lable at ScienceDirect

Placenta 34 (2013) 1014e1019

Contents lists avai

Placenta

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

The role of leukemia inhibitory factor in tubal ectopic pregnancy

T. Krishnan a, b, 1, A. Winship a, c, 1, S. Sonderegger a, 2, E. Menkhorst a, A.W. Horne d,J. Brown d, J.-G. Zhang e, N.A. Nicola e, S. Tong f, E. Dimitriadis a, c, *

a Prince Henry’s Institute of Medical Research, P.O. Box 5152, Clayton, Victoria 3168, Australiab Faculty of Medicine, Nursing & Health Sciences, Wellington Road, Monash University, Clayton, Victoria 3800, Australiac Department of Anatomy and Developmental Biology, Wellington Road, Monash University, Clayton, Victoria 3800, Australiad MRC Centre for Reproductive Health, University of Edinburgh, Queen’s Medical Research Institute, 47 Little France Crescent, Edinburgh EH16 4TJ,United Kingdome The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3052, Australiaf Department of Obstetrics and Gynaecology, University of Melbourne, Mercy Hospital for Women, 163 Studley Road, Heidelberg, Victoria 3084, Australia

a r t i c l e i n f o

Article history:Accepted 6 September 2013

Keywords:Leukemia inhibitory factorEctopicImplantationPlacentaEVTPregnancy

* Corresponding author. Prince Henry’s Institute o5152, Clayton, Victoria 3168, Australia. Tel.: þ61 3 9594

E-mail address: evdokia.dimitriadis@princehenrys1 Denotes equal contribution.2 Faculty of Medicine, Nursing and Health Sciences

Diseases (ACBD), Monash University, Alfred Centre,bourne, Victoria 3004, Australia.

0143-4004/$ e see front matter � 2013 Published byhttp://dx.doi.org/10.1016/j.placenta.2013.09.003

a b s t r a c t

Introduction: Ectopic pregnancy is unique to humans and a leading cause of maternal morbidity andmortality. The etiology remains unknown however factors regulating embryo implantation likelycontribute. Leukemia inhibitory factor (LIF) has roles in extravillous trophoblast adhesion and invasionand is present in ectopic implantation sites. We hypothesised that LIF facilitates blastocyst adhesion/invasion in the Fallopian tube, contributing to ectopic pregnancy.Methods: We immunolocalised LIF receptor (R) in tubal ectopic pregnancy (N ¼ 5). We used an oviductcell line (OE-E6/E7) to model Fallopian tube epithelial cells and a trophoblast spheroid co-culture model(HTR-8/SVneo cell line formed spheroids) to model blastocyst attachment to the Fallopian tube. Weexamined LIF signaling pathways in OE-E6/E7 cells by Western blot. The effect of LIF and LIF inhibition(using a novel LIF inhibitor, PEGLA) on first-trimester placental outgrowth was determined.Results: LIFR localised to villous and extravillous trophoblast and Fallopian tube epithelium in ectopicpregnancy. LIF activated STAT3 but not the ERK pathway in OE-E6/E7 cells. LIF stimulated HTR-8/SVneospheroid adhesion to OE-E6/E7 cells which was significantly reduced after PEGLA treatment. LIF pro-moted placental explants outgrowth, while co-treatment with PEGLA blocked outgrowth.Discussion: Our data suggests LIF facilitates the development of ectopic pregnancy by stimulating blas-tocyst adhesion and trophoblast outgrowth from placental explants. Ectopic pregnancy is usually diag-nosed after 6 weeks of pregnancy, therefore PEGLA may be useful in targeting trophoblast growth/invasion.Conclusion: LIF may contribute to the development of ectopic pregnancies and that pharmacologicallytargeting LIF-mediated trophoblast outgrowth may be useful as a treatment for ectopic pregnancy.

� 2013 Published by Elsevier Ltd.

1. Introduction

Ectopic pregnancy is the most common cause of maternalmortality in pregnancies during the first trimester, occurring at arate of 2% of all pregnancies in developed countries [1e3].Approximately 95% of ectopic pregnancies occur in the Fallopian

f Medical Research, P.O. Box4392; fax: þ61 3 9594 6125.

.org (E. Dimitriadis).

, Australian Centre for Blood99 Commercial Road, Mel-

Elsevier Ltd.

tube [4]. While the precise etiology of tubal ectopic pregnancy isunknown, factors critical for embryo implantation into the uterusmay also contribute to blastocyst implantation in the Fallopian tube[5].

LIF is a member of the interleukin-6 family of cytokines. It isindispensable for uterine blastocyst implantation in mice [6,7] andplays a critical role in implantation in women [8e10]. LIF is asecreted glycoprotein that signals via the gp130/LIFR complex toactivate the Janus tyrosine kinases (JAK). This, in turn, can activatedownstream signaling pathways, including the signal transducerand activator of transcription (STAT) pathway [11e15], or extra-cellular signal regulated kinase (ERK) [16]. In the human endome-trium and trophoblast, LIF primarily acts via STAT3 [17,18].

T. Krishnan et al. / Placenta 34 (2013) 1014e1019 1015

The blastocyst itself expresses LIFR and gp130 on the tro-phectoderm.We have previously demonstrated that LIF plays a rolein blastocyst adhesion to the endometrial epithelial cells [19e21].Additionally, LIF positively affects the development of the pre-implantation embryo in vitro [22,23]. LIF, LIFR and gp130 areexpressed in the Fallopian tube surface epithelium and LIF pro-duction has been reported to be increased at tubal ectopic im-plantation sites [24e27]. Together, this indicates a role for LIFsignaling in blastocyst attachment to the Fallopian tube epithelium,and the development of ectopic pregnancy.

Placental chorionic villous structures develop in human Fallo-pian tube ectopic implantation sites, as in uterine implantation[28]. LIF is important for trophoblast adhesion and invasion duringnormal placental development. LIF and LIFR mRNA are expressed inthe chorionic villi and decidua of first-trimester placenta inwomen[29,30]. LIF activates STAT3 in human primary invasive extravilloustrophoblast (EVT) cells and stimulates their adhesion to primaryendometrial epithelial cells and extracellular matrix proteins,including fibronectin and collagen IV, produced on the blastocystsurface [18]. LIF has also been shown to mediate EVT cell invasionrequired for the establishment of the placenta [18]. Thus, it ispossible that LIF may facilitate blastocyst adhesion and trophoblastinvasion into the Fallopian tube.

Given LIF’s role in uterine implantation and trophoblast inva-sion, and the presence of LIF at tubal ectopic implantation sites, wehypothesised that LIF facilitates blastocyst adhesion and tropho-blast invasion in the Fallopian tube to establish ectopic pregnancies.To address our hypothesis, we localised LIFR in human tubal ectopicimplantation sites.We examined the effects of LIF and LIF inhibitionusing a specialised LIF antagonist conjugated to polyethylene glycol(PEGLA) [31,32] on HTR-8/SVneo (trophoblast derived cell line) [33]spheroid adhesion to OE-E6/E7 (oviduct cell line) [34] cells andinvestigated whether this was mediated by pSTAT3. Using primaryfirst-trimester placentas we also studied the effect of LIF and LIFblockade on placental villous outgrowth.

2. Materials & methods

2.1. Tissue collection

2.1.1. Ectopic pregnancy implantation sitesWe collected fallopian tube biopsies from the ectopic implantation site from

participants (aged 18e45 years, N ¼ 5) undergoing surgery for an ectopic pregnancyat The Royal Infirmary Hospital, Edinburgh, Scotland. None of the women presentedacutely with hemodynamic shock, and all required serial serum beta-HCG and ul-trasound monitoring prior to diagnosis. None of the patients had a past history oftubal disease or ectopic pregnancy and no fetal hearts were seen on the diagnosticscans. Detailed characteristics of the patients are shown in Table 1. Written andinformed consent was obtained from all patients before sample collection. Ethicalapproval was obtained from Lothian Research Ethics Committee, Edinburgh, UnitedKingdom (04/S1103/20).

2.1.2. First-trimester placentaPlacental samples were collected from healthy women (n ¼ 6) undergoing first-

trimester termination of pregnancy (6e10 weeks) for psychosocial reasons. Tissueswere washed in 0.9% saline before transfer to DMEM/F12 medium 1:1 (Invitrogen).Informed consent was obtained from all participating patients. Ethics approval wasobtained from the Southern Health Human Research and Ethics committee, Victoria,Australia.

Table 1Clinical information for ectopic pregnancy implantation sites.

Sample hCG (IU/l) Serum progesterone (nmol/l) Gestational age (days)

1 5981 158.1 412 453 8.8 473 10,285 31.7 464 1082 23.9 525 508 7.1 44

2.2. Cell lines

The HTR-8/SVneo trophoblast cell line exhibits features of invasive trophoblastscells, such as HLA-G (EVT marker) and cytokeratin-7 [33]. Cells were maintained inRPMI medium (SigmaeAldrich) supplemented with 10% FCS at 37 �C, 5% CO2. Giventhe difficulty in obtaining primary Fallopian tube epithelial cells, we utilised the onlyavailable human oviduct cell line. Oviduct epithelial OE-E6/E7 cells [34] were a kindgift from Professor Calvin Lee (The University of Hong Kong). Cells were cultivated inDMEM/F12 medium containing 10% FCS at 37 �C, 5% CO2.

2.3. LIFR immunohistochemistry

Formalin fixed, paraffin wax-embedded sections were mounted on Snow CoatX-tra� charged slides (Surgipath Europe), de-waxed in xylene, rehydrated andsubjected to antigen retrieval by pressure cooking for 5 min in 10mM sodium citrate(pH 6.0), before blocking endogenous peroxidase with 3% hydrogen peroxidase(Sigma). An avidinebiotin block (Vector Laboratories) and protein block (Dako) wereperformed prior to overnight incubation with 2 mg/ml of rabbit anti-LIF-R antibody(Santa Cruz) or control rabbit IgG (Dako). Sections were incubated with biotinylatedsecondary antibody and ABC-Elite (Vector Laboratories). Positive immunostainingwas visualised using 3,3-diaminobenzidine (Vector Laboratories). Sections werecounterstained in Mayer’s Hematoxylin and mounted using Pertex (Cellpath PLC).

2.4. LIF and LIFR mRNA expression in OE-E6/E7 cells

Total RNA was isolated from cultured OE-E6/E7 cells using the RNeasy Minikit(QIAGEN) according to the manufacturer’s instructions. RNA samples were analysedby spectrophotometry (Nanodrop) using the absorbance ratio of A260/280 nm.cDNA was synthesised from total RNA (500 ng) using Superscript III reverse tran-scriptase (Invitrogen).

2.5. RT-PCR

PCR reactions were performed using PCR express machine (ThermoFisher Sci-entific) and GoTaq master mix (Promega) according to the manufacturer’s in-structions. OE-E6/E7 cDNAwas analysed for LIF, LIFR and 18s as previously described[35].

2.6. LIF enzyme-linked immunosorbent assay

A LIF ELISA (ELH-LIF-001; RayBiotech) was performed to quantify LIF secretionfrom OE-E6/E7 and HTR-8/SVneo cells. OE-E6/E7 and HTR-8/SVneo cells werecultured in 96-well plates in DMEM/F12 or RPMI media respectively, �10% FCS,before media was collected. Total protein was determined using BCA Protein AssayKit (Pierce). LIF was detected according to the manufacturer’s instructions.

2.7. STAT3 activation in OE-E6/E7 and HTR-8/SVneo cells

OE-E6/E7 and HTR-8/SVneo cells were serum starved for 24 h prior to treatment.HTR-8/SVneo cells were treated for 30 min with recombinant LIF (R&D Systems)diluted in 0.1% bovine serum albumin (BSA) (Sigma) in PBS (100 ng/ml). OE-E6/E7cells were originally treated for 30 min with 5, 50 or 100 ng/ml LIF. Subsequently,OE-E6/E7 cells were treated for 30 minwith either; vehicle control; LIF (100 ng/ml);PEGLA (100 ng/ml); or PEGLAþ LIF (cells pre-treated with PEGLA (100 ng/ml) for 1 hbefore the addition of LIF for 30 min (100 ng/ml)).

2.8. SDS-PAGE and Western blotting

Cells were washed with PBS and lysed in lysis buffer (50 mM Tris base, 150 mMNaCl, 2 mM EDTA, 2 mM EGTA, 25 mM NaF and 25 mM b-glycerophosphate; pH 7.5,2ml/ml protease inhibitor cocktail set; Calbiochem). Cellular proteinwas quantified bythe BCA Protein Assay Kit (Pierce). Equal amounts of protein were resolved on 8e10%sodium dodecyl sulfate (SDS)epolyacrylamide gel electrophoresis (PAGE) gels, trans-ferred to polyvinyl difluoride (PVDF) membranes (GE Healthcare), blocked with 5%non-fat dry milk in Tris-buffered saline (TBS) with 0.1% Tween-20 (Bio-Rad) beforebeing probed with polyclonal antibodies against pSTAT3 (1:1000), STAT3 (1:1000),pERK (1:1000) and b-actin (1:1000) (Cell Signaling) overnight at 4 �C, followed bythree wash steps. Membranes were incubated for 1 h at room temperature withsecondary antibodies (anti-rabbit IgG) conjugated to horse-radish peroxidase (HRP)(1:5000) (DakoCytomation) and signals were developed with enhanced chem-iluminescence Western blotting detection system reagent (Pierce).

2.9. Trophoblast spheroid-oviduct cell adhesion co-culture model

To determine the effect of LIF and LIF inhibition on the adhesive propertiesessential for the attachment of the blastocyst to the Fallopian tube, a co-culture modelwas established based on published methods [36]. OE-E6/E7 cells were grown toconfluence 48-well plate and serum starved for 24 h, and then treated for 24 h witheither: vehicle control; LIF (100 ng/ml); PEG (100 ng/ml); PEGLA (100 ng/ml);

Fig. 2. LIF did not activate STAT3 in OE-E6/E7 cells. Western blot analyses of pSTAT3and total STAT3 protein levels in LIF dose response treatments in OE-E6/E7 cells. OE-E6/E7 cells were treated with vehicle control, or 5, 50 or 100 ng/ml of LIF. Represen-tative Western blots of N ¼ 3 are shown.

T. Krishnan et al. / Placenta 34 (2013) 1014e10191016

PEGLAþ LIF (cells pre-treatedwith PEGLA (100 ng/ml) for 1 h before the addition of LIF(100 ng/ml)). Treatments were determined from previous studies [37]. For spheroidformation, HTR-8/SVneo cells were cultured in RPMI 10% FCS containing Methocel(SigmaeAldrich), in a Cellstar U-shaped 96-well Suspension Culture Plate (GreinerBio-One) and incubated at 37 �C for 24 h. Spheroids were extracted from wells andcentrifuged in 15 ml falcon tubes for 8 min at 200 rcf using a Rotina 380 centrifuge.Spheroids (18/well) were transferred into the 48-well plate containing treated OE-E6/E7 cells. Spheroid number was determined using a microscope before the plate wasincubated at 37 �C for 1 h. Co-culture wells were washed gently, twice with 150 mlserum-free DMEM/F12 media, and spheroids counted to determine the number ofadhered spheroids; expressed as a percentage of the original spheroid number. Assays(n ¼ 4) were performed in duplicate.

2.10. Placental villous explant culture

To determine the effect of LIF and LIF inhibition on trophoblast invasive prop-erties, an outgrowth assay was performed using primary villous tissue of first-trimester placentas (n ¼ 6) as in our previous publication [38]. Briefly, 30 villoustips (1 � 1 mm) were cultured overnight in serum-free DMEM/F12 medium.Collagen I (Rat tail, BD Bioscience) in serum-free medium was added to 48-wellplates (30 ml collagen/well) and incubated at 37 �C for 1 h, before villous explantswere seeded for 3 h on collagen drops, allowing anchorage. Anchoring villi (6 rep-licates/group) were treated with either: vehicle control; LIF (100 ng/ml); PEG(100 ng/ml); PEGLA (100 ng/ml); or PEGLA þ LIF (explants pre-treated with PEGLA(100 ng/ml) for 1 h before the addition of LIF (100 ng/ml)). After 48 h, anchoring villiwere photographed. Area of outgrowth (mm2), representing EVT outgrowth, wasmeasured and quantified using Adobe Photoshop (Adobe) imaging software.

2.11. Statistical analysis

All statistical analyses were performed using GraphPad Prism (GraphPad Soft-ware). Data was analysed by T-test when comparing two groups. Comparisons of 3or more groups were conducted using a 1-way ANOVA analysis. Data was expressedas a mean � SEM. Differences were considered significant at P < 0.05.

3. Results

3.1. LIFR is expressed at human tubal ectopic implantation sites

We immunolocalised LIFR to tubal ectopic implantation sites. Inthe placental villous, LIFR immunostainingwas evident in some celltypes in the stroma, as well as the cyto- and syncytiotrophoblast.LIFR staining was also present in the EVT within the trophoblastshell attached to the tube wall and cells of the Fallopian tube(Fig. 1).

3.2. LIF and LIFR mRNA expression and LIF secretion in OE-E6/E7oviduct cells

Prior to any functional studies, we determined the mRNAexpression of LIF and LIFR in oviduct epithelial OE-E6/E7 cells. LIFand LIFR mRNA were expressed in the cell line (data not shown).Endogenous LIF was secreted by the OE-E6/E7 cells at levels rangingbetween 125 and 137 pg/mg protein and fetal calf serum levels did

Fig. 1. LIFR localisation in human tubal ectopic implantation sites. A: Photomicrograph reprsites �10 magnification and B: high power image of (A) �100 magnification. A (inset): repreplacental villous (*), cytotrophoblast (CT), syncytiotrophoblast (ST), extravillous trophoblas

not affect LIF secretion levels (data not shown). Exogenous LIF hadno effect on pSTAT3 abundance in OE-E6/E7 cells (Fig. 2).

3.3. LIF blockade reduced STAT3 activation in OE-E6/E7 cells

Inhibition of endogenous LIF using PEGLA significantly reducedpSTAT3 compared with control and LIF treatment, with no alter-ations in STAT3 protein levels (P < 0.05, Fig. 3A, B). PEGLA alsoblocked exogenous LIF, as pSTAT3 was significantly reduced in OE-E6/E7 cells treated with PEGLA þ LIF, compared with LIF treatmentalone (P< 0.05, Fig. 3A, B). Conversely, LIF or PEGLA had no effect onpERK in OE-E6/E7 cells (Fig. 3C, D). We trialed PEG control andfound no effect on STAT3 or ERK activation (data not shown).

3.4. Effect of LIF and LIF inhibition on HTR-8/SVneo spheroidadhesion to OE-E6/E7 oviduct cells

Using a co-culture model to represent embryo adhesion to theFallopian tube, we examined the effect of LIF on HTR-8/SVneotrophoblast cell spheroid adhesion to an OE-E6/E7 oviduct cellmonolayer. While exogenous LIF treatment of OE-E6/E7 oviductcells had no significant effect on spheroid adhesion compared tovehicle or PEG control, PEGLA treatment alone significantlyreduced adhesion by 50 � 4% compared to vehicle (P < 0.01);84 � 12% compared to LIF (P < 0.001); and 63 � 13% compared toPEG (P < 0.01; Fig. 4). Co-treatment of oviduct epithelial cells withPEGLA þ LIF significantly decreased spheroid adhesion by 66 � 5%compared to the vehicle (P < 0.01), or 79 � 4% compared to PEGcontrol (P < 0.01) and 97 � 5% compared to LIF (P < 0.001; Fig. 4).

3.5. Effect of LIF and LIF inhibition on placental villous outgrowth

To determine the effect of LIF on placental villous outgrowtharea (mm2), we used primary first-trimester placental villous

esentative of LIFR immunohistochemical staining in human tubal ectopic implantationsent negative isotype IgG control. Representative of N ¼ 5. Stroma, Fallopian tube (FT),t (EVT).

Fig. 3. LIF inhibition reduced STAT3 activation, but did not affect ERK in OE-E6/E7 cells. A: Western blot analyses of pSTAT3 and total STAT3 protein in vehicle control, LIF, PEGLA, orLIF þ PEGLA-treated OE-E6/E7 cells. B: Mean value of vehicle control was arbitrarily set to 1 and optical densities were normalised to vehicle control, bars represent mean � SEM.*P < 0.05. C: Western blot analyses of pERK and total b-actin cellular protein in vehicle control, LIF, PEGLA, or LIF þ PEGLA-treated OE-E6/E7 cells. D: Mean value of vehicle controlwas arbitrarily set to 1 and optical densities were normalised to vehicle control, bars represent mean � SEM. Representative Western blots of N ¼ 3 are shown.

T. Krishnan et al. / Placenta 34 (2013) 1014e1019 1017

explants as a model and quantified percentage outgrowth. LIFtreatment of placental villous explants significantly increasedvillous outgrowth compared to vehicle control (144� 28%) and PEGcontrol (163 � 7%) respectively (P < 0.01; Fig. 5). By contrast,placental outgrowth area was significantly reduced by 132 � 42%in PEGLA-treated explants compared to LIF-treated explants(P < 0.05; Fig. 5). The effect of exogenous LIF on placentaloutgrowth was reversed by co-incubation of LIF with PEGLA, where

Fig. 4. LIF promotes trophoblast spheroid adhesion to oviduct epithelial cells, whichcan be reversed with LIF inhibition. Quantification of the HTR-8/SVneo trophoblastspheroid adhesion to vehicle control, LIF, PEG, PEGLA, or LIF þ PEGLA-treated OE-E6/E7cells. The number of attached spheroids is expressed as a percentage. Bars representmean � SEM from N ¼ 4 co-culture experiments. Mean value of vehicle control wasarbitrarily set to 100%. *P < 0.05, **P < 0.01, ***P < 0.001.

outgrowth was significantly reduced by 178 � 19% compared to theLIF treatment group (P < 0.01; Fig. 5). PEG treatment alone had noeffect on placental villous outgrowth compared with vehiclecontrol.

4. Discussion

This study is the first to investigate the role of LIF in thedevelopment of tubal ectopic pregnancy. We demonstrated thatLIFR localised to trophoblast cells and the Fallopian tube in humantubal ectopic implantation sites. Given that ectopic pregnancy is aphenomenon unique to humans and does not occur naturally orexperimentally in other animals [39e41], we investigated thefunctional role of LIF in the development of ectopic pregnancy us-ing an in vitro co-culture model to model blastocyst adhesion to theFallopian tube. LIF blockade using PEGLA reduced pSTAT3 levels inOE-E6/E7 oviduct epithelial cells and significantly impaired HTR-8/SVneo trophoblast spheroid adhesion to OE-E6/E7 cells. Usingprimary human placental villous explants, we also demonstratedthat LIF significantly promoted primary placental villousoutgrowth, while this was reversed using PEGLA. Together, theseresults suggest that LIF may promote blastocyst adhesion to theFallopian tube andmediate trophoblast invasion that contributes tothe development of ectopic pregnancy.

LIF is increased in the Fallopian tube at ectopic implantationsites compared to non-pregnant and healthy pregnant controls[26]. However, the presence of LIFR at ectopic implantation siteshas not been reported. In this study LIFR immunolocalized to tro-phoblasts in the placental villous, EVT cells and the Fallopian tubein tubal ectopic pregnancy sites. Trophoblast cells displayed asimilar spatial pattern of LIFR expression to that of invasive uterinetrophoblasts [29], supporting the rationale that LIF may play a rolein trophoblast invasion during ectopic implantation.

Fig. 5. The effect of LIF on primary placental villous outgrowth. A: Representative samples from each of vehicle control, LIF, PEG, PEGLA or LIF þ PEGLA-treated villous explants areshown (40-fold magnification). B: Quantification of the area of outgrowth in vehicle control, LIF, PEG, PEGLA or LIF þ PEGLA villous explant cultures. Bars represent mean � SEM ofcultures prepared from N ¼ 6 first-trimester placentas. Mean value of vehicle control was arbitrarily set to 100%. *P < 0.05, **P < 0.01.

T. Krishnan et al. / Placenta 34 (2013) 1014e10191018

We have previously shown that LIF increases trophoblast celladhesion to primary endometrial epithelial cells, the initiatingevent in embryo implantation in the endometrium [18]. Here, wedemonstrated that LIF also facilitates trophoblast cell line adhesionto an oviduct epithelial cell line, while PEGLA treatment signifi-cantly impaired adhesion. These preceding findings, together withour current data support our rationale that LIF may facilitate theestablishment of tubal ectopic pregnancy.

LIF did not activate STAT3 in OE-E6/E7 cells, likely becausepSTAT3 was endogenously high. However blockade of endogenousLIF using PEGLA reduced pSTAT3 abundance, confirming thatendogenous LIF contributed to pSTAT3 in OE-E6/E7 oviduct cells.OE-E6/E7 cells secreted moderate levels of LIF, regardless of serumpresence, which may have contributed to the high endogenouspSTAT3 levels detected at basal conditions. Since LIF also activatesthe ERK signaling pathway [16], we explored whether LIF inducedphosphorylation of ERK. However, there was no evidence for ERKactivation following LIF treatment in OE-E6/E7 cells. Factors otherthan LIF, known to activate the STAT3 pathway may also havecontributed to the high pSTAT3 abundance [42], as PEGLA did nottotally block pSTAT3 in OE-E6/E7 cells.

Since ectopic pregnancies have often implanted at the time ofdiagnosis, elucidating the factors that facilitate trophoblast cellinvasion is a critical requisite to both identify factors regulating EVTinvasion into the Fallopian tube and also in the identification ofnovel targets for the treatment of ectopic pregnancies. By inhibitingtrophoblast invasion at ectopic implantation sites, embryonicgrowth will likely be blocked, and may be utilised as a treatmentoption for ectopic pregnancy. The current primary medical man-agement of ectopic pregnancy is methotrexate, primarily used totreat women presenting with ectopic pregnancies where serumbhCG < 3000 IU/l [43]. However this is associated with teratogenicand other side effects [44]. Methotrexate has been shown to reduceplacental growth and EVT invasion due to reduced trophoblast cellproliferation [45]. In the present study we demonstrated that LIF

significantly increased first-trimester placental villous outgrowth,confirming a role for LIF in trophoblast cell invasion. This likelyoccurred via pSTAT3, as we have previously demonstrated that LIFsignals via pSTAT3 in EVTcells [18]. Our data also supports a role forLIF in trophoblast invasion in intrauterine pregnancies. Mostnotably, in the present study, PEGLA blocked primary first-trimester placental villous explant outgrowth, comparable to con-trol levels, suggesting that this may act as an alternative treatmentto block trophoblast invasion in the Fallopian tube in ectopicpregnancy.

Thus, our data supports our hypothesis that LIF may facilitatethe development of ectopic pregnancy, likely mediated by activa-tion of STAT3 and is the first study to indicate a functional role of LIFin ectopic pregnancies. While further investigations are required,we have demonstrated that there is potential for PEGLA to be usedto block placental growth in the Fallopian tube. Thus, pharmaco-logically targeted LIF blockade may serve as a new potential ther-apeutic in the treatment of tubal ectopic pregnancies thatcontribute to maternal mortality.

Acknowledgments

We acknowledge the support of the Victorian Government’sOperational Infrastructure Support Program and the AustralianGovernment NHMRC IRIISS. This work was supported by fellow-ships from the National Health and Medical Research Council(NHMRC) (Australia). ED and NAN were supported by NHMRCFellowships (#550905 and #637300, respectively). EM was sup-ported by NHMRC Early Career Fellowship (#611827), JGZ and NANwere supported by an NHMRC Program Grant (#461219). AWH issupported by an MRC Clinician Scientist Fellowship (G0802808).This work was supported by the Consortium for Industrial Collab-oration in Contraceptive Research Program of the ContraceptiveResearch and Development Program, Eastern Virginia MedicalSchool (Subproject CIG-02-82). The authors are grateful to Phil

T. Krishnan et al. / Placenta 34 (2013) 1014e1019 1019

Morgan of WEHI for his technical assistance in the production ofPEGLA. PHI Data Audit# 13-15.

References

[1] Farquhar C. Ectopic pregnancy. Lancet 2005;366:583e91.[2] Varma R, Gupta J. Tubal ectopic pregnancy. Clin Evid 2009;20:1406.[3] Menkhorst E, Zhang J, Sims N, Morgan P, Soo P, Poulton I, et al. Vaginally

administered PEGylated LIF antagonist blocked embryo implantation andeliminated non-target effects on bone in mice. PLoS One 2011;6:e19665.

[4] Brenner P, Roy S, Mishell D. Ectopic pregnancy. A study of 300 consecutivesurgically treated cases. JAMA 1980;15:673e6.

[5] Shaw J, Dey S, Critchley H, Horne A. Current knowledge of the aetiology ofhuman tubal ectopic pregnancy. Hum Reprod Update 2010;16:432e44.

[6] Cheng G, Chen J, Hernandez L, Alvord W, Stewart C. Dual control of LIFexpression and LIF receptor function regulate Stat3 activation at the onset ofuterine receptivity and embryo implantation. Proc Natl Acad Sci U S A2001;98:8680e5.

[7] Stewart C, Kaspar P, Brunet L, Bhatt H, Gadi I, Köntgen F, et al. Blastocystimplantation depends on maternal expression of leukaemia inhibitory factor.Nature 1992;359:76e9.

[8] Cullinan E, Abbondanzo S, Anderson P, Pollard J, Lessey B, Stewart C. Leukemiainhibitory factor (LIF) and LIF receptor expression in human endometriumsuggests a potential autocrine/paracrine function in regulating embryo im-plantation. Proc Natl Acad Sci U S A 1996;93:3115e20.

[9] Vogiagis D, Marsh M, Fry R, Salamonsen L. Leukaemia inhibitory factor in humanendometrium throughout the menstrual cycle. J Endocrinol 1996;148:95e102.

[10] Laird S, Tuckerman E, Dalton C, Dunphy B, Li T, Zhang X. The production ofleukaemia inhibitory factor by human endometrium: presence in uterineflushings and production by cells in culture. Hum Reprod 1997;12:569e74.

[11] Aghajanova L, Altmäe S, Bjuresten K, Hovatta O, Landgren B, Stavreus-Evers A.Disturbances in the LIF pathway in the endometrium among women withunexplained infertility. Fertil Steril 2009;91:2602e10.

[12] Bhatt H, Brunet L, Stewart C. Uterine expression of leukemia inhibitory factorcoincides with the onset of blastocyst implantation. Proc Natl Acad Sci U S A1991;88:11408e12.

[13] Classen-Linke I, Müller-Newen G, Heinrich P, Beier H, von Rango U. Thecytokine receptor gp130 and its soluble form are under hormonal control inhuman endometrium and decidua. Mol Hum Reprod 2004;10:495e504.

[14] Dimitriadis E, Menkhorst E, Salamonsen LA, Paiva P. Review: LIF and IL11 introphoblasteendometrial interactions during the establishment of pregnancy.Placenta 2010;24:S99e104.

[15] Dimitriadis E, Nie G, Hannan NJ, Paiva P, Salamonsen LA. Local regulation ofimplantation at the human fetalematernal interface. Int J Dev Biol 2010;54:313e22.

[16] Heinrich P, Behrmann I, Haan S, Hermanns H, Muller-Newen G, Schaper F.Principles of interleukin (IL)-6-type cytokine signalling and its regulation.Biochem J 2003;374:1e20.

[17] Dimitriadis E, Sharkey A, Tan Y, Salamonsen LA, Sherwin J. Immunolocalisa-tion of phosphorylated STAT3, interleukin 11 and leukaemia inhibitory factorin endometrium of women with unexplained infertility during the implan-tation window. Reprod Biol Endocrinol 2007;5:44e51.

[18] Tapia A, Salamonsen L, Manuelpillai U, Dimitriadis E. Leukemia inhibitoryfactor promotes human first trimester extravillous trophoblast adhesion toextracellular matrix and secretion of tissue inhibitor of metalloproteinases-1and -2. Hum Reprod 2008;23:1724e32.

[19] Sharkey A, Dellow K, Blayney M, Macnamee M, Charnock-Jones S, Smith S.Stage-specific expression of cytokine and receptor messenger ribonucleicacids in human preimplantation embryos. Biol Reprod 1995;53:974e81.

[20] Charnock-Jones D, Sharkey A, Fenwick P, Smith S. Leukaemia inhibitory factormRNA concentration peaks in human endometrium at the time of implan-tation and the blastocyst contains mRNA for the receptor at this time. J ReprodFertil 1994;101:421e6.

[21] van Eijk M, Mandelbaum J, Salat-Baroux J, Belaisch-Allart J, Plachot M, Junca A,et al. Expression of leukaemia inhibitory factor receptor subunits LIFR betaand gp130 in human oocytes and preimplantation embryos. Mol Hum Reprod1996;2:355e60.

[22] Hsieh Y, Tsai H, Chang C, Hsu L, Chang S, Lo H. Prolonged culture of humancryopreserved embryos with recombinant human leukemia inhibitory factor.J Assist Reprod Genet 2000;17:131e4.

[23] Sargent L, Martin K, Barlow D. The use of recombinant growth factors topromote human embryo development in serum-free medium. Hum Reprod1998;13:239e48.

[24] Wånggren K, Lalitkumar P, Hambiliki F, Ståbi B, Gemzell-Danielsson K, Stav-reus-Evers A. Leukaemia inhibitory factor receptor and gp130 in the humanFallopian tube and endometrium before and after mifepristone treatment andin the human preimplantation embryo. Mol Hum Reprod 2007;13:391e7.

[25] Ji Y, Chen L, Xu K, Yao J, Shi Y. Locally elevated leukemia inhibitory factor inthe inflamed fallopian tube resembles that found in tubal pregnancy. FertilSteril 2009;91:2308e14.

[26] Güney M, Erdemoglu E, Oral B, Karahan N, Mungan T. Leukemia inhibitoryfactor (LIF) is immunohistochemically localized in tubal ectopic pregnancy.Acta Histochem 2008;110:319e23.

[27] Keltz M, Attar E, Buradagunta S, Olive D, Kliman H, Arici A. Modulation ofleukemia inhibitory factor gene expression and protein biosynthesis in thehuman fallopian tube. Am J Obstet Gynecol 1996;175:1611e9.

[28] Koo HS, Bae JY, Kang IS, Koong MK, Kim HO, Cha SH, et al. Laparoscopicmanagement of early primary peritoneal pregnancy: a case report. Clin ExpReprod Med 2011;38(2):109e14.

[29] Sharkey A, King A, Clark D, Burrows T, Jokhi P, Charnock-Jones D, et al.Localization of leukemia inhibitory factor and its receptor in human placentathroughout pregnancy. Biol Reprod 1999;60:355e64.

[30] Kojima K, Kanzaki H, Iwai M, Hatayama H, Fujimoto M, Inoue T, et al.Expression of leukemia inhibitory factor in human endometrium andplacenta. Biol Reprod 1994;50:882e7.

[31] Fairlie W, Uboldi A, McCoubrie J, Wang C, Lee E, Yao S, et al. Affinity matu-ration of leukemia inhibitory factor and conversion to potent antagonists ofsignaling. J Biol Chem 2004;279:2125e34.

[32] White C, Zhang J, Salamonsen L, Baca M, Fairlie W, Metcalf D, et al. BlockingLIF action in the uterus by using a PEGylated antagonist prevents implanta-tion: a nonhormonal contraceptive strategy. Proc Natl Acad Sci U S A2007;104:19357e62.

[33] Hannan NJ, Paiva P, Dimitriadis E, Salamonsen LA. Models for study of humanembryo implantation: choice of cell lines? Biol Reprod 2010;82:235e45.

[34] Lee Y, Lee K, Xu J, Wang Y, Tsao S, Yeung W. Establishment and character-ization of an immortalized human oviductal cell line. Mol Reprod Dev2001;59:400e9.

[35] Shuya L, Menkhorst E, Yap J, Li P, Lane N, Dimitriadis E. Leukemia inhibitoryfactor enhances endometrial stromal cell decidualization in humans and mice.PLoS One 2011;6:e25288.

[36] Laib A, Bartol A, Alajati A, Korff T, Weber H, Augustin H. Spheroid-basedhuman endothelial cell microvessel formation in vivo. Nat Protoc 2009;4:1202e15.

[37] Marwood M, Visser K, Salamonsen LA, Dimitriadis E. Interleukin-11 and leu-kemia inhibitory factor regulate the adhesion of endometrial epithelial cells:implications in fertility regulation. Endocrinology 2009;150:2915e23.

[38] Sonderegger S, Yap J, Menkhorst E, Weston G, Stanton P, Dimitriadis E.Interleukin (IL)11 mediates protein secretion and modification in humanextravillous trophoblasts. Hum Reprod 2011;26:2841e9.

[39] Tutton D, Carr D. The fate of trophoblast retained within the oviduct in themouse. Gynecol Obstet Invest 1984;17:18e24.

[40] Pauerstein C, Eddy C, Koong M, Moore G. Rabbit endosalpinx suppressesectopic implantation. Fertil Steril 1990;54:522e6.

[41] Orsini M, McLaren A. Loss of the zona pellucida in mice, and the effect of tuballigation and ovariectomy. J Reprod Fertil 1967;13:485e99.

[42] Heinrich P, Behrmann I, Muller-Newen G, Schaper F, Graeve L. Interleukin-6-type cytokine signalling through the gp130/Jak/STAT pathway. Biochem J1998;334:297e314.

[43] Skubisz MM, Tong S. The evolution of methotrexate as a treatment for ectopicpregnancy and gestational trophoblastic neoplasia: a review. ISRN ObstetricsGynecol 2012;2012:637094.

[44] Finn M, Bowyer L. Women’s health, a core curriculum. Marrickville, N.S.W.,Australia: Elsevier Australia; 2005.

[45] Floridon C, Nielsen O, Byrjalsen C, Holund B, Kerndrup G, Thomsen SG, et al.Ectopic pregnancy: histopathology and assessment of cell proliferation withand without methotrexate treatment. Fertil Steril 1996;65(4):730e8.