationships of the chorioallantoic barrier are shown here by
light microscopy to guide interpretation of the electron micrographs shown i n
figures 9b-d. The maternal (mv j and fetal (fv j blood vessels are closely apposed.
Note that cytoplasmic extensions (arrows) from the maternal syncytial masses
( S ) , as well as endometrial stroma ( c t ) , separate the maternal and fetal blood
vessels. The rem'arkable feature is the absence of the trophoblast (see fig. 3 also).
The boxed area is magnified in figure 9b.
Ultrastructure of the chorioallantoic placental barrier near term. Isoodon macrourus
2 ; 12 mm CR. x 3,750.
Notice the nucleated red blood cells i n the fetal allantoic vessel. Starting on the
maternal side, the tissue layers that intervene between the maternal and fetal
blood streams are: ( 1 ) maternal endothelium; ( 2 j endometrial stroma; ( 3 ) syncytium which has originated at least in part from the maternal uterine luminal
epithelium; and ( 4 ) fetal endothelium. The boxed area is enlarged i n figure 9c.
Ultrastructure of the chorioallantoic placental barrier near term. Isoodon mucrourus
2; 12 mm CR. x 16,000.
The components of the several tissue layers i n the placental barrier within the
boxed area of figure 9b are here enlarged. The maternal endothelium ( M E ) has a
distinct basal lamina (bl). The endometrial stroma (CT) here contains portions
of cells, collagen fibrils (cf), and ground substance (gs). The syncytium ( S ) is
separated from the allantoic endothelium (FE) by a region that contains cytoplasmic processes that are probably of syncytial origin. Some artifactual separation may have occurred at the fetal-maternal interface.
Helen A. Padykula and J. Mary Taylor
2; 12 mm CR.
Ultrastructure of the chorioallantoic placental barrier near term. Isoodon macrourus
x 14,300.
This area taken from the lower right corner of figure 9b illustrates the close
association of microvillous processes from syncytial masses ( S ) with the basal
surface of the fetal allantoic endothelium (FE). The pleornorphic surfaces of the
syncytial cytoplasmic extensions suggest mobility and intense activity. RBC, fetal
red blood cell; CT, maternal stroma.
Helen A. Padykula and J. M a r y Taylor
Без категории
Размер файла
2 497 Кб
loss, ultrastructure, bandicoot, trophoblast, evidence, layer, placental, chorioallantoic, australia, marsupialiaperamelidae
Пожаловаться на содержимое документа