The Placenta and Intrauterine Programming

Authors


Correspondence to: Abigail L. Fowden, Department of Physiology, Development and Neuroscience, University of Cambridge, Physiology Building, Downing Street, Cambridge CB2 3EG, UK (e-mail: alf1000@cam.ac.uk).

Abstract

Intrauterine programming is the process by which the structure and function of tissues are altered permanently by insults acting during early development. In mammals, the placenta controls intrauterine development by supplying oxygen and nutrients, and by regulating the bioavailability of specific hormones involved in foetal growth and development. Consequently, the placenta is likely to have a key role in mediating the programming effects of suboptimal conditions during development. This review examines placental phenotype in different environmental conditions and places particular emphasis on regulation of placental nutrient transfer capacity and endocrine function by insults known to cause intrauterine programming. More specifically, it examines the effects of a range of environmental challenges on the size, morphology, blood flow and transporter abundance of the placenta and on its rate of consumption and production of nutrients. In addition, it considers the role of hormone synthesis and metabolism by the placenta in matching intrauterine development to the prevailing environmental conditions. The adaptive responses that the placenta can make to compensate for suboptimal conditions in utero are also assessed in relation to the strategies adopted to maximise foetal growth and viability at birth. Environmentally-induced changes in placental phenotype may provide a mechanism for transmitting the memory of early events to the foetus later in gestation, which leads to intrauterine programming of tissue development long after the original insult.

The pattern of intrauterine development and size at birth, in particular, are critical in determining life expectancy. They affect not only neonatal viability, but also adult rates of morbidity and mortality. Human epidemiological studies have shown that low birth weight is associated with an increased incidence of reproductive and neurological disorders in later life as well as with a greater risk of adult-onset cardiovascular and metabolic diseases, such as hypertension, coronary heart disease, obesity and type 2 diabetes (1). These associations are observed in populations of different age, gender and ethnicity and occur independently of current weight and life style factors, such as smoking and level of exercise (2). Similarly, intrauterine growth retardation induced either naturally or by a variety of different experimental techniques has been shown to cause abnormalities in postnatal physiological function in a wide range of species (3, 4). These techniques have included maternal stress, hypoxaemia, glucocorticoid administration, restriction of uterine blood flow, over- and undernutrition and isocalorific manipulation of the macro- and micro-nutrient content of the diet (4, 5). Taken together, the epidemiological and experimental findings have led to the concept that adult disease can originate in utero as a result of changes in development during suboptimal intrauterine conditions often associated with impaired foetal growth. The process by which insults during early development lead to permanent changes in tissue morphology and function is known as intrauterine programming (6).

In mammals, the placenta is the main interface between the foetus and mother and regulates intrauterine development in two main ways (5). First, it supplies the nutrients and oxygen required for tissue accretion and energy metabolism by the foetus. Second, it produces and metabolises a range of hormones with effects on foetal growth and development. These hormones act both directly on the foetal tissues and indirectly by metabolic actions in the mother and placenta to alter the foetal nutrient supply (7). Changes in the nutrient transfer capacity and/or endocrine function of the placenta in response to nutritional and other environmental conditions will, therefore, have important consequences for the growth and intrauterine programming of the foetus. However, compared to foetal tissues, relatively little is known about programming of the placenta per se (2, 7). This review examines the potential role of the placenta in intrauterine programming with particular emphasis on regulation of its nutrient transfer capacity and endocrine function by environmental factors shown experimentally to cause intrauterine programming.

Placental nutrient transfer capacity

The capacity of the placenta to deliver nutrients to the foetus depends on a range of factors including its size, morphology, blood flow, transporter abundance and its rate of consumption and production of nutrients (Fig. 1). These factors are closely inter-related during development and are responsive to environmental changes of maternal and foetal origin (5). Environmentally-induced changes in the placental nutrient capacity are, therefore, likely to be multi-factorial, although few studies have taken a comprehensive approach in determining the placental contribution to altered foetal nutrient delivery during suboptimal intrauterine conditions. To date, the most widely studied factor is placental size (2, 5, 8).

Figure 1.

 Schematic diagram showing the interaction between the mother, placenta and foetus in mediating the programming effects of environmental challenges on foetal growth and development.

Placental size

Placental size directly affects the capacity for nutrient transfer via changes in the surface area, which influence transport by both diffusion and transporter mediated processes (9). In many species, placental weight is positively correlated to foetal weight during normal conditions and when placental size is manipulated by removal of implantation sites, variations in litter size, ligation of umbilical vessels or by embryo transfer between breeds of different size (7, 9–13). Growth of the placenta is also compromised by most of the environmental challenges used experimentally to induce intrauterine growth restriction (IUGR) and programming of adult phenotype (Table 1). Longitudinal studies performed during human pregnancy have confirmed that, in cases of IUGR, placental volume is smaller than normal at 12 weeks of pregnancy and continues to remain small thereafter (32). Equally, the rate of placental growth between 17 and 20 weeks of gestation correlates strongly with anthropometric measurements of the foetus at 35 weeks (33), suggesting that placental development exerts a determining influence on size of the human infant.

Table 1.   Effects of Environmental and Genetic Factors on Placental Weight and Efficiency in Late Gestation in Different Species.
TreatmentSpeciesPlacental weightPlacental efficiencyaReference
  1. aMeasured as foetal to placental weight ratio.

OvernutritionSheepDecreaseIncrease(14)
Undernutrition
 ProteinRatDecreaseIncrease(15)
DecreaseNo change(16)
 CaloriesRatDecreaseNo change(16)
No changeDecrease(17)
Guinea PigDecreaseNo change(18)
SheepDecreaseIncrease(19)
DecreaseNo change(8)
Heat stressSheepDecreaseIncrease(20, 21)
Glucocorticoid administrationRatDecreaseIncrease(22)
SheepDecreaseIncrease(23)
MonkeyDecreaseIncrease(24)
HypoxiaGuinea pigDecreaseIncrease(25)
SheepNo changeNo change(26)
HumanNo changeDecrease(27)
Restricted uterine blood flowRatDecreaseNo change(28)
No changeNo change(29)
Gene manipulation
 Igf2 nullMouseDecreaseDecrease(30)
 Igf2P0 nullMouseDecreaseIncrease(30)
 Peg 3 nullMouseDecreaseIncrease(31)

Direct measurements of placental nutrient delivery during adverse conditions, such as maternal heat stress, glucocorticoid administration and dietary manipulation, have shown that both the quantity and specific types of nutrients supplied to the foetus are altered when placental growth is compromised (13, 20, 21, 23, 34). Similar changes in relative transfer of glucose and amino acids are observed in response to placental growth retardation induced by disruption of the placental specific transcript of the imprinted Igf2 gene in mice (30). In some instances, the placental supply of specific nutrients may be increased per gram of placenta, despite the reduction in placental size (30, 35). Indeed, most of the environmental challenges that restrict placental and foetal growth lead to an increase in placental efficiency measured as the foetal to placental weight ratio (Table 1). In part, this apparent increase in efficiency may reflect the functional reserve capacity of the placenta and be due to increased maternal nutrient availability and/or transplacental concentration gradients as a result of metabolic adaptations to the environmental insults in the mother and foetus (2, 8). Alternatively, it may reflect a genuine increase in the nutrient transfer capacity per gram of placenta following adaptations in the small placenta to help maintain foetal growth during suboptimal intrauterine conditions (Table 1).

Morphology

One mechanism by which the small placenta may become more efficient at nutrient transfer is by changes in its morphology. Alterations in the relative surface area for nutrient exchange, vascularity, barrier thickness and cell composition of the placenta will affect its transport characteristics and alter the foetal to placental weight ratio (9, 12). Many of the environmental insults that reduce placental weight also alter the gross morphology and ultrastructure of the placenta (18, 22, 36, 37). In sheep, gross placentome morphology in late gestation is influenced by nutritional conditions earlier in gestation and by more short-term changes in cortisol exposure close to term (5, 37–39). This has implications for nutrient transfer as the frequency of different placentome types affects the placental capacity to transfer glucose to the foetus when foetal cortisol concentrations are high (Fig. 2). Indeed, the reduction in placentome eversion observed in response to foetal cortisol infusion during late gestation may be an adaptation to maximise glucose transfer during adverse intrauterine conditions (37). Adaptations in gross placental morphology also occur when the genetic potential for placental growth is constrained by transferring a Thoroughbred (TB) embryo into the smaller pony uterus. In these circumstances, the chorionic villi appear to elongate, which increases the total microscopic area of foeto–maternal contact both in absolute terms and as a proportion of gross surface area, although there is no change in foetal to placental weight ratio compared to the TB in TB pregnancy (10).

Figure 2.

 Mean ± SEM rates of placental glucose transfer measured as umbilical glucose uptake per kg placenta in sheep foetuses infused intravenously for 5 days with either saline (n = 14) or cortisol (n = 14, 1–2 mg/kg/day) with respect to placentome morphology. (< 10% C/D types, open columns or > 30% C/D types, stippled columns, n = 7 in each group). *Significant difference by two-way anova with Tukey’s post-hoc test (P < 0.05). Data from (37, 39).

Utrastructurally, the changes in morphology induced by environmental challenges may reduce or enhance placental efficiency depending on the specific insult. In hypoxic conditions, there is a reduction in the thickness of the barrier between the maternal and foetal capillaries in the human and guinea pig placenta, which will facilitate diffusional exchange (25, 40). Conversely, undernutrition of guinea pigs during pregnancy increases barrier thickness and reduces not only placental weight, but also the relative size of the labyrinthine zone involved in nutrient exchange during late gestation (18). Similarly, in the growth retarded murine placenta lacking the Igf2P0 transcript specifically in the labyrinthine zone, there is an increase in barrier thickness and decreased passive permeability during late gestation (41). There are also reductions in the glycogen cell population in the junctional zone of the complete Igf2 null placenta (42). The reductions in the binucleate cell population induced in the ovine placenta by early cortisol exposure may also influence barrier thickness and the expansion of the foeto–maternal syncytium with consequences for diffusion of oxygen and nutrients (34). Changes in the rates of proliferation and apoptosis of specific cell types within the placenta induced by environmental and genetic factors are, therefore, likely to have an important role in foeto–placental growth and, hence, intrauterine programming. Imprinted genes, like Igf2, which are widely expressed in the placenta, may have an important role in these processes (43).

Blood flow

Blood flow is a critical determinant of placental function and foetal growth. In the vast majority of cases of unexplained human IUGR (i.e. not associated with congenital malformations, cigarette smoking or infections), there is evidence of increased resistance in the uterine arteries, even as early as 10–14 weeks of gestation (44). Morphological studies have demonstrated that physiological conversion of the spiral arteries supplying the placenta is deficient in these pregnancies, due to inadequate trophoblast invasion (45, 46). Consequently, maternal blood enters the placenta at an abnormally high rate causing damage to the placental villi (47). Deficient conversion also leads to retention of vasoreactivity in the spiral arteries and may exacerbate intermittent perfusion of the placenta (48). Fluctuations in oxygenation are a powerful inducer of placental oxidative stress, which leads to activation of stress-activated signalling pathways such as nuclear factor-κB and p38 mitogen-activated protein kinase (49). Endocrine and transport functions of the placenta are potentially severely affected. For example, there is evidence of increased placental expression of cyclooxygenase-2, leptin, pro-inflammatory cytokines and angiogenic regulators following induction of oxidative stress in vivo and in vitro (49, 50). By contrast, concentrations of placental growth factor are decreased.

Oxidative stress can also have a major effect on cell proliferation, especially if it occurs early in pregnancy when the placenta is growing rapidly. The trophoblast, smooth muscle cells surrounding the stem villus arteries and the foetal endothelial cells appear to be particularly sensitive to changes in oxygenation (51). Slower proliferation, and/or increased apoptosis, within the endothelial cells could account for the impoverished growth of foetal vessels seen in pathological placentas (52). This, in turn, will lead to increased resistance within the umbilical circulation and abnormal waveforms, in particular reversed end-diastolic flow, that are seen in the most severe cases of IUGR and that correlate strongly with foetal hypoxaemia.

During late gestation, placental vascularity increases in response to hypoxaemia and between breed embryo transfer (10, 26, 53). Conversely, when placental growth is compromised by heat stress of adult ewes or by overnutrition in adolescent sheep, placental vascularity decreases in association with reduced vascular endothelial growth factor (VEGF) and VEGF receptor expression in the placentomes (54, 55). Similar decreases in angiogenic factor are observed in term placentomes from adult ewes undernourished from early in gestation (54). However, weight specific rates of umbilical and uterine blood flow in late gestation are unaffected by heat stress, glucocorticoid administration, hypoglycaemia, acute fasting and more prolonged nutritional manipulations from earlier in pregnancy (14, 20, 21, 56, 57). These observations suggest that growth of the placenta is tightly correlated with growth of its blood vessels and, hence, with placental blood flow during late gestation. Foeto–placental growth is, therefore, matched to the capacity to supply oxygen and nutrients to the placenta for transfer to the foetus.

Transporter abundance

Another explanation for the increased efficiency of the small placenta is enhanced expression of nutrient transporters per unit surface area. Activity of the placental glucose and amino acid transport systems is influenced by a wide range of environmental factors including heat stress, hypoxia, under- and overnutrition as well as exposure to hormones such as glucocorticoids, growth hormone (GH) and leptin (21, 37, 58–61). For example, in the small placenta of the carunclectomised ewe, clearance of 3-O-methyl glucose, a non-metabolisable glucose analogue, is increased per kg placenta in late gestation (13). Similarly, placental glucose transfer measured as umbilical uptake per kg placenta is increased in the small placenta of ewes with prolonged hypoglycaemia when the reduced transplacental glucose concentration gradient is taken into account (56). These observations suggest that there are changes in the abundance or activity of the glucose transporters (GLUTs) when placental growth is compromised.

The primary placental GLUTs are GLUT1 and GLUT3, although GLUT8 has also been detected in human, rodent and ruminant placentas, particularly in early gestation (62–72). Expression of these GLUTs is affected by most of the nutritional and other insults known to cause intrauterine programming (Table 2). The responses are isoform specific and, in some cases, are also localised to particular placental membranes. For example, in the human diabetic placenta, an increase in GLUT1 abundance occurs in the basement membranes but not in the microvillous membranes at the foeto–maternal interface (64). Responses are also species specific and dependent on the type, duration and gestational age at onset of the insult (Table 2). In sheep, maternal hyperglycaemia initially increases placental GLUT1 abundance but eventually suppresses both isoforms when the stimulus is prolonged, whereas, lowering maternal glucose levels produces a sustained reduction in GLUT1 protein abundance but has no effect on GLUT3 levels, even after 40 days of hypoglycaemia (70). Reducing dietary intake in ewes during the period of maximal placental growth from 30 to 80 days has no effect on placental GLUT1 at 80 days but increases placental GLUT1 protein abundance later in gestation (66). By contrast, undernutrition late in gestation when placental weight has plateaued increases placental GLUT3 but not GLUT1 at term (69). Conversely, in rats, undernutrition near term reduces placental GLUT3 but not GLUT1 protein levels (17). Similar species and temporal differences in placental GLUT expression are observed in response to glucocorticoids. In rats, administration of synthetic glucocorticoids as a single injection on day 16 of pregnancy suppresses placental GLUT1 and GLUT3 protein abundance 5 days later but, when administration is continuous from days 15 to 20 of pregnancy, both isoforms are up-regulated on day 21 (71, 72). By contrast, raising cortisol levels for 5 days in sheep foetuses during late gestation has no effect on placental GLUT1 or GLUT3 mRNA expression (39).

Table 2.   Effects of Environmental Factors on Placental Glucose Transporter (GLUT) Abundance in Different Species during Late Gestation.
TreatmentSpeciesGlucose transporterChange in abundanceReference
DiabetesRatGLUT 1No change(28)
GLUT 3Increase(64)
HumanGLUT 1Increase(65)
GLUT3Decrease(66)
Maternal nutrition
 Over-nutritionRabbitGLUT 1No change(67)
 Under-nutritionSheepGLUT 1No change(58)
RatGLUT 1No change(17)
GLUT 3Decrease(17)
SheepGLUT 1Increase(68)
No change(69)
GLUT 3Increase(69)
 HyperglycaemiaRatGLUT 1No change(60)
SheepGLUT 1Decrease(70)
HumanGLUT1Decrease(71)
 HypoglycaemiaSheepGLUT 1Decrease(70)
GLUT 3No change(70)
Glucocorticoid administrationRat in vivoGLUT 1Increase(72)
GLUT 1Decrease(71)
Sheep in vivoGLUT 1No change(34)
Human in vitroGLUT 1Decrease(71)
HypothryoidismRatGLUT 1Decrease(73)
GLUT 3Increase(73)
HypoxiaMouse in vitroGLUT 1No change(74)
Human in vitroGLUT 1No change(75)
Human in vivoGLUT 1Decrease(27)
Reduce uterine blood flowRatGLUT 1Decrease(28)
No change(29)
GLUT 3No change(29)

GLUT1 and GLUT3 have different affinities and Km values and are found in different placental membranes depending on species and placental type (62). The Km of GLUT3 is two- to five-fold lower than the Km of GLUT1, which makes GLUT3 more efficient at transporting glucose at low concentrations. Since GLUT3 tends to be localised in membranes facing maternal blood or at the materno-foetal interface (62, 72), it is likely to be responsible for transplacental glucose transport and the rate of glucose delivery to the foetus. GLUT1, on the other hand, tends is be localised to the basal placental membranes and blood vessels, which suggests that it transports glucose into the placenta and, consequently, may have a role in regulating placental glucose consumption (62, 72). These observations are consistent with the effects of environmental challenges on placental GLUT expression (Table 2) and with the ontogenic increase in placental GLUT3 expression that occur in rats and sheep during late gestation when the foetal demand for glucose is rising most rapidly in absolute terms (76, 77). Certainly, in the labyrinthine specific Igf2P0 knockout mouse, there is up-regulation of placental expression of the Slc2a3 gene encoding GLUT3 in late gestation when the demand for glucose by the growing foetus exceeds the supply capacity of the small placenta (Fig. 3a). This helps maintains the glucose supply per gram of foetus at wild type values despite a 50% reduction in the surface area for nutrient transfer (30, 41).

Figure 3.

 Gene expression of (a) glucose transporters, GLUT1 and GLUT3 and (b) three isoforms of the System A amino acid transporters in mouse placenta from wild-type (WT, open columns) and placental specific Igf2P0 knockout mutants (P0, stippled columns) at day 16 of pregnancy. **Significantly different from wild-type (P < 0.05). Data from (30).

Compared to the GLUTs, less is known about the regulation of amino acid transporter abundance in the placenta during adverse intrauterine conditions. Expression of the System A amino acid transporter, SNAT2, is decreased in term placenta from rats fed a low protein diet during pregnancy and is increased in human placental cell lines after amino acid deprivation and cortisol treatment (78, 79). Reduced expression of taurine, leucine and System A transporters have also been reported in the human placenta of IUGR pregnancies (80–83). In human and rat placenta, at least nine different amino acid transporter systems have been identified with distinct functional characteristics yet overlapping substrate specificity (35). For the System A family of amino acid transporters, which transport small neutral amino acids, there are three known proteins in the placenta, each of which is encoded by a separate gene and regulated independently (30). Environmentally-induced changes in expression of a single amino acid transporters are, therefore, unlikely to have major effects on amino acid transport overall but could reduce availability of individual amino acids with consequences for the development of foetal tissues that depend on specific amino acids, such as the pancreatic β cells (84).

In mice, abundance of the placental amino acid transporters is also responsive to the foetal demands for nutrients. Expression of the Slc38a4 gene encoding a System A amino acid transporter is increased in the small mouse placenta lacking the placental specific Igf2 transcript in the labyrinthine zone (Fig. 3b). This response is associated with increased transport of methyl aminoisobutyric (MeAIB), a non-metabolisable amino acid analogue, and with normal foetal weight at day 16 of pregnancy (30, 43). Up-regulation of Slc38a4 expression and MeAIB transport is not seen when placental and foetal growth are both retarded in the complete Igf2 null, which lacks Igf2 in all foetal and placental tissues (30). Indeed, these placentas are less efficient and transfer less MeAIB per gram than the wild-types in association with reduced placental expression of several different amino acid transporters (30, 85).

These observations, together with those on glucose transport, suggest that the small mouse placenta can respond to nutrient demand signals from growing foetal tissues by increasing nutrient transporters abundance (Fig. 1). This adaptive response will counteract the reduced surface area for nutrient exchange and help maintain foetal growth for as long as possible, although the relative contribution of different nutrients to the total foetal requirement may differ from normal in these circumstances with consequences for tissue programming in utero (43). Since foeto–placental Igf2 gene expression is sensitive to several nutritional and endocrine challenges (86), changes in the interplay between placental and foetal Igf2 may explain, in part, the effects of environmental conditions on placental nutrient transporter abundance (5). However, the mechanisms by which insulin-like growth factors (IGFs) and nutrients alter the expression and activity of nutrient transporters at the cellular and molecular level remain unknown, although recent studies have implicated the mammalian target of rapamycin as a potential final common signalling pathway (80).

Nutrient consumption and production

Placental nutrient delivery to the foetus is also determined by the utilisation and production of nutrients by the placenta itself. In humans, ruminants and equids, the uteroplacental tissues use 40–70% of the glucose and oxygen taken up by the uterus (87, 88). They also produce lactate for foetal use and metabolise amino acids by transamination and deamination as part of a multi-organ system supplying essential and gluconeogenic amino acids to the foetus (89). Uteroplacental handling of both carbohydrates and amino acids is responsive to a range of environmental factors, including nutritional state, placental size, temperature and the concentrations of hormones, such as the glucocorticoids, catecholamines, GH and IGFs, in the foetal and maternal circulations (13, 34, 58, 61, 90).

In sheep, preterm exposure to cortisol increases uteroplacental glucose consumption and reduces the absolute amount and proportion of glucose taken up by the uterus that is delivered to the foetus (34). Fasting and moderate maternal undernutrition also reduce placental glucose consumption but have no effect on the partitioning of glucose between the uteroplacental and foetal tissues during short periods of hypoglycaemia (89). When maternal hypoglycaemia is more prolonged, the uteroplacental tissues conserve glucose for their own use and transfer proportionately less glucose to the foetus, although absolutes rates of glucose consumption are reduced in both the foetus and placenta (Fig. 4). However, after 2 months of hypoglycaemia, uteroplacental tissues are using proportionately less of the low rate of uterine uptake than when hypoglycaemia has lasted for 2–20 days, even though the percentage distribution between the foetal and uteroplacental tissues has still not returned to the values seen in normoglycaemic conditions (Fig. 4). These adaptations in uteroplacental glucose handling with the duration of hypoglycaemia may be due, in part, to temporal changes in placental GLUT1 expression (70).

Figure 4.

 Mean ± SEM rates of uterine glucose uptake (whole column), umbilical glucose uptake (open column) and uteroplacental glucose consumption (stippled column) together with the percentage distribution of the uterine uptake between the foetal and uteroplacental tissues in normoglycaemic, well fed ewes and in ewes made hypoglycaemic for varying periods by undernutrition (2 and 7 days) or maternal insulin infusion (20 and 65 days). Data from (56, 88, 91, 92) and A. L. Fowden (unpublished observations).

Placental handling of lactate and specific amino acids is also altered by nutrient and hormone availability and by growth retardation of the placenta itself (13, 14, 58, 61, 90). In sheep, placental lactate production and delivery to the foetus decreases in response to undernutrition, reduced uterine blood flow and foetal treatment with cortisol and IGF-I but increases above normal values after re-feeding fasted ewes and during maternal treatment with GH and IGF-I (34, 61, 89). Similarly, placental handling of the essential and gluconeogenic amino acids, in particular, is influenced by undernutrition, heat stress and treatment with glucocorticoids and IGF-I (20, 35, 61, 90). Foetal IGF-I treatment increases placental and foetal amino acid uptake wheras maternal treatment appears to stimulate placental amino acid uptake from the foetal circulation (90, 93). The differences in the placental response to maternal and foetal IGF-I administration suggest that there are complex interactions between the metabolic status of the foetus, placenta and mother in balancing nutrient distribution between the foeto–placental tissues to achieve optimal foetal growth. Certainly, environmentally-induced changes in the placental handling of essential amino acids are likely to have major implications for intrauterine programming as these amino acids are rate limiting for protein synthesis and tissue accretion in the foetus.

Placental endocrine function

The placenta produces a number of hormones, including steroids, peptides, cytokines, glycoproteins and eicosanoids, which are released into both the foetal and maternal circulation (Fig. 1). Some of these hormones, such as progesterone, placental lactogen and the placental variant of GH (GHv), have metabolic actions in the mother that favour glucose delivery to the foetus (7). In sheep, maternal concentrations of progesterone and placental lactogen are lower than normal during adverse conditions, such as heat stress, overfeeding and undernutrition (14, 21, 94). This may be due, in part, to the reduced size of the placenta (Table 1) but may lead to reduced placental delivery of glucose to the foetus independently of placental weight by altering partitioning of nutrients between the maternal and uteroplacental tissues. Placental production of GHv and cytokines, such as IGF-I and IGF-II, may also affect nutrient partitioning during pregnancy by endocrine actions on maternal metabolism and/or by paracrine actions on the placental capacity for simple and facilitated diffusion (61, 90, 93, 95). Receptors for these hormones are present in ovine and human placenta and ligand binding to these receptors can alter placental clearance of non-metabolisable glucose and amino acid analogues in vivo and in vitro (90, 93, 96). Maternal concentrations and placental expression of GHv per unit placental area are reduced in pregnant women with growth retarded infants (97), which suggests that changes in placental endocrine function may be a cause rather than a consequence of altered foetal growth.

Other placental hormones, such as CRH and the prostaglandins (PG)E2 and F, also affect the foetal supply of nutrients and oxygen but more indirectly by actions on foetal endocrine function, regional blood flow and myometrial contractility (98, 99). In sheep, horses and nonhuman primates, undernutrition during late gestation increases uteroplacental activity of PGH synthase and the production of PGF and PGE2 (100). These increments in PG synthesis are directly related to the fall in uteroplacental glucose consumption and can be reversed by restoring the normoglycaemia by re-feeding or glucose infusion into the fasted animal (99). Similarly, in mouse placenta in vitro, reduction in oxygen availability activates the enzymes responsible for PG synthesis (100). In sheep, horses and nonhuman primates, the placenta has also been shown to contain PG dehydrogenase (PGDH), an enzyme which converts biological active PGs into their inactive keto metabolites (98, 99). Uteroplacental output of these metabolites rises towards term and in response to undernutrition during late gestation in all three species (98, 99). In part, the nutritionally induced changes in PG production and metabolism may be due to the concomitant rise in glucocorticoid concentrations as cortisol has been shown to increase production of PGF and PGE2 by enhancing PGH synthase activity and decreasing PGDH activity in human and ovine placenta (99, 101). Sensitivity of placental PGDH and PGH synthase to nutritional and endocrine stimuli increases with increasing gestational age in parallel with the rise in foetal cortisol concentrations and the onset of myometrial contractile activity towards term (101). Environmentally-induced changes in the balance between placental PG production and metabolism may therefore have an important role in governing the length of gestation. Before term, these changes tend to help maintain pregnancy by favouring PG metabolism, whereas, close to term, they precipitate delivery of viable offspring by shifting the balance towards production of the primary PGs.

Placental exposure to glucocorticoids is affected not only by the foetal and maternal glucocorticoid concentrations but also by the activity of the enzyme 11β-hydroxysteroid dehydrogenase type 2 (11βHSD2), which converts active glucocorticoids to their inactive metabolites (102). This enzyme therefore limits placental and foetal exposure to the higher glucocorticoid concentrations found in the maternal circulation of most species. Its placental activity varies with species in relation to the magnitude of the transplacental glucocorticoid gradient (7) and is regulated by a wide range of environmental factors (Table 3). In particular, long-term nutritional and other stresses during pregnancy down-regulate placental 11βHSD2 activity, which will increase glucocorticoid exposure of the placenta (Table 3). In turn, this influences the placental production and metabolism of other glucocorticoid sensitive hormones such as the PGs, progesterone, oestrogens and placental lactogen (7, 99, 124). Increased foeto–placental glucocorticoid exposure will also impair foeto–placental growth directly by switching the cell cycle from proliferation to differentiation, although these changes may have beneficial effects on neonatal viability (124). Certainly, in human and rats, placental 11βHSD2 activity is positively correlated to birth weight whereas inhibition of this enzyme in rats during pregnancy retards foeto–placental growth and leads to abnormalities in cardiovascular and metabolic function in the adult offspring (102, 125). In rodents and humans, the activity and sensitivity of placental 11βHSD2 to stimuli appears to be sex-linked (115, 118), which may explain, in part, the gender specificity of the phenotype programmed in utero.

Table 3.   Effects of Environmental Factors on the Activity or Gene Expression of Placental 11β-Hydroxysteroid Dehydrogenase Type 2 in Different Species during Late in Gestation.
TreatmentSpeciesPlacental 11βHSD2 abundanceReference
AlcoholRatIncrease males; Decrease females(103)
InfectionHumanDecrease(104)
DiabetesRatDecrease(105)
Maternal undernutrition
 ProteinRatDecrease(106, 107)
 CaloriesRatDecrease(17)
SheepDecrease(108, 109)
Increase(110)
Maternal stress
 AcuteRatIncrease(111)
 ChronicRatDecrease(112)
Glucocorticoid administration
 In vitroRatDecrease(113)
HumanIncrease(114)
 Mother in vivoRatNo effect(115)
SheepDecrease(116)
MonkeyIncrease(117)
HumanIncrease(118)
 Foetus in vivoSheepDecrease(119)
RatNo effect(120)
Hypoxia
 In vitroHumanDecrease(121)
 Mother in vivoSheepNo effect(122)
 Foetus in vivoSheepNo effect(123)

In addition to glucocorticoids, the placenta inactivates a range of other hormones that regulate foetal metabolism and growth including catecholamines, IGFs and thyroxine. This will limit the effectiveness of these hormones in the foetal circulation. For example, placental type III deiodinase (D3) converts thyroxine to biological inactive reverse tri-iodothyronine and maintains a high foetal clearance rate of tri-iodothyronine, the most biological active thyroid hormone (7). Compared to 11βHSD2, much less is known about regulation of metabolism of these other hormones, although placental D3 activity is suppressed by increased cortisol levels (126). The ovine placenta also clears IGF-I from the umbilical circulation when foetal IGF-I concentrations are high but secretes IGF-I when foetal concentrations are low (96). Since foetal IGF-I concentrations are positively correlated to foetal body weight (86), these observations suggests that the placenta can adapt hormone production and metabolism in relation to foetal growth to maintain the optimal growth rate for the prevailing conditions in utero. The balance between hormone production and metabolism by the placenta, therefore, depends on the growth rate and gestational age of the foetus and on the precise nature of the environmental challenges to which the foeto–placental tissues are exposed during intrauterine development.

Conclusions

The nutrient transfer capacity and endocrine function of the placenta are responsive to environmental stimuli and can adapt to help maintain the foetal nutrient supply, particularly when the placenta is small (Fig. 1). The placenta may therefore act as a nutrient sensor (80), modifying nutrient and hormone availability to foeto–placental tissues in relation to the environmental conditions in a manner that maximises foetal growth, yet maintains pregnancy for long enough to ensure delivery of viable young. Environmentally-induced changes in placental development are likely to affect intrauterine development in several ways. First, they affect the absolute amount and specific types of nutrients available for foetal growth by changes in placental size, morphology, blood flow, transporter abundance and nutrient utilisation induced by the insult itself. Second, in some instances, the mismatch between the foetal drive for growth and the actual nutrient supply during adverse conditions can lead to adaptations in the placenta, which, although improving the total nutrient supply, may alter the relative contribution of specific nutrients with consequences for the programming of individual foetal tissues. Similarly, environmentally-induced changes in placental endocrine function may have beneficial effects in maintaining pregnancy and/or diverting nutrients from maternal to foetal use yet simultaneously have adverse sequelae for the development of foetal and maternal tissues, such as the mammary gland, with implications for nutrition and growth after birth. Environmentally-induced changes in placental phenotype may therefore provide a mechanism for transmitting the memory of early events to the foetus later in gestation and determining foetal tissue development long after the original insult. Therefore, the placenta has an important role in intrauterine programming and its phenotype may be a better marker of disease risk in later life than many of the other indices currently used to identify exposure to suboptimal conditions during intrauterine development (83). The epigenetic mechanisms that control placental nutrient transfer capacity and endocrine function at the molecular level remain largely unknown, but may involve changes in gene imprinting, DNA methylation, histone modifications and/or in mitochondrial oxidative function (43, 127) Further studies are required to determine whether any of these factors are important in programming the placenta.

Acknowledgements

The authors would like to thank their many colleagues who helped with this work and in the preparation of the manuscript. We are grateful to the BBSRC and the Wellcome Trust for their financial support.

Ancillary