lable at ScienceDirect

Placenta 31 (2010) 259–268

Contents lists avai

Placenta

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

Distinct Patterns of Gene-Specific Methylation in Mammalian Placentas:Implications for Placental Evolution and Function

H.K. Ng a, B. Novakovic a,b, S. Hiendleder c,d, J.M. Craig a,b, C.T. Roberts c, R. Saffery a,b,*

a Developmental Epigenetics, Murdoch Children’s Research Institute, Royal Children’s Hospital, Flemington Road, Parkville, Victoria, Australiab Department of Paediatrics, University of Melbourne, Parkville, Victoria 3052, Australiac JS Davies Epigenetics and Genetics Group, Animal Science University of Adelaide, Australia 5005d Robinson Institute, School of Paediatrics and Reproductive Health, University of Adelaide, Australia 5005

a r t i c l e i n f o

Article history:Accepted 12 January 2010

Keywords:DNA methylationEpigeneticsWnt signallingDNMT1Vitamin D 24-hydroxylase

* Corresponding author. Developmental EpigenResearch Institute, Royal Children’s Hospital, FlemingAustralia. Tel.: þ61 03 83416341.

E-mail address: richard.saffery@mcri.edu.au (R. Sa

0143-4004/$ – see front matter � 2010 Elsevier Ltd.doi:10.1016/j.placenta.2010.01.009

a b s t r a c t

The placenta has arisen relatively recently and is among the most rapidly evolving tissues in mammals.Several different placental barrier and structure types appear to have independently evolved commonfunctional features. Specific patterns of gene expression that determine placental development inhumans are predicted to be accompanied by specific profiles of epigenetic modification. However, thestratification of epigenetic modifications into those involved in conserved aspects of placental function,versus those involved in divergent placental features, has yet to begin. As a first step towards this goal,we have investigated the methylation status of a small number of gene-specific methylation eventsrecently identified in human placenta, in a panel of placental tissue from baboon, marmoset, cow, cat,guinea pig and mouse. These represent disparate placental barrier types and structures. In this study wehypothesized that specific epigenetic markings may be associated with placental barrier type or function,independent of phylogeny. However, in contrast to our predictions, the majority of gene-specificmethylation appears to track with phylogeny, independent of placental barrier type or other structuralfeatures. This suggests that despite the likelihood of epigenetic modification playing a role in thefunctioning and evolution of different placental subtypes, there is no evidence for an involvement of thegene-specific methylation profiles we have identified, in specifying these differences. Further studies,examining larger numbers of epigenetic modifications across phylogeny, are required to define the role ofspecific epigenetic modifications in the evolution of distinct placental structures.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

In addition to regulating nutrient exchange between maternal tofetal circulations, the placenta represents the first line of defense inprotecting the developing fetus from adverse environmentalexposures and the maternal immune system. Although allplacentas share a basic common function in modulating exchangeat the fetomaternal interface, it is clear that the human placenta isunique in structure and function [1].

Placentas can be classified in several ways, including by thenumber of cell layers separating maternal and fetal bloodstreams[2]. There is a maximum of six potential cellular layers, threematernal and three fetal. Epitheliochorial placentas contain all six

etics, Murdoch Children’ston Road, Parkville, Victoria,

ffery).

All rights reserved.

layers and are found in the horse, pig and ruminant species.However, the ruminant placenta is strictly synepitheliochorial inwhich trophoblast binucleate cells migrate away from the fetaltrophoblast layer and fuse with maternal uterine luminal epithelialcells to form a fetomaternal syncytium. Endotheliochorialplacentas, with one maternal layer and three fetal layers, arepresent in carnivores and some insectivores, lower primates, andmegachiropteran bats. Hemochorial placentas, found in higherprimates, rodents, microchiropteran bats and the armadillo, showa more extensive trophoblast invasion during development thatremodels maternal arteries, thereby maximizing blood flow to theintervillous space [3].

Each eutherian order/sub order has a distinctive placental shapeand type of fetomaternal attachment that have presumably evolvedto meet specific fetal needs. Therefore, despite the value of animalmodels in examining fundamental functions of the placentationprocess, they may be of limited value in fully dissecting molecularprocesses contributing to specific human pathologies such as preeclampsia (PE), intra uterine growth restriction (IUGR) and babies

H.K. Ng et al. / Placenta 31 (2010) 259–268260

small for gestational age (SGA) [4]. The mechanisms underlying thedistinct differences within eutherian placentas have remainedelusive but clearly include a unique gene expression profile indifferent species (reviewed in [5]). Limited evidence also supportsa role of distinct epigenetic modification in placental functioningand morphology across evolution.

It has been known for many years that the placenta displayssome unique epigenetic features compared to somatic cells. Thisincludes the lowest 5-methylcytosine content (w3.1%) of all humantissues except germ cells [6–8], comparable to that in humancancers [9–11], and a unique set of imprinted genes (expressed ina parent-of-origin manner) regulated by DNA methylation [12].Many of these genes are involved in regulation of growth, withdisruption of imprinting previously associated with IUGR [13], SGA[14], PE [15–17], and the use of assisted reproductive technologies[18–22].

More recently, non-imprinting-associated placenta-specificepigenetic modification has been described (eg. [23,24]). Thisincludes several instances of tumour suppressor (TS) gene silencingby histone modification and/or DNA methylation [25–28]. TS genesare generally unmethylated in healthy human somatic tissue butare often methylated and silenced in human cancers. Despite thefact that general disruption of DNA methylation by specific drugtreatment disrupts placental development and inhibits invasive-ness in trophoblast cell lines [29–31], for the most part, the functionof placenta-specific epigenetic modification remains to beelucidated.

In this study we investigated the association of a small numberof gene-specific methylation events, previously described inhumans, with distinct features of eutherian placentas such asbarrier type and shape. Genes of interest are involved in disparatefunctions including Wnt pathway signaling, vitamin D homeostasis,and maintenance of DNA methylation. These pathways play pivotalroles in modulation of trophoblast invasiveness, differentiation,and immune functioning and differing regulation of these genes viaepigenetic divergences may be associated with variations inpreviously described placental morphology and trophoblastfunction.

2. Methods

2.1. Tissue samples

Human tissues were harvested for research purposes with appropriate humanethics clearances from the Royal Women’s Hospital (03/51), Mercy Hospital forWomen (R07/15), and Monash Medical Centre (07084C). Full term baboon placentaltissue was provided by Dr Neroli Sunderland (Royal Prince Alfred Hospital, SydneyAustralia, approval 2009/028A). Full term marmoset placental tissue was obtainedfrom Monash Animal Services (Melbourne, Australia). Full term guinea pig placentawas obtained with ethical approval from the University of Adelaide Animal EthicsCommittee (M-73-2003). Day 153 cow cotyledon tissue was collected from thePlacenta fetalis with approval from the University of Adelaide Animal EthicsCommittee (S-094-2005A). Day 18 gestation mouse placental tissue was providedby Drs Patrick Western and Craig Smith (Murdoch Childrens Research Institute,Australia), and 4 week gestation cat placental tissue was obtained from Dr ChrisThurgood following routine desexing (Royal Society for the Protection and Care ofAnimals, Burwood, Victoria, Australia). These tissues were provided in line with theprinciple of Reduction, NHMRC Code of Practice for the Care and use of animals forscientific purposes, 7th Edition, 2004. For human placental tissue sampling, a coreof full thickness tissue was isolated and bisected into two layers from maternal andfetal sides. All placental tissues were processed by maceration with a scalpel bladeimmediately prior to DNA isolation.

2.2. Genomic DNA isolation

Tissue samples were incubated at 50 �C overnight with shaking in DNAextraction buffer [100 mM NaCl, 10 mM TrisHCl pH8, 25 mM EDTA, 0.5% SDS] with200 mg/mL Proteinase K. DNA was isolated by two rounds of phenol:chloroformextraction, followed by RNAse A treatment, precipitation in absolute ethanol con-taining 10% sodium acetate (3M, pH 5.2), and resuspended in 50 mL TE Buffer [10 mM

TrisHCl pH7.5, 1 mM EDTA]. DNA was quantitated and quality assessed by massspectroscopy on a Nanodrop (ThermoFisher Scientific, Waltham MA, USA), andstored at �20 �C.

2.3. Methylation analysis using bisulphite DNA sequencing

DNA samples were processed using the MethylEasy� bisulphite modification kit(Human Genetic Signatures, North Ryde, NSW, Australia) kit according to themanufacturer’s instructions. Sodium bisulphite selectively converts unmethylatedcytosine (C) residues to uracil nucleotides. This covalent modification is stablymaintained during subsequent PCR amplification where the uracil is converted tothe complementary thymidine nucleotide included in the amplification mix. PCRamplification was performed on converted genomic DNA using primers directed tobisulphite modified DNA (sequences listed in Supplementary Table 1). Assays foreach region of interest were designed following bioinformatic analysis to identifyequivalent regions to that previously assayed in humans (Fig. 1). In the event thatassay design parameters did not allow equivalent regions to be assessed, regions ofclose proximity were examined. Although bisulphite sequencing is relatively timeconsuming in comparison to other methods, and measures a small number of alleleswithin any tissue specimen, it remains the ‘gold standard’ approach for the parallelcharacterization of both DNA methylation level and distribution at specific loci. Thiswas an important consideration in the current study where both aspects of DNAmethylation were being assessed.

Amplification conditions were 95 �C for 5 min, 56 �C for 1 min 30 s and 72 �C for1 min 30 s � 40 cycles, 72 �C for 7 min. Resulting amplicons were cloned into TOPOTA Cloning kit (Invitrogen, Carlsbad, CA, USA) for automated fluorescent sequencingas previously described [32]. Data were analyzed using BiQ Analyser software [33]and clones showing less than 80% conversion or 90% homology to the referencesequence were not included in subsequent analyses.

3. Results

We used bisulphite sequencing to profile the methylationstatus of a small number of gene promoters in placental tissuefrom several eutherians, representing widely divergent placentalmorphologies, barrier types and phylogenetic distances. Usingthis approach, we found that patterns of methylation for the fourgenes analyzed could be broadly classed into 5 groups as detailedin Table 1.

3.1. Comparative methylation of Wnt pathway regulators

The Wnt signaling pathway plays a pivotal role in many cellularfunctions, including cell division, proliferation, differentiation andmotility [34]. Active Wnt signaling has been implicated in thesurvival, differentiation, and invasion of human trophoblasts withnuclear b–catenin accumulation apparent in extravillous tropho-blast cells [35].

Previous studies from our laboratory have shown that themajor promoter of the Adenomatous Polyposis Coli (APC) gene(negative regulator of Wnt signaling) is monoallelically methyl-ated specifically in human (hemomonochorial villous) placentaand purified trophoblasts, with no methylation apparent inmouse (hemotrichorial labyrinthine) placental tissues [27]. Weexamined the methylation status of this promoter in baboon(hemomonochorial villous), marmoset (hemomonochorialvillous), cow (synepitheliochorial cotyledonary), cat (endothe-liochorial zonary) and guinea pig (hemomonochorial labyrin-thine) placental tissue. Whereas strong evidence of APCmonoallelic methylation was found in baboon placental tissue,manifesting as intermediate (MED) methylation levels (MED-MONO class), almost complete methylation (HIGH) was apparentin marmoset placental tissue, with no evidence of a monoallelicpattern (HIGH class; Fig. 6B). In contrast, much lower levels ofmethylation (LOW) were detected in cow, cat and guinea pigplacental tissue, with weak evidence of a minor population ofhighly methylated alleles in each of these species (LOW-mono?class). Essentially no methylation of this region was detected inmouse placentas (LOW class; Figs. 1and 6B).

Fig. 1. Location of methylation assays used in the current study. Alignment of methylation assays was performed using the BLAT function of the UCSC genome browser (humangenome build GRCh37). Chromosomal location is listed along with the relative positions and level of sequence homology between the different methylation assays. The 50 end ofgenes of interest is shown along with the transcription start site (arrow). Red dashes denote individual CpG sites within each assay. (For interpretation of the references to colour inthis figure legend, the reader is referred to the web version of this article).

H.K. Ng et al. / Placenta 31 (2010) 259–268 261

We also examined methylation of the SFRP2 gene (representinganother class of negative regulators of Wnt signaling), previouslymeasured in human placental tissue [28]. When the correspondingregion was investigated in other eutherian placentas, highly vari-able methylation levels and profiles were found in different species.

This included intermediate mean methylation levels in baboon,marmoset and guinea pig placentas with strong evidence of mon-oallelic methylation (MED-MONO class) in marmoset with little orno evidence for monoallelic methylation in individual baboon andguinea pig placental tissue (MED-mono? class; Figs. 2 and 6B).

Table 1Classification of methylation profile according to mean absolute methylation leveland distribution pattern.

Class Absolute methylation levela

HIGH >70%MED 30–70%LOW <30%

Evidence for monoallelicmethylation

Criteriab

STRONG (MONO) Three or more instances of coincident fullymethylated plus fully unmethylated alleles

WEAK (mono?) At least one instance of coincident fully methylatedplus fully unmethylated alleles

NONE No examples of coincident fully methylated plusfully unmethylated alleles

Methylation profile was defined using a combination of absolute methylation levelin addition to the evidence for a population of monoallelically methylated alleleswithin in individual placental samples. Thus complete monoallelic methylation witha 50:50 ratio of fully methylated versus fully unmethylated alleles would bedescribed as MED-MONO.

a Absolute methylation cutoffs for each class are arbitrarily defined.b Fully methylated in this instance was defined was defined as >90% methylation,

while fully unmethylated was <10% methylated CpG sites within single alleles asdetermined by bisulphite sequencing.

H.K. Ng et al. / Placenta 31 (2010) 259–268262

Lower levels of methylation were seen in the cow, with someevidence for a minor population of highly methylated alleles (LOW-mono?), whereas this region in the cat and mouse placentas wereboth essentially unmethylated (LOW; Figs. 2 and 6B).

3.2. Comparative methylation of the vitamin D catabolic24-hydroxylase gene (CYP24A1)

Very little is known about the epigenetic regulation of enzymesmodulating vitamin D homeostasis at the fetomaternal interface.However, we recently identified placenta-specific methylation ofthe CYP24A1 gene in human placental tissue and purified firsttrimester cytotrophoblasts [23]. Given the likely importance ofvitamin D in proper placental development and functioning, fetaldevelopment, and in the modulation of maternal immune responseto the developing fetus, we speculated that epigenetic regulation ofthis gene may be conserved across all eutherians.

An examination of the methylation level in marmoset placentaltissue revealed a pattern of intermediate methylation with noevidence for monoallelic methylation (MED), in contrast to thatseen in human full term placenta and purified trophoblasts (MED-MONO) [23]. Surprisingly, similar levels were also observed inguinea pig placenta with weak evidence for a minor population ofhighly methylated alleles in some placentas (MED-mono?). Lowerlevels of methylation were observed in baboon, similar to that inthe mouse (LOW; Fig. 3). Essentially no methylation was detected inthe corresponding region of the CYP24A1 gene in cat and cowplacental tissue (LOW; Figs. 3 and 6B).

3.3. Primate-specific DNMT1 promoter methylation in eutherianplacental tissue

The family of DNA methyltransferases DNMT1, -3A, -3B and -3Lare responsible for the establishment and maintenance of DNAmethylation in mammals (reviewed in [36]). We have recentlydemonstrated high prevalence monoallelic methylation of theDNMT1 gene in human and baboon placental tissue [55]. In thisstudy we have confirmed high prevalence monoallelic methylationin human and marmoset placental tissue (MED-MONO; Fig. 4) anddemonstrated a similar pattern in baboon full term placentas (alsoMED-MONO). Much lower levels of methylation were seen in cow,cat, guinea pig and mouse placental samples (LOW; Fig. 4)

demonstrating a primate-specific pattern of promoter methylation(Fig. 5).

4. Discussion

Despite diverse structural differentiation and some functionalparameters in placentas from different placentas, all share severalcommon features. Interspecific comparison of the placental epi-genome therefore offers an unparalleled opportunity to identifymolecular mechanisms contributing to both conserved and diver-gent placental functions.

Previous limited epigenetic analysis has identified a generalhypomethylation of extraembryonic tissue in humans (lowergenomic 5-methylcytosine content) relative to somatic tissues dueto a more pronounced re-methylation of the embryonic lineagevery early in development [7,8,37–39]. This phenomenon is alsoevident in placental tissue of baboon, marmoset, guinea pig [55],mouse [39–41], rabbit [42], sheep [43], and cow [44,45], and whereexamined, is correlated with lower levels of methylation at repet-itive DNA sequences.

Accumulated data have demonstrated a link between genomicimprinting (an epigenetic phenomenon), the evolution of viviparityand the acquisition of a placenta in metatherian and eutherianmammals, but it is also clear that the spectrum of placentalimprinting varies between species. For example, genes imprinted inboth embryonic and extraembryonic tissues show extensiveconservation between mouse and human and are generally regu-lated by DNA methylation [12]. In contrast, many genes showplacenta-specific imprinting in mice, but not humans. Such genesare invariably regulated at the level of histone modification ratherthan DNA methylation [46].

Placenta-specific gene expression has been documented overmany years. This has revealed potential conserved gene expressionsignatures associated with placentation, but also interspecificdifferences in genes such as transcription factors that likely lead todifferential regulation of a wide range of downstream genes andcellular processes (reviewed in [5]). Despite intensive efforts tocatalogue genes specifically expressed in the placenta, no system-atic examination of genes showing placenta-specific down-regu-lation (or silencing) has been reported. Such silencing is usuallymediated by epigenetic modification.

Formation of any tissue, including the placenta, requiresa coordinated series of epigenetic modifications that preciselyregulate gene expression at key developmental time points(reviewed in [47]). DNA methylation is a widely studied epigeneticregulator of gene expression, and disruption of placental methyl-ation has been unequivocally linked to aberrant placental function.Specifically, 5-deoxy azacitidine (5-Aza; general DNA methylationinhibitor) administered to pregnant rats, disrupts trophoblastproliferation and pregnancy development [30,31], and 5-Azatreatment of choriocarcinoma-derived cell lines in vitro inhibitstrophoblast migration [29]. In addition, gene knockout studies ofDNA methyltransferase enzymes Dnmt1 or Dnmt3l are associatedwith placental defects [48,49].

Recent large scale-studies have begun to profile locus-specificplacental methylation on a genome-wide (or subgenomic) levelin humans [50–53]. Other studies have begun to characterizespecific methylation directly in the placenta in a variety of species(eg. [25,28,54]). However, the level of conservation of gene-specificmethylation across eutherians remains unclear. In the currentstudy, we measured methylation levels of genes previously iden-tified as methylated in human placenta, in a collection of placentaltissue from seven species representing a range of different placentastructure and barrier types indicative of variation across euthe-rians, including hemomono- and trichorial (primates and mouse

Fig. 2. Methylation of the APC gene in 7 eutherian placentas. Following bisulphite conversion of placental genomic DNA, regions of interest were amplified with primer sequencesspecific to converted DNA sequence. Amplified PCR products were then cloned into a DNA vector before transformation into E. coli. Individual colonies were then randomly selectedfor DNA isolation and sequencing. Biological replicate samples from independent placentas (number listed in parentheses) were tested for each species. Data presented for (A)human, (B) baboon, (C) marmoset, (D) cow, (E) cat, (F) guinea pig, and (G) mouse placental tissue. Grey horizontal bars denote CpG island regions associated with the gene ofinterest and transcription start site (raised arrows). Black bars show locations of associated exonic regions. Red dashes denote individual CpG sites within each assay whichcorrespond to circles below. Circled methylation data are presented for a single placental sample (as an example) for which 8–12 DNA sequences from individual alleles (horizontalrows) were generated. Due to space constraints, only 8 alleles from each placental sample are shown. Shaded circles denote methylated CpG sites within specific DNA clones (opencircles). Mean methylation (0–100%) for each CpG site in biological replicates (number in brackets) is shown as shaded bars below each site. (For interpretation of the references tocolour in this figure legend, the reader is referred to the web version of this article).

Fig. 3. Methylation of the SFRP2 gene in 7 eutherian placentas. Bisulphite sequencing was carried out as described in Fig. 2. Data presented for (A) human, (B) baboon, (C) marmoset,(D) cow, (E) cat, (F) guinea pig, and (G) mouse placental tissue. For details see Fig. 2.

H.K. Ng et al. / Placenta 31 (2010) 259–268264

Fig. 4. Methylation of the CYP24A1 gene in 7 eutherian placentas. Bisulphite sequencing was carried out as described in Fig. 2. Data presented for (A) human, (B) baboon, (C)marmoset, (D) cow, (E) cat, (F) guinea pig, and (G) mouse placental tissue. For details see Fig. 2.

H.K. Ng et al. / Placenta 31 (2010) 259–268 265

Fig. 5. Methylation of the DNMT1 gene in 7 eutherian placentas. Bisulphite sequencing was carried out as described in Fig. 2. Data presented for (A) human, (B) baboon, (C)marmoset, (D) cow, (E) cat, (F) guinea pig, and (G) mouse placental tissue. For details see Fig. 2.

H.K. Ng et al. / Placenta 31 (2010) 259–268266

Fig. 6. Placenta specific gene methylation levels according to phylogenetic relationship (A) Phylogenetic relationship of eutherians showing placenta barrier type (He, Hemochorial;Hm, Hemomonochorial; Hd, Hemodichorial; Ht, Hemotrichorial; Ep, Epitheliochorial; En, Endotheliochorial), placenta shape (C; Cotyledonary; Z, Zonary; D, Diffuse; Dd, Discoid; Bd,Bidiscoid), and type of interdigitation (T, Trabecular; Lb, Labyrinthine; F, Folded; L, Lamellar). Species examined in this study are highlighted in red. (B) Summary of methylation levelfor each of the four genes examined in this study for seven different species. Phylogenetic tree adapted from {Wildman, 2006 #1672}. Note that the classification of manateeplacental barrier type has recently been modified from He to En {Carter, 2008 #1673}. Classification of methylation profile is based on parameters listed in Table 1. (For inter-pretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

H.K. Ng et al. / Placenta 31 (2010) 259–268 267

respectively), synepitheliochorial (cow) and endotheliochorial(cat).

Our data indicate a conservation of gene methylation withinprimates (old and new world monkeys and apes). However, thedistribution and absolute levels of methylation showed somevariation in these closely related species, particularly in theCYP24A1 gene that is more highly methylated in humans relative tobaboon and marmoset, and the APC gene that shows an inversepattern of methylation.

The findings of the current study add a further dimension toprocesses potentially contributing to differential placental expres-sion, by demonstrating limited conservation of locus-specificmethylation patterns, even between closely related species. Littleconservation of the methylation profile in APC, SFRP2, DNMT1 orCYP24A1 genes was found in non-primate species, suggesting thatthese locus-specific methylation events arose relatively recently inevolution. However, some evidence for a reduced level of methyl-ation, or reduced incidence of monoallelic methylation (possiblyreflecting differing cellular composition in placental tissue) wasevident. This will require further investigation, preferably in puri-fied placenta-derived cell populations.

The consistency of DNA methylation patterns at specific lociwithin species, and the coordinated methylation of multiple geneswithin signaling pathways (eg. APC and SFRP2 in primates), impli-cates genetic factors in the establishment of placental methylationprofile. Future studies will be required to identify such factors tofully elucidate the role that such methylation plays in mediatingspecific aspects of placental development and function.

Finally, the mounting evidence for significant epigenetic discor-dance between human and animal placentas may limit the value of

animal models in unraveling the complex interplay between envi-ronment, epigenetic disruption, and phenotype likely to contributeto elevated risk of adverse pregnancy outcomes in humans.

Acknowledgements

We would like to thank Dr Neroli Sunderland (Royal PrinceAlfred Hospital, Australia) for full term placental tissue and CVSsamples, Drs Patrick Western and Craig Smith (Murdoch ChildrensResearch Institute, Australia) for mouse placental tissue, MonashAnimal Services (Australia) for Marmoset placental tissue, and DrChris Thurgood (RSPCA, Victoria) for feline placental tissue. Thiswork was funded by National Health and Medical Research Council(Australia) special grant to JC and RS. SH is a JS Davies Fellow.

Appendix. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.placenta.2010.01.009.

References

[1] Carter AM. Animal models of human placentation–a review. Placenta2007;28(Suppl. A):S41–7.

[2] Enders AC, Carter AM. What can comparative studies of placental structure tellus?–a review. Placenta 2004;25(Suppl. A):S3–9.

[3] Benirschke K, Kaufmann P. Pathology of the human placenta. Springer; 2000.[4] Soundararajan R, Rao AJ. Trophoblast ’pseudo-tumorigenesis’: significance and

contributory factors. Reprod Biol Endocrinol 2004;2:15.[5] Rawn SM, Cross JC. The evolution, regulation, and function of placenta-specific

genes. Annu Rev Cell Dev Biol 2008;24:159–81.

H.K. Ng et al. / Placenta 31 (2010) 259–268268

[6] Gama-Sosa MA, Wang RY, Kuo KC, Gehrke CW, Ehrlich M. The 5-methyl-cytosine content of highly repeated sequences in human DNA. Nucleic AcidsRes 1983;11(10):3087–95.

[7] Tsien F, Fiala ES, Youn B, Long TI, Laird PW, Weissbecker K, et al. Prolonged cultureof normal chorionic villus cells yields ICF syndrome-like chromatin deconden-sation and rearrangements. Cytogenet Genome Res 2002;98(1):13–21.

[8] Fuke C, Shimabukuro M, Petronis A, Sugimoto J, Oda T, Miura K, et al. Agerelated changes in 5-methylcytosine content in human peripheral leukocytesand placentas: an HPLC-based study. Ann Hum Genet 2004;68(Pt 3):196–204.

[9] Baylin SB. DNA methylation and gene silencing in cancer. Nat Clin Pract Oncol2005;2(Suppl. 1):S4–11.

[10] Issa JP. CpG island methylator phenotype in cancer. Nat Rev Cancer2004;4(12):988–93.

[11] Esteller M, Herman JG. Cancer as an epigenetic disease: DNA methylation andchromatin alterations in human tumours. J Pathol 2002;196(1):1–7.

[12] Wagschal A, Feil R. Genomic imprinting in the placenta. Cytogenet GenomeRes. 2006;113(1–4):90–8.

[13] Diplas AI, Lambertini L, Lee MJ, Sperling R, Lee YL, Wetmur J, et al. Differentialexpression of imprinted genes in normal and IUGR human placentas. Epige-netics 2009;4(4):235–40.

[14] Guo L, Choufani S, Ferreira J, Smith A, Chitayat D, Shuman C, et al. Altered geneexpression and methylation of the human chromosome 11 imprinted region insmall for gestational age (SGA) placentae. Dev Biol 2008;320(1):79–91.

[15] Kanayama N, Takahashi K, Matsuura T, Sugimura M, Kobayashi T, Moniwa N,et al. Deficiency in p57Kip2 expression induces preeclampsia-like symptomsin mice. Mol Hum Reprod 2002;8(12):1129–35.

[16] Yu L, Chen M, Zhao D, Yi P, Lu L, Han J, et al. The H19 gene imprinting in normalpregnancy and pre-eclampsia. Placenta 2009;30(5):443–7.

[17] Mutze S, Rudnik-Schoneborn S, Zerres K, Rath W. Genes and the preeclampsiasyndrome. J Perinat Med 2008;36(1):38–58.

[18] Gosden R, Trasler J, Lucifero D, Faddy M. Rare congenital disorders, imprintedgenes, and assisted reproductive technology. Lancet 2003;361(9373):1975–7.

[19] Halliday J, Oke K, Breheny S, Algar E, Amor D. Beckwith-Wiedemannsyndrome and IVF: a case-control study. Am J Hum Genet 2004;75(3):526–8.

[20] Li T, Vu TH, Ulaner GA, Littman E, Ling JQ, Chen HL, et al. IVF results in de novoDNA methylation and histone methylation at an Igf2-H19 imprinting epige-netic switch. Mol Hum Reprod 2005;11(9):631–40.

[21] Katari S, Turan N, Bibikova M, Erinle O, Chalian R, Foster M, et al. DNAmethylation and gene expression differences in children conceived in vitro orin vivo. Hum Mol Genet; 2009.

[22] Gomes MV, Huber J, Ferriani RA, Amaral Neto AM, Ramos ES. Abnormalmethylation at the KvDMR1 imprinting control region in clinically normalchildren conceived by assisted reproductive technologies. Mol Hum Reprod2009;15(8):471–7.

[23] Novakovic B, Sibson M, Ng HK, Manuelpillai U, Rakyan V, Down T, et al.Placenta-specific methylation of the vitamin D 24-hydroxylase gene: impli-cations for feedback autoregulation of active vitamin D levels at the fetoma-ternal interface. J Biol Chem 2009;284(22):14838–48.

[24] Zhang HJ, Siu MK, Wong ES, Wong KY, Li AS, Chan KY, et al. Oct4 is epige-netically regulated by methylation in normal placenta and gestationaltrophoblastic disease. Placenta 2008;29(6):549–54.

[25] Chiu RW, Chim SS, Wong IH, Wong CS, Lee WS, To KF, et al. Hypermethylationof RASSF1A in human and rhesus placentas. Am J Pathol 2007;170(3):941–50.

[26] Dokras A, Coffin J, Field L, Frakes A, Lee H, Madan A, et al. Epigenetic regulation ofmaspin expression in the human placenta. Mol Hum Reprod 2006;12(10):611–7.

[27] 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;268(1):56–62.

[28] Novakovic B, Rakyan V, Ng HK, Manuelpillai U, Dewi C, Wong NC, et al. Specifictumour-associated methylation in normal human term placenta and first-trimester cytotrophoblasts. Mol Hum Reprod 2008;14(9):547–54.

[29] Rahnama F, Shafiei F, Gluckman PD, Mitchell MD, Lobie PE. Epigenetic regu-lation of human trophoblastic cell migration and invasion. Endocrinology2006;147(11):5275–83.

[30] Serman L, Vlahovic M, Sijan M, Bulic-Jakus F, Serman A, Sincic N, et al. Theimpact of 5-azacytidine on placental weight, glycoprotein pattern andproliferating cell nuclear antigen expression in rat placenta. Placenta2007;28(8–9):803–11.

[31] Vlahovic M, Bulic-Jakus F, Juric-Lekic G, Fucic A, Maric S, Serman D. Changes inthe placenta and in the rat embryo caused by the demethylating agent 5-azacytidine. Int J Dev Biol 1999;43(8):843–6.

[32] Wong NC, Wong LH, Quach JM, Canham P, Craig JM, Song JZ, et al. Permissivetranscriptional activity at the centromere through pockets of DNA hypo-methylation. PLoS Genet 2006;2(2). e17.

[33] Bock C, Reither S, Mikeska T, Paulsen M, Walter J, Lengauer T. BiQ Analyzer:visualization and quality control for DNA methylation data from bisulfitesequencing. Bioinformatics 2005;21(21):4067–8.

[34] Eastman Q, Grosschedl R. Regulation of LEF-1/TCF transcription factors by Wntand other signals. Curr Opin Cell Biol 1999;11(2):233–40.

[35] Pollheimer J, Loregger T, Sonderegger S, Saleh L, Bauer S, Bilban M, et al.Activation of the canonical wingless/T-cell factor signaling pathway promotesinvasive differentiation of human trophoblast. Am J Pathol 2006;168(4):1134–47.

[36] Bestor TH. The DNA methyltransferases of mammals. Hum Mol Genet2000;9(16):2395–402.

[37] Gama-Sosa MA, Midgett RM, Slagel VA, Githens S, Kuo KC, Gehrke CW, et al.Tissue-specific differences in DNA methylation in various mammals. BiochimBiophys Acta 1983;740(2):212–9.

[38] Okahara G, Matsubara S, Oda T, Sugimoto J, Jinno Y, Kanaya F. Expressionanalyses of human endogenous retroviruses (HERVs): tissue-specific anddevelopmental stage-dependent expression of HERVs. Genomics2004;84(6):982–90.

[39] Santos F, Hendrich B, Reik W, Dean W. Dynamic reprogramming of DNAmethylation in the early mouse embryo. Dev Biol 2002;241(1):172–82.

[40] Chapman V, Forrester L, Sanford J, Hastie N, Rossant J. Cell lineage-specificundermethylation of mouse repetitive DNA. Nature 1984;307(5948):284–6.

[41] Rossant J, Sanford JP, Chapman VM, Andrews GK. Undermethylation ofstructural gene sequences in extraembryonic lineages of the mouse. Dev Biol1986;117(2):567–73.

[42] Manes C, Menzel P. Demethylation of CpG sites in DNA of early rabbittrophoblast. Nature 1981;293(5833):589–90.

[43] Beaujean N, Hartshorne G, Cavilla J, Taylor J, Gardner J, Wilmut I, et al. Non-conservation of mammalian preimplantation methylation dynamics. Curr Biol2004;14(7):R266–7.

[44] Santos F, Zakhartchenko V, Stojkovic M, Peters A, Jenuwein T, Wolf E, et al.Epigenetic marking correlates with developmental potential in cloned bovinepreimplantation embryos. Curr Biol 2003;13(13):1116–21.

[45] Hiendleder S, Mund C, Reichenbach HD, Wenigerkind H, Brem G,Zakhartchenko V, et al. Tissue-specific elevated genomic cytosine methylationlevels are associated with an overgrowth phenotype of bovine fetuses derivedby in vitro techniques. Biol Reprod 2004;71(1):217–23.

[46] Monk D, Arnaud P, Apostolidou S, Hills FA, Kelsey G, Stanier P, et al. Limitedevolutionary conservation of imprinting in the human placenta. Proc NatlAcad Sci U S A 2006;103(17):6623–8.

[47] Hemberger M. Epigenetic landscape required for placental development. CellMol Life Sci 2007;64(18):2422–36.

[48] Arima T, Hata K, Tanaka S, Kusumi M, Li E, Kato K, et al. Loss of the maternalimprint in Dnmt3Lmat�/� mice leads to a differentiation defect in theextraembryonic tissue. Dev Biol 2006;297(2):361–73.

[49] Li E, Bestor TH, Jaenisch R. Targeted mutation of the DNA methyltransferasegene results in embryonic lethality. Cell 1992;69(6):915–26.

[50] Christensen BC, Houseman EA, Marsit CJ, Zheng S, Wrensch MR,Wiemels JL, et al. Aging and environmental exposures alter tissue-specificDNA methylation dependent upon CpG island context. PLoS Genet 2009;5(8). e1000602.

[51] Chu T, Burke B, Bunce K, Surti U, Allen Hogge W, Peters DG. A microarray-based approach for the identification of epigenetic biomarkers for thenoninvasive diagnosis of fetal disease. Prenat Diagn; 2009.

[52] Papageorgiou EA, Fiegler H, Rakyan V, Beck S, Hulten M, Lamnissou K, et al.Sites of differential DNA methylation between placenta and peripheral blood:molecular markers for noninvasive prenatal diagnosis of aneuploidies. Am JPathol 2009;174(5):1609–18.

[53] Rakyan V, Down T, Thorne N, Flicek P, Kulesha E, Graf S, et al. An integratedresource for genome-wide identification and analysis of human tissue-specificdifferentially methylated regions (tDMRs). Genome Res; 2008.

[54] Guilleret I, Osterheld MC, Braunschweig R, Gastineau V, Taillens S, Benhattar J.Imprinting of tumor-suppressor genes in human placenta. Epigenetics2009;4(1):62–8.

[55] Novakovic B, Wong NC, Sibson M, Ng HK, Morley R, Manuelpillai U, et al. DNAmethylation-mediated down-regulation of DNA methylatransferase-1(DNMT1), is co-incident with, but not essential for, global hypomethylation inhuman placenta. J Biol Chem.; 2010 Jan 13 [Epub ahead of print].