Villous sprouting: fundamental mechanisms of

Human Reproduction Update 2000, Vol. 6, No. 5 pp. 485±494
Ó European Society of Human Reproduction and Embrology
Villous sprouting: fundamental mechanisms of
human placental development
M.Castellucci1,*, G.Kosanke2, F.Verdenelli1, B.Huppertz2 and P.Kaufmann2
1
Institute of Normal Human Morphology, Faculty of Medicine, Ancona, Italy and 2Department of Anatomy, University of
Technology, Aachen, Germany
There is increasing evidence that maldevelopment of the placental villous tree can play an important role in the
pathogenesis of various pregnancy diseases. In this review we present the most recent advances of cellular and
molecular mechanisms involved in the early formation of chorionic villi. In particular we focus our attention on the
structural events during early villous sprouting leading to the formation of the mesenchymal villi which are the
forerunners of all other villous types, i.e. immature intermediate villi, stem villi, mature intermediate villi and
terminal villi. Early villous sprouting starts as `hot spots' which are circumscribed areas consisting of highly
proliferating cytotrophoblastic and stromal cells. The post-proliferative cytotrophoblastic cells fuse with the
overlying syncytium leading to the formation of the trophoblastic sprouts. When villous mesenchyme invades the
trophoblastic sprouts, the latter are transformed into villous sprouts. The vascularization of the villous sprouts leads
to the formation of the mesenchymal villi, the most basic villous type. This process is repeated throughout pregnancy.
We analyse the in¯uence of various extracellular matrix molecules, e.g. tenascin and hyaluronic acid, on the
formation of hot spots and mesenchymal villi as well as the transformation of the latter in other villous types. We
present a critical survey on the data on vessel formation related to villous sprouting and morphogenesis of
mesenchymal villi as well as the expression of various angiogenic factors and their receptors.
Key words: angiogenesis/chorionic villi/extracellular matrix/growth factors/placenta
TABLE OF CONTENTS
Introduction
Villous sprouting: the key event for placental development
Structural events during early villous sprouting
Extracellular matrix
Angiogenesis, growth factors and growth factor receptors
Heterogeneity of mesenchymal villi
Conclusions
Acknowledgements
References
Introduction
Recent advances in non-invasive obstetric diagnostics such as
Doppler ultrasound have increased understanding in the pathogenesis of pregnancy diseases. It has been shown that several of
these conditions resulting in poor neonatal outcome are
aetiologically closely related to maldevelopment of the placental
villous trees (Macara et al., 1995, 1996; Fox, 1997; Kingdom and
Kaufmann, 1997; Kingdom, 1998; Benirschke and Kaufmann,
2000). This progress in clinical knowledge has raised urgent
demands for a better knowledge of the mechanisms of human
placental villous development.
The most important data concerning the morphology of the
various types of villi and their differentiation pathways have been
obtained by combinations of various techniques such as light and
electron microscopy, morphometry, immunohistochemistry and
in-situ hybridization. Experiments in animal models are completely missing. The reasons for this are the great structural
differences between the human placenta and those of most other
mammals. The only exceptions are some higher primates which,
however, are not easily available for obvious reasons. As a
consequence of this situation, experimental data concerning
speci®c villous functions and regulation of villous development
are largely lacking. This is deplorable since ~10% of human
pregnancies suffer from pregnancy complications associated with
unexplained villous maldevelopment, ®nally resulting in fetal and
neonatal problems ranging from untimely birth, via intrauterine
growth restriction, partly with mental retardation, to intrauterine
and neonatal death (Macara et al., 1996; Todros et al., 1999;
Benirschke and Kaufmann, 2000). There is increasing evidence
that some of these fetal problems, e.g. intrauterine growth
retardation, can play an important role in the development of
various diseases in later adult life (Barker, 1992; Barker et al.,
1993; Purdy and Metzger, 1996; Kadyrov et al., 1998).
*
To whom correspondence should be addressed at: Institute of Normal Human Morphology, Faculty of Medicine, Via Tronto 10/A, I-60131 Ancona,
Italy. Tel: ++39-071-2206086; Fax: ++39-071-2206087; E-mail: [email protected]
486
M.Castellucci et al.
Villous sprouting: the key event for placental
development
Villous development of the human placenta starts between days
12 and 18 post conception (p.c.). The massive trophoblastic
trabeculae of the placental anlage start to proliferate, forming
®nger-like trophoblastic protrusions (primary villi) into the
maternal blood surrounding the trophoblastic trabeculae (Boyd
and Hamilton, 1970). Only 2 days later embryonic connective
tissue derived from the extra-embryonic mesenchyme (Enders
and King, 1988) invades these villi, transforming them into socalled secondary villi. Beginning between days 18 and 20 p.c., the
®rst fetal capillaries can be observed in this mesenchyme (King,
1987; Demir et al., 1989). By de®nition, the appearance of
capillaries in the villous stroma marks the development of socalled tertiary villi (Boyd and Hamilton, 1970), the ®rst
generation of which are the mesenchymal villi (Castellucci and
Kaufmann, 1982; Castellucci et al., 1990; Benirschke and
Kaufmann, 2000). The latter are the ®rst structures providing
the morphological prerequisites for an effective materno-fetal
exchange of nutrients, gases and waste substances. Analysis of
these ®rst generations of mesenchymal villi between pregnancy
days 20 and 42 by electron microscopy (Demir et al., 1989) and
CD34 immunohistochemistry (Kaufmann and Kingdom, 2000)
revealed vasculogenesis (de-novo formation of capillaries out of
mesenchymal precursor cells, see below) as the underlying
mechanism of vessel formation.
The same developmental steps that lead to the formation of
mesenchymal villi, namely trophoblastic sprouting, mesenchymal
invasion and local fetal angiogenesis, are repeated throughout
pregnancy as long as the villous trees expand by villous sprouting.
Consequently, there is no formation of specialized villous types
other than from the pool of mesenchymal villi. Because of this,
the formation of mesenchymal villi plays a key role for the
development of the villous trees.
The importance of mesenchymal villi for the understanding of
villous development and maldevelopment is further underlined by
the fact that they directly or indirectly can differentiate into a
variety of specialized villous types, e.g. immature intermediate
villi, stem villi, mature intermediate villi, and terminal villi
(Figure 1) (Kaufmann et al., 1979; Castellucci and Kaufmann,
1982; Castellucci et al., 1990; Kaufmann and Castellucci, 1995;
Benirschke and Kaufmann, 2000).
Up to about the 5th week p.c., all placental villi belong to the
mesenchymal type (Castellucci et al., 1990). Thereafter, an
increasing number of these villi are transformed into immature
intermediate villi (Figure 1) which are characterized by an
impressive increase in villous diameter and by the presence of
numerous longitudinally oriented stromal channels (Castellucci
and Kaufmann, 1982). These immature intermediate villi are
®nally transformed into stem villi (Figure 1a) by means of a
stromal ®brosis starting around the centrally positioned fetal
vessels which at the same time are transformed into arteries and
veins (Figure 1b). The continuous loss of mesenchymal and
immature intermediate villi to the bene®t of stem villi is
compensated by continuous sprouting of mesenchymal villi along
the surfaces of all mesenchymal and immature intermediate villi
(Figures 1, 2, 3). In this way a rapid increase of total villous mass
and thus of materno-fetal exchange surface is provided.
Starting with the 23rd week p.c., a key event in placental
development takes place. The villous growth switches: mesenchymal villi start transforming into mature intermediate villi rather
than into immature ones (Figure 1) (Castellucci et al., 1990). The
mature intermediate villi differ from their immature counterparts
not only by the absence of stromal channels, lower cytotrophoblastic density and better fetal capillarization, but also by the fact
that they do not mature to stem villi. Rather, along their surfaces
they produce large numbers of highly capillarized terminal villi
(Figure 1b) which are the most effective structures for maternofetal diffusional exchange (Castellucci et al., 1990; Benirschke
and Kaufmann, 2000). Few remaining mesenchymal and
immature intermediate villi are located in the centres of the
villous trees, where they form a kind of poorly differentiated
growth reserve. The differentiation of mesenchymal villi into
mature intermediate villi takes place from about week 23 p.c.
until term (Figure 1).
Delayed switching towards mature intermediate villi may well
lead to the predominance of mesenchymal and immature
intermediate villi, to unlimited growth of villous trees and to an
extremely large but undifferentiated placenta, which is functionally insuf®cient. This situation is typically associated with clinical
features such as Rhesus incompatibility or persisting villous
immaturity in post-mature placentas, the latter situation normally
associated with intrauterine growth retardation (Benirschke and
Kaufmann, 2000).
By contrast, too early switching towards the mature intermediate villi leads to untimely stop of villous growth but
premature differentiation. The resulting placenta is unusually
small and despite its highly differentiated villi, may not provide
enough exchange surface. Typical examples comprise maturitas
praecox placentae (Becker, 1981), hypermaturity (Salvatore,
1968), and pregnancies complicated with intrauterine growth
retardation with Doppler high resistance index in the umbilical
arteries (Macara et al., 1995, 1996; Todros et al., 1999). These
cases of villous maldevelopment are common. They often result
in pregnancy loss or in poor neonatal outcome (Kingdom and
Kaufmann, 1997).
The cause for these cases of villous maldevelopment must be
sought in abnormal development of mesenchymal villi and/or
abnormal control of the developmental switch towards the one or
the other type of intermediate villi. The control of this switch is
still a mystery. By contrast, information has been obtained on the
developmental events of villous sprouting up to this switch.
Structural events during early villous sprouting
Beginning around the 7th week p.c., the formation of new
mesenchymal villi is no longer achieved from the original
blastocystic trabeculae since these are largely transformed into
villi. Rather, the surfaces of already existing mesenchymal villi
and their immature intermediate successors represent the sources
for newly sprouting villi (Figure 1). The early structural correlates
of this process are circumscribed areas, the so-called `hot spots'
(Figures 1, 4a) (Kosanke, 1994). Structurally, these are characterized by subtrophoblastic spots of highly cellular stroma, void
of stromal channels, sometimes separated from the remaining
stroma by an incomplete belt of macrophages. The trophoblast
covering these stromal spots shows increased numbers of
Villous sprouting in placental development
487
Figure 1. Schematic representation of the formation and differentiation of placental villi during early and late pregnancy. (Modi®ed from Castellucci et al., 1990,
with permission.) (a) Developmental pathway of placental villi during the 1st and 2nd trimester (upper half) as compared to the 3rd trimester (lower half). White
arrows: developmental steps of villous differentiation; black arrows: formation of new villi at the surface of pre-existing ones. During the ®rst and second trimester
especially at the surface of mesenchymal and immature intermediate villi, hot spots and subsequently trophoblastic sprouts are formed that ®rst differentiate into
villous sprouts and ®nally into new mesenchymal villi. The latter transform into immature intermediate villi and ®nally into stem villi. From the beginning of the
third trimester, mesenchymal villi preferentially differentiate into mature intermediate villi. The surface of the latter is passively protruded by elongating and
looping fetal capillaries resulting in protrusion of terminal villi. The latter are highly specialized places of materno-fetal exchange. Transformation of the
mesenchymal villi into immature intermediate villi is largely blocked in the third trimester of gestation. The remaining immature intermediate villi differentiate
into stem villi. Because of this, the base for the formation of new sprouts is reduced and the growth capacity of the villous trees slows down. (b) Histological
characteristics of the various villous types and their typical topographical relationships. The left part of the diagram represents the ®rst and second trimester, the
right part represents the third trimester. Note the immature intermediate villus (left) showing a `hot spot' which subsequently (top of the left villus) develops via
trophoblastic and villous sprouts into a new mesenchymal villus. Hot spots correspondingly mark the sites of future villous branching.
trophoblast cells. Application of proliferation markers such as Ki67 and proliferating cell nuclear antigen (PCNA) clones or
[3H]TdR incorporation reveals that these areas are characterized
by increased trophoblastic proliferation rates (Figure 4a). Some of
these spots, probably later stages of development, additionally
show local stromal proliferation which exceeds that of the
surrounding villous stroma by more than 200%. `Hot spots'
throughout gestation demonstrate the highest proliferation rates
found in the villous trees (Table I) (Kosanke, 1994; Kosanke et
al., 1995).
From serial sections of different developmental stages of
`hot spots' it becomes evident that they represent the ®rst
stages of sprouting of mesenchymal villi. It is obvious that
sprouting involves active growth processes of both tissue
components, since both trophoblast and stroma show increased
proliferative activity. The post-proliferative trophoblast cells
fuse with the syncytium leading to the formation of massive
syncytiotrophoblastic outgrowths, the so-called trophoblastic
sprouts (Boyd and Hamilton, 1970). In a next step, further
proliferating trophoblast cells, followed by the highly proliferative mesenchyme, invade these trophoblastic sprouts,
transforming them into villous sprouts. The ®rst step of this
process, i.e. the formation of unvascularized villous sprouts,
has been recently mimicked in vitro. We have cloned the
BeWo human choriocarcinoma cell line obtaining the clone
MC1 which is not invasive and shares phenotypic similarities
with human villous cytotrophoblast (Crescimanno et al., 1996).
Spheroids of these immortal trophoblast cells cultured in threedimensional collagen-I gels did not show any trophoblastic
outgrowth (M.Castellucci et al., unpublished observations). Coculture on agarose of MC1 cells with fetal ®broblasts resulted
in formation of spheroids containing a ®broblastic core.
Embedded in collagen type I, these spheroids produced
villous-like outgrowths with a ®broblastic core. Comparable
trophoblastic outgrowths were produced when MC1 spheroids
were cultured in the presence of ®broblast-conditioned
medium. These data suggest that secretory products of fetal
®broblasts induce trophoblastic and villous sprouting.
488
M.Castellucci et al.
Figure 2. Scanning electron micrograph of the outer surface of trophoblastic
and villous sprouts (drumstick-shaped appendices), connected to the surface of
an immature intermediate villus (below). Bar = 10 mm.
Figure 3. Semi-thin section of part of a ®rst trimester villus demonstrating the
transition of a stem villus (lower right part of the micrograph) to an offbranching immature intermediate villus (central part of the micrograph), which
continues into a mesenchymal villus (left part of the micrograph). The
developmental connections and the stromal characteristics of all three villous
types are evident in this picture. Bar = 40 mm. (From Castellucci et al., 1990,
with permission.)
By de®nition, the vascularization of a villous sprout leads to the
formation of a mesenchymal villus, the most basic villous type
(Figure 4b) (Castellucci et al., 1990; Benirschke and Kaufmann,
2000). Generally, vascularization can be achieved by three
different mechanisms (Risau and Lemmon, 1988): (i) vasculogenesis (de-novo formation of endothelium from mesenchymal
precursors), (ii) branching or sprouting angiogenesis (sprouting
of blind-ending capillaries, originating from pre-existing vascular
tubes in the neighbouring tissues), (iii) non-sprouting angiogen-
Figure 4. Immunostaining of placental tissue with the proliferation marker
MIB-1. The MIB-1 monoclonal antibody (Immunotech Co., Marseille, France)
reacts with normal and recombinant Ki-67 nuclear cell proliferation-associated
antigen. (a) Section of an immature intermediate villus. The densely packed
MIB-1-positive cytotrophoblastic nuclei located between the syncytiotrophoblast and the stromal channels (SC) mark a `hot spot`, the ®rst step of a newly
developing mesenchymal villus. Bar = 10 mm. (b) Histological section of a
mesenchymal villus. MIB-1-positive cells are mainly located at the base of the
mesenchymal villus and at its tip, the latter ending with a drumstick-like,
massive trophoblastic sprout (*). Note the contrast between the compact
stromal architecture of the mesenchymal villus and the loosely arranged
stroma of the immature intermediate villus. Bar = 20 mm. Sections were
immunostained by avidin-biotin complex method.
Villous sprouting in placental development
489
Table I. Numerical density of proliferating nuclei of different villous types and villous segments
Cytotrophoblast
(no. of MIB-1-positive
nuclei per 100 mm
basal lamina)
Stroma
(no. of MIB-1-positive
nuclei per 2000 mm2
stromal area)
Stem villi (1st trimester)
1.21 (6 0.28)
0.30 (6 0.06)
Stem villi (3rd trimester)
0.78 (6 0.90)
0.29 (6 0.12)
Immature intermediate villi
(1st trimester)
1.85 (6 0.71)
0.36 (6 0.15)
`Hot spots' of immature intermediate
villi (1st trimester)
2.78 (6 1.03)
1.17 (6 0.37)
Mesenchymal villi (1st trimester)
2.52 (6 1.08)
1.44 (6 0.78)
Mature intermediate villi
(3rd trimester)
0.69 (6 0.31)
0.64 (6 0.35)
Terminal villi (3rd trimester)
0.80 (6 0.44)
1.65 (6 1.47)
Data were gained by evaluation of immunohistochemical reactions on paraffin sections incubated
with the proliferation marker MIB-1 (mean 6 SD). Trophoblastic labelling is related to 100 mm
trophoblastic basal membrane, stromal labelling is related to 2000 mm2 stromal area. As for the `hot
spots', these data were found to be the mean dimensions of the `hot spots' observed in paraffin
sections of immature intermediate villi. Note the high density of proliferative activity in both
cytotrophoblastic and stromal nuclei of mesenchymal villi and `hot spots'.
Figure 5. Scanning electron micrograph of a vessel cast (plastic injection) of a
mesenchymal villus at 18 weeks post conception. Note the net-like
arrangement of capillaries resulting from branching angiogenesis and the
blind-ending capillaries near the villous tip representing newly developing
vessel sprouts. Bar = 14 mm.
esis (elongation of endothelial tubes by intussusceptive proliferation with subsequent protrusion of a vessel loop into the villous
sprout). Ultrastructural analysis (Demir et al., 1989) and
Figure 6. Term placenta. Cross-section of a villous sprout which is
characterized by large epithelioid connective tissue cells (CO) and capillary
sprouts. The latter consist of densely packed endothelial cells (E), connected
with each other by tight junctions, showing no or only minimal lumens (inset)
and surrounded by basal laminas. The latter sometimes form thick
convolutions (arrows). Bar = 3 mm; inset bar = 1 mm. (From Demir et al.,
1989, with permission.)
immunohistochemistry using CD34 antibodies and proliferation
markers (Kaufmann and Kingdom, 2000) showed absence of
isolated haemangioblastic cells but presence of single blindending tubes with endothelial mitoses. These data favour
sprouting angiogenesis as the prevailing mechanism for vascular-
490
M.Castellucci et al.
ization in all mesenchymal villi from the 7th week p.c. onwards.
The expression of vascular endothelial growth factor (VEGF; a
potent stimulator of sprouting angiogenesis) in trophoblastic
sprouts (Shiraishi et al., 1996) underlines this notion. Basic
®broblast growth factor (bFGF) also expressed in the neighbouring stroma (MuÈhlhauser et al., 1996) may support this process by
stimulating endothelial proliferation (Shreeniwas et al., 1991)
and/or by recruiting additional perivascular cells for expansion of
the vascular bed.
The mesenchymal villi are short, stubby structures extending
peripherally in a drum-stick-like massive syncytiotrophoblastic
Figure 7.
Figure 9.
Figure 7. Histological section of a ®rst trimester placenta immunostained for
tenascin using monoclonal BC-4 antibody (Castellucci et al., 1991) which
recognizes an epitope within the EGF-like sequence of the tenascin molecule.
BC-4 recognizes all current known tenascin isoforms. Immunostaining by
streptavidin-biotin. In the centre of the micrograph the stroma of a
mesenchymal villus is intensely labelled. The stroma of the immature
intermediate villi (iiv) is immunonegative for this extracellular matrix
molecule. Bar = 60 mm.
Figure 8. Staining for hyaluronic acid in ®rst trimester chorionic villi by a
binding probe consisting of a biotinylated form of the hyaluronate-binding
complex (b-PG). The reaction product is accumulated within the stroma of the
mesenchymal villi (arrows) and in the wall of the fetal vessels of the immature
intermediate villus (IIV). In the latter, HA is also present beneath the
trophoblastic covering (arrowheads) where capillaries of the paravascular
network are present. Note the absence of staining in most of the stroma of the
villous core of the immature intermediate villus. Bar = 50 mm.
Figure 8.
Figure 9. Mesenchymal villus of the ®rst trimester placenta branching off
from an immature intermediate villus (lower half) immunostained for bFGF
using a rabbit polyclonal antibody against natural bovine brain bFGF (British
Bio-technology, Oxford, UK) and avidin-biotin-peroxidase complex. Strong
immunoreactivity is present in the trophoblastic covering. The villous stroma
shows a positive reaction only in the distal half of the mesenchymal villus,
where, according to MuÈhlhauser et al. (1996), heparan sulphate proteoglycan
is co-localized. Bar = 20 mm. (From MuÈhlhauser et al., 1996, with permission.)
Villous sprouting in placental development
outgrowth (Figures 2, 3, 4b). The surrounding mantle of
trophoblast contains a largely complete layer of cytotrophoblast
beneath the syncytiotrophoblastic surface layer (Castellucci et al.,
1990). Cytotrophoblastic proliferation is mainly found near the
villous tip (Figure 4b) and at the base of the villus where the
resulting daughter cells locally may form multilayered clusters
(Kosanke, 1994; MuÈhlhauser et al., 1996). The stroma contains
longitudinally oriented, moderately branched capillaries peripherally ending blindly in compact aggregates of endothelial cells
(Figures 5, 6). Endothelial mitoses are evenly distributed all over
these capillaries (Kosanke, 1994).
Extracellular matrix
It has been well established that extracellular matrix (ECM) plays
a pivotal role in organogenesis, differentiation and tissue
remodelling (Hay, 1991). Indeed, at the cell surface, matrix
receptors link the ECM to the cell interior; the metabolism and
fate of the cell, its shape, and many of its properties and functions
are continuously related to and dependent on the composition and
organization of the matrix (Hay, 1991). Consequently particular
attention has been devoted to the study of ECM molecules
participating in villous differentiation and morphogenesis. This is
also important because the various villous types differ mainly
regarding structure of their stromal cores (see above), re¯ecting
differences in distribution of ECM molecules.
Recent data have shown that speci®c ECM proteins are
preferentially expressed in the stroma of the mesenchymal villi.
One of these ECM proteins is tenascin (Figure 7) (Castellucci et
al., 1991) which has been shown to be associated with cell
proliferation and migration, e.g. in wound healing and tumour
progression (Hauptmann et al., 1995; Ikeda et al., 1995;
Deryugina and Bourdon, 1996; Latijnhouwers et al., 1996;
Riedl et al., 1998). The limited cell spreading on tenascin and
its inhibition of ®bronectin mediated cell adhesion (ChiquetEhrisman et al., 1988) has suggested that tenascin, like SPARC
(secreted protein acidic and rich in cysteine) and thrombospondin,
is an anti-adhesion matrix protein (Sage and Bornstein, 1991). It
possibly reduces substrate attachment during the physiological
and pathological processes mentioned above, playing a pivotal
role in cell migration and tissue remodelling (Castellucci et al.,
1991; Murphy-Ullrich et al., 1991; Deryugina and Bourdon,
1996). Thus, the presence of tenascin nearly exclusively in the
stroma of mesenchymal villi (Figure 76) suggests a special role of
this protein in growth, stromal remodelling and angiogenesis of
this forerunner of all other villous types. It has been thought to
facilitate the migration of cytotrophoblast, ®broblasts and
endothelium, all of which are essential for mesenchymal villous
growth (Castellucci et al., 1991).
Hyaluronic acid (HA), a high molecular mass polysaccharide,
is a further important ECM molecule which plays a central role in
numerous morphogenetic processes such as cell motility,
proliferation and cell matrix adhesion (Laurent and Fraser,
1992). Its biological activity is mediated by HA receptors that
are present on the cell surface (Green et al., 1988; Underhill,
1992). These receptors belong to (i) the glycoprotein family
CD44, which represents a group of transmembrane proteins
(Underhill, 1992) and also probably to (ii) the RHAMM (receptor
for hyaluronan-mediated motility) family of receptors (Turley,
491
1992). In addition to its HA-binding capacity, CD44 is involved
in the degradation of HA by receptor-mediated internalization
and, subsequently, by the activity of acid hydrolases (Culty et al.,
1992; Underhill, 1992). By contrast, the second receptor group,
RHAMM may be involved in HA-mediated cell locomotion
(Turley, 1992).
HA has been found to be expressed in high quantities in the
entire stroma of mesenchymal villi throughout the ®rst half of
gestation (Figure 8). In contrast the other villous types as the
immature intermediate villi were stained for HA only around the
vessel walls and focally beneath the trophoblastic cover (Figure 8)
(M.Castellucci et al., unpublished observations). This subtrophoblastic expression of HA in immature intermediate villi is related
to the `hot spots' representing areas of high trophoblastic and
stromal proliferation rates (Kosanke, 1994; Kosanke et al., 1995).
These areas are involved in the protrusion of HA-rich mesenchyme into newly formed trophoblastic sprouts, thus facilitating the
formation of mesenchymal villi arising from the immature
intermediate ones (M.Castellucci et al., unpublished observations). Interestingly, in the ®rst trimester CD44 was only
expressed in few fetal macrophages (Hofbauer cells) in restricted
parts of mesenchymal villi. Macrophages of other villous types
were mainly negative. Intraplacental expression of the second HA
receptor family, RHAMM, has not yet been studied. Taken
together, these preliminary data are in agreement with previous
reports indicating that the expression of HA is inversely
correlated with the presence of CD44 in some developing organs
(Underhill et al., 1993). Thus, one may postulate that, in
agreement with the views presented by Toole (1991) in other
embryonic tissues, the large amounts of HA found within the
mesenchymal villi and neighbouring parts of immature intermediate villi are required as a medium through which mesenchymal cells and blood vessels migrate.
Angiogenesis, growth factors and growth factor
receptors
A crucial step in early human gestation is the establishment of an
ef®cient nutrient-waste exchange between fetal and maternal
blood circulation. Therefore, formation of new vessels is a
fundamental feature of the development of the placenta and in
particular in the morphogenesis of the mesenchymal villi.
Isolation of endothelial cells from placental vessels has been
recently achieved (Leach et al., 1994). Unfortunately no in-vitro
data are available at present on the morphogenetic interactions of
such endothelial cells with other cellular components of the
chorionic villus. However, some ultrastructural, immunohistochemical and in-situ hybridization studies on placental tissues
have been published on angiogenic processes in the chorionic
villi. Demir and co-workers (1989) identi®ed haemangioblastic
cells differentiating in placental villi at day 21 p.c., originating
from mesenchymal precursors. Immunohistochemically these
cells have been identi®ed using antibodies against CD34
(Kaufmann and Kingdom, 2000). These cells are normally closely
apposed to each other, forming roundish clusters or cord-like
strings which are considered to be the forerunners of the capillary
endothelium. Such groups of endothelial precursors have been
found to be present near the tips of the mesenchymal villi (Figure
6) (Demir et al., 1989).
492
M.Castellucci et al.
A recent report by MuÈhlhauser et al. (1996) has emphasized
these data. These authors demonstrate that basic ®broblast growth
factor (bFGF), an angiogenic factor involved in the recruitment of
haemangiogenic precursor cells, is co-distributed with its low
af®nity receptor heparan sulphate (HSPG) in the distal part of the
mesenchymal villous stroma (Figure 9) whereas in other villous
types this co-distribution is found only in the vessel walls. These
data are of particular interest because this bFGF-HSPG codistribution corresponds to the region where blindly ending
capillary sprouts were found (Figures 5, 6). Interestingly, binding
of bFGF to HSPG is necessary for the binding of this growth
factor to the high af®nity ®broblastic growth factor receptor
(FGF-R) and for its mitogenic activity (Yayon et al., 1991;
Turnbull et al., 1992; David, 1993). Moreover, bFGF bound to
HSPG is protected from proteolytic degradation and therefore acts
as a kind of growth factor reservoir in the ECM (Saksela et al.,
1988; Roghani and Moscatelli, 1992).
Shiraishi et al. (1996) showed that vascular endothelial growth
factor (VEGF) is most intensely expressed in the syncytiotrophoblast of villous sprouts, i.e. in the ®rst phase of villous sprouting.
In agreement with the reduction of villous sprouting with
advancing pregnancy, VEGF expression decreases until term (Li
et al., 1995). Unlike bFGF, VEGF possesses a signal peptide for
secretion (Senger et al., 1983; Connolly et al., 1989; Ferrara et al.,
1992). Like bFGF, it is thought to be inducible by hypoxia
(Nomura et al., 1995; Cao et al., 1996). This is of particular
interest because in the human placenta low intraplacental oxygen
partial pressure has been found in those stages of pregnancy
(Rodesch et al., 1992) and in those areas (Schuhmann et al., 1988;
Benirschke and Kaufmann, 2000) where villous sprouting and
mesenchymal villi prevail. The stimulating effect of hypoxia on
branching angiogenesis in the placenta has been proven
experimentally (Scheffen et al., 1990) as well as by analysis of
pathological specimens (Kadyrov et al., 1998).
Interestingly, Vuckovic et al. (1996) identi®ed KDR, one of the
two VEGF receptors, not only in endothelial cells but also in
endothelial cell pecursors. This may indicate that capillarization
of villous sprouts is not only based on endothelial proliferation
(angiogenesis) but additionally supported by recruitment of
surrounding mesenchymal cells (vasculogenesis).
In addition to KDR, ¯t-1, the second VEGF receptor, has been
identi®ed in human placental capillaries; according to
Crescimanno et al. (1995) it is expressed throughout pregnancy.
Studies conducted by Wilting et al. (1995a,b), on the chorioallantoic membrane of the chicken have shown that VEGF
stimulation of endothelial cells which express ¯t-1 and ¯k-1
(the non-human analogue of KDR) results in branching
angiogenesis and formation of capillary networks, similar to
those observed by us in immature intermediate villi and their
mesenchymal precursors. In contrast the closely related placentaderived growth factor (PlGF), another ligand of ¯t-1 receptor, is
down-regulated under hypoxic conditions (Shore et al., 1997) and
it does not stimulate branching angiogenesis (Wilting et al.,
1995b). A switch from branching to non-branching angiogenesis
in the human placental villi later in gestation (Kaufmann and
Kingdom, 2000) goes in line with decreasing expression of VEGF
(Li et al., 1995) and KDR (Vuckovic et al., 1996), but increasing
PlGF expression (Crescimanno et al., 1995; Khaliq et al., 1996).
Moreover, platelet-derived growth factor B (PDGF-B) was
suggested to be involved in placental angiogenesis. Holmgren and
co-workers (1991) reported that most capillary endothelial cells of
newly formed villi co-express this mitogenic growth factor and
PDGF-b receptor. This observation suggests that PDGF-B is
involved in placental angiogenesis by forming autostimulatory
loops in capillary endothelial cells promoting cell proliferation.
Last but not least, insulin seems to be closely related to villous
development. Throughout the ®rst trimester, insulin receptor is
expressed along the apical trophoblastic surface which is exposed
to the maternal blood, predominantly in sprouts and mesenchymal
villi (Desoye et al., 1994). These are the structures that are mainly
responsible for villous growth in this stage of pregnancy. In the
second and third trimester of pregnancy when the fetal pancreas
starts insulin secretion and when expansion of the villous trees is
mostly based on longitudinal growth and coiling of fetal
capillaries, insulin receptor expression switches to the luminal
surfaces of fetal capillaries (Desoye et al., 1994) suggesting that
from this period onwards also fetal insulin is involved in the
control of placental angiogenesis.
Vandenbunder et al. (1989) have shown that the c-ets1 protooncogene, which encodes a transcription factor, is highly
expressed within endothelial cells during blood vessel formation.
Based on these data Wernert et al. (1992) studied the expression
of c-ets1 during angiogenesis under different conditions in human
tissues. Adult tissues expressed c-ets1 only where angiogenesis
was resumed, e.g. in granulation tissue. These data emphasize the
pivotal role of c-ets1 in early angiogenesis. Of particular interest
is the fact that c-ets1 proteins are involved in the regulation of the
transcription of matrix-degrading protease genes (Wernert et al.,
1992). The expression of the latter is essential to ensure the
formation of new blood vessels (Mignatti et al., 1989; Montesano
et al., 1990; Montesano, 1992; Wernert et al., 1992). Vessels of
early human placental villi also express c-ets1 (Wernert et al.,
1992; Luton et al., 1997). This emphasizes the role of matrix
degrading proteases in villous angiogenesis and sprouting.
Heterogeneity of mesenchymal villi
By de®nition, every small diameter villus (calibre <100 mm)
branching off the large calibre immature intermediate villi is a
mesenchymal villus (Castellucci et al., 1990). Besides the usual
phenotype described above, two other phenotypes can be found.
Locally, clusters of obviously normal mesenchymal villi
showing all structural characteristics as described previously,
can be found. They do not show signs of proliferation or
expression of tenascin nor of the other molecules considered
above (Castellucci et al., 1991; Kosanke et al., 1993; Kosanke,
1994; MuÈhlhauser et al., 1996). One possible explanation for this
phenomenon is that these are resting stages of sprouting, thus
suggesting that villous differentiation and development do not
occur contemporarily in every mesenchymal villus and in every
villous tree. Topological analysis of the branching patterns of the
villous tree has demonstrated asymmetry of villous branching
patterns. This asymmetry was interpreted as a result of
asynchronous growth of mesenchymal villi (Kosanke et al.,
1993).
Moreover, groups of mesenchymal villi can be found which are
covered by clearly degenerative trophoblast, void of cytotropho-
Villous sprouting in placental development
blast or partly even embedded in ®brinoid. These show a highly
condensed ®brous stroma expressing high concentrations of
tenascin (Castellucci et al., 1991), but no evidence of stromal
or trophoblastic proliferation (Kosanke, 1994). This involution of
mesenchymal villi is probably due to a local overproduction of
sprouts that might negatively in¯uence the surrounding maternal
blood ¯ow causing turbulence or stasis in the local maternal blood
circulation. In this sense sprouting with subsequent expansion of
the villous trees and shaping of the intervillous space seems to
take place as a self-regulating process simply following the trial
and error principle (Benirschke and Kaufmann, 2000).
Conclusions
The data discussed above give clear evidence for the central role
of sprouts and mesenchymal villi for growth and differentiation of
the villous tree. We have presented further evidence that the
future differentiation of mesenchymal villi into immature
intermediate villi or into mature intermediate villi decides upon
the balance between growth and maturation of the placenta. We
are just beginning to understand some of the molecular
mechanisms controlling this switch. Bearing in mind that roughly
10% of all pregnancies suffer from different types of villous
maldevelopment resulting in poor physical and mental neonatal
outcome with its tremendous socio-economic impact (Barker,
1995; Kingdom 1998; Benirschke and Kaufmann, 2000), a better
understanding of the pathogenetic mechanisms of villous sprouting and maldevelopment is urgently required.
Acknowledgements
This review is based on projects which
the Italian Ministry of University and
the University of Ancona, the Deutsche
the Vigoni Program and the Biomed
Community.
were supported by grants from
Scienti®c Research (MURST),
Forschungsgemeinschaft (DFG),
I program of the European
References
Barker, D.J. (1992) The fetal origins of diseases of old age. Eur. J. Clin. Nutr.,
46 (Suppl. 3), S3-9.
Barker, D.J. (1995) Intrauterine programming of adult disease. Mol. Med.
Today 1, 418±423.
Barker, D.J., Gluckman, P.D., Godfrey, K.M. et al. (1993) Fetal nutrition and
cardiovascular disease in adult life. Lancet, 341, 938±941.
Becker, V. (1981) Allgemeine und spezielle Pathologie der Plazenta. In
Becker, V., Schiebler, T.H. and Kubli, F. (eds), Die Plazenta des
Menschen. Thieme, Stuttgart, pp. 251±393.
Benirschke, K. and Kaufmann, P. (2000) The Pathology of the Human
Placenta, 4th edn. Springer, New York.
Boyd, J.D. and Hamilton, W.J. (1970) The Human Placenta. Heffer,
Cambridge.
Cao, Y., Linden P., Shima, D. et al. (1996) In vivo angiogenic activity and
hypoxia induction of heterodimers of placenta growth factor/vascular
endothelial growth factor. J. Clin. Invest., 98, 2507±2511.
Castellucci, M. and Kaufmann, P. (1982) A three-dimensional study of the
normal human placental villous core. II. Stromal architecture. Placenta, 3,
269±286.
Castellucci, M., Scheper, M., Scheffen, I. et al. (1990) The development of the
human placental villous tree. Anat. Embryol., 181, 117±128.
Castellucci, M., Classen-Linke, I., MuÈhlhauser, J. et al. (1991) The human
placenta: a model for tenascin expression. Histochemistry, 95, 449±458.
Chiquet-Ehrismann, R., Kalla, P., Pearson, C.A. et al. (1988) Tenascin
interferes with ®bronectin action. Cell, 53, 383±390.
Connolly, D.T., Heuvelman, D.M., Nelson, R. et al. (1989) Tumor vascular
493
permeability factor stimulates endothelial cell growth and angiogenesis. J.
Clin. Invest., 84, 1470±1478.
Crescimanno, C., Marzioni, D., Persico, M.G. et al. (1995) Expression of
bFGF, PlGF and their receptors in the human placenta. Placenta, 16, A13.
Crescimanno, C., Foidart, J.M., Noel, A. et al. (1996) Cloning of
choriocarcinoma cells shows that invasion correlates with expression
and activation of gelatinase A. Exp. Cell Res., 227, 240±251.
Culty, M., Nguyen, H.A. and Underhill, C.B. (1992) The hyaluronan receptor
(CD44) participates in the uptake and degradation of hyaluronan. J. Cell
Biol., 116, 1055±1062.
David, G. (1993) Integral membrane heparan sulfate proteoglycans. FASEB J.,
7, 1023±1030.
Demir, R., Kaufmann, P., Castellucci, M. et al. (1989) Fetal vasculogenesis
and angiogenesis in human placental villi. Acta Anat., 136, 190±203.
Deryugina, E.I. and Bourdon, M.A. (1996) Tenascin mediates human glioma
cell migration and modulates cell migration on ®bronectin. J. Cell Sci.,
109, 643±652.
Desoye, G., Hartmann, M., Blaschitz, A. et al. (1994) Insulin receptors in
syncytiotrophoblast and fetal endothelium of human placenta.
Immunohistochemical evidence for developmental changes in
distribution pattern. Histochemistry, 101, 277±285.
Enders, A.C. and King, B.F. (1988) Formation and differentiation of
extraembryonic mesoderm in the rhesus monkey. Am. J. Anat., 181,
327±340.
Ferrara, N., Houck, K.A., Jakeman, L.B. et al. (1992) Molecular and
biological properties of the vascular endothelial growth factor family of
proteins. Endocrine Rev., 13, 18±32.
Fox, H. (1997) Pathology of the Placenta, 2nd edn. Saunders, London.
Green, S.J., Tarone, G. and Underhill, C.B. (1988) Distribution of hyaluronate
receptors in the adult lung. J. Cell Sci., 89, 145±156.
Hay, E.D. (1991) Cell Biology of Extracellular Matrix, 2nd edn. Plenum Press,
New York.
Hauptmann, S., Zardi, L., Siri, A. et al. (1995) Extracellular matrix proteins in
colorectal carcinomas. Expression of tenascin and ®bronectin isoforms.
Lab. Invest., 73, 172±182.
Holmgren, L., Glaser, A., Pfeifer-Ohlsson, S. et al. (1991) Angiogenesis
during human extraembryonic development involves the spatiotemporal
control of PDGF ligand and receptor gene expression. Development, 113,
749±754.
Ikeda, Y., Mori, M., Kajiyama, K. et al. (1995) Immunohistochemical
expression of tenascin in normal stomach tissue, gastric carcinomas and
gastric carcinoma in lymph nodes. Br. J. Cancer, 72, 189±192.
Kadyrov, M., Kosanke, G., Kingdom, J.C. et al. (1998) Increased fetoplacental
angiogenesis during ®rst trimester in anaemic women. Lancet, 352, 1747±
1749.
Kaufmann, P., Sen, D.K. and Schweikhart, G. (1979) Classi®cation of human
placental villi. I. Histology. Cell Tissue Res., 200, 409±423.
Kaufmann, P. and Castellucci, M. (1995) Development and anatomy of the
placenta. In Fox, H. and Wells, M. (eds), Haines and Taylor's Obstetrical
and Gynaecological Pathology. Churchill Livingstone, Edinburgh, pp.
1437±1476.
Kaufmann, P. and Kingdom, J.C. (2000) Development of the vascular system
in the placenta. In Risau, W. and Rubanyi, G. (eds), Morphogenesis of
Endothelium. Harwood, Amsterdam, pp. 255±275.
Khaliq, A., Li, X.F., Shams, M. et al. (1996) Localisation of placenta growth
factor (PlGF) in human term placenta. Growth Factors, 13, 243±250.
King, B.F. (1987) Ultrastructural differentiation of stromal and vascular
components in early macaque placental villi. Am. J. Anat., 178, 30±44.
Kingdom, J.C. (1998) Placental pathology in obstetrics: adaptation or failure
of the villous tree? Placenta, 19, 347±351.
Kingdom, J.C. and Kaufmann, P. (1997) Oxygen and placental villous
development: origins of fetal hypoxia. Placenta, 18, 613±626.
Kosanke, G. (1994) Proliferation, Wachstum und Differenzierung der
ZottenbaÈume der menschlichen Placenta. Shaker Publishers, Aachen.
Kosanke, G., Castellucci, M., Kaufmann, P. et al. (1993) Branching patterns of
human placental villous trees: Perspectives of topological analysis.
Placenta, 14, 591±604.
Kosanke, G., Korr, H., Kohnen, G. et al. (1995) Growth zones of the human
placental villous trees. Placenta, 16, A 38.
Latijnhouwers, M.A., Bergers, M., Van Bergen, B.H. et al. (1996) Tenascin
expression during wound healing in human skin. J. Pathol., 178, 30±35.
Laurent, T.C. and Fraser, J.R. (1992) Hyaluronan. FASEB J., 6, 2397±2404.
Leach, L., Bhasin, Y., Clark, P. et al. (1994) Isolation of endothelial cells from
human term placental villi using immunomagnetic beads. Placenta, 15,
355±364.
494
M.Castellucci et al.
Li, X-F., Whittle, M.J. and Ahmed, A. (1995) Localization of vascular
endothelial growth factor and its receptor in pregnancy. Placenta, 16,
A42.
Luton, D., Sibony, O., Oury, J.F. et al. (1997) The c-ets1 protooncogene is
expressed in human trophoblast during the ®rst trimester of pregnancy.
Early Hum. Dev., 47, 147±156.
Macara, L., Kingdom, J.C., Kohnen, G. et al. (1995) Elaboration of stem
villous vessels in growth restricted pregnancies with abnormal umbilical
artery Doppler waveforms. Br. J. Obstet. Gynaecol., 102, 807±812.
Macara, L., Kingdom, J.C., Kaufmann, P. et al. (1996) Structural analysis of
placental terminal villi from growth-restricted pregnancies with abnormal
umbilical artery Doppler waveforms. Placenta, 17, 37±48.
Mignatti, P., Tsuboi, R., Robbins, E. et al. (1989) In vitro angiogenesis on the
human amniotic membrane: requirement for basic ®broblast growth
factor-induced proteinases. J. Cell. Biol., 108, 671±682.
Montesano, R. (1992) Regulation of angiogenesis in vitro. Eur. J. Clin. Invest.,
22, 504±515.
Montesano, R., Pepper, M.S., Mohle-Steinlein, U. et al. (1990) Increased
proteolytic activity is responsible for the aberrant morphogenetic behavior
of endothelial cells expressing the middle T oncogene. Cell, 62, 435±445.
MuÈhlhauser, J., Marzioni, D., Morroni, M. et al. (1996) Codistribution of basic
®broblast growth factor and heparan sulfate proteoglycan in the growth
zones of the human placenta. Cell Tissue Res., 285, 101±107.
Murphy-Ullrich, J.E., Lightner, V.A., Aukhil, I. et al. (1991) Focal adhesion
integrity is downregulated by the alternatively spliced domain of human
tenascin. J. Cell Biol., 115, 1127±1136.
Nomura, M., Yamagishi, S., Harada, S. et al. (1995) Possible participation of
autocrine and paracrine vascular endothelial growth factors in hypoxiainduced proliferation of endothelial cells and pericytes. J. Biol. Chem.,
270, 28316±28324.
Purdy, L.P. and Metzger, B.E. (1996) In¯uences of the intrauterine metabolic
environment on adult disease: what may we infer from size at birth?
Diabetologia, 39, 1126±1130.
Riedl, S., Kadmon, M., Tandara, A. et al. (1998) Mucosal tenascin C content
in in¯ammatory and neoplastic diseases of the large bowel. Dis. Colon
Rectum, 41, 86±92.
Risau, W. and Lemmon, V. (1988) Changes in the vascular extracellular
matrix during embryonic vasculogenesis and angiogenesis. Dev. Biol.,
125, 441±450.
Rodesch, F., Simon, P., Donner, C. et al. (1992) Oxygen measurements in
endometrial and trophoblastic tissues during early pregnancy. Obstet.
Gynecol., 80, 283±285.
Roghani, M. and Moscatelli, D. (1992) Basic ®broblast growth factor is
internalized through both receptor-mediated and heparan sulfate-mediated
mechanisms. J. Biol. Chem., 267, 22156±22162.
Sage, E.H. and Bornstein, P. (1991) Extracellular proteins that modulate cellmatrix interactions. J. Biol. Chem., 266, 14831±14834.
Saksela, O., Moscatelli, D., Sommer, A. et al. (1988) Endothelial cell-derived
heparan sulfate binds basic ®broblast growth factor and protects it from
proteolytic degradation. J. Cell Biol., 107, 743±751.
Salvatore, C.A. (1968) The placenta in acute toxemia. A comparative study.
Am. J. Obstet. Gynecol., 102, 347±353.
Scheffen, I., Kaufmann, P., Philippens, L. et al. (1990) Alterations of the fetal
capillary bed in the guinea pig placenta following long-term hypoxia. In
Piiper, J., Goldstick, T.K. and Meyer, D. (eds), Oxygen Transfer to Tissue,
XII. Plenum Press, New York, pp. 779±790.
Schuhmann, R., Stoz, F. and Maier, M. (1988) Histometric investigations in
placentones (materno-fetal circulation units) of human placentae.
Trophoblast Res., 3, 3±16.
Senger, D.R., Galli, S.J., Dvorak, A.M. et al. (1983) Tumor cells secrete a
vascular permeability factor that promotes accumulation of ascites ¯uid.
Science, 219, 983±985.
Shiraishi, S., Nakagawa, K., Kinakawa, N. et al. (1996) Immunohistochemical
localization of vascular endothelial growth factor in the human placenta.
Placenta, 17, 111±121.
Shore, V.H., Wang, T.H., Wang, C.L. et al. (1997) Vascular endothelial
growth factor, placenta growth factor and their receptors in isolated
human trophoblast. Placenta, 18, 657±665.
Shreeniwas, R., Ogawa S., Cozzolino, F. et al. (1991) Macrovascular and
microvascular endothelium during long-term hypoxia: alterations in cell
growth, monolayer permeability, and cell surface coagulant properties. J.
Cell Physiol., 146, 8±17.
Todros, T., Sciarrone, A., Piccoli, E. et al. (1999) Umbilical Doppler
waveforms and placental villous angiogenesis in pregnancies complicated
by fetal growth restriction. Obstet. Gynecol., 93, 499±503.
Toole, B.P. (1991) Proteoglycans and hyaluronan in morphogenesis and
differentiation. In Hay, E.D. (ed.), Cell Biology of Extracellular Matrix.
Plenum, New York, pp. 305±341.
Turley, E.A. (1992) Hyaluronan and cell locomotion. Cancer Metastasis Rev.,
11, 21±30.
Turnbull, J.E., Fernig, D.G., Ke, Y. et al. (1992) Identi®cation of the basic
®broblast growth factor binding sequence in ®broblast heparan sulfate. J.
Biol. Chem., 267, 10337±10341.
Underhill, C. (1992) CD44: The hyaluronan receptor. J. Cell Sci., 103, 293±
298.
Underhill, C.B., Nguyen, H.A., Shizari, M. et al. (1993) CD44 positive
macrophages take up hyaluronan during lung development. Dev. Biol.,
155, 324±336.
Vandenbunder, B., Pardanaud, L., Jaffredo, T. et al. (1989) Complementary
patterns of expression of c-ets 1, c-myb and c-myc in the blood-forming
system of the chick embryo. Development, 107, 265±274.
Vuckovic, M., Ponting, J., Terman, B.I. et al. (1996) Expression of the
vascular endothelial growth factor receptor, KDR, in human placenta. J.
Anat., 188, 361±366.
Wernert, N., Raes, M.B., Lassalle, P. et al. (1992) c-ets1 proto-oncogene is a
transcription factor expressed in endothelial cells during tumor
vascularization and other forms of angiogenesis in humans. Am. J.
Pathol., 140, 119±127.
Wilting, J., Brand-Saberi, B., Huang, R. et al. (1995a) Angiogenic potential of
the avian somite. Dev. Dyn., 202, 165±171.
Wilting, J., Brand-Saberi, B., Kurz, H. et al. (1995b) Development of the
embryonic vascular system. Cell. Mol. Biol. Res., 41, 219±232.
Yayon, A., Klagsbrun, M., Esko, J.D. et al. (1991) Cell surface, heparin-like
molecules are required for binding of basic ®broblast growth factor to its
high af®nity receptor. Cell, 64, 841±848.
Received on January 24, 2000; accepted on July 18, 2000