| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Article |
Center for Reproductive Sciences, Department of Obstetrics, Gynecology and Reproductive Sciences (E.G., A.B., R.B.J.), Genomic Core, Comprehensive Cancer Center (D.G.G., D.H.M.), and Department of Clinical Pathology (C.J.Z.), University of California, San Francisco, San Francisco, California 94143-0556
Address all correspondence and requests for reprints to: Robert B. Jaffe, M.D., Center for Reproductive Sciences, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, San Francisco, California 94143-0556. E-mail: . jaffer{at}obgyn.ucsf.edu
Abstract
During placental development, vessel formation occurs initially by vasculogenesis and subsequently by branching and nonbranching angiogenesis. We investigated vascular endothelial growth factor (VEGF)-A, angiopoietin (Ang)-1 and -2 transcript profiles, and the protein products that they encode in placentas from normotensive pregnancies throughout pregnancy. In addition, we compared these genes in placentas from normotensive women and those with preeclampsia during the third trimester. Quantitative real-time PCR analysis demonstrated that VEGF-A and Ang1 mRNA increased in a linear pattern by 2.5 (not significant) and 2.8%/wk (P = 0.034), respectively, whereas Ang2 decreased logarithmically by 3.5%/wk (P = 0.0003). Ang2 mRNA was 400- and 100-fold higher than Ang1 and VEGF-A, respectively, in the first trimester and declined to 20-fold and 7-fold in the third. Ang2 protein (ELISA) decreased by 4.7%/wk (P = 0.0001), whereas Ang1 and VEGF-A were undetectable. In preeclampsia compared with normotensive pregnancy, only VEGF-A mRNA increased significantly, by 3-fold (P = 0.006). This increase may be related to low oxygen tension, as VEGF-A is up-regulated by hypoxia.
In situ hybridization and immunohistochemical studies revealed that VEGF-A was localized in cyto- and syncytiotrophoblast and perivascular cells, whereas Ang1 and Ang2 were only in syncytiotrophoblast and perivascular cells in the immature intermediate villi during the first and second trimesters, and mature intermediate and terminal villi during the third trimester.
These data suggest that these molecules may play important roles in placental biology and chorionic villus vascular development and remodeling in an autocrine/paracrine manner. The tight correlation between Ang2 mRNA and protein indicates that regulation of placental vascular development occurs at the transcriptional, and not translational, level.
SUCCESSFUL PREGNANCY requires the development of a complex maternal and fetal vascular network to support increasing oxygen and metabolic demands of the growing fetus. There are three stages in human placental vascular development: an initial stage of vasculogenesis, and subsequently branching and then nonbranching angiogenesis (1, 2). Vasculogenesis involves the de novo formation of new vessels, whereas angiogenesis refers to the formation of capillaries from preexisting vessels (1, 2). Vascular endothelial growth factor (VEGF)-A is a potent mitogen and survival factor for endothelial cells that has been shown to initiate vasculogenesis and angiogenesis by inducing endothelial cell proliferation, migration and sprouting, as well as promoting endothelial cell formation of tube-like structures (3, 4). The angiopoietins Ang1 and Ang2 act subsequently in the later stages of angiogenesis, in concert with VEGF-A (1, 2). Ang1, which possesses only weak endothelial cell mitogenic activity, causes endothelial maturation and vascular stabilization in its surrounding tissue (1, 2, 5, 6, 7, 8), whereas Ang2, the natural antagonist of Ang1 in endothelial cells, acts in a complementary manner with VEGF-A at the front of the invading vascular sprouts by blocking constitutive stabilization and maturation of vessels, allowing them to remain in a more plastic state in which they may be more responsive to a sprouting signal provided by VEGF-A (1, 2, 8, 9).
The development of a normal functioning placental vascular network requires a remarkable degree of coordination between different vascular endothelial cell-specific growth factors and cell types and is exquisitely dependent upon signals exchanged between these cells (Fig. 1
; Refs. 10, 11, 12). Therefore, we hypothesized that these three vascular endothelial cell-specific growth factors play major roles in vasculogenic, and branching and nonbranching angiogenic, processes and the transformation of one to the other during human pregnancy: VEGF-A by inducing vasculogensis and angiogenesis, Ang1, which acts subsequently, causing maturation of the vascular network, and Ang2 causing destabilization of this network, leading to the formation of branching and nonbranching angiogenesis. Moreover, because both VEGF-A and Ang2 gene expression are up-regulated by hypoxia, we investigated the transcript profiles of these genes and the protein products that they encode to gain insights into the biology of placental vascularization in hypertensive-hypoxic pregnancy (preeclampsia).
|
Tissue preparation
Human placental tissues were obtained from first (n =7), second (n =7) and third (n =10) trimesters of normotensive pregnancies as well as third trimester (n =5) pregnancies complicated by severe preeclampsia [with intrauterine growth restriction (IUGR)]. Preeclampsia was defined according to the National High Blood Pressure Education Program: Working Report on High Blood Pressure in Pregnancy (blood pressure 
140/90 mm Hg after 20 wk gestation with complete remission 6 wk postpartum and proteinuria 300 mg/24 h) (13, 14). Patients with chronic hypertension were excluded. The gestational age and birthweight of women with preeclampsia were 33.4 ± 4.0 wk and 2067.9 ± 980.1 g vs. 39.0 ± 1.7 wk and 3556.7 ± 537.0 g for normotensive pregnant women (P < 0.002). Infants were diagnosed with IUGR only when their birth weights were at or below the fifth percentile for their gestation (15, 16).
Placental tissue from the first and second trimesters was obtained after elective pregnancy termination (7–24 wk gestation). Terminations were performed by dilatation and evacuation, and gestational age was determined by fetal foot length. Tissue from third trimester placentas was obtained after vaginal or cesarean deliveries (25–41 wk gestation).
Tissues used for RNA and protein analyses were dissected and frozen immediately in liquid nitrogen and stored at -80 C. Tissues used for immunohistochemistry were fixed by immersion for 24 h in 4% paraformaldehyde/100 mmol/liter PBS and embedded in paraffin.
The study was approved by the Committee on Human Research, University of California (San Francisco, CA).
RNA and protein extraction
Total RNA was prepared from snap-frozen human placental tissue samples (weight 1–5 mg) by homogenization according to the method of Chomczynski and Sacchi (17), using TRIzol Reagent (Life Technologies, Inc., Rockville, MD). RNA was quantified by absorbance at 260 nm, analyzed by electrophoresis on a 1% agarose gel, and stored at -80 C.
Proteins were extracted from snap-frozen tissue samples (weight 1–5 mg) by homogenization in RIPA (20 mM Tris, 150 mM NaCl, 1 mM EDTA, 5 mM benzamidine, 50 mM sodium fluoride, 1 mM sodium orthoanadate, and Complete tablet protease inhibitors)/10% Nonidet P-40 buffer (Roche, Indianapolis, IN). Protein levels were assayed using a Bio-Rad Laboratories, Inc. (Hercules, CA) protein assay and were quantified by absorbance at 595 nm using a DU 640 spectrophotometer (Beckman Coulter, Inc., Palo Alto, CA) and stored at -80 C.
Real-time quantitative PCR (qPCR)
Reverse transcriptase reactions were performed by heating a 15-µl reaction mixture containing 2 µg total RNA and 0.5 µg oligodeoxythymidine (Life Technologies, Inc.) at 70 C for 10 min. After cooling, 25 U rRNAasin ribonuclease inhibitor (Life Technologies, Inc., Gaithersburg, MD) and 200 U Moloney’s murine leukemia virus ribonuclease reverse transcriptase (Promega Corp., Madison, WI) were added in a final 25-µl reaction mixture containing 10 mM deoxy-NTP (Life Technologies, Inc.) and 5 µl Moloney’s murine leukemia virus reaction buffer (Promega Corp.), incubated for 1 h at 42 C and heated 15 min at 70 C.
Relative expression levels were measured using the 5' fluorogenic nuclease assay in qPCR, using the 5' nuclease assay on the ABI PRISM 7700 (Applied Biosystems, Foster City, CA). For each gene, PCR was conducted in triplicate with 50-µl reaction volumes of 1x PCR buffer A (Applied Biosystems), 5.5 mM MgCl2, 0.4 µM of each primer, 200 µM each deoxy-NTP, 100 nM probe, and 0.025 U/µl Taq Gold (Applied Biosystems). In each experiment, a large master mix of the above components was made for each cDNA and aliquoted into each optical reaction tube. Each primer/probe set (5–10 µl) was then added, and PCR conducted using the following cycle parameters: 95 C, 12 min x 1 cycle (95 C, 20 sec; 60 C, 1 min) x 40 cycles.
Analysis was carried out using the sequence detection software supplied with the ABI PRISM 7700. This software calculates the threshold cycle (Ct) for each reaction and this Ct is used to quantify the amount of starting template in the reaction. The Ct values for each set of three reactions were averaged for all subsequent calculations. A difference in Ct values (
Ct) was calculated for each gene by taking the mean Ct of triplicates and subtracting the mean Ct of the control gene triplicates for each cDNA sample at the same concentration (18).
The primers chosen using Primer Express Software (Applied Biosystems) for sense and antisense, respectively, were: human VEGF-A cDNA (GenBank accession no. NM003376) 5'-CTCTACCTCCACCATGCCAAG-3' and 5'-AGACATCCA TGAACTTCACCACTTC-3'; human Ang1 cDNA (GenBank accession no. U83508) 5'-GCAACTGGAGCTGATGGACACA-3' and 5'-CATCTGCACAGTCT CTAAATGGT-3'; human Ang2 cDNA (GenBank accession no. AF004327) 5'-CAGATTTTGGACCAGACCAGTG-3' and 5'ACTGTATGTTGGATGATGTGCTTG-3' and human ß-glucuronidase (GUS) (GenBank accession no. NM000181) 5'-CTCATTTGGAATTTGCCGATT-53 and 5'-CCGAGTGAAGATCCCCTTTTA-3'. The primers were obtained from IDT (Integrated DNA Technologies, Coralville, IA). The TaqMan fluorogenic probe consisted of the following sequences: VEGF-A, 5'-FAM-TGGCAGAAGGAGGAGGGCAGAATC A-6-carboxy tetramethyl rhodamine-3' (TAMRA); Ang1, 5'-FAM-CAATCTTTGCACTAAAGAAG GTGTTTTACT-TAMRA; Ang-2, 5'-FAM-CCTAGAAAAGAAGGTGCTAGCTATG GAA-TAMRA; GUS, 5'-FAM-TGAACAGTCACCGACGAGAGTGCTGG-TAMRA. All of the probes were purchased from IDT.
Although gestational age ranges in the third trimester patients with preeclampsia were lower, the relative mRNA and protein values were distributed within the normal range of normotensive pregnancies, and thus likely were not influenced by gestational age. Moreover, no difference in relative VEGF-A and Ang transcription was found between placentas from spontaneous vaginal delivery and cesarean section.
RNA localization (in situ hybridization)
Human Ang1 (570-bp) NotI-EcoRI fragment was subcloned into PstI site (214-bp, corresponding to nucleotides 512–726) of pBluescript II KS+. In vitro transcription was carried out using T3 for generation of sense cRNA and T7 for generation of antisense cRNA from plasmid linearized with EcoR I and with BamH I, respectively. Human Ang2 (640-bp) EcoRI-HindIII fragments were subcloned into PstI and HindIII sites (162-bp corresponding to nucleotides 480–642) of pBluescript II KS+. In vitro transcription was carried out using T7 for generation of sense cRNA and T3 for generation of antisense cRNA from plasmids linearized with HindIII and BamHI, respectively. In vitro transcription of human VEGF-A (370-bp corresponding to nucleotides 1354–1723) EcoRI-BamHI fragment was carried out using T3 for generation of sense cRNA and T7 for generation of antisense cRNA from plasmids linearized with EcoRI and BamHI, respectively.
Pretreated placental sections were deparaffinized in xylene, rinsed in graded ethanols (70%, 90%, 95%, 100%, 5 min each), followed by 1x Tris-buffered saline (TBS). Slides were treated with proteinase K (2 µm/ml; Sigma, St. Louis, MO) for 20 min at 42 C. Acetylation of the slides was carried out for 10 min at room temperature in 0.1 mol/liter triethanolamine hydrochloride/acetic anhydride (Sigma). Slides were rinsed in 1x TBS for 10 min, dehydrated through graded ethanols, and dried at room temperature. Slides were then incubated with hybridization mixture [2x standard saline citrate (SSC), 10% dextran sulfate, 1x Denhardt’s solution (Sigma), 50% formamide, and 0.02% SDS] in a humidified oven for 30 min at 45 C. Hybridization with 1 ng/ml digoxigenin-uridine 5'-triphosphate (Roche) labeled antisense or sense probes and 100 µg/ml yeast tRNA (Sigma) was performed in 50% formamide in a humidified oven at 52–53 C overnight.
After overnight incubation, the slides were rinsed in 2x SSC for 20 min at room temperature, 0.2x SSC for 1 h at 55 C, 0.5x SSC for 20 min, 1x TBS for 10 min and 1% BSA (Sigma) at room temperature. The slides were then incubated for 1 h in a humidified chamber with 5 µl/ml antidigoxigenin alkaline phosphatase (Roche), and washed with 1x TBS for 10 min at room temperature. Slides were developed using 5-bromo-4-chloro-3-indoyl phosphate (Sigma) and nitroblue tetrazolium (Sigma) in N,N,dimethylformamide (Sigma). The sections were than incubated at room temperature overnight, rinsed and dehydrated in graded alcohols, cleared with xylene, and mounted with Cytoseal (Stephens Scientific, Inc., Riverdale, NJ).
Protein localization (immunohistochemistry)
Serial sections (6 µm) of paraformaldehyde-fixed, paraffin-embedded tissue were deparaffinized in xylene and rehydrated through graded ethanols (70%, 90%, 95%, 100%, 5 min each) and washed in TBS, pH 7.4. For antigen retrieval, sections were incubated in 10 mM citrate buffer (Sorensen’s citrate, pH 6.0), and heated in a microwave oven for 12 min at high power and 10 min at medium-low power. To inhibit endogenous peroxidase, sections were incubated for 30 min in 1% hydrogen peroxide in TBS and washed three times in TBS. Nonspecific binding was blocked by incubation with 3% normal horse serum and 3% normal goat serum in TBS for 20 min. Sections were than incubated overnight at 4 C in a humidified chamber with a rabbit polyclonal antibody directed against human Ang2 (a generous gift from Dr. G. Yancopoulos, Regeneron Pharmaceuticals, Inc., Tarrytown, NY) (2.2 µg/ml), a rabbit polyclonal antibody directed against human VEGF (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) (0.4 µg/ml) (19, 20, 21), a mouse monoclonal antibody directed against human CD34 (Zymed Laboratories, Inc., South San Francisco, CA) (4 µg/ml) used as an endothelial cell marker, a mouse monoclonal antibody directed against human CD68 (DAKO Corp., Glostrup, Denmark) (7.4 µg/ml) used as a macrophage cell marker, and a mouse monoclonal antibody directed against cytokeratin AE1/AE3 (Chemicon Inc., Temecula, CA) (1.6 µg/ml) used as an epithelial cell-specific marker to localize trophoblast cells.
A murine monoclonal antibody of the same IgG isotype, directed against the gp120 protein (20 µg/ml) and normal mouse serum (20 µg/ml), was used as control. In addition, a second negative control for Ang2 primary antibody was performed by preincubation of the antibody with Ang2 peptide (Regeneron Pharmaceuticals, Inc.) for 30 min at room temperature. After incubation, sections were washed three times in 1x TBS and incubated with biotinylated horse antimouse and antirabbit IgG (Vector Laboratories, Inc., Burlingame, CA). Sections were washed again with 1x TBS and incubated with avidin-biotin-peroxidase complex (Vector Laboratories, Inc.) for 30 min and washed with TBS again. Antibody binding was detected by liquid DAB (substrate kit, Zymed Laboratories, Inc.).
The appearance of brown reaction product was observed by light microscopy. The sections were lightly counterstained with hematoxylin, dehydrated in graded alcohols, cleared with xylene, and mounted with Cytoseal (Stephens Scientific, Inc., Riverdale, NJ).
Western blot
For Western blotting, 10 µg extracted protein were resuspended in 2x SDS sample buffer, boiled for 10 min, followed by centrifugation for 1 min at 10,000 rpm. Equal amounts of total protein (10 µg) were separated on 4–12% Bis-Tris gel by electrophoresis (Invitrogen, Carlsbad CA) and transferred to polyvinylidene difluoride (Amersham Pharmacia Biotech, Piscataway, NJ) for 1 h at room temperature. Membranes were blocked with 5% nonfat milk in 1x Tween (T)-TBS (10 mM Tris-HCl, pH 8.0; 150 mM NaCl; 0.1% Tween 20) for 1 h at room temperature and washed three times with 25 ml 1x T-TBS at room temperature for 5 min each. Membranes were incubated with Ang2 antibody (1:10,000) (Regeneron Pharmaceuticals, Inc.), the same antibody used for immunocytochemistry, at 4 C overnight with rocking. Membranes were washed three times for 5 min with 15 ml 1x T-TBS, and incubated with antirabbit-horseradish peroxidase secondary antibody (1:2,000) (BD PharMingen, San Diego, CA) diluted in 1x T-TBS and 5% nonfat milk for 1 h at room temperature. Membranes were washed again 3 times with 15 ml 1x T-TBS at room temperature for 5 min each, and antibody reaction was detected using the electrochemiluminescence detection kit (Amersham Pharmacia Biotech). Chemiluminescence was detected on x-ray film following 1-min exposure.
ELISA
Free VEGF-A, Ang1, and Ang2 protein levels were determined using two-site ELISA (22).
Statistical analysis
Statistical analysis was performed by SPSS-PC (23), using the paired t test. Probability of less than 0.05 was considered statistically significant. Changes in molecular markers over time were studied using linear and log-linear regression, with the amount of marker, or log of the amount, regressed against time. Calculations were carried out in Data Desk (24).
Results
VEGF-A, Ang1, and Ang2 mRNA expression
Using qPCR analysis, we were able to measure VEGF-A, Ang1, and Ang2 mRNA expression levels in all placental tissue studied during the first (n = 7), second (n = 7) and third (n = 10) trimesters of normotensive pregnancies and third (n = 5) trimester of pregnancies complicated by preeclampsia. Trend analysis demonstrated that Ang1 mRNA expression increased significantly by 2.8%/wk during pregnancy (Ang1 = 0.50 + 0.028 wk, r2 = 21%, P = 0.034) (Fig. 2A). While there were no significant changes (2.5%/wk) in VEGF-A mRNA levels across pregnancy (VEGF = 3.31 + 0.025 wk, r2=1%) (Fig. 2A), Ang2 mRNA expression decreased significantly by 3.5%/wk during pregnancy (Ang2 = 616 x 10(-0.035 wk), r2 = 51%, P = 0.0003) (Fig. 2B). Interestingly, Ang2 mRNA expression was 400-fold higher than Ang1 mRNA and 100-fold higher than VEGF-A mRNA in the first trimester and declined to 20-fold and 7-fold, respectively, in the third trimester (Table 1).
|
|
, Fig. 2C). VEGF-A, Ang1, and Ang2 mRNA localization
VEGF-A mRNA was localized by in situ hybridization principally in the villous cytotrophoblast (Langhans’ cells) and syncytiotrophoblast, in perivascular cells and stromal macrophages (Hofbauer cells) in the immature intermediate villi during the first and second trimesters, and mature intermediate and terminal villi during the third trimester of normotensive pregnancy (Fig. 3, Table 2). In contrast, Ang1 and Ang2 mRNA were localized principally in the villous syncytiotrophoblast and in perivascular cells and stromal macrophages in the immature intermediate villi during the first and second trimesters, and mature intermediate and terminal villi during the third trimester of normotensive pregnancy (Figs. 4 and 5 and Table 2). In preeclampsia, VEGF-A, Ang1, and Ang2 mRNA were localized principally in the syncytiotrophoblast and the fibrous stroma within the mature intermediate and terminal villi (Fig. 6, Table 2).
|
|
|
|
|
Ang2 protein levels decreased throughout pregnancy by 4.7%/wk (Ang2 = 8.66 x 10(-0.047 wk), r2 = 64%, P = 0.0001) (Fig. 2B). A 99% correlation was found between Ang2 mRNA expression and Ang2 peptide levels. Using Z-statistics to compare ratios, there is a 61% probability that the two ratios are equal.
Using Western blot analysis, the Ang2 antibody recognized a single specific band at 65 kDa corresponding to the recombinant Ang2 protein. Ang2 protein levels decreased throughout pregnancy (Fig. 2D).
In placentas from women with preeclampsia, there were not significant changes (3.0-fold increase) in Ang2 protein levels. Ang1 and VEGF-A protein levels were undetectable.
VEGF-A and Ang2 protein localization
Immunohistochemical staining for human VEGF-A demonstrated localization of the protein in both the cytotrophoblast (Langhans’ cells) and syncytiotrophoblast within the chorionic villi, endothelial cells, and some stromal macrophages (Hofbauer cells) (Fig. 3 and Table 1). Ang2 protein was detected principally in the villous syncytiotrophoblast and endothelial cells and some stromal macrophages (Fig. 5). These results are consistent with our mRNA localization analysis.
Endothelial cell, macrophage cell and cytokeratin localization was confirmed by staining with an antibody against human CD34 (Fig. 6D), CD68 (Fig. 4F), and cytokeratin AE1/AE3 (Fig. 4E), respectively. No immunohistochemical staining was observed in the controls (Figs. 3H and 4H). A reliable antibody against human Ang1 was not available for immunohistochemistry during the study period.
Discussion
Elucidation of the quantification and pattern of expression of the genes encoding VEGF-A, Ang1 and Ang2 has provided insights into the factors that regulate the development of vasculogenesis in the human placenta and the transition to a branching and subsequently nonbranching angiogenic phenotype (Fig. 1). These data also provide a plausible explanation for the mechanism by which differences in vascular development between the normoxic and hypoxic (preeclamptic) placenta occur.
Figure 1 implies that the transition from a vasculogenic to angiogenic phenotype is a reflection of the pattern of vascular endothelial-specific growth factor gene expression, likely acting via an autocrine/paracrine mechanism. When we analyzed quantitatively the pattern of these three vascular-specific growth factor genes during human pregnancy, we found similar increases in expression of VEGF-A and Ang1 (1–4%) in a linear pattern (Fig. 2A). These observations are consistent with previous morphologic studies of the maturation of the placental vasculature, which continues until term (25). These changes are paralleled by the known roles of VEGF-A in vasculogenesis and angiogenesis: it induces vasculogenesis and also is essential in the continued maturational processes involved in angiogenesis (26), i.e. branching and nonbranching angiogenesis. In these latter (angiogenic) processes, Ang1 also plays a central role in blood vessel stabilization induced by VEGF-A.
Interestingly, the dramatic 10-fold decrease in Ang2 mRNA expression throughout pregnancy in a log-linear pattern (Fig. 2B) has implications for the vasculogenic-angiogenic transformation: in the first trimester, when Ang2 levels are very high relative to VEGF-A and Ang1 (100- and 400-fold, respectively), Ang2 likely permits the fetal blood vessels to undergo remodeling. These changes allow the placenta to increase in size to meet the increased oxygen and metabolic demands of the growing fetus. The progressive decrease in Ang2 beginning at the end of the second trimester in a log-linear pattern may prevent the villous vessels from undergoing destabilization, allowing them to acquire a more plastic form that permits the transformation from branching to nonbranching angiogenesis (Fig. 1).
Two molecular mechanisms for Ang2-mediated blood vessel remodeling have been proposed. First, in a Tie2-dependent mechanism, Ang2 antagonizes Ang1-mediated phosphorylation of the Tie 2 receptor (9). Ang1 is responsible for blood vessel stabilization and maturation by recruitment of pericytes and smooth muscle cells (1). The vasculopathy phenotype in Tie2-deficient (27) and Ang1-deficient (6) mice is exacerbated in Ang2 embryonic lethal transgenic mice (9). This vasculopathy is characterized by endothelial cell detachment and disrupted blood vessel formation (9). Manipulation of Ang2 levels by direct placental transfection induces a similar phenotype in a murine model (Geva, E., M. P. Pallavicini, D. G. Ginzinger, D. H. Moore, P. C. Ursell, and R. B. Jaffe, unpublished observations). Second, a Tie2-independent mechanism based upon Ang2 modulation of endothelial cell adhesiveness has been suggested recently (28). Those data demonstrated that Ang2 binds to 
Vß5 and vitronectin integrins, which mediate migration and spreading of both endothelial and nonendothelial cells. The expression of Vß5 and vitronectin integrins by placental trophoblast cells (29, 30, 31) also is consistent with a Tie2-independent mechanism. Moreover, the decrease in Ang2 levels throughout gestation, as levels of VEGF-A and Ang1 are increasing, may explain why the placental villous vessels do not undergo regression even at term, consistent with previous morphologic studies in the human placenta (25).
The tight correlation between Ang2 mRNA and peptide concentrations indicates that regulation of placental vascular development in both normotensive and preeclamptic pregnancies occurs at the transcriptional, and not translational, level.
Our methodologic approach differs from those of Dunk et al. (32) and Goldman-Wohl et al. (33), who also studied expression of Ang1 and Ang2 in human placentas. Because we were unable to detect any protected species corresponding to Ang1 using up to 10 µg of total RNA with ribonuclease protection assays, as described by Dunk et al. (32), we used qPCR, as we predicted that we would be able to quantify even relatively small differences in mRNA expression between the various placental samples during pregnancy. This proved to be the case. Using RT-PCR, Dunk et al. (32) were able to detect Ang1 mRNA in only two of the nine third trimester placental samples studied. However, their RT-PCR results are in contrast to their in situ hybridization data, which demonstrate high levels of Ang1 mRNA expression. Furthermore, using a ribonuclease protection assay, there were no significant differences between the levels of Ang2 mRNA expression in third trimester placentas from normal pregnancy and those with severe IUGR/small for gestational age. This is likely related to Ang2 expression already having been significantly decreased. However, one might expect to find increased Ang2 expression in severe IUGR/small for gestational age because of hypoxia. In addition, third trimester placentas in the study by Dunk et al. (32) from patients who delivered at 27–36 wk gestation, were included in their normal placental population, although these premature deliveries may not have been associated with normal pregnancy.
Our expression and localization data are concordant with those of Goldman-Wohl et al. (33) in part. However, they found Ang2 mRNA expression only during the first trimester of pregnancy in syncytiotrophoblast cells using Northern blot and in situ hybridization analysis, perhaps related to the lower sensitivity of these techniques than qPCR.
For mRNA quantitation, several internal (constitutive) controls, such as glyceraldehyde 3-phosphate dehydrogenase (GAPDH), generally are used, assuming that gene expression does not change in relative abundance during development. However, when we analyzed GAPDH in our placental samples using qPCR, we observed differential expression during pregnancy (data not shown). These results have been described previously in vitro and in vivo in vascular endothelial and nonendothelial cells, and may be related to stress (e.g. oxidative stress, cellular proliferation), various growth factors (e.g. insulin, IL-2, platelet-derived growth factor), or pregnancy, as GAPDH expression during pregnancies in other species also does not remain constant (34). Therefore, for this study, it was apparent that GAPDH was an inappropriate control. When we studied GUS as a normalizing control, we were unable to measure any difference in expression between our placental samples during pregnancy, as well as in placentas from normotensive compared with hypertensive pregnancies. Therefore, we concluded that GUS was a more appropriate control for our vasculogenesis/angiogenesis study.
VEGF-A mRNA and peptide were localized by in situ hybridization and immunohistochemistry mainly in the villous cytotrophoblast, and also in syncytiotrophoblast and perivascular cells/endothelial cells, as has been described previously (35, 36), whereas Ang1 and Ang2 mRNA and protein were localized principally in the villous syncytiotrophoblast and perivascular/endothelial cells. VEGFR-1/Flt-1 mRNA and peptide have been localized in cytotrophoblast and syncytiotrophoblast cells (36, 37, 38), whereas VEGFR-2/KDR mRNA is expressed only in villous endothelial cells (36, 37, 38). Moreover, Tie2 (the receptor for both Ang1 and Ang2) mRNA and peptide have also been shown to be localized in cytotrophoblast and syncytiotrophoblast cells in addition to the villous endothelial cells (32, 33, 36). Therefore, there may be different autocrine/paracrine roles for these molecules in placental angiogenesis and trophoblast cell function. Hypoxia, as in preeclampsia, can induce increased expression of VEGF-A (3.1-fold) and Ang2 (1.5-fold) by trophoblast cells. Thus, the extent of oxygen saturation in maternal and fetal blood can regulate fetal vasculature, acting via the intravillous space.
Preeclampsia is a disease that adversely affects 3–5% of first pregnancies and is the leading cause of maternal death. In its severest form, the signs and symptoms occur in the second trimester and contribute significantly to maternal morbidity and fetal prematurity (13, 14). The principal cause of preeclampsia is still unknown, although it is likely that the disease is related to the presence of placental tissue.
During placentation, the cytotrophoblast has a unique ability to invade maternal blood vessels to replace the maternal endothelium within the spiral artery segments as far as the myometrium, and to remodel the tunica media of these arteries. To accomplish this, differentiating cytotrophoblast stem cells lose their epithelial phenotype and transform their cell-cell adhesion molecule phenotype (6ß4 and E-cadherin) to that of endothelial cells (VE-cadherin, platelet endothelial cell adhesion molecule-1, and Vß3) that are expressed during vasculogenesis and angiogeneis (39). The ability of cytotrophoblasts to proliferate and differentiate along the invasive pathway is regulated by oxygen tension: hypoxia stimulates cytotrophoblast proliferation, and relatively high oxygen tension promotes cytotrophoblast differentiation (e.g. expression of integrin 1) (40). Hypoxia in women with preeclampsia leads to increased trophoblastic proliferation and increased villous capillary growth and branching. The extensively branched capillaries are responsible for the characteristic deformation of the shape of the terminal villi (25, 41).
Analysis of placentas from patients who underwent pregnancy complicated by severe preeclampsia (with IUGR) revealed increases in VEGF-A (3-fold), P value of 0.006, and Ang2 (1.5-fold, not significant) expression compared with the placentas from normotensive women, whereas no change in Ang1 expression was found. This may play a pivotal role in placental angiogenesis (nonbranching) and may serve as a mechanism to compensate for the hypoxia occurring in preeclampsia. We hypothesize that this increase may be related to low oxygen tension, as hypoxia up-regulates both molecules. Our findings are consistent with the observation that hypoxia is a primary inducer of angiogenesis via activation of hypoxia-inducible transcription factor-1, which functions as a master switch to induce expression of several angiogenic factors, including VEGF-A, nitric oxide synthase, platelet-derived growth factor, and Ang2 (42, 43). Both hypoxia and VEGF-A were found to up-regulate Ang2 mRNA expression, whereas neither hypoxia nor VEGF-A induced Ang1 or Tie2 expression (42, 43). Treatment in vitro of a choriocarcinoma cell line [BeWo, which expresses both VEGF receptors (Flt-1 and KDR)] with VEGF-A165 did not affect Ang2 mRNA expression in a time- or dose-dependent manner (Geva, E., and R. B. Jaffe, unpublished observations), suggesting a direct effect of hypoxia on Ang2 transcription; this may maintain the vessels prone to be affected by angiogenic peptides.
In summary, our data suggest that VEGF-A, Ang1 and Ang2 gene expression in the human placenta during pregnancy may have critical autocrine/paracrine effects on vasculogenic-angiogenic (branching to nonbranching angiogenesis) transformation. The tight correlation between mRNA and protein concentrations of these vascular endothelial-specific growth factors indicates that regulation of placental vascular development in both normotensive and preeclamptic pregnancies occurs at the genetic level.
Acknowledgments
We thank John S. Rudge and George D. Yancopoulos (Regeneron Pharmaceuticals, Inc.) for helpful discussions and gifts of reagents; Donna Hylton (Regeneron Pharmaceuticals, Inc.) for performing the ELISAs; Christina Tzagarakis-Foster and Laurence Lamarcq (UCSF) for their invaluable suggestions and assistance; and Jean Perry (UCSF) for her assistance in obtaining placental tissues.
Footnotes
This work was supported, in part, by a Serono Foundation Fellowship in Reproductive Endocrinology (to E.G.).
Abbreviations: Ang, Angiopoietin; Ct, threshold cycle; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GUS, ß-glucuronidase; IUGR, intrauterine growth restriction; qPCR, quantitative PCR; SSC, standard saline citrate; TAMRA, A-6-carboxy tetramethyl rhodamine-3'; TBS, Tris-buffered saline; T-TBS, Tween-TBS; VEGF, vascular endothelial growth factor.
Received February 8, 2002.
Accepted May 28, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  M. Plaisier, E. Streefland, P. Koolwijk, V.W.M. van Hinsbergh, F.M. Helmerhorst, and J.J.H.M. Erwich Angiogenic Growth Factors and Their Receptors in First-Trimester Human Decidua of Pregnancies Further Complicated By Preeclampsia or Fetal Growth Restriction Reproductive Sciences, September 1, 2008; 15(7): 720 - 726. [Abstract] [PDF]
|
|
|
|
 |
|
 Y. Xue, R. Cao, D. Nilsson, S. Chen, R. Westergren, E.-M. Hedlund, C. Martijn, L. Rondahl, P. Krauli, E. Walum, et al. FOXC2 controls Ang-2 expression and modulates angiogenesis, vascular patterning, remodeling, and functions in adipose tissue PNAS, July 22, 2008; 105(29): 10167 - 10172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Ishimoto, K. Minegishi, T. Higuchi, M. Furuya, S. Asai, S. H. Kim, M. Tanaka, Y. Yoshimura, and R. B. Jaffe The Periphery of the Human Fetal Adrenal Gland Is a Site of Angiogenesis: Zonal Differential Expression and Regulation of Angiogenic Factors J. Clin. Endocrinol. Metab., June 1, 2008; 93(6): 2402 - 2408. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ishimoto, M. O. Muench, T. Higuchi, K. Minegishi, M. Tanaka, Y. Yoshimura, and R. B. Jaffe Midkine, a Heparin-Binding Growth Factor, Selectively Stimulates Proliferation of Definitive Zone Cells of the Human Fetal Adrenal Gland J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 4050 - 4056. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ishimoto, D. G. Ginzinger, T. Matsumoto, Y. Hattori, M. Furuya, K. Minegishi, M. Tanaka, Y. Yoshimura, and R. B. Jaffe Differential Zonal Expression and Adrenocorticotropin Regulation of Secreted Protein Acidic and Rich in Cysteine (SPARC), a Matricellular Protein, in the Midgestation Human Fetal Adrenal Gland: Implications for Adrenal Development J. Clin. Endocrinol. Metab., August 1, 2006; 91(8): 3208 - 3214. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Pau, I. Alonso-Muriel, R. Gomez, E. Novella, A. Ruiz, J. A. Garcia-Velasco, C. Simon, and A. Pellicer Plasma levels of soluble vascular endothelial growth factor receptor-1 may determine the onset of early and late ovarian hyperstimulation syndrome Hum. Reprod., June 1, 2006; 21(6): 1453 - 1460. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ishimoto, D. G. Ginzinger, and R. B. Jaffe Adrenocorticotropin Preferentially Up-Regulates Angiopoietin 2 in the Human Fetal Adrenal Gland: Implications for Coordinated Adrenal Organ Growth and Angiogenesis J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1909 - 1915. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Padavala, N. Pope, P. Beker, and I. Crocker An Imbalance Between Vascular Endothelial Growth Factor and its Soluble Receptor in Placental Villous Explants of Intrauterine Growth-Restricted Pregnancies Reproductive Sciences, January 1, 2006; 13(1): 40 - 47. [Abstract] [PDF]
|
|
|
|
 |
|
 J. R. Miles, C. E. Farin, K. F. Rodriguez, J. E. Alexander, and P. W. Farin Effects of Embryo Culture on Angiogenesis and Morphometry of Bovine Placentas During Early Gestation Biol Reprod, October 1, 2005; 73(4): 663 - 671. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Mondon, T.-M. Mignot, R. Rebourcet, H. Jammes, J.-L. Danan, F. Ferre, and D. Vaiman Profiling of oxygen-modulated gene expression in early human placenta by systematic sequencing of suppressive subtractive hybridization products Physiol Genomics, June 16, 2005; 22(1): 99 - 107. [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]
|
|
|
|
|
|
J. R. Miles, C. E. Farin, K. F. Rodriguez, J. E. Alexander, and P. W. Farin Angiogenesis and Morphometry of Bovine Placentas in Late Gestation from Embryos Produced In Vivo or In Vitro Biol Reprod, December 1, 2004; 71(6): 1919 - 1926. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. D. Albrecht, V. A. Robb, and G. J. Pepe Regulation of Placental Vascular Endothelial Growth/Permeability Factor Expression and Angiogenesis by Estrogen during Early Baboon Pregnancy J. Clin. Endocrinol. Metab., November 1, 2004; 89(11): 5803 - 5809. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Papazoglou, G. Galazios, M. I. Koukourakis, I. Panagopoulos, E. N. Kontomanolis, K. Papatheodorou, and E. Maltezos Vascular endothelial growth factor gene polymorphisms and pre-eclampsia Mol. Hum. Reprod., May 1, 2004; 10(5): 321 - 324. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Thadhani, W. P. Mutter, M. Wolf, R. J. Levine, R. N. Taylor, V. P. Sukhatme, J. Ecker, and S. A. Karumanchi First Trimester Placental Growth Factor and Soluble Fms-Like Tyrosine Kinase 1 and Risk for Preeclampsia J. Clin. Endocrinol. Metab., February 1, 2004; 89(2): 770 - 775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Tsatsaris, F. Goffin, C. Munaut, J.-F. Brichant, M.-R. Pignon, A. Noel, J.-P. Schaaps, D. Cabrol, F. Frankenne, and J.-M. Foidart Overexpression of the Soluble Vascular Endothelial Growth Factor Receptor in Preeclamptic Patients: Pathophysiological Consequences J. Clin. Endocrinol. Metab., November 1, 2003; 88(11): 5555 - 5563. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. O. Muench, J. V. Ratcliffe, M. Nakanishi, H. Ishimoto, and R. B. Jaffe Isolation of Definitive Zone and Chromaffin Cells Based upon Expression of CD56 (Neural Cell Adhesion Molecule) in the Human Fetal Adrenal Gland J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3921 - 3930. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) (n = 28) and Ang1 (
) (n = 27) mRNA expression relative to GUS in human placentas during pregnancy. B, Ang2 (n = 27) mRNA (
) expression relative to GUS and protein (
) (n = 20) levels in human placentas during pregnancy. C, VEGF-A, Ang1 and Ang2 mRNA expression in placentas from normotensive (n = 10) and severe preeclamptic (with IUGR) (n = 5) pregnancies. D, Representative Western blot analysis of Ang2 protein during pregnancy. A single major band at 65 kDa, corresponding to Ang2 wild-type (WT) protein, was detected. Ang2 protein levels decrease throughout gestation.
