Associations of DNA polymorphisms in growth hormone and its transcriptional regulators with growth and carcass traits in two populations of Brangus bulls

INTRODUCTION

Growth hormone (GH) has been used as a functional and positional candidate gene in genotype to phenotype association studies in several species, including bovines (i.e., Bos taurus and Bos indicus) and Homo sapiens (Ge et al., 2003; Beauchemin et al., 2006; Rudd et al., 2006). Rationale for choosing this hormone as a candidate gene includes its role in growth, lactation, carbohydrate metabolism, and many other aspects of homeorhesis (Ohlsson et al., 1998; Akers, 2006; Ayuk and Sheppard, 2006). Synthesis and secretion of GH are regulated by hypothalamic releasing factors, somatotrophic transcription factors, as well as a plethora of endocrine feedback signals (Giustina and Veldhuis, 1998; Pfaffle et al., 1999; Fodor et al., 2006).

The GH locus is on bovine chromosome 19; Lagziel et al. (2000) suggested that a sequence polymorphism in intron 4 of the GH gene could be used to differentiate humped (B. indicus) from humpless (B. taurus) cattle. Other comparisons among these types of cattle, particularly Angus and Brahman breeds, revealed phenotypic differences in carcass traits (Morrison, 2005). Admixed populations of British-B. taurus cattle (i.e., Angus or Hereford) and B. indicus cattle (i.e., Brahman) have been useful for various investigations, particularly quantitative trait locus (QTL) detection (Casas et al., 2003; Kim et al., 2003) and physiology delineations. In brief, Brahman cattle tend to have lower levels of adiposity and serum concentrations of leptin, but greater serum concentration of GH than do Angus cattle (Carroll, 1996; Thomas et al., 2002; Lopez et al., 2006).

Association studies and validation procedures using small numbers of DNA polymorphisms (i.e., markers) typically assume additivity, if the markers are within a gene or in linkage (Van Eenennaam et al., 2007). However, in admixed populations of B. taurus and B. indicus cattle, heterozygous genotypes appear to be advantageous for the prediction of physiological measurements and performance traits (Thomas et al., 2004; Pereira et al., 2005). This effect is probably due to allele interactions. Creating admixed populations with Zebu (i.e., B. indicus) cattle most likely augments this effect as these cattle were found to have high frequencies of homozygosity in single nucleotide polymorphisms (SNPs) discovered in sequencing projects of B. taurus cattle. This is particularly true for polymorphisms in genes of the GH axis (Beauchemin et al., 2006). Herein, we evaluated frequencies and associations of genotypes and haplotypes to phenotypes of bull development in a composite B. indicus x B. taurus breed of cattle (Brangus = 3/8 Brahman x 5/8 Angus).

MATERIAL AND METHODS

Animal model

Phenotype information on 434 Brangus bulls was collected based on guidelines described by the Beef Improvement Federation (BIF, 2006) for gain-tested bulls. From 1997 to 2003, the tested bulls were from the New Mexico State University Chihuahuan Desert Rangeland Research Center (CDRRC; progeny of 14 sires). From 2004 to 2005, the test included bulls from a cooperative breeding program (COOP; Mound Creek Ranch, Leona, TX; progeny of 34 sires). The two time periods were considered as separate populations/data sets. Complete pedigree information was known for all bulls, and bulls were registered with the International Brangus Breeders Association (San Antonio, TX). Subsets of bulls from reference sires in each breeding program were included in both performance tests from 2004 to 2005 (the COOP population). Reference sires included at least two bulls from the CDRRC breeding program and two bulls from the Mound Creek breeding program. A reference sire had to have at least two progeny at the two bull test locations. In the COOP population, contemporary groups were assigned based on location of weaning and location of post-weaning gain test. A brief description of the two breeding programs follows:

Chihuahuan Desert Rangeland Research Center

This Brangus herd was initiated in 1966; cows were managed as spring-calving range cattle with limited supplementation and assistance with dystocia. The annual breeding season was from May 1 to August 1 with artificial insemination and natural service matings. Within two days after birth of each calf, date of birth, birth weight, and calf gender were recorded and calves were assigned a unique identification number. Three to four months following birth, calves were vaccinated for clostridial, complex viral, and pasturella diseases. Vaccinations were repeated two to four weeks after the first vaccination. Calves were weighed at weaning at ~205 days of age. Post-weaned calves were given a two- to four-week acclimation period prior to the start of a gain test. The gain test ration was formulated to achieve a 1.5 kg head^-1 day^-1 gain with a corn-alfalfa-based diet (14.9% protein and 75% total dissolved nitrogen) for 112 days. Bulls were weighed every 28 days to obtain a linear plot of days versus gain to estimate average daily gain (ADG; slope of the line from y = mx + b). At ~365 days of age, bulls were given a breeding soundness exam to measure scrotal circumference and insure reproductive functionality. Simultaneously, carcass traits were measured with ultrasound.

Mound Creek Ranch

This organization has been breeding Brangus cattle for 17 years. For spring calving, breeding season was from May 21 to August 21 with the use of AI and natural service matings. Cows were maintained on Bermuda (Cynodon dactylon) and Bahia (Paspalum notatum) grass pastures and supplemented with hay of these grasses in winter as needed. Date, weight, and hair color code were recorded at birth and calves assigned their individual identification. Three to four months after birth, calves were provided creep feed and administered vaccinations for clostridial, complex viral, and pasturella diseases. At ~205 days of age, vaccines were re-administered, calves were de-wormed, and weaned. Post-weaned bull calves were weighed and allowed to graze Bermuda grass ad libitum. Prior to the start of this test, calves were acclimated with a medicated-receiving ration for seven days, followed by a two-week acclimation period with a ration formulated for a 1.5 kg head^-1 day^-1 gain. This was a corn-based supplement (12.0% CP and 70.7% TDN) to grazing on Bermuda grass. When the calves reached ~365 days of age, they were weighed, scrotal circumference recorded, and carcass traits measured via ultrasound. At this time, bulls were re-vaccinated for clostridial and respiratory diseases and de-wormed. Bulls were weighed at the beginning and end of the gain test and these data were used to estimate average daily gain (ADG = (final weight - initial weight)/number of days on test).

Carcass measurements

Carcass trait data included real-time ultrasound measurements of fat thickness at the 12th and 13th rib, longissimus muscle area/kg of body weight, and intramuscular fat percentage of the longissimus muscle by a Centralized Ultrasound Processing certified technician (Perkins et al., 1992).

DNA extraction and genotyping

Genotyping and DNA extraction were done using procedures described by Beauchemin et al. (2006). In brief, blood was collected with vacuum tubes coated with ethylenediaminetetraacetic acid, the post-centrifugation white blood cell supernatant (i.e., buffy coat) was recovered, and then DNA extracted with QIAamp DNA Blood Mini Kit (#51104, Qiagen, CA, USA). The genotypes were determined (Table 1). Allelic discrimination was accomplished using fluoro-probes and real-time PCR or PCR, followed by digestion of amplicon with target restriction enzyme and subsequent fragment separation using ethidium bromide-stained 3% NuSieve agarose gels and electronic imaging. Two of the sequence polymorphisms were assayed as SNP, GH leucine (C) to valine (G) SNP and Prop-1 histadine (A) to arginine (G). The GH sequence polymorphism in intron 4 was assayed as an RFLP and presented as a C/T SNP. The Hinf-I RFLP in Pit-1 was represented as an A/G SNP. Data were represented as SNP for ease of presentation as haplotypes. The two sequence polymorphisms in the GH gene were in intron 4 and exon V, which were 1547 and 1758 bp from the 5’ start of the gene.

Statistical analysis

Prior to statistical analyses, birth weight, 205-day weight, and 365-day weight traits were adjusted by age and age of dam according to BIF guidelines (2006). Statistical analyses were conducted for each population of bulls using SAS (SAS Inst. Inc., Cary, NC, Ver 9.1.3), which includes functions for genetic analyses (Saxton, 2004). Assumptions of normality of data distribution and equality of variances within contemporary group, sire, and genetic categorizations were tested (Littell et al., 2002). A sire was required to have at least two progeny to be included in these analyses. Allelic, genotypic, and haplotype frequencies were estimated using the program routines, proc allele and proc haplotype. Linkage disequilibrium was estimated for the two loci in the GH gene. This procedure yielded tests of chi-square distribution, correlation coefficient (r²), and Lewontin’s D’ estimate.

The genetic effect of the sequence polymorphisms was evaluated in prediction analyses using mixed model methodology (Littell et al., 1996). The model was:

where y_ijklmn = phenotypic value of trait,
µ = population mean,
A_i = fixed effect of genotype, haplotype, or probability of haplotype,
B_j = fixed effect of contemporary group,
C_k = fixed effect of year,
D_l = covariate of Julian birthday (i.e., age of bull),
E_m = fixed effect of the age of dam categories (BIF, 2006),
F_n = random effect of sire nested within contemporary group using the Z statistic to test if
Ho: s_w² = 0 (Littell et al., 1996), and
e_ijklmn = random residual error.

If the trait was adjusted for known environmental effects prior to analyses (e.g., age of dam adjustments for birth weight and age of calf adjustments for weaning weight), Julian birthday and age of dam terms were omitted from the model. Year was only included in predictions involving the CDRRC population, since the data collected from the COOP population was for a single year. The genetic terms (i.e., A_i) were evaluated in four formats: 1) fixed effect of genotype; this was the genotype of a single locus for Pit-1 and Prop-1; however, for GH, this term involved the combination of two loci (i.e., prior to combining these two genotypes, the interaction of the two genotypes was determined to be significant; P < 0.05), 2) fixed effect of a single haplotype, 3) fixed effect of the four haplotypes fitted simultaneously, and 4) effect of the probability of an individual possessing each of the four haplotypes fitted as covariates (i.e., these four haplotypes were fitted in the model simultaneously). If the variances of the genetic terms were heterogenous then a repeated measures analysis was used that accounted for that heterogeneity. The interactions of the four SNPs were also tested with this model. If the genetic term (i.e., A_i) or an interaction of the terms were detected as significant (P < 0.05) source(s) of variation, means were separated with pre-planned pair-wise comparisons of least squares means. To evaluate the proportion of variation attributed to a specific genotype, simple predictions were conducted using a general linear model and ascertaining the coefficient of determination (R²).

RESULTS

Growth and carcass traits (Table 2) and allelic and genotypic frequencies (Table 3) were recorded. Combined genotypes for the two loci in the GH gene and haplotypes were also recorded (Table 4). Haplotypes from the two GH gene sequence polymorphisms appeared to be in linkage disequilibrium in both populations of bulls (c² » 36, P < 0.001; r² » ± 0.4; D’ » ± 0.8).

Sire and year were significant (P < 0.05) sources of variation in the prediction of traits in the CDRRC population. Sire and contemporary group were significant predictors (P < 0.05) of traits in the COOP population. Age and age of dam were not significant sources of variation for traits that were not pre-adjusted for age of the animal and age of dam; however, these terms were included in these models because of their biological relevance. Pit-1 and Prop-1 genotypes were not significant sources of variation in prediction analyses in the CDRRC population (Table 5). However, in the COOP population, Pit-1 genotype was a significant (P < 0.05) source of variation in prediction of intramuscular fat (%), as was the Prop-1 genotype for prediction of scrotal circumference. Specifically, bulls of the Pit-1 GG genotype appeared to have greater (P < 0.05) intramuscular fat (%) than bulls of the AA or AG genotype, and bulls of the Prop-1 AA genotype appeared to have greater (P < 0.05) scrotal circumference than bulls of the AG or GG genotype (Table 5). Interactions were detected (P < 0.07) among genotypes of Pit-1 and Prop-1 and among the two genotypes with the combined genotypes in the GH gene in prediction of ADG. This interaction was detected in both populations; however, a mean separation test did not reveal a consistent observable trend inferring an advantageous genotype combination. Coefficient of determination from a simple prediction of traits with Pit-1 and Prop-1 genotypes accounted for less than 5% of the variation in these two populations.

The combined genotypes of the sequence polymorphisms in the GH gene were significant (P < 0.01) predictors of birth weight, scrotal circumference, and fat thickness in the CDRRC population. These genotypes also tended (P < 0.09) to predict ADG, while the significance level of genotype was 0.15 for 365-day weight and 0.18 in the prediction of intramuscular fat (%). In the COOP population, GH genotype was a significant predictor (P < 0.05) of intramuscular fat (%), fat thickness, and longissimus muscle area per unit of body weight. The mean separation test includes eight columns of genotypes for comparison (Table 6), which makes it difficult to select advantageous genotype(s) for each trait; however, in general, it appears that heterozygous (CT) genotypes at the intron 4 locus have the most advantageous (P < 0.05) production trait levels when paired with the CC genotype of the exon V locus in the CDRRC population and the heterozygous CG genotype of the exon V locus in the COOP population. Simple prediction analyses of these data suggested these combined genotypes accounted for ³27.7% of the variation in traits of the CDRRC population. The intron 4 locus accounted for ³20.0% of this variation; however, coefficients of determination were <5% in the COOP population.

When haplotype was tested as a single-fixed effect in prediction analyses, the CG haplotype was detected as a significant (P < 0.05) predictor of fat thickness and the TC haplotype was detected as a significant (P < 0.05) predictor of longissimus muscle area per unit of body weight in the CDRRC population (Table 7). The CC haplotype was a significant (P < 0.05) predictor of fat thickness in the COOP population. The CC haplotype was a significant (P < 0.05) predictor of longissimus muscle area per unit of body weight in the CDRRC population, whereas it predicted fat thickness in the COOP population. Models attempting to evaluate the four haplotypes simultaneously as fixed effects were unsolvable, which was probably due to the presence of rare haplotypes. Nonetheless, when haplotypes were fitted based on the probability of the individual possessing the four haplotypes as covariates, the four haplotypes predicted (P < 0.05) intramuscular fat (%) and longissimus muscle area per unit of body weight in the COOP population. The CC, CG, and TG haplotypes predicted (P < 0.05) intramuscular fat (%) and the TG haplotype predicted (P < 0.05) longissimus muscle area per unit of body weight in the CDRRC population.

DISCUSSION

Growth hormone is transcribed and translated in the somatotrophs of the anterior pituitary gland. Transcription is regulated by two powerful DNA-binding factors, known as Pit-1 and Prop-1. Pituitary secretion of GH is stimulated by hypothalamic secretion of GH-releasing hormone, but inhibited by hypothalamic secretion of somatostatin (Giustina and Veldhuis, 1998; Pfaffle et al., 1999; Fodor et al., 2006). An illustration of the many genes involved in the growth hormone-IGF-I endocrine axis/pathway was presented by Farber et al. (2006). Genetic selection for enhanced growth and body leanness has been associated with increased pituitary secretion of GH (Bunger and Hill, 1999; te Pas et al., 2001, 2004). Administration of recombinant GH enhances lactation and carcass lean meat yield (i.e., bovine somatotropin; Etherton and Bauman, 1998; Akers, 2006). Knowledge of these relationships provided rationale for targeting GH as both a functional and positional candidate gene relevant for genetic selection programs in Brangus cattle.

Detection of QTL and association testing of sequence polymorphisms in candidate genes are two approaches used to develop tools to improve complex traits of animal production (Andersson and Georges, 2004). In some initial candidate gene studies, which involved only a few sequence polymorphisms within or linked to a known gene, associations were typically tested in statistical models that assumed that the effects of alleles were additive. Data were presented based on regression on the number of favorable alleles (Van Eenennaam et al., 2007). This procedure was effective in association analyses involving additive SNPs within the µ-calpain gene that were in linkage disequilibrium (Page et al., 2002, 2004; White et al., 2005).

Crossbred cattle of the Angus and Brahman parent breeds are known to exhibit heterosis (Morrison, 2005). Thus, it is possible that in admixed, B. taurus x B. indicus cattle populations such as Brangus cattle, alleles are not additive. Thomas et al. (2004) and Pereira et al. (2005) suggested that heterozygous genotypes could be favorable in associations with performance traits or physiological measurements. It is possible that the alleles from these GH loci in these Brangus populations have some type of allele interaction that influences the phenotypes. Knowledge of these potential allelic interactions led to the design of the statistical models that we used in this study.

Extensive linkage disequilibrium should be expected in admixed/crossbreed populations (Zhao et al., 2003). These types of populations have proven to be useful for QTL detection, even though chromosomal coverage was somewhat limited at the time these initial studies were conducted (Casas et al., 2003; Kim et al., 2003). In general, preliminary evidence suggests cattle have a high level of linkage disequilibrium (~13 Mb; Khatkar et al., 2006). Linkage disequilibrium was detected among the two GH loci that we evaluated in the two populations in our study. This result was expected as the loci of these sequence polymorphisms were only separated by 211 bp. Additional research is needed to delineate the genetic effects of heterosis and allele interactions to define the role of linkage disequilibrium when testing for associations of genotypes with phenotypes in admixed populations. An analysis that involves identity by descent of haplotypes or haplotype blocks could help with these statistical challenges. Li et al. (2002) and Stone et al. (2005) successfully used these approaches for the detection of associations of haplotype with phenotype; however, the cattle evaluated in those studies were primarily composites of B. taurus breeds.

Associations involving transcriptional regulators of the GH gene, Pit-1 and Prop-1, were single locus associations in our study. Only in the COOP population involving 34 sires, was there enough of a distribution of these bi-allelic genotypes to detect an association. Typically, the minor allele frequency must be greater than 10% for association testing (Abecasis et al., 2001). Nonetheless, there must have been a large enough effect of the GG genotype on the Pit-1 and AA genotype for Prop-1 to detect significant association with intramuscular fat (%) and scrotal circumference in this population. Since the frequency of the GG genotype in Pit-1 was quite low, these results suggest that selective breeding for this genotype would effectively improve the intramuscular fat (%) trait. However, this may not be feasible biologically. The Pit-1 protein localized in regions of the somatotrope nucleus conducting transcription and expression of this gene in its wild-type form was not found to be correlated with pituitary secretion levels of GH in a study of growing wethers (Mancini et al., 1999; Thomas et al., 2000).

Most of our knowledge on the endocrine importance of Pit-1 was derived from mutated/non-functional protein studies. This is also true for the prophet of Pit-1, Prop-1, which binds upstream to Pit-1 in the GH gene promoter/enhancer region (Pfaffle et al., 1999; Guy et al., 2004). Pit-1 has not been found to be a strong predictor of growth and carcass traits or has lacked presence of at least one of the three genotypes in association studies involving beef cattle (Moody et al., 1996; Zhao et al., 2004; Curi et al., 2006).

Our study is one of the first reports of an association study involving Prop-1 in cattle. Showalter et al. (2002) described the DNA sequence mutations in Prop-1 and provided evidence from transcriptional activation assays suggesting that the histadine to arginine non-synonymous SNP at position 173 of the coding sequence greatly influences the ability of this protein to initiate transcription. Non-synonymous SNPs generally infer a gene structure to function relationship and have become a focal point of genomic projects attempting to use SNP data to predict phenotypes of complex traits (Crawford et al., 2005). Cumulatively, the data of this study, together with reports of QTL detection for ovulation rate in cattle (Kirkpatrick et al., 2000; Arias and Kirkpatrick, 2004), suggest that additional fine mapping and(or) candidate gene investigations are justifiable for this region of chromosome 7. The interactions of Pit-1 and Prop-1 with GH genotypes also suggest that the chromosomal regions containing these transcription factors may influence the GH gene through epistasis.

Growth hormone was one of the initial targets in candidate gene association studies in cattle (Taylor et al., 1998; Parmentier et al., 1999; Vukasinovic et al., 1999). Additional reports were published by Barendse et al. (2006), Gao et al. (2006) and Li et al. (2006). There is also evidence suggesting that the exon V non-synonymous SNP coding for leucine versus valine influences pituitary secretion of GH (Grochowska et al., 1999; Sorensen et al., 2002). In our study, the polymorphisms in intron 4 and exon V both appeared to be predictors of growth and carcass traits. Specifically, the heterozygous genotype appears advantageous in the COOP population, whereas the CTCC genotype is notable in predictions involving the CDRRC population. In these two populations, the effects of these two bi-allelic GH loci are more detectable in measures of body fat than for growth traits. This could be explained by the fact that these two breeding programs have been selecting for growth for many generations, but for carcass traits for only a few generations. Thus, there is probably less variation available for partitioning to genotype terms for the growth traits (i.e., 205- and 365-day adjusted weights) than for carcass traits. Note that scrotal circumference is included in these discussions of growth traits, as it has been found to be strongly correlated with growth traits in the CDRRC population (Thomas et al., 2002).

Results from our mixed model analyses suggested that haplotypes are significant predictors of carcass traits. However, an advantageous haplotype was not easily observed, nor were there any similar observable patterns among the two populations when the means were plotted. Since the model assumed allele interaction rather than additivity, haplotypes were fitted as covariates; so mean separation test(s) were not applicable. Haplotype analyses are becoming the norm in association studies, which will continue to develop to account for linkage disequilibrium and identification of tagging SNP (Wall and Pritchard, 2003; Zhao et al., 2003). The reports of Bosse et al. (2005) and Wagner et al. (2005) are examples of the use of haplotypes in association studies involving GH and(or) measures of body fat in humans. Our research demonstrates that haplotypes may contain important information for the prediction of body fat and muscling traits; however, the project needs additional chromosomal coverage to determine if these SNPs are important in a haplotype block on chromosome 19 and to unravel the challenges of determining the significance of a causal SNP mutation within a haplotype block.

A noteworthy observation in our study was the coefficient of determination (R² ³27.7%) in the prediction of growth and carcass traits based on genotypes or haplotypes in the CDRRC population. Most of this value was also attributed to the polymorphism in intron 4. This is probably a unique finding in a relatively small population, and initially, we would expect the non-synonymous SNP in exon V to be more biologically relevant. However, current literature suggests that polymorphisms in introns have substantial relevance, through mechanisms such as alternative splicing and exon shuffling (Roy and Gilbert, 2006). When these markers were tested in the COOP population with a larger number of sires, the coefficient of determination was minimal in the association tests of genotypes/haplotypes with phenotypes. However, though these GH bi-allelic loci were significant sources of variation in these sire-based predictions, the results did not reveal an advantageous haplotype. In summary, polymorphisms in GH and its transcriptional regulators appear to predict growth and carcass traits in Brangus bulls, particularly heterozygous GH genotypes.

ACKNOWLEDGMENTS

We thank Mound Creek Ranch and Cattle Brokers, Inc., for contributions of DNA and data from the Mound Creek bull development program. Research was funded and collaborations supported by the New Mexico Agricultural Experiment Station, the NIH-MBRS-RISE graduate education program, and the Western Education/Extension and Research Activity Committee for Beef Cattle Breeding Research.

REFERENCES

Abecasis GR, Cookson WO and Cardon LR (2001). The power to detect linkage disequilibrium with quantitative traits in selected samples. Am. J. Hum. Genet. 68: 1463-1474.

Akers RM (2006). Major advances associated with hormone and growth factor regulation of mammary growth and lactation in dairy cows. J. Dairy Sci. 89: 1222-1234.

Andersson L and Georges M (2004). Domestic-animal genomics: deciphering the genetics of complex traits. Nat. Rev. Genet. 5: 202-212.

Arias J and Kirkpatrick B (2004). Mapping of bovine ovulation rate QTL; an analytical approach for three generation pedigrees. Anim. Genet. 35: 7-13.

Ayuk J and Sheppard MC (2006). Growth hormone and its disorders. Postgrad. Med. J. 82: 24-30.

Barendse W, Bunch RJ, Harrison BE and Thomas MB (2006). The growth hormone 1 GH1:c.457C>G mutation is associated with intramuscular and rump fat distribution in a large sample of Australian feedlot cattle. Anim. Genet. 37: 211-214.

Beauchemin VR, Thomas MG, Franke DE and Silver GA (2006). Evaluation of DNA polymorphisms involving growth hormone relative to growth and carcass characteristics in Brahman steers. Genet. Mol. Res. 5: 438-447.

Beef Improvement Federation (BIF, 2006). http://www.beefimprovement.org/library/06guidelines.pdf. Accessed August 1, 2006.

Bosse Y, Bouchard L, Despres JP, Bouchard C, et al. (2005). Haplotypes in the phospholipid transfer protein gene are associated with obesity-related phenotypes: the Quebec family study. Int. J. Obes. 29: 1338-1345.

Bunger L and Hill WG (1999). Role of growth hormone in the genetic change of mice divergently selected for body weight and fatness. Genet. Res. 74: 351-360.

Carroll JA (1996). Pituitary gland development and somatotroph function in domestic livestock. Ph.D. dissertation, Texas A&M University, College Station.

Casas E, Shackelford SD, Keele JW, Koohmaraie M, et al. (2003). Detection of quantitative trait loci for growth and carcass composition in cattle. J. Anim. Sci. 81: 2976-2983.

Crawford DC, Akey DT and Nickerson DA (2005). The patterns of natural variation in human genes. Annu. Rev. Genomics Hum. Genet. 6: 287-312.

Curi RA, Palmieri DA, Suguisawa L, de Oliveira HN, et al. (2006). Growth and carcass traits associated with GH1/AluI and POU1F1/HinfI gene polymorphisms in Zebu and crossbred cattle. Genet. Mol. Biol. 29: 56-61.

Etherton TD and Bauman DE (1998). Biology of somatotropin in growth and lactation of domestic animals. Physiol. Rev. 78: 745-761.

Farber CR, Corva PM and Medrano JF (2006). Genome-wide isolation of growth and obesity QTL using mouse speed congenic strains. BMC Genomics 7: 102.

Fodor M, Kordon C and Epelbaum J (2006). Anatomy of the hypophysiotropic somatostatinergic and growth hormone-releasing hormone system minireview. Neurochem. Res. 31: 137-143.

Gao X, Xu XR, Ren HY, Zhang YH, et al. (2006). The effects of the GH, IGF-I and IGF-IBP3 gene on growth and development traits of Nanyang cattle in different growth period. Yi. Chuan 28: 927-932.

Ge W, Davis ME, Hines HC, Irvin KM, et al. (2003). Association of single nucleotide polymorphisms in the growth hormone and growth hormone receptor genes with blood serum insulin-like growth factor I concentration and growth traits in Angus cattle. J. Anim. Sci. 81: 641-648.

Giustina A and Veldhuis JD (1998). Pathophysiology of the neuroregulation of growth hormone secretion in experimental animals and the human. Endocr. Rev. 19: 717-797.

Grochowska R, Zwierzchowski L, Snochowski M and Reklewski Z (1999). Stimulated growth hormone (GH) release in Friesian cattle with respect to GH genotypes. Reprod. Nutr. Dev. 39: 171-180.

Guy JC, Hunter CS, Showalter AD, Smith TP, et al. (2004). Conserved amino acid sequences confer nuclear localization upon the prophet of Pit-1 pituitary transcription factor protein. Gene 336: 263-273.

Khatkar MS, Collins A, Cavanagh JA, Hawken RJ, et al. (2006). A first-generation metric linkage disequilibrium map of bovine chromosome 6. Genetics 174: 79-85.

Kim JJ, Farnir F, Savell J and Taylor JF (2003). Detection of quantitative trait loci for growth and beef carcass fatness traits in a cross between Bos taurus (Angus) and Bos indicus (Brahman) cattle. J. Anim. Sci. 81: 1933-1942.

Kirkpatrick BW, Byla BM and Gregory KE (2000). Mapping quantitative trait loci for bovine ovulation rate. Mamm. Genome 11: 136-139.

Lagziel A, DeNise S, Hanotte O, Dhara S, et al. (2000). Geographic and breed distribution of an MspI PCR-RFLP in the bovine growth hormone (bGH) gene. Anim. Genet. 31: 210-213.

Li C, Basarab J, Snelling WM, Benkel B, et al. (2002). The identification of common haplotypes on bovine chromosome 5 within commercial lines of Bos taurus and their associations with growth traits. J. Anim. Sci. 80: 1187-1194.

Li MH, Adamowicz T, Switonski M, Ammosov I, et al. (2006). Analysis of population differentiation in north Eurasian cattle (Bos taurus) using single nucleotide polymorphisms in three genes associated with production traits. Anim. Genet. 37: 390-392.

Littell RC, Milliken GA, Stroup WW and Wolfinger RD (1996). SAS^® system for mixed models. SAS Institute Inc., Cary.

Littell RC, Stroup WW and Freund RJ (2002). SAS^® for linear models. SAS Institute Inc., Cary.

Lopez R, Thomas MG, Hallford DM, Keisler DH, et al. (2006). Case study: Metabolic hormone profiles and evaluation of associations of metabolic hormones with body fat and reproductive characteristics of Angus, Brangus, and Brahman heifers. Prof. Anim. Sci. 22: 273-282.

Mancini MG, Liu B, Sharp ZD and Mancini MA (1999). Subnuclear partitioning and functional regulation of the Pit-1 transcription factor. J. Cell Biochem. 72: 322-338.

Moody DE, Pomp D, Newman S and MacNeil MD (1996). Characterization of DNA polymorphisms in three populations of Hereford cattle and their associations with growth and maternal EPD in line 1 Herefords. J. Anim. Sci. 74: 1784-1793.

Morrison DG (2005). A compilation of research results involving tropically adapted beef cattle breeds. South. Coop. Ser. Bull. 405. http://www.lsuagcenter.com/en/crops_livestock/livestock/beef_cattle/breeding_genetics/tropical+breeds.htm. Accessed August 1, 2006.

Ohlsson C, Bengtsson BA, Isaksson OG, Andreassen TT, et al. (1998). Growth hormone and bone. Endocr. Rev. 19: 55-79.

Page BT, Casas E, Heaton MP, Cullen NG, et al. (2002). Evaluation of single-nucleotide polymorphisms in CAPN1 for association with meat tenderness in cattle. J. Anim. Sci. 80: 3077-3085.

Page BT, Casas E, Quaas RL, Thallman RM, et al. (2004). Association of markers in the bovine CAPN1 gene with meat tenderness in large crossbred populations that sample influential industry sires. J. Anim. Sci. 82: 3474-3481.

Parmentier I, Portetelle D, Gengler N, Prandi A, et al. (1999). Candidate gene markers associated with somatotropic axis and milk selection. Domest. Anim. Endocrinol. 17: 139-148.

Pereira AP, Alencar MM, Oliveira HN and Regitano LCA (2005). Association of GH and IGF-I polymorphism with growth traits in a synthetic beef cattle breed. Genet. Mol. Biol. 28: 230-236.

Perkins TL, Green RD, Hamlin KE, Shepard HH, et al. (1992). Ultrasonic prediction of carcass merit in beef cattle: evaluation of technician effects on ultrasonic estimates of carcass fat thickness and longissimus muscle area. J. Anim. Sci. 70: 2758-2765.

Pfaffle RW, Kim C, Blankenstein O and Kentrup H (1999). GH transcription factors. J. Pediatr. Endocrinol. Metab. 12 (Suppl 1): 311-317.

Roy SW and Gilbert W (2006). The evolution of spliceosomal introns: patterns, puzzles and progress. Nat. Rev. Genet. 7: 211-221.

Rudd MF, Webb EL, Matakidou A, Sellick GS, et al. (2006). Variants in the GH-IGF axis confer susceptibility to lung cancer. Genome Res. 16: 693-701.

Saxton AM (2004). Genetic analysis of complex traits using SAS. SAS Institute Inc., Cary.

Showalter AD, Smith TP, Bennett GL, Sloop KW, et al. (2002). Differential conservation of transcriptional domains of mammalian prophet of Pit-1 proteins revealed by structural studies of the bovine gene and comparative functional analysis of the protein. Gene 291: 211-221.

Sorensen P, Grochowska R, Holm L, Henryon M, et al. (2002). Polymorphism in the bovine growth hormone gene affects endocrine release in dairy calves. J. Dairy Sci. 85: 1887-1893.

Stone RT, Casas E, Smith TP, Keele JW, et al. (2005). Identification of genetic markers for fat deposition and meat tenderness on bovine chromosome 5: development of a low-density single nucleotide polymorphism map. J. Anim. Sci. 83: 2280-2288.

Taylor JF, Coutinho LL, Herring KL, Gallagher DS Jr, et al. (1998). Candidate gene analysis of GH1 for effects on growth and carcass composition of cattle. Anim. Genet. 29: 194-201.

te Pas MF, Freriksen JW, van Bijnen AJ, Gerritsen CL, et al. (2001). Selection for growth rate or against back fat thickness in pigs is associated with changes in growth hormone axis plasma protein concentration and mRNA level. Domest. Anim. Endocrinol. 20: 165-184.

te Pas MF, Visscher AH and de Greef KH (2004). Molecular genetic and physiologic background of the growth hormone-IGF-I axis in relation to breeding for growth rate and leanness in pigs. Domest. Anim. Endocrinol. 27: 287-301.

Thomas MG, Carroll JA, Raymond SR, Matteri RL, et al. (2000). Transcriptional regulation of pituitary synthesis and secretion of growth hormone in growing wethers and the influence of zeranol on these mechanisms. Domest. Anim. Endocrinol. 18: 309-324.

Thomas MG, Enns RM, Hallford DM, Keisler DH, et al. (2002). Relationships of metabolic hormones and serum glucose to growth and reproductive development in performance-tested Angus, Brangus, and Brahman bulls. J. Anim. Sci. 80: 757-767.

Thomas MG, Enns RM, Silver GA and Garcia MD (2004). Growth hormone (GH) gene polymorphisms appear to differentially influence pituitary responsiveness to growth hormone releasing hormone (GHRH) among Bos taurus and Bos indicus-derivative cattle. Int. Plant Anim. Genome XII: 229 (Abstract). http://www.intl-pag.org/12/abstracts/P5j_PAG12_635.html. Accessed August 1, 2006.

Van Eenennaam AL, Li J, Thallman RM, Quaas RL, et al. (2007). Validation of commercial DNA tests for quantitative beef quality traits. J. Anim. Sci. (manuscript in press: E-2006-512).

Vukasinovic N, Denise SK and Freeman AE (1999). Association of growth hormone loci with milk yield traits in Holstein bulls. J. Dairy Sci. 82: 788-794.

Wagner K, Hemminki K, Israelsson E, Grzybowska E, et al. (2005). Association of polymorphisms and haplotypes in the human growth hormone 1 (GH1) gene with breast cancer. Endocr. Relat. Cancer 12: 917-928.

Wall JD and Pritchard JK (2003). Haplotype blocks and linkage disequilibrium in the human genome. Nat. Rev. Genet. 4: 587-597.

White SN, Casas E, Wheeler TL, Shackelford SD, et al. (2005). A new single nucleotide polymorphism in CAPN1 extends the current tenderness marker test to include cattle of Bos indicus, Bos taurus, and crossbred descent. J. Anim. Sci. 83: 2001-2008.

Zhao H, Rothschild MF, Fernando RL and Dekkers JC (2003). Tests of candidate genes in breed cross populations for QTL mapping in livestock. Mamm. Genome 14: 472-482.

Zhao Q, Davis ME and Hines HC (2004). Associations of polymorphisms in the Pit-1 gene with growth and carcass traits in Angus beef cattle. J. Anim. Sci. 82: 2229-2233.