Reviews in Endocrine & Metabolic Disorders 2005;6:173–182
C 2005 Springer Science + Business Media, Inc. Manufactured in The Netherlands.

Growth Hormone During Development Joy Osafo∗ , Yuhong Wei∗ , Gurvinder Kenth∗ and Cynthia Gates Goodyer Departments of Pediatrics and Experimental Medicine, McGill University, Montreal, Quebec, Canada

Key Words. growth hormone, development, human, liver, adipose, growth plate

Introduction Growth hormone (GH) is a member of an extensive GH/prolactin (PRL) family of peptides [1]. The evolution of this family in primates differs significantly from other mammals: whereas in most species there is a single GH gene and multiple prolactin-like genes, in the primate there are multiple GH-related genes and a single PRL gene [2,3].

Primate GH Gene Complex The human GH gene cluster consists of five highly conserved (>90%) genes spanning more than 48 kb on chromosome 17q22–24 [4]. The apparent result of a series of gene duplication events, they line up from 5
to 3
as GH N , PL-L (placental lactogen), PL-A, GH V (GH variant) and PL-B (Fig. 1). Normally these genes are expressed in very tissue-specific patterns: GH N in pituitary (somatotropes and somatolactotropes) and the four others in placenta (syncytiotrophoblasts and extravillous cytotrophoblasts). Onset of their expression is regulated by a complex upstream locus control region (LCR), at least in part by the establishment of large domains of acetylated and methylated chromatin that encompass portions of the LCR as well as the individual genes [4,5]. Three Pit-1 response elements within the LCR are critical for full activation of the GHN gene. For the four placentally-expressed genes, there is an additional common regulatory mechanism: conserved sequences called P elements, that lie 2 kb upstream of each of these genes, either enhance their expression in the placenta or inhibit expression in the pituitary [6].

Pituitary GHN First secreted by fetal somatotropes at ∼7–8 weeks of fetal life, immunoreactive GH serum levels rise rapidly,

reaching a very high peak at midgestation of ∼120 ng/ml. These levels decrease thereafter until birth but, even at term, plasma levels remain significantly higher than maternal or non-pregnancy levels until almost three months postnatally [7]. In vitro investigations of fetal pituitaries and clinical studies of GH secretory patterns in premature infants suggest that fetal somatotropes respond primarily to the stimulatory effects of growth hormone releasing hormone (GHRH) and that regulation by the corresponding hypothalamic inhibitory factor, somatostatin, only matures after birth [7,8]. Sex steroids have significant effects on GH synthesis and secretion; in puberty, sexually dimorphic patterns of GH secretion are established [9]. There are two major forms of circulating GHN , 22 K (191aa) and 20 K (176aa), due to an alternative splicing event [10]. The physiological regulation of the two forms is thought to be similar since their circulating ratios don’t vary significantly throughout postnatal life (∼85:15); fetal ratios have not been reported. While both act as full agonists at the GH receptor (GHR) and exert similar lipolytic activities, the 20 K form has demonstrated less of an acute insulin-like effect as well as less agonistic activity at the PRL receptor (PRLR) [10].

Placental GHV GHV (191aa) differs from GHN by 13 residues scattered throughout the peptide and by a unique N-linked glycosylation site; as a result, it has relatively high somatogenic (including lipolytic) but low lactogenic activity [10]. Beginning at ∼6 weeks of fetal life, it is secreted in a tonic fashion by the syncytiotrophoblast into the maternal, but not fetal, circulation, suggesting that it has no direct role in fetal development. Maternal levels of ∗ These

authors contributed equally to this chapter.

Address correspondence to: Cynthia Gates Goodyer, Endocrine Research Laboratory, McGill University-Montreal Children’s Hospital Research Institute, Room 415/1, 4060 Ste Catherine St. West, Montreal, Quebec, Canada H3Z 2Z3. E-mail: [email protected] 173

174

Osafo et al.

is no GH feedback at the level of the syncytiotrophoblast [14]. Glucose levels have been shown both in vivo and in vitro to negatively modulate GHV secretion. In addition, GHV concentrations are negatively correlated to prepregnancy BMI and maternal leptin levels suggesting that GHV , like GHN , is negatively regulated by adipose tissue. These clinical findings are important given the role of the GHV /IGF-1 axis in facilitating nutrient flow to the placenta and fetus. Whether other nutritional elements have effects, as has been shown for PL (see below), still needs to be examined as do the intracellular mechanisms involved.

PL Fig. 1. The human growth hormone (GH) cluster and its upstream locus control region (LCR) on chromosome 17q22-24. The pituitary GH gene (GH N ) is a black box while the placentally expressed genes (P L-L , P L-A, GH V and PL-B) are hatched boxes. Sites of transcription are indicated by bent arrows. The five DNase I hypersensitivity sites (HS) of the LCR are marked by straight arrows: HSI and II are involved in induction of GH N expression, HSIV (hatched line) is specific for the four placentally expressed genes, and HSIII and HSV are common to both. The three essential Pit-1 sites within HSI are indicated by ovals in the magnified section of the LCR. The P elements that regulate placental gene expression are represented by hatched circles. (Modification of figure from [4])

GHV increase throughout gestation, reaching a maximum of ∼14 ng/ml by 34–35 weeks [11]. Surprisingly, this placental hormone gradually replaces maternal GHN . Because maternal IGF-1 levels during the second half of gestation in both normal and intrauterine growth retarded (IUGR) pregnancies correlate with GHV but not GHN or PL, GHV is thought to become the major influence on maternal serum IGF-1 levels, placento-fetal nutritional supply and, subsequently, fetal growth [11,12]. Whether it also has direct effects on the placenta, through the syncytiotrophoblast GHR, is not known. The mechanisms regulating GHV production and secretion are also not yet clear [13]. The fact that there is no evidence of a granulated (secretory vesicle) storage form suggests that GHV production may be regulated more at the level of gene transcription than from a “readily-releasable pool”. GHV levels are known to be low in women carrying IUGR fetuses, likely due to a malfunctioning placenta, and they disappear rapidly at the onset of labour. Acute treatments with either GHRH or somatostatin have no effect, despite the fact that these peptides and their receptors have been detected in the syncytiotrophoblast [13]. Since GHV is secreted in a tonic fashion, chronic studies with the classical GH regulatory peptides need to be undertaken before we rule out their potential role. Interestingly, although GHV can shut off maternal GHN production, therapy of a GHN -deficient pregnant woman with recombinant 22 kD GH did not suppress placental GHV , suggesting that there

Placental lactogen, like GHN and GHV , is a 191aa (22 kD) protein. While it binds to both GHR and PRLR, it is primarily a “lactogenic” hormone: it is equipotent with PRL at the PRLR but only a weak agonist of GHR [3]. The three PL genes are coordinately induced within the first month of gestation. The PL-L gene is thought to be a pseudogene: although multiple mRNA variants are produced, no protein product has ever been reported. In contrast, PL-A and PL-B are very actively expressed, although PL-A gene transcription predominates (∼5-fold) [15]. PL is secreted in a tonic fashion into both the fetal (maximum ∼30 ng/ml at term) and maternal (maximum ∼3–6 ug/ml) compartments. Thus, PL can play a direct role in fetal development and, in the mother, in conjunction with GHV , can stimulate IGF-1 production and intermediate metabolism, resulting in an increase in nutrient availability for the feto-placental unit. Traditional hypothalamic factors, such as GHRH, somatostatin, thyrotropin releasing hormone and dopamine, do not influence PL production [16]. However, a number of nutritionally-related factors do, including HDL, apolipoproteins, T3, and retinoic acid. It is important to note that these effects were only observed after several (3–5) days of exposure and that syncytiotrophoblast PL mRNA and protein levels rose in parallel, suggesting that, like for GHV , regulation occurs predominantly at the transcriptional level. cAMP. DAG and phorbol esters all stimulate PL synthesis suggesting that multiple signaling cascades are involved.

PRL The human PRL gene, located on chromosome 6, consists of six exons and five introns [17]. The mature protein (199aa, 23 kD) is primarily produced by lactotropes in the anterior pituitary but PRL is also synthesized in lymphoid and decidual cells. Tissue-specific expression appears to be regulated by two distinct promoters, “pituitary” expression by a proximal promoter within ∼5 kb upstream of the transcriptional start site (TSS) while the “extrapituitary”

Growth Hormone During Development

promoter is located ∼8.3 kb upstream of the TSS [18]. Fetal serum levels of PRL are very low from 10–23 weeks of fetal life, probably a product of the pituitary somatolactotropes [19]. However, after 23 weeks, with the appearance of true lactotropes, there is a sharp rise to very high levels (∼200 ng/ml, comparable to hyperprolactinemic patients) during the last trimester [7,19]. After birth, serum PRL levels gradually fall, reaching the normal adult range by 4–6 weeks postnatal. Mechanisms regulating fetal PRL are as yet unknown although rising estrogens are thought to play a major role; classical hypothalamic factors (TRH, dopamine) have little to no effect in vivo [7,20]. During gestation, PRL is also produced by maternal decidua, regulated by the “extrapituitary” proximal promoter. Levels reach a peak in amniotic fluid by 20–25 weeks and then decline to term. Progesterone appears to be the major influence on these levels while TRH, dopamine and estrogen have little to no effect [7,21,22]. Decidual PRL appears to have a major role during early mouse development: the PRLR knockout female is infertile due to a complete failure of blastocysts to implant [21–23]. Studies in the rhesus monkey also implicate decidual PRL in regulation of amniotic and fetal extracellular fluid and electrolyte balance; interestingly, neither PL nor GH has the same effects [21]. The role of fetal pituitary PRL is as yet undefined [24].

GHR The GHR, a member of the Class I cytokine receptor family, is a 620aa single transmembrane protein that forms primarily homodimers [25]. The extracellular ligand binding domain of these dimers is specific for GH-related (GHN , GHV , PL) but not PRL hormones (Fig. 2). Interestingly, while PL binds to GHR, in some species it may in fact be antagonistic to GH: ruminant PL binds only one molecule of GHR, thus sequestering GHRs without activating them

175

[26]. However, it can form active heterotrimers with one molecule of GHR and one PRL receptor (PRLR), creating a chimeric receptor complex with unknown biological consequences in those cells expressing both types of receptors [27]. The cytoplasmic region of the GHR has multiple signaling domains [28]. Primarily through the binding and activation of Jak2 at the Box 1 region and subsequent phosphorylation of tyrosine sites along the intracellular domains of the GHR dimer, multiple SH2 binding sites are formed. These act as scaffolding for the binding of a number of different signaling molecules including STAT-1, -3 and -5, IRS-1 and -2, SOCS-1, -2, -3 and CIS as well as several phosphatases. The end result is regulation of several signaling cascades (ERK1/2, PI-3 kinase, calcium) as well as direct STAT regulation of multiple genes and major effects on target cell proliferation, differentiation, metabolism and expression. GHR are widely expressed at both the mRNA and protein level in human fetal tissues by the end of the first trimester [29–34]. These receptors appear to be functional in that they are capable of binding GH and exerting anabolic effects on a number of early to midgestation GH target cells [33,35]. Newborns with dysfunctional GHR (Laron dwarfs) or congenital GH deficiency are often born short although their birth weights are normal, likely due to increased fat deposition; a rapid decline in early postnatal growth is a hallmark of these two syndromes [36–38]. In contrast, GHR k/o mice are normal sized at birth but display numerous growth and metabolic defects by three weeks of postnatal life [39].

PRLR PRLR are Class I cytokine receptors with similar structural and activation systems as GHR. In contrast, however, PRLR bind and are activated by PRL as well as PL, GHN and GHV (Fig. 2). Although no newborns with congenital absence of PRL have been reported, fetal effects have not been observed in PRLR k/o mice [23,24], despite the fact that PRLR are widely expressed in rodent (and human) fetal tissues [40]. The major postnatal effects of PRL are on the reproductive system, mammary gland development and maternal behaviour [23].

Redundancy Issue

Fig. 2. The “redundancy” of the human GH/PL/PRL-receptor-IGF axes showing the relationship of the classical fetal growth hormones to either direct or indirect (IGF) actions on growth and development.

All four hormones, GH, GHV , PL and PRL, have been shown to stimulate production of the insulin-like growth factors (primarily IGF-1, occasionally IGF-2) in multiple fetal as well as postnatal target tissues, although not all of their effects are exerted through the IGFs (Fig. 2) [33,35,41,42]. The IGFs are ∼70 kD growth factors that

176

Osafo et al.

are essential for normal embryonic, fetal as well as postnatal growth [41,42]. IGF-2 is thought to be more actively involved during early embryogenesis and IGF-1 during later fetal and postnatal life; both exert their growth effects through a common Type 1 IGF receptor (IGF-1R). Thus, it seems likely that if one or two of the fetal “growth hormones” are deficient, the others can still stimulate production of the common biological endpoint, the IGFs/IGF1R, and maintain “normal” fetal growth and development. However, this case of natural redundancy is lost after birth, with loss of PL production and decreased PRL levels, such that GH must be present for normal postnatal growth. Defective Pit-1 gene function leads to multiple pituitary hormone deficiency (MPHD) syndromes because of the requirement for this transcription factor for terminal differentiation of three pituitary cell types: somatotropes, lactotropes and a subset of thyrotropes [43]. Interestingly, probands have normal birth weights and lengths although their growth rates fall drastically immediately after birth. These clinical findings raise the question of whether the placental “growth hormones” are actively promoting fetal growth in these individuals. Unfortunately, there are no published data on cord serum levels of PL or GH V levels in newborns with MPHD. There is also the issue that, since all five of the genes in the GH cluster have two Pit-1 sites in their proximal promoters, why would GH N expression be affected but not the PL or GHV genes? One possible explanation is that there are three essential Pit-1 sites in the LCR region specific for induction of pituitary GH N expression but none within the placental-specific domains of the LCR (N.E. Cooke, personal communications). In addition, of the two sites within the proximal promoters of the placental genes, only one appears to be functional [44] and Pit-1 concentrations within the placenta are low. Thus, Pit-1 may not be critical for either induction or expression of the placental hormones and, although this needs to be proven, placental hormones may be actively involved in regulating fetal growth in Pit-1 null individuals. A review of single and multiple deficiencies within the GH cluster has not produced any consistent clinical findings relative to fetal growth: a few probands are described as having IUGR but the majority have normal birth weights if not lengths, including two siblings thought to have deletion of the four functional genes in the cluster leaving only the pseudogene PL-L [45,46]. Neither “quadruple” hormone (GHN /GHV /PL/PRL) or “dual” receptor (GHR/PRLR) defects have been reported in the human. However, a double mutant mouse (90% GH deficiency, PRLR k/o), that effectively eliminates any PRL or PL effects and markedly diminishes the role of GH, was recently created [24]. Unfortunately, the offspring have only been studied from day 7 postnatal, but these early stage data suggest that fetal growth and metabolic

activity may be affected. The major developmental defects observed postnatally include decreased IGFs and insulin, lower postnatal weight gain curves and femur lengths, increased body fat and decreased lean body mass. The critical experiment still to be done, the crossing of GHR and PRLR null mice, has not yet been reported.

Effects on Major Target Tissues During Development Waters and Kaye [12] recently reviewed multiple roles for GH at the earliest stages in development (e.g. ovulation rate, preimplantation embryo development, embryonic development). The following is a brief review of the role of GH (and related) hormones during fetal and early postnatal life in three major target tissues: liver, adipose and epiphyseal growth plate. Liver GH is an anabolic hormone, primarily involved in carbohydrate and lipid metabolism in the liver. The known effects of GH on postnatal hepatocytes include stimulation of gluconeogenesis and glycogenolysis and regulation of several enzymes involved in glucose metabolism and transport (e.g. phosphoenolpyruvate kinase C, GLUT2) [47]. Two studies suggest that GH has a role in glucose metabolism in the fetal liver: primary cultures of midgestation (15–20 week) hepatocytes respond to GH treatment with a significant increase in glucose uptake [33] while newborns with GH deficiency are often observed to be hypoglycaemic [48]. GH also affects plasma cholesterol levels by increasing hepatic expression of LDL (low density lipoprotein) receptors and production of lipoproteins and triglycerides. GH treatment in both normal and GH deficient humans increases hepatic LDL receptors, resulting in reduced LDL cholesterol and increased lipoprotein A. The effects of GH on LDL receptors and LDL are direct as IGF-1 does not have the same effect as GH [49]. Whether GH also has effects on hepatic lipid metabolism in the fetus is not known. In humans, the liver bud can be detected at four weeks as an endodermal outgrowth from the foregut [50]. The bipotent hepatoblasts within the bud migrate into the septum transversum where, in response to a variety of signals, they differentiate, adopting a hepatic or biliary lineage. By the 7th week of fetal life, cordlike structures of hepatocytes can be observed. By 10 weeks, the liver has a fully developed structure, including portal triads and a functional bile duct/gall bladder system. In addition, the fetal liver has a unique function at this stage in development: it is a major site for haematopoiesis. Only in late gestation does the site of haematopoiesis move to the bone marrow leaving the

Growth Hormone During Development

liver to carry out the more traditional metabolic functions observed in the postnatal individual. Functional receptors for GH and PRL are present on hepatocytes as early as 8 weeks of gestation in the human fetus [29–33,40]. GH, PL and IGF-1 have been shown to stimulate proliferation of human fetal hepatocytes [35]. There is a hierarchy of liver enriched transcription factors (LETFs) that are involved in liver development, many of which are regulated by GH (HNF-3β, HNF-3γ , HNF-4α, C/EBPα, HNF-6, DBP, c-fos and c-jun) [47,51]. HNF-3α and β, that are involved in chromatin remodeling, are first expressed in the definitive endoderm, along with HNF-4α. HNF-3β appears to be essential for hepatic development as HNF-3β knockout mice do not form livers. HNF-4α knockout mouse embryos die during gastrulation; however, conditional HNF-4α knockout adult mice display defects in the expression of several key liverspecific genes, suggesting that it is important for both fetal and postnatal liver function [52]. HNF-6 appears in the liver primordium after HNF-4α, HNF-3 and HNF1β and is involved in the differentiation process by regulating the expression of several liver-specific genes (e.g. transthyretin, alpha fetoprotein); the liver is present in HNF-6 null mice but in an abnormal hypoplastic form [53]. In the mature liver, C/EBP and DBP are also actively involved in liver-specific gene regulation: fetal development in C/EBPα knockout mice is normal but they die soon after birth due to failure to accumulate glycogen in the liver while DBP is known to regulate several liver genes expressed in a circadian pattern [51]. Interestingly, a recent study in young adult mice showed that hepatic GHR mRNA levels were lowest in the morning (0800–1000) and higher at midday (1200–1400) and evening (2000–2200). In their GH-deficient counterparts, the circadian effect was abolished, suggesting that GH regulates expression of its own receptor in the liver [54]. In bovine hepatocytes, GH treatment can increase DBP expression [47]. Computer scanning programs for transcription factors have identified putative binding sites for DBP in a promoter regulating expression of the major liver-specific mRNA of GHR in humans, mice, sheep and the cow (C.G. Goodyer, unpublished data). These data collectively imply a role for GH in regulating its own receptor through DBP. Liver is the main source of the circulating IGF-1, IGFBP-3 (IGF binding protein 3) and ALS (acid labile subunit) both pre- and post-natally. Serum IGF-1 is relatively low during fetal life, rising progressively in parallel with IGF-2 but with levels ∼7-fold lower [55]. In keeping with these data, we have shown that human fetal hepatocytes respond to GH treatment with an increase in IGF-2 but not IGF-1 secretion, suggesting either an immaturity of the GHR or an adaptation to the in utero environment

177

[33]. IGFBP-3 production is also stimulated by GH but only beginning in the third trimester [42]. It is likely that, in addition, GH affects late gestation liver production of ALS: hepatic ALS expression is known to be influenced by GH postnatally and fetal serum ALS levels increase in parallel with IGFBP-3 [56]. IGF-1, IGF-2, IGFBP-3 and ALS levels rise slowly in the postnatal infant and child but dramatically in the adolescent, concurrent with the somatic growth spurt [41,42]. IGF-1 levels are higher in females before, during and post puberty than in males. The sex steroids appear to exert their effects on IGF-1 indirectly through GH since, in GHdeficient humans, the pubertal increase in IGF-1 levels in the bloodstream is not observed [57]. In addition to metabolic functions, the liver also acts as a detoxifying organ, a function carried out by a family of P450 enzymes (CYP) [57]. Most of these enzymes are expressed in a sexually dimorphic manner: they may be exclusively expressed in the male or female or expressed at different levels in response to the sex-linked GH secretory pattern. Sexually dimorphic influences are first observed during puberty. In rats, STAT5b has been identified as the intracellular mediator of the sexual dimorphic effects on GH on its target genes [58]: STAT5 activity (tyrosine phosphorylation, nuclear translocation) is high during a GH pulse and, in pubertal male rats, GH stimulation of STAT5b positively regulates male specific CYP genes in the liver. HNF-6 and HNF-3β are two LETFs that are regulated by GH in a sexually dimorphic pattern in the adult, being more highly expressed in the adult female [47,59]. In contrast, HNF-4α has been found to contribute to GH regulation of several sexually dimorphic liver P450 cytochrome genes, favouring the expression of certain male CYPs over female ones [59]. Whether GH influences expression of these CYP genes in the fetal hepatocyte is not known. cDNA microarray studies in the rat have identified multiple GH-regulated genes involved in lipid and carbohydrate metabolism as well as the production of transport binding proteins and detoxifying enzymes [60]. Similar experiments have not been carried out in the fetus, either rodent or human, limiting our knowledge of the relative importance of GH and related hormones in fetal liver development and function.

Adipose Tissue Adipose tissue, primarily referred to as white adipose tissue (WAT), is the major source for energy storage in the human body [61]. The majority of adipose tissue consists of lipid-filled adipocytes, interspersed with several other types of cells, including blood cells, endothelial cells, pericytes and fibroblast-like adipose precursor cells

178

Osafo et al.

that are capable of differentiating into mature adipocytes upon proper stimulation. As obesity, especially childhood obesity, which results from an excess of WAT, is becoming an epidemic health problem, it is important to understand the ontogeny and regulation of adipose tissue. Both clinical and experimental studies have clearly demonstrated that GH is a major regulator of WAT [62–64]. This section will review human adipose tissue development as well as GH effects on adipocyte differentiation and function. Our understanding of human adipose tissue development during gestation and early postnatal life is still limited. In the fetus, both white and brown adipose tissues are present, where brown fat cells are characterized by the unique expression of uncoupling protein 1 (UCP1) [65,66]. Brown fat development is primarily to ensure that enough UCP1 is synthesized so that thermoregulation is effective upon cold exposure in the extra-uterine environment. Brown fat reaches its maximal abundance (in neck, scapular, mediastinal and intra-abdominal regions) at the time of birth and is then normally undetectable after the immediate postnatal period. Human WAT becomes detectable around the 12th week of fetal life. First noticeable in the head and neck, deposition of WAT rapidly progresses to the trunk and then the limbs such that, by the beginning of the third trimester, adipose tissue is present in all of the expected subcutaneous and intra-abdominal regions. During the last trimester, WAT goes through a rapid expansion due to increased fat cell number (hyperplasia) as well as size (hypertrophy). While there are marked differences from one depot to another in terms of fat accumulation, there do not appear to be gender differences at this stage in development. Clinical studies suggest that fetal lipids in early gestation are derived from maternal fatty acids that cross the placenta, followed by a gradual shift to de novo synthesis from glucose in fetal tissues. Fetal fat mass is positively correlated with maternal IGF1 levels during the last trimester and can be significantly altered by changes in maternal nutrition [65–67]. Postnatally, fat cell size and number can also fluctuate markedly, depending on the nutritional intake and lifestyle of the individual child. Puberty is marked by substantial increases in adipocyte cell number and size as well as significant gender differences in fat distribution, with a preponderance in the upper body of males and in the lower body of females. GH appears to be involved in the triggering and/or modulating of the fat development process both in utero as well as postnatally. Our own studies have shown that GHR are expressed in human fetal subcutaneous fat by 16 weeks of fetal life [68]. Functional GHR are present on both preadipocytes (adipose precursor cells) and mature adipocytes [62]. A marked up-regulation of GHR gene

expression has been observed during adipocyte differentiation, a process whereby fibroblast-like preadipocytes are converted into lipid-filled adipocytes, using either human or mouse preadipocyte cell lines or primary cultures [61,69–71]. GH-deficient children have reduced fat cell numbers but enlarged fat cell volume, while giving GH replacement can normalize both disorders [62]. Treatment of GH-deficient adults with GH can significantly reduce their fat mass [62]. In addition, it is well recognized that GHR-defective Laron Syndrome patients progress towards obesity following birth [36,38]. GHR knockout (GHR-/-) dwarf mice also have an increased percentage of body fat, with most of the excess accumulating in subcutaneous WAT as well as interscapular BAT [64,72,73]. However, hypophysectomized pigs can still form adipose tissue, suggesting that GH is not obligatory for adipose tissue development and that other hormones can compensate [63]. These data demonstrate that, although the role of GH in regulating adipose tissue development may be redundant due to alternate effective hormones, adipose tissue is one of the direct target tissues for GH. In vitro studies of the effects of GH on adipocyte differentiation and function in combination with clinical observations have suggested two distinct functions for GH in adipose tissue: (1) a consistent antilipogenic and lipolytic effect on differentiated adipocytes characterized by a reduced fat cell volume; and (2) a controversial effect (either stimulatory or inhibitory) on adipocyte differentiation. The major chronic effects of GH on mature adipocytes include enhanced lipid breakdown (lipolysis) and reduced triglyceride (TG) accumulation (anti-lipogenesis) as well as a decrease in glucose uptake (anti-insulin effect) [61,62]. GH inhibits lipoprotein lipase (LPL) activity, the critical enzyme which hydrolyzes TGs from circulating VLDL and chylomicron particles and releases FFA Thus, it reduces FFA that can be taken up by adipocytes and used to generate lipid, ultimately reducing lipid accumulation. This inhibitory effect does not seem to happen at the gene expression level, and may be due to post-translational modification of enzyme activity. On the other hand, GH potently promotes lipolysis by prolonging hormone-sensitive lipoprotein lipase (HSL) activity, which is the rate-limiting enzyme in charge of breaking down TG in adipocytes. The mechanism by which GH regulates HSL is not clear yet. But it appears to be indirect, through upregulating the β-adrenergic receptor to activate HSL and by prolonging the cAMP-signaling cascade to maintain HSL activity [62]. Clinical studies of GH-/- and GHR-/- newborns have shown that they have well-developed fat depots despite deficiencies in birth length; these data suggest that, normally, GH is acting as a lipolytic hormone in the late gestation fetus [36].

Growth Hormone During Development

Adipose tissue mass is determined by both the fat cell number and the fat cell size. Although the effects of GH on reducing fat cell size are consistent, in vitro studies of the effects of GH on fat cell number have been very controversial because of conflicting data from different cell culture models [61,62]. Fat cells are generated from a process called adipogenesis, whereby preadipocytes, precursor cells that have committed to adipocyte lineage, further differentiate into lipid-filled functioning adipocytes under the control of an array of regulatory hormones. In the murine clonal cell line 3T3-F442A, GH promotes differentiation by priming the cells to the mitogenic effects of IGF-1, which then drives clonal expansion [74]. In the same cell line, GH helps adipocyte conversion by inducing an antimitogenic state and modifying the extracellular matrix [74]. In contrast, studies using cultured primary preadipocytes isolated from either rat or human adipose tissues showed that GH stimulates preadipocyte proliferation via stimulating IGF-1, but markedly inhibits the formation of new adipocytes as well as the expression of mature adipocyte marker enzymes [74]. This GH-induced suppression of adipocyte differentiation appears to be the direct effect of GH, not via IGF-1. The reason for such obvious differences between experimental outcomes in these two culture models is not known. It may be that different developmental stages of adipocytes have different responses to GH or that some subset of the cells in primary culture may have already been exposed to GH in vivo and, therefore, may respond differently. Further studies are needed to determine whether or not GH regulates adipocyte numbers and whether this effect begins in utero.

Epiphyseal Growth Plate Longitudinal bone growth results from tightly controlled endochondral ossification in the long bones beginning around 8 weeks of human fetal life [50,75]. The epiphyses and metaphyses of long bones originate from independent ossification centers and are separated by a growth plate. The growth plate, first visible around 6–7 weeks of fetal life, is a cartilaginous structure consisting of columns of proliferating and differentiating chondrocytes that can be divided into four regions: the germinal, proliferative, hypertrophic and calcifying zones. At puberty, epiphyseal fusion occurs, resulting in cessation of longitudinal bone growth. Regulation of bone growth is complex. Final height of an individual is influenced by many factors including nutrition, environment and psychosocial state while hormonal regulation plays an essential role for normal bone

179

growth beginning in utero [75] The major hormones that regulate long bone growth during fetal and postnatal life are GH, IGF-1, glucocorticoids and thyroid hormones (T3 and T4), as well as sex steroids during puberty. The focus here will be on GH. GH acts on its target tissues either directly or indirectly through IGF-1 and IGF-2 [42]. IGF-2 is known to be essential for normal embryonic and early fetal growth, whereas IGF-1 appears to have more of a growth-related role during late gestation, childhood and adult life. The original theory for the growth-promoting actions of GH and IGF-1 was the “somatomedin (IGF) hypothesis” [42]. This proposed that GH stimulated production of IGF-1 from liver, which then promoted growth at the epiphyseal plate. Subsequently, with the discovery that most tissues produce both IGF-1 and IGF-2, the “dual effector theory” developed, proposing that GH acts directly on the germinal zone of the growth plate to stimulate chondrocyte differentiation and, by local stimulation of IGF-1 production, the clonal expansion of these chondrocytes and hence long bone growth [42]. More recent findings from transgenic and knockout mouse models (showing that GH has IGF-1 independent effects on chondrocyte proliferation, that IGF-1 has specific effects on the size of hypertrophic chondrocytes, and that IGF-2 rather than IGF-1 is normally expressed by proliferating chondrocytes) have led to modifications of this latter theory [75,76]. GHR and IGF-1 receptors are expressed at all stages of growth plate chondrocyte differentiation in several species [75]. In sternal growth plates of 13–18 week fetuses and 1–5 month old humans, GHR protein has been detected in proliferative and hypertrophic chondrocytes as well as perichondral fibroblasts [34]. Both the fetal and infant chondrocyte GHR appear to be functional since they show specific binding of radiolabeled GH. Immunoreactive PRLR have been reported in the mesenchymal precartilage and maturing chondrocytes of the endochondral long bones, craniofacial bones, ribs and vertebrae [40]. The importance of GH and IGF-1 are supported by knockout models and natural human disorders [12,36,42,76]. Knockout mice models for IGF-1 or its receptor and children with either an IGF-1 or IGF-1R deletion all show severe IUGR as well as postnatal dwarfism. A mouse model for Laron Syndrome (GHR-/-) and a double knockout for both GHR and IGF-1 demonstrate growth retardation; the double knockout mice were much smaller than the single knockout cohorts and displayed significant IUGR. As stated earlier, GH- and GHR-deficient humans are often small at birth and rapidly display severe growth retardation postnatally. These data support the concept that both GH and IGF-1 play a major role in longitudinal bone growth both in utero and postnatally.

180

Osafo et al.

Whether there are in utero effects of PL has not been reported. From PRLR−/− mice studies, it appears that PRL does not have a major role in utero on the epiphyseal growth plate [23,40]. However, PRL may influence osteoblast activity: bone deposition since there is less ossification in calvaria of 18.5 day fetal PRLR−/− mice and a significant decrease in their bone formation postnatally.

Conclusions Present dogma suggests that, while pituitary-derived growth hormone (GH) is essential for normal postnatal growth and development, it has no role in the fetus. This widely published belief is based on both clinical studies and experimental animal models showing that loss of pituitary GH has no effect on the prenatal infant or animal. However, analysis of these natural and laboratory experiments is complicated by the fact that there are multiple members of the GH/prolactin (PRL) family, produced by alternate genes in different fetal tissues. Not only are the receptors as well as intracellular signaling pathways used by these various members often common, but the biological endpoints are also often the same. Thus, there is a redundancy factor that should be recognized in describing the role of GH in early stages of development. This review focuses on the redundancy issue and outlines the roles of the various GH/PRL members during development of three major GH target tissues (liver, adipose tissue and epiphyseal growth plate), with a special emphasis on the human.

Key Unanswered Questions Why did the GH gene duplicate in primates and the PRL gene in other mammals (i.e. is there some advantage of the somatogenic (GHN , GHV ) and mixed somatogenic/lactogenic (PL) hormones to the primate organism?)? Why doesn’t a Pit-1 null individual exhibit IUGR? What happens during fetal development if you cross GHR and PRLR null mice? What regulates GHV production by the human syncytiotrophoblast? What regulates GHR expression in each of the human target tissues, both preand post-natally? What is the role for GH during liver development? What is the role for GH in the preadipocyte (proliferation vs. differentiation?) and the differentiating adipocyte? What is the role for IGF-2 vs. IGF-1 during human fetal development, especially within the epiphyseal growth plate?

Acknowledgments This work was supported by grants from the Canadian Institutes of Health Research (C.G.G.). YW and GK are the

recipients of studentships from the Fonds de Recherche en Sant´e du Qu´ebec and the McGill University-Montreal Children’s Hospital Research Institute, respectively. Our apologies to those whose work has not been directly cited; due to space constraints, many reviews have been referenced instead of an exhaustive list of primary publications.

References 1. Goffin V, Shiverick KT, Kelly PA, Martial JA. Sequence-function relationships within the expanding family of prolactin, growth hormone, placental lactogen, and related proteins in mammals. Endocr Rev 1996;17:385–410. 2. Revol DM, Esquivel ED, Martinez D I, Saldana H. Expansion and divergence of the GH locus between spider monkey and chimpanzee. Gene 2004;336:185–193. 3. Soares MJ. The prolactin and growth hormone families: pregnancyspecific hormones/cytokines at the maternal-fetal interface. Reprod Biol Endocrinol 2004;2:51. 4. Ho Y, Liebhaber SA, Cooke NE. Activation of the human GH gene cluster: roles for targeted chromatin modification. Trends Endocrinol Metab 2004;15:40–45. 5. Kimura AP, Liebhaber SA, Cooke NE. Epigenetic modifications at the human growth hormone locus predict distinct roles for histone acetylation and methylation in placental gene activation. Mol Endocrinol 2004;18:1018–1032. 6. Norquay LD, Yang X, Sheppard P, Gregoire S, Dodd JG, Reith W, Cattini PA. RFX1 and NF-1 associate with P sequences of the human growth hormone locus in pituitary chromatin. Mol Endocrinol 2003;17:1027–1038. 7. Goodyer CG. Ontogeny of pituitary hormone secretion. In Collu R, Ducharme JR, Guyda H eds. Pediatric Endocrinology. New York: Raven Press; 1989;125–170. 8. Miller JD, Wright NM, Esparza A, Jansons R, Yang HC, Hahn H, Mosier HD, Jr. Spontaneous pulsatile growth hormone release in male and female premature infants. J Clin Endocrinol Metab 1992;75:1508–1513. 9. Leung KC, Johannsson G, Leong GM, Ho KK. Estrogen regulation of growth hormone action. Endocr Rev 2004;25:693– 721. 10. Boguszewski CL. Molecular heterogeneity of human GH: from basic research to clinical implications. J Endocrinol Invest 2003;26:274–288. 11. Wu Z, Bidlingmaier M, Friess SC, Kirk SE, Buchinger P, Schiessl B, Strasburger CJ. A new nonisotopic, highly sensitive assay for the measurement of human placental growth hormone: development and clinical implications. J Clin Endocrinol Metab 2003;88:804– 811. 12. Waters MJ, Kaye PL. The role of growth hormone in fetal development. Growth Horm IGF Res 2002;12:137–146. 13. Lacroix MC, Guibourdenche J, Frendo JL, Muller F, EvainBrion D. Human placental growth hormone–a review. Placenta 2002;23(Suppl A):S87–S94. 14. Lonberg U, Damm P, Andersson AM, Main KM, Chellakooty M, Lauenborg J, Skakkebaek NE, Juul A. Increase in maternal placental growth hormone during pregnancy and disappearance during parturition in normal and growth hormone-deficient pregnancies. Am J Obstet Gynecol 2003;188:247–251. 15. MacLeod JN, Lee AK, Liebhaber SA, Cooke NE. Developmental control and alternative splicing of the placentally expressed transcripts from the human growth hormone gene cluster. J Biol Chem 1992;267:14219–14226.

Growth Hormone During Development

16. Handwerger S, Datta G, Richardson B, Schmidt CM, Siddiqi T, Turzai L, Anantharamaiah GM. Pre-beta-HDL stimulates placental lactogen release from human trophoblast cells. Am J Physiol 1999;276:E384–E389. 17. Goffin V, Binart N, Touraine P, Kelly PA. Prolactin: the new biology of an old hormone. Annu Rev Physiol 2002;64:47–67. 18. Brar AK, Kessler CA, Handwerger S. An Ets motif in the proximal decidual prolactin promoter is essential for basal gene expression. J Mol Endocrinol 2002;29:99–112. 19. Asa SL, Kovacs K, Horvath E, Losinski NE, Laszlo FA, Domokos I, Halliday WC. Human fetal adenohypophysis. Electron microscopic and ultrastructural immunocytochemical analysis. Neuroendocrinology 1988;48:423–431. 20. Asa SL, Ezzat S. Molecular basis of pituitary development and cytogenesis. Front Horm Res 2004;32:1–19. 21. Tseng L, Mazella J. Prolactin and its receptor in human endometrium. Semin Reprod Endocrinol 1999;17:23–27. 22. Jabbour HN, Critchley HO. Potential roles of decidual prolactin in early pregnancy. Reproduction 2001;121:197–205. 23. Kelly PA, Binart N, Lucas B, Bouchard B, Goffin V. Implications of multiple phenotypes observed in prolactin receptor knockout mice. Front Neuroendocrinol 2001;22:140–145. 24. Fleenor D, Oden J, Kelly PA, Mohan S, Alliouachene S, Pende M, Wentz S, Kerr J, Freemark M. Roles of the lactogens and somatogens in perinatal and postnatal metabolism and growth: studies of a novel mouse model combining lactogen resistance and growth hormone deficiency. Endocrinology 2005;146:103–112. 25. Kossiakoff AA. The structural basis for biological signaling, regulation, and specificity in the growth hormone-prolactin system of hormones and receptors. Adv Protein Chem 2004;68:147–169. 26. Gootwine E. Placental hormones and fetal-placental development. Anim Reprod Sci 2004;82-83:551–566. 27. Gertler A, Djiane J. Mechanism of ruminant placental lactogen action: molecular and in vivo studies. Mol Genet Metab 2002;75:189– 201. 28. Rowland JE, Lichanska AM, Kerr LM, White M, d’Aniello EM, Maher SL, Brown R, Teasdale RD, Noakes PG, Waters MJ. In vivo analysis of growth hormone receptor signaling domains and their associated transcripts. Mol Cell Biol 2005;25:66–77. 29. Hill DJ, Riley SC, Bassett NS, Waters MJ. Localization of the growth hormone receptor, identified by immunocytochemistry, in second trimester human fetal tissues and in placenta throughout gestation. J Clin Endocrinol Metab 1992;75:646–650. 30. Simard M, Manthos H, Giaid A, Lefebvre Y, Goodyer CG. Ontogeny of growth hormone receptors in human tissues: an immunohistochemical study. J Clin Endocrinol Metab 1996;81:3097–3102. 31. Zogopoulos G, Figueiredo R, Jenab A, Ali Z, Lefebvre Y, Goodyer CG. Expression of exon 3-retaining and -deleted human growth hormone receptor messenger ribonucleic acid isoforms during development. J Clin Endocrinol Metab 1996;81:775–782. 32. Zogopoulos G, Albrecht S, Pietsch T, Alpert L, von Schweinitz D, Lefebvre Y, Goodyer CG. Fetal- and tumor-specific regulation of growth hormone receptor mRNA expression in human liver. Cancer Res 1996;56:2949–2953. 33. Goodyer CG, Figueiredo R, Krackovitch S, DeSouza Li L, Manalo J, Zogopoulos G. Characterisation of the growth hormone receptor in human dermal fibroblasts and liver during development. Am J Physiol Endocrinol Metab 2001;281:2313–2320. 34. Werther GA, Haynes K, Edmonson S, Oakes S, Buchanan CJ, Herington AC, Waters MJ. Identification of growth hormone receptors on human growth plate chondrocytes. Acta Paediatr Suppl 82 Suppl 1993;391:50–53. 35. Hill DJ. What is the role of growth hormone and related peptides in implantation and the development of the embryo and fetus. Horm

181

Res 1992(Suppl);1:28–34. 36. Laron Z. Laron syndrome (primary growth hormone resistance or insensitivity): the personal experience 1958–2003. J Clin Endocrinol Metab 2004;89:1031–1044. 37. Gluckman PD, Gunn AJ, Wray A, Cutfield WS, Chatelain PG, Guilbaud O, Ambler GR, Wilton P, Albertsson-Wikland K. Congenital idiopathic growth hormone deficiency associated with prenatal and early postnatal growth failure. The International Board of the Kabi Pharmacia International Growth Study. J Pediatr 1992;121:920– 923. 38. Rosenfeld RG, Rosenbloom AL, Guevara-Aguirre J. Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocr Rev 1994;15:369–390. 39. List EO, Coschigano KT, Kopchick JJ. Growth hormone receptor/binding protein (GHR/BP) knockout mice: A 3-year update. Mol Genet Metab 2001;73:1–10. 40. Freemark M. Ontogenesis of prolactin receptors in the human fetus: roles in fetal development. Biochem Soc Trans 2001;29:38–41. 41. Gluckman PD, Pinal CS. Regulation of fetal growth by the somatotrophic axis. J Nutr 2003;133:1741S–1746S. 42. Butler AA, Le Roith D. Control of growth by the somatropic axis: growth hormone and the insulin-like growth factors have related and independent roles. Annu Rev Physiol 2001;63:141–164. 43. Reynaud R, Saveanu A, Barlier A, Enjalbert A, Brue T. Pituitary hormone deficiencies due to transcription factor gene alterations. Growth Horm IGF Res 2004;14:442–448. 44. Nachtigal MW, Nickel BE, Klassen ME, Zhang WG, Eberhardt NL, Cattini PA. Human chorionic somatomammotropin and growth hormone gene expression in rat pituitary tumour cells is dependent on proximal promoter sequences. Nucleic Acids Res 1989;17:4327– 4337. 45. Rygaard K, Revol A, Esquivel-Escobedo D, Beck BL, Barrera-Saldana HA. Absence of human placental lactogen and placental growth hormone (HGH- V) during pregnancy: PCR analysis of the deletion. Hum Genet 1998;102:87–92. 46. Prager S, Wollmann HA, Mergenthaler S, Mavany M, Eggermann K, Ranke MB, Eggermann T. Characterization of genomic variants in CSH1 and GH2, two candidate genes for Silver-Russell syndrome in 17q24-q25. Genet Test 2003;7:259–263. 47. Eleswarapu S, Jiang H. Growth hormone regulates the expression of hepatocyte nuclear factor-3 gamma and other liver-enriched transcription factors in the bovine liver. J Endocrinol 2005;184:95–105. 48. Cornblath M, Ichord R. Hypoglycemia in the neonate. Semin Perinatol 2000;24:136–149. 49. Lind S, Rudling M, Ericsson S, Olivecrona H, Eriksson M, Borgstrom B, Eggertsen G, Berglund L, Angelin B. Growth hormone induces low-density lipoprotein clearance but not bile acid synthesis in humans. Arterioscler Thromb Vasc Biol 2004;24:349– 356. 50. Nishimura H. Atlas of Human Prenatal Histology. Igaku-Shoin, Tokyo, New York, 1983. 51. Hayashi Y, Wang W, Ninomiya T, Nagano H, Ohta K, Itoh H. Liver enriched transcription factors and differentiation of hepatocellular carcinoma. Mol Pathol 1999;52:19–24. 52. Parviz F, Matullo C, Garrison WD, Savatski L, Adamson JW, Ning G, Kaestner KH, Rossi JM, Zaret KS, Duncan SA. Hepatocyte nuclear factor 4alpha controls the development of a hepatic epithelium and liver morphogenesis. Nat Genet 2003;34:292–296. 53. Plumb-Rudewiez N, Clotman F, Strick-Marchand H, Pierreux CE, Weiss MC, Rousseau GG, Lemaigre FP. Transcription factor HNF6/OC-1 inhibits the stimulation of the HNF-3alpha/Foxa1 gene by TGF-beta in mouse liver. Hepatology 2004;40:1266–1274. 54. Itoh E, Iida K, Rincon JP, Kim DS, Thorner MO. Diurnal Variation in Growth Hormone Receptor Messenger Ribonucleic Acid in Liver

182

55.

56.

57. 58.

59.

60.

61. 62. 63.

64.

65.

Osafo et al.

and Skeletal Muscle of lit/+ and lit/lit Mice. Endocr J 2004;51:529– 535. Gohlke BC, Fahnenstich H, Dame C, Albers N. Longitudinal data for intrauterine levels of fetal IGF-I and IGF-II. Horm Res 2004;61:200– 204. Lewitt MS, Scott FP, Clarke NM, Wu T, Sinosich MJ, Baxter RC. Regulation of insulin-like growth factor-binding protein-3 ternary complex formation in pregnancy. J Endocrinol 1998;159:265–274. Gatford KL, Egan AR, Clarke IJ, Owens PC. Sexual dimorphism of the somatotrophic axis. J Endocrinol 1998;157:373–389. Choi HK, Waxman DJ. Plasma growth hormone pulse activation of hepatic JAK-STAT5 signaling: developmental regulation and role in male-specific liver gene expression. Endocrinology 2000;141:3245– 3255. Wiwi CA, Gupte M, Waxman DJ. Sexually dimorphic P450 gene expression in liver-specific hepatocyte nuclear factor 4alpha-deficient mice. Mol Endocrinol 2004;18:1975–1987. Flores-Morales A, Stahlberg N, Tollet-Egnell P, Lundeberg J, Malek RL, Quackenbush J, Lee NH, Norstedt G. Microarray analysis of the in vivo effects of hypophysectomy and growth hormone treatment on gene expression in the rat. Endocrinology 2001;142:3163– 3176. Ailhaud G, Grimaldi P, Negrel R. Cellular and molecular aspects of adipose tissue development. Annu Rev Nutr 1992;12:207–233. Nam SY, Lobie PE. The mechanism of effect of growth hormone on preadipocyte and adipocyte function. Obes Rev 2000;1:73–86. Louveau I, Gondret F. Regulation of development and metabolism of adipose tissue by growth hormone and the insulin-like growth factor system. Domest Anim Endocrinol 2004;27:241–255. Flint DJ, Binart N, Kopchick J, Kelly P. Effects of growth hormone and prolactin on adipose tissue development and function. Pituitary 2003;6:97–102. Poissonnet CM, LaVelle M, Burdi AR. Growth and development of adipose tissue. J Pediatr 1988;113:1-9.

66. Symonds ME, Mostyn A, Pearce S, Budge H, Stephenson T. Endocrine and nutritional regulation of fetal adipose tissue development. J Endocrinol 2003;179:293–299. 67. Clapp JF, III, Schmidt S, Paranjape A, Lopez B. Maternal insulinlike growth factor-I levels (IGF-I) reflect placental mass and neonatal fat mass. Am J Obstet Gynecol 2004;190:730–736. 68. Wei Y, Rhani Z, Goodyer C. Growth hormone receptor gene expression in human adipocytes. US Endocrine Society 86th Annual Meeting, 2004;P3–121. 69. Wei Y, Goodyer C. Growth hormone receptor (GHR) expression in the human adipocyte. US Endocrine Society 87th Annual Meeting. 2005;P3–83. 70. Zou L, Menon RK, Sperling MA. Induction of mRNAs for the growth hormone receptor gene during mouse 3T3-L1 preadipocyte differentiation. Metabolism 1997;46:114–118. 71. Wabitsch M, Braun S, Hauner H, Heinze E, Ilondo MM, Shymko R, De Meyts P, Teller WM. Mitogenic and antiadipogenic properties of human growth hormone in differentiating human adipocyte precursor cells in primary culture. Pediatr Res 1996;40:450–456. 72. Berryman DE, List EO, Coschigano KT, Behar K, Kim JK, Kopchick JJ. Comparing adiposity profiles in three mouse models with altered GH signaling. Growth Horm IGF Res 2004;14:309–318. 73. Li Y, Kelder B, Kopchick JJ. Identification, isolation, and cloning of growth hormone (GH)-inducible interscapular brown adipose complementary deoxyribonucleic acid from GH antagonist mice. Endocrinology 2001;142:2937–2945. 74. Gregoire FM, Smas CM, Sul HS. Understanding adipocyte differentiation. Physiol Rev 1998;78:783–809. 75. van der Eerden BC, Karperien M, Wit JM. Systemic and local regulation of the growth plate. Endocr Rev 2003;24:782–801. 76. Wang J, Zhou J, Cheng CM, Kopchick JJ, Bondy CA. Evidence supporting dual, IGF-I-independent and IGF-I-dependent, roles for GH in promoting longitudinal bone growth. J Endocrinol 2004;180:247–255.