Bovine-Specific Nucleotide Polymorphisms and mRNA Expression of the Growth Hormone Secretagogue Receptor 1a (GHSR1a) Gene and its Genetic Association with Growth and Carcass Traits

The growth hormone secretagogue receptor 1a (GHSR1a) is involved in many important functions including growth hormone (GH) secretion and appetite regulation and other important functions. We reveal herein, the unravelling of bovine-specific 5’untranslated region (5’UTR) microsatellite polymorphisms, a 3bp-indel in exon 1 (DelR242) and two different kinds of transcripts of the GHSR1a gene (spliced, without a microsatellite with in the 5’UTR (GHSR1a); and non-spliced, with the microsatellite (GHSR1b)). A number of 17 alleles ((TG)10~33 ) in the 5’UTR microsatellite was found in 11 cattle breeds. Furthermore, we found the DelR242 (3R) allele, a truncated 3arginine residue (3R) (major type: 4 arginine residues (4R)) within the intracellular loop 3 of GHSR1a protein in Japanese Shorthorn with a high frequency of 0.43 compared to the low frequency of 0.00~0.09 in other cattle breeds. We carried out a genetic association study between the 5’UTR microsatellite and growth and carcass traits in 1,285 steers. Statistical analysis revealed that the 5’UTR microsatellite locus had a significant additive effect on carcass weight (CW) and average daily gain (ADG). The 19-TG allele had a significantly desirable effect on these traits. We proposed a translational hypothesis that the association is due to differences in the secondary structure of GHSR1b mRNA among the GHSR1a gene haplotypes. We also examined age-related changes in the expressions of GHSR1a and GHSR1b in many cattle tissues. The GHSR1a mRNA expression in the arcuate nucleus of postweaning calves was more than 10-fold higher than those of pre-weaning calves and cows. In peripheral tissues, there were 3 marked differences in mRNA expression between cattle, humans and mice, as follows: (1) the GHSR1a mRNA expression in the liver is high in cattle and very low in humans and mice; (2) the GHSR1b mRNA expression in the liver is low in cattle and high in humans; (3) the GHSR1b mRNA expression in the pancreas is very high in cattle.


Bovine-Specific Nucleotide Polymorphisms of the GHSR1a Gene and Its Genetic Association with Growth and Carcass Traits
The growth hormone secretagogue receptor 1a (GHSR1a), also known as ghrelin receptor, is involved in many important functions including growth hormone (GH) secretion, appetite regulation, energy balance and other important functions [1,2]. In cattle, the GHSR1a gene was reported as a potential candidate gene when we detected growth trait QTLs in Japanese Black cattle using microsatellite DNA markers and half-sib regression analysis [3]. With respect to the bovine GHSR1a gene, it is of great interest that a polymorphic microsatellite ((TG)n ) is located within the 5'-flanking region of this locus [4], because no microsatellite had ever been found within the GHSR1a locus in either humans [5], mice [6] or rats [7]. However, there was no published report on nucleotide polymorphisms from the 5'-flanking region to the 3'-UTR nor on the transcriptional analysis of the 5'-UTR of the GHSR1a gene in cattle. Therefore, we revealed for the first time, novel nucleotide polymorphisms from the 5'-flanking region to the 3'UTR (~6 kb) and two different kinds of transcripts (spliced, without a microsatellite within 5'UTR (GHSR1a); and nonspliced, with the microsatellite (GHSR1b)) of the bovine GHSR1a gene ( Figure 1) [8]. The nucleotide sequencing of this gene (~6 kb) revealed 47 single nucleotide polymorphisms (SNPs), 4 indels and the two microsatellites ((TG)n) in 5'UTR and (GTTT ) n in Intron 1). A number of 17 alleles (10-TG to 33-TG) was found in the 5'UTR microsatellite locus in 11 cattle breeds. There were breed differences in allele frequencies and major alleles. Specifically, in Japanese Black cattle, the major alleles were 19-TG, 23-TG and 24-TG ; alleles 19-TG,  21-TG, 22-TG, 23-TG and 24-TG in European cattle breeds, and  alleles 10-TG and 22-TG in Philippine native cattle breeds (a mixture of Bos indicus and Bos taurus types). Short repeat number alleles 10-TG, 15-TG, 16-TG and 18-TG were found in the Philippine native cattle. The microsatellite TG-repeat sequences included one cytosine(C) instead of guanine (G) at the 9th repeat position from the 3'-end of this locus of 7 alleles ((TG)10~24 ( TC )(TG )8 ). This position was constant and independent of the TG-repeat number. These results suggest that the TG-repeat number of microsatellites increased with evolution from the 3'-end to the 5'-end direction of this sequence ( Figure 1). We investigated the genetic association between the 5'UTR microsatellite ((TG)n ) of the GHSR1a gene and growth and carcass traits in Japanese Black cattle [9]. We used a population of 1,285 Japanese Black steers in a progeny-testing program of the Livestock Improvement Association of Japan (LIAJ). Genetic association analysis between DNA markers, growth and carcass traits was carried out using a univariate model within the framework of a derivative-free restricted maximum likelihood algorithm as applied in the MTDFREML [10]. MTDFREML statistical analysis clearly revealed that the 19-TG allele, one of the four major microsatellite alleles, had a very significant additive substitution effect on carcass weight (CW) (P<0.0007), and average daily gain (ADG) (P<0.0002). Besides, the A allele of the nt-7(C>A) locus also had a significant effect on these traits It has been reported that GT repeat polymorphisms in Tilapia prolactin 1 (prl 1) 5'UTR promoter are associated with differences in prl 1 gene expression and the growth response of salt-challenged fishes [11]. The length of the TG-repeat in the P1 promoter region of the growth hormone receptor (GHR) gene was significantly related to growth and carcass traits in beef cattle [12,13]. Stepwise increase in repeat numbers from 0 to 21 for a CA-microsatellite located in the promoter of the human matrix metalloproteinase-9 gene has been reported to produce incremental surges in transcription rates [14]. Furthermore, a review of the simple sequence repeats (SSRs) in the 5'-UTR by Li et al. [15] revealed that the regulation of gene expression is affected by both transcription and translation. Therefore, a logical explanation for the association between 5'UTR microsatellite ((TG)n ) polymorphism and growth traits is through differences in transcriptional or splicing or translational levels of the GHSR1a gene.
The first, a "transcriptional hypothesis" is that the differences in DNA structure around the 5'UTR microsatellite region between  and [non-19-TG] simultaneously affect transcriptional levels of both the 1a and 1b mRNAs from the 1a gene. However, this hypothesis is hard to explain using the GHSR1b function because it has been reported that GHSR1b (the truncated receptor polypeptide) acts as a dominant-negative mutant of the GHSR1a (functional Ghrelin receptor) due to the formation of GHSR1a/GHSR1b heterodimer [16].
The second, a "splicing hypothesis" is that 5'UTR microsatellite ((TG)n ) in the intron affects the efficiency of alternative splicing of an adjacent exon in pre-mRNA (e.g. Cystic fibrosis [17,18] for review]). However, this hypothesis seems to have low potential because the GHSR1a mRNA is the only spliced mRNA as there are no other alternative splicing variants. Moreover, the position of 5'UTR microsatellite is neither adjacent to 5' splice site (GT) nor 3' splice site (AG) [(GT)-//-(111 bases)-//-(5'UTR microsatellite)-//-(95bases)-//-(AG)]. The third, a "translational hypothesis", is more intriguing and interesting as it relates to the differences in RNA secondary structure around the 5'UTR region of the 1b mRNAs between the [19-TG] and [non-19-TG], thus affecting translational levels of 1b mRNAs but not 1a mRNAs. McClelland et al. [19] reported that translational efficiency is inversely correlated with the stability of the mRNA secondary structure, the presence of base-pairing in the consensus Kozak sequence, the number of start codons in the 5'UTR and the length of the 5'UTR. In order to test this "translational hypothesis", we predicted the optimal RNA secondary structure of the GHSR1a and GHSR1b mRNAs using the Vienna RNA secondary structure server (http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.cgi; RNA fold web server) [20]. The mRNA sequences of 6 haplotypes (haplotypes frequency: > 0.04) were analyzed. The results showed that the secondary structure around the 5'UTR microsatellite region of the [19-TG]-1b mRNAs had a unique structure that was different from those of [non-19-TG], thus  seemed to have a small dominant negative effect on the translation efficiency of GHSR1b mRNA due to its unique structure. Therefore, the GHSR1b mRNA translation efficiency based on the secondary structure of the 5'UTR microsatellite regions seems to be in the following order: [non-19-TG] > . On the other hand, the optimal RNA secondary structures of the 1a mRNAs appeared to be almost the same among haplotypes and there seemed to be no differences in GHSR1a mRNA translation efficiency and GHSR1a protein level among these haplotypes. On the basis of these results, we proposed a translational hypothesis that differences in the RNA secondary structure of GHSR1b mRNAs among the 5'UTR microsatellite affect the functional level of the Ghrelin receptor (GHSR1a). The function of GHSR1b has been suggested to regulate GHSR1a expression in the form of GHSR1a/GHSR1b heterodimer [16]. Furthermore, the GHSR1b attenuates the constitutive activation of phosphatidylinositol-specific phospholipase C by ghrelin receptors but has no effect on ghrelin-stimulated extracellular signal [21].  [9]. This hypothesis should be validated by a molecular biological study in the future.
The DelR242 in Exon 1 is a 3-bp (AGG) indel (Figure 1). This mutation causes a truncated 3-arginine residue (3R) (normal type: 4 arginine residues (4R)) of the third loop of the intercellular loop (ICL) domain of the GHSR1a protein. The 3R is a fundamental allele of the GHSR1a gene in mice, rats and Chinese hamsters. The 4R is a fundamental allele in humans, pigs and sheep. Furthermore, in humans, two missense mutations (p.R237W and p.D246A) adjacent to the 4R region of ICL3 of the GHSR1a protein have been reported in short-stature families [22,23]. These GHSR1a gene mutations displayed a partial loss of constitutive activity in the GHSR1a receptor. We found an interesting breed difference in the 3R allele frequency which was high (0.43) in Japanese Shorthorn and low (0.09) in Holstein-Friesian cattle. This allele was never found in Japanese Black cattle (Figure 1). The 3R allele is widely distributed in many cattle breeds and may be an older allele than the 4R allele [8]. The DelR242 locus and/or the 3R allele may have some significant effects on growth and feeding behavior in Japanese Shorthorn cattle. To further investigate the reason for a higher frequency of the 3R allele of the GHSR1a gene in Japanese Shorthorn cattle than other cattle breeds, a research project is in progress (manuscript in preparation).
To find a species specific motif within the promoter region of the GHSR1a gene in cattle, we compared potential transcriptional regulatory sequences in approximately 2.6 kb of the 5'-flanking region the GHSR1a gene among cattle, humans and mice [8]. Six bovine specific motifs were identified as follows: (1) Apolipoprotein E_B1, (2) A-activator binding site (AABS), (3) S1 nuclease-hypersensitive site, (4) Nuclear protein factors and erythroid specific 1_CS1, (5) Nuclear protein factors I, and (6) Nuclear factor_E1. Apolipoprotein E (apoE) is a major constituent of very low density lipoprotein and can be found associated with all of the major classes of lipoprotein particles [24]. Furthermore, AABS motif is a binding site for C/EBP beta (CCAAT/ enhancer-binding protein beta) [25]. C/EBP beta is a transcriptional regulator of the UCP1 (uncoupling protein-1) gene, the specific marker gene of brown adipocytes responsible for thermo genic capacity [26]. mRNA expression of the GHSR1a gene in cattle may be more coordinated with lipoprotein metabolism than those of humans and mice.

Bovine-specific mRNA expression of the GHSR1a gene
Age-related changes in the GHSR1a mRNA expression have been reported in rats [27] and in mice [28]. Furthermore, comprehensive tissue distributions of the GHSR1a and/or GHSR1b mRNA expressions have been reported in humans [29] and in mice [28]. However, in cattle, no comprehensive tissue distributions of the GHSR1a and/or GHSR1b mRNA expressions in the arcuate nucleus, pituitary gland and other bovine tissues had been reported in cattle. Therefore, in order to develop a better understanding of the agerelated functions of GHSR1a and GHSR1b in the hypothalamus/ pituitary-mediated regulation of GH secretion and feeding/growth in cattle, we examined the age-related changes in the GHSR1a and GHSR1b mRNA expressions in several tissues including the arcuate nucleus and pituitary gland by real-time PCR [30].
Age-related changes in relative expression levels of the GHSR1a and GHSR1b mRNAs in five tissues including the arcuate nucleus are shown in Table 1. The expression level of GHSR1a mRNA in the arcuate nucleus during the post-weaning age was more than 10-fold higher than in pre-weaning calves and the mature cow. The expression level of GHSR1b mRNA did not change significantly among age groups. In the pituitary gland, the expression level of GHSR1a and GHSR1b mRNAs declined with age.  Ghrelin / GHSR stimulate appetite in the arcuate nucleus [1,31]. In the arcuate nucleus, the ghrelin-containing neurons send efferent fibers onto neuropeptide Y (NPY)-and agouti-related protein (AgRP)expressing neurons to stimulate the release of these orexigenic peptides and onto proopiomelanocortin (POMC) to suppress the release of anorexigenic peptide in rodents [32]. In sheep, offering feed ad libitum (resulting in greater ME intake), decreased hypothalamic mRNA expression of NPY and AgRP and tended to increase that of POMC compared with feed-restricted wethers [33]. The GHSR1a mRNA detected in NPY and GHRH neurons in the arcuate nucleus and GHSR1a are involved in the up-regulation of NPY and GHRH expression in the arcuate nucleus [34]. In cattle, absolute body weight gain (kg) per month is larger in the post-weaning period than during the pre-weaning phase and adulthood [35]. Voluntary feed intake increases significantly with age and reaches or exceeds 'adult' levels within 6 weeks after weaning [36]. Furthermore, voluntary feed intake per unit of metabolic weight (dry matter (g) / live weight (kg)/ day) from weaning to sexual maturity shows a steady decline with increasing weight [37]. Itoh et al. [38] and ThidarMyint et al. [39] observed that GH response to ghrelin and growth-hormone-releasing hormone (GHRH) stimulation in post-weaning calves is greater than in pre-weaning calves and cows, but no synergistic effects of ghrelin indicates that in the post-weaning period, the very high expression of GHSR1a mRNA and relatively lower expression of GHSR1b mRNA in the arcuate nucleus, dramatically amplify ghrelin signaling that stimulates the release of orexigenic peptides (eg, NPY, AgRP, GHRH). These conditions also suppress the release of anorexigenic peptides (eg, -melanocyte stimulating hormone) as well as the secretion of GH in cattle. Therefore, post-weaning calves exhibit a very high voluntary feed intake.
In peripheral tissues, there were 3 marked differences in mRNA expression between cattle (ruminants), humans and mice (monogastric animals), as follows: (1) GHSR1a mRNA expression in the liver is high in cattle and very low in humans and mice; (2) GHSR1b mRNA expression in the liver is low in cattle and high in humans; (3) GHSR1b mRNA expression in the pancreas is very high in cattle (Table 2).
Murata et al. [40] reported that GHSR1a mRNA is expressed in hepatoma cells and that ghrelin up-regulates the mRNA level of phosphoenolpyruvate carboxykinase (PEPCK), which is the ratelimiting enzyme of glyconeogenesis and modulates downstream molecules involved in insulin-signaling in humans. Furthermore, GHSR1a and ghrelin mRNAs are expressed in human T lymphocytes and monocytes, where ghrelin acts via GHSR1a to especially inhibit the expression of pro-inflammatory anorectic cytokines such as IL-6, IL-1β and TNF-α [41]. IL-6 is well-known to inhibit the expression of the glyconeogenic genes (phosphoenolpyruvate carboxylase-1(PCK-1) and glucose 6-phosphatase (G6PC)) in the liver via the signal transducer and activator of the transcription-3 (STAT-3) signal cascade [42]. Since hepatic and renal gluconeogenesis is crucially important in glucose metabolism in ruminants [43], the high expression of GHSR1a mRNA and the low expression of GHSR1b mRNA in the liver seem to be important for gluconeogenesis and represent the bovine-specific expression pattern for maintaining glucose homeostasis in cattle. The very high expression of splenic GHSR1b mRNA described may attenuate the inhibition of proinflammatory anorectic cytokines and leptin-induced anorectic cytokine expression in monocytes and T cells by the action of ghrelin via GHSR1a for maintaining the immune system and appetite control. Ghrelin is expressed in pancreatic islets and there appears to be a negative association between ghrelin and insulin secretion from the pancreas in vivo [44]. Ghrelin has been known to function as a potent inhibitor of pancreatic cholecystokinin (CCK)-induced exocrine secretion in rats [44], and the CCK A and CCK B /gastrin receptors, which are G protein-coupled receptors (GPCRs), are expressed in the pancreas in cattle [45]. Moreover, pancreatic polypeptide (PP) is expressed by PP cells in the endocrine pancreas and is released in response to meals as an anorexigenic peptide. A receptor with a high affinity for PP, the Y4 receptor, which is also a GPCR, is expressed in the human pancreas [46]. In addition, the GHSR1b and neurotensin receptor 1 (NTSR1) have been shown to be overexpressed in human non-small cell lung cancers (NSCLC), and that a heterodimer complex of these receptors (GHSR1b/NTSR1) functioned as a neuromedin U (NMU) receptor. A very high expression of GHSR1b mRNA in the pancreas may support the hypothesis that the GHSR1b alters the basal expression of GHSR1a by the GHSR1a -GHSR1b heterodimer formation [16]. Furthermore, other GPCRs expressed in the pancreas, such as the CCK A and CCK B /gastrin receptors and Y4 receptor, may interact with GHSR1b to alter basal expression of these receptors and to ensure a ready response to changes in nutritional / physiological body conditions. Further investigation on the relationship among these receptors is needed.
Bovine-specific nucleotide polymorphisms and mRNA expression of the GHSR1a gene described in this mini review will contribute to a better understanding of functions of the GHSR1a and GHSR1b.