Placental phenotype and the insulin‐like growth factors: resource allocation to fetal growth

Abstract

The placenta is the main determinant of fetal growth and development in utero. It supplies all the nutrients and oxygen required for fetal growth and secretes hormones that facilitate maternal allocation of nutrients to the fetus. Furthermore, the placenta responds to nutritional and metabolic signals in the mother by altering its structural and functional phenotype, which can lead to changes in maternal resource allocation to the fetus. The molecular mechanisms by which the placenta senses and responds to environmental cues are poorly understood. This review discusses the role of the insulin‐like growth factors (IGFs) in controlling placental resource allocation to fetal growth, particularly in response to adverse gestational environments. In particular, it assesses the impact of the IGFs and their signalling machinery on placental morphogenesis, substrate transport and hormone secretion, primarily in the laboratory species, although it draws on data from human and other species where relevant. It also considers the role of the IGFs as environmental signals in linking resource availability to fetal growth through changes in the morphological and functional phenotype of the placenta. As altered fetal growth is associated with increased perinatal morbidity and mortality and a greater risk of developing adult‐onset diseases in later life, understanding the role of IGFs during pregnancy in regulating placental resource allocation to fetal growth is important for identifying the mechanisms underlying the developmental programming of offspring phenotype by suboptimal intrauterine growth.

Abbreviations

AKT

protein kinase B

GLUT1

glucose transporter 1

IGF

insulin‐like growth factor

IGF1R

type 1 IGF receptor

IGF2R

type 2 IGF receptor

INSR

insulin receptor

Jz

junctional zone

LAT

L‐type amino acid transporter

Lz

labyrinthine zone

MAPK

mitogen‐activated protein kinase

mTORC1

mechanistic target of rapamycin

PI3K

phosphoinositide‐3 kinase

SNAT/Slc38a

sodium‐coupled amino acid transporter

Introduction

Intrauterine growth is a key determinant of lifespan. Babies born growth restricted or large for gestational age are at greater risk of perinatal morbidity and mortality than those of normal birth weight. Moreover, the ‘memories’ of an altered environment and growth in utero can stretch beyond the perinatal period to influence health much later in life. Epidemiological studies in humans have shown that babies grown abnormally due to poor maternal nutrition are at heightened risk of developing conditions such as type 2 diabetes, heart disease and obesity as adults, and of dying younger as a consequence (Gluckman et al. 2005; Jansson & Powell, 2006). Similarly, manipulating intrauterine growth experimentally by varying maternal food intake, dietary composition, oxygen availability, endocrine status or utero‐placental blood flow has been shown to programme cardiovascular, metabolic and endocrine function of the adult offspring in a wide range of mammalian species (Gluckman et al. 2005; McMillen & Robinson, 2005; Fowden et al. 2006).

As the interface between the mother and fetus, the placenta is one of the main determinants of intrauterine growth. It supplies all the nutrients and oxygen required for fetal growth as well as secreting hormones that influence maternal metabolism in favour of the fetal needs. Its morphological and functional characteristics, therefore, have an important role in determining the allocation of maternal resources to fetal growth. These characteristics include cell composition, surface area, barrier thickness, blood flow, vascularity, nutrient utilisation and the abundance and activity of the various transporter molecules (Fowden et al. 2009; Sandovici et al. 2012). Recent studies have shown that the placenta can respond to maternal nutritional and metabolic signals by altering these characteristics which, in turn, leads to changes in the placental capacity to supply resources to the fetus (Fowden et al. 2009; Sandovici et al. 2012). Thus, the placenta is a key mediator in linking maternal environmental conditions to development of the fetus (Burton et al. 2016; Sferruzzi‐Perri & Camm, 2016). However, the molecular mechanisms by which the placenta senses and responds to environmental cues during pregnancy are poorly understood. This review discusses the role of the insulin‐like growth factors (IGFs) in controlling placental resource allocation to intrauterine growth, particularly in relation to maternal environmental conditions during pregnancy. It focuses primarily on small laboratory animals, such as the mice, rats and guinea pigs that are most commonly used for these studies, but also draws on data from other species, including humans, where available.

The insulin‐like growth factors

The insulin‐like growth factors (IGFs), IGF1 and IGF2, are 7.5 kDa single‐chained polypeptides that promote growth, both before and after birth. They affect the metabolism, mitogenesis, survival and differentiation of a wide variety of cell types by binding to IGF receptors (IGF1R and IGF2R), insulin receptor (INSR) and a hybrid IGF1R–INSR receptor with varying affinity (Sferruzzi‐Perri et al. 2008; Fernandez & Torres‐Aleman, 2012; Harris & Westwood, 2012). Their actions are influenced by at least six different IGF binding proteins (IGFBP‐1 to IGFBP‐6) and numerous IGF‐related binding proteins, which alter access of the IGFs to their receptors and have been reviewed in detail elsewhere (Bach et al. 2005; Bach, 2015; Clemmons, 2016). The main signalling receptor for the IGFs is IGF1R, which activates the phosphoinositide‐3 kinase (PI3K)–protein kinase B (AKT) and mitogen‐activated protein kinase (MAPK) signalling pathways. IGF2 also binds to the IGF2R, which can lead to either IGF2 degradation or activation of the G‐protein‐coupled signalling pathway (Okamoto et al. 1990).

The Igf2 gene is subject to parental imprinting and only the paternal allele is expressed. It can be expressed by different promoters, of which P0 (Igf2P0) is specific to the placenta in mice (Moore et al. 1997). In mice, though largely not in humans, the Igf2r gene is also imprinted but in a reciprocal fashion to Igf2 with expression from the maternal allele (Monk et al. 2006). The IGFs (particularly IGF2), their receptors and signalling pathways are expressed by the placenta in many species and change in their abundance both developmentally and in response to environmental cues (Sferruzzi‐Perri et al. 2010). In many species, circulating IGF concentrations are higher during pregnancy than in the non‐pregnant animal and also change in the mother and fetus with proximity to delivery (Fowden, 2003; Sferruzzi‐Perri et al. 2010). IGF2 is more abundant than IGF1 in both the maternal and fetal circulations in all species studied to date (Fowden, 2003; Sferruzzi‐Perri et al. 2010). IGF2 is also more highly expressed than IGF1 by the placenta in all species studied to date (Sferruzzi‐Perri et al. 2010).

The effects of the insulin‐like growth factors on placental phenotype

The effects of the IGFs on the placenta have been studied directly in two main ways. First, they have been given exogenously either to placental cultures in vitro or to pregnant animals in vivo to study placental growth, transport and endocrine function. Secondly, the Igf genes, their receptors and key molecules in their downstream pathways have been under‐ or over‐expressed in genetically modified mice to determine the morphological and functional consequences for the placenta at different stages of pregnancy. While the functions of the placenta are common across species, its structure varies in terms of shape, organisation of trophoblast lineages, extent of invasion into the maternal uterus, and degree of interdigitation at the feto‐materno interface (reviewed in depth elsewhere; Carter, 2007; Wooding & Burton, 2008; Roberts et al. 2016). For instance, the human and non‐human primate placenta is composed of a series of highly branched structures, called villi. These contain a mesenchymal core that has fetal capillaries that are closely associated with an overlying syncytiotrophoblast layer. The syncytiotrophoblast is directly bathed in maternal blood and functions in both transport and hormone secretion. Cytotrophoblast cells, can fuse to form the syncytiotrophoblast or migrate from the villous tree into the decidua, where they invade and remodel uterine spiral arteries to promote blood flow to the placenta. The syncytiotrophoblast is also bathed in maternal blood in the mouse, rat and guinea pig placenta. However, the mouse placenta is arranged into two morphologically and functionally distinct regions: the labyrinth zone (Lz), which is responsible primarily for transport, and the junctional zone (Jz; also known as basal or interlobium region), which functions in uterine remodelling/invasion and hormone secretion. In ruminate species like the sheep and cow, the placenta is composed of individual placentomes that form at specialised sites called caruncles in the uterine wall. The overlaying trophoblast layer can be a syncytium (in sheep) or remains unicellular (columnar epithelium in cows) and there no invasion of the maternal blood vessels by trophoblast cells. However, in sheep some trophoblast cells migrate and fuse with caruncle epithelial cells and play an endocrine role.

Exogenous administration of IGFs

In vitro experiments

IGF1 and IGF2 prevent apoptosis and enhance proliferation and migration/invasion of human placental villous explants, primary trophoblast cultures and trophoblast cell lines from the first trimester and term (Table 1). IGF1 also promotes the proliferation, invasion and survival of first trimester human placental fibroblasts (Miller et al. 2005) and the differentiation of term trophoblast cells into syncytiotrophoblast (Bhaumick et al. 1992; Milio et al. 1994; Cohran et al. 1996). Similarly, IGF1 stimulates proliferation and migration of murine ectoplacental cone trophoblast in culture (Kanai‐Azuma et al. 1993) and early pregnancy porcine trophoblast cells (Jeong et al. 2014). Furthermore, IGF2 promotes differentiation of murine ectoplacental cone trophoblast and migration of ovine trophoblast cells in vitro (Kim et al. 2008). Using receptor and pathway inhibitors and IGF analogues with selectivity for particular receptors, some of the molecular mechanisms mediating the actions of IGFs on the human placenta have begun to be identified in vitro. IGFs appear to mediate their proliferative and anti‐apoptotic effects on trophoblast through activating IGF1R and triggering the MAPK and PI3K–AKT signalling pathways, respectively (Forbes et al. 2008). IGFs also induce trophoblast migration and invasion through IGF1R, and possibly INSR, with subsequent activation of MAPK and PI3K–AKT signalling pathways (Diaz et al. 2007; Shields et al. 2007; Forbes et al. 2008; Mayama et al. 2013). However, IGF2 may also signal via IGF2R and G_i proteins, MAPK and Rho GTPase pathways to trigger trophoblast migration and invasion (McKinnon et al. 2001; Shields et al. 2007; Harris et al. 2011). Thus IGFs promote the growth of different cell lineages in the placenta via multiple mechanisms (Fig. 1 A).

Table 1.

The impact of exogenous IGF1 or IGF2 on the placental phenotype and fetal outcome (where available)

IGF	System	Species	Treatment	Study	Placental size and morphology	Placental function	Fetal weight	References
IGF1	In vitro	Mouse	Primary ectoplacental cone trophoblast	First trimester	↑ Proliferation and migration			Kanai‐Azuma et al. (1993)
		Pig	Primary trophoblast cells	First trimester	↑ Proliferation and migration			Jeong et al. (2014)
		Human	First trimester primary trophoblast	First trimester	↑ Invasion via INSR and IGF1R activation of Akt			Mayama et al. (2013)
		Human	First trimester placental explant	First trimester	↑ Proliferation and syncytial formation via IGF1R‐mediated MAPK signalling,↓ apoptosis via IGF1R‐mediated PI3K signalling			Forbes et al. (2008, 2015)
		Human	First trimester placental trophoblast	First trimester	↑ Proliferation, migration			Hashimoto et al. (2010)
		Human	First trimester placental explant	First trimester	↑ Proliferation			Forbes et al. (2009, 2015)
		Human	First trimester placental explant	First trimester	↑ Migration			Lacey et al. (2002)
		Human	First trimester placental explant	First trimester	↑ Proliferation	↑ hCG, hPL		Maruo et al. (1995)
		Human	First trimester trophoblast	First trimester		↑ System A amino acid and glucose uptake		Kniss et al. (1994)
		Human	First trimester primary placental fibroblasts	First trimester	↑ Proliferation, invasion, ↓ apoptosis			Miller et al. (2005) *Ad‐IGF‐I
		Human	BeWo syncytial cell line		↑ Proliferation, invasion, ↓ apoptosis	↑ System A and System L amino acid transporter activity, Snat1, Snat2, Lat1, 4F2hc, GLUT1, GLUT3 and GLUT8, ↓ Lat2		Jones et al. (2013, 2014) *Ad‐IGF‐I
		Human	BeWo			↔ pGH		Zeck et al. (2008)
		Human	JEG‐3 choriocarcinoma cell line		↑ Proliferation, ↓ apoptosis	↑ P4, hCG secretion		Rak‐Mardyla & Gregoraszczuk, (2010)
		Human	JEG‐3		↑ Invasion via induction of adhesion and migration through IGF1R–PI3K and MAPK signalling			Diaz et al. (2007)
IGF1	In vitro	Human	BeWo			↑ System A amino acid transporter activity via PI3K signalling, ↔ Snat1 or Snat2		Fang et al. (2006)
		Human	BeWo, term explants and term perfused human placenta			↑ Glucose transport, GLUT1 membrane abundance		Baumann et al. (2014)
		human	Term human placenta	Term		↓ LPL activity in		Magnusson‐Olsson et al. (2006)
		Human	Term trophoblast	Term		↑ System A amino acid uptake		Bloxam et al. (1994); Karl (1995); Yu et al. (1998)
		Human	Term trophoblast and cell lines	Term	↑ Syncytialisation			Bhaumick et al. (1992); Milio et al. (1994); Cohran et al. (1996)
IGF2	In vitro	Mouse	Primary ectoplacental cone trophoblast	First trimester	↑ Differentiation into endocrine cells			Kanai‐Azuma et al. (1993)
		Sheep	Primary trophoblast	First trimester	↑ Migration			Kim et al. (2008)
		Human	First trimester HTR8_SVneo cell line	First trimester	↑ Migration via Rho GTPases			Qiu et al. (2005); Shields et al. (2007)
		Human	First trimester HTR8_SVneo cell line	First trimester	↑ Migration via signalling through IGF2R involving inhibitory G proteins and the MAPK pathway			McKinnon et al. (2001)
		Human	First trimester primary trophoblast	First trimester	↑ Migration/invasion			Irving & Lala (1995); Hamilton et al. (1998)
		Human	First trimester placental explant	First trimester	↑ Trophoblast proliferation and syncytial formation via IGF1R‐mediated MAPK signalling, ↓ apoptosis via IGF1R‐mediated PI3K signalling			Forbes et al. (2008, 2009, 2015)
		Human	JEG‐3 choriocarcinoma cell line		↑ Invasion via induction of adhesion and migration through INSR–PI3K and MAPK signalling			Diaz et al. (2007)
IGF2	In vitro	Human	SGHPL4 and first trimester villous explants	First trimester	↑ Proliferation, migration and invasion			Pollheimer et al. (2011)
		Human	First trimester primary placental fibroblasts	First trimester	↑ Proliferation and invasion, ↓ apoptosis			Miller et al. (2005) *Ad‐IGF‐II
		Human	First trimester and term trophoblast	First trimester	↓ Apoptosis, ↑ proliferation and survival against TNF‐α and IFN‐γ‐induced apoptosis			Hills et al. (2012)
		Human	First trimester placental trophoblast	First trimester		↑ Glucose uptake		Kniss et al. (1994)
		Human	First trimester placental trophoblast	First trimester		↑ Glucose and System A amino acid uptake		Kniss et al. (1994)
		Human	First trimester placental explant	First trimester	↑ Migration			Lacey et al. (2002)
		Human	In BeWo and term explants		↑ Proliferation, ↓ apoptosis and necrosis			Harris et al. (2011)
IGF1	In vivo	Mouse	D14	D17	↔ Weight, ↑ placental cross‐sectional area, Lz and fetal and maternal facing areas		↔ Weight or viability	Katz et al. (2009) *Ad‐IGF‐I
		Mouse uterine artery ligation	D16	D20	↔ Weight, ↑ placental thickness		↑ 27%, ↔ fetal viability	Abd Ellah et al. (2015) *nanoparticle targeted delivery to placenta: PLAC1‐IGF‐1
		Mouse uterine artery ligation	D18	D20	ND	↑ 4F2hc, Lat1, Lat2, GLUT8, GLUT9a/b, ↔ Snat1, Snat2, GLUT1	ND	Jones et al. (2013, 2014) *Ad‐IGF‐I
		Guinea pig	D20–37	D40	↔ Weight		↑ 6%, ↓ litter size	Sohlstrom et al. (2001)
IGF1	In vivo	Guinea pig	D20–38	D35	↑ 17% weight, ↓ placental and Lz area, ↔ Lz, Jz, FC, MBS, Troph Vd	↑ Glucose and System A amino acid transfer, Snat2 and prorenin activation, ↓ Igf2, ↔ Glut1, Igf1	↑ 15%	Sferruzzi‐Perri et al. (2007b); Standen et al. (2015)
		Guinea pig	D20–38	D62	↔ Weight, ↔ structure	↑ Glucose and System A amino acid transfer	↑ 17%, ↑ fetal viability	Sferruzzi‐Perri et al. (2006, 2007a)
		Guinea pig 30% UN	D20–37	D40	↑ 13% weight		↔	Sohlstrom et al. (2001)
		Rabbit Natural runt	D19	D21	↔ Weight		↑ 19%	Keswani et al. (2015) *Ad‐IGF‐I
		Sheep	D128, 4 h infusion	D128	ND	↑ Glucose transfer and lactate production, ↔ blood flow, urea or glucose transfer	ND	Liu et al. (1994)
		Sheep *Fetal infusion	D121–132	D132	↔ Weight, ↓ placentome number	↓ Glucose and System A amino acid transfer	↔	Bloomfield et al. (2002b)
		Sheep *Fetal infusion	D128, 4 h infusion	D128	ND	↓ Glucose transfer, lactate uptake and umbilical flow, ↔ urea transfer or serine uptake	ND	Harding et al. (1994); Jensen et al. (1999, 2000)
		Sheep Embolised *Fetal infusion	D128, 4 h infusion	D128	ND	↔ Glucose or urea transfer, lactate uptake and umbilical flow	ND	Jensen et al. (1999)
		Sheep Spontaneous growth restriction *Fetal infusion	D128, 4 h infusion	D128	ND	↔ Glucose or urea transfer, lactate uptake and umbilical flow	ND	Jensen et al. (1999)
		Sheep Embolised *Intra‐amniotic infusion	D110, D117, D124	D120–131	↔ But placentas no longer significantly different to untreated controls	↔ Glucose uptake, ↑ Glut1, Glut4, Systems Y+ and L transporters (Slc7a1 and Slc7a8), ↔ Glut3, Snat4, Slc7a5	↔ Weight, ↑ fetal growth rate and fetuses no longer significantly different to untreated controls	Eremia et al. (2007); Wali et al. (2012)
IGF2	In vivo	Mouse	D14, D16, D18 IGF2 (1 mg kg−1 day−1) or iRGD‐liposome with IGF2 (0.3 mg kg−1 day−1)	D18	↑ Weight		↔	King et al. (2016)
		Mouse IGF2P0	D14, D16, D18 treatment with iRGD‐liposome with IGF2 (0.3 mg kg−1 day−1)	D18	↔ Weight of Igf2P0 and WT		↑ Igf2P0 but not WT	King et al. (2016)
		Rat	D16–22	D22	↔ Weight, ↑ Jz		↔	Van Mieghem et al. (2009)
		Guinea pig	D20–37	D40	↑ 9% weight		↑ 7%	Sohlstrom et al. (2001)
		Guinea pig	D20–38	D35	↔ Weight and structure	↔ Glucose or System A amino acid transfer, Glut1, Snat2, Igf1 and Igf2,↑ prorenin activation	↔	Sferruzzi‐Perri et al. (2007b); Standen et al. (2015)
		Guinea pig	D20–38	D62	↔ Weight, ↑ Lz area, Vd, Vol, SA, ↓ Jz Vd, ↔ BT	↑ Glucose transfer, ↔ System A amino acid transfer	↑ 11% weight and ↑ fetal viability	Sferruzzi‐Perri et al. (2006, 2007a)
		Guinea pig 30%UN	D20–37	D40	↔ Weight		↔	Sohlstrom et al. (2001)
Leu27‐IGF2	In vivo	Mouse	D13–19	D19	↔ Weight	↔ System A amino acid transfer, ↓ litter System A amino acid variability	↔ Weight, ↓ variability in fetal weight	Charnock et al. (2016)
		Mouse eNOS–/–	D13–19	D19	↔ Weight		↑	Charnock et al. (2016)
		Guinea pig	D20–38	D62	↔ Weight, ↑ Lz vd, Troph, MBS Vd and Vol and SA, ↓ Jz area, Vd, Vol, FC Vd, Vol and BT	↑ Glucose and System A transfer and prorenin activation	↑ 11%	Sferruzzi‐Perri et al. (2008)

^* Highlights details about the administration of IGF. For in vivo studies, exogenous IGF was administered to the mother, unless stated otherwise. Genes are written in lower case and proteins are written in capital. Abbreviations: Ad, Adenoviral‐mediated; BT, barrier thickness; D, day; FC, fetal capillaries; GLUT, glucose transporter; hCG, human chorionic gonadotrophin; hPL, human placental lactogen; IGF1/Igf1, insulin‐like growth factor‐1; IGF2/Igf2, insulin‐like growth factor‐2; Jz, junctional zone; LAT, L‐type amino acid transporter; Lz, labyrinthine zone; MAPK, mitogen activated protein kinase; MBS, maternal blood space; ND, not determined; P4, progesterone; PI3K, phosphoinositol 3‐kinase; pGH, placental growth hormone; Prl; prolactin‐related hormone; SA, surface area; SNAT/Snat, sodium‐coupled amino acid transporter; UN, undernutrition; vol, volume; vd, volume density. Search terms used to find studies listed in the table: trophoblast, placenta, fetus, insulin‐like growth factor, IGF and/or transport.

Figure 1. Impact of exogenous IGFs on the placenta.

A, the effect of exogenous IGFs on placental human trophoblast in vitro. Proposed signalling pathways mediating the actions of IGFs are shown. B, the effect of exogenous maternal IGFs on the mouse, rat and/or guinea pig placenta in vivo. Dashed lines indicate a potential interaction (A) or impact (B) of IGF1. IGF, insulin‐like growth factor; IGF1R, type 1 IGF receptor; IGF2R, type 2 IGF receptor; INSR, insulin receptor; Jz, junctional zone; Lz, labyrinthine zone; MAPK, mitogen‐activated protein kinase; PI3K, phosphoinositol 3‐kinase.

In addition to stimulating placental growth, both IGFs stimulate glucose and System A amino acid uptake, and IGF1 increases System L activity but reduces lipoprotein lipase activity in human trophoblast in vitro (Table 1). However, these changes in nutrient uptake do not always track with the expression of the transporter genes or proteins, suggesting that the IGFs may also affect post‐transcriptional/translational mechanisms (Fang et al. 2006; Jones et al. 2013, 2014). Indeed, IGF1 was recently shown to stimulate glucose transporter capacity by increasing the translocation of glucose transporter 1 (GLUT1/SLC2A1) to the trophoblast plasma membrane (Baumann et al. 2014). In culture, IGF1 prevents the release of the vasoconstrictors prostaglandin E and F and thromboxane by the term human placenta and reduces the agonist‐mediated vasoconstriction of human myometrial arteries (Siler‐Khodr et al. 1995; Corcoran et al. 2012). In vivo, these effects could increase utero‐placental blood flow and substrate transfer in late gestation. Both IGF1 and IGF2 also enhance trophoblast endocrine capacity in culture. IGFs increase the secretion of hormones including progesterone, human chorionic gonadotrophin and placental lactogen in vitro although others, like placental growth hormone may not be affected (Maruo et al. 1995; Zeck et al. 2008; Rak‐Mardyla & Gregoraszczuk, 2010). In addition, IGF2 simulates the differentiation of hormone‐producing murine and ovine trophoblast in vitro (Kanai‐Azuma et al. 1993; Kim et al. 2008). Thus, IGFs have the capacity to promote growth, hormone secretion and substrate transport capacity of the placenta.

In vivo experiments

Treatment of guinea pig dams with either IGF1 or IGF2 in early–mid pregnancy increases fetal weight near term (Table 1; Sferruzzi‐Perri et al. 2006). With exogenous IGF1, placental Lz area and Igf2 gene expression is reduced during the treatment, even though fetal weight is increased already in mid pregnancy (Sohlstrom et al. 2001; Sferruzzi‐Perri et al. 2007b; Standen et al. 2015). Whilst there is no sustained effect of either IGF on placental weight, IGF2 increases the volume and surface area of the transport Lz, near term (Table 1; Sferruzzi‐Perri et al. 2006). Development of the placental exchange region was further enhanced when the IGF2R‐selective synthetic analogue Leu²⁷‐IGF2 was administered maternally (Sferruzzi‐Perri et al. 2008). In mice, maternal Leu²⁷‐IGF2 treatment from day 13 of pregnancy halves the number of fetuses naturally growth restricted within the litter near term (Charnock et al. 2016). Taken together, these findings suggest that maternal IGF2 in early gestation may act, in part, via the IGF2R to enhance functional development of the placenta with beneficial impacts on fetal growth. However, caution is warranted as part of the effects of Leu²⁷‐IGF2 could be due to the displacement of endogenous IGF2 and its subsequent interaction with IGF1R and INSR in the placenta.

Exogenous IGFs also modify the functional capacity of the placenta to supply resources for fetal growth. In the late pregnant ewe, increasing IGF1 in the fetal circulation increases amino acid and glucose uptake by the placenta but may reduce materno‐fetal transfer of these substrates, lactate production and the number of placentomes (Table 1). Increasing IGF1 in the maternal circulation also alters placental metabolic function in the pregnant ewe near term; glucose transfer capacity and lactate production are enhanced by an acute infusion of IGF1 (Liu et al. 1994). In guinea pigs, placental delivery of glucose and/or neutral amino acids to the fetus is increased in late gestation by chronic maternal IGF treatment in early–mid pregnancy (Table 1). This enhanced placental transfer in late gestation is partly due to increased expression of nutrient transporters (System A amino acid; SNAT2/Snat2/Slc38a2) by IGF1 in mid pregnancy and improved development of the exchange region by IGF2 in late pregnancy (Sferruzzi‐Perri et al. 2006, 2007b). In mice, the variability in System A amino acid transport capacity and conceptus weight within the litter is abolished by maternal Leu²⁷‐IGF2 (Charnock et al. 2016) and data suggest that IGFs may have most benefit for improving growth of the smallest pups. Indeed, maternal Leu²⁷‐IGF2 improves the weight of fetuses that are growth restricted due to a lack of the endothelial nitric oxide gene and reduces the number of pups below the fifth centile of the wild‐type population in late gestation (Charnock et al. 2016). In addition to improving placental transport function, exogenous IGFs also affect endocrine capacity in vivo (Fig. 1 B). Maternal IGF2 treatment simulates the development of the endocrine Jz of the rat placenta (Van Mieghem et al. 2009) and exogenous IGF1 and IGF2 increase placental pro‐renin activation in guinea pigs (Standen et al. 2015). Thus, IGFs may also increase fetal resource supply through changing placental endocrine function and thus maternal adaptations to pregnancy, but further studies are warranted.

To circumvent possible confounding effects of systemic IGF treatment on the mother, approaches are being developed to target IGFs to the placenta. In mice, adenoviral‐mediated site‐specific intraplacental transfer of the Igf1 gene on day 14 of pregnancy increases the area of the placenta and the size of the Lz and of the maternal‐ and fetal‐facing areas 3 days later, although there is no change in conceptus weight (Table 1; Katz et al. 2009). In response to liposome‐mediated targeting of IGF2 to the mouse placenta, placental growth is also increased although fetal weight is not affected (King et al. 2016). In rabbits, the weight of natural runt fetuses in the litter is increased 2 days following placental Igf1 transgene delivery without a change in placental weight, but how it impacts on structure and function of the placenta remains unknown (Keswani et al. 2015). These data suggest that targeting of IGF delivery to the placenta may prove an effective method of improving placental function and thus fetal growth, particularly when feto‐placental growth is impaired.

Genetic manipulation of the IGF system

In mice, knockout of the Igf2 gene in the entire conceptus or within the fetal or trophoblast cell lineages leads to placental and fetal growth restriction, with the greatest reduction in growth seen with ubiquitous Igf2 loss (Table 2). Similarly, a heterozygous deficiency in the PI3K‐p110α (Pik3ca; homozygous deficiency is lethal) or complete ablation of the AKT1 (Pkba) or MAPK1 (Erk2) genes causes feto‐placental growth restriction (Cho et al. 2001; Hatano et al. 2003; Yang et al. 2003; Yung et al. 2008; Kent et al. 2012; Sferruzzi‐Perri et al. 2016). In contrast, over‐expressing the Igf2 gene through activating the normally silent maternal gene copy in the H19 null, increasing IGF2 availability via Igf2r ablation, or deletion of the PI3K signalling inhibitor (Pten) results in over‐growth of the fetus and placenta (Leighton et al. 1995; Ludwig et al. 1996; Louvi et al. 1997; Ripoche et al. 1997; Church et al. 2012). Deletion of the Igf1, Igf1r or Insr genes in mice also leads to fetal growth restriction, but placental weight is unaffected (DeChiara et al. 1990; Baker et al. 1993; Louvi et al. 1997). This suggests that the growth‐promoting effect of IGF2 in the mouse placenta occurs independently of IGF1R and INSR, possibly through an unknown, distinct placenta‐specific receptor (XRp) (Louvi et al. 1997). However, evidence from H19 null mutants suggests that IGF1R could contribute to the control of placental growth in mice as the first exon of the H19 gene encodes miR‐675, which targets Igf1r for reduced expression (Keniry et al. 2012). Overgrowth of the H19 null placenta (Leighton et al. 1995; Esquiliano et al. 2009; Angiolini et al. 2011; Church et al. 2012) is thus thought to be due to biallelic Igf2 via imprinting mechanisms, as well as enhanced Igf1r expression through loss of miR‐675 (Keniry et al. 2012). Taken together, these data highlight the importance and complexity of the IGF system in controlling conceptus growth in mice.

Table 2.

The effect of genetically manipulating IGF abundance and/or signalling on feto‐placental growth in mice

		Placenta
Manipulation	Approach	Size	Morphology	Function	Fetal weight	Reference
Deficiency of IGF and downstream signalling
Global IGF1 KO	Igf1 –/–	D18/19 ↔			D18/19 ↓ 40%	Baker et al. (1993)
Global IGF2 KO	Paternal Igf2 –	D15 ↓ 47% D18/19 ↓ 20–30%	D15 ↓ Jz GlyT D18/19 ↔ Lz and Jz Vd ↓ Jz GlyT	D17 ↓ EAAT1, EAAT2 (Jz), EAAT3 (Jz), EAAT4,↑ CAT1, ↔ 4f2hc D18/19 ND	D15 ND D18/19 ↓ 40% ↑ fetal loss	DeChiara et al. (1990, 1991); Baker et al. (1993); Liu et al. (1993); Lopez et al. (1996); Matthews et al. (1999); Esquiliano et al. (2009); Church et al. (2012); Kent et al. (2012)
Global IGF2 KO	Paternal transmission LacZDMR2 –	D16 ↓ 27% D19 ↓ 40%	D16 ND D19 ↓ Lz vd and volume of all Lz components, SA, FC length and diffusing capacity, ↑ Jz vd and BT	D16 ↔ System A and glucose transport, Snat1, Snat2, Snat4 D19 ↓ System A transfer and passive permeability and Snat2, ↔ glucose transport, Snat1 and Snat4	D16 ↓ 24% D19 ↓ 52%	Constancia et al. (2005); Coan et al. (2008)
Fetal specific IGF2 KO	Inner cell mass Igf2 –	D17 ↓ 14%			D17 ↓ 27%	Gardner et al. (1999)
Placental trophoblast specific IGF2 KO	Trophechoderm Igf2 –	D17 ↓ 21%			D17 ↓ 12%	Gardner et al. (1999)
Placental Lz specific IGF2 KO	Paternal transmission Igf2P0 –	D16 ↓ 20% D17 ↓ 24% D19 ↓ 35%	D16 ↔ Lz or Jz Vd ↓ Lz Trophoblast, GlyT D19 ↓ SA, trophoblast, FC volume, FC length, diffusing capacity, ↑ BT ↔ Lz or Jz vd and umbilical artery flow	D16 ↑ System A and glucose transport, Snat4, Glut3, ↓ passive permeability, calbindin, ↔ Snat1, Snat2, Glut1 D19 ↑/↔ System A transport, ↑ glucose and calcium transport, ↓ passive permeability, ↔ Snat1, Snat2, Snat4, calcium transport, calbindin, PMCA1, TRPV6	D16 ↔/↓ 4% D17 ↓ 24% D19 ↓ 24%	Constancia et al. (2002, 2005); Sibley et al. (2004); Coan et al. (2008); Dilworth et al. (2010, 2013); Kusinski et al. (2011); Sferruzzi‐Perri et al. (2011)
Global IGF1R KO	Igf1r –/–	D19 ↔	D18/19 ↔ Jz GlyT	D17 ↓ EAAT2 (Jz), EAAT3 (Lz and Jz), ↑ CAT1, ↔ EAAT1, EAAT4	D19 ↓ 55%	DeChiara et al. (1990); Louvi et al. (1997); Matthews et al. (1999); Esquiliano et al. (2009)
Global INSR KO	Insr –/–	D15 ↔ D18/19 ↔	D15 ↔ Jz GlyT D18/19 ↔ Jz GlyT		D15 ND D18/19 ↓ 10%	Louvi et al. (1997); Esquiliano et al. (2009)
PI3K p110α (Pik3ca)	Kinase dead heterozygote Pik3ca‐D933A	D16 ↓ 9% D19 ↓ 12%	D16 ↓ Lz vol, FC vol, FC length, MBS vol, SA, diffusing capacity, ↑ BT, ↔ Jz D19 ↓ Lz vol, FC vol, FC length, Troph vol, SA, diffusing capacity, ↑ BT, ↔ Jz	D16 ↑ glucose and System A transfer per unit SA, ↔ Glut1, Glut3, Snat1, Snat2, Snat4 D19 ↑ glucose and System A transfer per unit SA, ↑ Prl3b1, ↔ Glut1, Glut3, Snat1, Snat2, Snat4	D16 ↓19% D19 ↓11%	Sferruzzi‐Perri et al. (2016)
Global decreased AKT signalling through increased PTEN	Prl2 –/–	D17 ↓ 22%	D17 ↓ Jz, GlyT and Lz	D17 ↓ passive transport	D17 ↓ 17%	Dong et al. (2012)
Global decreased AKT1 signalling	Pkba –/– (exons 4–8 deleted)	D17 ↓ 33% D19 ↓ 45%	D17 ↓ thickness, GlyT, Lz vessel density, length, area	D17 ↓ pAKT D19 ↓ total AKT, pAKT, ↑ Akt2 and Akt3	D17 ↓ 17%	Yang et al. (2003; Yung et al. (2008)
Global decreased AKT1 signalling	Pkba –/– (exon 1 deleted)	D18 ↓ 30%	D18 ↔ Lz and Jz Vd	D18 ↓ pAKT, ↔ pAKT	D18 ↓ 22% weight and ↑ fetal loss	Cho et al. (2001); Kent et al. (2012)
Global decreased MAPK signalling	Erk2 –/–	D11 ↓	D11 ↓ Lz thickness, FC development	↓ MAPK signalling	D11 ↓ weight and ↑ fetal loss	Hatano et al. (2003)
Over‐expression of IGF and downstream signalling
Global IGF2 over‐expression	Maternal Igf2r –	D16 ↑ 40% D18 ↑ 25%			D16 ↑ 40% D18 ↑ 40%	Ludwig et al. (1996); Louvi et al. (1997)
Global IGF2 over‐expression*	Maternal H19Δ13–	D15 ↑ 37% D16 ↑ 30% D18 ↑ 60% D19 ↑ 45%	D15 ↑ Jz GlyT D16 ↑ volume of all placental components, ↑ SA, diffusing capacity, ↔BT D18 ↑ Jz GlyT D19 ↑ volume of all placental components, SA, diffusing capacity	D15 ↑ AKT1 ↔ pAKT, pERK1/2 D16 ↓ glucose transfer, passive permeability and Glut3, ↔ Glut1, Snat1, Snat2, Snat4 D19 ↓ glucose and System A transfer, passive permeability and Snat4, ↔ Glut1, Glut3, Snat1, Snat2	D15 ↑ 30% D16 ↑ 12% D18 ↑ 20% D19 ↑ 23%	Leighton et al. (1995); Esquiliano et al. (2009); Angiolini et al. (2011); Church et al. (2012)
Global increased IGF2 and signalling via AKT	Double KO of maternal H19 and Pten +/–	D16 ↑ 65% D19 ↑ 80%	D16 ↑ Jz, GlyT D19 ↑ Jz, GlyT	↑ pAKT and IGF2	D16 ↑ 31% D19 ↑ 31%	Church et al. (2012)
Global increased pAKT	Pten +/–	D16 ↑ 22% D19 ↑ 22%	D16 ↑ Jz, GlyT D19 ↑ Jz, GlyT	↑ pAKT, ↔ IGF2	D16 ↑ 19% D19 ↑ 7%	Church et al. (2012)

^* H19 null has biallelic expression of Igf2 combined with absence miR675 (encoded by H19). Genes are written in lower case and proteins are written in capital. Abbreviations: AKT, protein kinase B; BT, barrier thickness; D, day; ERK, extracellular signal‐regulated kinase; FC, fetal capillaries; GLUT/Slc2a, glucose transporter; GlyT, trophoblast glycogen cells; IGF1/Igf1, insulin‐like growth factor‐1; IGF2/Igf2, insulin‐like growth factor‐2; Jz, junctional zone; LAT, L‐type amino acid transporter; Lz, labyrinthine zone; MAPK, mitogen activated protein kinase; MBS, maternal blood space; ND, not determined; p, phosphorylated; PI3K, phosphoinositol 3‐kinase; SA, surface area; SNAT/Snat, sodium‐coupled amino acid transporter; Vd, volume density. Search terms used to find studies listed in the table: placenta, fetus, insulin‐like growth factor, IGF, PI3K, ERK, MAPK, knock out, deficiency and/or transgenic.

Genetic manipulations of Igf2, Igf2r and the downstream signalling pathways also affect the morphology of the placenta (Table 2). For instance, loss of Igf2 (complete and Igf2P0 null), Pik3ca, Pkba or Erk2 gene expression causes defective Lz formation. In particular, Lz volume/thickness, exchange surface area and vascularisation are all reduced and the interhaemal barrier to diffusion of gases such as oxygen is greater in the placenta of all these mutants (Table 2). In contrast, in the H19 null, the Lz surface area is increased in line with the placentomegally observed (Angiolini et al. 2011). In addition, IGF2 affects the formation of endocrine cells in the placenta. In particular, loss or gain of Igf2 or the PI3K–AKT signalling pathway causes a disproportionate decrease or expansion of the glycogen cells in the Jz, whereas Igf1r or Insr nulls show no changes in Jz glycogen cell abundance (Table 2). Collectively, the available data suggest that IGF2 acts via both the PI3K–AKT and MAPK pathways to attain normal placental weight and Lz structure, and through PI3K–AKT signalling to drive placental glycogen cell formation in mice (Fig 2 A).

Figure 2. The effect of genetically manipulating IGF2 expression or signalling on placental phenotype in mice.

A and B, the effect of complete loss of IGF2 (A) and the effect of partial loss of IGF2 (B), either by deleting the placental‐exclusive isoform, Igf2P0 or through a constitutive heterozygous deficiency of PI3K‐p110α. Dashed line indicates a potential interaction of IGF2 with receptor. Line with a round head indicates parameters reduced by loss of IGF2 signalling. Loss of IGF2 signalling leads to reductions in placental development and transport function (A). Partial loss of IGF2 signalling also leads to reductions in placental development, but is associated with adaptive up‐regulation in transport function (B). AA, amino acids; IGF, insulin‐like growth factor; IGF1R, type 1 IGF receptor; Lz, labyrinthine zone; GLUT, glucose transporter; MAPK, mitogen‐activated protein kinase; PI3K, phosphoinositol 3‐kinase; SNAT, sodium‐coupled amino acid transporter; XRp, unknown placental‐specific IGF receptor.

Placental function also changes when the IGF system is genetically modified in mice (Table 2). The passive permeability of the placenta to hydrophilic nutrients/solutes is reduced in the complete Igf2 null, placental‐specific Igf2P0 null and H19 null (Constancia et al. 2002; Sibley et al. 2004; Coan et al. 2008; Angiolini et al. 2011). The complete Igf2 null placenta transports less neutral amino acid (methyl aminoisobutyric acid, MeAIB) via the System A transporters in association with reduced SNAT2/Snat2/Slc38a2 expression (Constancia et al. 2005). There is also reduced abundance of System X_AG ^– and System Y⁺ transporters, responsible for placental transfer of cationic and anionic amino acids, in the complete Igf2 and the Igf1r null (Matthews et al. 1999). In contrast, the Igf2P0 null placenta transports more neutral amino acids via System A, as well as more glucose and calcium in late gestation (Table 2). Up‐regulation of placental transport capacity is associated with increased expression of SNAT4/Snat4/Slc38a4 and GLUT3/Glut3/Slc2a3 by the Igf2P0 deficient placenta. In contrast to Igf2, there is little or no information on the capacity of the Igf1 or Insr null placenta to supply nutrients to the fetus. In the complete Igf2 null, placental and fetal growth restriction occurs concurrently and becomes evident in mid‐gestation (Table 2; Baker et al. 1993; Constancia et al. 2005). In the Igf2P0 null, placental weight is reduced at a similar time in gestation, but fetal growth only becomes restricted much closer to term and to a lesser extent than in the complete Igf2 null (Baker et al. 1993; Constancia et al. 2002, 2005). Liposome‐mediated targeting of IGF2 to the placenta has recently been shown to increase the weight of Igf2P0 null mouse fetuses near term (Table 1; King et al. 2016). Collectively, these findings suggest that the Igf2P0 null placenta compensates for its defective development and compromised permeability by adaptively up‐regulating its nutrient transport systems and thereby minimises the degree of fetal growth restriction, relative to the complete Igf2 null. The Pik3ca heterozygote deficient placenta also transfers glucose and amino acids via System A transporters with increased efficiency in compensation for its impaired development, which is associated a less severe reduction in fetal weight close to term than earlier in gestation (Sferruzzi‐Perri et al. 2016). Moreover, the naturally small placenta that supports more fetal mass per gram shows increased expression of Igf2P0 coupled with a preservation of Lz growth and with increased placental System A transport capacity and SNAT2/Snat2/Slc38a2 abundance compared to the large placenta in the litter (Coan et al. 2008). In contrast, the over‐grown H19 null placenta shows diminished neutral amino acid and glucose transport, which is thought to limit fetal over‐growth and avoid an excessive drain of maternal resources into the fetus (Angiolini et al. 2011). Thus, IGF2 in the placenta is important for fine‐tuning nutrient supply to the fetus (Fig. 2 B).

In addition to effects on placental transport, the Igf2 gene may also affect the endocrine function of the placenta with consequences for maternal physiology during pregnancy. Evidence for this stems from associations between altered placental Jz formation in H19 and Igf2P0 null mutants and raised circulating glucose, insulin and/or corticosterone in phenotypically wild‐type dams (Petry et al. 2010; Sferruzzi‐Perri et al. 2011). Thus, IGF2 has an important role in nutrient allocation to the fetus. By regulating placental phenotype, it balances the fetal genetic drive for growth with the maternal ability to supply the required resources, thereby optimising both offspring and maternal fitness.

IGFs as environmental signals in regulating placental resource allocation to fetal growth

IGFs may also play an important role in changing placental resource allocation to the fetus in environmentally challenged pregnancies. As Igf1 expression is relatively low in the placenta, studies have largely focused on placental expression of Igf2 and activation of its signalling pathways (Table 3). However, since the signalling pathways are responsive to both IGFs, the placenta can also respond to changes in circulating IGF1 and IGF2 induced by nutritional or other environmental cues.

Table 3.

The effect of maternal environmental challenge on fetal growth and placental structure, function and IGF signalling

			Placenta
Maternal manipulation	Species	Timing	IGF and signalling	Size	Morphology	Function	Fetal weight	Reference
Nutrient restriction
20% UN	Mouse	D3 to 19	D16 ↑ IGF1R ↓ Igf2P0 and PI3K signalling D19 ↓ Igf2P0 and PI3K signalling	D16 ↓ 6% D19 ↓ 9%	D16 ↔ Lz but ↓ Jz and GlyT D19 ↓ Lz (MBS and FC vols and SA), ↔ BT	D16 ↓ Glut1 D19 ↑ System A amino acid transport, ↑ Glut1, Snat2, ↓ Snat4	D16 ↔ D19 ↓ 13%	Coan et al. (2010); Sferruzzi‐Perri et al. (2011)
10–30% UN	Guinea pig	–D28 to D60	D35/40 ↓ Igf2, ↔ Igf1	D35 ↓ 20% D60 ↓ 30%	D35 ↓ Jz volume ↔ Lz, but ↓ MBS, SA and ↑ BT D60 ↓ Lz volume, MBS, FC, SA, ↑ BT, ↔ Jz		D35 ↓ 29% D60 ↓ 35%	Roberts et al. (2001); Olausson & Sohlstrom (2003)
30% UN	Sheep	D22 to 135	D135 ↔ Igf2	D135 ↓ 19%, altered placentome distribution			D135 ↓ 12%	Osgerby et al. (2002, 2004)
50% UN	Sheep	–D60 to D30	D78 ↑ insulin–IGF signalling (pAKT and pERK1/2)	D78 ↓ 29%	D78 ↑ vascularity		D78 ↔	Zhu et al. (2007b)
50% UN	Sheep	D28 to 78	D78 ↑ insulin–IGF signalling (pARK1/2, ↔ pAKT) ↔ mTORC1 signalling D135 ↔	D78 ↓ 21% D135 ↔		D78 ↑ Glut3, GLUT1, Fatp4 D135 ↑ Fatp4	D78 ↓ 26% D135 ↔	Ma et al. (2011)
UN gradual decrease to full food withdrawal	Sheep	D83 to 90	D90 ↔ Igf2 D135 ↓Igf2	D90 ↓ 22% D135 ↔		D90 ↔ Glut1, Glut3 D135 ↔ Glut1, Glut3	D90 ↔ D135 ↔	McMullen et al. (2005)
UN 50%	Cow	D30 to 125	D125 ↑ insulin–IGF signalling (pAKT and pERK1/2) D250 ↔	D125 ↓ 27% D250 ↓ 20%	D125 ↑ vascularity D250 ↔ vascularity		D125 ↔ D250 ↔	Zhu et al. (2007a)
70% UN	Baboons	D30 to 165	D90 ↓ Igf2, IGF2R, ↑ IGF1R, ↔ Igf1 or IGF1 D120 ND D165 ↓ insulin/IGF‐I, MAPK (IRS‐1, AKT S6K, ERK‐1) and mTOR signalling	D90 ↔ D120 ↔ D165 ↓ 20%		D90 ND D120 ↓ System A amino acid transport, ↔ system L amino acid transport, GLUT1, TAUT, SNAT1, SNAT2, SNAT4, LAT1, LAT2 D165 ↓ System A and L amino acid transport, GLUT1, TAUT, SNAT2, LAT1, LAT 2	D90 ↔ D120 ↔ D165 ↓ 19%	Li et al. (2007); Kavitha et al. (2014); Pantham et al. (2015)
Low protein diets
16% vs. 20% protein (0.80CT)	Mouse	D3 to 19	D16 ↔ Igf2, H19 D19 ↔ Igf2, H19	D16 ↑ 5% D19 ↑ 5%	D16 ↓ Lz/Jz ratio D19 ↓ Lz/Jz ratio	D16 ↑ glucose transport, Glut1, ↔ System A amino acid transport D19 ↓ System A amino acid transport, Snat4, ↔ glucose transport,	D16 ↔ D19 ↔	Coan et al. (2011)
8% vs. 20% protein (0.40CT)	Mouse	D3 to 19	D16 ↑ total Igf2 ↔ Igf2P0, H19 D19 ↔ Igf2, H19	D16 ↔ D19 ↑4%	D16 ↔ D19 ↔	D16 ↑ glucose transport, Snat2, ↔ System A amino acid transport D19 ↓ Snat1, Snat4, ↔ glucose and System A amino acid transport	D16 ↔ D19 ↓ 9%	Coan et al. (2011)
9% vs. 17% protein (0.53CT)	Rat	D1 to 22	D22 ↑ Igf1 ↓ Igf2 ↔ Igf1r, Igf2r, Insr	ND			D22 ↓ 8%	Nusken et al. (2011)
6% vs. 20% protein (0.30CT)	Rat	D1 to 21	D14 ↓ Lz Igf2, Insr in female and ↑ Lz IGF2, ↓ Igf1r in male D16 ↓ Lz Igf2 in female and male D21 ↓ Lz IGF2 in male and female	D14 ↓ 25% D18 ↓ 12% D21 ↔	D14 ↓ Lz and Jz vol D18 ↓ Lz vol, ↑ trophoblast stem cells and Lz sinuosoidal GiT, ↓ spongiotrophoblast and GiT cells, ↔ Jz D21 ↔ Lz ↓ Jz		D14 ↓ 21.5 D18 ↓ 27 D21 ↓ 14%	Gao et al. (2012,b, 2013)
4% vs. 18% protein (0.22CT)	Rat	D2 to 21	D19 and D21 ↓ mTOR D21 ↓ PI3K signalling (pAKT‐T308)	D15–19 ↔ D21 ↓ 12.5%	ND	D19 and D21 ↓ Systems A and L amino acid transport, LAT1, LAT2, SNAT2, ↔ glucose transport, SNAT4	D15–19 ↔ D21 ↓ 21%	Jansson et al. (2006); Rosario (2011); Pantham et al. (2016)
Obesogenic diets
2.5 times fat	Mouse	–D28 to D1	D13 ↓ Igf2, Mtor, ↔ Igf1 D18 ↓ Igf2, Igf2r, ↔ Igf1	D13 ↓ 20% D18 ↑15% in males, ↔ females	↔ Lz	D13 ↓ Snat1, Glut1, ↔ Cd36 D18 ↓ Cd36, ↔ Snat1, Glut1	D13 ↓ 28% D18 ↓ 15%	Sasson et al. (2015)
2.5 times fat	Mouse	–D28 to D18	D13 ↑ Igf1r, ↓ Igf2, Igf2r, Mtor, ↔ Igf1 D18 ↓ Igf2, Igf2r, ↔ Igf1	D13 ↓ 20% in males D18 ↔ males or females	↔ Lz	D13 ↓ Snat1, Glut1, ↔ Cd36 D18 ↓ Cd36, ↔ Snat1, Glut1	D13 ↓ 25% D18 ↓ 25%	Sasson et al. (2015)
2.5 times fat	Mouse	D1 to 18	D13 ↑ Igf1r, ↓ Igf2, Igf2r, Mtor, ↔ Igf1 D18 ↓ Igf2, Igf2r ↔ Igf1	D13 ↓ 20% in males D18 ↑ 15% in males	↔ Lz	D13 ↓ Snat1, Glut1, Cd36 D18 ↓ Glut1, Cd36 ↔ Snat1	D13 ↓ 28% D18 ↓ 28%	Sasson et al. (2015)
5.3 times fat	Mouse	–D84 to D19	D15 ↑ Igf2 and Igf2r male, ↔ female D19 ↔ Igf2 and Igf2r	D15 ↔ D19 ↔		D15 ↑ Lz Snat2 in male, ↑ Lz Snat4 in female D19 ↔	D15 ↔ D19 ↓ 8% in males	King et al. (2013)
6 times fat	Mouse	D1 to 15	D15 ↑ Igf1, ↓ Irs1 in males, ↔ Igf2, Igf2P0, Igf2r, H19	D15 ↑ 7%	D15 ↔ Lz or vascularity	D15 ↓ Slc22a1, ↑ Slc22a2 *sexually dimorphic response of placenta	D15 ↔	Gallou‐Kabani et al. (2010); Gabory et al. (2012)
2.5 times fat	Rat	D1 to 21	D21 ↔ mTORC1 signalling	D21 ↔	D21 ↓ Jz		D21 ↓ 5%	Mark et al. (2011)
5–6 times fat	Rat	–D49 to D21	D21 ↑ mTORC1 signalling, ↔ Insulin‐IGF signalling (pAKT or pMAPK)	D21 ↔		D21 ↓ SNAT1, ↔ Systems A and L amino acid transport and LPL activity, SNAT2, SNAT4, GLUT1, GLUT3, GLUT9, FATP4, FATP6, LPL	D21 ↑ 7%	Gaccioli et al. (2013)
3 times fat and 5 times sugar diet	Mouse	–D42 to D18	D18 ↓ mTORC1 signalling, ↔ Insulin‐IGF PI3K (pAKT, IRS1, PI3K‐p85)	D18 ↔			D18 ↔	Lager et al. (2014)
4 times fat and 1.3 times sugar	Mouse	–D20 to D19	D19 ↑ Insulin/IGF‐PI3K (p‐IRS1, pAKT‐T308) and mTORC1 signalling, ↔ MAPK	D19 ↔		D19 ↑ Systems A and L amino acid transport, SNAT2, LAT1, GLUT1, GLUT3, FATP6, ↔ SNAT4, LAT2, CD98, FAT/CD36, FATP2, FATP4	D19 ↑ 18%	Diaz et al. (2015); Rosario et al. (2015, 2016)
3 times fat and 5 times sugar diet	Mouse	D1 to 19	D16 ↑ Igf2, IgfP0, H19, Insulin/IGF–PI3K signalling (PI3K‐p110α, pAKT), ↓ INSR, ↔ mTORC1 or MAPK D19 ↑ Insulin/IGF–PI3K signalling (PI3K‐p110α, pAKT, pMAPK), ↔ Igf2, Igf2P0, H19, INSR or mTORC1	D16 ↓ 11% D19 ↓ 8%	D16 ↓ Lz FC ↑ BT D19 ↓ Lz, MBS, BT, SA and ↓ GlyT	D16 ↑ glucose and System A amino acid transport, Glut3, Snat2 D19 ↑ FATP1, ↔ glucose and System A amino acid transport	D16 ↓ 9% D19 ↔	Sferruzzi‐Perri et al. (2013)
50% greater food intake	Sheep	–D60 to D135	D70–75 ↓ pIRS1, mTORC1 signalling, pMAPK in the arterial tissues, ↔ INSR, IGF1R D165 ND	D70–75 ↓22% D165 ↔	D70–75 ↑ arteriole diameters, ↓ vessel density D165 ↔	D70–75 ↑ Fatp1, Fatp4, Cd36, Lpl D165 ↑ GLUT3, FATP1, Fatp4, Cd36, ↔ Lpl	D70–75 ↑ 20–26% D165 ↔	Zhu et al. (2009, 2010); Ma et al. (2010); Tuersunjiang et al. (2013)
Hypoxia
13%	Mouse	D1 to 19	D19 ↑ Insulin‐IGF (↑ pAKT) and mTORC1 signalling	D19 ↑10%	D19 ↑ Maternal arterial and venous blood space	ND	D19 ↓ 12% weight and litter size	Matheson et al. (2015);
13% hypoxia	Mouse	D11 to 16	D16 ↓ Igf2, ↔ Igf2P0, altered pAKT (depending on site phosphorylated)	D16 ↔	D16 ↑ Lz ↑ MBS, trophoblast vol, SA exchange	D16 ↔ System A amino acid amino acid or glucose transport, Gluts and Snats	D16 ↔	Higgins et al. (2015)
13% hypoxia	Mouse	D14 to 19	D19 ↑ Igf2, Igf2P0, altered insulin‐IGF signalling (↓ INSR, IGF1R, PI3K‐p85α, PI3K‐p110α but ↑ pAKT)	D19 ↔	D19 ↑ FC volume and density, ↓ BT	D19 ↑ glucose transport, Snat1, ↔ System A amino acid amino acid transport	D19 ↓ 5%	Higgins et al. (2015)
12% hypoxia	Mouse	D14.5 to 18.5	D18.5 ↓ Igf2r and Igf2, Igf1r in females	D18.5 ↔	D18.5 ↓ Lz blood space, ↑ tissue in females	D18.5 ↓ Glut1, ↑ Snat1 in females, ↔ Glut3	D18.5 ↓ 6.5%	Cuffe et al. (2014)
10% hypoxia	Mouse	D14 to 19	D19 ↓ Insulin‐IGF signalling (↓ INSR, IGF1R, PI3K‐p85α, PI3K‐p110α and pAKT), ↔ Igf2, Igf2P0	D19 ↔	D19 ↓ Lz vd, MBS volume, SA exchange, ↑ Jz vd, trophoblast vol and BT	D19 ↓ System A amino acid transport, ↔ glucose transport but altered uterine artery vasoreactivity	D19 ↓ 21%	Higgins et al. (2015); Skeffington et al. (2015)
Endocrine disruption
Corticosterone 83 μg g–1 day–1	Mouse	D11 to 16	D16 ↓ pAKT, ↔ Igf2, Igf2P0, INSR, IGF1R, mTORC1 signalling	D16 ↓ 6%	D16 ↓ FC vol and Vd, ↑ MBS and Troph Vd, ↔ SA, BT	D16 ↓ Glut1, Glut3, Snat1, Snat2, ↔ glucose or System A amino acid transport and Snat4	D16 ↓ 7%	Vaughan et al. (2012, 2015)
Corticosterone 81 μg g–1 day–1	Mouse	D11 to 19	D19 ↓ mTORC1 signalling, ↔ Igf2, Igf2P0, INSR, IGF1R, pAKT	D19 ↓ 12%	D19 ↔ FC, MBS, Troph, SA, BT	D19 ↓ glucose and System A amino acid transport, ↑ Snat1, ↔ Glut1, Glut3, Snat2, Snat4	D19 ↓ 19%	Vaughan et al. (2012, 2015)
Dexamethasone 24 μg g–1 day–1	Mouse	D13 to 16	D16 ↓ MAPK1 D18 ↔ MAPK1 D16 and D18 ↔ Igf2	D16 ↓ 20% female only D18 ↔	D16 ↓ Jz area female only D18 ND	D16 and D18 ↔ Glut1, Glut3, Snat1, Snat2, Snat4	D16 ↓ 20% D18 ↔	Cuffe et al. (2011)
Dexamethasone 24 μg g–1 day–1	Rat	D13 to 20	D20 ↓ pAKT in Jz	D20 ↓ 50%		D20 ↓ Prls in Jz, ↑ Prls in Lz	D20 ↓ 22%	Ain et al. (2005)
Diabetes via streptozotocin administration neonatally	Rat		D20 ↑ Igf1, Igf2, Igf2r, IGF1R kinase and autophosphorylation activity, ↔ Igf1r, Insr	D21 ↑ 22%	D21	D21 ↑ glycerol and FFA release	D21 ↑ 13%	Hauguel‐de Mouzon et al. (1992); Martinez et al. (2008); White et al. (2015)
Diabetes via streptozotocin administration 1 week before mating	Rat	–D7 to D21	D21 ↓ Insr, Irs1, Igf2, Igf2r, ↔ Irs2, Igf1r	D21 ↑ 22%	D21 ↑ Lz, ↑ lacunae	D21 ↓ Glut1, ↑ Lpl, ↔ Glut3, Snat2, Snat4, Lat1	D21 ↑ 5% or ↔	Cisse et al. (2013)
Insulin resistance via heterozygous p110α deficiency	Mouse		D16 ↓ PI3K signalling D19 ↓ PI3K signalling	D16 ↔ D19 ↑15%	D16 ↓ Lz Troph vol, ↓ BT D19 ↑ Jz vol ↑ SA diffusing capacity	D16 ↓ glucose transfer, Snat1, ↔ System A amino acid transfer, Glut1, Glut3, Snat2, Snat4 D19 ↓ glucose transfer, Glut1, Snat1, Snat2, Prls, ↔ System A amino acid transfer, Glut3, Snat4	D16 ↔ D19 ↔	Sferruzzi‐Perri et al. (2016) *Depends on fetal genotype
Other manipulations affecting conceptus growth
Restriction of utero‐placental blood flow
Uterine ligation	Mouse	D18	D20 ↓ Igf1, Igf2	D20 ↔	D20 ↓ Lz depth, vol, vessel area	D20 ↓ Slc5a9, Slc7a10, 4F2hc, Lat1, Lat2, Snat2, GLUT1, GLUT8, ↑ Snat1, ↔ GLUT3, GLUT9	D20 ↓ 11%	Habli et al. (2013); Jones et al. (2013, 2014)
Uterine ligation	Rat	D17	D20 ↓ Igf2	D20 ↓ 8%			D20 ↓ 20%	Price et al. (1992);
Uterine ligation	Rat	D18 or D19	D20 ↓ IGF1R, ↔ INSR	D20 ↔ or ↓ 25%	D20 ↑ diameter, ↔ Lz vd	D20 ↓ GLUT1, ↔ GLUT3	D20 ↓ 7% or 27% weight and ↓ litter size	Das et al. (1998); Reid et al. (2002); Wlodek et al. (2005)
Uterine ligation	Rat	D19	D22 ↓ Igf1, ↔ Igf2, Insr, Igf1r, Igf2r	ND			D22 ↔	Nusken et al. (2011)
Uterine ligation	Guinea pig	D30	D55–60 ↔ Igf1, Igf2	D55–60 ↔ or 37%		D55–60 ↓ System A amino acid transfer, ↔ glucose transfer	D55–60 ↓ 7% or 38%	Jansson & Persson, (1990); Carter et al. (2005)
Placental embolism	Sheep	D113 to 120	D131 ↓ IGF1R, ↔ IGF‐I	D131 ↓ 30%			D131 ↓ 21%	Bloomfield et al. (2002a); Shaikh et al. (2005)
Placental embolism	Sheep	D103 to 109	D131 ↔ Mtor	D131 ↓ 43%		D131 ↓ Glut1, Slc7a1, Slc7a8, ↔ Glut3, Glut4, Snat4, Slc7a5	D131 ↓ 20%	Wali et al. (2012)
Uterine carunclectomy	Sheep	–D70	D130–134 ↑ Igf2, ↔ Igf1, Igf1r, Igf2r	D130–134 ↓ 30–40%	D130–134 altered distribution of placentome types and ↓ placentome number but ↑ individual weight of placentomes, trophoblast and maternal capillary volume and SA of placentomes	D130–134 ↓ Fatp4, ↔ Glut1, Glut3, Glut4, Slc7a1, Slc7a5, Snat1, Snat4, Fatp1, Cd36, Fabp5	D130–134 ↓ 26%	Zhang et al. (2016b)
Hyperthermia	Sheep	D39 to 125	D55 ↑ IGF2 D90 ↑ IGF1, mTORC! signalling, ↓ pAKT, ↔ MAPK D135 ↑ pAKT, pMAPK dys‐regulated mTORC1 signalling (↑ mTORC1 but ↓ p70 phosphorylation)	D55 ↔ D90 ↓ 24% D135 ↓ 58%		D135 ↑ Slc7a5, Slc7a8, uterine blood flow, trans‐placental oxygen diffusion, ↓ branched amino acid and glucose transport, ↔ utero‐placental oxygen uptake	D55 ↔ D90 ↔ D135 ↓ 47%	Thureen et al. (1992); Ross et al. (1996); Anderson et al. (1997); Regnault et al. (2003, 2005, 2007); de Vrijer et al. (2004, 2006); Arroyo et al. (2009, 2010)
Alcohol consumption	Rat	–D4 to D4	D20 ↓ Lz Igf1, and Lz Igf1r in males, ↔ Igf2 ↑ Jz Igf2, ↔ Jz Igf1, Lz or Jz Igf2r	D20 ↔	D20 ↑ length and width, ↓ Lz and ↑ Jz and ↑ GlyT in females	D20 ↓ Lz Snat2, ↔ Lz Snat1, Snat4, Glut1, Glut3 and ↓ Jz Glut1 in males, ↑ Jz Glut1 in females	D20 ↓ 7%	Gardebjer et al. (2014)

^* Highlights additional information on the study presented in the Table. Gestational age: mouse ∼20 days, rats ∼23 days, guinea pigs ∼70 days, sheep ∼150 days, cows ∼283 days, baboons ∼183 days. Genes are written in lower case and proteins are written in capital. Abbreviations: BT, barrier thickness; D, day; ERK, extracellular signal‐regulated kinase; FATP, fatty acid transport protein; FC, fetal capillaries; GLUT, glucose transporter; GiT, giant trophoblast cells; GlyT, trophoblast glycogen cells; IGF1/Igf1, insulin‐like growth factor‐1; IGF2/Igf2, insulin‐like growth factor‐2; Jz, junctional zone; LAT, L‐type amino acid transporter; LPL, lipoprotein lipase; Lz, labyrinthine zone; MAPK, mitogen activated protein kinase; MBS, maternal blood space; mTOR, mechanistic target of rapamycin; p, phosphorylation; PI3K, phosphoinositol 3‐kinase; Prl, prolactin‐related hormone; SA, surface area; SNAT/Snat, sodium‐coupled amino acid transporter; UN, undernutrition; vol, volume; vd, volume density. Search terms used to find studies listed in the table: placenta, fetus, insulin‐like growth factor, IGF, nutrient restriction, undernutrition, low protein diet, high sugar, high fat, obesogenic, IUGR, PI, hypoxia, uterine ligation, corticosterone, dexamethasone carunclectomy, heat stress and/or diabetes.

Maternal nutrition

Undernutrition

In mice, guinea pigs and baboons, undernutrition restricts placental growth in association with a decrease in the expression of Igf2 and/or signalling via the PI3K–AKT and MAPK pathways (Table 3). There are also reductions in placental vascularisation, exchange surface area, Jz volume and glycogen cell abundance and/or a greater barrier to diffusion with maternal undernutrition in mice and guinea pigs, morphological parameters that were altered similarly by a genetic deficiency in Igf2, Pik3ca, Pkba and Erk2 (Tables 2 and 3). Together, these studies suggest that decreases in IGF2 expression and signalling within the placenta could underlie the growth and morphological defects observed with maternal undernutrition in these species. In larger animals, the expression of Igf2 and its signalling machinery reduces, is unchanged or even increases in response to undernutrition (Table 3). For instance, signalling via MAPK and PI3K–AKT in the placenta is up‐regulated in nutrient‐restricted ewes and heifers (Zhu et al. 2007a,b; Ma et al. 2011). In these models, changes in signalling relate to a normalisation of placental weight or an increase in placental cotyledon vascularity. They also correlate with a maintenance or restoration of fetal weight in later gestation, despite an exposure to undernutrition. In ewes of a moderate condition, which have the smallest placentas supporting more mass of fetus per gram, placental expression of Igf2 is greatest (Osgerby et al. 2003). These studies therefore suggest that in larger species, there is morphological adaptation of the placental to an adverse maternal nutritional state through increasing Igf2 and growth signalling locally.

In the undernourished sheep placenta with increased PI3K–AKT and MAPK signalling, the expression of glucose and fatty acid transporters is also increased (Ma et al. 2011). However, in undernourished baboons, diminished Igf2 expression and signalling in the placenta accompanies reductions in Systems A and L amino acid transporter capacity and glucose transporter gene expression (Pantham et al. 2015, 2016). Taken together, these studies suggest that IGF2 and the PI3K–AKT and MAPK signalling pathways could also mediate changes in placental transport function during undernutrition. In mice, despite a 20% reduction in maternal food intake and placental growth restriction earlier in gestation, fetal weight is normal until just prior to term (Coan et al. 2010). This maintenance of fetal growth relates to an initial preservation of Lz development in earlier gestation and an adaptive up‐regulation of System A amino acid transporter capacity and SNAT2/Snat2/Slc38a2 expression near term, by the growth restricted undernourished placenta. However, in mice lacking the placental‐specific Igf2 isoform (Igf2P0), these adaptations to maternal undernutrition fail to occur. The development of the placental exchange region is compromised earlier in gestation, there is no up‐regulation of amino acid transport or SNAT2/Snat2/Slc38a2 expression and reduced SNAT4/Snat4/Slc38a4 abundance near term in Igf2P0 null placentas compared to wild‐type in undernourished mice (Sferruzzi‐Perri et al. 2011). As a result, fetal growth is restricted earlier in gestation and more adversely affected near term by undernutrition, in Igf2P0 nulls. The Igf2P0 transcript is, therefore, a major determinant of the environmental modification of placental phenotype with undernutrition in mice. The expression of genes involved in glucose, neutral amino acid and fatty acid transport, as well as the IGF signalling pathways in the human placenta are modified by the diet and physical activity of the mother during pregnancy (Brett et al. 2015). Thus, the IGF system may also be important for modifying resource capacity of the human placenta in response to changes in the maternal environment.

Low‐protein diets

During rodent pregnancy, consumption of an iso‐calorific low protein diet has inconsistent impacts on both placental weight and placental Igf2 expression (Table 3; Sferruzzi‐Perri & Camm, 2016). However, the nature of the specific effect appears to depend on the degree of protein deprivation, stage of pregnancy studied and sex of the conceptus (Jansson et al. 2006; Coan et al. 2011; Nusken et al. 2011; Gao et al. 2012). Despite the contrasting results, placental Igf2 expression seems to track positively with the weight of the placenta in mice and rats (Coan et al. 2011; Nusken et al. 2011; Gao et al. 2012). For instance, in pregnant mice, low protein diets cause placentomegaly and the degree of placental weight increase relates to the level of Igf2 up‐regulation at first appearance of growth enhancement (Coan et al. 2011). The variation in placental growth and Igf2 expression observed in different models of protein deficiency could be caused by the content and source of carbohydrate used to maintain calorie intake. Nevertheless, taken together, these findings suggest that at least part of the changes in placental growth seen with protein deprivation could be mediated through local changes in Igf2.

There are also changes in placental transport capacity with gestational protein malnutrition. For instance, in response to a diet with 8% protein, the mouse placenta adaptively transports more glucose to the fetus on day 16 of pregnancy (Coan et al. 2011). This up‐regulation occurs when placental Igf2 expression is also increased and when fetal growth is maintained despite maternal protein deprivation (Coan et al. 2011). A few days later, however, glucose transport is unchanged, System A amino acid transporter abundance is reduced and Igf2 expression no longer increased in the placenta by a low protein diet, and fetal growth restriction ensues (Coan et al. 2011). These data suggest that placental Igf2 may be important for adapting nutrient supply to the fetus in response to maternal protein malnutrition in mice. However, there is evidence that pathways downstream of Igf2 may also be important. For instance, the mechanistic target of rapamycin (mTORC1) mediates the mitogenic and metabolic actions of IGFs (Jansson et al. 2012b). In rats, protein deprivation reduces mTORC1 signalling, Systems A and L amino acid transport, and SNAT2/Snat2/Slc38a2, and L‐type amino acid transporter (LAT1/lat1/Slc7a5 and LAT2/Lat2/Slc7a8) expression by the placenta, prior to the appearance of placental and fetal growth restriction (Jansson et al. 2006; Rosario et al. 2011). These findings suggest that down‐regulation of signalling pathways like mTORC1 and amino acid transporters in the placenta could link maternal protein restriction to decreases in fetal growth. The availability of protein and specific amino acids during pre‐implantation rodent development is linked to alterations in the expression of genes within the H19‐Igf2 locus, mTORC1 signalling and trophoblast cell formation and differentiation with consequences for feto‐placental phenotype in late gestation (Kwong et al. 2006; Van Winkle et al. 2006; Eckert et al. 2012; Watkins et al. 2015). Thus, changes in Igf2 expression and its signalling pathways could be responsive to the availability of nutrients from the earliest stages of development.

Diets with excess sugar and/or fat

The expression of Igf2 and its signalling pathways in the placenta are inconsistently altered by diets with excess sugar and/or fat (Table 3). Weight of the conceptus may also be reduced, increased or unchanged, depending on the level of fat in the diet, the amount of simple sugars consumed and the timing of the dietary manipulation (Table 3 and reviewed in Sferruzzi‐Perri & Camm, 2016). Part of these variations in Igf2 expression and conceptus growth could be due to the differences in protein and micronutrient intake, as species like mice and rats control their calorie intake tightly (Keesey & Hirvonen, 1997). In mice fed a diet containing 2.5 times the fat of the controls, placental weight is reduced in early pregnancy in association with decreases in the expression of Igf2 and signalling machinery, including Mtor (Sasson et al. 2015). These placental changes accompanied reductions in the expression of System A amino acid transporter, SNAT1/Snat1/Slc38a1, glucose transporter, GLUT1/Glut1/Slc2a1, and/or fatty acid translocase, CD36, depending on the length of high fat feeding and whether the diet was eaten before pregnancy (Sasson et al. 2015). In over‐nourished ewes, placental weight is reduced in mid‐gestation in association with decreased activity of the IGF signalling pathway (including activation of IRS1 and mTORC1) and changes in vessel size and density in the placenta (Zhu et al. 2009; Ma et al. 2010). However, fetal weight is increased along with fatty acid transporters and translocases in the placenta, suggesting that alternative signalling pathways may be activated to adapt placental nutrient supply to the fetus in ewes with excess food intake (Zhu et al. 2010; Tuersunjiang et al. 2013).

In other studies, increases in the placental IGF system are coupled with improved placental resource allocation to the fetus in dams fed obesogenic diets (King et al. 2013; Sferruzzi‐Perri et al. 2013; Diaz et al. 2015; Rosario et al. 2015, 2016). For instance, in mice, consumption of a high sugar and fat diet from day 1 of pregnancy initially causes conceptus growth restriction and morphological defects in the placental Lz. However, fetal weight normalises by term, despite the persistence of placental growth and morphological defects through adaptive up‐regulation of glucose and neutral amino acid transport to the fetus by the placenta (Sferruzzi‐Perri et al. 2013). Up‐regulation of transport capacity relates to increased expression of GLUT3/Glut3/Slc2a3 and SNAT2/Snat2/Slc38a2, as well as elevated expression of the placental‐specific Igf2 isoform and PI3K–AKT signalling in the placenta in dams fed a diet with excess sugar and fat. Obesogenic diets fed from before pregnancy also increase placental nutrient transporter capacity (glucose, Systems A and L amino acid and fatty acids) in line with greater Igf2 or PI3K–AKT and mTORC1 signalling; however, responses varied with the precise composition of the diet and, possibly, fetal sex (King et al. 2013; Aye et al. 2015; Diaz et al. 2015; Rosario et al. 2015, 2016). The expression of IGF signalling machinery (receptors, AKT, mTORC1) and nutrient transporters is also altered in the placenta from obese women; however, the specific nature of these changes appears to depend on the level of maternal body fat mass, gestational weight gain and whether macrosomia is observed (Jansson et al. 2012a; Brett et al. 2016; Martino et al. 2016). Taken together, these findings suggest that obesity and obesogenic diets alter placental phenotype in association with changes in placental Igf2 system and fetal growth.

Maternal hypoxia

In mice, hypoxia typically reduces fetal growth in a severity‐dependent manner without a change in placental weight (Table 3 and reviewed in Sferruzzi‐Perri & Camm, 2016). However, if the hypoxic challenge commences early in pregnancy, placentomegaly is observed in associated with greater maternal blood spaces and activation of the PI3K–AKT and mTORC1 signalling pathways in the placenta (Matheson et al. 2015). Even though placental weight may not be altered when maternal hypoxia commenced later in pregnancy, placental expression of the IGF system and capacity to supply resources to the fetus is altered (Table 3). In particular, placental expression of IGF receptors, INSR and PI3K isoforms is decreased in response to 5 days of 13–10% maternal hypoxia in late mouse gestation, and in 10% hypoxia this effect is due to reductions in maternal food intake (Cuffe et al. 2014; Higgins et al. 2015). However, expression of Igf2, Igf2P0 and activated AKT increases with 13% hypoxia, but is unchanged or even decreased in response to 5 days of 12–10% hypoxia near term (Cuffe et al. 2014; Higgins et al. 2015). In the 13% hypoxic mouse placenta showing increases in Igf2 expression and signalling, there are beneficial changes in Lz structure including improved vascularisation, maternal blood spaces and a thinner diffusion barrier to exchange, changes that would optimise oxygen delivery to the fetus near term (Higgins et al. 2015; Matheson et al. 2015). There is also greater placental glucose uptake and transport and maintained delivery of neutral amino acids to the fetus when 13% hypoxia occurs in the last third of pregnancy (Higgins et al. 2015). In contrast, in the 12–10% hypoxic placenta with unchanged or decreased expression of the IGF2 system, the morphology of the placental Lz is compromised, with reductions in maternal blood spaces and surface area and a greater barrier to diffusion, changes that would further limit fetal oxygen supply in hypoxic dams (Cuffe et al. 2014; Higgins et al. 2015). Moreover, placental glucose uptake and transport capacity are not up‐regulated or even reduced (less GLUT1/Glut1/Slc2a1) and delivery of neutral amino acids diminished, in dams exposed to 12–10% hypoxia, depending on whether food intake is reduced and the sex of the fetus (Cuffe et al. 2014; Higgins et al. 2015). In culture, 1% hypoxia reduces the outgrowth of mouse ectoplacental cone trophoblast in association with diminished Igf2 expression (Pringle et al. 2007). Hypoxia (1% oxygen) also diminishes the expression of PI3K–AKT and mTORC1 signalling in human trophoblast cell lines (Yung et al. 2012a) and modulates IGF1 and IGF2 signalling in early pregnancy placental mesenchymal stem cells (Youssef et al. 2014; Youssef & Han, 2016). Placental expression of the PI3K–AKT and mTORC1 signalling pathways and GLUT1/Glut1/Slc2a1 expression are decreased in women at 3100 m above sea level who deliver growth‐restricted babies (Zamudio et al. 2006; Yung et al. 2012a). In addition, inducing endoplasmic stress in the mouse placental Jz genetically is associated with defects in PI3K–AKT and mTORC1 signalling, altered IGF2 glycosylation and bioactivity, and with feto‐placental growth restriction (Yung et al. 2012b). Taken together, these findings suggest that activating IGF2 and/or PI3K–AKT signalling in the placenta may be critical for adapting placental resource allocation to the fetus during hypoxia in late pregnancy. They also suggest that the placenta may integrate signals of oxygen and nutrient availability through the IGF2 system to adapt its phenotype and optimise maternal resource supply to fetal growth. Indeed, the mouse Igf2 gene harbours a hypoxia‐responsive element in its promoter (Feldser et al. 1999), as well as CHORE motifs, which bind the glucose‐responsive transcription factor, MLX (Hunt et al. 2015). Therefore, the availability of oxygen and nutrients in utero could have direct effects on placental Igf2. Nutritional and hypoxic challenges alter the concentration of hormones like the glucocorticoid stress hormone and insulin, in the maternal circulation (Sferruzzi‐Perri et al. 2011; Cuffe et al. 2014). Thus, changes in placental phenotype may reflect alterations in the metabolic and endocrine state of the mother.

Maternal endocrine challenges

Endocrine challenges can affect maternal metabolism and utilisation of nutrients and thus the partitioning of resource to the conceptus in pregnancy (Vaughan et al. 2011). Administering corticosterone or the synthetic glucocorticoid dexamethasone to rodents for 3–7 days reduces fetal and placental weights during gestation (Table 3; Vaughan et al. 2011). In mice, corticosterone decreases AKT and mTORC1 activation in association with reductions in feto‐placental weight, Lz vascularisation, and glucose and System A amino acid transporter capacity; however, the specific nature of these effects depends on when in pregnancy the over‐exposure occurs (Table 3). Administering the synthetic glucocorticoid dexamethasone reduces the expression of MAPK and weight of placenta in female, but not male, conceptuses and there is no change in glucose and SNAT/Snat/Slc38a amino acid transporters irrespective of fetal sex (Cuffe et al. 2011). In mice, placental Igf2 expression is unaffected by maternal administration of corticosterone and dexamethasone even though the conceptus may be growth restricted (Cuffe et al. 2011; Vaughan et al. 2012, 2015). Whereas, restrain stress increases placental Igf2 but does not alter offspring weight in mice (Pankevich et al. 2009). In rats, dexamethasone decreases placental Igf2 and the level of activated AKT, particularly in the endocrine Jz (Ain et al. 2005). In dexamethasone‐treated rats, there are reductions in the expression of prolactin‐related family genes by the Jz in late gestation, which may influence the maternal adaptations to pregnancy and, thus, alter the fetal supply of nutrients indirectly (Ain et al. 2005). The expression of Igf2 by the term human placenta is also altered in women with elevated plasma cortisol during pregnancy due to emotional distress (Mina et al. 2015). Glucocorticoid response elements have been identified in the human Igf1 gene promoter (He et al. 2016); however, very little is known about whether glucocorticoids could have direct effects on placental Igf2 expression. Collectively these findings suggest that reductions in the placental IGF2 system and the functional phenotype of the placenta could link elevated maternal glucocorticoids to decreases in fetal growth.

In rats, pre‐existing maternal diabetes also alters the expression of the IGF system in the placenta, as well as materno‐fetal resource allocation; however, the direction of change depends on how long the dam was insulin deficient/dysglycaemic. For instance, Igf expression and IGF1R activation are elevated in association with greater glycerol and free fatty acid transfer by the placenta and an increase in fetal weight by 13% in rats that are diabetic from neonatal life (White et al. 2015). Placental lipid transport capacity is also increased in rat dams that are diabetic for 1 week prior to pregnancy (increase in placental lipoprotein lipase); however, the expression of Igf2 and the IGF signalling machinery is decreased and the expression of GLUT1/Glut1/Slc2a1 reduced in association with a more minor increase in fetal weight (by 5%) (Cisse et al. 2013). Genetically inducing maternal insulin insensitivity by a global heterozygous deficiency in PI3K‐p110α signalling capacity in the mouse dam is associated with improved placental Lz development (larger surface area and thinner barrier to diffusion), but reduced glucose transport and expression of nutrient (GLUT1/Glut1/Slc2a1, SNAT1/Snat1/Slc38a1, SNAT2/Snat2/Slc38a2) and prolactin‐related family genes near term (Sferruzzi‐Perri et al. 2016). However, the specific nature of these placental changes depended on whether the conceptus itself was heterozygous for the PI3K‐p110α deficiency (Sferruzzi‐Perri et al. 2016). Moreover, there is no effect of maternal heterozygous deficiency in PI3K‐p110α signalling on fetal weight in this model, irrespective of fetal genotype (Sferruzzi‐Perri et al. 2016). Taken together, these studies suggest that the IGF2–PI3K‐p110α system plays an important role in modulating fetal nutrition and growth in response to maternal insulin deficiency and/or insensitivity, by acting at the level of placental transport phenotype.

Other environmental challenges affecting conceptus growth

The expression of Igfs, receptors and signalling machinery in the placenta also changes in response to insults that affect the placental capacity to supply the fetus with nutrients. Such insults include reduced utero‐placental blood flow, heat stress and alcohol consumption (Table 3). Reducing both maternal oxygen and nutrient supply to the conceptus using uterine artery ligation in mice, rats and guinea pigs, or placental embolism in sheep, reduces placental expression of components of the IGF system in association with defects in placental Lz structure and in transporter expression and activity of the glucose and Systems A and L amino acid transporters (Table 3). The extent of these changes, however, depends on timing of the insult in the pregnancy. In sheep, removal of uterine caruncles prior to pregnancy is associated with increased placental Igf2 expression and an adaptive increase in placentome size, trophoblast and maternal capillary volume and surface area, although total placental mass and fetal weight are reduced (Table 3; Zhang et al. 2016b). Acute exogenous IGF1 does not alter nutrient metabolism by the embolised sheep placenta (Table 1; Jensen et al. 1999). However, several doses of intra‐amniotic IGF1 increase glucose and Systems A and L amino acid transporter expression by the embolised placenta and improves feto‐placental growth in vivo (Table 1; Eremia et al. 2007; Wali et al. 2012). Moreover, in mice with uterine artery ligation, targeting of IGF1 to the placenta using a nanoparticle or adenoviral‐mediated approach increases the abundance of glucose and Systems A and L amino acid transporters in the placenta, placental width and fetal growth (Table 1; Jones et al. 2013, 2014; Abd Ellah et al. 2015). These findings highlight the therapeutic potential of IGFs for improving the capacity of the placenta to supply nutrients to the fetus in compromised pregnancies.

In ewes, heat stress reduces placental growth and glucose transport capacity, as well as altering the expression of IGF1 and IGF2, and AKT, mTORC1 and MAPK signalling pathways during gestation (Table 3). In rats, alcohol consumption during the peri‐conceptional period leads to late gestational fetal growth restriction but no change in placental weight (Gardebjer et al. 2014). However, Lz development and Igf1, Igf1r and SNAT2/Snat2/Slc38a2 expression are decreased, but Jz glycogen cell formation, Igf2 and GLUT1/Glut1/Slc2a1 may be increased in response to peri‐conceptional alcohol exposure near term (Gardebjer et al. 2014). This suggests there can be programmed changes in the conceptus leading to alterations in the IGF system and the structural and functional phenotype of the placenta that link the maternal environment from the earliest stages of pregnancy to fetal growth near term.

Conclusions and perspectives

Thus, IGFs are important regulators of placental resource allocation to fetal growth both developmentally and in response to environmental manipulations known to programme the ill health of offspring. They increase placental morphogenesis, substrate transport and hormone secretion, which, in turn promotes fetal growth either directly via the supply of nutrients and oxygen or indirectly via the maternal metabolic adaptation to pregnancy and the availability of nutrients for transplacental transport. In response to environmental challenges, the IGFs (particularly IGF2) and their signalling pathways change in line with the alterations in placental structure and function, and thereby link changes in the maternal environment to fetal substrate supply and growth during pregnancy with implications for developmental programming. The environmentally induced changes in the IGF system and placental phenotype may be beneficial (obesogenic diets, moderate hypoxia and low protein diets) or detrimental (e.g. severe oxygen and nutrient deprivation and glucocorticoid excess) to resource allocation to the fetus depending on the type, severity and timing of the challenge during pregnancy (Fig. 3). The beneficial effects of IGF treatments on placental phenotype show promising therapeutic potential for improving fetal growth in situations in which placental growth is impaired without major maternal compromise, particularly when the treatment with IGF1 or IGF2 is targeted directly to the placenta. However, efforts to understand the regulation of endogenous placental IGF expression may also be fruitful, particularly in the case of Igf2, which appears to be most important for mediating adaptive responses locally in mice. These findings are important in the context of human pregnancy as dysregulated expression of the IGFs and signalling components are often reported in the human placenta associated with abnormal fetal growth (Abu‐Amero et al. 1998; Sheikh et al. 2001; Gratton et al. 2002; Gurel et al. 2003; Laviola et al. 2005; Scioscia et al. 2006; Street et al. 2006; Trollmann et al. 2007; Akram et al. 2008; Yung et al. 2008; Colomiere et al. 2009; Borzsonyi et al. 2011; Street et al. 2011; Demendi et al. 2012; Jansson et al. 2012a; Iniguez et al. 2014; Nawathe et al. 2016; Zhang et al. 2016a). However, it is important to note that several causes of environmental, maternal and fetal origin can lead to changes in placental phenotype and fetal growth in humans (Gaccioli & Lager, 2016). Thus studies of animal models showing alterations in the expression of IGFs and their signalling pathways provide insight but further information is required on the natural conditions of variable placental phenotype among humans.

Figure 3. The effect of different environmental manipulations on the placental IGF system and resource allocation phenotype in the mouse.

A, manipulations that down‐regulate IGF2 signalling. B, manipulations that up‐regulate IGF2 signalling. AKT, protein kinase B; IGF, insulin‐like growth factor; Lz, labyrinthine zone. ^*Note Igf2P0 is required for the placenta to up‐regulate amino acid transport to the fetus in response to maternal undernutrition.

Additional information

Competing interests

None declared.

Author contributions

All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

A.N.S.‐P. is funded by a Royal Society Dorothy Hodgkin Research Fellowship.

Biography

Amanda Sferruzzi‐Perri received her Bachelor of Science degree with Honours and PhD degree from the University of Adelaide, Australia (in 2001 and 2007, respectively). In 2008 she received a C. J. Martin Overseas Biomedical Fellowship from the NH&MRC to undertake research at the University of Cambridge, UK. Through the award of a Next Generation Fellowship from the Centre for Trophoblast Research in 2011 and a Dorothy Hodgkin Research Fellowship from the Royal Society in 2014, she has been using a variety of strategies to decipher the role of insulin‐like growth factors and their signalling pathway, PI3K, in maternal–placental fetal interactions governing pregnancy success.