lable at ScienceDirect

Placenta 36 (2015) 809e820

Contents lists avai

Placenta

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

Laminin a4 (LAMA4) expression promotes trophoblast cell invasion,migration, and angiogenesis, and is lowered in preeclamptic placentas

N. Shan a, X. Zhang a, X. Xiao b, H. Zhang a, C. Tong a, X. Luo a, Y. Chen a, X. Liu a, N. Yin a,Q. Deng a, H. Qi a, *

a Department of Obstetrics and Gynecology, The First Affiliated Hospital, Chongqing Medical University, Chongqing 400016, Chinab Laboratory of Lipid and Glucose Research, The First Affiliated Hospital, Chongqing Medical University, Chongqing 400016, China

a r t i c l e i n f o

Article history:Accepted 15 April 2015

Keywords:LamininLAMA4PlacentaTrophoblastPreeclampsiaMAPKp38JNKERKMethylation

* Corresponding author. Department of ObstetricAffiliated Hospital, Chongqing Medical University, Notrict, Chongqing 400016, China. Tel.: þ86 1380 837 6

E-mail address: qihongbocq@gmail.com (H. Qi).

http://dx.doi.org/10.1016/j.placenta.2015.04.0080143-4004/© 2015 Published by Elsevier Ltd.

a b s t r a c t

Introduction: The laminin a4 subunit (LAMA4) has been shown to promote migration, proliferation, andsurvival of various cell types. This study investigated LAMA4's role in trophoblast cells during placentaldevelopment.Methods: LAMA4 expressionwas immunohistochemically assessed in the first trimester and term humanplacentas. LAMA4 siRNA was applied to silence LAMA4 expression in extravillous explants and HTR8/SVneo cells. Hypoxiaereoxygenation (H/R) conditions were applied to mimic preeclampsia. LAMA4expression and trophoblast cell invasion, migration, and tube formation (a measure of angiogenesis)were assessed in HTR8/SVneo cells. The p38 mitogen-activated protein kinase (MAPK) inhibitorSB203580 was used to study the mechanism underlying LAMA4 activity. LAMA4 promoter methylationwas assessed by bisulfite-sequencing polymerase chain reaction (PCR) or methylation-specific PCR.Results: LAMA4 levels in preeclamptic placentas were significantly lower than those in controls. LAMA4silencing significantly inhibited extravillous explant outgrowth as well as HTR8/SVneo cell invasion andmigration. H/R conditions significantly lowered LAMA4 expression. Application of either H/R conditionsor LAMA4 silencing both significantly decreased HTR8/SVneo cell invasion, migration, and tube forma-tion, decreased MMP2 and MMP9 expression, and increased TIMP2 expression. SB203580 significantlyreduced LAMA4 expression. LAMA4 silencing significantly decreased p-p38, p-c-Jun N-terminal kinase(JNK), and p-extracellular signal-regulated kinase (ERK) expressions; by contrast, H/R conditions inducedsignificant upregulation of p-p38 and p-ERK but decreased p-JNK. LAMA4 promoter methylation was notsignificantly altered in preeclamptic placentas compared to controls.Conclusions: LAMA4 expression is lowered in preeclamptic placentas and promotes trophoblast cellinvasion, migration, and angiogenesis. H/R conditions decrease LAMA4 expression and appear todecouple the positive relationship between LAMA4 expression and p38 and ERK activation.

© 2015 Published by Elsevier Ltd.

1. Introduction

Preeclampsia (PE) is a pregnancy-specific disorder characterizedby hypertension and proteinuria. It affects 5e7% of all healthypregnancies and causes approximately 50,000 maternal deathsannually [1]. Although numerous studies have attempted to eluci-date this complex multifactorial disease, the pathogenesis of PE isstill not fully understood.

s and Gynecology, The First. 1 Youyi Road, Yuzhong Dis-116.

After embryo implantation, cytotrophoblast (CTB) cells prolif-erate and differentiate into syncytiotrophoblasts (STBs) orextravillous trophoblasts (EVTs) [2]. EVTs infiltrate the endomyo-metrium to anchor the placenta and penetrate uterine spiral ar-teries to remodel the vasculature for adequate blood supply andnutrients [3]. The invasion of uterine decidua and maternalvasculature by trophoblast cells, especially EVTs, is a critical processfor successful establishment of maternalefetal circulation.

The precise pathogenesis of PE is debatable, but oxygendisruption and abnormal trophoblast invasion seem to be the keycontributing factors [4]. An elevated level of oxidative stress isencountered in PE. The gestational-stage-dependent variation inoxygen tension is not simply the result of blood vessel development

N. Shan et al. / Placenta 36 (2015) 809e820810

in the maternalefetal interface, but it plays important roles inmodulating trophoblast cell behavior to maintain normal placen-tation. Studies in HTR8/SVneo cells have produced inconsistentresults on how oxygen levels affect cell invasion. Reports fromGraham et al. strongly support the invasion-promoting function oflow oxygen tension via elevating expression of cell surface uPAR inHTR8/SVneo cells [5]. Increasing evidence suggests that hypo-xiaereoxygenation (H/R) within the placenta leads to reactive ox-ygen species (ROS) production, and these H/R-induced ROS (such ashydrogen peroxide) can cause apoptosis and resulting cellular in-vasion [6].

The extracellular matrix (ECM) consists of fibrous proteins,proteoglycans, and glycoproteins such as fibronectin, laminin,entactin, tenascin, and thrombospondin. Interactions of basementmembrane constituents with cell surface receptors can influencecytoskeletal organization and intracellular signaling, therebyregulating various cell functions, including attachment, differenti-ation, and proliferation. In particular, laminins are the main non-collagenous glycoprotein found in the basement membrane, andthey are composed of one a-, one b-, and one g-chain, forming ashape of a cross [7]. Laminin is the first basement membranecomponent to appear during the early stages of embryonic devel-opment, and it promotes cell attachment and angiogenesis [8,9].The a4 subunit (laminin a4 (LAMA4)) is a component of laminin-8(which is composed of a4, b1, and g1) and laminin-9 (which iscomposed of a4, b2, and g1), and it has been found in various tis-sues of mesenchymal origin, endothelial basement membranes,and certain epithelial basement membranes [10e12]. Several arti-cles have associated increases in circulating laminins with variouscancers [13e16], and LAMA4 appears to promote migration, pro-liferation, and survival of endothelial cells, blood cells, and cancercells [17e19].

Although trophoblast cells and cancer cells share some com-mon invasive abilities [20], the role of LAMA4 in the placenta hasremained largely unresolved. Therefore, we set out to study therole of placental LAMA4 during pregnancy by measuring theexpression of serum LAMA4 in PE patients. Then, we evaluated itseffect on trophoblast invasive and migratory abilities through RNAinterference (RNAi) and in vitro cell invasion and migration assayson a human placental extravillous explant culture. We furtherused an H/R cycle to mimic the pathological conditions of PE. Inaddition, we combined the p38 mitogen-activated protein kinase(MAPK) pathway inhibitor SB203580 to decipher the molecularmechanisms of the LAMA4-p38 MAPK signaling pathway introphoblast cells. Finally, we investigated LAMA4's promotermethylation to determine any epigenetic mechanism(s) that mayaffect LAMA4's expression in PE placentas. Our findings supportLAMA4 downregulation in PE placentas and LAMA4's role inpromoting invasion, migration, and angiogenesis of trophoblastcells.

2. Materials and methods

2.1. Ethics statement

Ethical approval was granted by the ethics committee (international reviewboard (IRB)) of the First Affiliated Hospital of Chongqing Medical University(Chongqing, China). All womenwere informed and signed a consent form to donatetheir placentas for scientific purposes.

2.2. Patients and tissue collection

According to the American College of Obstetrics and Gynecology, PE is defined asa pregnancy-specific syndrome that occurs after 20 weeks of gestation in previouslynormotensive women, and it is characterized by a systolic blood pressure of�140 mm Hg and/or a diastolic blood pressure of �90 mm Hg in combination withproteinuria (0.3 g or more of protein in a 24-h urine collection, which usually cor-responds to 1þ or greater on a urine-dipstick test). All subjects with chronic medical

disorders such as diabetes mellitus, cardiovascular disease, collagen disorders,chronic renal diseases, chronic hypertension, andmetabolic diseases were excluded.

Human placental tissues from 6 to 8 weeks (defined as “first trimester”) wereobtained from healthy women undergoing legal abortion by cesarean section fornonmedical reasons; normal term (n ¼ 20) and PE (n ¼ 22) placentas were collectedafter cesarean section. Considering patient safety, we collected samples from nearthe placental bed and not from the inner myometrium. Specifically, samples of thematernal placental surface close to the basal decidua (approximately 1e2 cm3 and100-mg net weight) were collected after cesarean section. Tissues were dissectedfrom the maternal side from five cotyledons free of visible infarction, calcification,hematoma, or tears, and they were located midway between the umbilical cordinsertion site and the peripheral edge of the placenta. Immediately after collection, aportion of each tissue sample was snap-frozen and stored at �80 �C, while theremaining portion was collected and washed in 0.9% saline for immunohisto-chemistry (IHC), fixed in 3% formaldehyde, and embedded with paraffin at roomtemperature.

2.3. Cell culture and H/R application

The human EVTcell line HTR8/SVneo has beenwidely used as a model for EVT inthe first trimester. HTR8/SVneo cells used in this study were kindly provided by Dr.Charles H. Graham (Kingston, ON, Canada). HTR8/SVneo cells were routinelycultured in Roswell Park Memorial Institute (RPMI) 1640 (Gibco, MA, USA) mediacontaining 10% fetal bovine serum (FBS) (Gibco, MA, USA). Cells maintained understandard culture conditions (5% CO2 at 37 �C) were used as the normal control. H/Rintervention (two 24-h cycles consisting of 8 h at 2% oxygen followed by 16 h atstandard culture conditions) was performed as previously described [21].

2.4. RNAi and p38 MAPK inhibition

For LAMA4 RNAi, cells were transfected with 100-nM LAMA4 siRNA (50-CAGG-GAUUUAUGCAGAAAUTT-30 , GenePharma, Shanghai, China; Genbank ID forLAMA4:NM_000006.12) or the control siRNA (a universal negative control, Gene-Pharma) with Lipofectamine™ 2000 (Invitrogen, Carlsbad, CA, USA), and thetransfection procedures were performed according to the manufacturer's protocol.The transfection efficiency was >90% as measured by fluorescence-labeled siRNA.SB203580 is a specific inhibitor of p38 MAPK, and it was dissolved in dimethylsulfoxide (DMSO) as previously described [22]. Cells were preincubated with 5-mMSB203580 (Invitrogen, Carlsbad, CA, USA) for 90 min.

2.5. Immunohistochemistry

IHC was performed as described previously [23]. Sections were incubated with aprimary LAMA4 antibody (1:500; SigmaeAldrich Co), cytokeratin-7 (CK7, 1:500;Santa Cruz Biotechnology Co.), and human leukocyte antigen G (HLA-G) (1:500;Abcam) overnight at 4 �C. Then, the sections were incubated with horseradishperoxidase-conjugated anti-rabbit secondary antibody (1:1000; Santa CruzBiotechnology Co.). Staining was performed by incubation with diaminobenzidinechromogen solution (DAB). Hematoxylin was used as a nuclear counter stain. Anegative control was performed using nonimmune rabbit interleukin (Ig)G. For eachsample, experiments were performed in triplicate.

2.6. Western blotting

Proteins extracted from tissues or HTR8/SVneo cells were subjected to Westernblotting analysis. The membranes were blotted using primary antibodies to LAMA4(1:1000; SigmaeAldrich Co. #SAB4501719), E-cadherin (1:1000; Cell SignalingTechnology #3195), vimentin (1:1000; Cell Signaling Technology #5741), p38(1:1000; Cell Signaling Technology #4631), JNK (1:1000; Cell Signaling Technology#4008), extracellular signal-regulated kinase (ERK) (1:1000; Cell Signaling Tech-nology #4370), MMP-9 (1:500; Santa Cruz Biotechnology Co. sc-10737), MMP-2(1:500; Santa Cruz Biotechnology Co. sc-10736), TIMP1 (1:500; Santa CruzBiotechnology Co. sc-5538), TIMP2 (1:500; Santa Cruz Biotechnology Co. #sc-21735), and b-actin (1:1000; Santa Cruz Biotechnology Co. #sc-130656) at 4 �Covernight, and they were then incubated with a horseradish peroxidase-conjugatedsecondary antibody (1:5000; Santa Cruz Biotechnology Co.). Densitometric analysiswas performedwith enhanced chemiluminescence reagents and a Chemi-doc imageanalyzer (Bio-Rad). All experiments were repeated in triplicate.

2.7. RNA extraction and quantitative real-time reverse-transcriptase polymerasechain reaction

Total RNA was extracted from cells or tissues using TRIzol regent (TakaraBiotechnology, Japan), and it was purified following the manufacturer's instructions.Total RNA was reverse-transcribed into cDNA with Superscript II First-Strand Syn-thesis System for quantitative reverse-transcriptase polymerase chain reaction (RT-PCR) (Takara Biotechnology, Japan). Primers are listed in Table 1. The PCR conditionsconsisted of an initial preincubation step at 95 �C for 30 s, followed by 39 cycles at95 �C for 5 s and 60 �C for 30 min. The mean threshold cycle values were normalized

Table 1Primers for quantitative RT-PCR

Gene symbol Primer sequence (50e30)

LAMA4 F GGAAAATAAGCGAGGCACCGR AGCCACAGAGGCAGAACCGA

a-tubulin F CCAAGTTCTGGGAGGTGATCTGR TTGTAGTAGACGTTGATGCGCTC

N. Shan et al. / Placenta 36 (2015) 809e820 811

to a-tubulin, and the relative mRNA levels were analyzed using the 2-DDCt method.The experiments were performed in triplicate.

2.8. Placental villous explant culture

Eighteen cases of first-trimester placental villi, including six cases each of 6, 7,and 8 gestational weeks, were dissected from the tips of villi and then explanted aspreviously described [24]. The explants were cultured in serum-free DMEM/F12media with 500-nM siRNA, or 5-mM SB203580 for 90 min before treatment [22],100-IU/ml penicillin, and 100-mg/ml streptomycin under normal or H/R conditionsover three 24-h cycles. The migratory distance of EVTs e defined as the length fromthe cell column base to the outgrowth tip e was measured over the full duration ofthe three 24-h cycles with ImageJ software. Then, the differences among the groupswere calculated. The experiment was repeated three times, and four explants wereincluded per group each time.

2.9. Migration and invasion assays

Invasion assays were performed to assess the effects of various pretreatmentson trophoblast invasion in 24-well plates with Matrigel (BD Biosciences)-coatedTranswell inserts as described previously [23]. Migration assays were performedusing the insert without Matrigel coating. HTR8/SVneo cells (1 � 105) with(including siRNA and SB) or without (the control group) treatments were addedto each insert with 200 ml of serum-free medium plated in the upper compart-ment, and 600 ml of medium with 10% FBS in the lower chamber, and thencultured for 48 h. The H/R group was treated under 37 �C for 48 h (two 24-hcycles of 8 h of hypoxia and 16 h of reoxygenation). The inserts were fixedin methanol and dyed with crystal violet. The invaded and migratory cellswere observed and recorded under a microscopy (Olympus IX51, Japan) at amagnification of 200� in 10 random fields. All experiments were repeated intriplicate.

2.10. Immunofluorescence staining

HTR8/SVneo cells with different treatments were fixed in ice methanol andblocked with 1% bovine serum albumin (BSA). The primary antibody used was theanti-LAMA4 antibody (1:100). Then, the cells were incubated with a fluoresceinisothiocyanate (FITC) secondary antibody (Zhongsan Golden Bridge Crop) for 1 h.The nuclei were stained with propidium iodide (PI, 3 mg/mL) dye for 8e10 min.Images were acquiredwith an Olympus BMX-60microscope equippedwith a cooledcharge-coupled device sensi-camera (Cooke, Auburn Hills, MI, USA) and Slidebooksoftware (Intelligent Imaging Innovations, Denver, CO, USA).

Table 2Clinical characteristics of the preeclamptic and control groups.

Category Preeclampsia Control

2.11. Tube formation

BD Matrigel was diluted with a serum-free medium at a ratio of 1:2 with cooledpipettes and distributed in a 24-well plate (150 ml per well) on ice. After the Matrigelmixture solidified, HTR8/SVneo cells (1.0� 105 per well) after various pretreatmentswere gently added to each of the triplicate wells followed by normal culture or H/Rintervention for 48 h. Over this experimental period, the plates were monitored bycapturing digital images from each well (�200). At the end of the experimentalperiod, we measured and summed the tube perimeters (mm) of all tube-likestructures across five randomly selected microscopic fields (�40) and thendivided by five to determine a mean total perimeter (mm) per field. The tube for-mation index was then calculated by dividing this mean total perimeter (mm) perfield by the square millimeter area of the field [25].

Number 22 20Patient age (years) 27.25 ± 2.38 27.59 ± 2.54Gestational age (weeks) 35.82 ± 1.40** 38.65 ± 1.04Body mass index (BMI; kg/m2) 24.97 ± 1.07 25.62 ± 0.79Proteinuria (g/24 h) 2.32 ± 0.13** 0.04 ± 0.01Systolic blood pressure (mm Hg) 150.14 ± 6.18** 99.75 ± 9.24Diastolic blood pressure (mm Hg) 106.91 ± 8.52** 73.5 ± 9.19Neonatal birth weight (g) 2338.05 ± 111.45** 3405 ± 187.18Placental weight (g) 436.14 ± 18.14** 524.95 ± 10.29

*p < 0.05, **p < 0.01.

2.12. Cell viability assay

Cell viability was examined by an MTT assay. Briefly, 5 � 103 cells with varioustreatments (as described before) were seeded in 96-well plates for 44 h. Then, theculture medium was changed, and 100-ml MTT reagent (3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide; Apllygen Corp., Beijing, China) was added.The MTT reagent was gently removed 4 h later, and 100-ml DMSO was added toeach well. The optical density (OD) was measured at an absorbance of 570 nmusing a microplate reader (Model 550; BioRad). Experiments were performed intriplicate.

2.13. Flow cytometric analysis of apoptosis

The apoptotic rates of the various pretreated cells were measured by an AnnexinV-FITC and PI Apoptosis Detection Kit (Key-GenBiotech, Jiangsu, China) as previouslydescribed [25]. Cells were seeded on six-well plates and cultured under normal or H/R conditions for 48 h. Then, cells were harvested, washed with phosphate-bufferedsaline (PBS), and stained with a binding buffer. Apoptosis rates were quantified witha FACS Vantage SE flow cytometer (BD Biosciences, San Jose, CA, USA). Cells thatexcluded PI and were positive for Annexin V-FITC binding were deemed apoptoticcells.

2.14. Enzyme-linked immunosorbent assay procedure

The concentrations of LAMA4 in serum from healthy pregnant (N ¼ 20) and PE(N ¼ 20) women were measured by a commercial enzyme-linked immunosorbentassay (ELISA) kit (R&D Systems, Abingdon, UK) according to the manufacturer'sprotocol. Each experiment was performed at least three times with triplicatesamples.

2.15. DNA methylation analysis

Genomic DNA samples obtained from the normal and PE placentas were purifiedusing DNAzol (Invitrogen). Purified DNA was treated with sodium bisulfite (Sigma,Phoenix, USA), and then it was analyzed by bisulfite-sequencing PCR (BSP) ormethylation-specific PCR (MSP) as previously described [26]. The base sequenceof LAMA4 was as follows: GAGGGAGAGAGAGGCTGGAACTCTCCCTTACTCTCGCACTTTCCCTTTATTTCCTAAGCTGCTGGTTGCGCAGCCACCTCGGGATACTGCACACGGAGAGGAGGGAAAATAAGCGAGGCACCGCCGCACCACGCGGGAGACCTACGGAGACCCACAGCGCCCGAGCCCTGGAAGAGCACTACTGGATGTCAGCGGAGAAATGGCTTTGAGCTCAGCCTGGCGCTCGGTTCTGCCTCTGTGGCTCCTCTGGAGCGCTGCCTGCTCCCGCGCCGCGTCCGGGGACGACAACGCTTTTCCTTTTGACATTGAAGGGAGCTCAGCGGTTGGCAGGCAAGACCCGCCTGAGACGAGCGAACCCCGCGTGGCTCTGGGACGCCTGCCGCCTGCGGCCGAGGTACAGTGTCCCTGCCATTGCCACC. The methylation-specific primers are InF: 50- GAGGGA-GAGAGAGGTTGGAAT and InR: 50- AATAACAATAACAAAAACACTATACCTC. Amplifiedbisulfite PCR products were subcloned into a TA vector system (Promega) accordingto the manufacturer's instructions. DNA sequencing was performed on five indi-vidual clones (Sangong, China). The PCR products were confirmed by agarose gelelectrophoresis and visualized using ethidium bromide staining.

2.16. Statistical analysis

Each experiment was performed in triplicate. All values are expressed as themean ± standard error of mean (SEM). The data were analyzed for statistical sig-nificance using GraphPad Prism software. Independent t-tests were used for inter-group comparisons of continuous variables. Statistical differences among multiplegroups were evaluated by one-way analysis of variance (ANOVA) followed by theleast significant difference multiple-comparisons test as appropriate. P-value of<0.05 was considered to be statistically significant.

3. Results

3.1. Clinical characteristics

Differences in these parameters between the healthy pregnantwomen and PE women were evaluated. The women in earlypregnancy were similar in age, mean arterial pressure (MAP), andbody mass index (BMI). The characteristics of these patients in thethird trimester are shown in Table 2. The two groups were similar inage, BMI, and parity (%); however, it was not possible to match thegestational ages between PE patients and normal women on ac-count of the differing dates of miscarriage between the two groups.

Fig. 1. Dynamic LAMA4 Expression in Placentas. (A) Immunohistochemistry (IHC) of LAMA4 in human villi. (B) CK7 expression in human villi. (C and I) Negative controls withnonimmune rabbit IgG. (D) LAMA4 strongly expressed in human decidua in the first trimester. (E) HLA-G expression from the identical sample of human decidua. (F) CK7 expressionin human decidua. (G) LAMA4 expression was moderate in the term placenta. (H) LAMA4 was rarely expressed in the preeclamptic placentas. Scale bar ¼ 100 mm. Abbreviations:CTB, cytotrophoblasts; STB, syncytiotrophoblasts; EVT, extravillous trophoblasts, PE, preeclamptic.

N. Shan et al. / Placenta 36 (2015) 809e820812

Women with PE had significantly higher MAP and proteinuria aswell as significantly lower mean birth weight and placental weightcompared with healthy women (Table 1).

3.2. LAMA4 expressed in placental trophoblasts and endothelialcells during the first trimester

We first evaluated LAMA4 protein expression in human pla-centas at different stages of pregnancy using IHC. We used HLA-Gstaining to identify EVT cells (Fig. 1E) and CK7 (Fig. 1B) fortrophoblast and stromal cells, respectively. LAMA4 was intensely

Fig. 2. LAMA4 Expression in Placentas and Maternal Serum. (A, B) LAMA4 expression was ddetermined by Western blotting (n ¼ 3, p < 0.01) and real-time RT-PCR (n ¼ 3, p < 0.01).preeclamptics (N ¼ 20) and healthy controls (N ¼ 20, p > 0.05). Abbreviations: PE, preecla

and specifically expressed in trophoblasts and endothelial cells,including EVT cells during the first trimester (Fig. 1A and D). Inmaternal decidua, LAMA4 was strongly stained in glandularepithelium and some invading EVT cells, and it was found to bemoderately expressed in decidual cells (Fig. 1D). In the term pla-centas, LAMA4 was also found to be expressed in trophoblasts andendothelial cells (Fig. 1G), but it showed reduced expression in PEplacentas (Fig. 1H).

Based on Western blotting and qRT-PCR, the expression ofLAMA4 in term placentas (N ¼ 20) was significantly higher thanthat in PE (N ¼ 22). As shown in Fig. 2, both the mRNA (n ¼ 3,

ecreased in placentas of preeclamptics (N ¼ 22) compared to healthy controls (N ¼ 20)(C) LAMA4 expression in the maternal serum was not significantly different betweenmptic.

N. Shan et al. / Placenta 36 (2015) 809e820 813

p < 0.001, Fig. 2A) and protein (n ¼ 3, p < 0.001, Fig. 2B) levels weresignificantly lower in PE placentas compared with the controlgroup. However, the concentration of LAMA4 protein in the serumof patients (N ¼ 20) with PE (N ¼ 20) was not significantly differentfrom the healthy controls (p > 0.05; Fig. 2C).

3.3. LAMA4 siRNA inhibits invasion, migration, and tube formation

siRNA targeting LAMA4 significantly decreased LAMA4 proteinexpression in HTR8/SVneo cells as validated by Western blotting(n ¼ 3, p < 0.001; Fig. 3A and B), qRT-PCR (n ¼ 3, p < 0.001; Fig. 3C),and IF (Fig. 7). Compared with the control group, LAMA4 siRNAsignificantly decreased the percentage of cells that invaded (n ¼ 3,p < 0.001; Fig. 4A and C) or migrated (n ¼ 3, p < 0.001; Fig. 4B andD) to the other side of the filter.

MMP2 and MMP9 have been shown to play important roles introphoblast invasion by degrading the ECM [27]. We furtherexamined MMP2 and MMP9 and their specific inhibitors TIMP2and TIMP1, respectively, by Western blotting. Consistent with theinvasive experimental findings, we found that the expressionlevels of MMP2 (n ¼ 3, p < 0.001) and MMP9 (n ¼ 3, p < 0.001)decreased (Fig. 3A and D), while expressions of TIMP2 (n ¼ 3,p < 0.001) increased, after LAMA4 siRNA transfection, but therewas no significant change in TIMP1 expression (n ¼ 3, p > 0.05)(Fig. 3F and G).

We also observed that tube formation significantly decreasedafter LAMA4 siRNA transfection for 48 h (n¼ 3, p < 0.01; Fig. 6A andB), and it was not secondary to reduced viability and/or apoptosis(n ¼ 3, p > 0.05; Fig. 5AeC).

3.4. H/R conditions suppress invasion, migration, and tubeformation

Under H/R conditions, we observed significantly decreased cellinvasion and migration accompanied by a lower expression ofLAMA4. Western blotting showed that LAMA4 protein was down-regulated by approximately 70.27% in HTR8/SVneo cells under H/Rcompared with untreated control (n ¼ 3, p < 0.001; Fig. 3A and B)that was validated by IF (Fig. 7). qRT-PCR assay also showed thatLAMA4mRNA levels were significantly reduced after H/R treatment(n ¼ 3, p < 0.001; Fig. 3C). The results also revealed that expression

Fig. 3. Expression of LAMA4, MMP2, MMP9, TIMP1, and TIMP2 after Various Treatments. (Ablotting. (B, C) LAMA4 expression was decreased after various interventions as compareRepresentative immunoblotting of MMP2, MMP9, TIMP1, and TIMP2 normalized to b-actin

levels of MMP2 (n ¼ 3, p < 0.01) and MMP9 (n ¼ 3, p < 0.05)decreased (Fig. 3A, D, and E), while those of TIMP1 (n¼ 3, p < 0.001)and TIMP2 (n ¼ 3, p < 0.001) increased, under H/R conditions(Fig. 3A, F, and G). As shown in Fig. 4, the results showed that, afterexposure to the H/R environment for 48 h, both the invasion(reduced 49.71%) and migration (reduced 45.64%) of HTR8/SVneocells decreased compared with the control group (both n ¼ 3,p < 0.001).

Next, we used flow cytometric analysis to quantitatively inves-tigate the effect of H/R on apoptosis in HTR8/SVneo cells. Fig. 5shows that the H/R environment dramatically raised cellapoptosis. The mean apoptotic rate for HTR8/SVneo cells in the H/Rand control groups were 17.20 ± 3.55% and 1.888 ± 0.31%, respec-tively (n ¼ 3, p < 0.001; Fig. 5A and C). As revealed in Figs. 5B and6A, C, and E, a reduction in tube formation and viability of HTR8/SVneo cells was induced by H/R intervention (p < 0.001). As ourflow cytometric analysis of apoptosis revealed that H/R conditionsdecreased the cell number by 17.20 ± 3.55%, we added an additional18% cells to the cell invasion/migration/tube formation assays un-der H/R conditions in order to control for the confounding higherapoptosis rate under H/R conditions. Specifically, HTR8/SVneo cells(1.18 � 105, not 1.00 � 105) were added to each insert in the uppercompartment during Transwell assays under H/R conditions.Despite this 18% increase in cell count under H/R conditions, wefound that the cell invasion/migration/tube formation stilldecreased, indicating that H/R conditions decrease cell invasionindependent of the higher apoptosis rate under H/R conditions(Supplement 1).

After 72-h culture in vitro, LAMA4 siRNA (119.37 ± 38.42 mm)and H/R (70.96 ± 20.14 mm)-treated explants displayed asignificant reduction in the outgrowth and distance of migrationcompared with the control group (241.39 ± 24.34 mm) (p < 0.001)(Fig. 8).

3.5. Inhibiting the p38 MAPK signaling pathway suppresses LAMA4expression

To explore the mechanism underlying LAMA4 activity introphoblasts, we chose the classic oxidative stress signaling p38MAPK pathway. We noted that, compared to the control group,the p38 inhibitor SB203580 markedly reduced LAMA4 expression

) Expression of LAMA4, MMP2, MMP9, TIMP1, and TIMP2 were examined by Westernd to the control group according to Western blotting and real-time RT-PCR. (DeG)and comparisons of protein concentrations after various treatments.

Fig. 5. Effects of Different Treatments on HTR8/SVneo Cell Viability and Apoptosis. (A) Flow cytometric analysis of apoptosis across different treatments, including normal controls,H/R, LAMA4 silencing, and SB. (B) MTT assay showed no significant differences in viability after LAMA4 silencing and SB treatment by analysis of variance (ANOVA). (C) Significantdifferences in cell apoptosis were observed after H/R treatment (n ¼ 3, p < 0.001).

Fig. 4. Invasive and Migratory Abilities of HTR8/SVneo Cells Decreased with LAMA4 Silencing and H/R Conditions. (A, B) The invasive and migratory abilities of HTR8/SVneo cellsdecreased after exposure to H/R conditions and LAMA4 silencing (both n ¼ 3, p < 0.001). (C, D) Statistical bar graphs of the invasion and migration assays. Data were analyzed byanalysis of variance (ANOVA) to assess significant differences. Original images were taken at 200� magnification.

N. Shan et al. / Placenta 36 (2015) 809e820814

N. Shan et al. / Placenta 36 (2015) 809e820 815

in HTR8/SVneo cells (n ¼ 3, p < 0.001, Fig. 9A and B). LAMA4silencing also significantly decreased the expression of p-p38(n ¼ 3, p < 0.001, Fig. 9A and B). In addition, we investigatedthe expression of two other MAPK proteins e p-JNK and p-ERK e

and we found that silencing LAMA4 expression decreases the levelof p-JNK (n ¼ 3, p < 0.01) and p-ERK (n ¼ 3, p < 0.05) (Fig. 9A, C,and D).

Moreover, H/R conditions induced a nearly twofold higher levelof p-p38 in HTR8/SVneo cells (Fig. 9A and B). We also investigatedp-JNK and p-ERK protein expressions after H/R exposure, and weshowed that H/R treatment induced significantly higher p-ERK

Fig. 6. Tube Formation of HTR8/SVneo Cells under Different Treatments. (A) Nontreated gstructures. Bars ¼ 100 mm. (E) Bar graphs showing quantification in the tube formation as[2]). **p < 0.01 compared with the nontreated control group.

expression (n ¼ 3, p < 0.001) along with significantly decreased p-JNK expression (n ¼ 3, p < 0.01) (Fig. 9A, C, and D).

3.6. DNA methylation of LAMA4 between normal and PE placentas

It has been reported that epigenetic silencing of the laminin a3chain may lead to the inability to synthesize the basement mem-brane, and it may affect cancer cell invasion [28], sowe investigatedthe status of LAMA4 DNAmethylation between normal term and PEplacentas using MSP and BSP. All samples showed a hyper-methylated status of the LAMA4 promoter, and they revealed that

roup. (B) siLAMA4 group. (C) H/R group. (D) SB group. Arrows indicate capillary-likesay. The tube formation index was expressed as tube perimeter (mm) per area (mm

N. Shan et al. / Placenta 36 (2015) 809e820816

the LAMA4 promoter was not significantly altered in PE placentascompared with that in the normal ones (n ¼ 15 in each group,1.293 ± 0.2699% vs. 1.414 ± 0.2550%, respectively, p ¼ 0.7463;Fig. 10).

4. Discussion

In the present study, we provided several lines of evidenceto demonstrate that LAMA4 promotes trophoblast invasion,migration, and angiogenesis, providing novel insight into thepathogenesis of PE. Laminin and fibronectin have been previouslydemonstrated in the pericellular basement membrane of humandecidual cells and placental villi [29]. In the current study, LAMA4was localized in human first-trimester placental villi to syncytio-trophoblast cells and in the decidua to EVT cells. This finding sug-gests that LAMA4 has a role in regulating trophoblast function andplacentation.

Trophoblast research over the past few decades has underlinedthe striking similarities between the proliferative, migratory, and

Fig. 7. Immunofluorescent Detection of LAMA4 Proteins and Nuclei in HTR8/SVneo Cells undLAMA4 expression (green) decreased after siRNA transfection. (GeI) LAMA4 expression (greSB treatment. Cell nuclei stained by propidium iodide (red) (n ¼ 3 in triple). Original imag

invasive properties of placental cells and those of cancer cells [20].MMPs play an important role in the invasion process of these cells,as they degrade ECM proteins so that the cell is free to move [30].The activity of these proteases involves interaction with TIMPs thatregulate the activating proteolytic cleavage of the pro-domain ofthe MMPs in the extracellular space [31]. Here, we focused onMMPs because they are highly linked with cell invasion and tissueremodeling in a variety of biological systems, including EVT cells[32]. LAMA4 siRNAwas shown to decrease the expression of MMP2and MMP9, which was accompanied by an increased TIMP-2expression, in HTR8/SVneo cells and extravillous explants, sug-gesting that LAMA4 plays a key role in trophoblast invasion andmigration.

Reduced invasion of trophoblasts and incomplete conversion ofspiral arteries are key pathological features of PE [33]. Previousstudies have revealed that ROS concentrations are higher in PEplacentas than in normotensive placentas [6]. Excessive oxidativestress leads to higher cell apoptosis, inhibition of proliferation, andincreased invasion of trophoblasts [34]. In this study, HTR8/SVneo

er Different Treatments. (AeC) LAMA4 expression (green) in normal control cells. (DeF)en) decreased after 48 h of H/R. (JeL) LAMA4 expression (green) was not changed afteres were taken at 400� magnification.

N. Shan et al. / Placenta 36 (2015) 809e820 817

cells and the extravillous explant culture model were exposed tooxidative stress-inducing H/R conditions to mimic PE conditions.Both HTR8/SVneo cells and extravillous explants displayed a dra-matic decrease in invasion and migration when exposed to H/Raccompanied with reduced LAMA4 expression. Although theobserved decreases in cell invasion and migration may be partiallyattributable to increased apoptosis and decreased cell viabilityafter H/R exposure, our previous trophoblast research supports thatH/R conditions have adverse effects on invasive and migratoryfunctions of trophoblasts independent of apoptosis and cellviability [23,35]. Moreover, we did not find a negative association

Fig. 8. Placental Villous Explants after 72-h Culture In Vitro. (A) Nontreated group. (B) siLAMtrophoblast cell migration distance in placental villous explants (n ¼ 5). **p < 0.01 compar

between serum LAMA4 expression and PE. Therefore, thedecreased LAMA4 expression was closely associated with excessiveoxidative stress in the PE placenta but not with secretion into theserum.

Angiogenesis is critical to successful fetal outcomes, as theplacental blood flow is dependent on placental vascularization.Lack of placental vascular development may contribute to inade-quate CTB invasion as observed in PE [36]. Laminins are the mainproteins involved in the angiogenesis of endothelial cells, so weinvestigated their function in trophoblast angiogenesis. We foundthat HTR8/SVneo cells showed a dramatic reduction in tube

A4 group. (C) H/R group. (D) SB group. Bars ¼ 100 mm. (E) Graphical representation ofed with nontreated control group.

N. Shan et al. / Placenta 36 (2015) 809e820818

formation under H/R conditions or LAMA4 silencing, indicatingLAMA4's role in placental angiogenesis.

Some studies have shown that MAPKs are involved in lamininsignaling [9]. MAPK cascades transduce various spectrums ofextracellular and intracellular stimuli in gene expression, andthey are essential for invasion, proliferation, differentiation, andapoptosis in trophoblast cells [37e39]. Extracellular signal-regulated kinase1/2 (ERK1/2), p38 MAPK, and JNK are three keyMAPK intermediaries, which are parallel to one another [40]. Here,we demonstrated that LAMA4 silencing significantly reduced theexpression of p-p38, p-JNK, and p-ERK, and p38 inhibition bySB203580 resulted in significant LAMA4 downregulation. Thesetwo findings suggest that there is a positive relationship betweenLAMA4 expression and MAPK signaling activation in HTR8/SVneocells. However, after exposure to H/R conditions, LAMA4 down-regulation was accompanied by significantly higher p-p38expression, significantly higher p-ERK expression, and signifi-cantly decreased p-JNK expression. These findings suggest that H/R decouples the positive relationship between LAMA4 expressionand p38 and ERK activations in HTR8/SVneo cells. Obviously,further investigation on the relationships between LAMA4 (andother laminins) and the major signaling pathways involved in in-vasion, viability, proliferation, differentiation, and apoptosis introphoblast cells under both normal and H/R conditions is stillneeded.

Epigenetic modification of genes in placental trophoblasts hasbeen reported in previous studies [41]. DNA methylation of sometumor-suppressor genes such as Maspin, RASSF1A, and adenoma-tous polyposis coli (APC) have been investigated in human pla-centas [42e44]. In humans, the post-replicative DNA modification

Fig. 9. MAPK Signaling Pathway under Different Treatments. (A) Protein expressions of pimmunoblotting of p-p38, p-JNK, and p-ERK normalized with b-actin and comparisons of p

occurs predominantly on cytosines that precede a guanosine in theDNA sequence (the CpG dinucleotide) [45]. These dinucleotides canbe clustered in small stretches of DNA termed CpG islands that areoften associated with promoter regions. The importance of DNAmethylation for gene expression, and especially for transcriptionalsilencing, has been mainly associated with CpG islands in gene-promoter regions that are non-methylated [46]. In the presentstudy, we used the MSP and BSP methods to investigate the CpGstatus of LAMA4, but we failed to find any significant difference inmethylation between normal and PE placentas. Further studies,especially on primary cells, are needed to elucidate themechanismsunderlying epigenetic modifications of the LAMA4 gene.

In conclusion, the present study supports LAMA4's role in pro-moting invasion/migration and angiogenesis in the trophoblast. H/R conditions significantly decrease LAMA4 expression and appearto decouple the positive relationship between LAMA4 expressionand p38 and ERK activations, suggesting one possible pathophysi-ological mechanism underlying PE. Further studies are needed toinvestigate the relationships between LAMA4 (and other laminins)and the major signaling pathways involved in trophoblast invasion,viability, proliferation, differentiation, and apoptosis as well asepigenetic modifications associated with LAMA4 expression.

Author contributions

N.S, XM.Z, QY.D, and NL.Y were involved in the acquisition ofdata. XQ.X and H.Z analyzed and interpreted the data. X.L, XR.L, andY.C drafted the article. C.T and HB.Q revised the article. Hongbo Qidesigned the paper and approved the version to be published. Allauthors approved this version for publication.

-p38, p-JNK, and p-ERK were examined by Western blotting. (BeD) Representativerotein concentrations after various treatments.

Fig. 10. DNA Methylation of Placental LAMA4. CpG sites from þ5312 to þ5669 bps (relative to the transcription start site) of LAMA4 were included in the bisulfite genomic sequencing assay. A total of 33 CpG sites in the normal andpreeclamptic (PE) placentas were included, respectively. Each square represents an individual CpG site with the yellow portions representing methylation-positive results and the blue portions representing methylation-negativeresults. The sizes of the yellow portions indicate fractional methylation. The data are presented as the mean values from the placental samples.

N.Shan

etal./

Placenta36

(2015)809

e820

819

N. Shan et al. / Placenta 36 (2015) 809e820820

Conflicts of interest

The authors declare that there are no conflicts of interestregarding the publication of this paper.

Acknowledgments

We thank Dr. Yubin Ding and Dr. Qian Zhang for their sugges-tions regarding the design of the study. We are grateful for theexcellent technical assistance from the Key Laboratory for MajorObstetric Diseases of Guangdong Province and the Key Laboratoryof Diagnostic Medicine designated by theMinistry of Education andChongqing Medical University. This work was supported by theNational Natural Science Foundation of China (grant nos. 81070502,81170585, 81100444, 81300509, and 81300508) and the NationalKey Clinical Department of China.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.placenta.2015.04.008.

References

[1] Khan KS,Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO analysis of causesof maternal death: a systematic review. Lancet 2006 Apr 1;367(9516):1066e74.

[2] Ji L, Brkic J, Liu M, Fu G, Peng C, Wang YL. Placental trophoblast cell differ-entiation: physiological regulation and pathological relevance to preeclamp-sia. Mol Asp Med 2013 Oct;34(5):981e1023.

[3] Zhou Y, Fisher SJ, Janatpour M, Genbacev O, Dejana E, Wheelock M, et al. Humancytotrophoblasts adopt a vascular phenotype as they differentiate. A strategy forsuccessful endovascular invasion? J Clin Invest 1997 May 1;99(9):2139e51.

[4] Pennington KA, Schlitt JM, Jackson DL, Schulz LC, Schust DJ. Preeclampsia: mul-tiple approaches for a multifactorial disease. Dis ModelMech 2012 Jan;5(1):9e18.

[5] Graham CH, Forsdike J, Fitzgerald CJ, Macdonald-Goodfellow S. Hypoxia-mediated stimulation of carcinoma cell invasiveness via upregulation ofurokinase receptor expression. Int J Cancer 1999 Feb 9;80(4):617e23.

[6] Zhou X, Zhang GY, Wang J, Lu SL, Cao J, Sun LZ. A novel bridge betweenoxidative stress and immunity: the interaction between hydrogen peroxideand human leukocyte antigen G in placental trophoblasts during pre-eclampsia. Am J Obstet Gynecol 2012 May;206(5)(447):e7e16.

[7] Engvall E, Wewer UM. Domains of laminin. J Cell Biochem 1996 Jun 15;61(4):493e501.

[8] Malinda KM, Kleinman HK. The laminins. Int J Biochem Cell Biol 1996Sep;28(9):957e9.

[9] Givant-Horwitz V, Davidson B, Reich R. Laminin-induced signaling in tumorcells: the role of the M(r) 67,000 laminin receptor. Cancer Res 2004 May15;64(10):3572e9.

[10] Iivanainen A, Sainio K, Sariola H, Tryggvason K. Primary structure and expressionof a novel human laminin alpha 4 chain. FEBS Lett 1995May 29;365(2e3):183e8.

[11] Lefebvre O, Sorokin L, Kedinger M, Simon-Assmann P. Developmentalexpression and cellular origin of the laminin alpha2, alpha4, and alpha5chains in the intestine. Dev Biol 1999 Jun 1;210(1):135e50.

[12] Petajaniemi N, Korhonen M, Kortesmaa J, Tryggvason K, Sekiguchi K,Fujiwara H, et al. Localization of laminin alpha4-chain in developing and adulthuman tissues. J Histochem Cytochem 2002 Aug;50(8):1113e30.

[13] Kosanam H, Prassas I, Chrystoja CC, Soleas I, Chan A, Dimitromanolakis A, et al.Laminin, gamma 2 (LAMC2): a promising new putative pancreatic cancerbiomarker identified by proteomic analysis of pancreatic adenocarcinomatissues. Mol Cell Proteomics 2013 Oct;12(10):2820e32.

[14] Kuratomi Y, Sato S, Monji M, Shimazu R, Tanaka G, Yokogawa K, et al. Serumconcentrations of laminin gamma2 fragments in patients with head and necksquamous cell carcinoma. Head Neck 2008 Aug;30(8):1058e63.

[15] Aghcheli K, Parsian H, Qujeq D, Talebi M, Mosapour A, Khalilipour E, et al.Serum hyaluronic acid and laminin as potential tumor markers for uppergastrointestinal cancers. Eur J Intern Med 2012 Jan;23(1):58e64.

[16] Rosa H, Parise ER. Is there a place for serum laminin determination in patientswith liver disease and cancer? World J Gastroenterol 2008 Jun 21;14(23):3628e32.

[17] DeHahn KC, Gonzales M, Gonzalez AM, Hopkinson SB, Chandel NS,Brunelle JK, et al. The alpha4 laminin subunit regulates endothelial cell sur-vival. Exp Cell Res 2004 Mar 10;294(1):281e9.

[18] Doi M, Thyboll J, Kortesmaa J, Jansson K, Iivanainen A, Parvardeh M, et al.Recombinant human laminin-10 (alpha5beta1gamma1). Production, purifi-cation, and migration-promoting activity on vascular endothelial cells. J BiolChem 2002 Apr 12;277(15):12741e8.

[19] Vainionpaa N, Lehto VP, Tryggvason K, Virtanen I. Alpha4 chain laminins arewidely expressed in renal cell carcinomas and have a de-adhesive function.Lab Invest 2007 Aug;87(8):780e91.

[20] Ferretti C, Bruni L, Dangles-Marie V, Pecking AP, Bellet D. Molecular circuitsshared by placental and cancer cells, and their implications in the prolifera-tive, invasive and migratory capacities of trophoblasts. Hum Reprod Update2007 Mar-Apr;13(2):121e41.

[21] Hung TH, Chen SF, Li MJ, Yeh YL, Hsieh TT. Differential effects of concomitantuse of vitamins C and E on trophoblast apoptosis and autophagy betweennormoxia and hypoxia-reoxygenation. PLoS One 2010;5(8):e12202.

[22] Xiong Y, Liebermann DA, Tront JS, Holtzman EJ, Huang Y, Hoffman B, et al.Gadd45a stress signaling regulates sFlt-1 expression in preeclampsia. J CellPhysiol 2009 Sep;220(3):632e9.

[23] Yang Z, Bai B, Luo X, Xiao X, Liu X, Ding Y, et al. Downregulated Kruppel-likefactor 8 is involved in decreased trophoblast invasion under hypoxia-reoxygenation conditions. Reprod Sci 2014 Jan;21(1):72e81.

[24] Zhang Q, Chen Q, Lu X, Zhou Z, Zhang H, Lin HY, et al. CUL1 promotes trophoblastcell invasion at the maternal-fetal interface. Cell Death Dis 2013;4:e502.

[25] Luo X, Yao ZW, Qi HB, Liu DD, Chen GQ, Huang S, et al. Gadd45alpha as anupstream signaling molecule of p38 MAPK triggers oxidative stress-inducedsFlt-1 and sEng upregulation in preeclampsia. Cell Tissue Res 2011Jun;344(3):551e65.

[26] Guo X, Jing C, Li L, Zhang L, Shi Y, Wang J, et al. Down-regulation of VEZT geneexpression in human gastric cancer involves promoter methylation and miR-43c. Biochem Biophys Res Commun 2011 Jan 14;404(2):622e7.

[27] Staun-Ram E, Goldman S, Gabarin D, Shalev E. Expression and importance ofmatrix metalloproteinase 2 and 9 (MMP-2 and -9) in human trophoblast in-vasion. Reprod Biol Endocrinol 2004 Aug 4;2:59.

[28] Ii M, Yamamoto H, Taniguchi H, Adachi Y, Nakazawa M, Ohashi H, et al. Co-expression of laminin beta3 and gamma2 chains and epigenetic inactivationof laminin alpha3 chain in gastric cancer. Int J Oncol 2011 Sep;39(3):593e9.

[29] Wewer UM, Faber M, Liotta LA, Albrechtsen R. Immunochemical and ultra-structural assessment of the nature of the pericellular basement membrane ofhuman decidual cells. Lab Invest 1985 Dec;53(6):624e33.

[30] Yan C, Boyd DD. Regulation of matrix metalloproteinase gene expression.J Cell Physiol 2007 Apr;211(1):19e26.

[31] Murphy G, Houbrechts A, Cockett MI, Williamson RA, O'Shea M, Docherty AJ.The N-terminal domain of tissue inhibitor of metalloproteinases retains met-alloproteinase inhibitory activity. Biochemistry 1991 Aug 20;30(33):8097e102.

[32] Anacker J, Segerer SE, Hagemann C, Feix S, Kapp M, Bausch R, et al. Humandecidua and invasive trophoblasts are rich sources of nearly all human matrixmetalloproteinases. Mol Hum Reprod 2011 Oct;17(10):637e52.

[33] Verlohren S, Geusens N, Morton J, Verhaegen I, Hering L, Herse F, et al. In-hibition of trophoblast-induced spiral artery remodeling reduces placentalperfusion in rat pregnancy. Hypertension 2010 Aug;56(2):304e10.

[34] Seeho SK, Park JH, Rowe J, Morris JM, Gallery ED. Villous explant culture usingearly gestation tissue from ongoing pregnancies with known normal out-comes: the effect of oxygen on trophoblast outgrowth and migration. HumReprod 2008 May;23(5):1170e9.

[35] Li J, Luo X, Xiao X, Zhang X, Qi H, Liu X, et al. Decreased expression of Wiskott-Aldrich syndrome protein family verprolin-homologous protein 2 may beinvolved in the development of pre-eclampsia. Reprod Biomed Online 2014Jan;28(1):70e9.

[36] Carmeliet P, Jain RK. Molecular mechanisms and clinical applications ofangiogenesis. Nature 2011 May 19;473(7347):298e307.

[37] Fitzgerald JS, Busch S, Wengenmayer T, Foerster K, de la Motte T,Poehlmann TG, et al. Signal transduction in trophoblast invasion. ChemImmunol Allergy 2005;88:181e99.

[38] Daoud G, Amyot M, Rassart E, Masse A, Simoneau L, Lafond J. ERK1/2 and p38regulate trophoblasts differentiation in human term placenta. J Physiol 2005Jul 15;566(Pt 2):409e23.

[39] Perez-Perez A, Maymo J, Duenas JL, Goberna R, Calvo JC, Varone C, et al. Leptinprevents apoptosis of trophoblastic cells by activation of MAPK pathway. ArchBiochem Biophys 2008 Sep 15;477(2):390e5.

[40] Johnson GL, Lapadat R. Mitogen-activated protein kinase pathways mediatedby ERK, JNK, and p38 protein kinases. Science 2002 Dec 6;298(5600):1911e2.

[41] Grigoriu A, Ferreira JC, Choufani S, Baczyk D, Kingdom J, Weksberg R. Cellspecific patterns of methylation in the human placenta. Epigenetics 2011Mar;6(3):368e79.

[42] Dokras A, Coffin J, Field L, Frakes A, Lee H, Madan A, et al. Epigenetic regulationof maspin expression in the human placenta. Mol Hum Reprod 2006Oct;12(10):611e7.

[43] Chiu RW, Chim SS, Wong IH, Wong CS, LeeWS, To KF, et al. Hypermethylation ofRASSF1A in human and rhesus placentas. Am J Pathol 2007 Mar;170(3):941e50.

[44] Wong NC, Novakovic B, Weinrich B, Dewi C, Andronikos R, Sibson M, et al.Methylation of the adenomatous polyposis coli (APC) gene in human placentaand hypermethylation in choriocarcinoma cells. Cancer Lett 2008 Sep8;268(1):56e62.

[45] Herman JG, Baylin SB. Gene silencing in cancer in association with promoterhypermethylation. N Engl J Med 2003 Nov 20;349(21):2042e54.

[46] Bird A. DNA methylation patterns and epigenetic memory. Genes Dev 2002Jan 1;16(1):6e21.