Placental Hormones and the Control of Fetal Growth

The weight of a child at birth is a critical determinant of neonatal morbidity and mortality and the risk of metabolic disease throughout the life span: growth retarded [intrauterine growth retardation/small for gestational age (SGA)] infants are predisposed to hypoglycemia, hypocalcemia, hypothermia, asphyxia, and cognitive dysfunction; those who have rapid catch-up weight gain in early childhood are at increased risk for adult-onset type 2 diabetes, hypertension, preeclampsia, dyslipidemia, and ischemic heart disease (1). Likewise, large for gestational age (LGA) infants are predisposed to childhood weight gain, type 2 diabetes, and the metabolic syndrome (2). Despite a wealth of investigation, the molecular mechanisms controlling fetal weight gain and growth remain poorly understood. The paper by Mannick et al. (3) in this month’s Journal provides new insight into the roles of placental hormones in the regulation of fetal growth in normal and pathological pregnancies. The placenta controls fetal growth through delivery of blood, oxygen, and nutrients to the fetus and clearance of fetal waste products including carbon dioxide and urea. The substrates for uteroplacental transport are derived from maternal nutrient stores (4). In early and mid-gestation, maternal food intake increases 10–15%, intestinal calcium absorption doubles, and there is a 60% rise in first-phase insulin secretion. Because insulin sensitivity is preserved during the first half of pregnancy, the increase in insulin secretion promotes lipogenesis and limits fatty acid oxidation, facilitating maternal fat storage. In mid- and late gestation, maternal food intake and fat mass escalate, but maternal metabolism is transformed by the emergence of insulin resistance. This facilitates maternal utilization of free fatty acids as an energy source, sparing glucose, amino acids, essential fatty acids, and ketones for placental-fetal transport and fetal growth. The metabolic adaptations of pregnancy are orchestrated by hormones produced by the placenta and maternal pituitary gland, which undergo dramatic changes during gestation (4). After involution of ovarian sex steroid production by wk 6, placental estrogen and progesterone production increases exponentially to term. Concurrently, there are progressive increases in prolactin (PRL), produced by the maternal pituitary gland and decidua, and human chorionic somatomammotropin (CSH, also called human placental lactogen), which has structural similarities to GH and PRL. CSH binds with high affinity to PRL receptors (Kd 0.1 nm) but with low affinity to human GH receptors (Kd 770 nm), suggesting that it functions as a lactogen rather than as a somatogen during pregnancy. Interestingly, levels of pituitary GH (GH1) decline sharply during early and mid-pregnancy and are no longer detected in maternal serum after 24 wk gestation. GH1 is replaced by a variant GH [GH2 or placental GH (PGH)] produced by the syncytiotrophoblast. In contrast to pituitary GH, the variant GH is secreted tonically and in increasing amounts throughout the second half of gestation, reaching levels (20–40 ng/ml) comparable to those in acromegaly. PGH binds with high affinity to GH receptors but with low affinity to PRL receptors. Thus, the mother in mid- and late gestation is bathed in high levels of two lactogenic hormones (PRL, peak 150–180 ng/ml; and CSH, peak 5000–7000 ng/ml) and a nearly pure somatogen (PGH). CSH is also secreted directly into the fetal circulation, reaching levels approximating 20–30 ng/ml, whereas fetal pituitary PRL levels can climb as high as 200–300 ng/ml; in contrast, PGH is detected only in maternal blood. Previous studies had found low maternal levels of PGH and CSH in pregnancies complicated by hypertension, preeclampsia, and intrauterine growth retardation (5, 6). Mannick et al. (3) extend this work by exploring the relation between PGH and CSH expression and birth weight in normal term pregnancies. The expression of PGH and CSH transcripts was assessed by a sophisticated fluorescent RT-PCR assay coupled with gene-specific restriction analysis. Their study has a number of strengths. Most important, the cohort was a homogeneous population of healthy Caucasian women without known underlying metabolic or pregnancy-related diseases. Mothers with diabetes or toxemia and infants with fetal anomalies or genetic or infectious diseases were excluded. This design eliminates a number of major complicating factors and may provide a window into the physiological, rather than strictly pathological, role of PGH and CSH in fetal growth regulation. Five of the 15 women in the SGA group were smokers, but the authors have controlled for this confounder in their multiple regression analysis. Their findings are novel. First, they show that half of the PGH transcripts detected in term placenta encode proteins with C termini distinct from the major GH2 product (GH2-1); the functions of these alternative proteins are unknown, but they merit further scrutiny. Second, the expression of GH2-1 is reduced in placentas from pregnancies of SGA babies but is normal in placentas of LGA pregnancies. Third, the predominant CSH transcript (CSH1) is increased in placentas of LGA pregnancies but is unchanged in pregnancies of SGA infants. These findings dovetail nicely with previous studies by Caufriez et al. (7), who found high levels of CSH in the blood of pregnant mothers with gestational diabetes and low levels of PGH in pregnant mothers giving birth to SGA infants. Similar findings were recorded by Ursell et al. (8) and McIntyre et al. (9). Thus, the major placental syncytiotrophoblast protein members of the GH/CSH hormone family are regulated differentially, at least in late gestation. How can the differential regulation of PGH and CSH be explained, and what is its pathophysiological significance? Some insight may be gleaned by consideration of this study’s cohort. Auxological differences in the infants were asymmetric; birth weight (19.1–35%) and abdominal circumference (8.6–11.4%), reflecting fat deposition, were affected to a far larger degree in the LGA and SGA infants than birth length (5.9–7.9%) or head circumference (5.7–7.1%). Moreover, differences in placental weight (19.6–30%) were comparable to differences in birth weight. These findings bespeak effects of PGH and CSH on fetal weight gain during the last trimester, when fetal fat stores and liver glycogen deposition increase exponentially. Maternal weight and fat storage appear to play important roles; mothers of SGA children had lower prepregnancy body mass indices (BMIs) and less pregnancy weight gain than mothers of appropriate for gestational age children; the converse was true of mothers of LGA children. In contrast to some other studies (1), the height of the mother and the height and weight of the father had no relation to birth weight. In sum, these observations suggest that changes in fetal growth associated with differential PGH and CSH gene expression may be mediated or modulated by maternal nutritional status. How do maternal nutritional status and placental hormone expression interact to regulate fetal growth? One hypothesis proposes that polymorphic or epigenetic regulation of PGH and CSH expression could alter the expression of other critical hormones or growth factors such as insulin or IGF-I, change the distribution and availability of maternal nutrients, and/or affect directly the replication and/or growth of fetal tissues. Because PGH is an insulin antagonist and potent lipolytic hormone that stimulates IGF-I production during pregnancy (7, 10), a reduction in PGH expression would be expected to reduce lipolysis and circulating IGF-I levels. This in turn could reduce fetal weight by reducing maternal nutrient availability and limiting placental nutrient transfer (11). Conversely, increases in CSH expression have been postulated to induce insulin resistance and lipolysis and to increase maternal IGF-I expression (5); these in theory would increase maternal nutrient availability and placental transfer and promote fetal growth. But this hypothesis is problematic for several reasons. First, a reduction in lipolysis caused by PGH deficiency should be accompanied by an increase in maternal fat stores; but many women who give birth to SGA infants, like those in this and the McIntyre et al. study (9), have low BMIs. Second, the evidence that CSH is lipolytic and IGF-I-tropic is weak at best. Nearly all studies showing lipolytic effects of CSH were performed using placental extracts obtained before the discovery of PGH; more recent studies using recombinant hormones show that the lactogens may inhibit lipolysis (12) and promote adipogenesis (13). Because the molecular weights of CSH and PGH are nearly identical, it is conceivable that CSH purified from human placenta contained amounts of PGH sufficient to induce lipolysis in vivo and in vitro. Moreover, mice with lactogen resistance resulting from a knockout of the PRL receptor, which binds placental lactogen as well as PRL, have normal insulin sensitivity and normal plasma IGF-I and IGF binding protein-3 levels (14, 15). An alternative or complementary hypothesis is grounded in the idea that changes in PGH and/or CSH expression represent responses to, or effects of, underlying metabolic or genetic disturbances that alter uteroplacental function. In this schema, PGH expression may decline directly as a consequence of uteroplacental insufficiency resulting from an underlying disease like hypertension or systemic lupus erythematosus or a generalized disorder of placental development. Alternatively, or in addition, the fall in PGH may be an indirect consequence of maternal undernutrition: decreased maternal weight or pregnancy weight gain would in theory increase maternal insulin sensitivity and plasma adiponectin, which reduces PGH expression in trophoblast cells in vitro (16). A fall in PGH expression might limit maternal lipolysis, thereby compensating in part for a reduction in maternal fat stores. Both mechanisms could explain the reduction in PGH and the fall in maternal IGF-I in pregnancies associated with fetal growth retardation. Conversely, an increase in maternal BMI or gestational weight gain would increase maternal fat stores and reduce maternal insulin sensitivity, predisposing to glucose intolerance in the pregnant state. The resulting relative or absolute hyperglycemia would increase placental weight, possibly through induction of IGF-II (11), and fetal weight, through induction of fetal hyperinsulinemia (1, 11). The increase in maternal fat would reduce plasma adiponectin; this may increase CSH expression because adiponectin suppresses CSH as well as PGH (16). Increases in CSH in the fetal circulation could promote hyperinsulinemia by induction of β-cell replication (14, 17), thereby increasing fetal weight gain. Increases in maternal CSH and induction of maternal β-cell replication and insulin production could serve to compensate for maternal insulin resistance and defend against, or reduce the severity of, gestational diabetes (14, 18, 19). Definitive evidence supporting these hypotheses is currently lacking. Indeed, many have questioned the roles of CSH and PGH in human physiology because some women with deletions of the CSH and/or PGH genes have had normal pregnancies and have given birth to children of normal birth weight (4). These findings suggest that neither CSH nor PGH per se is required for normal pregnancy; however, it is possible that other hormones, such as pituitary GH and/or PRL, might compensate for the absence of PGH or CSH. There are no clinical models of combined deficiencies of lactogen and somatogen action in pregnancy or postnatal life, but studies of a lactogen-resistant/GH-deficient mouse (15, 20) provide evidence that the lactogens and somatogens act through overlapping but distinct mechanisms to regulate metabolism and growth. It should also be noted that a review (4) of published cases (n = 11) of CSH and PGH deletions showed variable fetal growth failure (mean birth weight, 2791 ± 203 g; range, 1270–3600 g), suggesting incomplete compensation in some cases or the presence of other factors that can maintain pregnancy and fetal development in the absence of CSH and/or PGH. Disclosure Summary: The author has nothing to disclose. Body mass index chorionic somatomammotropin large for gestational age placental GH prolactin small for gestational age