The Role of Placental Hormones in Mediating Maternal Adaptations to Support Pregnancy and Lactation

During pregnancy, the mother must adapt her body systems to support nutrient and oxygen supply for growth of the baby in utero and during the subsequent lactation. These include changes in the cardiovascular, pulmonary, immune and metabolic systems of the mother. Failure to appropriately adjust maternal physiology to the pregnant state may result in pregnancy complications, including gestational diabetes and abnormal birth weight, which can further lead to a range of medically significant complications for the mother and baby. The placenta, which forms the functional interface separating the maternal and fetal circulations, is important for mediating adaptations in maternal physiology. It secretes a plethora of hormones into the maternal circulation which modulate her physiology and transfers the oxygen and nutrients available to the fetus for growth. Among these placental hormones, the prolactin-growth hormone family, steroids and neuropeptides play critical roles in driving maternal physiological adaptations during pregnancy. This review examines the changes that occur in maternal physiology in response to pregnancy and the significance of placental hormone production in mediating such changes.


INTRODUCTION
Pregnancy is a dynamic and precisely coordinated process involving systemic and local changes in the mother that support the supply of nutrients and oxygen to the baby for growth in utero and in the subsequent lactation. Inappropriate adaptation of maternal physiology may lead to complications of pregnancy, such as gestational diabetes, preeclampsia, fetal growth restriction, fetal overgrowth and pre-term birth; which can have immediate consequences for fetal and maternal health. Furthermore, these pregnancy complications can also lead to long-term health consequences for the mother and infant. Altered fetal growth is associated with an increased risk of the offspring developing obesity, type-2 diabetes and cardiovascular disease in adulthood (Hales and Barker, 2001;Barker, 2004;Fowden et al., 2006). Moreover, women who develop gestational diabetes or preeclampsia are more likely to develop type-2 diabetes or cardiovascular disease in later life (Kim et al., 2002;Petry et al., 2007). Maternal adaptations to pregnancy are largely mediated by the placenta; the functional interface between the mother and fetus that secretes hormones and growth factors into the mother with physiological effects. This review aims to provide an overview of the physiological changes that occur in the mother in response to pregnancy and to discuss the role of key placental hormones in mediating such adaptations. In particular, this review focuses on the importance of the prolactin-growth hormone family (e.g., prolactin, placental lactogen and growth hormone), steroids (estrogens and progesterone) and neuropeptides (serotonin, melatonin and oxytocin) in adaptations of maternal physiology during pregnancy. Where possible, this review draws upon findings in women and animal models, including rodents and sheep. However, differences exist between species in the specific hormones produced by the placenta, the access of these hormones to the maternal circulation, and the relative proportion of conceptus mass to maternal size (hence constraint on the mother to provide resources for fetal growth; Haig, 2008;Carter, 2012;Fowden and Moore, 2012). Where such differences between species exist, these have been highlighted and discussed as necessary in the relevant sections. Nevertheless, although some effects described may not be applicable to all species, the different animal models of pregnancy still provide novel insight into the fundamental mechanisms of maternal adaptation during gestation.

ADAPTATIONS IN MATERNAL PHYSIOLOGY DURING PREGNANCY AND LACTATION
Most tissues and organs in the mother respond to the pregnant state. Changes include alterations in size, morphology, function and responsiveness of tissues and organs to hormonal and metabolic cues. These changes arise in the cardiovascular, pulmonary, immune, and metabolic systems of the mother (Figure 1). Some of these changes are seen from very early in pregnancy, prior to the establishment of a fully functional placenta, highlighting that non-placental factors may also be important (Paller et al., 1989;Drynda et al., 2015). The specific nature of changes in maternal physiology depends on the stage of the pregnancy and appears to track with alterations in the metabolic requirements of the mother versus the developing fetus.
Alterations in the maternal cardiovascular system begin very early in gestation (Chapman et al., 1998) and ultimately lead to systemic vasodilation and increased blood perfusion of maternal organs, including the gravid uterus. Systemic vascular resistance is reduced by 25-30% and accompanied by a 40% increase in cardiac output during human pregnancy; while in mice, blood pressure decreases by 15% and cardiac output is increased by 48% (Bader et al., 1955;Kulandavelu et al., 2006;Soma-Pillay et al., 2016). Renal blood flow and glomerular filtration rates are also increased (Davison and Dunlop, 1980;Soma-Pillay et al., 2016). The renin-angiotensin-aldosterone system (RAAS) which is a major determinant for sodium balance during gestation, is progressively upregulated toward term with associated plasma volume expansion (Elsheikh et al., 2001;Tkachenko et al., 2014). This rise in blood volume, which is required to cope with the oxygen requirements of the maternal organs and the conceptus growth, plateaus by the late gestation, resulting in an increase in total blood volume by approximately 30% at the end of pregnancy (Chang and Streitman, 2012). There is also an increase in the numbers of red blood cells in the mother during pregnancy, due to proliferation of erythroid progenitors in the spleen (Bustamante et al., 2008). Pulmonary function is also altered and encompasses changes in ventilation rates and blood gases. For instance, lung tidal volume and minute ventilation increases by 30-50% (Hegewald and Crapo, 2011). As a result of increased oxygen consumption during hyperventilation, there is greater carbon dioxide production, which leads to chronic respiratory alkalosis that is compensated by an increased renal excretion of bicarbonate (Weinberger et al., 1980). Overall, these adaptations ensure the well-being of the mother, while also providing an adequate blood flow to the placenta for fetal nutrition, oxygenation and maturation.
There are also alterations in maternal metabolic and endocrine state during gestation. In early pregnancy, the maternal pancreatic β-cell mass expands due to both hyperplasia and hypertrophy of islets, which for example in rats, results in a >50% increase (Ackermann and Gannon, 2007;Rieck and Kaestner, 2010). The threshold for glucose-stimulated insulin production is also lowered and maternal circulating insulin concentration is greater compared to the non-pregnant state. In early pregnancy, when fetal demands are relatively low, whole body maternal insulin sensitivity is unchanged or increased and there is accumulation of energy reserves in the mother. In particular, early pregnancy is associated with adipocyte hypertrophy, increased lipogenesis and lipid storage and relates to improved insulin sensitivity of white adipose tissue in the mother (Hadden and Mclaughlin, 2009;Mcilvride et al., 2017). Interestingly, in pregnant mice, brown adipose stores of the dam also switch to a white adipose tissue-like phenotype in early gestation (Mcilvride et al., 2017). Additionally, glycogen accumulates in the liver, which also increases in size from early gestation (Bustamante et al., 2010). In contrast, late pregnancy is associated with diminished maternal tissue insulin sensitivity and a concomitant increase in lipolysis and hepatic gluconeogenesis (Freemark et al., 2002;Lain and Catalano, 2007;Musial et al., 2016). Despite the pregnancy-related rise in leptin and insulin concentrations, maternal appetite increases in pregnancy (Villar et al., 1992;Douglas et al., 2007;Hadden and Mclaughlin, 2009;Díaz et al., 2014). Together, these metabolic and endocrine alterations increase lipid and glucose availability for the rapidly growing fetus in late gestation. Intriguingly in rodents, whole body responsiveness to insulin starts to improve near term, which may be important for conserving nutrients for maternal use, as parturition and lactation approach (Musial et al., 2016). There are also notable changes in maternal bone metabolism during pregnancy. In particular, intestinal calcium absorption is enhanced in the mother during pregnancy via upregulation of 1,25-dihydroxyvitamin D levels, improved renal conservation and increased calcium mobilization from the maternal skeleton (Hellmeyer et al., 2006). These processes support the supply of calcium for the formation, growth and mineralization of the fetal skeleton (King, 2000;Kalkwarf and Specker, 2002).

PLACENTAL HORMONES THAT MEDIATE MATERNAL ADAPTATIONS TO PREGNANCY, PARTURITION AND LACTATION
The placenta is a highly active endocrine organ during gestation; secreting a variety of hormones with physiological effects in the mother. Placental hormones include members of the prolactin and growth hormone family, steroid hormones and neuroactive hormones. The function of these hormones in driving physiological changes during pregnancy has been assessed in two main ways. First, the expression and activity of the hormones have been manipulated in vivo by either exogenously administering or genetically manipulating the expression of hormones and hormone receptors to study the physiological consequences for the animal. Secondly, hormones have been manipulated similarly in cultured cells and tissue explants to inform on the cellular and molecular mechanisms by which they modulate function. The effects of hormones in non-pregnant animals have been included as they provide information on the baseline of physiological changes that occur in the absence of hormone expression/activity, which is especially important in the case of some placental-derived hormones, where analyses in the pregnant state have not been conducted.

Prolactin (PRL)-Growth Hormone (GH) Family
The PRL-GH family is one of the main families of hormones secreted by the placenta during gestation. Members of this family consist of prolactin (PRL) (Handwerger et al., 1992), placental lactogens (PLs) (Wiemers et al., 2003), PRL-like hormones (Wiemers et al., 2003), proliferins (PLF) (Lee et al., 1988), proliferin-related proteins (PRP) (Jackson et al., 1994) and growth hormone (GH). Between mammalian species, there are differences in the number and type of family members expressed by the placenta [reviewed elsewhere (Linzer and Fisher, 1999;Soares, 2004;Soares et al., 2007)]. For instance, in the mouse and rat, the placenta expresses all these members except for PRL and GH whereas the human placenta only expresses GH and PL genes. In mice and rats, expression of the individual PRL-GH family members vary spatially and temporally in the placenta (Dai et al., 2002;Simmons et al., 2008;Urbanek et al., 2015). The anterior pituitary also produces PRL and GH; however this is diminished by mid-pregnancy, when placental hormone production predominates (Bridges, 2015). In several species including rodents and humans, PRL is additionally produced by the decidua during pregnancy. The family members share structural similarity to one another and may bind, with varying affinity to PRL and GH receptors (PRLR and GHR, respectively), which are widely expressed by tissues in the body (Haig, 2008;Trott et al., 2008;Ben-Jonathan and Hugo, 2015). As the PRL-GH members also exert similar functions, these have been presented in a grouped fashion in the text and tables (Tables 1, 2). However, where possible, the roles of individual family members of the PRL-GH in physiological changes have been described.
Studies performed both in vivo and in vitro support a role for the PRL-GH family in mediating maternal metabolic adaptations to pregnancy (Tables 1, 2). PRL, PRL-like proteins and PL, principally via the PRL receptor, induce β-cell mass expansion by both increasing β-cell proliferation and reducing apoptosis of islets in vivo and in vitro ( Table 2; PRL/PL/GH; Brelje et al., 1993;Huang et al., 2009). PRL and PL also increase insulin secretion during pregnancy, particularly in response to glucose, by enhancing the expression of glucose sensors (glucokinase, hexokinase and glucose transporter-2) and activating the serotonin biosynthesis pathway in pancreatic islets ( Table 2; PRL/PL/GH; Nielsen, 1982;Brelje et al., 1989Brelje et al., , 1993Weinhaus et al., 1996;Sorenson and Brelje, 1997;Arumugam et al., 2014). Moreover, PL protects β-cells against streptozotocininduced cell death in mice (Fujinaka et al., 2004). GH may also be important for modulating pancreatic insulin production (Billestrup and Nielsen, 1991;Brelje et al., 1993). However, GH from the placenta appears to be primarily important in the acquisition of insulin resistance and shifting metabolic fuel use from glucose to lipid in the mother during pregnancy ( Table 1; PRL/PL/GH; Horber and Haymond, 1990;Goodman et al., 1991;Galosy and Talamantes, 1995;Barbour et al., 2002;Dominici et al., 2005;Boparai et al., 2010;Liao et al., 2016b;Sairenji et al., 2017). Placental GH reduces insulin receptor expression and signaling, as well as, diminishes the abundance of the insulin-sensitive glucose-transporter, GLUT-4, in the skeletal muscle (Barbour et al., 2004;Kirwan et al., 2004). Insulin receptor abundance and signaling in the liver is also reduced in response to increased GH abundance in transgenic mice (Dominici et al., 1999). In white adipose tissue, GH also disrupts the insulin signaling pathway, and inhibits insulin action on glucose uptake and lipid accumulation (Del Rincon et al., 2007). In part, the effects of GH may be mediated through insulin-like growth factor-1 (IGF1), which is primarily secreted from the liver in response to GH and exerts lipolytic effects during pregnancy (Randle, 1998;Sferruzzi-Perri et al., 2006;Del Rincon et al., 2007). Insulin-like growth factor-2 (IGF2), which is not directly regulated by GH, but is secreted by the placenta is also important for modulating the sensitivity of β cells to glucose (Tables 1, 2; IGF2; Casellas et al., 2015;Modi et al., 2015) and maternal insulin and glucose concentrations during pregnancy (Petry et al., 2010;Sferruzzi-Perri et al., 2011). Polymorphisms/mutations in the PRL-GH family of genes and receptors have been reported in human pregnancies associated with gestational diabetes and fetal growth restriction (Rygaard et al., 1998;Le et al., 2013). Moreover, loss of PRLR signaling in β-cells causes gestational diabetes mellitus (GDM) in mice (Banerjee et al., 2016). Taken together, the production of PRL-GH family of hormones by the placenta appears to be important in regulating both insulin production and sensitivity of the mother in response to pregnancy.

Steroid Hormones
The placenta is a primary source of steroid hormones during pregnancy. Placental steroid hormones include estrogens and progesterone (Costa, 2016;Edey et al., 2018). In species like rodents, the corpus luteum continues to contribute to the circulating pool of steroid hormones during pregnancy, whereas in other species such as humans and ruminants, the placenta serves as the main source (Costa, 2016). Physiological effects of progesterone are mediated predominately by nuclear receptors (PR-A, PR-B) although membrane bound-type receptors (mPR) enable non-genomic actions. Steroid hormones are implicated in pregnancy complications such as gestational diabetes and preeclampsia. High progesterone and estrogen concentrations have been reported for women with gestational diabetes (Branisteanu and Mathieu, 2003;Qi et al., 2017). Moreover, placental estrogen and progesterone levels are reduced in preeclamptic patients compared with healthy pregnant women (Açikgöz et al., 2013).
Studies performed in vivo, suggest placental steroid hormones may be important in driving the changes in insulin sensitivity and glucose metabolism of the mother during pregnancy ( Table 3). Hyperinsulinemic-euglycemic clamp studies in women and rodents highlight a role for progesterone in reducing maternal insulin sensitivity during pregnancy. Progesterone administration decreases the ability of insulin to inhibit glucose production by the liver, and diminishes insulin-stimulated glucose uptake by skeletal muscle and to a lesser extent in the adipose tissue of non-pregnant animals ( Table 3; Progesterone; Leturque et al., 1984;Ryan et al., 1985;Kim, 2009). In contrast, exogenous estrogen increases whole body insulin sensitivity in non-pregnant state ( Table 3; Estrogen; Ahmed-Sorour and . Similarly, genetic deficiency of ERα or aromatase (Cyp19), which is involved in estrogen production, reduces hepatic and whole body insulin sensitivity and impairs glucose tolerance in non-pregnant mice (Takeda et al., 2003;Bryzgalova et al., 2006). Loss of the estrogen receptor or estrogen production is also associated with increased body weight, adiposity and hepatic lipogenesis ( Table 3; Estrogen; Takeda et al., 2003;Bryzgalova et al., 2006). Progesterone and estrogen also exert opposite effects on food intake in vivo ( Table 3). In particular, estrogen depresses food intake in part via induction of leptin production by adipose tissue, whereas progesterone increases food intake by enhancing NPY and reducing CART expression by the hypothalamus (Table 3; Fungfuang et al., 2013;Stelmanska and Sucajtys-Szulc, 2014). Estrogen and progesterone however seem to have similar effects on the pancreas; they both appear to induce islet hypertrophy and/or increase pancreatic insulin levels and glucose-stimulated secretion in vivo (Table 3; Costrini and Kalkhoff, 1971;Bailey and Ahmed-Sorour, 1980). Nevertheless, there is some evidence that progesterone may inhibit the PRL-induced proliferation

Hormones
Expression level

In vivo effects References
Estrogen Low Non-pregnant Estrogen receptor knockout ER −/− (ERKO, BERKO or viral-mediated ER suppression mouse): ↓ glucose tolerance; whole body and hepatic insulin sensitivity; insulin-stimulated glucose uptake by skeletal muscle; blood adiponectin, testosterone; sexual behavior ↑ body weight; abnormalities in vascular smooth muscle cells (ion channel function); systolic and diastolic blood pressure; arterial pressure; heart failure; hepatic lipid biosynthesis; adipose tissue mass; blood glucose, insulin, leptin Zhu et al., 2002;Bryzgalova et al., 2006;Ribas et al., 2011 Aromatase knockout CYP19-/-(mouse): ↓ glucose and insulin tolerance; glucose oxidation; lean body mass ↑ body weight; adipocyte volume; blood glucose and testosterone Yeh et al., 2002;Takeda et al., 2003 Pregnancy and lactation and insulin secretion of β cells in vitro (Table 4; Progesterone; Sorenson et al., 1993). Furthermore, in rodent models of type 1 and 2 diabetes mellitus, estrogen supplementation protects pancreatic β-cells from oxidative stress, lipotoxicity and apoptosis (Table 3; Estrogen; Tiano and Mauvais-Jarvis, 2012). Therefore, both estrogen and progesterone play roles in regulating insulin and glucose homeostasis, lipid handling and appetite regulation, which may be important in promoting metabolic changes in the mother during pregnancy. High Exogenous (rat islets): ↔ PRL-induced β cell proliferation or insulin secretion Sorenson et al., 1993 Exogenous (human umbilical venous endothelial cells): ↓ folic acid induced anti-angiogenic action ↑ leptin expression; differentiation Lee et al., 2017 Exogenous (human, mouse, rat vascular smooth muscle cells): ↓ proliferation (↑ cell cycle arrest at G1) ↑ eNOS activity Haynes et al., 2000;Hisamoto et al., 2001;Takahashi et al., 2003 Exogenous ( Work conducted both in vitro and in vivo indicate that estrogen and progesterone may also facilitate some of the cardiovascular changes that accompany pregnancy (Tables 3,  4). Estrogen attenuates the vasoconstrictor responses of blood vessels, impairs vascular smooth muscle cell proliferation and calcium influx, and increases vasodilatory nitric oxide synthase activity in vitro (Table 4; Estrogen; Takahashi et al., 2003). It also increases uterine artery angiogenesis and amplifies the vasodilatory impact of vascular endothelial growth factor on isolated rat uterine vessels (Storment et al., 2000;Jobe et al., 2010). In non-pregnant mice, deficiency of the ERβ gene leads to defects in vascular smooth muscle function, hypertension and signs of heart failure ( Table 4; Estrogen; Zhu et al., 2002;Fliegner et al., 2010). Conversely, estrogen supplementation appears to protect the heart and vasculature from pressure overload or vessel injury (Zhang et al., 1999;Zhu et al., 2002;Fliegner et al., 2010). Progesterone also exerts cardiovascular effects. It stimulates nitric oxide synthesis by human umbilical vein endothelial cells in vitro and by rat abdominal aorta and mesenteric arteries in vivo (Tables 3, 4; Progesterone; Chataigneau et al., 2004;Simoncini et al., 2004). It also decreases blood pressure, when infused into ovariectomised ewes and protects against vascular injury in non-pregnant mice (Pecins-Thompson and Keller-Wood, 1997;Zhang et al., 1999). In culture, progesterone induces hypertrophy and inhibits apoptosis of rodent cardiomyocytes (Morrissy et al., 2010;Chung et al., 2012). Thus, via its impacts on cardiomyocytes, progesterone may mediate the pregnancyinduced growth of the mother's heart in vivo. In late pregnancy, the murine heart shifts to use fatty acids, rather than glucose and lactate, as a metabolic fuel. In part, this metabolic shift is proposed to be mediated by progesterone during pregnancy, which inhibits pyruvate dehydrogenase activity in ventricular myocytes . Thus, placental-derived progesterone and estrogen may mediate part of the changes in the maternal cardiovascular system during pregnancy.
In many mammalian species, progesterone levels decline just before parturition and this is associated with the initiation of labor. Indeed, in rodents, inhibition of progesterone synthesis or administration of a progesterone antagonist results in premature delivery of the neonate (Table 3; Progesterone; Fang et al., 1997;Kota et al., 2013). In humans, circulating progesterone levels continue to be high until birth. Commencement of labor is therefore proposed to be related to a functional withdrawal of progesterone activity in the myometrium of women (Brown A. G. et al., 2004;Norwitz and Caughey, 2011). In experimental animals, progesterone reduces the production of prostaglandins and decreases the expression of contraction-associated genes including oxytocin and prostaglandin receptors, gap junction proteins and ion channels in the myometrium (Table 3; Progesterone; Fang et al., 1997;Soloff et al., 2011;Edey et al., 2018). Together, these progesterone-mediated actions decrease contractility of uterine smooth muscle cells and maintain uterine quiescence until term. In contrast to progesterone, estrogen levels rise prior to term and estrogen promotes the expression of contraction-associated genes and contraction of the myometrium ( Table 4; Estrogen; Nathanielsz et al., 1998;Di et al., 2001;Chandran et al., 2014). Therefore, in many species, the high ratio of estrogen to progesterone in the maternal circulation is thought to contribute the onset of labor. Parturition is associated with an influx of inflammatory cells and release of pro-inflammatory cytokines, including interleukin (IL)-1β and tumor necrosis factor (TNF)-α, in the myometrium, cervix and fetal membranes (Golightly et al., 2011). In mice, progesterone reduces the expression of pro-inflammatory cytokines, including IL-1β and IL-6 by the uterus and trophoblast and may modulate the abundance of myometrial monocytes (Table 3; Estrogen; Edey et al., 2018). Progesterone also decreases the ability of LPS to induce pro-inflammatory cytokine secretion by human myometrium and placental explants (Youssef et al., 2009;Garcia-Ruíz et al., 2015). It also diminishes the ability of estrogen to induce the infiltration of macrophages and neutrophils into the uterus, and decreases LPS-induced leukocyte adhesion to human umbilical vein cells (Simoncini et al., 2004). Thus, it is perhaps not surprising that progesterone receptor null mice demonstrate chronic uterine inflammation, particularly in response to estrogen treatment (Table 3; Estrogen; Lydon et al., 1995). There is also evidence that placental steroids participate in cervical softening, by regulating the expression of matrix remodeling enzymes as well as leukocyte infiltration and function (Chinnathambi et al., 2014;Gopalakrishnan et al., 2016;Berkane et al., 2017). In addition to regulating the events leading to parturition, recent data suggest that during the course of pregnancy, both estrogen and progesterone contribute to the maternal tolerance of the fetus by modulating proliferation and cytokine expression of CD4 and CD8 T cells and enhancing the suppressive function of T-regulatory cells (Mao et al., 2010;Robinson and Klein, 2012;Lissauer et al., 2015).
Additionally, both estrogen and progesterone are key stimulators of mammary gland development. For instance, progesterone stimulates proliferation of mammary stem cells and mammary epithelium (Tables 3, 4; Progesterone; Joshi et al., 2010;. In mice, deficiency of the progesterone receptor restricts mammary gland development, whereas exogenous progesterone induces ductal side branching and lobuloalveolar differentiation and development (Table 3; Progesterone; Plaut et al., 1999;Joshi et al., 2010). In addition, both estrogen and progesterone may have indirect effects on mammary gland development by regulating prolactin secretion from the pituitary gland (Rezaei et al., 2016).
Maternal behavior during and after birth are regulated by the steroid hormones. Estrogen stimulates maternal nurturing behavior in numerous species, including rats, mice, sheep and primates (Bridges, 2015). In particular, maternal care is induced by estrogen treatment, whereas the converse happens when ERα expression is suppressed; deficiency of ERα increases the latency to pup retrieval and reduces the length of time dams spend nursing and licking their pups (Table 3; Estrogen; Ribeiro et al., 2012). Findings from animal models suggest that progesterone plays a role in regulating anxiety and depression-related behavior. For instance, exogenous progesterone stimulates anti-anxiety and anti-depressive actions in mouse dams ( Table 3; Progesterone; Koonce and Frye, 2013). In contrast, progesterone withdrawal increases these types of behaviors (Gulinello et al., 2002). Thus, placental-derived steroids may modulate several aspects of maternal physiology which are beneficial to both pregnancy and post-partum support of the offspring.

Neuroactive Hormones
One major target of placental hormones is the maternal brain and related neuroendocrine organs such as the hypothalamus and pituitary glands. These neuroendocrine effects enable the mother to respond and adapt accordingly to her environment, so as to mitigate the adverse effects of stress and maintain homeostasis (Voltolini and Petraglia, 2014). Neuroactive hormones also prepare and enable the future mother to adequately care for her young (Lévy, 2016). In addition to their impact on the maternal neuroendocrine system, these hormones have additional functions in vivo and in vitro functions as well, which are detailed in Tables 5, 6, respectively.

Melatonin and Serotonin
Melatonin and its precursor, serotonin, are tryptophan-derived hormones with well-known neuroendocrine impacts. In humans, circulating concentrations of melatonin and serotonin increase as pregnancy advances (Lin et al., 1996;Nakamura et al., 2001). In the non-pregnant state, melatonin and serotonin are primarily produced by the pineal gland and the brain, respectively. However, the enzymes involved in melatonin and serotonin biosynthesis are also expressed by the human placenta throughout gestation (Iwasaki et al., 2005;Soliman et al., 2015;Laurent et al., 2017). The mouse placenta similarly expresses the enzymes needed for serotonin synthesis , although work is required to assess if melatonin synthesizing enzymes are also expressed. The rat placenta does not produce melatonin de novo due to the lack of synthesizing enzymes (Tamura et al., 2008). However, the same study demonstrated that conditioned medium from cultured term rat placentas stimulated melatonin release by the maternal pineal gland (Tamura et al., 2008). These findings suggest that placental-derived factors may indirectly regulate melatonin levels by the mother during pregnancy. Placental expression of melatonin, serotonin and their respective enzymes, also remains to be investigated in other species such as rabbits and sheep, which are commonly used in pregnancy-related

Nonpregnant
Exogenous (rat): ↓ food and fluid intake; blood pressure; blood calcium ↑ latency to first meal post-fast; bone formation Arletti et al., 1989Arletti et al., , 1990Petersson et al., 1996;Elabd et al., 2007 Exogenous (diet-induced obese rats): ↓ weight gain ↑ glucose and insulin tolerance; adipose tissue lipolysis and fatty acid β-oxidation  Comai et al., 2015). Serotonin is thus a major regulator of maternal mood and behavior (Angoa-Pérez and Kuhn, 2015). For instance, genetically-induced serotonin deficiency leads to increased maternal aggression, lower pup retrieval and greater pup cannibalization, which reduces postnatal survival of offspring in mice (Angoa-Pérez et al., 2014). There is some evidence that serotonin and melatonin may also impact maternal feeding behavior. For example, increased serotonin signaling reduces food intake in pregnant cows (Laporta et al., 2015;Weaver et al., 2016Weaver et al., , 2017Hernández-Castellano et al., 2017). Similarly, exogenous melatonin lowers food intake in pregnant rats (Nir and Hirschmann, 1980;Jahnke et al., 1999;Singh et al., 2013). These negative effects on maternal food intake suggest that peak serotonin and melatonin concentrations in late pregnancy may serve to control the maternal appetite and prevent excessive weight gain.
Another key function of melatonin and serotonin is glucose homeostasis and the regulation of steroid synthesis ( Table 5; Serotonin and Melatonin). In mice, loss of melatonin or serotonin signaling leads to glucose intolerance and insulin resistance, with consequences for blood glucose and insulin concentrations in both the non-pregnant and pregnant state (Contreras-Alcantara et al., 2010;Kim et al., 2010;Owino et al., 2016). However, these neuroactive hormones appear to have differential effects on the pancreas ( Table 6; Serotonin and Melatonin). Serotonin promotes pancreatic β-cell proliferation in vitro (Kim et al., 2010), and is thus important for pancreatic β-cell mass expansion during pregnancy in mice (Goyvaerts et al., 2016). In contrast, melatonin reduces insulin release by rodent pancreatic islets in vitro (Mühlbauer et al., 2012). Non-pregnant mice with deficient serotonin signaling have impaired lipid handling and excessive lipid accumulation in association with reduced adipose aromatase expression and circulating estrogen (Zha et al., 2017). Similarly, treating placental-derived trophoblast cells with norfluoxetine, a selective serotonin-reuptake inhibitor, inhibits aromatase activity and estrogen secretion in vitro (Hudon Thibeault et al., 2017). Supplementation of melatonin in non-pregnant humans reduces circulating triglycerides and cholesterol levels, but effects of lipid handling in pregnancy are unknown (Mohammadi-Sartang et al., 2017). Melatonin also modulates steroid production. For instance, melatonin treatment in pregnant cows reduces circulating estrogen and progesterone (Brockus et al., 2016), while lack of melatonin signaling raises blood corticosterone in mice (Comai et al., 2015).
Given melatonin's additional effects on regulating the circadian rhythm (Mühlbauer et al., 2009), there is some weak evidence for its role in the timing of parturition (Yellon and Longo, 1988;González-Candia et al., 2016). Melatonin can either enhance or reduce uterine myometrial contractility depending on the species (   Exogenous (mouse osteoblast and osteoclast): ↑ proliferation; differentiation Tamma et al., 2009Tamma et al., 2009Tamma et al., , 2010. Both melatonin and serotonin are also important for lactation, specifically for mammary gland development and milk nutrient content (Okatani et al., 2001;Xiang et al., 2012;Laporta et al., 2014a,b). For instance, mammary gland proliferation and calcium transport is impaired in pregnant mice with genetically-induced serotonin deficiency (Laporta et al., 2014a,b).
Conversely, supplementation of a serotonin precursor increases mammary calcium transporter expression and milk calcium content in lactating mice and cows (Laporta et al., 2013a(Laporta et al., ,b, 2015Weaver et al., 2016Weaver et al., , 2017Hernández-Castellano et al., 2017). In contrast to serotonin, increased melatonin signaling is associated with reduced ductal growth and branching, as well as impaired terminal end bud formation in the non-pregnant state (Xiang et al., 2012). Thus, during lactation, these mice with increased melatonin signaling have impaired mammary gland lobuloalveolar development and reduced milk protein content, which reduces the weight of suckling pups (Xiang et al., 2012). Indeed, a recent study showed antenatal melatonin supplementation further exacerbated the growth restriction of offspring and raised circulating maternal cortisol in a sheep model of fetal growth restriction (González-Candia et al., 2016). Nevertheless, melatonin supplementation during pregnancy confers significant beneficial neuroprotective effects on the fetus and enhances maternal antioxidant capacity (Miller et al., 2014;González-Candia et al., 2016;Castillo-Melendez et al., 2017). Therefore, while melatonin supplementation shows promise for use in the clinic, particularly for enhancing the neurodevelopmental outcomes of offspring in growth compromised pregnancies, the potential adverse outcomes for both mother and child must also be considered and should be assessed in further studies.

Oxytocin
Another key neuroendocrine factor is oxytocin. Oxytocin is widely known for its role in triggering maternal nursing behavior (Bosch and Neumann, 2012). This is mediated by oxytocin's actions on the maternal brain, as well as, the mammary glands. Indeed, a greater rise in circulating oxytocin concentrations from early to late pregnancy in pregnant women, is associated with a stronger bond between a mother and her infant (Levine et al., 2007). Concurrently, placental expression of oxytocin also peaks at term in humans . The rat placenta also produces oxytocin (Lefebvre et al., 1992), while placental expression in other species remains unclear. Reduced oxytocin signaling decreases maternal nurturing behavior such as pup retrieval in rats (Van Leengoed et al., 1987). It also decreases the willingness of female voles to care for, groom and lick unrelated pups (Keebaugh et al., 2015). Low oxytocin signaling can additionally impair social bonding in voles and mice (Ferguson et al., 2000;Takayanagi et al., 2005;Lee et al., 2008;Keebaugh et al., 2015), while high levels builds trust and cooperation in a group setting to facilitate group survival in humans (Declerck et al., 2010;De Dreu et al., 2010). Moreover, a lack of oxytocin disrupts mammary gland proliferation and lobuloalveolar development, which impairs milk release from the mammary tissues in mice (Nishimori et al., 1996;Wagner et al., 1997). Therefore, high oxytocin levels enable the mother to bond better and protect her newborn, when it is most vulnerable.
Oxytocin is also important in the process of parturition ( Table 6; Oxytocin); it stimulates the contraction of smooth muscle cells in the myometrium (Ayar et al., 2001;Arrowsmith and Wray, 2014), by inducing calcium influx and stimulating prostaglandin release (Wilson et al., 1988;Voltolini and Petraglia, 2014;. Cardiovascular effects of oxytocin include its ability to significantly lower blood pressure in nonpregnant rats (Petersson et al., 1996). There is also some evidence that oxytocin induces anti-inflammatory and antioxidant effects in the heart under hypoxic conditions in non-pregnant rats (Gutkowska and Jankowski, 2012). Nevertheless, the specific cardiovascular effects of oxytocin in pregnancy remain to be explored.
Studies performed in non-pregnant rodents show that oxytocin also affects metabolic function in vivo (Table 5; Oxytocin). In particular, loss of oxytocin reduces glucose and insulin tolerance and increases adiposity (Camerino, 2009), whereas exogenous oxytocin has the reverse effect (Deblon et al., 2011). Studies are however, required to determine whether the rise in oxytocin in late pregnancy (Levine et al., 2007) may serve to improve insulin sensitivity in the mother in preparation for the metabolic requirements of delivery and lactation. There is some evidence that oxytocin may additionally play a role in controlling energy expenditure and thermoregulation during pregnancy. Even with a similar diet and activity level to control mice, oxytocin-deficient mice become obese due to reduced energy expenditure from poor thermoregulation in the nonpregnant state (Chaves et al., 2013). Furthermore, exogenous oxytocin in non-pregnant mice causes a rise in body temperature (Mason et al., 1986;Tamma et al., 2009). Nevertheless, whether oxytocin may play a role in controlling heat dissipation due to the increased maternal energy expenditure during pregnancy requires exploration. Exogenous oxytocin also reduces food intake in non-pregnant rats (Arletti et al., 1989(Arletti et al., , 1990. However, the role of oxytocin in appetite regulation during pregnancy remains to be explored. There is also evidence for oxytocin's possible involvement in maternal bone metabolism and calcium homeostasis during pregnancy and lactation. For instance, oxytocin stimulates both bone resorption and bone formation by osteoclasts and osteoblasts respectively in vitro (Tamma et al., 2009). Moreover, oxytocin administration in rats reduces circulating calcium with an overall skew toward bone formation (Elabd et al., 2007). These findings may suggest that the peak in circulating oxytocin toward term promote the restoration of depleted maternal skeletal calcium stores.

Other Neuroactive Hormones
In addition to the aforementioned melatonin, serotonin and oxytocin, the human placenta also produces neuroactive hormones such as kisspeptin and thyrotropin-releasing hormone (TRH), which may function in adapting maternal physiology to support pregnancy (Bajoria and Babawale, 1998;De Pedro et al., 2015). In humans, circulating kisspeptin rises throughout pregnancy to concentrations 10,000-fold that of the nonpregnant state, with the placenta speculated as a major source (Horikoshi et al., 2003). In the non-pregnant state, kisspeptin can both stimulate and impede glucose stimulated insulin secretion in mice (Bowe et al., 2009;Song et al., 2014). The nature of the effect may partly relate to differences in the actions of kisspeptin isoforms on pancreatic islets (Bowe et al., 2012). Kisspeptin may also have effects on the maternal cardiovascular system, given its reported vasoconstrictive effects on vascular smooth muscle cells and fibrotic effects on the heart in non-pregnant rats (Mead et al., 2007;Zhang et al., 2017). Studies in humans highlight the importance of regulating kisspeptin production during gestation; increased placental kisspeptin is associated with pre-eclampsia (Whitehead et al., 2013;Matjila et al., 2016) and reduced circulating kisspeptin is observed in women with hypertension and diabetes during pregnancy (Cetković et al., 2012;Matjila et al., 2016). Like the human, the murine placenta produces kisspeptin. Although a kisspeptin-deficient mouse has been established, previous work has been focused on fetoplacental outcomes, with no examination of maternal physiology (Herreboudt et al., 2015). Studies are required to determine the consequences of abnormal placental kisspeptin on the maternal physiology during pregnancy.
In the non-pregnant state, hypothalamic TRH stimulates release of thyroid-stimulating hormone and PRL from the pituitary (Hershman et al., 1973;Vale et al., 1973;Askew and Ramsden, 1984). However, during pregnancy, the placenta serves as an additional source of TRH (Bajoria and Babawale, 1998). Excess TRH in pregnancy raises blood concentrations of thyroidstimulating hormone and PRL in humans, rhesus monkeys, sheep and rats (Thomas et al., 1975;Azukizawa et al., 1976;Roti et al., 1981;Moya et al., 1986;Lu et al., 1998). Conversely, a lack of TRH reduces blood PRL in mice (Rabeler et al., 2004;Yamada et al., 2006). Thyroid hormones are necessary for optimal brain development as well as thyroid function (Miranda and Sousa, 2018). Impaired TRH signaling is associated with anxietylike and depressive-like behavior in non-pregnant mice (Zeng et al., 2007;Sun et al., 2009) and there is some evidence which suggests a link between thyroid dysfunction and poor maternal mood during pregnancy in humans (Basraon and Costantine, 2011). However, whether any direct causal relationship between placental hormones, like TRH and perinatal depression remains unclear. Additionally, TRH is implicated in glucose homeostasis and appetite regulation. For example, mice with TRH deficiency are hyperglycaemic, due to an impaired insulin response to glucose (Yamada et al., 1997). Reduced TRH signaling also impedes leptin production and ghrelin acylation, which results in less energy conservation during fasting and a lower body mass in the non-pregnant state (Groba et al., 2013;Mayerl et al., 2015). Investigations are warranted to identify whether TRH may contribute to the regulation of glucose handling and appetite in the mother during pregnancy.

Additional Hormones
The placenta also produces numerous other hormones with pleiotropic effects. Several key ones, which have been implicated in pregnancy failure or disorders of pregnancy such as hypertension, hyperglycemia and hypercalcemia, are discussed here. The hormones presented here are by no means exhaustive and were selected primarily on their major associations with abnormal maternal physiology during pregnancy. The gonadotropin, chorionic gonadotropin (CG); transforming growth factor β (TGF β) family member, activin; angiogenic factor, relaxin; bone metabolism-associated parathyroid hormone-related protein (PTHrP) and energy homeostasis regulator, leptin are reviewed (Tables 7, 8).

Chorionic Gonadotropin (CG)
CG, is secreted by the human (hCG) and equine (eCG) placenta, although hCG has been more extensively studied. hCG is a large glycoprotein composed of α and β subunits, of which the α subunit identical to luteinizing hormone (LH), follicle stimulating hormone (FSH) and thyroid stimulating hormone (TSH). As a result, hCG can interact with LH, FSH and TSH receptors. In women, hCG is secreted from the trophoblast from very early in gestation and is thought to be the first placental hormone to act on the mother (Ogueh et al., 2011). Indeed, maternal circulating hCG concentrations peak in the first trimester and then decline toward term (Ogueh et al., 2011). In early pregnancy, hCG maintains corpus luteum allowing the continued secretion of ovarian progesterone and estrogens until the steroidogenic activity of the fetal-placental unit can compensate for maternal ovarian function (Fournier et al., 2015). In particular, hCG increases the abundance of lowdensity lipoprotein receptor and thus uptake of cholesterol for steroidogenesis. It also enhances the expression and/or activity of steroidogenic enzymes including 3β-hydroxysteroid and aromatase. There is also some evidence which suggests hCG may inhibit factors that promote luteal demise, such as the prostaglandins. The high levels of hCG in early pregnancy are also sufficient to bind to the TSH receptor and may act to increase maternal thyroid hormone production, which as mentioned previously, may exert effects in the mother and fetus.
CG may also play important autocrine and paracrine roles at the maternal-fetal interface. Administration of hCG antisera prevents implantation in marmoset in vivo (Hearn et al., 1988). Recent proteomic analysis of estrogen and hCG treated human endometrial epithelial cells demonstrates that hCG targets pathways involved in metabolism, basement membrane and cell connectivity, proliferation and differentiation, cellular adhesion, extracellular-matrix organization, developmental growth, growth factor regulation and cell signaling (Greening et al., 2016). Such pathways are likely to be important for placental development, as attenuating hCG signaling disrupts trophoblast differentiation in vitro (Shi et al., 1993). In contrast, supplementing human trophoblast cells with hCG increases their differentiation, migration, invasion and adhesion to uterine epithelial cells, and decreases their leptin secretion in vitro (Table 8; hCG; Shi et al., 1993;Prast et al., 2008;Lee C. L. et al., 2013;Chen et al., 2015). hCG also promotes angiogenic vascular endothelial growth factor secretion by both trophoblast and endometrial epithelial cells (Islami et al., 2003a;Berndt et al., 2006) and enhances endothelial tube formation and migration (Zygmunt et al., 2002). Furthermore, hCG is key in suppressing the maternal immune system from mounting a response against paternal antigens carried by the allogenic conceptus. Administration of hCG in a mouse model of spontaneous abortion significantly reduces the number of fetal resorptions due to improved immune tolerance of the fetus (Schumacher et al., 2013). In vitro, hCG enhances proliferation of immunosuppressive uterine natural killer cells (Kane et al., 2009), and the production of immunosuppressing  Karaplis et al., 1994 Pregnancy and lactation Infusion of PTH/PTHrP receptor antagonist or antibody against PTHrP: ↓ decidual apotosis ↑ decidualization; uterine weight Vanhouten et al., 2003 Mammary-specific PTHrP deletion (mouse): ↓ blood and milk PTHrP; blood vitamin D; urinary cAMP; bone turnover; lactation-associated bone loss ↑ bone mass Williams et al., 1998 Bone-specific PTHrP deletion (mouse): ↑ skeletal fragility Kirby et al., 2011 High
IL-10 by B cells (Fettke et al., 2016). hCG can also modulate the immune system even in a non-pregnant state, as shown by its efficacy in preventing the development of autoimmune diabetes in a mouse model (Khil et al., 2007). In pregnancy, hCG additionally inhibits the contractile function of smooth muscle cells in the uterus to help sustain myometrial quiescence Eta et al., 1994), so as to prevent premature expulsion of the fetus. Glycosylation of hCG affects its biological activity and half-life (Fournier et al., 2015). Given its involvement with multiple systems, it is perhaps unsurprising that abnormal concentrations of hCG and hCG glycoforms have been linked with pregnancy complications such as fetal growth restriction and preeclampsia (Chen et al., 2012). However, whether the abnormal concentrations of hCG are cause or consequence of the disorders remains to be determined.

Activins
Activins are members of the TGFβ family and were first discovered for their role in stimulating FSH production and determining estrus cyclicity and fertility in mice (Ahn et al., 2004;Sandoval-Guzmán et al., 2012). Activin signaling promotes the decidualization, as well as, apoptosis of endometrial stroma cells (Table 8; Activins; Tessier et al., 2003;Clementi et al., 2013;Yong et al., 2017); processes that accommodate implantation and conceptus development (Peng et al., 2015). Additionally, activin A enhances steroid production, invasion and apoptosis of human trophoblast in vitro (Ni et al., 2000;Yu et al., 2012;Li et al., 2015). However, activins may also be of importance in modulating the physiology of the mother during pregnancy ( Table 7; Activins). In normal human pregnancy, activin A concentrations gradually rise during gestation and peak at term (Fowler et al., 1998). The placenta is thought to be the main source of activin A in the maternal circulation during pregnancy, given the rapid clearance after delivery of the placenta (Muttukrishna et al., 1997;Fowler et al., 1998). A similar rise of activin in the maternal circulation is observed in pregnant ewes (Jenkin et al., 2001), while the circulating profiles in other species remain undetermined. Nevertheless, in mice, impaired activin signaling leads to poor pregnancy outcomes such as fewer viable pups (Clementi et al., 2013;Peng et al., 2015). However, there is evidence that an increase in activin may also be pathological and detrimental to pregnancy outcome. For instance in pregnant mice, infusion of activin A or plasmid overexpression of activin A results in the development of a preeclamptic phenotype; dams display hypertension and proteinuria, in addition to growth restriction and greater in utero deaths (Kim et al., 2008;Lim et al., 2015). The maternal hypertension observed likely results from pathological concentrations of activin A inducing vascular endothelial dysfunction (Yong et al., 2015). In the non-pregnant state, activins are also important for renal glomeruli development (Maeshima et al., 2000), as well as, for bone, fat and muscle metabolism (Yogosawa et al., 2013;Ding et al., 2017;Goh et al., 2017). The possible contributions of activin to these latter functions in pregnancy are currently unclear. Therefore, the impact of activin signaling on these other body systems during pregnancy remains to be determined.

Relaxin
Relaxin is a potent vasodilator (Danielson et al., 1999), and regulates hemodynamics in both the non-pregnant and pregnant state ( Table 7; Relaxin; Conrad et al., 2004). In pregnant women, circulating relaxin concentration peaks in the first trimester, declines in the second trimester and is maintained until delivery in the third trimester (Quagliarello et al., 1979;Seki et al., 1985). In contrast, circulating relaxin peaks toward term in mice, rats, guinea pigs and hamsters Renegar and Owens, 2002). In pregnant mice, relaxin deficiency leads to proteinuria, suggesting a particular role of relaxin in modulating renal function during pregnancy (O'sullivan et al., 2017). In addition, relaxin-deficient mice remain sensitive to vasoconstrictors such as angiotensin and endothelin, and are hypertensive during pregnancy (Marshall et al., 2016a;Mirabito Colafella et al., 2017). During pregnancy, relaxin-deficient mice also display stiffer uterine vessels and fetal growth is retarded (Gooi et al., 2013). Relaxin also enhances capillarisation and glucose uptake of skeletal muscles in nonpregnant mice (Bonner et al., 2013). Taken together, these data highlight the importance of relaxin in mediating changes in

Hormones Expression level
In vitro effects References

Parathyroid Hormone-Related Protein (PTHrP)
During pregnancy, the placenta serves as an additional source of PTHrP Emly et al., 1994), a key hormone involved in bone metabolism ( Table 7; PTHrP). PTHrP concentrations in the maternal blood rise throughout gestation in humans (Gallacher et al., 1994;Ardawi et al., 1997;Hirota et al., 1997) and correlate with the rise in maternal circulating calcium during pregnancy (Bertelloni et al., 1994). However, excessively high circulating PTHrP can lead to hypercalcaemia during pregnancy (Winter and Appelman-Dijkstra, 2017). PTHrP increases maternal bone resorption, thereby enabling calcium transfer from mother to fetus for bone development (Salles, 2016). Thus, it is perhaps not surprising that complete knockout of PTHrP in mice is lethal at birth in association with abnormal bone development (Karaplis et al., 1994). Carrying one defective PTHrP copy is enough to also impede bone development and reduce snout length in mice (Amizuka et al., 1996). Mammary-specific PTHrP deletion increases maternal bone mass and protects against lactation-associated bone loss by reducing bone turnover in mice (Williams et al., 1998;Vanhouten et al., 2003). However, deleting bone-specific PTHrP increases skeletal fragility, both in the non-pregnant and pregnant state (Kirby et al., 2011). PTHrP infusion of lactating goats increases mammary gland uptake calcium, phosphorous and magnesium for transfer in milk to the neonate (Barlet et al., 1992). These findings imply that a fine balance of PTHrP production by gestational and maternal tissues must be achieved for appropriate regulation of maternal bone metabolism and offspring calcium requirements during pregnancy and lactation.
Placental-derived PTHrP may also exert additional effects on the placenta and the mother which are beneficial for offspring development and growth. PTHrP stimulates the proliferation, differentiation, outgrowth and calcium uptake of trophoblast in vitro (Table 8; PTHrP; Hershberger and Tuan, 1998;El-Hashash and Kimber, 2006). In vivo, blocking PTHrP signaling during mouse pregnancy leads to excessive uterine growth and decidualization in association with a decrease in decidual cell apoptosis (Williams et al., 1998;Vanhouten et al., 2003). Moreover, over-expression of PTHrP impairs mammary gland branching morphogenesis (Wysolmerski et al., 1995;Dunbar et al., 2001). These studies highlight a possible important regulatory role of PTHrP in the control of decidualization and mammary gland development in vivo. In non-pregnant mice, PTHrP enhances pancreatic β-cells proliferation and insulin secretion whilst it inhibits islet cell apoptosis (Vasavada et al., 1996;Porter et al., 1998;Cebrian et al., 2002;Fujinaka et al., 2004). It also increases renal plasma flow and glomerular filtration rate, and exerts proliferative effects on renal glomerular and tubule cells in rodents (Izquierdo et al., 2006;Romero et al., 2010). Additionally, in vitro studies show PTHrP can induce relaxation of uterine arteries (Meziani et al., 2005). However, the significance of PTHrP on glucose-insulin dynamics and renal and vascular function of the mother during pregnancy remains to be investigated.

Leptin
Leptin is an abundant circulating hormone involved in regulating appetite. In the non-pregnant state, the adipose tissue is the exclusive source of circulating leptin. During pregnancy in humans, baboons and mice, concentrations of leptin rapidly rise throughout gestation, peaking toward term (Highman et al., 1998;Henson et al., 1999;Malik et al., 2005). The rise in leptin positively correlates with increases in maternal body fat (Highman et al., 1998). In humans, blood leptin rapidly falls to non-pregnant concentrations within 24 h of delivery, indicating that the placenta contributes to the main rise of leptin in pregnancy (Masuzaki et al., 1997). In particular, leptin is produced by the human placental trophoblast cells (Masuzaki et al., 1997). A similar post-pregnancy decline and placental trophoblast expression is seen in baboons (Henson et al., 1999). However, this is not the case for mice, as the murine placenta does not produce leptin (Malik et al., 2005). Nevertheless, leptin studies in mice still provide useful knowledge about pregnancy-related effects of leptin (Table 7; Leptin). For instance, leptin in pregnancy helps prepare the mother for lactation, as a deficiency results in impaired mammary gland development, which is detrimental for lactation post-delivery (Mounzih et al., 1998;Malik et al., 2001). Another significant effect of leptin in pregnancy observed through mouse studies is leptin resistance, whereby the dam increases her food intake in mid-pregnancy to meet increased energy demands despite an increase in circulating leptin, which in the non-pregnant state would lead to satiety (Mounzih et al., 1998). In contrast, excessive leptin significantly decreases maternal food intake and restricts feto-placental growth (Yamashita et al., 2001). Leptin exposure of rat and human islets and cultured insulinoma cells significantly decreases insulin production in vitro, demonstrating that leptin may be directly involved in glucose metabolism (Table 8; Leptin; Kulkarni et al., 1997). Indeed dysfunctional leptin signaling in FIGURE 2 | Summary of expression profiles, interactions and maternal physiological effects of placental-derived hormones. PRL, prolactin; PL, placental lactogen; PLF, proliferins; PRP, proliferin-related proteins; GH, growth hormone; GHRH, growth hormone releasing hormone; IGF1/2, insulin-like growth factor-1/2; E2, estrogen; P4, progesterone; MEL, melatonin; SER, serotonin; KISS, kisspeptin; OXY, Oxytocin; TRH, thyrotropin-releasing hormone; RELAX, relaxin; ACTIV, activin; CG, chorionic gonadotropin; LEP, leptin; PTHrP, parathyroid hormone-related protein.
pregnancy leads to the spontaneous development of a gestational diabetic phenotype in db/+ mice, who are heterozygous for the leptin receptor ( Table 7; Leptin; Yamashita et al., 2001). Further in vitro studies on placental explants or trophoblast cultures highlight a potential for leptin to be involved in immune modulation and placental hormone production, given its stimulatory effects on HLA-G and hCG expression ( Table 8; Leptin; Chardonnens et al., 1999;Islami et al., 2003a,b;Barrientos et al., 2015). Additional effects of leptin on the placenta are thoroughly reviewed elsewhere (Schanton et al., 2018). Therefore, placental leptin can have systemic effects on the mother in pregnancy.

CONCLUSION
Pregnancy represents a unique physiological paradigm; there are dynamic and reversible changes in the function of many organ systems in the mother that are designed to support offspring development. In part, these changes are signaled via the placental secretion of hormones, which in turn, alter in abundance, interact with one another and exert wide effects on maternal tissues during pregnancy. For instance, steroid hormones modulate most systems of the mother throughout pregnancy. However, they also alter the production of other hormones, such as prolactin and placental lactogens, which in turn, may contribute to the physiological changes in the mother (Figure 2). However, further work is required to better define how placental hormones elicit their actions in the mother, as well as, identify the extent to which they interplay with hormones produced by maternal tissues. As the endocrine and metabolic state of the mother is also influenced by her environment, maternal conditions such as poor nutrition and obesity may modulate placental hormone production and pregnancy adaptations. Indeed, previous work has shown that an obesogenic diet during pregnancy alters the expression of PRL/PL genes in the placenta in association with mal-adaptations of maternal metabolism in mice (Musial et al., 2017). Further studies are nonetheless needed to assess the interaction of the maternal environment with placental endocrine function. Placental hormones are also released into the fetal circulation, where they may have direct impacts on fetal growth and development (Freemark, 2010). Investigations exploring the importance of placental endocrine function on fetal growth, independent of the mother, will require future examination. Collectively, further studies on the nature and role of placental endocrine function in maternal adaptations and fetal growth will undoubtedly provide novel insights into understanding of the potential causes of obstetrical syndromes such as gestational diabetes and preeclampsia that are marked by maternal physiological maladaptation.

AUTHOR CONTRIBUTIONS
TN and HY substantially contributed to the conception of the work, drafting and revision of the manuscript, preparation of the tables and approved of the final version. JL-T substantially contributed to the conception of the work, drafting and revision of the manuscript, preparation of the figures and approved of the final version. AS-P substantially contributed to the conception of the work, critical revision of the manuscript for intellectual content and approved of the final version.

ACKNOWLEDGMENTS
TN was supported by the Marie Skłodowska-Curie Individual Fellowship from the European Union; HY was supported by an A * STAR International Fellowship from the Agency for Science, Technology and Research; JL-T was supported by the Newton International Fellowship from the Royal Society; AS-P was supported by the Dorothy Hodgkin Research Fellowship from the Royal Society.