THE GROWTH HORMONE-INSULIN-LIKE GROWTH FACTOR AXIS

GH is synthesised in the anterior pituitary, where it is store in secretory granules. It is the most abundant hormone in the pituitary accounting for 25% of the gland's hormones (1). GH is a single polypeptide chain of 191 amino acids with 2 disulphide bridges between amino acids 53-165 and 282-189, respectively. The GH variant 20-kDa accounts for 10-20% of the total content of pituitary GH, results from a different mRNA by alternative splicing of intron B that encodes amino acids in position 32-46. There is another form 17-kDa, which is produced by the same mechanism (2).

In man two GH genes are localised on chromosome 17. GH-N gene is expressed in the pituitary and encodes GH. It contains 5 exons and 4 introns. The GH-V gene is expressed in the placenta. There is considerable structural homology between GH, prolactin and hCG (3).

In the fetus GH is produced by the pituitary gland from the end of the 1st trimester of gestation and circulating concentrations of GH can reach relatively high levels (4). The biological role of GH during fetal life has been the source of many studies. Early studies with anencephalic foetuses and neonates suggested that GH was not primarily involved in human fetal growth (5). However, the identification of the GH receptor (GHR) and its widespread localisation in fetal tissues support a functional role for both GH and its receptor (6-9). There is growing evidence for a role of GH in fetal development although it may contribute only to approximately 20% of fetal size (10).

Growth Hormone Receptor (GHR)

GHR expression begins in fetal life. Hill et al demonstrated immunoreactive GHR in 14-16 weeks fetuses (8) and in 15-20 week human fetuses Werther found growth plate chondrocytes to express GHR (9). Human fetal GHR expression appears to be regulated by the V3 promoter whereas in postnatal life GHR expression is regulated by hepatic V1 and V4 promoters and concomitant in vitro studies have demonstrated that these GHR are functional (10, 11).

The cDNA encoding the GHR was cloned in 1987 (6). In humans the GHR is composed of 620 amino acid (aa) containing an extracellular (246 aa), a transmembrane (24 aa) and cytoplasmic (350 aa) domains (7). In addition to the membrane-bound receptor, a soluble GH binding protein, identical to the extracellular domain of the full-length receptors has been isolated in several species including humans (12). The availability of GHR cDNA and the identification of key intracellular proteins, tyrosine kinases, phosphatases and other transducers have allowed major advances in the understanding of the molecular aspects and the signalling pathways used by this receptor (13).

Regulation of GH secretion

Several factors including hormones, neurotransmitters and metabolic modulators are known to regulate GH in man. There is extensive literature on the role of each of these factors that have been reviewed elsewhere (14) and in chapter 5c of EndoText. The best charcterised are summarized in Table 1, including systemic factors and those locally produced in the hypothalamus which are the primary regulating factors, GH releasing hormone (GHRH) and somatostatin (SRIF) (14). Initially it was thought that GH was only under the regulation of these two hypothalamic factors, however a recently identified small peptide hormone, ghrelin, also plays a role in GH secretion. The main regulatory factors are discussed in some detail below.

GHRH

GHRH is the most important regulator of GH under physiologic conditions (15). GHRH stimulates GH synthesis by increasing both GH gene transcription and GH release (16-18). GHRH binds to its specific receptor, the GHRH receptor (GHRHR) which was cloned from a pituitary cDNA library (19). The GHRHR, is a member of the seven transmembrane domain G-protein receptor superfamily (20). The critical role of GHRHR in human growth has been now fully documented with 1) the identification of functional receptors in fetal pituitary gland (21) and 2) the demonstration of molecular defects in the extracellular domain that are associated with profound GH deficiency and severe growth retardation (22, 23).

Somatostatin

Krulich et al postulated the existence of somatostatin in 1968 using hypothalamic extracts that were able to inhibit the GH secretion (24) but it was finally isolated by Brazeau et al in 1973 (25) and subsequently the gene sequence was characterised in 1984 (26). Somatostatin binds to a family of specific receptors and inhibits adenyl cyclase via Gi, with net Ca influx reduction (14). Five somatostatin receptors which are regulated in a tissue specific manner have been cloned (27, 28). All subtypes are expressed in pituitary tumors and in normal fetal pituitary tissue (28, 29). It is worth noting that the primary function of somatostatin is to inhibit GH release but not its synthesis.

GH secretagogues (GHSs)

Met-enkephalins are able to stimulate GH secretion in a specific manner (30). This observation led in the 1990s to the design of synthetic analogues that were not opiodergic but which were active on GH release (30). These growth hormone releasing peptides, such as GHRP-6 and GHRP-2, are quite potent stimulators of GH release compared to natural met-enkephalins precursors (31-33).

Ghrelin

Ghrelin is an acylated peptide (MW of 3.3-kDa) and was first identified and characterised from the rat stomach (34). It is the endogenous ligand for the GH secretagogue receptor. Ghrelin is involved in a novel system for regulating GH release in a dose dependent manner in humans. Takaya et al demonstrated that the GH response to 0.2-μg/kg ghrelin was similar to 1.0 μg/kg GHRH and had a similar effect compared to GHRP-2, one of the most potent GH secretagogues (GHSs), indicating that per mol ghrelin is more potent for GH release (35, 36). Ghrelin levels in healthy lean children are similar to healthy lean adults but are reduced in obese children (37). Ghrelin levels are independent of gender and pubertal status but are negatively associated with obesity and insulin levels (37).

Thyroid hormones

GH spontaneous nocturnal secretion is low in hypothyroidism and hyperthyroidism (38). The rate and the amount of GH released are reduced in adolescent with untreated thyrotoxicosis compared with normal controls GH determined by spontaneous GH secretory profile (39). It has been proposed that the reduced GH release stimulated by GHRH in hyperthyroidism may be explained in part by an increase in hypothalamic somatostatin tone with a concomitant decrease in GHRH (40).

Hypoglycemia

GHRH stimulation and insulin-induced hypoglycaemia exert additive effects on GH release (41), which is consistent with a proposed somatostatin withdrawal during hypoglycaemia. The clinically inhibitory effect of glucose on GH release may be due to a discharge of hypothalamic somatostatin (14, 41, 42)

Glucocorticoids

The effect of glucocorticoids on growth is well established and many studies have examined the GH secretory profiles in prepubertal and pubertal children undergoing long-term glucocorticoid treatment or exposed to excess cortisol secretion (14, 43, 44). In these children GH responses assessed by GH spontaneous secretion and various pharmacological stimuli are decreased (14). The blunted GH response to stimulation tests in conditions of chronic exposure is well documented; in contrast the acute administration of glucocorticoids to normal subjects induces a transient increase in plasma GH levels reflecting a dual mode of action on GH secretion (44).

Sex steroid

Androgens without the actions of GH are insufficient in the human to drive the fully normal tempo of puberty, as observed in hypopituitary boys who were replaced with testosterone but not with GH showing a long pubertal growth period (45). Studies in healthy pubertal boys have demonstrated the close relationship between rising serum androgen concentrations and increased GH peak amplitude (46). Serum GH concentration rises throughout puberty in both sexes and in healthy girls the increase in GH concentrations is proportional to the increase in serum estradiol concentration (47).

The IGF system

The IGF system is composed of two IGF peptides, two specific receptors, a family of binding proteins and a glycoprotein named the acid-labile subunit (Figure 1). The insulin-like growth factors (IGF-I and IGF-II) are single chain polypeptides with structural similarity to proinsulin. In 1957 Salmon and Daughaday identified the "sulphation factor", later renamed "somatomedin" in the 1970s and referred as IGFs nowadays (48, 49). IGFs are essential regulators of normal fetal and postnatal growth (50-54). IGF-I and IGF-II, as many other growth factors, exert physiologic effects virtually in every organ and tissue during fetal and postnatal life (50-54). The development of IGF knockout models and the identification of two patients with IGF-I gene defects have provided clear evidence on the important role that circulating and locally produced IGF-I play in human and animal fetal and postnatal growth and development (54-56).

Figure 1. Schematic representation of the GH-IGF axis.

Schematic representation of the GH-IGF axis.

IGF-I is a basic peptide with 70 amino acids and a molecular weight of 7.6-kDa (57). The human IGF-I peptide is the product of a single gene (~ 90-kb), that is located on the long arm of chromosome 12. It contains six exons and two promoters (57). The IGF-I transcribed by exon 1 is ubiquitously expressed, while the IGF-I transcribed by exon 2 is found exclusively in the liver, increasing with the onset of GH dependent linear growth (57). The expression of the IGF-I gene is determined by the activity of its promoters and by transcription factors that stimulate or inhibit their activity. The most potent regulator of IGF-I expression in postnatal life is GH (57). On the other hand IGF-I mediates growth hormone negative feedback (58). Nutritional status and supply of dietary energy and protein are also regulators of IGF-I and possibly the main regulators of IGF-I expression in fetal life (59, 60).

IGF-II is a neutral peptide with 67 amino acids and molecular weight of 7.4-kDa (57) and is the product of a single gene (~ 30-kb) located on the short arm of chromosome 11. It contains nine exons (57, 61) with four different promoters (57). The mature IGF-II peptide is encoded by exons 7, 8 and 9. IGF-II is an imprinted gene, with only the paternal allele being expressed while the maternal allele is inactive. IGF-II gene is linked to H-19 gene, a tumour suppressor gene. Overall it is considered that the regulation of IGF-II expression is GH-independent.

The IGF-1 Receptor (IGF-1R) is a hetero-tetrameric glycoprotein of two -subunits (61). The human IGF-1R is the product of a single-subunits and two  gene (~100-kb), that is located on chromosome 15 and has 22 exons (61). The -subunit corresponds to the extracellular cysteine-rich domain necessary for -subunit that contains a shortIGF-I recognition and binding and the extracellular domain, a transmembrane domain, and an intracellular domain with tyrosine kinase activity. Signaling activation of the IGF-1R is regulated through binding of the IGF ligands. IGF-I binds to the IGF-1R with high affinity, with dissociation constants in the range of 0.2-1.0 nM. The affinity of IGF-II for the IGF-1R is several folds lower, and the affinity of insulin for the IGF-1R is 100-fold lower (61, 62).

IGF-binding proteins (IGFBP) and acid-labile subunit (ALS): The IGFs are found in association with specific IGFBPs in blood and extracellular fluids. Six IGFBPs have been characterized IGFBP-1 through to IGFBP-6. All six IGFBPs have at least 50% homology among themselves and 80% homology among different species and are products of different genes with different chromosomal location (61). Most of the homology is found in the amino-and carboxy-terminal regions with a distinct mid-region among the six IGFBPs (Baxter, 1994). The main functions of this family of proteins are 1) To extend IGFs half-life in the circulation 2) To transport the IGFs to the target cells and to modulate the biological actions of the IGFs (61).

Most of the IGFs in circulation are found in a 150-kDa or ternary complex consisting of IGF-I or IGF-II, IGFBP-3 and acid-labile subunit (ALS) (63, 64). The human ALS gene is located on chromosome 16 and encodes a mature protein of 578 amino acids (65). GH is the main regulator of the ternary complex. In humans, ALS is undetectable in fetal serum at 27 weeks of gestation, but IGFs and IGFBPs can be detected from the first trimester of gestation. Serum IGFs and IGFBPs are present from early fetal life as determined from cord blood and are related to fetal size (66, 67). ALS is present at the end of fetal life, and increases five-fold from birth to puberty with little change during adulthood (68).

Recently, the development of an animal model for ALS deficiency (the ALS-KO mouse) and the identification of patients with inactivating mutations in the IGFALS gene have provided the opportunity to assess the physiological role of this protein in the circulating IGF system (69).

Domené et al. (70) described a 17-year-old boy with delayed onset of puberty and slow pubertal progress. Childhood medical history was unremarkable and psychomotor and neurological development was normal. He was first referred at 14 years of age for evaluation of growth and pubertal delay, at which time his height was 145.2 cm (-2.05 SDS) and his weight was 35.9 kg. He was Tanner stage 1 for both sexual development and pubic hair. Both testes were 3 ml in volume. Bone age was 12.5 years. Growth hormone responses to provocative tests were normal, but there was marked reduction of both IGF-I and IGFBP-3, which remained unchanged after stimulation with growth hormone. ALS was undetectable in the serum before and after growth hormone stimulation. Sequencing of exons 1 and 2 of the IGFALS gene revealed a 1-bp deletion, 1338delG, which involved 1 of 5 consecutive guanines at positions 1334-1338. This frameshift point mutation resulted in the substitution of a lysine for a glutamic acid at codon 35 (E35K) and the appearance of a premature stop codon at position 120 of the precursor form of the acid-labile subunit. The patient was thought to be homozygous for the mutation; parental DNA was not available because the patient had been adopted. After 6 months of treatment with recombinant human growth hormone, reduction in the subscapular skinfold thickness was observed but there was no beneficial effect on either growth velocity or the serum levels of IGF-I, IGFBP-3, and ALS. Three additional patients have been identified with ALS gene defects associated to a diagnosis of idiopathic short stature (ISS; 71). These findings underlie the important physiological role of ALS in the maintenance of the circulating IGF-I reservoir and in maintaining normal serum concentrations of IGF-I and IGFBP-3. However, the clinical cases reported to date suggest that the ternary complex has a modest role in the regulation of linear growth.

The "somatomedin hypothesis"

The original "somatomedin hypothesis" postulated that somatic growth was controlled by pituitary GH and mediated by circulating IGF-I expressed exclusively by the liver (47). The discovery that most tissues produce IGF-I in the late 1980's supporting a role of autocrine/paracrine IGF-I led investigators to modify the original hypothesis to what is known today as the "dual effector" theory (72). Recent experiments using gene deletions and transgenic technologies have revealed new information that again has led investigators to revisit both hypotheses (72-75). These experimental studies have shown that the liver is the principal source of IGF-I in the circulation but is not required for postnatal body growth. This finding indicates that autocrine/paracrine IGF-I but not liver-derived IGF-I (endocrine IGF-I) is the major determinant of postnatal body growth (72-75). However, lack of liver-derived IGF-I results in disproportional organ growth. An early study showed that GH increases liver size in proportion to body weight (76). Thus, increased liver size in the LI-IGF-I -/- mice (inactivation of IGF-I gene in the liver) is a direct result of the increased GH levels. Decrease kidney size in LI-IGF-I -/- mice is also a result of the decreased serum IGF-I levels (76).