**2. Family of GH genes**

## **2.1 The GH gene locus**

The locus of the human genome where the genes belonging to the GH family are found is located on the long arm of chromosome 17, in the region q22–24 [16–18] (**Figure 2**), spanning 48 kb of DNA in the order: 5′-(hGH-1/-hCS-5/hCS-1/hGH-2/ hCS-2)-3′ [19, 20]. These genes encode normal human GH (GH-N) and its variant (GH-V) as well as the group of chorionic genes: hCS-L, hCS-A, and hCS-B. These are genes highly conserved in evolution, being the homology among them 5%. The hGH/ hCS gene locus seems that it has been evolved by duplication mechanisms [20]. They come from a common ancestral gene with that of Prolactin (PRL) from which they diverged about 300–400 million years ago, although the gene coding for PRL is found on chromosome 6. All these genes have five exons separated by four introns [20] (**Figure 2**). Among them, GH-1 (now known as GH-N) is expressed in the pituitary gland and peripheral tissues, while the other four seem to be expressed only in the placenta, although N-glycosylated GH-related peptides have been found in human pituitary extracts, suggesting that the hGH-2 gene (now known as GH-variant [GH-V), or other unknown GH-related genes, could be expressed too at the pituitary level, and perhaps in other tissues, since the hGH-N gene lacks the consensus sequence for N-glycosylation observed in some nonplacental GH-related products [21, 22].

## **2.2 Molecular GH heterogeneity**

**Figure 2** is only a schematic description of the hGH gene family, but pituitary and tissue GH heterogeneity is really high. For example, in the human pituitary,

#### **Figure 2.**

*hGH/hCS gene locus localization, expression of the genes of this family and products resulting from it. As noted above, the hGH/hCS gene family locus is located on the long arm of chromosome 17 (q22–q24). The figure also represents the 5*′*—3*′ *arrangement of the genes that make up this family, of which GH-N is outlined in the upper rectangle (indicated by a black arrow). In this gene, it can be seen the organization in exons (I–V) separated by introns (A–D). Note that, although the GH-N gene is expressed practically throughout the body (pituitary and multiple tissues and organs), the other four genes seem to be expressed only in the placenta, although as we will see, this is not exactly the case, at least for the GH-V gene.*

#### **Figure 3.**

*Heterogeneity of human pituitary (PIT) GH. Western blot showing the high heterogeneity of GH-N obtained from a human pituitary extract. Note the dimer 44 kDa formed by GH 22 kDa, as well as the number of GH-N variants whose biological significance is unknown. The 5 kDa GH form is not shown in this figure.*

many isoforms of GH-N can be found. Apart from the main GH products, 22 kDa and the 20 kDa form of GH, the latter resulting from the alternative splicing of the mRNA of the GH-N gene, there are several posttranslationally modified forms of GH (N-acylated, deamidated, and O-glycosylated forms of monomeric GH), as well as non-covalent and disulfide-linked oligomers up to pentameric GH [23, 24]. The high heterogeneity of pituitary GH can be seen in **Figure 3**.

Furthermore, in human pituitary extracts, some significant amounts of GH variants of lower molecular weight (17 kDa and 5 kDa) can be found. These originate from the selective cleavage of the bond between amino acids 43 and 44, leading to the production of fragments 1–43 and 44–191 in the somatotrophic cells themselves and have also been found in human plasma and tissues and in some different species [15, 25–30]. Since these forms can be originated by specific cleavage of the main 22 kDa GH form, it is likely that they appear in tissues as a result of posttranslational modification of the hormone. Interestingly, while human GH 22 kDa has both insulin-like and diabetogenic effects, these proteolytically generated fragments show opposite effects: the short 5 kDa GH form potentiates the effects of the insulin, while the 17 kDa form shows diabetogenic activity [28–30]. Since these two isoforms can be generated under acidic conditions, it has been postulated that they may play a significant physiological role because the hormone in cells is exposed to an acidic environment [25]. Similar conclusions have been drawn regarding the effects of these 17 and 5 kDa GH variants at the end of the last century. According to their data [29–31], these GH fragments, obtained by recombinant DNA technology, have potent in vivo effects on glucose homeostasis in rodents, but cannot stimulate body growth.

The great heterogeneity of forms produced from GH-N is due to the fact that in the main molecule, there are a series of points susceptible of hydrolysis by proteases, which would be another point in favor of considering GH as a prohormone, which undergoes selective proteolysis, both in the pituitary and in specific tissues, giving rise to peptides with specific biological activities. For example, there are two fragments of GH, GH 177–191 and GH 95–133, of which the first shows a lipolytic activity identical to that of the 22 kDa GH-N form, while the second is a potent mitogen. For its part, the 24 kDa GH-N form caused by retention of signal peptide, like the 12 kDa form originating from the previous one, shows greater somatogenic and lactogenic potency than the 22 kDa GH-N itself.

The alternative splicing of the major pituitary GH-N form 22 kDa occurs in approximately 10% of the transcripts in eukaryotes (**Figure 4**).

As a consequence of the cleavage of part of intron III in the 20 kDa GH-N form, the amino acids located from position 32 to 46 have disappeared (**Figure 5**).

The pituitary expression of these two proteins (GH-N 22 kDa and 20 kDa) reflects the alternative utilization of a major (B) and a minor (B′) splice acceptor site in exon 3 of the hGH-N transcript, although it has been postulated that, in addition to the importance of sequences in the immediate vicinity of the two alternative splice

#### **Figure 4.**

*Alternative splicing of the GH-N gene. After transcription of the GH-N gene, a part of intron III is cleaved (in 10% of transcripts, white rectangle), giving rise to a shorter GH-N form: The 20 kDA GH-N form.*

*Growth Hormone Gene Family and Its Evolution DOI: http://dx.doi.org/10.5772/intechopen.108412*

#### **Figure 5.**

*Schematic description of the primary structure of GH-N 22 kDa. The molecule is a polypeptide formed by 191 amino acids arranged in a single chain in which there are two disulfide bridges (S–S) that join the cysteines located at positions 53 and 182 with those located at positions 165 and 189 (red lines). The alternative splicing removes amino acids 32–46 (represented by yellow circles) giving rise to the GH-N 20 kDa form.*

acceptor sites, additional sequences further away in the transcript also contribute to this alternative splicing selection. These more distal sequences would not act individually, but instead would interact in such a way that the net level of alternative splicing in exon 3 would be dictated by the overall higher-order structure of the hGH-N transcript [32]. Curiously, a recent study concludes that although there is an important GH-N expression in retina, no splicing variants have been detected [33]. Our group found expression of GH and its receptor in hippocampal neural stem cells, but we did not differentiate whether the main GH form or the GH 20 kDa or both [34] were expressed. In the case of retina, it seems to be a particular situation, because both GH-N 22 kDa and 20 kDa bind and activate the same receptor (GHR); both forms exhibit similar, but not identical physiological activities, although the kinetics of signaling by GH-N 22 kDa and 20 kDa is different being weaker that of 20 kDa than of the major form, which can justify the different biological activities exhibited by these GH-N forms [35]. Furthermore, in humans, 20 kDa GH is a weaker agonist for the receptor of PRL than 22 kDa GH [36]. Other differences between both GH-N forms are related to their half-life in plasma. GH 20 kDa has a lower affinity for GH-binding proteins (GHBP) than GH 22 kDa [37] and fails to form a 1:1 complex with this binding protein [37]. The internalization rate of GH 20 kDa is less than that of GH 22 kDa [38]. Lastly, GH 20 kDa shows a stronger tendency to aggregate than GH 22 kDa [38–40]. The amplitude of GH secretory bursts is negatively correlated to percentage of body fat while testosterone plasma levels positively affect the secretory burst mass of the hormone [41]. Regarding the existence of the C-terminal disulfide bridge of GH, we know that it has been conserved throughout the evolution, although its role is unknown [42], but they seem to be fundamental for the maintenance of the active conformation of the hormone.

#### **Figure 6.**

*Structure of 22 kDa GH-N showing the disposition of its 4 α-helix. Site 1 (outlined in a circle) is the one that first binds to the extracellular domain (ECD) of GHR-1, allowing the receptor to dimerize and Site 2 (outlined in a dotted line circle) of GH can bind to this second receptor (GHR-2). After that, the biological effects of the hormone begin.*

GH has a complex tertiary structure, like other polypeptide hormones. Tertiary structure plays a role in how the hormone regulates receptor activation. GH is a long chain four α-helix bundle proteins (**Figure 6**).

A notable feature of their tertiary structure is that it contains no symmetry that might support equivalent binding environments for the receptors. How the two receptors bind to the asymmetric hormone was first revealed from the crystal structure of human growth hormone bound to the ECD of its receptor (hGHR). The characteristic arrangement of the four antiparallel α-helix is essential when it comes to produce the binding of GH to its receptor. Since this binding occurs in 1:2 ratio (one GH molecule and two receptor molecules) in each molecule of GH there are two receptor recognition epitopes, located at opposite ends of the nucleus of α-helix, site 1, and site 2. The structure shows that the two ECDs binding to site 2 and site 1, respectively, use essentially the same set of residues to bind to two sites on opposite faces of the hormone (**Figure 6**). This binding is characterized by extraordinary local and global plasticity at the binding surfaces. The two binding sites have distinctly different topographies and electrostatic character, leading to different affinities for the receptor ECDs. The high-affinity site, site 1, is always occupied first by ECD1. This sequence of events is required because productive binding of ECD2 at site 2 of the hormone requires additional contacts to a patch of the C-terminal domain of ECD1. The binding of ECD2 is the programmed regulatory step for triggering biological action, and it involves a set of highly tuned interactions among binding interfaces in two spatially distinct binding sites. The energetic relationships between the ECD1-ECD2 contacts and the hormone-ECD2 site 2 interactions are known to be important.

In the case of the 20 kDa GH-N, the reduced affinity exhibited by the receptor of the 22 kDa form suggests that conformational changes occurring in it affecting recognition epitopes.

As **Figure 2** shows, another form of GH, the GH variant or GH-V is expressed primarily in the placenta. This GH-V has an identical size that the pituitary 22 kDa GH-N, although both differ from each other by 13 amino acids dispersed throughout GH-V. However, the differences between GH-N and GH-V in terms of their somatogenic and metabolic activities are small. While secretion of pituitary GH-N is pulsatile under control from the hypothalamus, the secretion of placental GH-V is tonic and increases progressively in maternal blood during the second and third trimester [24].

Despite its placental expression, GH-V has been found in human testis [43–45], where most likely it plays an auto/paracrine function in reproduction (spermatogenesis). In addition to the testis it is likely, as indicated above, that GH-V, or some unknown gene related to it, can be expressed in other territories, including the pituitary.
