|Home | About | Journals | Submit | Contact Us | Français|
Maternal endothelial dysfunction in preeclampsia is associated with increased soluble fms-like tyrosine kinase-1 (sFlt-1), a circulating antagonist of vascular endothelial growth factor and placental growth factor. Angiotensin II (Ang II) is a potent vasoconstrictor that increases concomitant with sFlt-1 during pregnancy. Therefore, we speculated that Ang II may promote the expression of sFlt-1 in pregnancy. Here we report that infusion of Ang II significantly increases circulating levels of sFlt-1 in pregnant mice, thereby demonstrating that Ang II is a regulator of sFlt-1 secretion in vivo. Furthermore, Ang II stimulated sFlt-1 production in a dose- and time-dependent manner from human villous explants and cultured trophoblasts but not from endothelial cells, suggesting that trophoblasts are the primary source of sFlt-1 during pregnancy. As expected, Ang II–induced sFlt-1 secretion resulted in the inhibition of endothelial cell migration and in vitro tube formation. In vitro and in vivo studies with losartan, small interfering RNA specific for calcineurin and FK506 demonstrated that Ang II–mediated sFlt-1 release was via Ang II type 1 receptor activation and calcineurin signaling, respectively. These findings reveal a previously unrecognized regulatory role for Ang II on sFlt-1 expression in murine and human pregnancy and suggest that elevated sFlt-1 levels in preeclampsia may be caused by a dysregulation of the local renin/angiotensin system.
The human placenta is principally a vascular organ that functions to achieve a physiological union of the maternal and fetal blood supplies. A major physiological role of the placenta is to develop an extensive vascular network allowing for nutrient, waste, and gas exchange between the maternal and fetal circulations. To accomplish this, the placenta produces a variety of angiogenic factors including vascular endothelial growth factor (VEGF) and placental growth factor.1 Cytotrophoblasts invade the maternal spiral arterioles and acquire properties of endothelial cells in a process termed pseudovasculogenesis.2 The balance between proangiogenic and antiangiogenic factors modulates this vascular remodeling. VEGF promotes angiogenesis through interaction with 2 high-affinity receptors, VEGF receptor-1 (VEGFR-1, also known as Flt-1) and VEGF receptor-2 (VEGFR-2, also known as KDR/Flk-1). The production of antiangiogenic factors increases toward the end of gestation, when continued placental growth is no longer needed. One such antiangiogenic factor is a soluble form of VEGFR-1 (sFlt-1) that binds to circulating VEGF and placental growth factor, thereby inhibiting their angiogenic activities.
It was first proposed in 1997 that low levels of VEGF in maternal decidua may be a contributory factor in preeclampsia because of increased levels of endogenous sFlt-1, which may antagonize the beneficial effects of VEGF in this disorder.3 Subsequently, a number of groups demonstrated the production of sFlt-1 from human placental explants1,4 and cultured cytotrophoblasts.5,6 Circulating levels of sFlt-1 increase significantly during the last 2 months of pregnancy7–9 but are present only in trace quantities in sera from nonpregnant women. Soluble Flt-1 is increased in the maternal circulation in preeclampsia before onset of clinical manifestations9,10 and correlates with severity of the disease.11 Over-expression of sFlt-1 in pregnant rats results in a preeclampsia-like phenotype.12 Furthermore, conditioned medium (CM) from human placental explants obtained from uncomplicated pregnancies induces in vitro angiogenesis, whereas CM from preeclamptic placentas inhibits angiogenesis.4,12 More importantly, removal of sFlt-1 from preeclamptic placenta CM promotes angiogenesis, suggesting that placental vascular growth may be restricted in preeclampsia because of an imbalance of pro- and antiangiogenic factors.1,4 However, the factors regulating the synthesis and secretion of sFlt-1 during pregnancy are largely unknown.
The presence of a local renin/angiotensin system has been detected in many tissues, including the brain, kidney, adrenal gland, and heart.13 In the placenta, the local generation of Ang II is responsible for the activation of Ang II type 1 (AT1) receptors present on trophoblasts.14,15 In mice, the renin/angiotensin system undergoes major changes in response to pregnancy.15 One of the most striking changes is a 5- to 10-fold rise in the levels of circulating prorenin and renin,15 which is associated with increased renin gene expression in placenta. As the increase in placental renin gene expression is concomitant with the increase in sFlt-1, we speculated that the production of Ang II in the placenta could regulate sFlt-1 expression. In this study, we report that in pregnant mice, infusion of Ang II induces sFlt-1 secretion, demonstrating that Ang II is a regulator of sFlt-1 secretion in vivo via AT1 receptor activation and calcineurin signaling.
HTR-8/SVneo cells, an immortalized line of human trophoblasts,16 were cultured in RPMI medium 1640 (Invitrogen, Grand Island, NY) supplemented with 5% FBS and 1% antibiotic–antimycotic solution. Human umbilical vein endothelial cells (HUVECs) were grown in endothelial growth medium (EGM)-2 medium (Cambrex). All cells were cultured at 37°C in a humidified incubator with a 5% CO2 atmosphere.
For sFlt-1 induction, HTR-8/SVneo cells were plated in 12-well plates at 1×105 cells per well overnight. The next day, cells were changed to serum-free medium and treated with Ang II for 4 days. To ensure that an adequate concentration of Ang II was maintained throughout the experiments, fresh Ang II was added each day. The cells and their CM were collected and centrifuged at room temperature for 15 minutes at 1000g. Cell pellets were collected for Western blotting. In some cases, cells were preincubated with 100 nmol/L losartan (Merck & Co Inc) or FK506 (Fujisawa Pharmaceutical, Osaka, Japan) for 30 minutes before addition of Ang II. To deplete sFlt-1 from CM, CM was first incubated with monoclonal anti-Flt-1 antibody (1:250 dilution) for 2 hours at 4°C before adding Protein-A beads overnight as previously described.4
Human placental tissue was collected as previously described.17,18 After dissection, 5 small fragments (15 to 20 mg wet weight) of placental villi were dissected from the placenta, teased apart, and placed on 24-well plates. Placental villous explants were cultured in phenol red–free DMEM medium (Sigma Ltd, Poole, UK) supplemented with 100 μg/mL streptomycin, 100 U/mL penicillin, and 0.2% BSA, in 5% CO2 at 37°C. Ang II was incubated with villous fragments for up to 72 hours at 37°C. For the inhibitor studies, losartan was added at 100 nmol/L for 30 minutes before stimulation with Ang II. After 24 hours, the conditioned media were collected and stored at −20°C for further analysis.
Soluble Flt-1 levels in sera and CM were measured using either a commercial kit (R&D Systems, Minneapolis, Minn) or an in-house ELISA.
Total RNA was isolated by TRIzol reagent as described previously.16 Superscript 1-step RT-PCR with Platinum Taq kit (Invitrogen) was used for reverse transcriptase-polymerase chain reaction (RT-PCR) according the protocol provided by the vendor. Flt-1 primer sequences for RT-PCR were as described19: sense primer, 5′-TTTGCATAGCTTCCAATAAAGTTG-3′; antisense primer, 5′-CATGACAGTCTAAAGTGGTGGAAC-3′. β-Actin was used as internal control and primer sequences as following: sense primer, 5′-GCTCTGGCTCCTAGCACCAT-3′; antisense primer, 5′-CCACCGATCCACACAGAGTAC-3′. RT-PCR products were run on 2% agarose gels and quantified using Bio-Rad Quantity One software. Flt-1 mRNA expression was represented by the ratio Flt-1/β-actin mRNA.
Following stimulation villous explant, tissue samples were homogenized using a PreCellys 24 homogenizer in 1 mL of TRIzol, and the RNA was extracted, further purified, and subjected to DNase-1 digestion on RNeasy mini columns (Qiagen). Total RNA (1 μg) was reverse transcribed with oligo-dT18 primers using the cDNA Synthesis Kit (Bioline) for 70 minutes at 48°C. Real-time PCR was performed using SYBR green PCR SensiMix (Quantace) on a Corbett Research Rotagene RG-3000 under the following conditions: 95°C, 10 seconds; 58°C, 15 seconds; 72°C, 15 seconds; and fluorescence was quantified after 85°C, 15 seconds, for 40 cycles. Triplicate cDNA samples and standards were amplified with primers specific for sFlt-1 (sense, 5′-AGGAGATGCTCCTCCCAAA-3′; antisense, 5′- GTGCAGGGATCCTCCAAAT-3′; 147-bp product) or β-actin, as described previously.20 The mean threshold cycle (CT) for each sample was compared against standard curves produced using 6 serial dilutions of a plasmid containing a C-terminal fragment of sFlt-1 or β-actin cDNA, and relative sFlt-1 expression was calculated following normalization to β-actin.
Proteins were analyzed by Western blots as described.16
The in vitro cell migration assay was preformed as described.21 Briefly, HUVECs (2.5×104 per well) were prepared with 500 μL of EGM-2 basal medium and added to the inserts of the 24-well plates. CM was used as a chemoattractant and added to the outside of the insert chamber in triplicate (BD Biosciences). After 24 hours of incubation at 37°C, migrated cells were stained with Quick-diff and counted. Four to 10 fields were counted and the mean recorded.
The endothelial cells in vitro tube formation assay was performed as described12 using 96-well plates coated with growth factor-reduced Matrigel (BD Biosciences) and incubated at 37°C for 30 minutes. HUVECs, at a density of 2×104 cells per well, were incubated with CM and plated on precoated 96-well plates and incubated at 37°C in a CO2 incubator. Images were taken after 4 hours of incubation, and the total tube length per field was measured using Metaphore software. Five to 10 fields were used to determine the mean tube length per field.
HTR cells (1.0×105) cells were transfected with 3 μg of small interfering RNA (siRNA) specific for calcineurin catalytic subunit α (CN) or nonspecific target (DHARMACON) using RNAiFect transfection reagent (Qiagen). After 48 hours posttransfection, cells were cultured in serum-free medium and treated with Ang II for 72 hours. Secreted sFlt-1 levels in cell culture were measured by ELISA.
C57BL/6J pregnant or nonpregnant mice (18 to 22 g; Harlan, Indianapolis, Ind) were anesthetized with sodium pentobarbital (50 mg/kg IP), and osmotic minipumps (Alzet model 2001; Alza, Palo Alto, Calif) were implanted subcutaneously in the nape of the neck. All animal studies were reviewed and approved by the Animal Welfare Committee, University of Texas Houston Health Science Center. Ang II was delivered at a rate of 1.5 mg/kg body weight per day into 10-day pregnant mice or nonpregnant female mice for 7 days.22–25 Control mice were infused with saline. For calcineurin pathway experiments, the calcineurin inhibitor FK506 (2 mg/kg body weight per day) was given daily by intraperitoneal injection 24 hours after the minipump was implanted until mice were euthanized. Mouse systolic blood pressure was measured using mouse tail-cuff (AD instruments, Colorado Springs, Colo).
All values are expressed as the mean±SEM. Statistical significance was determined by ANOVA test using GraphPad Prism software, and significance set at a probability value of P<0.05.
Circulating levels of sFlt-1 rise during the second half of pregnancy.6,8,19,26 To determine whether Ang II can stimulate sFlt-1 production in vivo, Ang II was infused into pregnant (gestation day 10) and nonpregnant mice and the levels of sFlt-1 measured in the maternal circulation at different time points. Ang II infusion resulted in significantly increased circulating levels of sFlt-1 in the pregnant mice by day 5 (gestation day 15) and sFlt-1 levels remained significantly above the saline-treated pregnant animals (Figure 1). In contrast, Ang II infusion in nonpregnant mice did not result in increased levels of sFlt-1 in the circulation or adverse outcome. These results provide the first direct evidence for the ability of Ang II to stimulate an increase in sFlt-1 secretion in the maternal circulation during normal pregnancy, a finding consistent with the view that during pregnancy sFlt-1 is primarily derived from trophoblasts.4
As infusion of Ang II induced the release of sFlt-1 in pregnant mice, we investigated the ability of Ang II to stimulate the release of sFlt-1 from normal human placenta. Placental villous explants were treated with increasing concentrations of Ang II (10 to 100 nmol/L) for 48 and 72 hours, and the CM was analyzed by ELISA. Ang II induced a concentration- and time-dependent increase in the levels of sFlt-1 (Figure 2A). Pretreatment of villous explants with 100 nmol/L losartan, an AT1 receptor specific antagonist, blocked Ang II–induced release of sFlt-1 observed at 72 hours (Figure 2B). Consistent with these findings, real-time PCR revealed a ≈3-fold increase in sFlt-1 transcripts after 24 hours of Ang II stimulation, and this was attenuated by pretreatment with losartan (Figure 2C), indicating that Ang II via AT1 receptor stimulates an increase of sFlt-1 mRNA abundance in human placenta.
Previous reports have shown that sFlt1 is detected in the trophoblasts and endothelium of placenta.4,27,28 To identify the cell types responsible for Ang II–mediated sFlt-1 release in the placenta, levels of sFlt-1 were measured in endothelial cells and trophoblasts CM following Ang II stimulation. Interestingly, Ang II induced a time- and concentration-dependent increase in sFlt-1 in an immortalized human trophoblast line (HTR-8/SVneo cells) but not in HUVECs (Figure 3A and 3B), demonstrating that Ang II only induces sFlt-1 expression in trophoblast, which is inhibited by losartan (Figure 3C through through3F).3F). The increase in the level of secreted sFlt-1 was accompanied by an increase in the abundance of sFlt-1 protein and mRNA levels (Figure 3D through through3G).3G). These results indicate that Ang II mediated AT1 receptor activation results in an increase in the synthesis of sFlt-1 by human trophoblasts but not in endothelial cells.
Calcineurin is among many signaling molecules that may function downstream of AT1 receptor activation. To test the involvement of calcineurin in sFlt-1 production, pregnant mice were infused with Ang II with or without daily subcutaneous injection of FK506, a well-known selective calcineurin inhibitor.29,30 As shown in Figure 4A, the increased sFlt-1 secretion induced by Ang II at gestation days 15 and 17 was completely inhibited by FK506, indicating that the induction of sFlt-1 by Ang II is mediated through calcineurin signaling in vivo. We also found that FK506 is able to attenuate the elevated systolic blood pressure and reduced fetal growth resulting from Ang II infusion (Table). Consistently, we showed that sFlt-1 induction by Ang II in human trophoblasts was inhibited by siRNA specific for calcineurin catalytic subunit α (Figure 4B). As shown in the inset of Figure 4B, siRNA specific for calcineurin catalytic subunit α successfully reduced calcineurin α protein levels (Figure 4B, inset) in human trophoblasts. Similarly, we found that Ang II–mediated induction of sFlt-1 secretion and sFlt-1 protein and mRNA expression in cultured human trophoblasts was inhibited by FK506 (Figure 4C and 4D). These results indicate that calcineurin signaling functions downstream of AT1 receptor activation and is essential for Ang II–mediated stimulation of sFlt-1 synthesis and secretion from human trophoblasts.
It has been shown that placental-derived sFlt-1 is responsible for reduced angiogenesis.4 To determine whether the sFlt-1 secreted by trophoblasts is biologically active, human trophoblast cells were treated with Ang II, and the effect of CM on endothelial cell migration and tube formation was assessed. Trophoblast CM induced endothelial cell migration and tube formation; however, CM from trophoblast treated with Ang II produced 50% less migration (Figure 5B) and reduced tube formation (Figure 5C). CM from trophoblasts treated with Ang II in the presence of losartan or FK506 induced similar levels of endothelial cell migration and tube formation to that seen with CM of untreated trophoblasts. These results indicate that blockade of AT1 receptor or calcineurin signaling overrode the inhibitory properties of the CM from Ang II–treated trophoblasts (Figure 5). Moreover, removal of sFlt-1 from Ang II–treated trophoblast CM using anti–sFlt-1 anti-body (Figure 5A) eliminated the antimigratory (Figure 5B) and antiangiogenic activity (Figure 5C), proving that sFlt-1 accounted for the reduced angiogenic activity seen in CM from Ang II–treated trophoblasts.
The regulation of angiogenesis in pregnancy is tightly controlled; however, the molecular mechanism of sFlt-1 induction in pregnancy is unknown. Here we report that Ang II stimulates an increase in circulating sFlt-1 in pregnant mice, human placental villous explants, and trophoblasts but not from endothelial cells. Ang II induces sFlt-1 release via the AT1 receptor and calcineurin pathway. Furthermore, we demonstrate that Ang II–induced sFlt-1 reduces the ability of trophoblast-derived CM to promote angiogenesis. Taken together, our findings identify a previously unrecognized regulatory role for Ang II in the regulation of sFlt-1 expression during pregnancy.
A number of groups have shown that a complete renin/angiotensin system is present in the placenta of humans and other mammals.15,31 Analysis of pregnant mice revealed that renin gene expression is developmentally regulated during normal pregnancy, achieving maximal values near the end of gestation.15 Ang II is also increased in the maternal circulation during normal human pregnancy.32–34 Using pregnant mice, villous explants, and human trophoblasts, we provide in vivo and in vitro evidence for Ang II as a positive modulator of sFlt-1 synthesis and secretion via AT1 receptor activation. These results imply that the local renin angiotensin system is intact in placental villous explants and is capable of generating endogenous Ang II and activating AT1 receptors to induce sFlt-1 synthesis and secretion. Furthermore, we showed that FK506, a calcineurin specific inhibitor,24–25 inhibited Ang II–mediated sFlt-1 production in pregnant mice and by cultured human trophoblasts. Moreover, siRNA specific for calcineurin subunit α mRNA completely abolished the Ang II–induced sFlt-1 secretion. Thus, the results obtained from both in vivo and in vitro experiments using FK506 or siRNA specific for calcineurin subunit α suggest that calcineurin signaling is required for Ang II–mediated sFlt-1 induction in trophoblasts during pregnancy. In other experiments, we have shown that Ang II treatment of human trophoblasts results in the activation of nuclear factor of activated T cells (NFAT), a well-known transcription factor and substrate of calcineurin. Ang II–mediated activation of NFAT is blocked by both FK506 and cyclosporin A (see the online data supplement, available at http://circres.ahajournals.org). Our earlier studies have shown that Ang II is capable of inducing NFAT nuclear translocalization and activation in human trophoblasts.16 Altogether, our results suggest that Ang II–induced synthesis and secretion of sFlt-1 involve AT1 receptor activation, calcineurin signaling, NFAT nuclear translocation, and the transcriptional activation of the Flt-1 gene.
Toward the end of gestation, continued placental vascular development is terminated, a process accomplished in part through the production of antiangiogenic factors. Multiple groups have demonstrated that sFlt-1, an antagonist to the action of VEGF and placental growth factor, is secreted from cytotrophoblasts in late pregnancy.6,8,19,26 The circulating levels of sFlt-1 are relatively low during the first trimester, begin to rise during the second trimester, and achieve relatively high levels during the third trimester. Thus, we propose that increased production of Ang II in the placenta late in the pregnancy stimulates production of antiangiogenic factors such as sFlt-1 to reduce continued placenta vascular development. We have provided the direct evidence that increased trophoblast sFlt-1 secretion by Ang II functions to antagonize angiogenesis as was earlier reported for preeclamptic CM.4 However, the biological significance of Ang II–mediated sFlt-1 secretion during pregnancy requires further investigation.
It is well known that serum concentrations of sFlt-1 are significantly higher in women with preeclampsia.35–39 However, Ang II levels are not elevated in women with preeclampsia. Thus, the following question remains: what factors contribute to the increased sFlt-1associated with preeclampsia? We40–42 and others43–47 have shown that autoantibodies present in the serum of women with preeclampsia are capable of activating AT1 receptors. In addition, these autoantibodies are able to induce the synthesis and secretion of plasminogen activator inhibitor-1, a feature that may also contribute to the shallow trophoblast invasion associated with preeclampsia.40 In this study, infusion of Ang II resulted in an increased arterial blood pressure and induced restricted fetal growth in pregnant mice (Table), which are 2 hallmark features of preeclampsia. We speculate that the AT1 receptor activating autoantibodies associated with preeclampsia may be responsible for the increased sFlt-1 observed in this disorder.
Sources of Funding This work was supported by NIH grants HL076558 (to Y.X.) and HD34130 (to R.E.K.). This work was also supported by grants from the British Heart Foundation, European Union, and Medical Research Council (to A.A.). S. Ahmad, P.W.H., and A.A. belong to the European Vascular Genomics Network, a Network of Excellence supported by the European Community's Sixth Framework Programme for Research Priority 1 Life Sciences, Genomics and Biotechnology for Health (contract LSHM-CT-2003-503254).