| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Medical Research Council, Human Reproductive Sciences Unit, Edinburgh EH3 9ET, United Kingdom
b Department of Obstetrics and Gynecology of the University of Ulm, 89075 Ulm, Germany
c Regeneron Pharmaceuticals, Tarrytown, New York 10591
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
female reproductive tract, growth factors, pregnancy, syncytiotrophoblast
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Two processes, vasculogenesis and angiogenesis, are involved in placental development. Whereas vasculogenesis is responsible for establishment of the primitive vascular network, during angiogenesis the existing vasculature is remodeled [1]. Thus, differentiation and development of the fetoplacental vasculature is due to vasculogenesis and angiogenesis, whereas conversion of the spiral arterioles is a remodeling process associated with angiogenesis. Coordination of these diverse developmental steps appears to be mediated by locally acting angiogenic growth factors such as the placenta growth factor (PlGF) and vascular endothelial growth factor (VEGF) family and the angiopoietin (Ang-1 and Ang-2) family [2–6]. Both growth factor families act via their tyrosine kinase receptors [7]. Although PlGF binds only to the Flt receptor, VEGF uses the Flt and KDR receptors. The physiological role of PlGF in placenta development is at present unclear [8], however, for VEGF, it is known that binding to the receptor KDR causes endothelial cell differentiation and proliferation [9], whereas binding to Flt mediates endothelial cell interaction and tubule formation [10]. Tie receptors (Tie-1 and Tie-2) constitute other tyrosine kinase receptors involved in vascular development. Tie-1 is present on the trophoblast [11], but the ligand for this receptor and its function during placentation is unknown. The Tie-2 receptor is used by the angiopoietins. Ang-1 activates the Tie-2 receptor and stabilizes newly formed vessels by recruiting perivascular support cells such as pericytes [12, 13]. Ang-2 is a competitive antagonist to Ang-1 at the Tie-2 receptor, which may cause destabilization of the vasculature [14]. Therefore, VEGF mediates vasculogenesis, whereas the angiopoietins are more involved in the process of vascular remodeling during angiogenesis.
Although the regulation of vascular development in the embryo is quite well understood, little is known about the complex molecular regulation of placenta development, especially the role of the angiopoietins. VEGF appears to play a central role in regulation of placental angiogenesis [15]. It is suspected that VEGF drives fetoplacental vascular development [16] whereas the angiopoietins are believed to be associated with reworking of the maternal vessels during placentation [17]. Besides these effects on the fetal and maternal endothelium, both growth factor families have been suggested to promote trophoblast growth and migration [18, 19].
Elucidating the mechanisms that control normal placentation is important for understanding the pathophysiology of conditions associated with the impairment of vascular development during placentation such as intrauterine growth retardation [20–23] and pre-eclampsia [24–27]. Obtaining normal human placentae throughout pregnancy is difficult, especially in the second trimester. In several studies, first-trimester placentae originating from terminations were used, whereas third-trimester placentae have been obtained by caesarean delivery. However, second-trimester placentae are, in general, only available from miscarriages, which would probably not fulfill the criteria of normal placentation, and therefore, they have not been included in studies. To obtain information during all stages of placentation, we investigated placentae of the marmoset, a commonly used nonhuman primate model in reproduction in which it has been demonstrated that the fine structure of the fully differentiated placenta corresponds to that of the human [28]. To elucidate a comprehensive and conclusive picture of the spatial and temporal expression of potential molecular regulators of placenta development throughout gestation, the gene expression and protein localization of VEGF, angiopoietins, and their receptors were investigated using in situ hybridization and immunocytochemistry supplemented by Northern blot analysis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experiments were carried out in accordance with the Animals (Scientific Procedures) Act of 1986. Placentae were collected over a period of several years from pregnant animals in the colony breeding stock, as they were being replenished by younger animals. Animals were sedated using 100 µl ketamine hydrochloride (Parke-Davis Veterinary, Pontypool, Gwent, UK) and killed humanely with an i.v. injection of 400 µl euthetal (sodium pentobarbitone, Rhone Merieux, Harlow, Essex, UK). After cardiac exsanguination via a heparinized syringe, placentae were removed immediately and either fixed in 4% paraformaldehyde or transferred to Tri-reagent (Molecular Research Center, Inc., Cincinnati, OH) for RNA extraction. After 24 h, the fixed placentae were dehydrated and embedded in paraffin according to standard procedures.
Dating the Placentae
Eleven placentae were collected during different time points of gestation: at 7 wk (n = 1), 10 wk (n = 3), 12 wk (n = 1), 14 wk (n = 2), 17 wk (n = 1), and >19 wk (near term; n = 3). In all cases, morphological dating of the placenta according to the method of Merker et al. [28] was used to confirm the gestational age. Different developmental steps of marmoset placentation are illustrated in Figure 1. In the marmoset, the length of gestation is 20–21 wk. During 7–10 wk, cytotrophoblast cones with mesenchymal cores penetrate deep into the continuous mass of the syncytial trophoblast. Maternal vessels appear with a thickened basal lamina and triangular shaped endothelial cells (Fig. 1, a and b). Opening of maternal vessels leads to the first visible lacunae. Trabeculae consist of an inner mesenchymal layer in which small fetal capillaries have developed (Fig. 1, c and d). First erythroblasts appear in the chorionic vessels. The mesenchymal layer is surrounded by cytotrophoblast cells, and the trabeculae are covered with a continuous syncytial trophoblast layer. Through branching of these trabeculae, a labyrinth develops. Although maternal vessels persist in the marmoset placenta, increased numbers of gaps in the maternal vessel walls are found, through which the labyrinth (intertrabecula space) is filled with maternal blood. At this point, a hemodichorial placenta has developed. From 14 wk onward, a change from the hemodichorial to a hemomonochorial placenta occurs by a slow disappearance of cytotrophoblast cells. Hematopoietic foci containing densely packed cells of the red blood cell line are numerous (Fig. 1, e and f) but subsequently decline in number. Chorionic vessels of the chorion plate are well developed. Near term (Fig. 1, g and h), a complete hemomonochorial placenta exists. Hematopoietic foci and cytotrophoblast cells have disappeared completely. The trabeculae consists only of fetal capillaries, few mesenchymal cells, and a thin syncytial trophoblast layer. Numerous chorion vessels have developed in the chorion plate.
|
Immunocytochemistry
Tissue sections were dewaxed in xylene, rehydrated in descending concentrations of ethanol, and placed into water. Except for VEGF, antigen retrieval was achieved by pressure cooking (Clypso pressure cooker, Tefal, Essex, UK) in 0.01 M citrate buffer pH 6.0 for 6 min. Sections were left for 20 min in hot buffer before cooling in water. Endogenous peroxidase activity was quenched with a 30-min incubation in 3% hydrogen peroxide in methanol at room temperature. To reduce nonspecific staining, sections were incubated for 20 min in normal rabbit serum (NRS, 1:5 dilution) in Tris-buffered saline (TBS) containing avidin, washed in TBS, and incubated for another 20 min in NRS biotin (avidin/biotin blocking kit, Vector Laboratories, Burlingame, CA). After 2 washes in TBS, slides were exposed overnight at 4°C to the primary antibodies: anti-VEGF (rabbit-anti-human VEGF polyclonal, 1:50 dilution), anti-Flt (rabbit-anti-human Flt, polyclonal, 1:200 dilution), anti-KDR (mouse-anti-human KDR, monoclonal, 1:300), anti-Ang-1 (rabbit-anti-human Ang-1, polyclonal, 1:100 dilution), anti-Ang-2 (goat-anti-human Ang-2, polyclonal, 1:20 dilution), and anti-Tie-2 antibody (mouse-anti-rat Tie-2 monoclonal, 1:500 dilution). Antibodies for VEGF (sc-152), FLT (sc-316), KDR (sc-6251), Ang-1 (sc-9044), and Ang-2 (sc-7015) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The Tie-2 antibody was prepared by Regeneron Pharmaceuticals (11E11-H11-E7; Tarrytown, NY). Slides were washed 2 times in TBS followed by incubation for 30 min at room temperature with the secondary antibody (swine-anti-rabbit-biotinylated, rabbit-anti-mouse-biotinylated or rabbit-anti-goat-biotinylated antibody respectively, 1:200 dilution, DAKO A/S, Glostrup, Denmark). For signal amplification, the avidin-biotin complex (ABC) method was used. After incubation with ABC-horseradish peroxide (HRP, DAKO) for 30 min at room temperature and 2 washes in TBS-Tween, signal enhancement was achieved by exposing the slides to biotinylated tyramide (GenPoint kit, DAKO) for 20 min at room temperature. Slides for VEGF, Ang-2, and Tie-2 were then incubated in ABC-alkaline phosphatase (AP) for 30 min. Visualization was performed using 500 µl/slide nitro blue tetrazolium (NBT) solution containing 45 µl NBT substrate (Boehringer-Mannheim, Mannheim, Germany), 10 ml NBT buffer, 35 µl 5-bromo-4-chloro-3-indolyl phosphate and 10 µl levamisole. For FLT, KDR, and Ang-1, tyramide enhancement followed by ABC-HRP was repeated 3 times. After the last tyramide incubation, ABC-AP was used before final signal detection with NBT substrate as described above.
In Situ Hybridization
In situ hybridization was performed as described previously [29, 30] using cRNA probes for human VEGF, Flt, KDR, Ang-1, Ang-2, and Tie-2. The known marmoset sequence for VEGF, Flt, and KDR shows 97%–98% homology to the human. The marmoset sequences for angiopoietins are unknown. Because of the specificity of the hybridization signals and the identical protein location, we were justified in using the human probes for marmoset tissue. For each probe and slide a sense and antisense was carried out. Sense and antisense probes were prepared using an RNA transcription kit (Ambion, Austin, TX) and labeled with 35S-uridine 5'-triphosphate (NEN, Boston, MA). Deparaffinized sections were treated with 0.1 N HCl and then digested in proteinase K (5 µg/ml, Sigma, St. Louis, MO) for 30 min at 37°C. After prehybridization for 2 h at 55°C for all probes, subsequent hybridization was performed in a moist chamber overnight at 55°C. High stringency posthybridization washings and RNAse treatment were used to remove excess probe. Slides were then dehydrated, dried, and dipped in Ilford G5 liquid emulsion (H.A. West, Edinburgh, UK). Exposure times were 4–5 wk. Slides were subsequently developed (Kodak D19 developer, Kodak, Rochester, NY) and fixed (Kodak GBS, Kodak). All slides were counterstained with hematoxylin (Richard-Allan, Richland, MI), dehydrated, and mounted.
Northern Blot Analysis
Total RNA was isolated from individual tissue specimens using Tri-reagent (Molecular Research Center, Inc., Cincinnati, OH) according to the manufacturer's instructions. Ten-microgram aliquots of total RNA from each placental sample were loaded onto 1.0% formaldehyde-agarose gels. Ethidium bromide was incorporated at a concentration of 0.2 µg/ml. After electrophoresis, the gel was photographed under UV light to visualize the 18S and 28S ribosomal RNA bands to confirm equal loading of the lanes. RNA was then transferred to a Nytran SuperCharge nylon membrane (Schleicher & Schuell, Inc., Keene, NH) using a rapid downward transfer system (Turboblotter; Schleicher & Schuell), and the transfer was continued for 4–5 h. After transfer, membranes were UV cross-linked using UV Stratalinker 1800 (Stratagene, La Jolla, CA). Membranes were hybridized to random primed 32P-labeled probes at 65°C overnight. The Prime-It II random primer labeling kit (Stratagene) was used to make the probes. The filter was washed at 65°C in 2x SSC, 0.1% SDS solution, air dried, and exposed to autoradiograph film for 3–10 days.
Probes
The hVEGF probe was a 543-base pair (bp) reverse transcriptase-polymerase chain reaction (RT-PCR) fragment amplified from human lung total RNA. The primers for the first PCR reaction were 5'-CGGTCGGGCCTCCGAAACCATGAAC-3' and 5'-CGAAACCCTGAGGGAGGCTCCTTCCTC-3'. The second (nested) PCR was done using primes 5'-CCGAAACCATGAACTTTCTGCTGTCTTGG-3' and 5'-CTCCTTCCTCCTGCCCGGCTCAC-3'. The sequence of the PCR product was confirmed by sequencing.
The hVEGF-R1 (hFlt1) probe was a 641-bp PCR fragment that hybridizes to base pairs 22–663 of the Flt mRNA. This comprises the region of the encoding extracellular domain of the Flt protein. Thus the probe recognizes both full-length and soluble Flt. The PCR template was pMT21.hFLT1.full. The primers used for PCR were 5'-GGGGTCCTGCTGTGCG-3' and 5'-GAGATAGTTTGTCTTATAC-3'.
The hVEGF-R2 (hKDR1) probe was a 692-bp PCR fragment. The PCR template was pBS-hFLK1.full. The primers used for PCR were 5'-CCTAGTGTTTCTCTTGATCTGCCCAGG-3' and 5'-CACATTTAGTTCAGTTCTTGC-3'.
The hAng-1 probe was a 2.0-kilobase (kb) XhoI fragment of pKS+/hTL1 containing the full-length coding sequence of human angiopoietin 1.
The hAng-2 probe was a 2.3-kb EcoRI fragment of pKS+/hTL2 containing the full-length coding sequence of human angiopoietin 2.
The hTie-2R probe was a 1-kb BamHI fragment of pKS+/hTL2 comprising the ectodomain of human Tie2 receptor (aa 70–380).
Analyses
Slides were analyzed qualitatively under lightfield and quantitatively under darkfield conditions at 20x magnification. Because VEGF showed expression in both fetal and maternal placenta tissue, and differences in expression levels could not be distinguished by Northern blots, the grain density (number of grains/µm2) was measured for VEGF as a value for gene expression as described previously [29, 30]. Six randomly chosen fields in each the maternal and fetal sites of the placenta in 4 sections per animal were analyzed. The image analyses system was set up to identify mRNA expressing cells (maternal decidua cells or fetal cytotrophoblast cells, respectively), to outline them and measure the grain density within these cells. Tissue background density was measured for each antisense slide and subtracted from the measurements. In each compartment, the mean density was calculated for the number of cells and fields assessed per animal.
Statistical Analysis
Differences between grain density of VEGF-expressing maternal and fetal cells were determined using a two-tailed, unpaired t-test. Differences were considered to be significant when P < 0.05. The tests were performed using SPSS version 6.1 (SPSS Inc., Chicago, IL) for the Macintosh computer. All values are given as mean ± SEM.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
VEGF Expression
VEGF mRNA and protein were expressed in the cytotrophoblast cells of the trabeculae (Fig. 2, a and b) and in the maternal decidual cells (Fig. 3, a and e). The grain density in maternal decidual cells (0.62 ± 0.004 grains/µm2) was significantly higher than in the fetal cytotrophoblast cells (0.46 ± 0.004 grains/µm2). This pattern of mRNA expression remained throughout gestation (Fig. 3, b, c, and e). The VEGF mRNA was evenly distributed within the trophoblast. With the disappearance of the cytotrophoblast at term, the VEGF expression in the trabeculae appeared to decline, however, high mRNA levels remained detectable in decidua cells (Fig. 3d). Northern blot analysis revealed no differences in mRNA expression from 10 to 17 wk, and a decrease at term (Fig. 4).
|
|
|
KDR Expression
The KDR receptor mRNA and protein were mainly localized in the fetal vasculature (Fig. 2, c and d, and Fig. 3j). Early during gestation (7–10 wk), KDR mRNA was also expressed at high levels in the mesenchymal cores in the endothelial precursor cells (Fig. 3, f and g). High expression was found later at 17 wk in the endothelium of vessels of the chorion plate (Fig. 3h). In fetal capillaries of the trophoblast, KDR expression appeared to be lower than expression in chorion vessels (Fig. 3h). In term placentae, KDR mRNA expression remained in the fetal vasculature (Fig. 3i). Northern blot analyses showed increasing total mRNA expression from 10 wk to term (Fig. 4).
Flt Expression
Flt receptor mRNA and protein were exclusively found in syncytial trophoblast cells (Fig. 2, e and f, and Fig. 3o). In the 7-wk placenta (Fig. 3k), Flt expression was low, but a marked increase was observed in the syncytial trophoblast near the fetal maternal interface of the 10-wk placentae (Fig. 3l). Flt expression was evenly distributed in the thin syncytial trophoblast layer covering the trabeculae of the 17-wk placenta (Fig. 3m) and remained high throughout gestation (Fig. 3n). Northern blot analysis showed 2 hybridization sites. The hybridization band for the membrane-bound Flt receptor above 28S showed increasing levels of mRNA from 10 wk until term. The second band, sFlt, was found between the 28S and 18S, showing a strong signal at 17 wk and at term.
Ang-1 Expression
Ang-1 mRNA and protein were mainly expressed in the syncytial trophoblast (Fig. 5, a and b). Early during placentation (7 wk), Ang-1 mRNA was found in the mesenchymal cores of cytotrophoblast cones at low levels (Fig. 6a). From 10 wk onward, Ang-1 was exclusively localized in the syncytial trophoblast lining the trabeculae (Fig. 6, b and d). An increase in mRNA expression was observed in the 17-wk placenta. In the term placentae, the same expression pattern remained (Fig. 6c). Northern bolt analyses revealed relatively low Ang-1 mRNA expression in the 10-wk placentae, and a marked increase during gestation until term (Fig. 7).
|
|
|
Ang-2 Expression
Ang-2 mRNA and protein were localized exclusively in the endothelial cells of maternal vessels (Figs. 5, c and d, and 6h). In situ hybridization showed high expression from 7 to 17 wk (Fig. 6, e and f), whereas Ang-2 expression appeared to be decreased at term (Fig. 6g). In the Northern blots (Fig. 7), high mRNA levels were detected at 10 wk, and decreased levels in 14-wk and term placentae. A strong signal was also found in the 17-wk placenta.
Tie-2 Expression
Tie-2 mRNA and protein were found in the wall of chorion vessels and cells of the chorion plate, and in the endothelium of fetal capillaries and maternal vessels (Fig. 5, e and f). During the early placentation period at 7 wk, Tie-2 was localized in the mesenchymal cells of the cytotrophoblast cones (Fig. 6i). From 14 wk (Fig. 6j) onward, high mRNA levels were found in the chorion vessels and mesenchymal cells of the chorion plate. To a lesser degree, Tie-2 mRNA was found in fetal capillaries, and maternal vasculature and cytotrophoblast cells. At term (Fig. 6k), the mRNA expression remained evenly distributed in the chorion plate. Northern blot analysis showed an overall low expression of Tie-2 (Fig. 7) with increasing amounts from 14 wk until term.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
During early placentation in the marmoset the syncytial trophoblast invades uterine epithelium and subepithelial maternal arterioles. The cytotrophoblast follows this growth direction by the formation of cytotrophoblastic cones toward the syncytial trophoblast. VEGF mRNA was highly expressed in the maternal decidua, whereas its receptor, Flt, was highly expressed in the invading syncytial trophoblast. Similar localization of VEGF and Flt have been shown in the human placenta [19, 31, 32] and in placentae of other species [2, 3, 33–35]. Thus, there is strong evidence that maternal VEGF, acting in a paracrine route by binding to the Flt receptor, is involved in the regulation of trophoblast invasion. This pattern of VEGF and Flt expression remained in the marmoset throughout gestation, even though invasion ceases at a certain time point during placentation. Within the trophoblast tree, VEGF was localized in cytotrophoblast cells and Flt in the neighboring syncytial trophoblast, suggesting that VEGF/Flt may play an important role in coordination of trophoblast differentiation and migration in addition to regulating trophoblast invasion.
Understanding the role of VEGF is complicated because the Flt receptor also exists in a soluble form. In the mouse placenta, it was demonstrated that the Flt pre-mRNA can be alternatively spliced, resulting in the membrane-bound Flt receptor and a soluble form (sFlt) that comprise the ligand-binding domain of the Flt receptor [36], thus becoming a potent endogenous antagonist of VEGF [33]. As an endogenous antagonist, sFlt may be important in the regulation of placenta angiogenesis by balancing VEGF (and PlGF) actions [36, 37], whereas the membrane bound form may promote trophoblast development. The existence of sFlt in the primate is confirmed by our Northern blots. Our in situ hybridizations do not distinguish between the membrane bound and the soluble Flt, but showed expression only in the cytotrophoblast. Thus, a proportion of the signal must be sFlt. Ligand-binding studies in the human [8] showed that VEGF was binding exclusively to endothelial cells, but no binding was detectable in the trophoblast. Using a specific probe for mRNA encoding sFlt, it was found that the trophoblast expresses only sFlt [8]. Although it remains questionable whether VEGF is involved in regulation of trophoblast development, its actions are modulated by an endogenous antagonist, sFlt, in the primate placenta. It appears that placentation depends on a dynamic equilibrium between angiogenic factors and endogenous antagonists. Imbalance within this equilibrium may be a cause of abnormal vacularization because it is found in pathological conditions such as intrauterine growth retardation (IUGR) [21]. IUGR is associated with elevated PlGF and decreased levels of VEGF and sFlt. PlGF is believed to increase trophoblast proliferation, whereas decreased VEGF may cause insufficient microvascular growth at the terminal villi, which is characteristic of IUGR [21].
Marmoset placenta cannot be dissected into fetal and maternal tissue, thus the relative contribution to the RNA in Northern blots are unknown. However, it is rational to assume that the fetal placenta makes up a large proportion in the extracted RNA. The Northern blots give an approximate index of any substantive changes as long as expression is restricted only to the maternal or to the fetal placenta. For the factors that are expressed only in the fetal placenta (Flt, KDR, Ang-1, and Tie-2) or only in the maternal tissue (Ang-2), Northern blots have detected expression differences throughout gestation. For VEGF mRNA, which was expressed in fetal and maternal placenta, levels detected in the Northern blots appeared to be low throughout pregnancy. This may be because VEGF expression was highest in maternal tissue, which is likely to make up a relatively small proportion of the total tissue in RNA. Here, grain density measurements give a more accurate representation for the expression levels than do Northern blots. The decreased VEGF mRNA levels at term are consistent with the disappearance of the cytotrophoblast cells at this time.
With ongoing placentation in the marmoset, the cytotrophoblast deeply penetrates the syncytial trophoblast, the typical trabeculae are formed, and a labyrinth develops. Ang-1 was highly expressed in the syncytial trophoblast, whereas its receptor, Tie-2, was located in the cytotrophoblast. This suggests that Ang-1 regulates trophoblast growth by a paracrine route, and is supported by findings in the human placenta, in which Ang-1 and Tie-2 have also been detected in the cyto/syncytiotrophoblast bilayer [18]. By using a recombinant Tie-2-Fc to inhibit Ang-1 action, it was demonstrated that Ang-1 stimulated trophoblast growth and migration [18], seemingly in a dose-dependent manner. The results of the current study showed increasing gene expression of Ang-1 during gestation. Thus, Ang-1 may trigger the in-growth of the cytotrophoblast cones in the syncytial trophoblast, while the relatively high levels later during placentation may be required for branching of the trabeculae and shaping the labyrinth.
Simultaneous with trophoblast growth, the fetoplacental vasculature develops, which includes differentiation of endothelial cells, proliferation, tubulus formation, and stabilization. From their localization, VEGF and KDR appear to be the main regulators of this process. KDR was highly expressed early during placentation in the mesenchymal cells of the trophoblast from which endothelial cells differentiate to form fetal capillaries. KDR may be required to drive this process in the placenta, because this receptor has been shown to be responsible for endothelial cell differentiation [9]. The high KDR expression in the mesenchymal cells found by in situ hybridization appeared to contrast with low mRNA levels detected with Northern blot analysis at that time. These findings may be explained by the differences in the techniques. For Northern blot analysis, tissue is homogenized and RNA is extracted, reflecting total RNA. In contrast, in situ hybridization localizes mRNA expression within individual cells in a tissue section. Thus, the Northern blot may indicate low expression in a tissue when, in fact, the mRNA is highly expressed in a minority of cells.
After differentiation of endothelial cells, the first fetal capillaries are formed in the trophoblast, which begin to increase in number by the process of angiogenesis accompanying the ongoing branching of the trophoblast. Within the trophoblast, VEGF was localized in cytotrophoblast cells, indicating a paracrine mechanism for VEGF acting on neighboring endothelial cells through the KDR receptor to mediate growth of fetal blood vessels. A comparable role for the VEGF/KDR ligand-receptor pair was suggested in the human placenta in which KDR was also expressed in fetal capillaries [19], and VEGF in trophoblast cells [38]. In the current study, a decrease of KDR expression in fetal capillaries was seen at midgestation, which may occur because after differentiation of endothelial cells and early capillary formation in fetal placenta, the process of vasculogenesis is complete. However, with ongoing gestation, the number of fetal capillaries and the number of chorion vessels increase. This increase in the total number of blood vessels further explains the increase in KDR mRNA levels at 17 wk and at term found in the Northern blots.
While fetal capillaries of the trophoblast are thin-walled in order to allow oxygen diffusion, chorion vessels are stabilized by a thick wall of pericytes, smooth muscle cells, or both, to fulfill their task of collecting and draining all fetal placental blood. In these vessels, Tie-2 was highly expressed, whereas in fetal capillaries, only low Tie-2 levels were observed. Furthermore, the Northern blot analyses confirmed that Ang-1 and Tie-2 mRNA were highly expressed during the later stages of placentation. These results imply that here, the Ang-1/Tie-2 ligand-receptor pair acts on the fetal vasculature, especially on chorion vessels, in order induce stability. In the human, relatively high levels of Ang-1 expression were found in the media of stem villous vessels at term [18], which was also consistent with its reported role in vessel maturation and stabilization [12].
To attain uteroplacental circulation, maternal vessels have to be transformed. In the marmoset placenta, a proportion of maternal vessels persists throughout pregnancy, but in several areas, maternal vessels are opened, forming gaps through which maternal blood is released into the labyrinth [28, 39]. This opening of maternal vessels is a crucial step during placentation and requires remodeling of the vascular wall. A unique pattern of Ang-2 expression was discovered in this study. Throughout gestation, Ang-2 was exclusively located in maternal endothelial cells together with its receptor, Tie-2. High levels of Ang-2 mRNA were detected by in situ hybridization and Northern blot analysis during early gestation, at the time when maternal remodeling mainly takes place. Ang-2 may be a possible candidate for induction of maternal vascular transformation, because Ang-2 destabilizes the vasculature [12]. In the human, high Ang-2 levels were found in the first trimester placenta. Ang-2 was localized in the syncytial trophoblast, and Tie-2 in maternal endothelium, suggesting a paracrine mechanism during maternal remodeling [17]. In contrast, in the marmoset, both Ang-2 and Tie-2 are expressed in the maternal endothelium, indicating an autocrine regulation. These findings lead us to propose that in the marmoset, the maternal vasculature itself appears to control this remodeling process, which is necessary for successful placentation.
In conclusion, this study reveals a detailed picture of the spatial and temporal expression of angiogenic factors in the primate placenta, indicating for the first time the interrelationship between the 2 major angiogenic families, VEGF and angiopoietins. The results suggest that the VEGF/Flt and Ang-1/Tie-2 ligand-receptor pairs may be involved in trophoblast invasion and development; VEGF/KDR and Ang-1/Tie-2 ligand-receptor pairs may control fetoplacental vascular differentiation, development, and stabilization; and the Ang-2/Tie-2 ligand-receptor-pair may be necessary for remodeling of the maternal vasculature. Furthermore, expression of sFlt suggests that successful vascular development during placentation is regulated by and dependent upon angiogenic factors and their endogenous antagonists.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported in part by a grant to C.W. from Deutsche Forschungsgemeinschaft. 
2 Correspondence and current address: C. Wulff, Department of Obstetrics and Gynecology of the University of Ulm, Sekretariat Prof. Kreienberg, Pritwitzstrasse 43, 89075 Ulm, Germany. FAX: 49 731 500 27602; christine-wulff{at}onlinehome.de
Accepted: October 30, 2001.
Received: June 8, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D Herr, M Rodewald, H M Fraser, G Hack, R Konrad, R Kreienberg, and C Wulff Regulation of endothelial proliferation by the renin-angiotensin system in human umbilical vein endothelial cells Reproduction, July 1, 2008; 136(1): 125 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Sengupta, P G L Lalitkumar, A R Najwa, D S Charnock-Jones, A L Evans, A M Sharkey, S K Smith, and D Ghosh Immunoneutralization of vascular endothelial growth factor inhibits pregnancy establishment in the rhesus monkey (Macaca mulatta) Reproduction, June 1, 2007; 133(6): 1199 - 1211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Hoffmann, J.-J. Feige, and N. Alfaidy Expression and Oxygen Regulation of Endocrine Gland-Derived Vascular Endothelial Growth Factor/Prokineticin-1 and Its Receptors in Human Placenta during Early Pregnancy Endocrinology, April 1, 2006; 147(4): 1675 - 1684. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Geva, D.G. Ginzinger, D.H. Moore II, P.C. Ursell, and R.B. Jaffe In utero angiopoietin-2 gene delivery remodels placental blood vessel phenotype: a murine model for studying placental angiogenesis Mol. Hum. Reprod., April 1, 2005; 11(4): 253 - 260. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J Rowe, C. Wulff, and H. M Fraser Angiogenesis and microvascular development in the marmoset (Callithrix jacchus) endometrium during early pregnancy Reproduction, July 1, 2004; 128(1): 107 - 116. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
