Endocrinology of Human Pregnancy and Fetal-Placental Neuroendocrine Development

Endometrial Receptivity

The ovarian steroid hormones, estradiol, and progesterone, prepare the uterus for pregnancy by inducing structural and functional changes in the endometrium (see Chapter 10 ). The endometrium has two functional layers, the stratum functionalis (closest to the uterine cavity) and the stratum basalis (adjacent to the myometrium). The stratum functionalis is responsive to estradiol and progesterone and forms a proliferative phenotype in response to estradiol or a secretory/decidualized phenotype in response to estradiol and progesterone. The secretory/decidualized stratum functionalis is shed at menstruation, miscarriage, and parturition. In the following cycle, estrogen induces the development of a new stratum functionalis from cells in the stratum basalis. Pregnancy is established in the postovulatory secretory/decidualized endometrium during the luteal phase of the menstrual cycle.

Receptivity of the endometrium to embryo adhesion and implantation has temporal and spatial characteristics. In humans, peak endometrium receptivity occurs between days 21 and 24 of the menstrual cycle. This is referred to as the implantation window. During this time, the probability of successful implantation of a healthy embryo is ∼85%. In contrast, implantation before or after the implantation window has a low probability (∼11%) of establishing a successful pregnancy. Analysis of the human endometrial transcriptome during the implantation window has revealed numerous genes that are up- and downregulated in comparison with late proliferative phase endometrium. ^,

Under the influence primarily of progesterone, the postovulatory endometrium transforms to the secretory phenotype. It thickens and becomes highly vascularized, and the glandular epithelium secretes glycoproteins and other factors to produce an intrauterine environment that favors survival of the floating embryo. The luminal epithelial cells also produce chemokines, growth factors, and cell adhesion molecules (CAMs) that attract the embryo to specific dome-shaped docking sites known as pinopodes. The tips of the pinopodes express chemokines and CAMs that attract the embryo and appear to be the preferred site for embryo adherence. Areas between pinopodes produce a repellent molecule, MUC-1, that prevents embryo adhesion. Embryo adhesion initiates the invasion of trophoblasts across the epithelium. The interaction also triggers the process of decidualization in the underlying endometrial stroma that is necessary for subsequent implantation and placentation.

Decidualization of the Endometrial Stroma

Late in the luteal phase, the endometrial stroma undergoes a morphologic and functional transformation referred to as decidualization. ^, Stromal fibroblasts in the endometrium proliferate and differentiate into large polyhedral epithelioid cells with high levels of glycogen and lipids. The cells also produce a tough pericellular capsule composed of collagen, laminin, fibronectin, and heparan sulfate proteoglycans.

In a nonfertile cycle, decidualization begins around blood vessels during the midsecretory phase and progressively encompasses the entire endometrial stroma during the latter part of the secretory phase. In a fertile cycle, decidualization is also triggered by signals from the luminal epithelium in response to blastocyst adhesion.

Decidualized endometrial stromal cells communicate with adjacent epithelial cells, immune cells, vascular endothelial cells, and trophoblast cells of the invading embryo to produce a microenvironment that is conducive to the establishment and continuation of pregnancy. Decidual cells promote the growth and remodeling of uterine spiral arterioles and regulate the local immune cell ecosystem to produce an immune-privileged tolerogenic site conducive to implantation of the semiallogeneic embryo. The specific cohort of resident immune cells in the decidua also restricts the passage of fetal antigens to the maternal compartment to further protect the conceptus from attack by the maternal immune system. ^, The decidual microenvironment also promotes the development of robust uterine arteriolar vasculature, providing the developing placenta access to the maternal circulation.

Although decidualized stromal cells surround the implanting blastocyst and are essential for successful implantation, the fully decidualized endometrium limits the depth of embryo invasion into the uterine wall and is hostile to de novo implantation. Placenta accreta is a condition characterized by excessive and sometimes life-threatening trophoblast invasion and is thought to be caused by invasion of the embryo into uterine scar tissue (usually from a previous Cesarean section) that lacks decidua.

Decidualization is dependent on progesterone produced by the corpus luteum (CL) and cyclic adenosine monophosphate (cAMP) signaling in endometrial stromal cells. In decidualized endometrial stromal cells, progesterone and cAMP induce production of multiple regulatory factors, especially prolactin, insulin-like growth factor (IGF)-binding protein 1 (IGFBP1), and cytokines such as interleukin (IL)-15 (IL-15) that participate in a complex cross-talk between endometrial epithelial cells and resident immune cells to create a microenvironment conducive for blastocyst implantation. Prolactin and IGFBP1 are thought to promote implantation by stimulating the proliferation and invasive properties of trophoblasts. ^, IL-15 promotes the differentiation of resident immune cells, especially uterine natural killer (uNK) cells, to produce an immune microenvironment that is tolerant of the semiallogeneic conceptus.

In a nonfertile cycle, the decidualized stratum functionalis is shed at menstruation, in response to progesterone withdrawal due to CL regression. In a fertile cycle, chorionic gonadotropin (CG) produced by the embryo prevents CL regression, leading to sustained progesterone support for the decidualized endometrium, which is necessary for the establishment and maintenance of pregnancy.

Implantation

After syngamy, an intrinsic development program is initiated in the early embryo (see Chapter 9 ). The sequence of cell divisions and differentiation is not dependent on the hormonal milieu of the fallopian tube or the uterus. By the third to fourth day after fertilization, the embryo comprises a solid ball of cells encapsulated by the translucent remnant of the zona pellucida. On about the fifth day after fertilization, a fluid-filled cavity, the blastocele, forms within the embryo, which at this stage is referred to as a blastocyst. The outer layer of blastocyst cells adjacent to the zona pellucida is known as the trophectoderm. These cells directly interact with maternal tissue and eventually give rise to the placenta and chorion. The fetus and the amnion and mesenchymal and vascular components of the placenta are derived from a group of cells, known as the inner cell mass, lying under the trophectoderm at one end of the blastocele.

The human blastocyst enters the uterus around the fifth day after fertilization and floats freely for 2 to 3 days. By this stage, the blastocyst escapes the confines of the zona pellucida (i.e., it hatches) and is in an optimal condition to implant into a receptive endometrium.

Almost immediately after fertilization, hormonal signals from the conceptus are transmitted to the mother. Studies of the secreted proteins (the secretome) produced by the preimplantation embryo show that the profile of secreted proteins changes every 24 hours and with different embryonic stages. This suggests that the embryo communicates with the uterine epithelium via a specifically orchestrated sequence of secretions. Any delay in embryo development or its transit through the oviduct increases the likelihood of implantation failure. Given this delicate balance, it is not surprising that the highest rate of pregnancy loss occurs during the periimplantation period.

The embryo develops and implants in a polarized fashion. After hatching from the zona pellucida, trophoblast cells overlying the inner cell mass of the embryonic pole are the first to interact with the uterine epithelium, most likely at the pinopodes ^, ( Fig. 11.1 ). The trophoblast cells proliferate and secrete proteases that degrade the extracellular matrix between endometrial cells, forming a path for the blastocyst to enter the uterine stroma. At this stage, the trophoblast cells are referred to as cytotrophoblasts and differentiate along two distinct paths: villous and extravillous. The extravillous cytotrophoblasts become highly invasive and form columns penetrating the basal membrane beneath the endometrial stroma and into the myometrium. Eventually, the entire embryo is embedded in the uterine stroma and anchored by columns of extravillous cytotrophoblasts. During this time, some villous cytotrophoblasts fuse their plasma membranes to become a single multinucleated cell known as the syncytiotrophoblast, which becomes the outer layer of the placental villi (see below).

As with adhesion, implantation also involves a paracrine dialogue between the embryo and the endometrial stoma. Some of the molecules involved in the paracrinology of implantation include ^, :

• Leukemia inhibitory factor

• Interleukin-11

• Heparin-binding epidermal growth factor-like growth factor

• Prostaglandins, especially prostacyclin and prostaglandin E ₂

• The homeobox transcription factors, especially HOXA10 and HOXA11 ^,

• Metalloproteinases

• The Wnt and Indian hedgehog signaling pathways ^,

Placentation

Early in the implantation process some cytotrophoblast cells aggregate and form migrating columns that penetrate into the inner third of the myometrium. The invading columns of cytotrophoblasts target maternal blood vessels and, via interstitial and endovascular routes, completely surround and occlude spiral arterioles. They then displace maternal endothelial cells and vascular smooth muscle cells to create a low-resistance arteriolar system by increasing vessel diameter. Poiseuille’s law of fluid dynamics dictates that flow through a cylinder is proportional to its radius multiplied by the fourth power. As a consequence of cytotrophoblast invasion, the average radius of human uterine arterioles is around 10-fold greater during pregnancy compared with the nonpregnant state. Such a change increases flow 10,000-fold. Cytotrophoblasts within the spiral arterioles also prevent maternal vasomotor control of the spiral arterioles. Through this process, the supply of nutrients and oxygen to the conceptus is optimized and the capacity for maternal restriction of uterine blood flow is subverted.

Maternal blood flowing from the dilated spiral arterioles fills spaces, known as lacunae, that form between the invading columns of cytotrophoblast cells. At around the same stage of placental development, finger-like projections of chorionic villi form in the lacunae and become bathed in maternal blood. Chorionic villi have a central core of loose connective tissue with an extensive capillary network linked with fetal circulation. Surrounding the cores are the outer syncytiotrophoblast and inner cytotrophoblast cells that form a barrier between the maternal and fetal circulations. This arrangement is referred to as hemochorial placentation since the maternal blood is in direct contact with the syncytiotrophoblast and extravillous cytotrophoblasts.

Placentation across all eutherian mammals is characterized by high angiogenic activity and blood vessel growth, especially at the site of placental attachment. The cytotrophoblast cells produce several angiogenic factors, including platelet-derived endothelial cell growth factor, vascular endothelial growth factor (VEGF), angiopoietin-1, and angiopoietin-2. ^, In addition, two potent inhibitors of angiogenesis have been isolated from mouse placenta. The presence of antiangiogenic factors during placentation is thought to prevent maternal endothelial cells from resealing the ends of spiral arterioles that have been occupied by cytotrophoblasts. In addition, antiangiogenic factors may prevent the overgrowth of maternal and fetal vessels, thereby preventing maternal blood vessels from entering the fetal compartment and fetal vessels from extending beyond the uterus.

Angiogenic events occurring during implantation and placentation are thought to be critical factors in the etiology of hemodynamic disorders in pregnancy. The extent and depth of the placental incursion appear to be the sum of the intrinsic proinvasive characteristics of cytotrophoblasts and the physical and biochemical barriers mounted by the maternal tissues. Imbalances in this equation can lead to failed implantation and pathological conditions. The arteriole remodeling process is complex and requires the cytotrophoblast cells to express specific adhesion molecules and adopt an endothelial phenotype. If this process is impaired the depth of invasion and remodeling may be limited, leading to placental ischemia ( Fig. 11.2 ). Abnormal spiral artery remodeling is associated with fetal growth restriction and hypertensive disorders of pregnancy. ^, In preeclampsia, cytotrophoblast invasion is restricted to the superficial decidual segments, leaving the myometrial spiral arterioles undisturbed and still responsive to maternal vasomotor control. To compensate, especially during the later stages of pregnancy when nutritional requirements of the fetus increase, the placenta is thought to secrete factors into the maternal circulation that increase maternal systemic blood pressure.

Immune Tolerance of the Conceptus

Because cytotrophoblasts are genetically and immunologically distinct from maternal tissue, implantation represents an extraordinary breach of maternal immune defenses. In 1953, Sir Peter Medawar in his studies of graft rejection proposed 3 mechanisms by which the embryo is protected from maternal immune rejection: (1) anatomical separation between fetal cells and maternal immune cells; (2) modulation of the maternal immune cells at the maternal-fetal interface; and/or (3) global modulation of the maternal immune system to produce refractoriness to allogenic fetal cells.

It is now clear that a physical barrier between maternal immune cells and fetal cells expressing alloantigens does not exist, thus eliminating Medawar’s first mechanism. Medawar’s second and third mechanisms are supported by clinical and animal data. Although maternal lymphocytes with the capacity to attack and eliminate fetal cells exist throughout gestation, ^, they are suppressed during pregnancy. This tolerogenic state is exemplified clinically by reduced severity of autoimmune disease during pregnancy and its relapses after parturition, and experimentally by increased tolerance for engraftment of paternally derived tumor cells in sites outside of the uterus during pregnancy in mice. The tolerogenic state of pregnancy is thought to involve a subset of T cells known as regulatory T (T-reg) cells that promote tolerance to nonself antigens, inhibit inappropriate immune responses against self-antigen, and whose population increases during pregnancy. Interestingly, T-reg cell expansion is synergistically driven by exposure to fetal alloantigen and by the endocrine milieu of pregnancy, especially by progesterone. Progesterone also stimulates Fas-ligand production by decidual cells that induce apoptosis of activated maternal T lymphocytes. Progesterone also stimulates decidual stromal cells to produce cytokines, especially IL-15, that induce uNKs with immunosuppressive activity. ^,

Cytotrophoblast cells avoid detection by maternal immune cells by producing a distinct human leukocyte antigens (HLA) profile involving polymorphic molecules, especially HLA-G, that do not trigger NK cell activation. Cytotrophoblast cells also may inhibit the proliferation of maternal T cells by catabolizing tryptophan, an essential amino acid needed for proliferating T cells. ^,

Chorionic Gonadotropin

One of the first endocrine signals from the embryo to the mother is mediated by a gonadotropin-like glycoprotein hormone, chorionic gonadotropin (CG; also known as hCG in humans) (see Chapter 2 ). CG is produced by trophoblast cells of the early embryo and binds to luteinizing hormone (LH) receptors on CL cells, where it acts as a super-gonadotropin to prolong the longevity and steroidogenic function of the CL. Importantly, CG prevents regression of the CL that would otherwise occur in a nonfertile cycle and as such sustains progesterone synthesis required to prevent menstruation and establish pregnancy. This is referred to clinically as the maternal recognition of pregnancy.

CG is biologically similar to pituitary LH and structurally similar to LH and follicle-stimulating hormone (FSH). It is a heterodimer composed of α and β subunits. The α-subunit of CG, LH, and FSH are identical, whereas the β-subunits differ. The α subunit of CG is produced by cytotrophoblast cells especially as they differentiate during implantation. The β subunit of CG is primarily produced by the syncytiotrophoblast but can also be detected in mature cytotrophoblast cells just before they fuse to form the syncytiotrophoblast. The syncytiotrophoblast produces both subunits and is the principal source of CG. As a potent luteotropin, CG acts via the LH receptor and stimulates cells of the CL to produce and release progesterone.

In normal pregnancies, CG is detectable (by measuring β-CG) 9 to 11 days after the midcycle LH peak, which is around 8 days after ovulation and only 1 day after implantation. This has clinical utility when it is important to determine the presence of pregnancy at an early stage. In early pregnancy, circulating CG levels double every 2 to 3 days, and concentrations of CG rise to peak values by 60 to 90 days of gestation. Thereafter, CG levels decrease to a plateau that is maintained during the remainder of the pregnancy ( Fig. 11.3 ). Assayable LH and FSH levels in the maternal blood are virtually undetectable throughout pregnancy.

Actions of CG may not be limited to maintaining progesterone production by the CL. ^, Much of the increased maternal thyroid activity that occurs in pregnancy has been attributed to CG, which binds specifically to thyroid cell membranes and displaces thyroid-stimulating hormone (TSH). CG also influences the development and function of the fetal adrenals and testes. In addition, CG may have actions on the maternal reproductive tract, including the decidual response, relaxin production by the CL, and relaxation of uterine smooth muscle. ^, LH receptors have been detected in fetal membranes. Extravillous cytotrophoblasts also produce hyperglycosylated CG and monomeric β-subunit that also act via the LH receptor to affect other tissues including some malignancies. ^,

Production of CG by the invading blastocyst may contribute in a paracrine manner to the implantation process. In vitro and in vivo studies in nonhuman primates demonstrate that CG promotes decidualization of the endometrial stroma. The hormone affects endometrial cells at the implantation site before its levels are detectable in the circulation. Studies suggest that CG produced by the blastocyst prolongs the window of implantation by inhibiting endometrial IGFBP1 production, augmenting angiogenesis at the implantation site by increasing VEGF expression, modulating local cytokine and chemokine expression, augmenting local protease activity and, via an autocrine effect on cytotrophoblasts, augmenting the invasive potential of the blastocyst. ^,

Gonadotropin-Releasing Hormone

The human placenta produces GnRH, which is identical to that produced by the hypothalamus. Levels of GnRH in the circulation of pregnant women are highest in the first trimester and correlate closely with CG levels. The close relation between GnRH and CG suggests a role for GnRH in regulating CG production. GnRH stimulates the production of both the α-subunit and β-subunit of CG in placental explants and specific GnRH-binding sites are present in the human placenta. Other regulators of pituitary gonadotropin, such as progesterone, inhibin, and activin, also influence placental CG production (at least in vitro ).

Activins and Inhibins

Activin and inhibin are disulfide-linked homo- and heterodimeric proteins belonging to the transforming growth factor -β (TGF-β) superfamily. Inhibin is a heterodimer composed of an α-subunit and one of two β-subunits, βA or βB. Inhibins (αβA and αβB) inhibit pituitary FSH secretion. In contrast, activins, composed of βA or βB homodimers (βA-βA and βB-βB), stimulate FSH production. Both hormones affect target cell function via specific cell surface receptors. The bioavailability of activin is restricted by follistatin, which binds to activin and prevents it from interacting with its receptor on target cells.

Inhibins and activins are produced by the human placenta. Each of the subunits is expressed in the syncytiotrophoblast and the levels of expression do not change with advancing gestation. Activin-A is also produced by the CL, decidua, and fetal membranes. The placenta also produces follistatin. These factors are secreted into the maternal and fetal circulations and amniotic fluid and their production varies with the stage of gestation.

Although the exact function of the inhibin-activin system in human pregnancy is not known, several studies indicate their involvement, and those of other TGF-β family members, in placental development and function, and in the pathogenesis of adverse pregnancy conditions. Levels of inhibin-A and activin-A in the maternal circulation can be indicative, albeit with relatively weak predictive value, of pathologies such as placental tumors, hypertensive disorders of pregnancy, intrauterine growth restriction, fetal hypoxia, Down syndrome, fetal demise, preterm delivery, and intrauterine growth restriction.

Because inhibin and activin are synthesized by cytotrophoblast cells, they may be involved in autoregulating placental CG production by modulating local GnRH activity. In vitro studies have shown that inhibin decreases GnRH-stimulated CG production by placental cell cultures, whereas antibody blockade of inhibin activity increases GnRH release and causes a parallel rise in CG secretion. Thus, it is possible that inhibin exerts an autocrine effect on placental CG secretion by suppressing GnRH action. In contrast, activin augments the GnRH-induced release of CG in cultured trophoblast cells, an effect that can be reduced by the addition of inhibin. Thus, at least in vitro , activin and inhibin contribute to the regulation of CG secretion in a manner similar to their effect on hypothalamic-pituitary gonadotropin secretion.

Corticotrophin-Releasing Hormone

First identified in the hypothalamus, corticotrophin-releasing hormone (CRH) is a 41-amino acid peptide that stimulates the expression and processing of proopiomelanocortin (POMC) by pituitary corticotropes, and the secretion of adrenocorticotropic hormone (ACTH), a key POMC derivative. The human placenta, fetal membranes, and decidua also produce CRH that is identical to that produced by the hypothalamus. CRH produced by the placenta can be detected from the seventh week of pregnancy and increases progressively until term, rising more than 20-fold in the last 5 to 7 weeks of pregnancy. Placental CRH is released mainly into the maternal compartment. Levels of CRH in the maternal circulation can be detected as early as 15 weeks of gestation and then increase through gestation, reaching maximum levels of 1 to 10 ng/mL at term. Remarkably, this is about 1000-fold higher than peripheral CRH levels in nonpregnant women.

Actions of CRH are mediated by two CRH receptors (CRH-Rs), CRH-R1 and CRH-R2. ^, A CRH binding protein (CRH-BP) also exists, and for most of pregnancy it is present in excess of CRH in the maternal circulation. As CRH-BP binds CRH with greater affinity than the CRH receptor, it is thought to sequester and therefore suppress CRH activity. Thus, for most of the pregnancy, the bulk of the placental CRH is considered to be inactive. However, during the last 4 weeks of pregnancy CRH-BP levels decrease markedly. This coincides with the exponential increase in placental CRH production, which could result in a dramatic increase in CRH bioavailability ( Fig. 11.4 ).

Despite the elevated concentrations of CRH during pregnancy, secretion of ACTH from the maternal pituitary does not increase concordantly. In fact, pituitary ACTH levels remain low throughout pregnancy. The lack of CRH stimulation could be due to inhibition by the CRH-BP. However, maternal ACTH production remains low late in gestation when CRH increases and CRH-BP decreases. In vivo studies have shown that responsiveness of the maternal pituitary to CRH is markedly attenuated during pregnancy, and in vitro studies have shown that CRH downregulates CRH-R expression in pituitary corticotropes.

In vitro studies indicate that agents that increase CRH production by the hypothalamus also increase CRH production by placental cells. These agents include prostaglandins (PGs) E ₂ and F _2α (PGE ₂ and PGF _2α ), norepinephrine, acetylcholine, vasopressin, angiotensin-II, oxytocin (OT), interleukin (IL)-1, and neuropeptide-Y. In contrast, progesterone and nitric oxide donors inhibit placental CRH expression in vitro . ^,

Production of CRH by the placenta is increased by cortisol. This is in contrast to hypothalamic CRH, which is decreased by cortisol via a classic negative feedback loop. This stimulatory action has been observed in vivo in women who receive glucocorticoid treatment during the third trimester, ^, and in vitro in cultured cytotrophoblast cells. The stimulation of placental CRH production by cortisol may result in a positive feedback endocrine loop. Placental CRH may stimulate ACTH production by the fetal pituitary, which would increase cortisol secretion by the fetal adrenals. Fetal adrenal cortisol could then further stimulate placental CRH production. The marked rise in placental CRH during the last 10 weeks of pregnancy could be due to such a positive feedback interaction and this endocrine loop may be involved in the process of parturition (discussed later in the chapter).

CRH also influences fetal adrenal steroidogenesis by directly increasing dehydroepiandrosterone sulfate (DHEA-S) production ^, ( Fig. 11.5 ). The capacity for CRH to act as an adrenal androgen secretagogue also has been demonstrated in vivo in adult men. Placental CRH and fetal adrenal DHEA-S increase concordantly during the third trimester. The placental CRH-fetal adrenal endocrine axis may play a key role in the regulation of human parturition (discussed later in the chapter).

Several actions have been ascribed to placental CRH in the control of human pregnancy. CRH may serve an autocrine-paracrine function within the placenta by regulating the expression and processing of POMC. Placental CRH may be part of the fetal-placental stress response mechanism. The placenta is comparable to the hypothalamus in its production of CRH in response to stress. Neurotransmitters and neuropeptides activated in response to stress stimulate placental CRH release in vitro . ^, The physiological implications of this are that the fetus may mount a stress response via placental CRH. This may be critical in conditions such as preeclampsia, placental vascular insufficiency, and intrauterine infection.

The CRH family of neuropeptides includes a group of structurally and functionally related proteins known as urocortins ^, that have a high affinity for both CRH receptors. The urocortins are expressed by the syncytiotrophoblast and extravillous cytotrophoblasts and by the decidua and fetal membranes. ^, Circulating levels of urocortins are low compared with CRH during pregnancy. Urocortin stimulates ACTH and PG production by trophoblast cells and causes vasodilation of the uteroplacental vasculature via activation CRH-R2. In women with decreased uterine artery blood flow during midgestation, circulating urocortin levels are reduced in proportion to the increase in uterine artery resistance. In vitro studies suggest that this activity is compromised in preeclampsia. Studies in animal models also show vasodilatory effects of urocortin on placental blood flow and suggest that one of its main functions is to protect the fetus from hypoxic insults. ^, Studies of term myometrial cells show that urocortin acting via CRH-R2 increases contractility suggesting that it plays a role in parturition.

Thyrotropin-Releasing Hormone

A substance similar to the hypothalamic thyrotropin-releasing hormone (TRH) has been found in the human placenta. It stimulates the release of pituitary thyroid stimulating hormone (TSH) release in the rat both in vitro and in vivo , but is not identical to hypothalamic TRH. To date, a placental TSH has not been identified. Whether placental TRH plays a role in stimulating fetal or maternal pituitary TSH remains to be ascertained. The thyroid-stimulating activity of the placenta has been ascribed to CG.

Placental Somatotropins

In most eutherian mammals the placenta expresses members of the growth hormone (GH)-PL gene family. The genes are encoded by a 66 kb segment of chromosome 17 that includes five closely related genes: GH1, GH2, CSH1, CSH2, and CSHL1 , each derived from the duplication of a common ancestral gene. GH1 encodes pituitary GH and is expressed only in the pituitary. The other four are expressed exclusively in the placenta. GH2 encodes a placental GH (PGH) variant, which differs from pituitary GH by 13 amino acids. CSH1 and CSH2 are identical and encode PL.

Placental lactogen

Placental lactogen is a single-chain polypeptide of 191 amino acids with 96% homology with GH. It can be detected in the placenta from around day 18 of pregnancy and in the maternal circulation by the third week of pregnancy. Low levels of PL (7–10 ng/mL) are present in the maternal circulation by 20 to 40 days of gestation. Thereafter, PL levels in the maternal circulation increase exponentially, reaching levels of 5 to 10 μg/mL at term. In contrast to the elevated levels of PL in the maternal circulation, concentrations of PL in the fetal circulation range from 4 to 500 ng/mL at midgestation and only 20 to 30 ng/mL at term. Thus, PL is preferentially secreted into the maternal compartment.

In normal pregnancy, PL is first synthesized by the cytotrophoblasts of the developing placenta during the first 6 weeks of pregnancy. Thereafter, expression switches to the syncytiotrophoblast, which eventually becomes the exclusive source of PL. The extent of PL expression by the syncytiotrophoblast does not change during the course of pregnancy, although total placental production increases substantially. Therefore, the rise in PL production is thought to be due to the increase in placental mass. Maternal PL levels rise concordantly with an increased amount of syncytiotrophoblast tissue as gestation advances ( Fig. 11.6 ). After delivery of the placenta, the half-life of the disappearance of circulating PL is 9 to 15 minutes. To maintain circulating concentrations, this would imply placental production of 1 to 4 g of the hormone per day at term. Thus, the production of PL represents one of the major metabolic and biosynthetic activities of the syncytiotrophoblast. PL is expressed by all types of trophoblastic tissue and has even been detected in the urine of patients harboring trophoblastic tumors, in men with choriocarcinoma of the testis, and in the serum and urine of women with molar pregnancies.

Factors that regulate PL production have been assessed in cultured cytotrophoblast cells. Insulin and growth hormone-releasing factor (GHRF) stimulate PL secretion, whereas somatostatin (SS) inhibits its secretion. The presence of PL, PGH, SS, and GHRF in the same cell suggests that another autoregulatory loop analogous to the hypothalamic-pituitary axis operates within the placenta. In the third trimester, maternal PL and GHRF levels are closely correlated. Interestingly, SS expression is maximal in early pregnancy and decreases during the second and third trimesters, a pattern opposite to that of PL. Thus, locally produced GHRF and SS may regulate placental PL production. Several studies have demonstrated changes in maternal PL levels in response to metabolic stress. Specifically, prolonged fasting at midgestation and insulin-induced hypoglycemia raise maternal PL concentrations. However, PL levels do not change in association with normal metabolic fluctuations during a typical 24-hour period.

The initial identification of PL was based on its lactogenic activity in bioassays, suggesting that PL acts as a lactogen in human pregnancy. Indeed, PL binds to the prolactin receptor with relatively high affinity (Kd 0.1 nM). In contrast, its affinity for the GH receptor is lower (Kd 770 nM). Thus, PL may function mainly as a lactogen during pregnancy with only minimal activity as a somatogen (promotes growth). However, administration of PL to nonpregnant women in sufficient quantities to mimic pregnancy levels did not induce lactation. This does not rule out PL as a lactogen, as its actions on the mammary gland may be dependent on other factors in the endocrine milieu of pregnancy (e.g., estrogens and progesterone). Nonetheless, the in vivo lactogenic properties, if any, of PL in human pregnancy remain to be established. It should be noted that maternal prolactin levels increase significantly in the later stages of pregnancy and, together with estrogen and progesterone, are likely sufficient to induce mammary growth and lactation.

Placental Growth Hormone

Two forms of PGH have been identified in the human placenta, both of which are expressed by the syncytiotrophoblast. The smaller, 22 kDa form is almost identical to pituitary GH, differing by only 13 amino acids. The larger 26kDa PGH is a splice variant that retains intron 4. The extent of PGH production is significantly less than that of PL, and PGH is not secreted into the fetal compartment. Levels of PGH in the maternal circulation are approximately 1000-fold less than PL and can be detected only after midgestation (between 21 and 26 weeks). During the third trimester, maternal PGH levels increase exponentially, in concert with PL, and reach a maximum of approximately 20 ng/mL by term. In the first trimester, pituitary GH is measurable and secreted in a highly pulsatile manner. However, pituitary GH production decreases progressively from about week 15, and by 30 weeks cannot be detected. During the same period, nonpulsatile secretion of PGH by the placenta increases markedly and becomes the dominant GH of pregnancy.

Function of PL and PGH

Studies of PL and PGH deficiency have revealed their potential synergistic roles in human pregnancy. In all cases of complete PL deficiency (i.e., no detectable PL in maternal blood or in the placenta) pregnancy and fetal development were normal. However, deficiency of both PL and PGH due to a mutation in the GH/PL gene cluster is associated with severe fetal growth retardation. Pregnancies in which only PGH is deficient have not been identified. These experiments of nature indicate that PL is not necessary for normal pregnancy, whereas PGH is an important regulator of fetal growth. As PGH does not enter the fetal compartment, it likely influences fetal growth via effects on the mother. This is consistent with the thesis that PL and PGH modulate maternal metabolism to meet fetal energy requirements.

Maternal food intake and intestinal calcium absorption increase during the first trimester, and insulin secretion immediately after feeding almost doubles. In the second half of pregnancy, maternal cells become increasingly resistant to insulin, i.e., a diabetogenic state, that is thought to be promoted by the combined GH-like and contra-insulin activity of PGH and PL. This decreases glucose uptake and increases free fatty acid release. The net effect is increased availability of free fatty acids, glucose, and amino acids for fetal consumption. Decreased glucose mobilization into maternal cells would increase the supply of glucose available for the fetus, which is especially important for fetal brain growth and development. Free fatty acids can cross the placenta, and the increased ketones induced by their metabolism are also an important energy source for the fetus. Thus, during the second half of pregnancy PGH and PL direct maternal metabolism toward the mobilization of maternal energy resources to furnish the needs of the developing fetus. Soon after birth, insulin resistance reverts to the normal nonpregnant state, suggesting that maternal glucose homeostasis is influenced by hormonal factors produced by the fetus-placenta.

This fetal-maternal hormonal interaction represents an example of genetic conflict. The fetus, through natural selection, acquires traits, for example, PGH and PL, that favor the extraction of resources from the maternal compartment. Conversely, mothers have evolved mechanisms to counteract fetal demand. Disorders on either side of this equation lead to pathophysiology. For example, in women with gestational diabetes, insulin secretion is insufficient to balance the decrease in insulin sensitivity and consequently blood glucose levels increase leading to an oversupply of glucose in the fetus resulting in macrosomia.