Pituitary Growth Hormone Secretion and Cell Growth Hormone Production: Regulation of Their Secretion and Their Signaling Pathways

Abstract

Growth hormone (GH) performs very diverse functions in the organism, and this is the reason by which the regulation of the secretion of this hormone is very complex; although the primary regulators are growth hormone-releasing hormone (GHRH) and somatostatin, it is in turn regulated mainly by adrenergic and cholinergic pathways, and other factors can act directly on its secretion, particularly on the somatostatin, thus affecting the pituitary secretion of GH. In this chapter, we will analyze the transcription of GH gene and how GH release is affected by different neurotransmitters, metabolic substrates, feeding and fasting, and other hormones, placing special emphasis on why pituitary secretion of GH is sexually dimorphic.

1. Introduction

Growth hormone (GH) is produced mainly in the anterior pituitary gland (AP), but its expression also takes place in many other territories, such as the brain, the gonads, retina, adrenals, where this hormone, or its derivatives formed by enzymatic cleavage, performs very diverse and specific roles, although how GH is regulated at these levels is not well known.

In the anterior pituitary gland (AP), GH is the most abundant hormone, and it is synthesized in eosinophilic cells, the somatotrophs developed due to the action of the transcription factor Prop-1 (homeobox protein prophet of Pit-1), whose expression leads to the pituitary development of Pit-1-specific lineages. The Pit-1 t transcription factor determines the specific cellular expression of GH (and also prolactin (PRL) and thyroid-stimulating hormone (TSH)).

GH can be found in the pituitary as early as 8 weeks of fetal development, being a key factor for the development and maturation of multiple organs and tissues, such as the brain. This fetal GH does not seem to play any role in the growth of the fetus, despite the high growth velocity during the fetal period. In fact, anencephalic fetuses are of normal height at birth. A high growth rate also occurs after birth, up to 6–10 months of age, after which growth slows down; these changes after birth are shown in Figure 1.

Figure 1.

Changes in growth velocity throughout life. Note that when puberty approaches and sex steroids increase, they act on the synthesis and release of pituitary GH, so the growth velocity increases with respect to childhood. However, from the end of puberty, GH secretion progressively decreases with aging.

Growth hormone secretion is pulsatile in nature; this is related to the optimal induction of physiological effects at the peripheral level: target tissues for GH appear to be more sensitive to the frequency at which this hormone arrives during a period of time than to the total amount of GH secreted during a similar period [1]. A clear pulsatility of GH only appears after birth; the majority of the GH pulses are associated with slow wave sleep, a period in which the amplitude of GH secretion reaches the maximum [2]. However, sleep processes do not appear to exert a predominant influence on GH release, since when sleep and circadian processes are misaligned, the blunting of the sleep-related GH pulse is counteracted, as in individuals deprived of sleep, by a compensatory mechanism that promotes GH pulses during wakefulness [3].

GH secretion is sexually dimorphic from the puberty, due to gender-related free estradiol (fE₂) levels in the brain [4]. In humans, plasma fE₂, but not testosterone (T), levels were shown to be strongly correlated with total and pulsatile GH release rates [5]. However, it cannot be ruled out whether there is an imprinting effect of sex steroids on the hypothalamic structures that govern the underlying hypothalamic-somatotrophic rhythm (HSR), as occurs in rat [6]. In women, the episodic secretion of GH is more frequent, but of a lower amplitude, while plasma GH values during trough periods are slightly higher than in men. Therefore, the fraction of GH secreted in peak episodes is lower in women, although it changes according to the phase of the menstrual cycle once puberty is established, reaching higher values in the late follicular phase of the menstrual cycle.

2. Regulation of pituitary GH secretion

In 1985, Plotsky and Vale demonstrated in rats that the episodic secretion of GH was under a dual hypothalamic control exerted by growth hormone-releasing hormone (GHRH) and somatostatin (SS) [7]. According to this model, the determinant of pulsatile GH secretion would be the rhythmic alternate release of hypothalamic GHRH and SS. The existence of a similar intrinsic hypothalamic-somatotroph rhythm (HSR) that governs the pituitary GH secretion was also demonstrated by our group in normal humans [8]. Intrinsic HSR most likely remains functionally active throughout life. However, plasma GH levels are highly age-dependent and show a different pattern of secretion in men and women. Both age and gender physiologically modify the standard pattern of activity, as plasma GH changes reflect. In addition, a number of factors, such as neurotransmitters, other hormones, and metabolic intermediates (amino acids, glucose, and free fatty acids), mainly acting at on the hypothalamic or pituitary level, or both, but also on the pituitary, can modify, both acutely and chronically, the pattern of GH release. Therefore, GH control has traditionally been classically considered as extraordinarily complex, although its complexity has increased over time due to the discovery of new factors that contribute significantly to modify the pituitary secretion of GH, as we will see more ahead. This also reflects the complexity of the multiple actions that GH exerts in the organism, far beyond its role in growth.

2.1 GHRH

Reichlin [9] was the first to propose the existence of a hypothalamic GHreleasing factor. However, it took several years to isolate and characterize this putative factor [10, 11]. The initial isolation of GHRH was facilitated by the existence of human tumors ectopically producing growth hormone-releasing activity [12, 13].

GHRH is a peptide that belongs to the brain-intestinal peptides family, the glucagon secretin family. Two main GHRH forms of GHRH, composed of 40 and 44 amino acids, have been described in the human hypothalamus, occurring as a result of the alternative processing of RNA during the transcription of a GHRH gene located on chromosome 20. These are the result of alternative processing of RNA from the transcript of a single GHRH gene located on chromosome 20. Once it was demonstrated that the GH-releasing activity of GHRH was located between amino acids 1–29, it was possible to produce shorter forms of the peptide (e.g., GHRH-29, known as GRF-29) to be used in clinical applications, such as a provocative test. Since the removal of amino acids 30–44 does not affect the GH-releasing activity of the peptide, shorter forms (e.g., GRF-29) have been synthesized for clinical purposes. On the contrary, modifications at the N-terminal end of the molecule markedly decrease its biologic activity. The half-life of GRF-29 in plasma is very short, about 2 minutes, as it is very rapidly inactivated by circulating proteases. This has been led to the development of potent GHRH agonists and antagonists., as it will be shown later.

GHRH-producing neurons involved in GH control are located primarily in the arcuate nucleus of the hypothalamus from where they reach the median eminence and release the peptide into the primary plexus of the hypothalamic-pituitary portal vascular system. In the pituitary, GHRH binds to its receptor on somatotrophs and induces a rapid, dose-dependent release of GH, but also the transcription of the GH gene as well as the differentiation, growth, and proliferation of these cells.

In addition to this, GHRH-producing cells have also been found in other hypothalamic and extra-hypothalamic structures. These cells, however, do not project to the median eminence but to several other hypothalamic and extrahypothalamic regions. This suggests that apart from its GH-releasing activity, the peptide also plays a neuromodulator role in the central nervous system. In addition, there is GHRH production in numerous tissues and organs, such as the gonads.

GHRH is encoded by a gene located in chromosome 20 (20q11.2), which consists of five exons separated by four introns. Exon I is not coding; exon II codes for the signal peptide and a small N-terminal connecting peptide; exon III encodes the majority of the mature GHRH molecule (including the biologically active peptide); exon IV encodes most of the C-terminal peptide; and exon V encodes the rest of the C-terminal peptide and also contains 3′ untranslated sequences [14] (Figure 2).

Figure 2.

Structure of the GHRH gene and transcription of GHRH. As the figure shows, the gene-encoding GHRH is composed by five exons (I–V) and four introns (A–D). GHRH is synthetized as a precursor (pre-pro-GHRH) formed by 108 amino acids (including the signal peptide), which is proteolytically cleaved and gives rise to the mature GHRH (blue rectangle) together with a C-terminal peptide formed by 31 amino acids. As the figure shows, the precursor of GHRH is produced by exon III and a small portion of exon IV (black arrows 3 and 4).

A variant of the GHRH precursor has been described. It is formed by RNA processing that excludes three nucleotides at the beginning of exon V, leading to the absence of serine-103 in the precursor molecule. This precursor will undergo identical processing to that of the 108 amino acids precursor, with the only difference that in this case the C-terminal peptide will be made up of only 30 amino acids [14].

An interesting aspect, little clarified, is how the transcription of the GHRH gene is regulated. KO mice for the homeobox transcriptional factor Gsh-1 exhibit a dwarf phenotype and abolished GHRH expression [15], and Gsh-1 binds on multiple binding sites of GHRH gene promoter. Co-expression of the transcription factor conditional cAMP response element binding (CREB) protein significantly enhances Gsh-1-induced expression of the GHRH gene, suggesting a cooperative role for the coactivator protein. The Gsh-1 homeobox protein is key for the expression of the GHRH gene.

And what is the role of CREB in the transcription of GHRH gene? In mice in which CREB is lost in the brain, but not in the pituitary, the amount of GHRH peptide is reduced, indicating that CREB is required for efficient production of GHRH in the hypothalamus [16]. This would explain the relationships between CREB and Gsh-1 for hypothalamic GHRH transcription.

Another GHRH gene transcription factor is the nuclear factor of activated T cells (NFAT). Site-directed mutagenesis experiments demonstrated the direct binding of NFAT at five sites of the GHRH promoter, suggesting that NFAT is involved in the depolarization-induced (with high potassium) transcriptional activation of GHRH gene in the hypothalamic neurons producing it [17].

Apart of its hypothalamic production, GHRH is a peptide produced in many human tissues and organs, where it plays a very important physiological and etiopathological role in addition to its main role in the synthesis and release of GH.

2.2 Somatostatin (SS)

At the end of the 1960s, McCann’s group observed that bovine hypothalamic extracts were able to inhibit the release of GH from pituitary cultures. Five years later, Guillemin’ group [18, 19, 20] reported the identification of the factor responsible for that effect: GH-release-inhibiting factor or SS. Several studies soon demonstrated that SS was not only a GHrelease-inhibiting factor but also a strong inhibitor of a number of secretory events, endo- or exocrine, whose analysis is beyond the objective of this chapter.

Although SS is a tetradecapeptide (SS-14), a form SS-28 has been found in pituitary portal blood, also capable of inhibiting pituitary GH release. Structurally, SS is a tetradecapeptide, although a 28-amino-acid form is also present in pituitary portal blood. Both molecular forms apparently have similar GH-release-inhibiting activity.

SS-producing neurons involved in GH control are found mainly are mainly located in the anterior periventricular and paraventricular nuclei of the hypothalamus. Their axons project to the median eminence, where SS is released into the hypothalamic-pituitary portal circulation to antagonize in a dose-dependent manner the GH-releasing effect of GHRH, and also, although in less degree, the proliferative activities of this GH-releasing peptide.

As with GHRH, SS is synthesized in the form of a precursor, pre-pro-somatostatin, a 116 amino acid peptide (including the 24 amino acid signal peptide) whose biologically active portion is at the C-terminal end. The gene that codes for SS is located on chromosome 3 and consists of two exons separated by an intron [21, 22]. In it, a cAMP response sequence is found between nucleotides −48 and −41 [23]. The transcription of the gene is stimulated by CREB, phosphorylated after an increase in cAMP levels, and the subsequent activation of protein kinase A (PKA). The coding sequence has a high degree of conservation in exons.

The proteolytic processing of pro-somatostatin gives rise to two physiologically important variants, SS-14 and SS-28. SS-14 is a tetradecapeptide that is formed after cleavage by endoproteases of the precursor between amino acids 101 and 102, while SS-28 corresponds to the sequence of the SS-14 form with an N-terminal extension of 14 amino acids [24] (Figure 3). The proteolytic cleavage of pro-somatostatin in the basic domain (Arg/Lys) or the monobasic domain (Arg) of the C-terminal end gives rise to SS-14 and SS-28, respectively (Figure 3). Both molecular variants are encoded by exon 2.

Figure 3.

Proteolytic processing of pre-pro-somatostatin. Proteolytic processing of pre-pro-somatostatin releases the signal peptide (formed by 24 amino acids) and then the connection peptide (green rectangle, formed by 64 amino acids). As indicated SS-28 is formed by the 14 amino acids extension (yellow rectangle) to the N-terminal end of SS-14. A: monobasic domain. B: basic domain.

Both SS-14 and SS-28 are widely distributed in the body, being particularly abundant throughout the brain, but also in the gastrointestinal tract (particularly in the pancreas and stomach), which may be related to its inhibitory functions of endocrine and exocrine secretions. SS also acts as a neurotransmitter and regulator of the immune system. Most likely, these are the reasons for the existence of five receptors (SST1 to SST5) with different localizations and biological activities. Since the fact that SS has anti-proliferative and anti-angiogenic effects and given that its half-life is very short (1–3 min), as well as the wide spectrum of its biological responses, led to the need to develop selective stable receptor agonists that are bind strongly to selective SS receptor subtypes, therefore, in addition to blocking GH secretion (e.g., treatment of Acromegaly), these SS agonists are being used for the therapy of neuroendocrine tumors, diabetic complications (nephropathy, retinopathy), anti-neoplastic therapies. More recently, it has been shown that these agonists have anti-inflammatory and anti-nociceptive effects [25].

Before reviewing how GHRH and SS interact in the control of pituitary synthesis and release, we will describe how the pituitary receptors for these hormones are stimulated. Most likely this is different than how they are stimulated in other tissues in which they are expressed.

2.2.1 Pituitary GHRH receptor

The GHRH receptor (GHRHR) is a 423 amino acid protein (including the signal peptide) that belongs to a subfamily of G-protein-coupled receptors [26, 27]. Its producing gene is located to the short arm of chromosome 7 (7p13–p21) near the epidermal growth factor receptor (EGFR) gene [28].

The gene has a very complex structure [26]. Unlike other G-protein-coupled receptors, which do not have introns, in this gene there is a coding sequence interrupted by several introns. An alternative ribonucleic acid (RNA) processing has been described that results in the insertion of 41 amino acids just before the beginning of the sixth transmembrane domain. This type of processing does not appear to be exclusive to GHRHR [29], although its physiological significance is unknown.

Like the rest of the receptors that use G proteins for intracellular signaling, the GHRHR has seven transmembrane hydrophobic domains, linked to each other by six loops, three intra-cytoplasmatic and three extracellular.

The N-terminal extracellular domain of GHRHR contains a site for N-glycosylation as well as six cysteine residues and an aspartate residue that are conserved in this receptor family, while the third intracellular loop and C-intracellular domain contain several potential phosphorylation sites, which may regulate signaling and receptor internalization.

GHRHR mRNA has been detected in the pituitary, gonads, placenta, kidney, hypothalamus, and many brain regions, although it is expressed predominantly in the anterior pituitary due to its dependence of Pit-1 (also known as GHF-1).

GHRH interaction with its GHRHR on the somatotroph leads to the activation of the associated G_s protein, allowing the α_s subunit to separate from the dimer ßγ. The activation of G_s protein will then have a double effect: (1) It induces the release of GH stored in secretory granules by allowing the entry of Ca⁺⁺ into the cell, and (2) the α_s subunit activates the adenylyl cyclase (AC), which leads to increased cellular levels of cAMP [26, 30, 31]. This, in turn, activates the protein kinase A (PKA), which induces a cascade of phosphorylation and activation of the transcription factor CREB [32]. The activation of CREB induces the pituitary-specific transcription factor Pit-1 [33, 34, 35]. Therefore, Pit-1 represents a pathway by which GHRH can increase GH synthesis in the pituitary. Interestingly, Pit-1 also regulates GHRHR synthesis, as if there were some kind of feedback positive between them, and also induces its own expression. In fact, the GHRHR is not expressed in the pituitary of dw/dw mice that lack functional Pit-1 [30]. In turn, the non-activation of CREB leads to somatotroph hypoplasia and dwarfism in mice.

Other signaling pathways activated after the GHRH-GHRHR interaction involve the Phospholipase C/Inositol phosphatide/Protein kinase C (PLC/IP/PKC) pathway, which is perhaps also responsible for the increase of Ca⁺⁺ for GH secretion in somatotrophs. In this case, the Ca⁺⁺ would come from the endoplasmic reticulum. This effect would occur through the stimulation of phospholipase C (PLC) through the complex G_βγ of heterotrimeric proteins G. Activated PLC produces diacylglycerol (DAG) and inositol triphosphate (IP3) that leads to the release of Ca⁺⁺ from the endoplasmic reticulum. However, this effect has not been seen in some inferior species [36]; another signaling pathway observed in a wide variety of cell types as a mediator of GHRH effects is the mitogen-activated protein kinase (MAPK) pathway, which may mediate the effects of GHRH on somatotroph cell proliferation [37], although it cannot be ruled out that the proliferative effect is mediated by c-fos activated by CREB. These pathways activated by the interaction GHRH-GHRHR are shown in Figure 4.

Figure 4.

Interaction GHRH-GHRHR at the pituitary level and synthesis and secretion of GH. After interacting with its receptor, GHRH activates the associated G_s protein, and the α_s subunit is separated from the ßγ dimer (1). This allows: (1) the entry of Ca++ into the cell (1′), and the activation of adenylyl-cyclase (AC) (1) leading to increased formation of cAMP from ATP (2). cAMP activates protein kinase A [PKA] that phosphorylates CREB (3). Phosphorylated CREB translocates to nucleus (4) and induces Pit-1, which, in turn, leads to the transcription of the GH gene (5) and also regulates the synthesis of GHRHR (5′). On the other hand, the G_βγ complex of heterotrimeric proteins G activates PLC (6) producing DAG and IP3 (7) that leads to the release of Ca⁺⁺ from the endoplasmic reticulum (7′) for the secretion of GH. Another mediator of the effects of GHRH is the mitogen-activated protein kinase (MAPK) pathway, which may mediate the effects of GHRH on the proliferation of somatotrophic cells (8).

2.2.2 Pituitary SS receptor

The wide variety of actions of SS and the different organs and tissues in which this hormone acts explain the existence of different types of receptors for this exocrine and endocrine inhibitory hormone. The five types of receptors for SS (SSTR1 to SSTR5) belong to the DRY family of G-protein-coupled receptors, whose common characteristic is the presence of an Asp-Arg-Tyr sequence, located near the boundary between the third transmembrane domain and the second intracellular loop, which is very important for transmission of the SS signal [38, 39]. Each of the receptors is encoded by a different gene, located on different chromosomes, and their coding region lacks introns (as is the case with most G-protein-coupled receptors).

SS receptors show a homology between them that varies between 45% and 61%, and 100 residues are the same in all of them. The highest degree of homology occurs in the sequences corresponding to the seven transmembrane helices, while the NH2 and COOH ends are the most divergent both in sequence and in length [39].

The SSTR2 receptor is the prototype of the SS receptors and is the one that mediates the inhibition that SS exerts on GH secretion in the pituitary gland [40]. This receptor has the same affinity for SS-14 as for SS-28 [39].

The binding of SS with its receptor induces the activation of the related G protein (Figure 5), which in the case of the SSTR2 receptor is a Gi protein. The activation of the Gi protein will produce an inhibition of the activity of AC, a reduction in the entry of Ca⁺⁺ through voltage-gated channels, and the appearance of rectifying potassium currents [39, 41], which leads to hyperpolarization of the somatotropic membrane. The net product of these actions would be the inhibition of the transcription of CREB-dependent genes [42], which counteracts the stimulatory effect of GHRH. However, in the control of GH, SS acts basically by inhibiting the release of the hormone, while its effect on the synthesis of GH would be of minor importance. SS also exerts anti-proliferative effects on mediated somatotrophs, through the activation of a tyrosine phosphatase dependent on the SSTR2 receptor [43, 44].

Figure 5.

Interaction of SS-SSTR2 at the pituitary level and its effects on GH secretion. In the pituitary gland, binding of SS to its pituitary receptor SSTR2 (1) activates (2) the Gi-related inhibitory protein. This leads to inhibition (3, −) of AC. Consequently, ATP cannot give rise to cAMP (4, =). The lack of the necessary cAMP prevents (5, =) the activation of PKA, which then cannot initiate the phosphorylation cascade (6, =) that would activate CREB, and non-phosphorylated CREB cannot translocate to the cell nucleus (7, =) to initiate the transcription of specific genes, including GH (8, =). Then, the secretion of GH would be blocked (=). Furthermore, activated Gi blocks (9, −) the channels involved in the entry of Ca⁺⁺ into the cell (X). This Ca⁺⁺ is necessary for the release of GH. Moreover, activated Gi leads to the appearance (10, +) of rectifying potassium currents, which leads to hyperpolarization of the somatotropic membrane. Overall, SS blocks GH secretion and partially affects GH synthesis. +: stimulates; −: inhibits; =: interrupts or cuts; X: disrupts; blue arrows: stimulates. Red arrows: inhibits.

SS receptors undergo desensitization phenomena after prolonged exposure to agonists [39], which seems to depend on the phosphorylation of serine and threonine residues, located in the third intracytoplasmic loop, by the action of the enzyme BARK (ß-adrenergic receptor kinase) [39].

Once the basic components that generate a secretory rhythm of GH have been established, the task should be to understand how this rhythm may begin, how it can be modulated by factors external to said system, and what could be the relationships between its basic components: GH, GHRH, and SS. That is, is there a signal associated with GHRH that causes the interruption of SS release or is SS release linked to the termination of GHRH secretion, as initially thought? Is the end of GHRH secretion directly dependent on the increase in GH produced by the peptide, or is there a GH-associated signal that stimulates SS release and blocks GHRH secretion?

Before answering these questions, there is the need to analyze how the pituitary transcription of GH is regulated by transcription factors, as well as the structural organization of Pit-1, the main one of these factors. After that, the mechanisms that regulate the hypothalamic secretion of GHRH and SS will be reviewed. Given that they are neurosecretory products, it seems logical that neurotransmitters play the main role in such regulation.

2.3 Regulation of the transcription of the gene GH-N

Pit-1 is the most important of the transcription factor involved in controlling the expression of the GH-N gene. It is a highly conserved protein, belonging to the family of transcription factors with POU domains that are a subclass of the homeobox genes involved in cell development and differentiation processes [42, 45]. In man, the gene encoding Pit-1 is located on chromosome 20 and is made up of six exons and five introns.

Pit-1 is not only expressed in the pituitary but also in many other territories where it may be involved in the control of cell proliferation. In the pituitary, Pit-1 not only controls the expression of GH and PRL, but also that of the ß chain of TSH, that of the GHRH gene, and, curiously, that of its own gene. Pit-1 is also, as we have already described, a key factor for the development, differentiation, and survival of somatotrophs. The absence of Pit-1, due to mutations, produces alterations in cell development and hormonal synthesis [46, 47]. On the contrary, the overexpression of this factor has been related to the appearance of pituitary adenomas [48, 49].

Basal transcription of the Pit-1 gene is self-regulated by two binding sequences for Pit-1 itself (Figure 6): an activating sequence located at position −55 and an inhibitory sequence, located at position +15. In addition, the transcription of this gene would be regulated by factors capable of increasing the levels of cAMP and activating cAMP response-element-binding (CREB). Likewise, the activation of the protein kinase C-dependent signaling pathway seems to play an important role, having described the existence of a response element to AP-1 (activating protein-1), which exerts an inhibitory effect on transcription [50].

Figure 6.

Schematic structure of the promoter of Pit-1. The human Pit-1 promoter is Pit-1-dependent and autoregulated. Activating sequence of basal Pit-1 self-transcription (blue figure, +) is located at position −55 bp in the promoter, while a second sequence of Pit-1, in this case inhibitor of self-transcription (red figure, −), is located at position +15 bp. The transcriptional activity is negatively regulated by Oct-1 and mediated by the octamer-binding site OTF (red figure). Intracellular levels of cAMP positively affect Pit-1 transcription acting through two cAMP-responsive elements (CRE9 localized in the proximal promoter region). PTF: pituitary transcription factor, a cell-type-specific TATA element that seems not to be conserved in the human promoter. TRE: TPA-responsive element, located at position −490 in the proximal promoter region, whose possible function could be to mediate a complete transcriptional shut-off of the human Pit-1 gene when the DNA-binding activity of the Pit-1 protein is inhibited by mitotic phosphorylation. (TPA: 12-O-tetradecanoylphorbol-13-acetate responsive element).

Three isoforms of Pit-1 have been described, of which two of them with molecular weights of 31 kDa and 33 kDa are jointly called Pit-1 or Pit-1a and represent the majority of Pit in the pituitary. The third isoform, Pit-2 or Pit-1b, appears by alternative RNA processing; it retains its ability to bind DNA and may be involved in the differentiation process of somatotrophs [51].

Structurally, three regions are distinguished in Pit-1: (1) a DNA-binding zone located at the C-terminal end, in which two domains can be differentiated, a homeodomain (POU_HD) of 60 amino acids found throughout the family of transcription factors with POU domains, and a specific POU domain (POU_s) of 75 amino acids located at the N-terminus of the POU_HD domain. The POU_HD domain is the one that comes into contact with the DNA and is the one that contains the Pit-1 nuclear transfer signal. For its part, the POU_s domain does not directly contact the DNA, but surely increases the binding of POU_HD to it, stabilizing the DNA-protein complex [42]. (2) A region responsible for the activation of transcription, which comprises the 72 amino acids from the N-terminal end of the molecule and binds to site 1 of the response element. (3) A “bridge” region located between the two above.

In addition to Pit-1, GH transcription in man is increased by GHRH and glucocorticoids and inhibited by SS and activin [46]. GHRH increases intracellular cAMP that produces an increase in the synthesis of CREB proteins that bind to the two CREs (proximal and distal) of the GH promoter (Figure 7) increasing the transcription of the gene. On the contrary, both SS and activin counteract these actions by inhibiting the synthesis of cAMP. In the case of glucocorticoids, their effect is produced by the direct action of their receptor on the response elements located at the promoter and first exon level (Figure 7), facilitating the access of other transcription factors (especially Pit-1 and CREB) to their specific binding sequences [42].

Figure 7.

Structural organization of the GH promoter. The figure shows the location of the DNA response elements for the different transcription factors of the GH-N gene. GRE: glucocorticoid response element. CRE: cAMP response elements (dCRE: distal and pCRE: proximal). 1 and 2 represent, respectively, the binding sites for the DNA-binding domain and for the transactivation domain of the specific transcription factor Pit-1. NF-1 (nuclear factor-1) represents the binding sequence of the silencers of said family (found in rat). Sp-1 (specificity protein 1) provides GHRH responsiveness in the GH gene through activation of adenylate cyclase. The lower part of the figure shows the location of the second GRE at intron A of the gene. Ex: Exon. In: Intron. bp: base pairs.

2.4 Neurotransmitter regulation of GHRH and SS secretion

Our group demonstrated that GH responses to an exogenous GHRH challenge and other stimuli for GH secretion, in humans, were strongly dependent on the functional status of the intrinsic HSR at the time of testing [8]. From this it is concluded that: (a) GHRH was not able to interrupt HSR the hypothalamic rhythm that governs GH secretion, and (b) stimuli capable of altering this rhythm must act mainly by inhibiting the release of SS, rather than by stimulating endogenous GHRH. In fact, a continuous infusion of GHRH does not alter the pulsatile GH secretion. The maintenance of pulsatile GH secretion when GHRH is continuously infused supports these conclusions.

On this basis, the role of the main neurotransmitters in the control of GH secretion was reevaluated.

2.4.1 Adrenergic pathways and GHRH and SS secretion

Central adrenergic pathways play an important role in the control of GH in mammals [52]. Pharmacological blockade of catecholamine synthesis or release leads to the abolition of pulsatile GH secretion, an effect reversed by the administration of α₂-adrenergic receptor agonists. Conversely, pharmacological blockade of α₁-adrenergic receptors does not modify the latter phenomenon [53].

For years, experimental evidence led to the postulate that GH secretion induced by central α₂-adrenergic stimulation was mediated by increased GHRH release.

However, consistent with our observations described above [8], GH responses to any challenge must depend on the following: (a) potentiation of endogenous GHRH release, while SS tone is physiologically low, or (b) a primary inhibitory effect on SS release, accompanied by a secondary stimulation of GHRH secretion. In the first case, the responses would be sporadic, depending on the functional status of HSR at the time of the test. In the second case, a significant GH release should always be observed. The rationale for this is also based on the fact that, although there is no GH secretion in the absence of GHRH, the GH-releasing effect of this peptide cannot be observed in the presence of a physiologically or pharmacologically elevated hypothalamic SS production tone. On this basis, we demonstrated in humans [54], dogs [55, 56], and rats [57] that α2-adrenergic pathways primarily act in GH control by inhibiting SS release, rather than inducing GHRH release, although hypothalamic release of this peptide is also stimulated by α2-adrenergic pathways. Blockade of α2-adrenergic receptors has been shown to stimulate SS release in rabbits [58], so there appears to be no interspecies difference in the role of α2-adrenergic pathways in GH control. Both the α1- and β-adrenergic systems antagonize the α2-effect, although the former appears to have no physiological relevance in man.

β-Adrenergic antagonists enhance GH responses to GHRH [59, 60] and other stimuli, such as insulin-induced hypoglycemia [59], which logically is due to inhibition of hypothalamic release of SS.

In summary, the adrenergic system performs two antagonistic functions in the regulation of human GH: facilitator, mediated by α2-adrenergic receptors that act mainly by inhibiting the release of SS and secondarily by inducing the release of GHRH; and inhibitory, dependent on β-adrenergic activity, which stimulates SS secretion and inhibits GHRH release [61, 62].

2.4.2 Dopaminergic pathways and GHRH and SS secretion

Studies on the role of dopaminergic pathways in GH secretion have yielded contradictory results; stimulant and inhibitory effects have been described. Vance et al. [63] reported that dopamine (DA) increased GH secretion caused by GHRH in humans, suggesting an inhibitory effect of DA on the hypothalamic release of SS, exerted at the level of the median eminence. However, administration of centrally acting DA agonists, such as bromocriptine [63] and CV 205–502 [64], has also been shown to enhance both spontaneous GH secretion and GHRH-induced release. Therefore, DA would presumably inhibit SS release at least at two levels (the hypothalamus proper and the median eminence).

In vitro studies have indicated that DA stimulates SS secretion from the median eminence; furthermore, an increase in SS release in the pituitary portal blood has been observed after intraventricular injection of DA in rats [54]. On the other hand, our group has reported that central DA receptor blockade with metoclopramide was able to enhance the GH response to GHRH during periods of physiologically increased delivery of SS to the pituitary, but not when the SS tone appeared to be low [65]. This suggested a stimulatory role for DA in hypothalamic SS release, but also that this could be mediated by an indirect mechanism: negative DA modulation of NA release. This hypothesis is not inconsistent with the aforementioned studies [63, 64] showing that DA agonists centrally facilitate GH release. In fact, centrally acting DA agonists, such as bromocriptine, induce a biphasic pattern of GH secretion: inhibition followed by rebound stimulation.

In summary, everything suggests that the central role of DA in the control of GH in humans depends mainly on its effects on adrenergic transmission to SS neurons. Therefore, DA would act as a modulator rather than a direct regulator of GH secretion.

2.4.3 Cholinergic pathways and GHRH and SS secretion

Evidence indicates that cholinergic pathways play a key role in the control of GH secretion in both humans and laboratory animals [54, 66, 67]. Inhibition of central cholinergic pathways with muscarinic receptor-blocking drugs (atropine, pirenzepine) strikingly decreases GH release induced by a number of physiologic and pharmacologic stimuli [68, 69]; however, these have no effect on the GH secretion elicited by insulin-induced hypoglycemia. Conversely, enhancement of cholinergic tone with muscarinic cholinergic agonists potentiates GHRH-stimulated GH secretion and stimulates basal GH release as well.

It has been hypothesized that cholinergic synapses may be the final pathway for a variety of stimuli inducing GH secretion. The administration of anti-SS antibodies blocks the inhibitory effect of atropine on GHRH-elicited GH release in normal rats. Moreover, the acetylcholinesterase inhibitor pyridostigmine increases GH responses to a maximal stimulating dose of GHRH even when there is an abnormally high SS tone, as in obesity. Therefore, the facilitating role of cholinergic pathways on GH secretion appears to be mediated by an inhibition of hypothalamic SS release. In fact, while muscarinic cholinergic agonists stimulate basal GH secretion and its response to GHRH, muscarinic antagonists inhibit both responses [68, 69], immunoneutralization of GHRH by a GHRH-specific antibody does not affect the secretion of GH induced by pyridostigmine [70].

Because α₂-adrenergic and muscarinic cholinergic pathways appear to play an equally important role in GH neuroregulation in humans, both by inhibiting hypothalamic SS release, it has been investigated the functional relationships between these two neurotransmitter pathways. Our findings obtained showed that while α₂-adrenergic activation was able to overcome the inhibitory effect of muscarinic cholinergic blockade on GHRH-induced GH secretion, blockade of α₂₋adrenergic receptors counteracted the stimulating action of the pyridostigmine on said response [61]. Given the above, we speculated that in the control of GH, the α₂-adrenergic neurons are located distally to the cholinergic neurons, so that the latter contribute to GH secretion by modulating the functional activity of the former, as occurring in peripheral tissues. The fact that known stimuli for SS inhibition acting through adrenergic neurons (e.g., galanin and insulin hypoglycemia) are able to overcome the inhibitory effect of atropine on GH secretion is consistent with this theoretical scheme, which, moreover, does not exclude the possibility that the cholinergic input directly reaches SS neurons and inhibits them, as acetylcholine does in hypothalamic cultures in vitro.

Thus, the degree of cholinergic activity likely determines the amount of catecholamines released into the synaptic cleft of SS neurons. This, in turn, activates either SS-stimulating β-adrenergic receptors (responsive to low concentrations of catecholamines) or SS-inhibiting α₂-adrenergic receptors (responsive only to high catecholamine concentrations). To better understand the actions of these three neurotransmission pathways on GH secretion, see Figure 8.

Figure 8.

Main control of GH by neurotransmitters. The main role is played by adrenergic signals (CA) to somatostatin-producing neurons (SS). Depending on the amount of CA released into the synaptic cleft, inhibitory α₂-R receptors (high CA input) or stimulatory β₂-R adrenoceptors (low CA supply) are activated in SS neurons (see text). This, in turn, would be positively modulated by ACh supply to CA neurons, but negatively by DA. However, a direct inhibitory effect of ACh on SS neurons, or a stimulatory effect of DA on SS release in the median eminence (?), cannot be excluded. On the other hand, the stimulatory effect of CA on GHRH release, mediated by stimulation of α₂-adrenergic receptors in the GHRH neuron, appears to be less important than the inhibition of SS. Therefore, SS appears to be the main determinant of the secretion pattern of GH. Blue arrows: stimulation. Red arrows: inhibition.

2.4.4 Other neurotransmitters acting on GHRH and SS secretion

Given the complexity of hypothalamic structures and the relationships between different neurotransmitters in the control of hypothalamic functions, it seems logical that neurotransmitters other than those analyzed may play a role in the control of the hypothalamic discharge of GHRH and SS release into portal blood. This is the case for serotonin, gamma-aminobutyric acid (GABA), nitric oxide (NO), and endogenous opioids.

2.4.4.1 Serotonin

In rats, serotonin stimulates GH secretion, whereas in humans, stimulation or inhibition or even no effect has been described [53]. These discrepancies could be due to the low specificity of the drugs used, but also to the fact that serotonin acts on multiple types of receptors [71]. However, serotonin fibers have been found in periventricular nucleus and the arcuate nucleus where the SS and GHRH-producing neurons are located, respectively, so this neurotransmitter must play some role in the control of GH secretion. Indeed, administration of an anti-GHRH antibody suppresses serotonin-induced GH secretion in rats [72] and hypoglycemia-induced GH release in man is abolished by administration of serotonin antagonists [73]. In turn, the administration of L-tryptophan, a precursor of serotonin, increases the release of GH.

2.4.4.2 Gamma aminobutyric acid (GABA)

GABA is an important neurotransmitter whose actions at the central level are mainly inhibitory. In 1984, GABA was reported to modulate the secretion of AP hormones secretion by acting at both the hypothalamic and pituitary levels [74]. Intraventricular administration of GABA in rats stimulated LH and GH release. This effect was attributed to the fact that GABA induced an elevation of hypothalamic NA and median eminence DA levels as well as AP DA levels. The authors concluded that GABA plays a physiological role in the control of AP hormone secretion, primarily through hypothalamic action [74]. These data supported a previous study in which a dual action of GABA was postulated in the control of the hormonal secretion of the AP gland: one mediated via the central nervous system and the other exerted directly at the level of the AP [75]. Inhibition of GABA degradation and blockade of GABA transmission as well as administration of GABA and GABA mimetic drugs have all been shown to affect GH secretion. However, there are many controversial findings. For example, ancient studies indicated that GABA inhibited the release of SS [76, 77]. The effects of GABA may depend on the site of action within the hypothalamic-pituitary axis and the hormonal milieu. What is clear is that GABA participates in the regulation and actions of a very important GH secretagogue, such as ghrelin. In addition, long-term GHRH administration increases GABA levels in all brain regions [78], and GHRH activates GABA receptors in the cerebral cortex [79]. In all, these data suggest that GABA participates in the control of GH secretion but acts as a modulator of the actions of other neurotransmitters or peptides.

2.4.4.3 Nitric oxide (NO)

The gaseous neurotransmitter nitric oxide (NO) is synthesized from arginine by the action of nitric oxide synthase (NOS). NO performs a number of very diverse actions in the body, both at the level of the central nervous system and in the periphery. NO plays a stimulatory role in vivo and in vitro on GH secretion in acromegalic patients [80]. In rats, NO stimulates secretion of GHRH therefore increasing secretion of GH; however, it was also found that GHRH, in turn, increases production of NO in somatotrophs blunting GH secretion [81]. In dogs, inhibition of NO blunts GHRH-induced GH secretion, suggesting that NO acts by decreasing hypothalamic SS release [82]. However, years before, Aguila [83] argued that the hypothalamic increase in NO induced by GHRH would act on SS neurons increasing both the synthesis and release of SS. More recent studies conducted in fetal human primary cell cultures indicated that NO stimulated GH secretion, probably via the cGMP pathway [84]. This is consistent with further data demonstrating that a strong GH secretagogue, such as ghrelin, requires the activation of the NOS/NO pathway, and its subsequent GC/cGMP signal transduction pathway, as essential steps to induce GH secretion from somatotrophs [85].

In summary, it appears that NO plays a role in mediating the GH response to neuroendocrine factors, but its role appears to change depending on environmental conditions, including mitochondrial functioning, either in the hypothalamus or in the pituitary gland itself.

2.4.4.4 Endogenous opioids

There are three major types of endogenous opioid peptides that can be considered as neurotransmitters: endorphins, enkephalins, and dynorphins, which are respectively derived from three different precursor proteins: pro-opiomelanocortin (POMC), preproenkephalin A and B [86], although this number increased after the discovery of endomorphin-1 and endomorphin-2 [87].

The endogenous opioid peptides play many different physiological and pharmacological effects in humans, including neuroendocrine actions. In the case of the control of GH secretion, it has been shown that β-endorphin, administered intravenously or directly into the rat hypothalamus, increases GH secretion, an effect mediated by α₂-adrenergic pathways, via stimulating GHRH and inhibiting SS release, because β-endorphin antiserum reduced the stimulatory effect of clonidine on GH release [88]. Treating rats with an antiserum against GHRH inhibits the GH stimulatory response to β-endorphin [89], and the inhibitory effect of SS on GH secretion is antagonized by endogenous opioids [90].

At this point, there is the need to remark that exogenous non-peptide opioids have different effects on GH secretion when administered to man. In general, whereas acute opioid administration increases GH secretion, the effects of chronic opioid administration are much more complex. For example, intrathecal administration of opioids in chronic pain patients inhibits GH secretion, but the response is affected by sex, body composition, and insulin resistance [91].

2.5 Regulation of GH by metabolic substrates

GH is a metabolic hormone that plays an important role in regulating carbohydrate, fat, and protein metabolism in humans. Therefore, it is logical that metabolic substrates play a modulator role in the control of GH secretion.

2.5.1 Glucose and GH secretion

Glucose is a nutrient that the brain uses almost exclusively to provide energy [92], so glycemic regulation is key to maintaining normal brain function.

Basically, acute hyperglycemia inhibits GH secretion, either in basal conditions or in response to a number of stimuli acting on the central nervous system. Conversely, the acute decrease in plasma glucose concentrations leads to a rapid GH discharge. GH is one of the four counter-regulatory hormones that oppose the hypoglycemic actions of insulin (the other three are catecholamines, glucagon, and cortisol).

The opposite effects of hypo- and hyperglycemia seem to take place in the hypothalamus in SS-producing neurons. Hypoglycemia leads to the inhibition of SS secretion, while hyperglycemia induces SS release as cholinergic agonism prevents hyperglycemic blockade of GHRH-induced GH secretion [53]. The increase in GH induced by hypoglycemia appears to be almost totally refractory to cholinergic blockade [53]. Since hypoglycemia strongly activates adrenergic transmission, an inhibition of SS release mediated by α₂-adrenoceptors [54, 61] would favor GH secretion, also positively acting on the secretion of GHRH. The former is most likely dependent on increased glucose-induced SS release, as cholinergic agonism prevents hyperglycemic blockade of GHRH-induced GH secretion [53]. The latter, in turn, appears to be mainly due to by inhibition of SS secretion. The fact that the increase in GH induced by hypoglycemia appears to be almost totally refractory to cholinergic blockade [53] does not invalidate such a possibility. Since hypoglycemia strongly activates adrenergic transmission, an inhibition of SS release mediated by α₂-adrenoceptors [54, 61] would favor GH secretion. Furthermore, as seen above, an enhancement of endogenous GHRH release would also take place. In fact, in mice, it has been seen that hypoglycemia activates GHRH neurons [93]. However, also in mice, it has been reported that hypoglycemia produces a decrease in GH release [94, 95], most likely mediated by specific activation of neurons producing Neuropeptide Y [96]. Therefore, it is not yet clear, at least in mice, what is the role of glucose detection in GHRH neurons in controlling GH release during hypoglycemia due to the possible involvement of other central systems, such as the neurons that produce NPY.

As it is logical, hyperglycemia has to act in the opposite way to hypoglycemia. Consequently, as described, hyperglycemia increases the hypothalamic release of SS [97] and GHRH-induced GH secretion drastically decreases [98]. Furthermore, GH content in the pituitary is decreased in diabetic rats.

However, GH responses to diabetes are different in rats and humans. In humans, type I diabetes presents with increased pulsatile GH secretion [99, 100], perhaps because there is a greater pituitary sensitivity to GHRH [101]; however, in type II diabetes, GH secretion is negatively affected [102, 103, 104]. These differences are most likely due to the different fat mass in both types of diabetes and its consequences on insulin secretion, suggesting that insulin may be a regulator of GH release in both types of diabetes.

2.5.2 Free fatty acids (FFA) and GH secretion

FFA can totally block GH responses to a number of stimuli, including that of exogenous GHRH [105]; therefore, it has been postulated that the inhibition of GH release induced by FFA is mediated by an increased secretion of hypothalamic SS, although the data obtained in vitro point to a direct pituitary inhibition also [106]. In humans, the GH response to combined administration of pyridostigmine plus GHRH is markedly altered when plasma FFAs are increased by administration of a lipid-heparin infusion [107]. This favors the possibility of a main inhibitory action of FFAs on somatotrophs, which could be exerted by modifying the bilayer structure of the membranes and, therefore, affect their ability to detect stimulating signals (e.g., GHRH) or activate subsequent secretory mechanisms (GHRH or increased intracellular Ca²⁺). Interestingly, the response of other pituitary hormones to their hypothalamic releasing factors appears to be unaffected by the increase in plasma FFA in humans. Therefore, the inhibitory action of FFAs on GH secretion appears to be very specific. This is curious, because an important metabolic effect of GH is to induce the release of FFA from adipose tissues [108, 109]. Obese subjects show increased plasma FFA levels, while GH secretion is practically abolished, and if we suppress circulating FFA with antilipolytic drugs, obese patients regain normal GH secretion [110, 111]. Elevated levels of FFA are associated with insulin resistance and insulin hypersecretion; therefore, the decrease in GH in obesity may be caused by altered insulin feedback, rather than being a consequence of elevated plasma FFA levels, although there is no doubt about the effect of FFAs on the pituitary response to GHRH and SS. The opposite situation is seen in fasting: GH secretion increases and stimulates FFA release [108, 109].

In total, these relationships between GH and FFA appear to be related to the need to maintain metabolic homeostasis.

2.5.3 Amino acids and GH secretion

Given the clear and important anabolic effects that GH exerts on protein synthesis and metabolism, it is at least surprising that only two basic amino acids, arginine and ornithine, exert a powerful effect on GH secretion in humans. Arginine clearly increases the maximal GH responsiveness to GHRH in humans [112, 113], even when an elevated SS tone seems to exist (e.g., in the elderly). Furthermore, pyridostigmine does not potentiate GH response to arginine [114]. Hence, the inhibition of hypothalamic SS release appears to be the mechanism involved in the GH-releasing effect of this amino acid. Furthermore, arginine proceeds from citrulline and can be transformed into NO by the action of NOS, which suggests that another mechanism of action of this amino acid in GH secretion may be secondary to the increase in NO synthesis at the central level.

The effect of arginine on GH secretion is much more powerful than that of ornithine.

The basic effects of metabolic substrates on GH secretion are schematically depicted in Figure 9.

Figure 9.

Basic control of GH by metabolic substrates. Perhaps the main regulator is plasma glucose levels, given the effects of GH as a counterregulatory hormone. Hypoglycemia is detected by glucose-sensitive neurons and leads to the release of catecholamines (CA). These have a dual effect: they inhibit the release of SS (red arrow, −) and stimulate the release of GHRH (blue arrow, +). In somatotrophic cells, there are receptors for GHRH (GHRHR) and SS (SSR) that when stimulated induce respectively the synthesis and release of GH (GHRH, blue arrow, +) or the inhibition of the secretion of this hormone (SSR, red arrow, −), although in the case of SSR, it cannot be ruled out that they also act negatively on the synthesis of pituitary GH (?). The effects of FFA seem to be more varied, facilitating the release of SS, like hyperglycemia (blue arrow, +), and also acting at the level of somatotrophs, inhibiting the activation of GHRH and/or preventing the stimulating effect of GHRH on GH release (red arrow, −). Arginine appears to directly inhibit SS release (red arrow, −), allowing GHRH release. Although not depicted in the figure, the pituitary SS receptor is type 2, SSTR2.

2.6 Regulation of GH by other hormones and peripheral and central peptides

Endocrine products mainly related to GH control are those that play a general permissive role in the body, such as thyroid hormones and glucocorticoids, or those related to maturational changes, such as sex steroids, but also other hormones, such as Insulin and IGF-I play an important role in the control of GH synthesis and release. Furthermore, as with other pituitary hormones, GH secretion is self-regulated. In addition, as GH plays multiple important roles in practically all the tissues of the body, especially of a homeostatic metabolic regulatory nature, a series of central and peripheral peptides, some of which can be considered as authentic hormones, clearly participate in the modulation of GH synthesis and secretion, as we will analyze in this section.

2.6.1 Thyroid hormones and GH secretion

Thyroid hormones are permissive hormones needed for a physiological response to any other hormone. Hypothyroidism is known to be associated with impaired linear growth, which is easily normalized by the administration of replacement doses of thyroid hormones. Consistent with it, there is a decrease in plasma levels of IGF-I, which are normalized after the treatment with thyroid hormones. Both spontaneous GH secretion and GH responses to classic stimuli are markedly reduced in clinical hypothyroidism in humans or in induced hypothyroidism in rats [115]. There are several explanations for these facts. First, the lack of thyroid hormones strongly affects the hypothalamic synthesis of GHRH [116]. Although this directly leads to impaired GH synthesis, additional effects of hypothyroidism may depend on the lack of both the positive modulation that thyroid hormone induces on the number of GHRH-receptor sites in somatotrophs and its facilitatory role on GHRH binding to its receptors. Furthermore, GH-secretory mechanisms are also impaired in hypothyroidism, not only because a decrease in the hypothalamic content of GHRH [116, 117] but also because SS appears to be increased, at least in the rat [117, 118, 119].

Here at this point, it is of interest to note that the administration of GH to hypothyroid children with GH deficiency can increase hypothyroidism, something that does not occur when GH is administered once the hypothyroidism has been corrected.

2.6.2 Glucocorticoids and GH secretion

The effects of glucocorticoids on GH synthesis and secretion are complex and reflect the pluripotential nature of these hormones. On the one hand, glucocorticoids are essential for the maintenance of GH secretion; thus, patients with adrenocortical insufficiency present GH deficiency that can be corrected by substitutive treatment with these steroids [120]. However, excess glucocorticoids decrease GH secretion and longitudinal growth [121, 122]. Acute administration of glucocorticoids leads to a GH-secreting response that lasts for a few hours, although after this there is a total blockade of the release of the hormone [123, 124]. This is probably all due to the fact that glucocorticoids stimulate transcription of the GH gene [125] and increase the stability of its mRNA [126], as well as they increase the expression rate of the GHRH receptor in somatotrophs.

The negative effect of excess glucocorticoids would depend on its induction of SS release, mediated by an increase in the responsiveness of SS neurons to β-adrenergic stimulation, as it happens in the periphery (Figure 10). It may also be that glucocorticoids directly modulate SS synthesis, since there is a glucocorticoid response element in the SS gene.

Figure 10.

Control of SS by glucocorticoids. Glucocorticoids (GC) stimulate the expression of β-adrenoceptors (β-R) in somatostatin neurons (SS neuron) and favor their response to low levels of catecholamines (1′ and 2), inducing the release of SS. Furthermore, GCs can stimulate the transcription of the SS gene (3). The result is an increased discharge of SS into the portal blood.

2.6.3 Sex steroids and GH secretion

The existence of a pattern of sexual dimorphism in GH secretion, first described in rat [127], has been a clearly well-established concept for many years [128]. Sexual differentiation occurs during fetal life; sexually different fetal sex steroid hormones regulate striking differences in the number of GHRH and SS neurons, their responses to sex steroids once the puberty begins, the adult hypothalamic synapse and its organization, and perhaps the number of somatotrophs and their responsiveness [129]. In humans of both sexes, spontaneous GH secretion is low during childhood, and it is gradually increasing until just before puberty, when the release of the hormone increases strongly [130, 131] and sexual dimorphism begins to manifest. Therefore, it appears that sex steroids explain this maturational change in the pattern of GH secretion. The question is: where does this action of sex steroids take place?

After puberty, GH secretion is greater in women than in men, although this can be modified depending on the phase of the menstrual cycle; the late follicular phase leads to GH secretion greater than that seen in both the early follicular phase and the luteal phase.

A direct pituitary effect of sex steroids on GH secretion cannot be excluded; however, they most likely act primarily on the hypothalamic level, modulating the rhythmic interaction of GHRH-SS and, in particular, the availability of SS [61]. The difference between hypothalamic levels of free estradiol (fE2) in man and women can explain the differences in the secretory pattern of SS and, therefore, why GH release is sexually dimorphic.

In women, there is a large variation in plasma E2 levels throughout the menstrual cycle. We will have to bear in mind that only fE2 is biologically active, and that only in this form can it cross the blood-brain barrier. In fact, most of E2, but also of T, circulates bound to sex hormone-binding globulin (SHBG), although its affinity for this carrier is lower than that of T. However, plasma levels of SHBG are approximately two times higher in women than in men, because E2 increases the hepatic synthesis of this globulin. This, together with the fact that T circulates in levels of ng/ml, while E2 does so in pg/ml, suggests that free testosterone (fT) reaches higher levels in the hypothalamus than fE2, but this would be compensated by the hypothalamic aromatization from T to E2.

The data in rats clearly indicate that E2 replacement therapy rapidly reverses the decrease in SS mRNA observed after oophorectomy [132], although a direct effect of E2 on the SS gene is unlikely. E2 can affect both the biosynthesis and turnover of catecholamines [133] in hypothalamic areas involved in the control of SS, as well as the responsiveness of α₂-adrenoceptors in SS neurons. This, together with our findings demonstrating the role of the adrenergic system in the regulation of SS [54, 61], allows to hypothesize that the action of E2 on SS neurons depends on its effects on adrenergic transmission to these neurons [4]. This is consistent with data indicating that, at the peripheral level, catecholamines appear to be integral signaling components for maintaining steroid sensitivity in some reproductive tissues [134]. In addition, E2 inhibits the hepatic synthesis of IGF-I, thus reducing the effects of this inhibitor of both SS and GH.

On the other hand, T seems to stimulate the release of SS, but it depends on its central aromatization to E2. Therefore, the different levels of free E2 fE2 related to gender could explain the dimorphism in GH secretion [4]. Indeed, in humans, plasma fE2 but not T, were shown to be strongly correlated with rates of total and pulsatile GH release [5]. Support for this approach is given by the fact that there is great aromatase activity in the hypothalamus, as well as by the fact that it is E2 and not T that modifies the activity of a series of enzymes involved in synthesis and turnover of catecholamines. This explains why in children with delayed puberty the administration of T produces an increase in the pituitary reserve of GH, which is not observed when a non-aromatizable androgen, such as oxandrolone, is administered. Furthermore, studies in pubertal boys and adult men given the antiestrogen tamoxifen abrogate the stimulatory effect of T on GH [135, 136].

In summary, higher levels of fE2 in the hypothalamus, in women, can cause decreased biosynthesis and/or release of SS. If fE2 exerts a positive effect on the production of catecholamines at the hypothalamic level, they will also act positively on the production of GHRH. In GHRH neurons, there is an estrogen receptor ERα whose deletion in mice delays female puberty, since in females a group of GHRH neurons changes their phenotype to start producing Kiss1 (a key peptide in gonadotropic regulation). Therefore, a direct action of estrogens on neurons with the dual GHRH/Kiss1 phenotype modulates growth and puberty and may direct the sex differences in endocrine function observed during pubertal transition [137].

In women, episodic GH secretion is more frequent, and plasma GH values during trough periods are somewhat higher than in men. Therefore, in general, the amount of GH secreted in peak episodes is higher in women. But this can change throughout the menstrual cycle, as described before. Pulsatile GH release in women is doubled in the late follicular stage [138], and when estrogens are administered for superovulation, GH release is significantly increased [139]. Moreover, treatment of men with diethylstilbestrol (a potent estrogen) induces a change in the pulsatile pattern of GH release similar to that commonly seen in females [140]. It is important to remark that in humans, as in the rat [141], sex steroids would act mainly during the fetal and early neonatal stages, impregnating the hypothalamus so that from puberty there is an increase in the pulsatile secretion of GH, determined by an increase in the amplitude of the peaks rather than a change in secretory frequency [142]. In adults, the modification of the levels of sex steroids is not accompanied by significant modifications of the secretion of the hormone [4, 135].

The fact that estrogens produce an alteration in GH secretion may be due to the affectation of the feedback mechanisms through which IGF-I acts on GHRH and SS, in this case extraordinarily diminished by these steroids [143]. Alteration in GH output in response to estrogens is thought to occur as a consequence of altered negative feedback, wherein the actions of IGF-I on hypothalamic components of the GH-axis are greatly diminished [143]. This is something paradoxical [144], with evidence suggesting a complex relationship between IGF-I, estrogens, and GH output [129]. At low doses, estrogens increase GH output (while IGF-I production is high), while high doses of estrogens can decrease bioactive IGF-I levels, preventing the inhibitory effect of IGF-I on GH secretion. Furthermore, the route by which estrogens are administered clearly affects their effects on GH secretion. The pulsatile secretion of GH is very different when E2 is administered orally than when it is administered via transdermal patches, in both premenopausal and postmenopausal women. In two studies, oral administration of estradiol caused an increase in total daily GH secretion and its pulsatile release and a decrease in plasma IGF-I levels, whereas transdermal estrogen treatment increased circulating levels of IGF-I in post-menopausal to that seen in pre-menopausal women without altering the pulsatile release of GH [145, 146]. Therefore, it is likely that oral administration of E2 leads to a decrease in plasma IGF-I levels, consequently reducing its inhibitory effects, at the pituitary and hypothalamic levels, on GH secretion. Therefore, it is likely that the decrease in GH secretion observed at menopause is not due to the reduction in estrogen levels [146], since, furthermore, neither oral nor transdermal administration of E2 in menopausal women is able to reverse the decreased GH secretion associated with aging even when stimulated with GHRH. Moreover, the route of estrogen administration significantly alters the effect of estradiol on GH release. Assessment of pulsatile GH output in pre- and postmenopausal women following oral versus transdermal estradiol treatment revealed divergent effects on GH release. In two studies, oral administration of estradiol caused an increase in total daily GH secretion and pulsatility and a decrease in plasma IGF-I levels. Increased 24 h mean and pulsatile GH release and decreased circulating levels of IGF-I, whereas transdermal estrogen treatment increased circulating levels of IGF-I in post-menopausal to that seen in pre-menopausal women without altering the pulsatile release of GH [145, 146]. Thus, it is thought that oral administration of estrogen increases GH release through lowering circulating levels of IGF-I, thereby reducing negative feedback. These observations further indicate that the reduced GH output following menopause is not due to a reduction in circulating levels estrogen [146]. To this extent, neither oral nor transdermal estrogen treatment in postmenopausal women can reverse age-related reductions in spontaneous or GHRH-stimulated GH and IGF-I release. However, a recent randomized, double-blind, controlled study, performed in 60 healthy postmenopausal women conclude that while there was a E2 decrease when giving anastrozole (an inhibitor of aromatase activity), but no changes in E2 levels were observed when fulvestrant (a selective estrogen receptor-α antagonist) was administered. In spite of it, GH rebound after SS infusion declined markedly during both estrogen-deprivation schedules, suggesting a previously unrecognized dependence of hypothalamic-pituitary GH regulation on low levels of endogenous estrogen during menopause [147].

As we have seen throughout this section, although it is clear that sex steroids play a more than important role in the control of GH secretion, with E2 as the main actor, there are many controversies regarding both the mechanism of action and the effects they produce. Most likely this is a reflection of the great complexity of the world of GH and the number of factors involved in it.

In a study from our group carried out in young healthy volunteers of both sexes, it was concluded that testosterone acts on the release of GH at the supra-hypophyseal level and that its action is mediated by its aromatization to E2 [148]. Therefore, the association between reduced GH release at menopause and low circulating estrogen levels must be thoroughly analyzed in order to understand the mechanisms that may link these concepts.

Another point of interest is the fact that in women there is no cortico-adrenal production of estrogens and there is production of androgenic precursors that give rise to testosterone in the liver, as in men, which, although small, can help in GH secretory control.

A summary of these concepts is shown in Figure 11.

Figure 11.

Control of SS by sex steroids. Depending on the sex, the testicles or the ovaries secrete sex steroids in the blood (1), testosterone (T) or estradiol (E2), respectively, but also low amounts of E2 (testicles) and T (ovaries). While T can cross the blood-brain barrier (BBB, 2), only free E₂ (fE2, 2) can reach the brain. In the hypothalamus, T is aromatized to E2 (3), so the amount of fE2 is greater in women (4) than in men. fE2 modulates hypothalamic synthesis and catecholamine turnover (CA, 5), so the amount of catecholamines that reach SS neurons [5] is high enough to stimulate α2-adrenoceptors (6), inhibitors of the synthesis (7) and release of SS (8), but stimulators of GHRH synthesis (6) and release (9). E2 induces the hepatic synthesis of SHBG (10) and binds to it (11), as does T (11), but with lower affinity than T. Also in the liver, E2 decreases the synthesis of IGF-I (red arrow), a stimulator of SS release (12).

2.6.4 Insulin and GH secretion

Insulin and GH play some opposite actions at the metabolic level in the human body. Therefore, it is logical to think that insulin has effects on pituitary GH secretion.

As is known, chronic GH secretion counteracts insulin effects leading to peripheral insulin insensitivity, resulting in pancreatic hypersecretion of insulin that attempts to compensate for its loss of sensitivity [149]. On the contrary, as already defined, weight gain leads to poor GH secretion correlated with an increase in circulating insulin. On these basis, it appears that GH release is tightly controlled by prolonged changes in energy intake and supposed insulin release. There are many situations that support these inverse relationships in various metabolic conditions. For example, fasting, anorexia nervosa, and type-1 diabetes are associated with decreased insulin levels and increased GH secretion, while hyperphagia and obesity are associated with increased insulin secretion and decreased GH secretion.

A series of many evidences points to a direct relationship between hypoinsulinemia and excess increased secretion of GH. Intensive insulin treatment in type-1 diabetic patients reverses GH hypersecretion [99], whereas calorie restriction to reverse hyperinsulinemia [150] results in the recovery of GH secretion, presumably in response to a reduction in insulin levels [151].

Insulin inhibits the expression of the GH gene in isolated pituitary cells [152, 153], as well as acts selectively on its receptors in the pituitary gland [154] in such a way that it suppresses GH release from isolated somatotrophs [154, 155]. These indicate that, although systemic insulin resistance exists [154], somatotrophs remain sensitive to insulin in obesity. Therefore, insulin can modulate GH release in relation to weight gain and promote continued suppression of GH release in obesity, independent of the inhibitory effect of plasma insulin. In contrast, in vivo deletion of somatotroph insulin receptors leads to increased pituitary GH content and release [155]. These effects occur independently of the development of systemic insulin resistance, suggesting that somatotrophs remain insulin sensitive in obesity. Thus, insulin can modulate GH release in relation to weight gain and promote sustained suppression of GH release in obesity, independent of systemic insulin resistance. In contrast, in vivo removal of insulin receptors from somatotrophs leads to an increase in the content and release of GH in the pituitary [155]. The suppressive actions of insulin on GH release occur independently of IGF-I, since the selective inactivation of the insulin receptor and IGF-I receptor (Insr and IgfIr) genes in mouse somatotrophs does not affect these effects of insulin [156], despite the fact that in the liver, insulin facilitates the expression of IGF-I, which suppresses the synthesis and secretion of GH. Therefore, it is likely that insulin can provide critical feedback to alter GH release in relation to long-term metabolic requirements. It should be noted that loss of somatotroph-specific insulin receptor expression does not completely reverse the suppression of GH release seen during diet-induced weight gain and thus leads to obesity [156].

The actions of insulin on GH release may not be restricted to the pituitary. Insulin receptors are expressed throughout the hypothalamus, including the arcuate and periventricular nuclei [157, 158, 159, 160]; therefore, insulin can modulate GH release by acting through the hypothalamus, although this is not yet well known. As insulin does not predict ultradian GH release patterns, the actions of insulin at the hypothalamus level to mediate GH production may be limited to infradian regulation of maximal GH production [130].

2.6.5 IGF-I and GH secretion

IGF-I is the main mediator of the actions of GH in the body and shares many effects with it, although it also has specific effects of its own. IGF-I is produced in virtually any tissue, but plasma IGF-I is synthesized and released from the liver. There are specific relationships between IGF-I and GH, since IGF-I is part of a long feedback loop by which it inhibits GH secretion, acting both at the hypothalamic level, where it stimulates the release of SS, and at the pituitary level, inhibiting GH secretion directly. Even at the level of the median eminence, specific binding sites for IGF-I have been identified where it acts by inhibiting GHRH secretion [161, 162]. At the pituitary level, IGF-I inhibits the transcription of GH and Pit-1 genes, both under basal conditions and after stimulation with GHRH [41, 163].

2.6.6 Ghrelin, Klotho, and GH secretion

At the end of the past century, Bowers’ group [164] tried to identify how peptides derived from endogenous opioids could stimulate GH secretion. They synthesized a Met-enkephalin derivative, without opioid activity, which they named GHRP-6 (growth hormone-releasing peptide-6) [165]; it exhibited GH-releasing activity in vivo, in humans, acting synergistically with GHRH without binding to any known receptor. After the synthesis of this GH secretagogue, many other were synthesized, until a specific receptor for these synthetic compounds was identified in 1996 [166]. This receptor was a GTP-binding protein (GHSR-1a) present in the pituitary and arcuate and infundibular hypothalamus of several species, including humans [166], and its detection implied that a natural hormone had to exist. Indeed, this hormone was found in the stomach of rats and was called ghrelin [167].

Since we recently published an extensive review of the actions of ghrelin and klotho [168], only a few of the many important actions of these peptides will be discussed in this chapter.

2.6.6.1 Ghrelin and GH secretion

Ghrelin is a 28 amino acid peptide, which is in fact a true pleiotropic hormone. In spite of its peptide structure, ghrelin must bind to an octanoyl group on the third amino acid to acquire full biological activities [169].

The discovery of ghrelin and the fact that it acts synergistically with GHRH caused a change in the concept of basic regulation of GH; that is, in addition to the classical GHRH-SS interaction, GH secretion from the pituitary gland is also modulated by ghrelin produced by specific cells located in oxyntic cells in the stomach (Figure 12A). Ghrelin is an orexigenic hormone that appears in the blood during fasting and reaches the CNS to transmit a hunger signal. This explains why in anorexia nervosa there is an increase in plasma ghrelin concentrations, while in obese subjects it is reduced, and also why GH secretion is increased in anorexia nervosa patients and during fasting, while it is reduced or absent in obesity, but also in the elderly [170] (Figure 12A). Therefore, it seems that ghrelin appeared in evolution to stimulate eating behavior and optimize the use of digested food by promoting the secretion of an anabolic hormone, such as GH [171].

Figure 12.

Control of GH secretion by ghrelin and GH secretion. A: In fasting situations, the empty stomach (1) produces ghrelin in P/DI cells (2), which is released to the blood (3, 4) from where it reaches the pituitary gland (5) and induces the synthesis and secretion of GH (6). Ghrelin also reaches the hypothalamus (7), where it stimulates (8) appetite neurons and GHRH and inhibits SS. Plasma ghrelin also inhibits IGF-I (9), therefore inhibiting its inhibitory effects on the pituitary synthesis and secretion of GH (10). In turn, IGF-I inhibits the increase in circulating ghrelin (11) produced by fasting. Blue arrows: stimulation. Red arrows: inhibition. Yellow arrow: synthesis and secretion of GH. Black arrow: hypothalamic effects of ghrelin. +: stimulation. −: inhibition. B: The active form of ghrelin is acyl ghrelin, obtained after an acylation performed by GOAT, which links after linking a fatty acid side chain to serine 3. Acyl ghrelin can then bind to its receptor GHSR-1a and activate it. This figure has been modified from reference 169, Front Endocrinol. 2021.

As indicated above, the active form of ghrelin that induces GH secretion, by stimulating GHRH secretion and inhibiting SS and IGF-I secretion (Figure 12A), and many other physiological functions, such as inhibiting cell death in cardiac and endothelial cells [172], is the acylated form (acyl ghrelin), produced by attaching an octanoyl group to amino acid 3 (serine) carried out by ghrelin O-acyl transferase (GOAT) (Figure 12B) [173, 174].

Ghrelin stimulates GH release: (1) directly on the pituitary somatotrophs and (2) antagonizing SS and inducing GHRH secretion. In addition, ghrelin decreases plasma levels of IGF-I, therefore inhibiting the negative effect of IGF-I on GH secretion.

In humans, the physiological nocturnal rise in GH was not inhibited by the infusion of a powerful analog of SS, such as octreotide [175, 176]. This indicates that the effect of ghrelin was not affected by SS. On the other hand, patients with inactive GHRH receptor, due to mutations, still had rhythmic GH secretion, suggesting that another factor, different to GHRH, was inducing pituitary GH secretion [177]. However, a non-sense mutation affecting ghrelin receptor in humans was found to be associated with short stature [178]. Furthermore, in rats, intravenous (iv) administration of ghrelin during a GH peak induced a marked increase in plasma GH [179], but the immunoneutralization of GHRH led to a virtual absence of ghrelin-induced GH secretion; however, when ghrelin was administered during a physiological trough period, the GH response was clearly diminished. Both effects indicate that ghrelin acts synergistically with GHRH while inhibits hypothalamic SS. In fact, ghrelin expression has been found in the hypothalamic arcuate nucleus [180].

Ghrelin and its receptor are also expressed in the pituitary [181, 182], so pituitary ghrelin may play an auto/paracrine role in the regulation of GH release. GHRH infusion increases pituitary ghrelin mRNA levels, suggesting that GHRH may be a regulator of pituitary ghrelin production [183]. Therefore, pituitary ghrelin can act physiologically on GH secretion, enhancing the response of somatotrophs to GHRH.

Ghrelin, as GHRH, acts through a G-protein-coupled receptor (GPCR), but in this case the activation of this receptor leads to the stimulation of the activity of phospholipase C (PLC) that induces the formation of IP3 and DAG (Figure 13); both IP3 and DAG induce an increase in cytosolic Ca²⁺ that facilitates the release of GH [179]. Ghrelin requires activation of the NOS/NO pathway and its subsequent guanylate cyclase (GC)/cGMP signal transduction pathway to induce GH release from the pituitary [85]. Chronic treatment with ghrelin produced an upregulation of GH transcription levels, as well as that of two isoforms of Na⁺ channels, sensitive to blockade with tetrodotoxin (TTX), expressed in somatotrophs, such as NaV1.1 and NaV1.2. This indicates that ghrelin also regulates the expression of the Na⁺ channel gene in somatotrophs (Figure 13) [184].

Figure 13.

Ghrelin control of GH secretion. Mechanisms of action of ghrelin in the control of GH secretion. Hypothalamic ghrelin (coming from the circulation and/or synthetized in the hypothalamus) stimulates the release of GHRH into the portal blood from where GHRH reaches the somatotrophs (green arrow) and induces synthesis (blue arrow) and release (green arrow) of GH, but also that of pituitary ghrelin that also participates in the synthesis of GH (blue arrow). Hypothalamic ghrelin antagonizes SS by itself (red arrow) but also through a mechanism mediated by its stimulation of AMPK (blue arrow). At the pituitary level, ghrelin from the hypothalamus (green arrow) or from the systemic circulation stimulates the release of GH via activation of the PLC that activates the intracellular Ca²⁺ supply and the transcription of Na⁺ channels necessary for the release of GH. Likewise, ghrelin after the activation of its membrane receptor (GPCR) activates the NOS/NO pathway, also necessary for the release of GH. Taken from reference [169], Front Endocrinol. 2021.

Interestingly, the hypothalamic enzymatic complex AMP-activated protein kinase (AMPK), involved in the hypothalamic control of energy and metabolic homeostasis, is also activated by ghrelin. Therefore, AMPK also participates in the control of GH secretion, and as its blockade or its functional impairment inhibits ghrelin- or GHRH-induced GH secretion, most likely by increasing SS production tone [185] (Figure 13).

It is well known that aging is associated with a decrease in GH secretion from the second decade of life [186]. Something similar happens with ghrelin [170, 187], although the pituitary ghrelin receptor does not decline with aging and the GH response to ghrelin is still seen in the elderly and there is an age-related decline [188].

Therefore, what is the reason why the secretion of an orexigenic hormone, such as ghrelin, and also that of an anabolic hormone, such as GH, is lost with aging?

Gastric ghrelin synthesis and secretion increase during fasting and decrease during feeding [189]. This is the reason why chronic intake of high-calorie diets, prolonged ingestion of high fats, and obesity lead to a reduction in gastric ghrelin production and secretion [189, 190], while a low protein supply significantly increases plasma ghrelin [190].

Adrenergic hormones stimulate the release of gastric ghrelin by acting directly on the β1 receptors of ghrelin-producing cells, very rich in this type of adrenergic receptors [191]. Therefore, fasting acts on gastric ghrelin secreting cells through the sympathetic nervous system, something that appears to be logical given the relationships between this autonomic system and the hypothalamic production of GHRH and SS for controlling GH secretion. The administration of muscarinic agonists also increases plasma ghrelin concentrations [192], as does vagus nerve excitation in the gastric mucosa [193].

Furthermore, and given the relationships between ghrelin, SS, and GH secretion, it stands to reason that SS [194] and GH [195] inhibit gastric ghrelin secretion, although ghrelin also stimulates pancreatic SS production [196], to inhibit insulin release [197, 198] and contribute to glucose homeostasis [197].

Since, as we have seen, metabolic factors play an important role in the control of GH secretion, it is logical that they also participate, via hormonal modulation, in the control of gastric ghrelin secretion, as can be seen in detail in Ref. [169]. Perhaps the only hormones to be discussed in this section are leptin and IGF-I. Leptin, a hormone produced by adipocytes, which acts by decreasing food intake; that is, it is an anorexigenic hormone, and it is logical that it acts by decreasing plasma ghrelin and inhibiting its actions at the central level [199].

Similarly, logical are the relationships between IGF-I and ghrelin. Plasma IGF-I concentration significantly determines plasma ghrelin concentrations, existing a negative correlation between them [200]. This has been found in children and adolescents [201, 202]. On this basis, the highest concentration of ghrelin was observed in GH-deficient children in whom there was low availability of free IGF-I [203], the bioactive form of IGF-I. That is, low plasma levels of IGF-I induce the synthesis and secretion of ghrelin, while in turn ghrelin decreases plasma levels of IGF-I.

2.6.6.2 Klotho and GH secretion

Although klotho was identified in 1997 as an antiaging agent, klotho-deficient mice exhibit growth retardation [204], implying that this transmembrane protein is also involved in the control of GH secretion. The kidneys are the main source of klotho, where it is stimulated by insulin [205] and IGF-I [206]. In contrast, klotho inhibits both receptor activation and intracellular signaling of these hormones [207], thus acting as a positive regulator of GH secretion.

Klotho induces GH secretion by activating the extracellular signal regulated kinase 1/2 (ERK1/2) pathway in AP somatotrophs [208]. Relationships between klotho and GH can be seen in untreated GH-deficient children and adults; in these patients, the plasmatic levels of klotho are low, while treatment with GH normalizes this klotho deficit, through a mechanism mediated by the activation of the Akt-mTOR pathway [206], a key pathway in GH signaling.

In addition to its production in the kidney, klotho is also produced in the somatotrophs, most likely to modulate auto/paracrine GH secretion. Interestingly, high production of klotho has been observed in pituitary adenomas that do not secrete GH, suggesting that klotho is also produced in other pituitary cells which are also a source of this still little known hormone that plays so many different roles in the body.

2.6.7 Neuropeptide Y (NPY) and GH secretion

NPY is a 36 amino-acid peptide belonging to the family of pancreatic polypeptides. It is expressed in the hypothalamus in neurons of the arcuate, paraventricular, and periventricular nuclei [209]. NPY exerts an important effect on the regulation of energy intake and expenditure, something that, as we have already analyzed, once again involves GH, so NPY has to play a role in regulating the secretion of this hormone, although this role is still little known. Ancient studies carried out in rodents showed that NPY inhibited the GH axis [210], most likely due to the fact that NPY neurons in the arcuate nucleus connect with periventricular SS neurons [211]. In fact, after 9 hours of fasting, it was observed, in mice, that the hypothalamic levels of NPY increased as did those of SS [95], which suggested that NPY would be the factor responsible for the inhibition of GH secretion, mediated by SS, associated with fasting. There are five receptors for NPY. Among them the Y1 receptor is mainly expressed at the post-synaptic level, while the Y2 receptor is expressed in NPY neurons [212, 213].

A study carried out in mice indicates that activation of Y1 receptor inhibits the secretion of GH during fasting, probably through induction of SS release, whereas Y2 receptor has no effect on GH secretion in fasting conditions, but maintains pulsatile GH release in mice fed “ad libitum” [96].

All these data indicate that NPY neurons play a very important role in the control of energy intake and expenditure, as indicated above. This concept is reinforced by the fact that most of these neurons co-express NPY and Agouti-related peptide (AgRP, another important regulator of metabolic homeostasis), and more importantly, there is an insulin/NPY network that regulates energy homeostasis, so that the lack of insulin signaling in NPY neurons induces an increase in energy reserves and obesity and also produces an alteration of GH/IGF-I axis. This is a clear example of how GH control is key to proper homeostasis of metabolism and food consumption, and how these, in turn, control GH production.

3. Facts and perspectives

It is now clear that pituitary secretion of GH is mainly subject to a tonic inhibitory control exerted by SS, contrary to what was thought after the discovery of GHRH. Therefore, only the factors that can inhibit the production of SS will be responsible for the release of GH induced by GHRH, whose synthesis and release into the portal circulation are inhibited by SS. All this was identified in the last decade of the last century. However, the great complexity of the hypothalamus, in terms of the large number of existing neuronal types that can release very different compounds, suggests that more neurotransmitters with the capacity to act on the SS and, consequently, on the hypothalamic-somatotropic axis can still be identified, although, logically, their actions will always be more secondary than those of the factors that have been described in this chapter.

4. Conclusions

In this chapter, the main factors involved in the control of GH secretion were analyzed, including the explanation of why it is sexually dimorphic or its relationship with feeding or fasting.

Although GH has traditionally been considered a merely metabolic hormone with specific actions on the longitudinal growth of the organism until the end of puberty, a series of studies carried out in recent years show that this hormone is highly complex in terms of the number of actions so diverse that it exerts in the organism. This has led to the control of its secretion being very diverse, although it primarily depends on a control exerted negatively from the hypothalamus by somatostatin and an inducing effect of its release carried out by GHRH, although it also depends on the nutritional status, by a primarily gastric peptide such as ghrelin.

As is logical, given the multiplicity of GH actions, a series of hormonal, metabolic factors, or factors simply indicative of the functional state of a gland or tissue, or neurotransmitters, will condition the rate and type of GH secretion. Basically, these factors will act primarily on the secretion of somatostatin, activating it and, consequently, inhibiting the secretion of GHRH-stimulating peptide, or inhibiting SS so that GHRH can be released into the hypothalamic-pituitary portal circulation, and stimulate the synthesis and pituitary secretion of GH. The basic control of SS secretion depends on the adrenergic pathways that reach the neurons that produce it, being the α₂-adrenergic inhibitory and the β-stimulants of its secretion, although some other factors, such as arginine or ghrelin, may act directly on SS neurons, while others, such as thyroid hormones, may act by regulating the number and responsiveness of GHRH receptors in the pituitary gland.

Conflict of interest

The authors declare no conflict of interest.