lable at ScienceDirect

Placenta 31 (2010) 448e455

Contents lists avai

Placenta

journal homepage: www.elsevier .com/locate/placenta

Expression of AT1R, AT2R and AT4R and Their Roles in ExtravillousTrophoblast Invasion in the Human

P.J. Williams a, H.D. Mistry a, B.A. Innes b, J.N. Bulmer b, F. Broughton Pipkin a,*

a School of Clinical Sciences, University of Nottingham, City Hospital Nottingham, Nottingham NG5 1PB, United Kingdomb Institute of Cellular Medicine, Newcastle University, Newcastle-upon-Tyne, Tyne and Wear NE2 4HH, United Kingdom

a r t i c l e i n f o

Article history:Accepted 23 February 2010

Keywords:ReceptorAngiotensinType 1Type 2Placenta

* Corresponding author. Tel.: þ44 1158 2 31892; faE-mail address: Fiona.broughton-pipkin@nottingha

0143-4004/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.placenta.2010.02.014

a b s t r a c t

The placental renin-angiotensin system (RAS) is active from early pregnancy and may have a role inplacentation. Angiotensin II (AngII) acts via binding to receptor types AT1R and AT2R. Recently smallerpeptide members of the angiotensin family have been recognised as having biological relevance.Angiotensin (3-8) (AngIV) has a specific receptor (AT4R) and evokes hypertrophy, vasodilatation andvascular inflammatory response. The aim of this study was to characterise placental expression of AT1R,AT2R and AT4R, and to determine whether AngII and AngIV regulate extravillous trophoblast (EVT)invasion, apoptosis and proliferation. Placental samples were obtained from women undergoing electivesurgical termination of pregnancy (TOP) at 8e10 weeks gestation (early TOP), 12e14 weeks gestation(mid TOP) or at delivery following normal pregnancy or with pre-eclampsia (PE). Immunohistochemistryand qRT-PCR were performed to determine placental mRNA and protein expression of AT1R, AT2R andAT4R at all gestational ages. EVT invasion following culture with AngII or AngIV was assessed in earlyplacental tissue using Matrigel invasion assays. Invasion was assessed on day 6 of culture and placentalexplants were harvested for immunohistochemical analysis of apoptosis and proliferation. The resultsfrom qRT-PCR and immunohistochemistry showed placental AT1R expression which did not vary withgestation. The highest levels of expression of AT2R were found in early and mid TOP placentae comparedto term pregnancy. Expression of AT4R was increased in term placentae, with a significant reduction inPE placentae. Moreover, culture with AngIV or AngII increased EVT invasion from placental explants,which showed increased trophoblast proliferation and reduced apoptosis. This study has characterisedexpression of AT4R and AT1R and AT2R in human placenta throughout normal pregnancy and in PE. BothAngIV and AngII may play an important role in normal pregnancy.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

The circulating renin-angiotensin system (RAS) is important forregulation of blood pressure and electrolyte balance. Until recentlyAngII was thought to be the major bioactive peptide, howeverrecently it has emerged that the shorter peptides, produced byaminopeptidase mediated cleavage of AngII, including angiotensin(3-8) (AngIV), also have roles in regulating cardiovascular function.There are two major angiotensin receptors, angiotensin II receptor1 (AT1R) and angiotensin II receptor 2 (AT2R), which have similarbinding affinities for AngII [1]. AngII mediates most of its effects viabinding with AT1R, ultimately triggering vasoconstriction, prolif-eration, angiogenesis or inflammation [1]. AT2R is predominantly

x: þ44 1158 2 31908.m.ac.uk (F. Broughton Pipkin).

All rights reserved.

expressed in fetal tissues [2], and AngII binding increases apoptosis,causes vasodilation and is involved in fetal tissue development [3].

AngIV appears to mediate various effects in different tissues viabinding with high affinity to its specific receptor, the AT4R [4].AngIV can also bind to AT1R and AT2R with low affinity. The activesite of AT4R has been identified to be an insulin-regulatedaminopeptidase (IRAP) [5], also known as cysteine aminopepti-dase, oxytocinase or placental leucine aminopeptidase (P-LAP).AT4R expression has been localised in both endothelial [6] andsmooth muscle cells [7] indicating physiological roles in regu-lating blood flow. AngIV can increase blood flow via a mechanismmediated by AT4R and nitric oxide [8]. Due to its localization inextravillous trophoblast (EVT) Ino et al. [9] suggested that thisreceptor may be involved in regulating the invasion of EVT duringplacentation.

In addition to the circulating RAS, tissue based systems arefound in heart, brain and reproductive tissues. The fetalematernalinterface comprises both the fetal placental tissue RAS and the

Table 1Clinical details of term pregnancy subject groups. Data are expressed as mean[standard deviation].

Normal termpregnancy

Pre-eclampsia P value

Number of subjects 10 10 n/aMaternal age (years) 28 [8] 32 [6] nsGestational age (weeks) 39.4 [1.1] 37.6 [2.6] P < 0.05Baby weight (kg) 3.4 [0.4] 2.9 [0.2] nsPlacental weight (g) 708 [112] 618 [145] nsSystolic blood

pressure (mmHg)115 [4] 159 [8] P < 0.000

Diastolic bloodpressure (mmHg)

75 [3] 98 [5] P < 0.000

Proteinuria (g/l) n/a 1.0 [0.6] n/aPlatelets (� 109/l) n/a 189 [43.9] n/aCreatinine (mmol/l) n/a 64 [9.7] n/a

P.J. Williams et al. / Placenta 31 (2010) 448e455 449

maternal decidual tissue RAS. Factors involved in stimulatingdecidualisation, including oestrogen and progesterone, have beenidentified as playing a role in regulating the local RAS, which isthought to be important in spiral artery remodelling.

Plasma renin concentration and activity, as well as AngII levelsare increased in pregnancy, but vascular responsiveness to AngII isdecreased [10]. In contrast, pre-eclamptic patients exhibitexaggerated pressor responses to AngII, although circulatingconcentrations are lower compared to control pregnancies [10].Although the underlying mechanism remains to be elucidated,there is growing evidence to indicate that dysregulation of both thetissue based and circulating RAS may be involved in the patho-physiology of PE [10].

The intervillous space is exposed to altering oxygen gradientsduring the first 10e12 weeks of gestation [11] during which timeEVT invades through the decidua. Following this period of EVTinvasion into uterine tissues, placental oxygen tension rapidlyincreases due to the maternal perfusion of the villous tips [12]. Thisoccurs episodically, and is predicted to have effects similar to thoseseen in hypoxia/reperfusion, leading to the development of localplacental oxidative stress. In the normal early placenta, antioxi-dants protect the placenta from undue damage due to oxidativestress [11].

The exact mechanism underlying the development of PEremains unknown, but it has been suggested that shallowtrophoblast invasion resulting in inadequate spiral artery trans-formation may underlie its pathogenesis [13]. Increased mito-chondrial generation of reactive oxygen species (ROS) andsynthesis through xanthine oxidase (XO) [14] and NADPH oxidaseresult in increased levels of ROS in PE and this is combined withdecreased expression of antioxidants [15]. Locally-generated AngIIis a potent stimulus to NADPH oxidase secretion [16]. NADPHoxidase is composed of a number of subunits, the catalyticmachinery of the enzyme being provided by gp91phox. A family ofgenes homologous to gp91phox known as Nox 1-5 has beendiscovered; Nox 4 is expressed in fetal tissues, placenta andvascular cells [17].

The objectives of the current study were two-fold. The first wasto examine the placental expression of AT1R, AT2R and AT4R, relatethis to the expression of NADPH oxidase and XO throughout normalpregnancy and compare the expression with that in PE. The secondwas to investigate a potential role for these receptors in normalplacental development.

2. Materials and methods

After local Ethical committee approval and with appropriate informed consent,placental tissue was obtained fromwomen undergoing elective surgical terminationof pregnancy (TOP) during the 1st trimester (early TOP; n ¼ 10; gestational age8.8 � 0.9 weeks [Mean � standard deviation]) and early 2nd trimester (mid TOP;n ¼ 10; gestational age 12.9 � 0.9 weeks) and at delivery in the third trimester from10 women with normal term pregnancy (gestational age 39.4 � 1.1 weeks) and 10women with PE (gestational age 37.6 � 2.6 weeks) who under went Caesareansection prior to labor, as previously described in Ref. [15]. Clinical data for normalterm pregnancy and PE subjects is shown in Table 1. Biopsies for immunohisto-chemistry were fixed in formalin and wax embedded, and 5 mM serial sections wereprepared. Biopsies for quantitative real-time PCR (qRT-PCR) were snap frozen andstored at �80 �C for generation of cDNA. Placental explants were prepared as out-lined below.

2.1. Immunohistochemistry

Five rabbit polyclonal antibodies and one murine monoclonal antibody (HLA-G)were employed for the immunostaining of paraffin-embedded placental sections.The optimal dilution for each antibody was selected on the basis of maximal specificreactivity and minimal background staining. Single immunohistochemical labellingwas performed using the Dako Envision visualisation system following the manu-facturer's instructions (Dako UK Ltd, Cambridgeshire, UK). Following microwaveheated pretreatment in trisodium citrate buffer (10 mM sodium citrate, pH 6.0)

sections were incubated for 30 min in normal swine serum (Dako) to block non-specific binding. Slides were then incubated with either anti-angiotensin II type 1receptor (Abcam PLC, Cambridge, UK; ab47162; 1:250 for 2 h), anti-angiotensin IItype 2 receptor (Abcam; ab19134; 1:500 for 4 h), anti-IRAP (kind gift from ProfessorS Keller, University of Virginia; 1:500 for 1 h) anti-NOX4 (Abcam; ab60940; 1:1000for 2 h), anti-xanthine oxidase (H-110) (Santa Cruz Biotechnology, Inc, Heidelberg,Germany; sc-20991; 1 in 200 for 2 h) or anti-HLA-G (Serotec; MCA2043; 1 in 200 for1 h). Specificity of stainingwas confirmed using positive control tissue. Kidney tissuewas used to confirm AT1R, AT2R and NOX4 staining, placental tissue was used forAT4R and HLA-G, liver for xanthine oxidase, breast carcinoma for caspase 3 andtonsil for Ki67. Negative controls were also performed in each staining run using noprimary antibody.

2.2. Analysis

All slides were analysed by the same observer (PW) who was blinded to preg-nancy stage or outcome. For analysis of placental sections, digital images of 5randomly selected medium-power (�200) fields were captured on NIS-ElementsF2.20 (Nikon UK Ltd, Surrey, UK). Using ImageScope (Aperio Technologies Ltd,Bristol, UK) the total percentage of positive labelled cells per �200 magnificationfield was determined using the ‘positive pixel count’ function. Results are expressedas ‘positivity’ which takes into account both the number of positive pixels and theintensity of staining. Assessment was performed of both trophoblast and villousstroma. A visual check was performed to ensure accurate discrimination of immu-nolabelled regions.

2.3. Preparation of mRNA

Total RNA from placental biopsies (50e100 mg) was obtained using 1 ml of Tri-reagent (Sigma, Poole, UK) according to the manufacturer's instructions. FollowingRNA extraction, 1 mg of each sample was reverse transcribed in a 20 ml reaction usingQuantiTect Reverse transcription kit (Qiagen, Crawley, UK) which includeda genomic DNA elimination reaction, using a Primus 96 advanced gradient ther-mocycler (Peqlab Ltd, Fareham, UK). The conditions used to generate first strandcDNA were 42 �C (15 min) and 95 �C (3 min). RNA concentration and quality wereverified by gel electrophoresis and spectrophotometrically using a standardconversion factor of one absorbance unit at 260 nm per 40 mg RNA/ml. RNAconcentration was adjusted by diluting with nuclease-free water (Ambion AppliedBiosystems, Warrington, UK) to a standard concentration of 1 mg/ml. All sampleswere stored at �80 �C until use.

2.4. Standard curve generation

Standards for AT1R, AT2R, AT4R and for the housekeeping gene TATA box bindingprotein (TBP) were made from cDNA obtained from a randomly selected controlplacenta using semiquantitative polymerase chain reaction (PCR). The method usedoligonucleotide primers to AT1R (F- TCTCAGCATTGATCGATACC and R-TGACTTTGGCTACAAGCATT; Melting temperature 57 �C), AT2R (F-TATGGCCTGTTTGTCCTCATand R-CATTGGGCATATTTCTCAGG; 55 �C), AT4R (F-TTCACCAATGATCGGCTTCAG andR-CTCCATCTCATGCTCACCAAG; 60 �C) and TBP (F-TATAATCCCAAGCGGTTTGC andR-GAATATGGTGGGGAGCTGTG; 58 �C) genes generating specific intron-spanningproducts.

The PCR program comprised an initial denaturation stage (95 �C, 15 min),amplification (stage I, 94 �C (30 s); stage II, melting temperature (30 s); stage III,72 �C (1 min)), and final extension (72 �C, 7 min; 8 �C ‘hold’). The PCR mixture (finalvolume 20 ml) contained 7 ml nuclease-free water (Ambion), 10 ml thermo-start PCRmaster mix (ABgene, Epsom, UK), 1 ml forward and reverse primers and 1 ml cDNA.The annealing temperature and cycle number of both primers were optimized and

P.J. Williams et al. / Placenta 31 (2010) 448e455450

used in their linear range. Agarose gel electrophoresis (2.0e2.5%) and ethidiumbromide staining confirmed the presence of both the product and TBP at theexpected sizes. The resultant PCR product was extracted (QIAquick gel extraction kit,Qiagen), sequenced, and results cross referenced against the Genbank website todetermine specificity of the target gene. After confirmation that the product wasspecific to the target gene, extracted PCR products were resuspended in nuclease-free water and a 10-fold serial dilution was performed. Standards were stored at�80 �C until use in quantitative real-time PCR (qRT-PCR).

2.5. Quantitative real-time PCR

qRT-PCR was used to quantify the expression of AT1R, AT2R and AT4R relative tothe housekeeping gene TBP. PCR reactions, set up in duplicate, were carried out in20 ml volumes consisting of 10 ml Fast SYBR Green master mix (Applied Biosystems),500 nM forward primer, 500 nM reverse primer, and 7 ml nuclease-free water. Real-time PCR was performed on all samples in a 7500 Fast Real-Time PCR System(Applied Biosystems) at 95 �C (20 s) followed by 40 cycles of 95 �C (3 s), 60 �C (30 s).Melt curve analysis was performed at 95 �Ce60 �C. Three negative control reactionswere carried out with each set of samples analyzed: (1) no RNA template but RT andpolymerase provided; (2) RNA and polymerase provided but no RT; and (3) RNA andRT provided but no polymerase. The crossing thresholds (Ct) of the target geneswere normalized to the housekeeping gene TBP, and relative quantification per-formed using the efficiency-corrected quantification model [18].

2.6. Invasion assays

The effect of AngII and AngIV on the ability of EVT from placental explants (7weeks gestational age) to invade through Matrigel growth factor reduced basementmembrane (BD supplied by VWR, Leicestershire, UK), was assessed as describedpreviously [19]. Briefly, chorionic villous tips were minced to approximately0.5 mm3 and resuspended in culture medium (Dulbecco's modified Eagle medium:F12 containing 10% FBS [FBS], penicillin/streptomycin, and amphotericin B [all fromSigma]). To assess the effect of AngII and AngIV on EVT invasion, exogenous AngII(10�10 M, 10�12 M or 10�14 M; BaChem UK Ltd, Merseyside, UK) or AngIV (10�6 M,10�9 M or 10�12 M; BaChem) was added to the culture media; these concentrationsbeing chosen as representing physiological ranges [20,21]. Each experiment wasperformed in duplicate and performed on 3 separate occasions. Invasion assays wereincubated for 6 days in standard tissue culture conditions 37 �C 5% CO2 (20% O2) inair incubator. To account for the considerable variation between different placentalsamples, mean cell counts for each invasion assay with explants from a singleplacenta were normalized with respect to the control values for that experiment. Ithas been previously demonstrated that the invaded cells have an EVT phenotype[22]. To determine whether AngII and AngIV altered trophoblast cell viability andnumbers (apoptosis and proliferation), at the end of the invasion assay explantswere fixed in 10% neutral-buffered formalin for 24 h and processed into paraffinwax. Serial 3 mm sections were immunolabelled using a standard avidin-biotinperoxidase method (Vector Laboratories, Peterborough, UK) with rabbit anti-activecaspase 3 (pretreatment citrate buffer pH 6.0 with 10 min microwave heating, 1:400dilution for 45 min; R&D Systems, Abingdon, UK) and mouse anti-Ki67 (pretreat-ment pressure cooking for 1 min in citrate buffer, pH 6.0, 1:200 dilution for 30 min;Novocastra Laboratories, Newcastle-upon-Tyne, UK). Due to the small size of theplacental explants the entire surface area of tissue was evaluated at �100magnification.

2.7. Statistical analysis

All analyses were performed using SPSS for Windows version 16.0 (Chicago, IL,USA). The KolmogoroveSmirnov test was used to assess the distribution of data.Depending on distribution of data, between-group comparisons were made usingeither one way analysis of variance (ANOVA) or the KruskaleWallis test as appro-priate, and if significant were followed by Bonferroni post hoc test orManneWhitney U tests, respectively. Dose response effects for invasion assays weredetermined using Kendall's Tau B test. A value of P¼ 0.05 was considered significant.Unless otherwise stated all data are expressed as mean (standard error of mean).

3. Results

3.1. Immunohistochemistry

3.1.1. Expression of AT1R, AT2R and AT4R in normal pregnancyPositive immunostaining for AT1R was present in placental

tissue from early TOP (0.36 � 0.02 positivity), mid TOP (0.36 � 0.02positivity) and term pregnancy (0.43 � 0.06 positivity) andremained at near constant levels throughout gestation (Fig. 1A).AT1R expression was predominant in syncytiotrophoblast andHofbauer cells in villous stroma in early and mid pregnancy, with

staining of endothelial cells of villous vessels present in termpregnancy. In contrast, AT2R positive immunostaining decreasedsignificantly between early TOP, mid TOP and term pregnancyplacenta (P ¼ 0.001). AT2R positivity reduced from both early TOP(0.23 � 0.02 positivity) and mid TOP (0.16 � 0.01 positivity)compared with term pregnancy placenta (0.07 � 0.2 positivity)(Fig. 1B). In addition expression of AT2R was decreased in mid TOPcompared with early TOP (P ¼ 0.014). In early TOP AT2R stainingwas identified in both villous and extravillous trophoblast EVT(Fig. 1). Expression of AT4R was present in placenta throughoutgestation, with staining predominant in syncytiotrophoblast andalso present in EVT in early TOP (Fig. 1). AT4R positivity did not varybetween early TOP (0.56 � 0.02 positivity) and mid TOP(0.48 � 0.04 positivity), but expression was increased in termpregnancy placenta (0.69 � 0.07 positivity) compared to mid TOP(Fig. 1C).

3.1.2. Expression of xanthine oxidase and NADPH oxidaseLevels of XO positivity were increased significantly in term preg-

nancy (0.15 � 0.003 positivity) compared to both earlyTOP (3.0 � 10�4 � 8.0 � 10�5 positivity) and mid TOP(6.0 � 10�4 � 2.5 � 10�4 positivity) (Fig. 2A). Xanthine oxidasepositivity was intense in both syncytiotrophoblast and also endo-thelial cells lining villous vessels. Positive staining for NADPHoxidasefollowed the similar pattern (Fig. 2B) with increased expression inplacenta from term pregnancy (6.8 � 10�2 � 1.4 � 10�2 positivity)compared to early TOP (9�10�4� 2.3�10�4 positivity) andmidTOP(3.7 � 10�3 � 9.7 � 10�4 positivity). NADPH oxidase expressionsignificantly increasedduringearlypregnancybetweenearlyandmidTOP samples (Fig. 2B). In contrast to expressionofXO,NADPHoxidaseexpression was predominantly localised to villous cytotrophoblast.

3.1.3. Pre-eclampsiaLevels of positivity of AT4R were significantly reduced in

placentae from PE (0.36 � 0.06 positivity) compared with termpregnancy (Fig. 1C). In contrast, expression of both AT1R and AT2Rdid not differ between placentae from term pregnancy and PE(0.45� 0.036 and 0.11�0.02 positivity). Placental expression of XO(0.21 � 0.02 positivity) and NADPH oxidase (0.14 � 0.03 positivity)both significantly increased in PE compared with term pregnancy(Fig. 2).

3.2. qRT-PCR for AT1R, AT2R and AT4R

Isolated total RNA quality was verified by an average A260/A280ratio of 1.97. An A260/A280 ratio greater than 1.8 is considered anacceptable indicator of RNA quality [23]. Expression of TBP mRNAdid not vary between groups (Table 2).

AT1RmRNAexpressiondidnot changeduringpregnancy (Table2).Furthermore, there was no difference in AT1R mRNA expressionbetween term pregnancy and PE (Table 2). AT2R expression wasreduced significantly in term pregnancy compared to both early TOPandmid TOP, see Table 2. Expression of AT4RmRNAwas significantlyincreased in termpregnancycompared tobothearly TOPandmidTOP(Table 2). In addition AT4R mRNA was significantly lowered inplacenta fromwomenwith PE (Table 2).

3.3. Effects of AngII and AngIV on EVT invasion

Culture in AngIV produced a dose dependent increase in themean invasion index of EVT (Fig. 3A). When compared to controlwells, incubationwith AngIV at concentrations of 10�6 M increasedinvasion by 74% (P < 0.0001) at 10�9 M by 56% (P < 0.0001) and at10�12 M by 22% (P ¼ 0.01) (Fig. 3A). Culture in AngII also showed

Fig. 1. A) AT1R, B) AT2R, C) AT4R immunostaining in placenta from 1) Early TOP, 2) Mid TOP, 3) Term pregnancy and 4) Pre-eclampsia, and D) Photomicrographs to confirmangiotensin receptor expression on invasive HLA-G positive EVT. Serial sections of placenta from Early TOP show immunostaining for 1) AT1R, 2) AT2R, 3) AT4R and 4) HLA-G.Regions outlined in red indicate HLA-G positive cells serving to confirm the location of the EVT population within the serial sections. In photomicrographs positive cells appearbrown; magnification �200 in AeC; magnification �400 in D. In graphs data are expressed as mean positivity per �200 field � SEM.

Fig. 2. Placental expression of A) XO and B) NADPH oxidase in 1) early TOP, 2) Mid TOP, 3) term pregnancy and 4) pre-eclampsia. In photomicrographs positive cells appear brown;magnification �200. In graphs data are expressed as mean � SEM positivity per �200 field of multiple observations.

Table 2Efficiency-corrected relative mRNA expression ratio for AT1R, AT2R and AT4R inplacenta.

Gene Early TOP Late TOP Termpregnancy

Pre-eclampsia

AT1R 2.66 (0.93) 2.27 (0.50) 4.13 (1.12) 5.68 (2.36)AT2R 3.67 (0.93) 2.61 (0.33) 0.81 (0.21) 1.11 (0.29)

P ¼ 0.001 vsTerm

P ¼ 0.041 vsTerm

AT4R 5.59 (1.37) 3.01 (0.62) 13.4 (2.10) 3.27 (0.73)P ¼ 0.002 vsTerm

P < 0.0001 vsTerm

P < 0.0001 vsterm

TBPa 3.68 (0.26) 3.68 (0.22) 3.54 (0.19) 4.44 (0.26)

a TBP expressed asmRNA concentration� 10�9 ng/ml. Data are expressed asmeanof duplicates � SEM.

P.J. Williams et al. / Placenta 31 (2010) 448e455452

a trend towards increasing EVT invasion, but this failed to reachsignificance (P > 0.5; data not shown).

3.4. Trophoblast cell apoptosis and proliferation following culture inAngII and AngIV

Following culture with AngIV immunolabelling for caspase 3showed a dose dependent decrease (P ¼ 0.03; Fig. 3B). The meancaspase 3 positivity of placental explants was reduced followingculture with AngIV at 10�6 M (71.23 � 3.91; P ¼ 0.05 vs control),10�9M (46.97� 7.12; P¼ 0.02 vs control) and 10�12M (15.52� 4.24;P¼0.01 vs control). In contrast, immunopositivity forKi67exhibiteda dose dependent increase (P ¼ 0.003). Ki67 mean positivity ofplacental explants after culturewithAngIV (Fig. 3C)was increased at

Fig. 3. Effect of culture in AngIV on A) the invasive capacity of placental villous explants, B)Placental explants were cultured with 1) DMEMF12 culture medium alone (control) and withphotomicrographs positive cells appear brown; magnification �200. In graphs data are exp

10�6 M (1670 � 371; P < 0.0001 vs control), 10�9 M (54.9 � 136;P¼ 0.007 vs control) and 10�12 M (127� 9.13; P¼ 0.040 vs control).

Culture of placental explants with AngII also had a dosedependent effect on caspase 3 expression (Fig. 4A). AngII reducedcaspase 3mean positivity at concentrations of 10�10 M (91.26� 5.8;P ¼ 0.05 vs control), 10�12 M (75.3 � 15.8; P ¼ 0.03 vs control) and10�14 M (15.3 � 0.70; P ¼ 0.003 vs control). The mean positivity ofKi67 of placental explants was increased following culture in AngII(Fig. 4B) at concentrations of 10�10 M (2621 � 123; P < 0.0001 vscontrol), 10�12 M (625 � 165; P ¼ 0.007 vs control) and 10�14 M(289 � 112; P ¼ 0.01 vs control).

4. Discussion

This study demonstrates that the placental RAS has a potentialrole in regulating EVT invasion. Furthermore, the present studyshows the relation of placental AT1R, AT2R and AT4R expressionwith that of the pro-oxidant enzymes XO and NADPH oxidase,providing an insight into their expression throughout pregnancy,and furthermore how their expression is altered in PE.

The angiotensin type AT1R is widely expressed in adulthood andmediates a wide variety of actions, while the AT2R is expressed atmuch lower levels. However, AT2R expression is much higher infetal and uteroplacental tissues than in adult tissues leading to thesuggestion that this receptor is important for growth and devel-opment [4]. It is generally accepted that the response to AngII isdetermined by the balance in expression of AT1R and AT2R [4,24]. Ifthis is so, the physiological effects of the relatively constantexpression of placental AT1Rs throughout pregnancy may bemarkedly modulated by the changing expression of the AT2Rs. The

caspase 3 expression of placental explants and C) Ki67 expression of placental explants.AngIV at concentrations of 2) 10�6 M AngIV, 3) 10�9 M AngIV and 4) 10�12 M AngIV. Inressed as mean invasion index normalized to controls � SEM (n ¼ 3).

0

20

40

60

80

100

120

Control 10-10M 10-12M 10-14MAngiotensin II concentration

No

rm

alised

casp

ase 3

po

sit

ivit

y

0

500

1000

1500

2000

2500

3000

Control 10-10M 10-12M 10-14MAngiotensin II concentration

No

rm

alised

Ki6

7 p

osit

ivit

y

P=0.001

P=0.03

A

B

Fig. 4. Effect of culture in AngII on A) caspase 3 expression and B) Ki67 expression of placental explants. Data shown are expressed as mean positivity normalized to controls � SEM(n ¼ 3, except for 10�10 M where n ¼ 1).

P.J. Williams et al. / Placenta 31 (2010) 448e455 453

high levels of expression of AT2R during early pregnancymay relateto the role of this receptor in the regulation of apoptosis, cellularproliferation and angiogenesis, all processes which are needed forsuccessful placentation [25].

We have confirmed previous reports of AT4R expression inhuman placenta [9,26] and now show high levels of expression atboth mRNA and protein level throughout pregnancy. Moreover,AT4R was expressed at higher levels than either AT1R or AT2R,indicating that the AT4R may have an important role in normalpregnancy. The higher expression of AT4R during normalpregnancy compared to AT1R may represent a control mechanismto ensure that AngIV predominantly binds with AT4R, and toprevent AngIV binding with AT1R which may lead to pathology. Forexample, in adults, high concentrations of AngIV can causesystemic and renal vasoconstriction that is completely abrogated bylosartan, indicating a low-affinity AT1R-mediated response [27,28].The balance in expression between AT4R and AT1R, similar to thatbetween AT2R and AT1R, may govern the functional role that AngIVhas in pregnancy and placental development. The high levels ofexpression of AT4R, with intense staining located in the invasiveEVT cell population, during early pregnancy suggest that thisreceptor may have a role in regulating placental development andthat deficient AT4R expression in early pregnancy may lead to thedevelopment of PE at term; our data support this suggestion sinceangiotensin IV stimulated EVT invasion in a dose dependentfashion. The finding of increased EVT invasion following culturewith AngIV is in contrast to the report of Cohen et al. [29], whoshowed that binding of a substituted AngIV molecule to IRAPreduced cytotrophoblast invasion. They used LVV-hemorphin 7

(LVV-H7) as ligand for IRAP. Although LVV-H7 has been identifiedas a potent AT4R ligand [30], it is structurally different from AngIV[31]. The altered structure of ligands used in these studies may haveresulted in different interactions with the IRAP active site of AT4Rand led to the generation of altered downstream signallingpathways.

Increased EVT invasion following culture with AngIV concurswith the demonstration that Ikaros, a transcription factor thatregulates AT4R, is highly expressed in the invasive EVT cell pop-ulation and that suppression of Ikaros results in decreased migra-tion and invasion of EVT cells [32].

No alteration in expression of AT1R was identified between termnormal pregnancy and PE. This is in agreement with the report byHerse et al. [33] of placental AT1R remaining constant, however, theydescribed upregulation of decidual AT1R in PE. Expression of AT2Ralso did not vary between normal pregnancy and PE, although thereduction in AT1R expression seen in normal pregnancy was notsignificant in pre-eclampsia. Dysfunction of the RAS has been impli-cated in the underlying deficiency of uteroplacental transfer in PE[34]. The present results indicate that altered expression of AT1R andAT2R are not involved in this dysregulation, and that altered pressorand vascular responsiveness to AngII in PE is due to the developmentof autoantibodieswhich facilitate bindingofAngIIwithAT1R [35]. It isunlikely that AT1R autoantibodies are able to interfere with AngIV/AT4R binding due to the structural differences between thesereceptors [36]; although this needs to be formally investigated.

We have shown high levels of AT4R expression in normal earlypregnancy and, for the first time, significantly reduced expressionin pre-eclamptic tissues near term, compared with normal term

P.J. Williams et al. / Placenta 31 (2010) 448e455454

placenta. We speculate that, if expression of AT4R is also low inthe first trimester in some women, this may be associated withthe reduced trophoblast invasion found in PE [13]. This may alsobe associated with decreased AngIV-mediated vasodilatation [37].To further worsen the impact of reduced AT4R expression in PE,AngIV may also be more available to bind with AT1R, stimulatingfurther vasoconstriction [38] which may contribute to theincreased oxidative stress characteristic of PE. We have confirmedprevious reports of increased XO and NADPH oxidase immunos-taining in PE [14,39], apparently co-localising with angiotensinreceptors. Locally-generated AngII is a potent stimulus to NADPHoxidase secretion [16]; it is not known whether AngIV shares thiseffect.

Xanthine oxidase and NADPH oxidase expression in placentawere similar, with increased amounts present in term pregnancycompared to both early and mid TOP. Increased expression of XOin term pregnancy concurs with the previous report by Many andcoworkers, who also reported a modest increase in expressionwith gestation [14]. Increased placental expression of XO andNADPH oxidase in PE also confirm studies by others [14,40], andextend these by offering a possible explanation for increasedexpression of the pro-oxidant enzymes due to alterations in theRAS in PE. Due to deficient EVT invasion and spiral arteryremodelling in PE there is increased vasoconstriction within thematernal-placental arteries and this leads to intermittentplacental perfusion [41,42] and ischemia/reperfusion in theplacenta [43]. Reperfusion following ischemia stimulates theconversion of xanthine oxidoreductase to favour production ofXO. Increased levels of XO culminate in increased superoxideproduction with consequential increase of oxidative stress,characteristic of PE [44].

There are a number of possible limitations of the current study.EVT invasion increased following treatment with AngIV. We havenot formally assessed the role of AngIV binding to AT1R on EVTinvasion; rather we have assumed that AngIV acted via AT4R andthat AngII acted via AT1R. Future work using the specific AT1R,AT2R and AT4R blockers, Losartan, PD123319 and divalinal AngIV,respectively, would enable a more in depth evaluation of the rolethat each of these receptors has in regulating EVT invasion. Inaddition for invasion assays only placentae of 7 weeks gestationalage were used, and it is important to note that effects of AngII andAngIV may be altered if other time points were used. The use of 7weeks gestational age placentae was chosen as trophoblastinvasion has been shown to decrease with gestational age [45],therefore, any effects of AngII or AngIV on EVT invasion would bereadily detectable.

The current results indicate a role for members of the RAS andtheir receptors in the establishment of the placenta. Such a role hasbeen suggested since the demonstration that the early secondtrimester chorion synthesises very high concentration of renin [46].Our results suggest that studies of the regulation of AT4R mightshed further light on processes of trophoblast development andits possible inadequate development in such conditions such aspre-eclamspia and spontaneous miscarriage.

In summary, we have shown expression of AT4R as well as AT1R,and AT2R in early TOP, mid TOP and term pregnancy. We have alsoshown for the first time that AngIV and AngII are capable of stim-ulating EVT invasion, with AngIV being more potent than AngII;both also regulate markers of placental apoptosis and proliferation.Furthermore reduction in expression of AT4R indicates that thisreceptor may have a role in the development of PE.

Funding

PW was supported by a Wellcome Trust VIP Fellowship.

Acknowledgements

The authors wish to gratefully acknowledge the assistance ofstaff at the Queen's Medical Centre, Nottingham and at the RoyalVictoria Infirmary, Newcastle-upon-Tyne for their help with samplecollection, and Professor Susanna Keller, University of Virginia, USA,for providing the rabbit polyclonal anti-IRAP antibody used forimmunohistochemistry and Dr Gendie Lash, Newcastle Universityfor advice on invasion assays.

References

[1] Chung O, Kuhl H, Stoll M, Unger T. Physiological and pharmacological impli-cations of AT1 versus AT2 receptors. Kidney Int Suppl 1998;67:S95.

[2] Ozono R, Wang ZQ, Moore AF, Inagami T, Siragy HM, Carey RM. Expression ofthe subtype 2 angiotensin (AT2) receptor protein in rat kidney. Hypertension1997;30:1238.

[3] Grishko V, Pastukh V, Solodushko V, Gillespie M, Azuma J, Schaffer S.Apoptotic cascade initiated by angiotensin II in neonatal cardiomyocytes: roleof DNA damage. Am J Physiol Heart Circ Physiol 2003;285:H2364.

[4] de Gasparo M, Catt KJ, Inagami T, Wright JW, Unger T. International union ofpharmacology. XXIII. The angiotensin II receptors. Pharmacol Rev2000;52:415.

[5] Albiston AL, McDowall SG, Matsacos D, Sim P, Clune E, Mustafa T, et al.Evidence that the angiotensin IV (AT(4)) receptor is the enzyme insulin-regulated aminopeptidase. J Biol Chem 2001;276:48623.

[6] Hall KL, Venkateswaran S, Hanesworth JM, Schelling ME, Harding JW. Char-acterization of a functional angiotensin IV receptor on coronary microvascularendothelial cells. Regul Pept 1995;58:107.

[7] Hall KL, Hanesworth JM, Ball AE, Felgenhauer GP, Hosick HL, Harding JW.Identification and characterization of a novel angiotensin binding site incultured vascular smooth muscle cells that is specific for the hexapeptide (3-8)fragment of angiotensin II, angiotensin IV. Regul Pept 1993;44:225.

[8] Coleman JK, Krebs LT, Hamilton TA, Ong B, Lawrence KA, Sardinia MF, et al.Autoradiographic identification of kidney angiotensin IV binding sites andangiotensin IV-induced renal cortical blood flow changes in rats. Peptides1998;19:269.

[9] Ino K, Kikkawa F, Suzuki T, Kajiyama H, Shibata K, Nomura S, et al. Expressionof placental leucine aminopeptidase and adipocyte-derived leucine amino-peptidase in human normal and malignant invasive trophoblastic cells. LabInvest 2003;83:1799.

[10] Shah DM. The role of RAS in the pathogenesis of preeclampsia. Curr HypertensRep 2006;8:144.

[11] Jauniaux E, Watson AL, Hempstock J, Bao YP, Skepper JN, Burton GJ. Onset ofmaternal arterial blood flow and placental oxidative stress. A possible factor inhuman early pregnancy failure. Am J Pathol 2000;157:2111.

[12] Huppertz B, Gauster M, Orendi K, Konig J, Moser G. Oxygen as modulator oftrophoblast invasion. J Anat; 2009.

[13] Redman CW, Sargent IL. Latest advances in understanding preeclampsia.Science 2005;308:1592.

[14] Many A, Hubel CA, Fisher SJ, Roberts JM, Zhou Y. Invasive cytotrophoblastsmanifest evidence of oxidative stress in preeclampsia. Am J Pathol2000;156:321.

[15] Mistry HD, Wilson V, Ramsay MM, Symonds ME, Broughton Pipkin F. Reducedselenium concentrations and glutathione peroxidase activity in preeclampticpregnancies. Hypertension 2008;52:881.

[16] Griendling KK, Minieri CA, Ollerenshaw JD, Alexander RW. Angiotensin IIstimulates NADH and NADPH oxidase activity in cultured vascular smoothmuscle cells. Circ Res 1994;74:1141.

[17] Cheng G, Cao Z, Xu X, van Meir EG, Lambeth JD. Homologs of gp91phox:cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 2001;269:131.

[18] Pfaffl MW. A new mathematical model for relative quantification in real-timeRT-PCR. Nucleic Acids Res 2001;29:e45.

[19] Scaife PJ, Bulmer JN, Robson SC, Innes BA, Searle RF. Effector activity ofdecidual CD8þ T lymphocytes in early human pregnancy. Biol Reprod2006;75:562.

[20] Jankowski V, Vanholder R, van der Giet M, Tolle M, Karadogan S, Gobom J,et al. Mass-spectrometric identification of a novel angiotensin peptide inhuman plasma. Arterioscler Thromb Vasc Biol 2007;27:297.

[21] Plovsing RR, Wamberg C, Sandgaard NC, Simonsen JA, Holstein-Rathlou NH,Hoilund-Carlsen PF, et al. Effects of truncated angiotensins in humans afterdouble blockade of the renin system. Am J Physiol Regul Integr Comp Physiol2003;285:R981.

[22] Lash GE, Otun HA, Innes BA, Bulmer JN, Searle RF, Robson SC. Inhibition oftrophoblast cell invasion by TGFB1, 2, and 3 is associated with a decrease inactive proteases. Biol Reprod 2005;73:374.

[23] Sambrook J, Fritsch EF, Maniatis T. Molecular cloning. A laboratory manual.New York: Cold Spring Harbor Laboratory Press; 1989.

[24] Berk BC. Angiotensin type 2 receptor (AT2R): a challenging twin. Sci STKE2003;2003:PE16.

P.J. Williams et al. / Placenta 31 (2010) 448e455 455

[25] Ray J, Jurisicova A, Caniggia I. IFPA Trophoblast Research Award Lecture: thedynamic role of Bcl-2 family members in trophoblast cell fate. Placenta2009;30(Suppl. A):S96.

[26] Yamahara N, Nomura S, Suzuki T, Itakura A, Ito M, Okamoto T, et al. Placentalleucine aminopeptidase/oxytocinase in maternal serum and placenta duringnormal pregnancy. Life Sci 2000;66:1401.

[27] Gardiner SM, Kemp PA, March JE, Bennett T. Regional haemodynamic effects ofangiotensin II (3-8) in conscious rats. Br J Pharmacol 1993;110:159.

[28] Li XC, Campbell DJ, Ohishi M, Yuan S, Zhuo JL. AT1 receptor-activated signalingmediates angiotensin IV-induced renal cortical vasoconstriction in rats. AmJ Physiol Renal Physiol 2006;290:F1024.

[29] Cohen M, Wuillemin C, Chai SY, Bischof P. Effects of ligands or substrate ofinsulin-regulated aminopeptidase (IRAP) on trophoblast invasion. J Pept Sci2008;14:883.

[30] Sardinia MF, Hanesworth JM, Krebs LT, Harding JW. AT4 receptor bindingcharacteristics: D-amino acid- and glycine-substituted peptides. Peptides1993;14:949.

[31] Vanderheyden PM. From angiotensin IV binding site to AT4 receptor. Mol CellEndocrinol 2009;302:159.

[32] Yamamoto E, Ito T, Abe A, Sido F, Ino K, Itakura A, et al. Ikaros is expressed inhuman extravillous trophoblasts and involved in their migration and invasion.Mol Hum Reprod 2005;11:825.

[33] Herse F, Dechend R, Harsem NK, Wallukat G, Janke J, Qadri F, et al. Dysregu-lation of the circulating and tissue-based renin-angiotensin system inpreeclampsia. Hypertension 2007;49:604.

[34] Irani RA, Xia Y. The functional role of the renin-angiotensin system inpregnancy and preeclampsia. Placenta 2008;29:763.

[35] Herse F, Staff AC, Hering L, Muller DN, Luft FC, Dechend R. AT1-receptorautoantibodies and uteroplacental RAS in pregnancy and pre-eclampsia. J MolMed 2008;86:697.

[36] Thomas WG, Mendelsohn FA. Angiotensin receptors: form and function anddistribution. Int J Biochem Cell Biol 2003;35:774.

[37] Fyhrquist F, Saijonmaa O. Renin-angiotensin system revisited. J Intern Med2008;264:224.

[38] Dupont AG, Yang R, Smolders I, Vanderheyden P. IRAP and AT(1) receptormediated effects of angiotensin IV. J Intern Med 2009;265:401.

[39] Cui XL, Brockman D, Campos B, Myatt L. Expression of NADPH oxidase isoform1 (Nox1) in human placenta: involvement in preeclampsia. Placenta2006;27:422.

[40] Karabulut AB, Kafkasli A, Burak F, Gozukara EM. Maternal and fetal plasmaadenosine deaminase, xanthine oxidase and malondialdehyde levels inpre-eclampsia. Cell Biochem Funct 2005;23:279.

[41] Hung TH, Skepper JN, Charnock-Jones DS, Burton GJ. Hypoxia-reoxygenation:a potent inducer of apoptotic changes in the human placenta and possibleetiological factor in preeclampsia. Circ Res 2002;90:1274.

[42] Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascularresponse to placentation in pregnancies complicated by pre-eclampsia and bysmall-for-gestational age infants. Br J Obstet Gynaecol 1986;93:1049.

[43] Hung TH, Skepper JN, Burton GJ. In vitro ischemia-reperfusion injury in termhuman placenta as a model for oxidative stress in pathological pregnancies.Am J Pathol 2001;159:1031.

[44] Mills TA, Wareing M, Shennan AH, Poston L, Baker PN, Greenwood SL. Acuteand chronic modulation of placental chorionic plate artery reactivity byreactive oxygen species. Free Radic Biol Med 2009;47:159.

[45] Lash GE, Otun HA, Innes BA, Bulmer JN, Searle RF, Robson SC. Low oxygenconcentrations inhibit trophoblast cell invasion from early gestation placentalexplants via alterations in levels of the urokinase plasminogen activatorsystem. Biol Reprod 2006;74:403.

[46] Symonds EM, Stanley MA, Skinner SL. Production of renin by in vitro culturesof human chorion and uterine muscle. Nature 1968;217:1152.