at SciVerse ScienceDirect

Placenta 33 (2012) e3ee10

Contents lists available

Placenta

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

Maternal to offspring resource allocation in plants and mammals

J.F. Gutierrez-Marcos a,*, M. Constância b,c,d, G.J. Burton c,d

a School of Life Sciences, University of Warwick, Wellesbourne Campus, Coventry CV4 7AL, UKbMetabolic Research Laboratories, Department of Obstetrics and Gynaecology, University of Cambridge, Cambridge CB2 0SW, UKcNational Institute of Health Research, Cambridge Biomedical Research Center, Cambridge, UKdCentre for Trophoblast Research, Department of Physiology, Development and Neuroscience, University of Cambridge, Cambridge CB2 3EG, UK

a r t i c l e i n f o

Article history:Accepted 30 August 2012

Keywords:PlacentaEndospermImprintingNutrients

* Corresponding author. Tel.: þ44 2476575077.E-mail addresses: j.f.gutierrez-marcos@warwick.a

jmasmc2@cam.ac.uk (M. Constância), gjb2@cam.ac.uk

0143-4004/$ e see front matter Crown Copyright � 2http://dx.doi.org/10.1016/j.placenta.2012.08.006

a b s t r a c t

Appropriate allocation of resources to the offspring is critical for successful reproduction, particularly inspecies that reproduce on more than one occasion. The offspring must be provisioned adequately toensure its vigour, whereas the parent must not become so depleted such that its survival is endangered.In both flowering plants and mammals specialised structures have evolved to support the offspringduring its development. In this review we consider common themes that may indicate conservation ofnutrient transfer function and regulation by genomic imprinting across the two kingdoms.

Crown Copyright � 2012 Published by Elsevier Ltd. All rights reserved.

1. Introduction

The success of reproduction in plants and mammals relies toa large extent on the balanced allocation of resources to thedeveloping embryo. Remarkably, the development and function ofplacental tissues in plants and mammals is in part epigeneticallyregulated. In this review, we discuss the similar strategies that haveevolved in both kingdoms to support maternal nutrient transfer tothe offspring, their regulation by genomic imprinting and wepropose a hitherto unrecognized role for placentally-derivednutrients on ensuring the epigenetic stability of the offspring.

2. Placental/endosperm tissues support nutrition of theoffspring in plants and mammals

The seeds of flowering plants are the result of a double fertil-ization event involving union of two sperm cells with a haploid eggand sister diploid central cell to form a diploid embryo and a trip-loid endosperm, respectively [1,2]. Hence, the endosperm has thesame genetic composition as the embryo (Fig. 1). Whereas theembryo develops into a mature plant, the development of theendosperm is confined to the seed stage. The endosperm is a keyfeature of flowering plants, however its evolutionary origin is notyet clear. It is likely that the endosperm arose from a secondsuccessful fertilization event, which resulted in an altruistic

c.uk (J.F. Gutierrez-Marcos),(G.J. Burton).

012 Published by Elsevier Ltd. All

embryo that could act as a nourishing placenta-like tissue. This issupported by the fact that basal seed-producing plants like conifers(Gymnosperms) produce supernumerary embryos that providenutritive support to the development of the embryo [3], and by therecent finding inwater-lilies (Nymphaeales), an ancestral floweringplant, that developing embryos are nourished by a reduced diploidendosperm [4].

Traditionally the endosperm has been considered a simpletissue that stores nutrients and supports the establishment of theembryo during germination. However, the role of the endosperm ismore complex than initially thought. Upon fertilization of thecentral cell gamete, the endosperm rapidly divides through a waveof synchronous nuclear divisions to form a coenocyte or syncytiumwhile the embryo remains as zygote [5]. At this stage of seeddevelopment nutrients can be freely transported to the developingembryo from the maternal tissues (Fig. 2A). The endosperm hasnuclei arranged within a thin layer of cytoplasm at the periphery ofthe syncytium, which in some plant species undergoes synchro-nous cellularization until the cavity is filled. Cellularization of theendosperm constitutes a key stage in endosperm patterning [6],and thereafter the endosperm differentiates into various cell types,of which the nutrient-transfer cells are the first to become distin-guishable and required to facilitate the active transport of nutrientsto the embryo (Fig. 2B). Transfer cells are characterized by anextensive network of plasma membrane forming wall ingrowths,and are rich in mitochondria and rough endoplasmic reticulum, aswell as abundant sugar translocating enzymes that convert sucroseinto hexose sugars [6]. The action of invertases in establishinga sugar flux in thematernalefilial interface is of vital importance, asthey regulate the development not only of the endosperm but also

rights reserved.

Fig. 1. Schematic diagram highlighting the strategies adopted in mammals and plantsto translocate maternal resources to embryos. In plants, a sperm cell fertilises thehaploid egg cell and the homo-diploid central cell gametes to give rise to a diploidembryo and a triploid endosperm. Endosperm transfer cells contribute to the activetransport of nutrients to the developing embryo. In mammals, haploid sperm and eggcell gametes fuse to produce a diploid zygote, which later differentiates into a placentawith trophoblast cells actively translocating maternal nutrients to the embryo. Bothendosperm and placenta function are regulated epigenetically by imprinted genes andenvironmental cues.

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10e4

of the maternal surrounding tissue. The evolutionary origin ofendosperm transfer cells is not yet understood but in species withpersistent endosperms their function is critical for reproductivesuccess and possibly in plant speciation [2].

During the final stages of seed development the endosperm actsas a nutritional reservoir to sustain embryo growth post-germination and the establishment of the next generation, muchin the same way as the yolk and albumin of a reptilian or avian egg.By contrast in mammals, maternal nutrients are supplied duringpregnancy via the placenta, and after birth through lactation. Thereis no doubt that the mammalian extraembryonic membranes andplacenta have evolved from the reptilian egg. Indeed viviparity isseen in many reptilian species, in particular those living at highaltitudes or in latitudes thought to allow for better thermoregula-tion during development [7]. Seed formation and oviparity argu-ably place a greater short-term drain on parental resources thanviviparity, when the nutrients can be provided over the course ofgestation. However, in mammals it is likely that the burden variesgreatly depending on the length of pregnancy and the reproductivestrategy employed. Thus inmice, where gestation is 21 days and thetotal conceptus weight of a litter represents 30% of maternal weight

Fig. 2. Schematic diagram highlighting different strategies to translocate maternal resourcesmaternal resources are translocated to the embryo directly from vascular networks located itissues into the embryo by specialized transfer cells (green) located at the maternalefilial intunderlying maternal (grey) and fetal (blue) capillary networks. D, In haemochorial placetranslocated directly to fetal (brown) capillary networks.

at term, the burden will be high compared to in humans whereconceptus weight represents only 5% of maternal weight andgestation lasts for 9 months. These differences may explainwhy theepigenetic regulation by imprinting of placental function is morestringent in mice than in humans [8,9].

During early development of eutherian mammals, the first celldivisions of the fertilised egg create the morula, a ball of diploidtotipotent cells. The first cell lineage divergence occurs inconjunction with the transformation of the morula into the blas-tocyst, which comprises an outer wall of trophoblast cells thatcontribute to the placenta, and the inner cell mass from which theembryo develops. The placenta therefore has the same geneticstructure as the embryo (Fig. 1). In all species, the trophoblast of theplacenta will interact with the uterus following implantation, butthe nature of that interaction varies widely across species [10]. Inthose with non-invasive epitheliochorial placentas the trophoblastabuts the uterine epithelium, and nutrient exchange takes placebetween the underlying maternal and fetal capillary networks(Fig. 2C). In the most invasive forms the trophoblast erodes thematernal epithelium and endothelium so that maternal bloodbathes the trophoblast. This is the haemochorial form of placen-tation, and is seen in rodents and the higher primates, includingman (Fig. 2D). For many years it was assumed that these differentforms represented an evolutionary progression, but molecularphylogenetics have revealed that the epitheliochorial form is anacquired state that has arisen in several different orders byconvergent evolution [11,12]. The selective pressures that havefavoured this development are uncertain, but greater control overresource allocation is a possibility.

Linked to the difference in invasiveness are changes in thehistological nature of the trophoblastic epithelium. In epi-theliochorial placentas the trophoblast generally remains cellular,whereas in the haemochorial forms it undergoes a syncytialtransformation into a polarised, multinucleated epithelium devoidof intercellular clefts (Fig. 2B and C). Despite the different numberof cell layers between the maternal blood and the trophoblast,comparative studies have shown considerable functional similari-ties between the different placenta types [10]. Placental exchangeoccurs by simple diffusion (e.g. respiratory gases, free fatty acids),facilitated diffusion (e.g. glucose), active transport (e.g. aminoacids), and receptor-mediated endocytosis (e.g. immunoglobulins).Most relevant to the discussion here is the transport of glucose andamino acids, as capacity will be dependent upon the insertion ofspecific transmembrane proteins. Transport of glucose is facilitated

to developing embryos in flowering plants and mammals. A, In syncytial endosperms,n maternal tissues. B, In cellular endosperms, resources are translocated from maternalerface. C, In epitheliochorial placentas, maternal resources are translocated between thentas maternal resources are in direct contact with the trophoblast (orange) and are

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10 e5

by the presence of members of the SLC2A (GLUT) family of trans-porters in the apical and basal membranes of the trophoblast. Theyhave been localised in the syncytiotrophoblast of the human androdent haemochorial placentas [13,14], and the uterine andtrophoblast epithelia of the ovine and bovine epitheliochorialplacentas [15]. Amino acid transfer involves a number of differenttransporter proteins [16]. These can be classified into two maingroups, accumulative transporters that are capable of transportingagainst a concentration gradient and lead to an increase inconcentration within the trophoblast, and amino acid exchangersthat are able to alter the balance of different classes of amino acidswithout changing the overall concentration [17]. Of these, theSystem A transporters are particularly important, as they alonemediate the accumulative uptake of amino acids that can be uti-lised by exchangers. A number of isoforms of the System A trans-porters exist, and interestingly the gene encoding one of these,Slc38a4, is imprinted [18]. Changes in the expression of thesetransporters, including Slc38a4, that can be considered beneficiallyadaptive have been reported in mice following maternal under-nutrition or genetic manipulations that induce placental-fetal-

Table 1Extraembryonic-specific imprinted gene expression in the mouse and representative exa

Imprinted region Gene Function

Mouse Chr 2 Sfmbt2 Chromatin regulatory proteinof the Polycomb group

Mouse Chr 6(Peg 10 cluster)

Tfpi2 Protease inhibitor with roles incellular matrix remodelling

Ppp1r9a Regulator of actin cytoskeletondynamics

Mouse Chr 7(Igf2 cluster)

Ins2 Hormone (insulin)Igf2(P0) Placenta-specific isoform of the

growth factor Igf2

Mouse Chr 7(Kcnq1 cluster)

Th Tyrosine hydroxylaseTh (Pa2 and a3) Placenta-specific isoforms of ThTssc4 Potential tumour suppressor geneTspan32 Transmembrane protein with roles

in haematopoietic-cell functionAscl2 Transcription factor

Mouse Chr 7 Ano1 Subunit of calcium activated chloridechannels

Mouse Chr 8 Gab1 Signalling mediator of receptor tyrosinekinases

Mouse Chr17(Igf2r cluster)

Slc22a3 Organic cation transporter

Slc22a2 Organic cation transporterMaize Chr 4 Meg1 Signalling peptide

Maize Chr 5 Mee1 Signalling peptideMaize Chr 10 Fie1 Chromatin regulatory protein of the

Polycomb groupMaize Chr 6 Mez1 Chromatin regulatory protein of the

Polycomb groupRice Chr 8 Fie Chromatin regulatory protein of the

Polycomb groupArabidopsis Chr 1 MEA Chromatin regulatory protein of the

Polycomb groupArabidopsis Chr 1 PHE1 Transcription factorArabidopsis Chr 2 FIS2 Chromatin regulatory protein of the

Polycomb groupArabidopsis Chr3 MPC RNA binding proteinArabidopsis Chr 5 FH5 Cytoskeleton

Abbreviations: Pl, placenta; YS, Yolk Sac; VYS, Visceral Yolk Sac; En, endosperm; Em, embgenes expressed in the murine placenta see [25]; for imprinted expression in both embrywww.mousebook.org/catalog.php?catalog¼imprinting; For details about imprinted genemouse was recently questioned on the basis of contaminating maternal cells (for detail

mismatch in supply and demand [19,20]. Equally, in humansthere is an inverse relationship between placental amino aciduptake and birth weight [21], which is again suggestive of anadaptive response.

3. Epigenetic regulation of placental/endospermdevelopment in plants and mammals

The development and function of the mammalian placenta andplant seed endosperm are in part regulated by genomic imprinting,an epigenetic phenomenon that leads to changes in gene expres-sion depending on parental inheritance [22]. Genomic imprinting isthought to have evolved independently in these organismsto regulate the transfer of nutrients and thus the fate of theoffspring [23].

Imprinting in plants is largely confined to placental endospermtissues during early embryo development [24]. Mutant screens inArabidopsis and molecular screens in maize performed a decadeago identified the first group of imprinted genes in plants [25]. Mostgenes reported in these studies where imprinted in the endosperm

mples in plants.

Imprintedexpression

Expressedallele

Loss-of-function phenotypein placenta/seed

Pl, YS Mat Unknown

Pl, YS Mat Unknown

Pl, YS Mat Unknown

VYS Pat UnknownPl (labyrinthine-specific)

Pat Growth restriction; reduced surfacearea for nutrient transport; impaireddiffusional exchange of nutrients

Pl Mat UnknownPl Mat UnknownPl Mat UnknownPl Mat Unknown

Pl Mat Absence of spongiotrophoblast layerand an expansion of the giant celllayer

Pl Mat Unknown

Pl Pat Underdeveloped labyrinthine layer,with reduced number of trophoblastcells

Pl Mat Impaired transplacental flux ofmonoamines

Pl Mat UnknownEn (Transfercell-specific)

Mat Defects on endosperm transfer celldevelopment, with severe embryogrowth restrictions

En/Em Mat UnknownEn Mat Defects on endosperm development

En Mat Defects on imprinting regulation

En Mat Unknown

En/em Mat Endosperm over-proliferation

En Pat UnknownEn Mat Endosperm over-proliferation

En Mat Defects on endosperm proliferationEn Mat Defects on endosperm cellularisation

ryo; Mat, maternal expression; Pat, paternal expression. For details about imprintedonic and extra-embryonic tissues and details about imprinting functions see http://s in plants see [31]. NOTE: This table excludes genes whose imprinted status in thesee [27]).

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10e6

and displayed a maternal parent-of-origin pattern of expression(Table 1). However, recent molecular screens in Arabidopsis, maizeand rice using high-throughput sequencing has revealed over 300genes to be imprinted in the endosperm with near-equal distri-bution of genes showing either paternal or maternal expression[26e32]. Surprisingly, only a handful of plant imprinted genesshow functional overlap between different species; an AID/BRIGHTDNA-binding domain protein, an SRA-Domain methylcytosine-binding protein known as VARIANT IN METHYLATION5 (VIM5)and a Flavin-containing monooxygenase (YUCCA10) involved inhormone biosynthesis. The low conservation observed forimprinted genes in plants could be explained by differences inreproductive stage(s) analysed and to technical limitations (e.g. lackof molecular markers or depth of the studies). Nonetheless, the factthat a number of imprinted genes are conserved across highlydivergent plant species suggests that imprinted-specific expressionin plants confers a selective advantage, as is thought to be the casein mammals.

Although imprinting in plants is mainly confined to the endo-sperm, there are recent examples for imprinting in embryos(Table 1). The first identified plant embryo imprinted gene, Mee1, isalso imprinted in the endosperm and its parent-of-origin expres-sion correlates with a reduction in DNA methylation [33]. There arecurrently over 100 transcripts in Arabidopsis [34] andmore than 40transcripts in maize [30] that show a parent-of-origin pattern ofexpression in early developing embryos as well as in endosperms.However, it is unclear why their imprinted expression does notpersist thoughout embryo development when they are stable in theendosperm. It is tempting to speculate that early developingembryo and endosperm may share common mechanisms to regu-late imprinting that are no longer maintained in the embryo at latestages of seed development. Nevertheless, imprinting in plantembryosmay be a transient effect since parent-of-origin expressionhas not been detected in mature embryos [26,27] or mature plants[32,35,36]. Alternatively, genomic imprinting in plant embryoscould be a feature of the transient extra-embryonic cell lineage,known as the suspensor, which plays an important role in thetranslocation of nutrients from surrounding tissues.

Similar to mammals, plants have many imprinted genes derivedfromnon-coding regions [37]. Almost half of the imprinted genes inmaize are long-non-coding RNAs (lncRNAs), some of which derivefrom intergenic regions of other imprinted genes [32]. However, itis not clear if these lncRNAs play a regulatory role in imprinting inplants, as is the case in mammals [38]. Most mammalian imprintedlncRNAs are located in clusters and are regulated by imprintingcontrol elements, whereas plant imprinted lncRNAs are dispersedin different genomic regions [32]. Plants however also possess anabundant number of imprinted non-coding small RNAs in theendosperm [39,40]. These small RNAs are transcribed by a plant-specific RNA polymerase (pol IV), amplified by an RNA-dependantRNA polymerase (RDR2) and processed into small interferingRNAs (siRNAs) by Dicer-like protein (DCL3). In Arabidopsis, it hasbeen shown that siRNAs can direct de novo DNA methylation todiscrete loci in the genome by the action of an Argonaute protein(AGO4) and the Domains Rearranged Methyltransferase2 (DRM2)[41]. Although the function of these imprinted siRNAs is unknown,an intriguing possibility is that they are able to move tosurrounding tissues and induce de novo DNA methylation changes[42]. If this hypothesis is correct, imprinted siRNAs could act non-cell autonomously to epigenetically target discrete genomicregions during early stages of embryo development, perhaps todirect epigenetic modifications at imprinted loci.

Imprinting in mammals is thought to have co-evolved with thedevelopment of placental structures and the emergence ofmaternalefetal interactions. Accordingly, imprinting is found in

eutherians (placental mammals) and in marsupials, which havea rudimentary placenta, and is absent in egg-laying mammals andoviparous species. Several lines of evidence support the notion thatimprinting plays a ‘special’ role in the placenta. The early 1980’smouse embryological experiments that led to the discovery ofimprinting revealed profound effects in extra-embryonic tissues,mostly of the trophoblast lineage. Genetic studies of mice carryingpartial or complete uniparental disomies are also often implicatedin some aspect of placental function and/or embryonic growthregulation. These initial findings have now been confirmed by theidentification of over 150 imprinted genes/transcripts, many ofwhich have been shown to be expressed in the placenta [43,44].Interestingly, a number of genes have been found to be imprintedonly in the placenta or to contain placental-specific promoters, ofwhich the vast majority are maternally expressed (Table 1) [45,46].Importantly, mouse imprinting knock-out studies frequently showaltered placental development and function (Table 1 only refers toa subset of these; see [44,45,47] for full details). In broad terms,imprinted genes acting in the placenta can be classified in twoclasses: those that are essential for differentiation of placental celltypes and often result in embryonic lethality due to the severity ofthe placental defects; and those that are important regulators ofplacental size and mass, with consequences for nutrient transfer tothe fetus.

Imprinted expression in mammals is highly tissue-specific anddevelopmentally regulated. It occurs in placenta, embryo and ina variety of adult tissues. The epigenetic mechanisms that regulateimprinting are not unique to imprinted loci but act throughout thegenome to regulate gene expression. It is the parental allele-specificmarks that regulate imprinting. Imprinted centres (ICs) orchestratethe clustering of imprinted genes in the mammalian genome andcan behave in different ways. Non-coding RNAs that originate fromICs, and higher-order chromatin interactions established at ICs byinsulator proteins like CTCF are examples of the two main mecha-nisms that regulate imprinting at many mammalian loci. Bothmechanisms involve DNA methylation and histone modificationsbut their hierarchy and interactions with ICs are not fully under-stood. Interestingly, there might be differences in the evolution ofthe imprinting mechanisms between the placenta and the embryo.The observation that maintenance of placental imprinting mightoccur independently of DNA methylation and that this epigeneticmark is not always required for imprinted expression inmarsupials,led to the hypothesis that an ancestral mechanism, based onhistone modifications, was initially limited to the short-livedplacenta [48]. Imprinting in embryonic structures would havearisen subsequently, with the addition of DNA methylation toconfer longer-term epigenetic stability [48].

Is imprinting regulated in a similar fashion in mammals andplants? The current data for mammals indicate that imprinting iserased in primordial germ cells, is then reset in either male orfemale gametes and after fertilization is maintained in embryonicand adult tissues. However, plants set aside their germline untilflower development and do not undergo a genome-wide epigeneticreprogramming as is the case in mammals. Surprisingly, despitethese differences there is evidence in plants for erasure andresetting of imprints, but this cycle takes place during a narrowwindow of early embryo development, and not in endosperm [33].

Differential DNA methylation of parental alleles is the keyepigenetic mark for most imprinted genes identified in mammalsand plants. The differential methylation in mammals is establishedin gametes and after fertilization [49](known as gametic andsomatic methylation, respectively). However, in plants differentialmethylation is established mainly in the central cell gamete. InArabidopsis this is due to the action of a DNA glycosylase DEMETER(DME), which removes cytosine methylation from discrete loci

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10 e7

associated with transposons and repeat sequences [50e52] and bythe repression of the maintenance DNA methyltransferase (MET1)[27]. Therefore, maternally expressed genes are commonly associ-ated with demethylated regions proximal to transposable elements(TEs) and repetitive sequences and are preferentially expressed inthe endosperm but silenced in vegetative tissues. However,expression of some imprinted loci in Arabidopsis depends on theFERTILIZATION INDEPENDENT SEED (FIS) Polycomb-group repres-sive complex [24]. The FIS PcG complex deposits trimethylation onthe lysine 27 of histone H3 (H3K27me3) and functions in thecentral cell and endosperm, where it is necessary for regulatingearly seed development. Intriguingly, most of the imprinted genesregulated by the FIS complex also depend on DNA methylation[27,52,53]. This silencing mechanism appears to be more prevalentin Arabidopsis paternally expressed genes, like PHERES1 (PHE1),which depends on the demethylation of a downstream repetitiveregion that modulates the binding of the FIS complex to itspromoter [54,55]. Therefore, genomic imprinting in plants dependson at least two distinct epigenetic mechanisms. Furthermore,mutants defective in both DNA and H3K27me3 methylation havemajor effects on seed size and embryo development, thus sup-porting the view that imprinting in plants modulate the transfer ofnutrients to the embryo.

Why did imprinting evolve independently in plants andmammals? The origin of the imprinting mechanisms in mammalsis attributed to the silencing of foreign DNA elements [37]. Simi-larly, genomic imprinting in plants is thought to be a consequenceof the recruitment of epigenetic silencing mechanisms to target TEsand repeats during sexual reproduction [50]. This hypothesis issupported by the fact that most imprinted genes identified inArabidopsis are associated with TEs and repeats, which are acti-vated by DME-mediated demethylation in the endosperm [26,56].Demethylation of TEs and repeats in the endosperm leads to theproduction of non-coding sRNAs that could direct silencing of theseelements in the embryo. Therefore, imprinting in the endospermmay have originated as a by-product of silencing transposableelements in the embryo [50]. Although this so-called ‘defencehypothesis’ explains the origin of imprinting mechanisms in bothorganisms, it does not provide an explanation for the evolutionaryorigin of imprinting. Two theories have been proposed to addressthis caveat that proposes imprinting as a major evolutionary forceon the distribution of nutrients to the offspring. The first theoryproposes that imprinting resulted from an intergenomic conflictover the distribution of resources from mother to offspring [57,58].In agreement with this theory, imprinting occurs in placentaltissues in both mammals and flowering plants, and a significantnumber of imprinted genes in both species affect the growth of theoffspring. An alternative theory postulates that genomic imprintingis the result of natural selection to increase offspring fitness byenhancing the genetic integration of coadapted offspring andmaternal traits [59]. In agreement with this theory, there area significant number of imprinted genes in mammals and plantswith maternal-specific expression that regulate maternal-offspringinteractions [60].

4. Regulation of resource allocation by genomic imprinting inmammals

A major function of mammalian imprinting is the regulation ofmaternal investment in pregnancy [61]. The imprinted insulin-likegrowth factor 2 (Igf2) gene plays a central role in this process as itcombines and balances the genetic control of supply (throughexpression in the placenta) with the genetic control of fetal demandfor nutrients (i.e. through expression in the fetus) [62]. Igf2 isa paternally expressed gene and encodes one of the most potent

placental-fetal growth factors in all vertebrate species (Igf2 is partof insulin/IGF signalling system, which also contains anotherimprinted gene, Igf2 type 2 receptor, Igf2r, that regulates Igf2 levelsnegatively). The function of Igf2 in maternalefetal nutrient allo-cation has been mainly revealed through studies in mouse models(described below).

4.1. The Igf2 P0 and the placental supply of nutrients

Igf2 is highly expressed in all placental cell types as well as in themajority of fetal organs (Fig. 3A). With the discovery of a mouseplacenta-specific promoter (Igf2 P0) [63] it became possible toseparate the direct actions of Igf2 in the placenta from that of fetaltissues using genetic engineering in mice [64]. Imprinted expres-sion and placental-specificity of the P0 promoter is, at least in part,determined by DNA methylation. In fetal tissues the P0 promoter issilenced by DNA hypermetylation on both parental alleles (Fig. 3A).In the placenta the paternally derived promoter is unmethylatedand active specifically in labyrinthine trophoblast cells, whereas thematernal allele is methylated (Fig. 3A). The Igf2 P0 deficient miceshow early placental growth restriction that affects all layers of theplacenta [64]. Fetuses become growth restricted in late gestationdue to the placental insufficiency in supplying nutrients. Theseelegant genetic experiments were the first to demonstrate a crucialrole for an imprinted gene in the placental supply of maternalnutrients to the mammalian fetus. This work also highlightedimportant compensatory responses of the small P0 placenta to fetaldemands for growth, which occurred during mid-gestation (theseadaptations included increases in glucose, System A amino acid andcalcium transport) [18,65,66]. Igf2 P0 seems to act as a sensor ofenvironmental signals and is thought to play a role in placentaladaptive responses to sub-optimal conditions (reviewed in [66]).Accordingly, Igf2 P0 is responsive to a wide range of environmentalstimuli, including dietary composition and glucocorticoid concen-trations, and the placental adaptations that enhance fetal System Aamino acid supply during maternal undernutrition were recentlyshown not to occur in Igf2 P0 mutants [66].

4.2. Fetal Igf2 promoters and the demand for nutrients

Igf2 produced from fetal promoters (P1-P3 in the mouse)controls the growth of fetal organs and coordinates whole bodygrowth with the growth of the diverse organs via endocrinemechanisms (Igf2 is secreted into the bloodstream at very highconcentrations, in particular in the fetal period). The level of fetaldemand for maternal nutrients through the placenta can thus bedetermined, at least in part, by the proliferative and endocrineactions of fetal Igf2. Circulating Igf2 is thus an attractive candidatefor a ‘fetal signal’ of nutrient demand. Mice with reduced fetal Igf2and therefore circulating Igf2 influences transplacental resourceallocation from mother to the fetus by compromising placentallabyrinthine growth and vasculogenesis (Sandovici & Constância,unpublished observations).

5. The supply and demand hypothesis of mammalianimprinting

The supply and demand concept illustrated here for Igf2 P0/Igf2can be extended to the imprinted gene cluster surrounding Igf2 andto other imprinted genes and imprinted clusters (e.g. Igf2r, Peg 10;Table 1 and Fig. 3) (and can perhaps provide a functional expla-nation for clustering in the genome) [62]. For example, the domi-nant positive signals by Igf2 are negatively counteracted by a varietyof maternally expressed supply/demand suppressors in the vicinityof Igf2 (e.g. Cdkn1c, Phlda2). The balance between positive (by

Fig. 3. Placental-embryonic interactions are regulated by genomic imprinting. A, Left panel, in situ mRNA hybridization in a mouse embryo at E13.5 showing that Igf2 is highlyexpressed in the placenta and in fetal organs (except the brain) (courtesy of Mr. Paul Smith, The Babraham Institute). Right panel, imprinting of the placental-specific Igf2 promoter(P0) is controlled by DNA methylation. P0 is only found unmethylated and active from the paternal allele in the placenta, and methylated and inactive in fetal tissues. The other fetaland placental Igf2 promoters (P1-P3) are unmethylated on both alleles, with imprinted expression mainly controlled by the Imprinting Control Region (ICR) upstream of the H19gene [63]. B, Left panel, in situ mRNA hybridization in a maize seed at 10 days after fertilization showing the specific expression of Meg1 in endosperm transfer cells. Right panel,imprinting of Meg1 in the endosperm is controlled by the interplay between DNA demethylation of the maternally inherited alleles and the transcriptional activation of MRP1. Inembryo, both parental alleles are demethylated but not expressed due to the absence of MRP1. C, Schematic diagram describing the interaction between fetal, placental andmaternal signals in mammals. Imprinted genes are important genetic regulators of the balance between placental supply and fetal demand systems for maternal nutrients.Expression of imprinted genes in the fetus and the placenta influence transplacental resource allocation in a variety of ways, and include signalling of mismatch between demandand supply of nutrients (see text for details). Shown below are some of the key imprinted genes with roles in allocation of maternal resources.

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10e8

paternally expressed imprinted genes) and negative (by maternallyexpressed genes) effects on supply and demand are proposed to becontrolled by differential epigenetic modifications at ICs ina parent-specific manner. The regulation of the pivotal balancebetween supply and demand of nutrients in mammals might bea specific selected function of some imprinted genes (Fig. 3C),which may have co-evolved with placentation [62]. In support ofthis hypothesis, a network of co-regulated imprinted genesinvolved in the control of embryonic growth has been identified[67]. The transcription factor Zac1/Plagl1 is a mainmodulator of thisnetwork that includes responsive genes such as Igf2, H19, Cdkn1cand Dlk1 [67].

But how do imprinted genes regulate placental nutrient transferin mammals? Several lines of evidence show that they can exertmultiple levels of control, from modulation of the growth, size andmorphology of placenta, to effects on thickness and permeabilityproperties of the exchange barrier for nutrient transport, to more‘indirect’ roles such as regulation of placental hormone secretionthat manipulate maternal physiology [66]. It is becoming increas-ingly clear that many imprinted genes are expressed in the endo-crine layer of the murine placenta, where they may play importantroles (in the human placenta, both transport and endocrine

functions are provided by the syncytiotrophoblast) [66]. A diversityof phenotypes, mainly related to altered differentiation potentialand growth of endocrine cells, have indeed been described inimprinted mouse knock-outs that are thought to impair thesecretion of placental hormones into the maternal circulation, butthis principle remains largely untested (reviewed in [66]).

6. Imprinting in plants and the endosperm supply ofnutrients

In plants, most of the imprinted genes functionally character-ized appear to regulate different aspects of endosperm develop-ment (Table 1). The imprinted genes MEA and FIS2, both membersof the Polycomb Repressive Complex 2 (PRC2), regulate endospermproliferation in Arabidopsis and mutations in these genes confermaternal effects on seed development [25]. Interestingly, in bothmutants embryos overproliferate [68,69] resembling the embryoovergrowth phenotype observed in the mouse Igf2r mutant [70].Other imprinted genes such as FH5 and MPC regulate the balancebetween endosperm proliferation and cellularization, and bothhave a major impact on embryo growth [71,72](Table 1). Theseobservations have given support to the hypothesis that imprinted

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10 e9

genes in plants regulate endosperm development and thus thesupply of nutrients to the developing embryo. The recent charac-terization of an imprinted gene,Maternally expressed gene 1 (Meg1),which directly regulates the transfer of maternal nutrientsstrengthens the view that imprinting in the endosperm directlyinfluences offspring nourishment [73]. Intriguingly, Meg1 encodesa small signalling peptide, which is expressed in endosperm soonafter fertilisation, prior to the differentiation of the transfer tissue[74]. At this critical stage of early seed development, Meg1 controlsthe expression of the single-myb transcription factor MRP1,through a feedback mechanism to regulate transfer cell geneexpression and subsequent differentiation [73]. Interestingly,during early stages of endosperm development the maternallyderived Meg1 promoter is not methylated and active specifically intransfer cells by action of MRP1, whereas the paternal allele ismethylated and transcriptionally inactive (Fig. 3B). However, at latestages of endosperm development both paternal and maternalalleles are demethylated and transcriptionally active, thus indi-cating that methylation is the main regulatory pathway for Meg1imprinting. Significantly, Meg1 e like other imprinted genes inmammals e acts in a dosage-dependant manner and loss ofimprinting leads to the over-proliferation of transfer tissues, anincrease in nutrient supply, and consequently larger seeds.

A role for genomic imprinting in regulating nutrient demand inplants has never been demonstrated. However, it is possible thatlike in mammals (Fig. 3D) imprinted genes in plant embryos couldinfluence endosperm development to regulate nutrient supplyaccording to demand. Several lines of evidence indicate that thedevelopment of the embryo and endosperm in plants is tightly co-ordinated by short-range signalling peptides [75]. The recentidentification of the small, secreted peptide Mee1, which isimprinted in the embryo, opens a tantalising possibility for theexistence of imprinted embryo factors that might regulate theallocation of maternal resources by influencing endospermdevelopment.

7. Future challenges

Modifications to the balance between the supply and demand ofnutrients have major implications for fetal growth and the intra-uterine programming of adult disease in mammals. It will beimportant to fully identify the co-regulated networks of imprintedgenes that act in the placenta and the fetus, and how these aremodified epigenetically by environmental factors, in particularmaternal nutrition. Because of thewidespread occurrence of tissue-specific imprinting in mammals, it is conceivable that a number ofnew genes will be found to be specifically imprinted within certaincell types of the placenta. Functional studies are needed to betterunderstand the function of placenta-specific imprinting networkson supply transfer to the fetus and their impact on maternalphysiology, a promising area of research that is currently poorlyunderstood.

While the plant endosperm is considered a key regulator ofnutrient supply to the developing embryo, we still remain ignorantof the precise mechanisms that regulate this process. One tanta-lising possibility is that, as in mammals, endosperm imprintedgenes contribute to regulating maternal nutrient supply whileembryonic imprinted genes may regulate nutrient demand(Fig. 3D). This theory could be easily tested by the systematicfunctional characterization of imprinted genes in plants.

The functional analysis of imprinted genes in both mammals andplants could potentially uncover a hitherto unrecognised role forplacental-derived nutrients on shaping the epigeneticmakeup of theoffspring. In such cases, the regulation of placental function through

imprinting could minimise nutritive fluctuations during earlyembryogenesis, thusensuring theepigenetic stabilityof theoffspring.

Conflict of interest

None declared.

Acknowledgements

Work in the authors’ groups is funded by the Royal Society (JGM),ESF/RTD, Framework COST Action (FA0903) (JGM), the Biotech-nology and Biological Sciences Research Council (JGM,MC, GJB), theMedical Research Council (MC) and the Wellcome Trust (GJB).

References

[1] Berger F. Endosperm development. Current Opinion in Plant Biology 1999;2(1):28e32.

[2] Costa LM, Gutierrez-Marcos JF, Dickinson HG. More than a yolk: the short lifeand complex times of the plant endosperm. Trends in Plant Science 2004;9(10):507e14.

[3] Friedman WE. Evidence of a pre-angiosperm origin of endosperm: implica-tions for the evolution of flowering plants. Science 1992;255(5042):336e9.

[4] Friedman WE. Hydatellaceae are water lilies with gymnospermous tenden-cies. Nature 2008;453(7191):94e7.

[5] Olsen OA. ENDOSPERM DEVELOPMENT: cellularization and cell fate specifi-cation. Annual Review of Plant Physiology and Plant Molecular Biology 2001;52:233e67.

[6] Costa LM, Gutierrez-Marcos JF, Brutnell TP, Greenland AJ, Dickinson HG. Theglobby1-1 (glo1-1) mutation disrupts nuclear and cell division in the devel-oping maize seed causing alterations in endosperm cell fate and tissuedifferentiation. Development 2003;130(20):5009e17.

[7] Blackburn DG. Reptilian viviparity: past research, future directions, andappropriate models. Comparative Biochemistry and Physiology A-Molecularand Integrative Physiology 2000;127(4):391e409.

[8] Morison IM, Ramsay JP, Spencer HG. A census of mammalian imprinting.Trends in Genetics 2005;21(8):457e65.

[9] Monk D, Arnaud P, Apostolidou S, Hills FA, Kelsey G, Stanier P, et al. Limitedevolutionary conservation of imprinting in the human placenta. Proceedingsof the National Academy of Sciences of the United States of America 2006;103(17):6623e8.

[10] Wooding FP, Burton GJ. Comparative placentation. structures, functions andevolution. Berlin: Springer; 2008. p. 301.

[11] Wildman DE, Chen C, Erez O, Grossman LI, Goodman M, Romero R. Evolutionof the mammalian placenta revealed by phylogenetic analysis. Proceedings ofthe National Academy of Sciences of the United States of America 2006;103(9):3203e8.

[12] Carter AM, Mess A. Evolution of the placenta in eutherian mammals. Placenta2007;28(4):259e62.

[13] Knipp GT, Audus KL, Soares MJ. Nutrient transport across the placenta.Advanced Drug Delivery Reviews 1999;38(1):41e58.

[14] Illsley NP. Glucose transporters in the human placenta. Placenta 2000;21(1):14e22.

[15] Wooding FB, Fowden AL, Bell AW, Ehrhardt RA, Limesand SW, Hay WW.Localisation of glucose transport in the ruminant placenta: implications forsequential use of transporter isoforms. Placenta 2005;26(8e-9):626e40.

[16] Battaglia FC, Regnault TR. Placental transport and metabolism of amino acids.Placenta 2001;22(2e3):145e61.

[17] Cleal JK, Lewis RM. The mechanisms and regulation of placental amino acidtransport to thehumanfetus. JournalofNeuroendocrinology2008;20(4):419e26.

[18] Constancia M, Angiolini E, Sandovici I, Smith P, Smith R, Kelsey G, et al.Adaptation of nutrient supply to fetal demand in the mouse involves inter-action between the Igf2 gene and placental transporter systems. Proceedingsof the National Academy of Sciences of the United States of America 2005;102(52):19219e24.

[19] Angiolini E, Coan PM, Sandovici I, Iwajomo OH, Peck G, Burton GJ, et al.Developmental adaptations to increased fetal nutrient demand in mousegenetic models of Igf2-mediated overgrowth. FASEB Journal: Official Publi-cation of the Federation of American Societies for Experimental Biology 2011;25(5):1737e45.

[20] Burton GJ, Fowden AL. Review: the placenta and developmental program-ming: balancing fetal nutrient demands with maternal resource allocation.Placenta 2012;33(Suppl):S23e7.

[21] Godfrey KM, Matthews N, Glazier J, Jackson A, Wilman C, Sibley CP. Neutralamino acid uptake by the microvillous plasma membrane of the humanplacenta is inversely related to fetal size at birth in normal pregnancy. JournalOf Clinical Endocrinology And Metabolism 1998;83(9):3320e6.

[22] Reik W, Walter J. Genomic imprinting: parental influence on the genome.Nature Reviews Genetics 2001;2(1):21e32.

J.F. Gutierrez-Marcos et al. / Placenta 33 (2012) e3ee10e10

[23] Feil R, Berger F. Convergent evolution of genomic imprinting in plants andmammals. Trends in Genetics: TIG 2007;23(4):192e9.

[24] Kohler C, Wolff P, Spillane C. Epigenetic mechanisms underlying genomicimprinting in plants. Annual Review of Plant Biology 2012;63:331e52.

[25] Raissig MT, Baroux C, Grossniklaus U. Regulation and Flexibility of genomicimprinting during seed development. The Plant Cell 2011;23(1):16e26.

[26] Gehring M, Missirian V, Henikoff S. Genomic analysis of parent-of-originallelic expression in Arabidopsis thaliana seeds. PloS One 2011;6(8):e23687.

[27] Hsieh TF, Shin J, Uzawa R, Silva P, Cohen S, Bauer MJ, et al. Regulation ofimprinted gene expression in Arabidopsis endosperm. Proceedings of theNational Academy of Sciences of the United States of America 2011;108(5):1755e62.

[28] Luo M, Taylor JM, Spriggs A, Zhang H, Wu X, Russell S, et al. A genome-widesurvey of imprinted genes in rice seeds reveals imprinting primarily occurs inthe endosperm. PLoS Genetics 2011;7(6):e1002125.

[29] McKeown PC, Laouielle-Duprat S, Prins P, Wolff P, Schmid MW, Donoghue MT,et al. Identification of imprinted genes subject to parent-of-origin specificexpression in Arabidopsis thaliana seeds. BMC Plant Biology 2011;11:113.

[30] Waters AJ, Makarevitch I, Eichten SR, Swanson-Wagner RA, Yeh CT, Xu W,et al. Parent-of-origin effects on gene expression and DNA methylation in themaize endosperm. The Plant Cell 2011;23(12):4221e33.

[31] Wolff P, Weinhofer I, Seguin J, Roszak P, Beisel C, Donoghue MT, et al. High-resolution analysis of parent-of-origin allelic expression in the ArabidopsisEndosperm. PLoS Genetics 2011;7(6):e1002126.

[32] Zhang M, Zhao H, Xie S, Chen J, Xu Y, Wang K, et al. Extensive, clusteredparental imprinting of protein-coding and noncoding RNAs in developingmaize endosperm. Proceedings of the National Academy of Sciences of theUnited States of America 2011;108(50):20042e7.

[33] Jahnke S, Scholten S. Epigenetic resetting of a gene imprinted in plantembryos. Current Biology: CB 2009;19(19):1677e81.

[34] Nodine MD, Bartel DP. Maternal and paternal genomes contribute equally tothe transcriptome of early plant embryos. Nature 2012;482(7383):94e7.

[35] Fujimoto R, Taylor JM, Shirasawa S, Peacock WJ, Dennis ES. Heterosis ofArabidopsis hybrids between C24 and Col is associated with increasedphotosynthesis capacity. Proceedings of the National Academy of Sciences ofthe United States of America 2012;109(18):7109e14.

[36] Groszmann M, Greaves IK, Albertyn ZI, Scofield GN, Peacock WJ, Dennis ES.Changes in 24-nt siRNA levels in Arabidopsis hybrids suggest an epigeneticcontribution to hybrid vigor. Proceedings of the National Academy of Sciencesof the United States of America 2011;108(6):2617e22.

[37] Barlow DP. Methylation and imprinting: from host defense to gene regula-tion? Science 1993;260(5106):309e10.

[38] Barlow DP. Genomic imprinting: a mammalian epigenetic discovery model.Annu Rev Genet 2011;45:379e403.

[39] Mosher RA. Maternal control of Pol IV-dependent siRNAs in Arabidopsisendosperm. New Phytologist 2010;186(2):358e64.

[40] Mosher RA, Melnyk CW, Kelly KA, Dunn RM, Studholme DJ, Baulcombe DC.Uniparental expression of PolIV-dependent siRNAs in developing endospermof Arabidopsis. Nature 2009;460(7252):283e6.

[41] Law JA, Jacobsen SE. Establishing, maintaining and modifying DNA methyla-tion patterns in plants and animals. Nature Reviews Genetics 2010;11(3):204e20.

[42] Mosher RA, Melnyk CW. siRNAs and DNA methylation: seedy epigenetics.Trends in Plant Science 2010;15(4):204e10.

[43] Frost JM, Moore GE. The importance of imprinting in the human placenta.PLoS Genetics 2010;6(7):e1001015.

[44] Lefebvre L. The placental imprintome and imprinted gene function in thetrophoblast glycogen cell lineage. Reproductive Biomedicine Online 2012.

[45] Fowden AL, Coan PM, Angiolini E, Burton GJ, Constancia M. Imprinted genesand the epigenetic regulation of placental phenotype. Progress in Biophysicsand Molecular Biology 2011;106(1):281e8.

[46] Okae H, Hiura H, Nishida Y, Funayama R, Tanaka S, Chiba H, et al.Re-investigation and RNA sequencing-based identification of genes withplacenta-specific imprinted expression. Human Molecular Genetics 2012;21(3):548e58.

[47] Bressan FF, De Bem TH, Perecin F, Lopes FL, Ambrosio CE, Meirelles FV, et al.Unearthing the roles of imprinted genes in the placenta. Placenta 2009;30(10):823e34.

[48] Lewis A, Mitsuya K, Umlauf D, Smith P, Dean W, Walter J, et al. Imprintingon distal chromosome 7 in the placenta involves repressive histonemethylation independent of DNA methylation. Nature Genetics 2004;36(12):1291e5.

[49] Ferguson-Smith AC. Genomic imprinting: the emergence of an epigeneticparadigm. Nature Reviews Genetics 2011;12(8):565e75.

[50] Hsieh TF, Ibarra CA, Silva P, Zemach A, Eshed-Williams L, Fischer RL, et al.Genome-wide demethylation of Arabidopsis endosperm. Science 2009;324(5933):1451e4.

[51] Kinoshita T, Miura A, Choi Y, Kinoshita Y, Cao X, Jacobsen SE, et al. One-waycontrol of FWA imprinting in Arabidopsis endosperm by DNA methylation.Science 2004;303(5657):521e3.

[52] Gehring M, Huh JH, Hsieh TF, Penterman J, Choi Y, Harada JJ, et al. DEMETERDNA glycosylase establishes MEDEA polycomb gene self-imprinting by allele-specific demethylation. Cell 2006;124(3):495e506.

[53] Weinhofer I, Hehenberger E, Roszak P, Hennig L, Kohler C. H3K27me3profiling of the endosperm implies exclusion of polycomb group proteintargeting by DNA methylation. PLoS Genetics 2010;6(10).

[54] Makarevich G, Villar CB, Erilova A, Kohler C. Mechanism of PHERES1imprinting in Arabidopsis. Journal of Cell Science 2008;121(Pt 6):906e12.

[55] Villar CB, Erilova A, Makarevich G, Trosch R, Kohler C. Control of PHERES1imprinting in Arabidopsis by direct tandem repeats. Molecular Plant 2009;2(4):654e60.

[56] Gehring M, Bubb KL, Henikoff S. Extensive demethylation of repetitiveelements during seed development underlies gene imprinting. Science 2009;324(5933):1447e51.

[57] Haig D, Westoby M. Parent specific gene expression and the triploid endo-sperm. American Naturalist 1989;134:147e55.

[58] Trivers R, Burt A. Kinship and genomic imprinting. Results and Problems inCell Differentiation 1999;25:1e21.

[59] Wolf JB, Hager R. A maternal-offspring coadaptation theory for the evolutionof genomic imprinting. PLoS Biology 2006;4(12):e380.

[60] Luo M, Bilodeau P, Koltunow A, Dennis ES, Peacock WJ, Chaudhury AM. Genescontrolling fertilization-independent seed development in Arabidopsis thali-ana. Proceedings of the National Academy of Sciences of the United States ofAmerica 1999;96(1):296e301.

[61] Constancia M, Kelsey G, Reik W. Resourceful imprinting. Nature 2004;432(7013):53e7.

[62] Reik W, Constancia M, Fowden A, Anderson N, Dean W, Ferguson-Smith A,et al. Regulation of supply and demand for maternal nutrients in mammals byimprinted genes. The Journal of Physiology 2003;547(Pt 1):35e44.

[63] Moore T, Constancia M, Zubair M, Bailleul B, Feil R, Sasaki H, et al. Multipleimprinted sense and antisense transcripts, differential methylation andtandem repeats in a putative imprinting control region upstream of mouseIgf2. Proceedings of the National Academy of Sciences of the United States ofAmerica 1997;94(23):12509e14.

[64] Constancia M, Hemberger M, Hughes J, Dean W, Ferguson-Smith A, Fundele R,et al. Placental-specific IGF-II is a major modulator of placental and fetalgrowth. Nature 2002;417(6892):945e8.

[65] Dilworth MR, Kusinski LC, Cowley E, Ward BS, Husain SM, Constancia M, et al.Placental-specific Igf2 knockout mice exhibit hypocalcemia and adaptivechanges in placental calcium transport. Proceedings of the National Academyof Sciences of the United States of America 2010;107(8):3894e9.

[66] Sandovici I, Hoelle K, Angiolini E, Constancia M. Placental adaptations to thematernal-fetal environment: implications for fetal growth and developmentalprogramming. Reproductive Biomedicine Online 2012.

[67] Varrault A, Gueydan C, Delalbre A, Bellmann A, Houssami S, Aknin C, et al.Zac1 regulates an imprinted gene network critically involved in the control ofembryonic growth. Developmental Cell 2006;11(5):711e22.

[68] Grossniklaus U, Vielle-Calzada JP, Hoeppner MA, Gagliano WB. Maternalcontrol of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis.Science 1998;280(5362):446e50.

[69] Luo M, Bilodeau P, Dennis ES, Peacock WJ, Chaudhury A. Expression andparent-of-origin effects for FIS2, MEA, and FIE in the endosperm and embryoof developing Arabidopsis seeds. Proceedings of the National Academy ofSciences of the United States of America 2000;97(19):10637e42.

[70] Barlow DP, Stoger R, Herrmann BG, Saito K, Schweifer N. The mouse insulin-like growth factor type-2 receptor is imprinted and closely linked to the Tmelocus. Nature 1991;349(6304):84e7.

[71] Fitz Gerald JN, Hui PS, Berger F. Polycomb group-dependent imprinting of theactin regulator AtFH5 regulates morphogenesis in Arabidopsis thaliana.Development 2009;136(20):3399e404.

[72] Tiwari S, Schulz R, Ikeda Y, Dytham L, Bravo J, Mathers L, et al. MATERNALLYEXPRESSED PAB C-TERMINAL, a novel imprinted gene in Arabidopsis, encodesthe conserved C-terminal domain of polyadenylate binding proteins. ThePlant Cell 2008;20(9):2387e98.

[73] Costa LM, Yuan J, Rouster J, Paul W, Dickinson H, Gutierrez-Marcos JF.Maternal control of nutrient allocation in plant seeds by genomic imprinting.Current Biology: CB 2012;22(2):160e5.

[74] Gutierrez-Marcos JF, Costa LM, Biderre-Petit C, Khbaya B, O’Sullivan DM,Wormald M, et al. Maternally expressed gene1 Is a novel maize endospermtransfer cell-specific gene with a maternal parent-of-origin pattern ofexpression. The Plant Cell 2004;16(5):1288e301.

[75] Marshall E, Costa LM, Gutierrez-Marcos J. Cysteine-rich peptides (CRPs)mediate diverse aspects of cell-cell communication in plant reproduction anddevelopment. Journal of Experimental Botany 2011;62(5):1677e86.