Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Genetics

Identification of single nucleotide polymorphisms in genes involved in digestive and metabolic processes associated with feed efficiency and performance traits in beef cattle12

 

This article in JAS

  1. Vol. 91 No. 6, p. 2512-2529
     
    Received: Aug 15, 2012
    Accepted: Mar 07, 2013
    Published: November 25, 2014


    3 Corresponding author(s): miller@uoguelph.ca
 View
 Download
 Share

doi:10.2527/jas.2012-5756
  1. M. K. Abo-Ismail*†,
  2. M. J. Kelly*‡,
  3. E. J. Squires*,
  4. K. C. Swanson,
  5. S. Bauck# and
  6. S. P. Miller 3
  1. Centre for Genetic Improvement of Livestock, Department of Animal and Poultry Science, University of Guelph, Guelph, Ontario, Canada, N1G 2W0
    Department of Animal and Poultry Science, Damanhour University, Damanhour, Egypt
    Queensland Alliance for Agriculture and Food Innovation University of Queensland, St Lucia, QLD 4072, Australia
    Animal Sciences Department, North Dakota State University, Fargo 58108-6050
    GeneSeek, 4665 Innovation Drive, Suite 120, Lincoln, NE 68521

Abstract

Discovery of genetic mutations that have a significant association with economically important traits would benefit beef cattle breeders. Objectives were to identify with an in silico approach new SNP in 8 genes involved in digestive function and metabolic processes and to examine the associations between the identified SNP and feed efficiency and performance traits. The association between SNP and daily DMI, ADG, midpoint metabolic weight (MMWT), residual feed intake (RFI), and feed conversion ratio (FCR; the ratio of average daily DMI to ADG) was tested in discovery and validation populations using a univariate mixed-inheritance animal model fitted in ASReml. Substitution effect of the T allele of SNP rs41256901 in protease, serine, 2 (trypsin 2; PRSS2) was associated with FCR (–0.293 ± 0.08 kg DMI kg–1 BW gain; P < 0.001) and RFI (–0.199 ± 0.08 kg; P < 0.01) and although not significant in the validation population, the phase of association remained. In the cholecystokinin B receptor (CCKBR) gene, genotypes in rs42670351 were associated with RFI (P < 0.05) whereas genotypes in rs42670352 were associated with RFI (P = 0.002) and DMI (P < 0.05). Substitution of the G allele in rs42670352 was associated with DMI (–0.236 ± 0.12 kg; P = 0.055) and RFI (–0.175 ± 0.09 kg; P = 0.05). Substitution of the G allele of SNP rs42670353 was associated with ADG (0.043 ± 0.02 kg/d; P < 0.01) and FCR (0.114 ± 0.05 kg BW gain kg–1 DMI; P < 0.05). In the validation dataset, SNP rs42670352 in gene CCKBR was significant for RFI and DMI and had the same phase of associations; SNP rs42670353 was significantly associated with FCR with same phase of association and the C allele in SNP rs42670351 was validated as decreasing DMI, RFI, and FCR. Substituting the G allele of SNP rs42670352 in CCKBR2 was associated with decreasing DMI and RFI in the validation study. New SNP were reported in genes PRSS2 and CCKBR, being associated with feed efficiency and performance traits in beef cattle. The association between these SNP with fertility, carcass, and meat quality traits must still be tested.



INTRODUCTION

A small improvement in feed efficiency would have a significant influence on the profitability of the beef production system (Herd et al., 2003). Residual feed intake (RFI) is one of the acceptable traits for improving feed efficiency in feedlot cattle (Wulfhorst et al., 2010). Estimates of the genetic variation in feed efficiency (Archer et al., 1999; Arthur et al., 2001; Schenkel et al., 2004) indicate that RFI is moderately heritable providing an opportunity for selection although the difficulty of recording feed intake has been reported as a major limitation (Arthur et al., 2001) toward implementation of selection for improved RFI. Accordingly, other criteria to evaluate feed efficiency such as DNA markers have been considered (Barendse et al., 2007; Nkrumah et al., 2007; Sherman et al., 2008a,b, 2010).

Herd et al. (2004) and Richardson and Herd (2004) proposed that processes such as digestion, body composition development, metabolism, including biological processes such as ion pumping, proton leakage, and protein turnover, activity, and thermoregulation contribute to the variation in RFI. There is evidence suggesting that inadequate production of specific digestive enzymes could be responsible for limitation in digestive efficiency. Genes such as pancreatic α amylase (AMY2B) is known as a primary enzyme responsible for starch digestion in cattle fed high-concentration diets (Swanson et al., 2002). In pigs, the concentration of the pancreatic trypsin enzyme was positively associated with ADG and negatively associated with feed conversion ratio (FCR: the ratio of average daily feed intake to ADG; Van den Borne et al., 2007). Therefore, cationic trypsin (LOC780933), pancreatic anionic trypsinogen or protease, serine, 2 (trypsin 2; PRSS2; Le Huerou et al., 1990), and pancreatic trypsin inhibitor (PTI; Ascenzi et al., 2003) are potential candidate genes. In human, polymorphisms in CT activation peptides were associated with pancreatitis (Chen and Ferec, 2000; Teich and Mossner, 2008; Kereszturi et al., 2009). Cholecystokinin (CCK) B receptor (CCKBR) regulates effect of CCK (Le Meuth et al., 1993; Le Dréan et al., 1999; Rehfeld et al., 2007). Uncoupling protein 2 (UCP2) provided a link between mitochondrial respiration and feed efficiency in beef cattle (Kolath et al., 2006; Sherman et al., 2008b) and obesity and insulin secretion (Zhang et al., 2001). Pyruvate carboxylase (PC) is one of the key enzymes playing a potential role in gluconeogenesis (in liver and kidney), lipogenesis (in adipose tissue and lactating mammary gland), and insulin signaling pathway (in pancreatic islets; Jitrapakdee and Wallace 1999; Greenfield et al., 2000; Velez and Donkin, 2005; Haga et al., 2008). The adenosine triphosphatase (ATPase), H+ transporting, lysosomal 56/58 kDa, V1 subunit B2 (ATP6V1B2) is involved in transmembrane transport, hydrolase activity, proteolysis, and generation of precursor metabolites and energy (Jefferies et al., 2008; Appendix 1). In addition, other isoform of Vacuolar-type H+-ATPase is involved in the regulation of insulin secretion from pancreatic β-cells (Sun-Wada et al., 2006). Genes involved in these processes (Appendix 1) are good candidates for improving feed efficiency.

Identifying new SNP in these candidate genes significantly associated with RFI would be beneficial. Therefore, objectives were to discover new SNP in 8 genes involved in digestive and metabolic processes and examine the relationships between these SNP and feed efficiency and performance traits and validate these associations in a more recent group of cattle.


MATERIALS AND METHODS

Discovery Study

Phenotypic Data Collection.

The study was approved from The University of Guelph Animal Care Committee based on the recommendations outlined in the Canadian Council on Animal Care (1993) guidelines. Animals were born in 1 of 3 University of Guelph cooperative herds, the University of Guelph Elora Beef Research Centre (EBRC), University of Guelph New Liskeard Agriculture Research Station, and the Agriculture Agri-Food Canada Kapuskasing Experimental Farm, or purchased from commercial sources. Calves were weaned at approximately 200 d of age and were involved in various postweaning trials at the EBRC with different nutritional treatments over time from 1998 to 2007. Phenotypes were collected from an average of 660 crossbred animals, heifers (40), steers (363), and bulls (257). Average breed contributions were Angus (AN; 41%), Simmental (SM; 24%), Piedmontese (PI; 11%), Charolais (CH; 8%), Gelbvieh (GV; 4%), and Limousin (LIM; 1%) determined by pedigree information on the ancestors. Body weight was recorded a number of times over the course of trials, with most trials recording BW at least every 4 wk. The ADG of the animals was calculated as a linear regression coefficient of their BW on the actual 112 d of measurement using nlme package from R software (Pinheiro et al., 2011). The R2 for all growth curves averaged from 0.85 to 0.99. Midpoint metabolic weight (MMWT) was calculated as the midpoint BW to the power 0.75. Daily DMI data were acquired by 2 automated feeding systems: Calan-gate (American Calan, Northwood, NH; Ferris et al., 2007) and Insentec (Insentec, Marknesse, the Netherlands; Chapinal et al., 2007) systems where DMI data was filtered to exclude outlier records or days due to mechanical problems. The DMI was calculated for each animal as total DMI divided by number of days during the test period. The RFI was calculated from the difference between the average of the actual daily DMI and the expected DMI of the animal (Koch et al., 1963; Arthur et al., 2001). Expected DMI was determined through the regression coefficients estimated from the data using a multiple phenotypic regression model as follows:in which yijk is the DMI for animal k during the feeding period, µ is the overall mean, β1 is the regression on ADG as determined through a linear regression of BW on days of trial as described above, β2 is the coefficient of the linear regression on MMWT, Sexi is the effect of the ith sex, TTY is the effect of jth treatment × trial × year (38 levels), and eijk is the residual random effect associated with animal k and is the resulting RFI used in further analyses. The descriptive statistics of the traits are given in Table 1.


View Full Table | Close Full ViewTable 1.

Descriptive statistics in feedlot beef cattle for performance and feed efficiency in the training and validation datasets

 
Mean
SD
Minimum
Maximum
Trait1 Training Validation Training Validation Training Validation Training Validation
ADG, kg d–1 1.81 1.70 0.38 0.39 0.60 0.71 3.29 3.30
MMWT, kg 103.30 92.39 14.94 11.70 68.32 53.25 157.70 128.10
DMI, kg d–1 9.89 9.81 1.60 1.76 5.38 4.18 15.64 15.54
RFI, kg d–1 –0.12 –0.07 0.89 1.13 –5.62 –3.70 3.84 3.35
FCR, kg gain kg–1 DM 5.68 6.09 1.52 1.87 2.68 3.11 18.74 16.76
1MMWT = midpoint metabolic weight; RFI = residual feed intake; FCR = feed conversion ratio.

Single Nucleotide Polymorphism Discovery.

An in silico study was conducted to discover SNP within genes AMY2B, LOC780933, PRSS2, PTI, and CCKBR, UCP2, PC, and ATP6V1B2 using the available expressed sequence tags (EST) or whole genome shotgun (WGS) traces in GenBank (Benson et al., 2005). The SNP were discovered in the candidate genes in silico in 3 main steps. The first step was to obtain the EST or WGS required for the alignment as follows: a) The reference sequence (cDNA) in FASTA format was acquired from the GenBank in the National Center for Biotechnology Information (NCBI), b) The reference sequence was aligned with the cow sequence using the basic local alignment search tool (BLAST; Zhang et al., 2000) at http://blast.ncbi.nlm.nih.gov, c) the Traces-EST (ftp://ftp.ncbi.nih.gov/repository/dbEST/) or WGS databases (Benson et al., 2012) and the MegaBlast program (Morgulis et al., 2008) using the default parameters were used, and d) traces were selected and acquired from the Trace archive including standard chromatography files (SCF).

The second step was the SNP identification process. In this step, a DNA sequence assembly software called Sequencher (Gene Codes Corporation, Ann Arbor, MI) was used to align the acquired sequences with the reference sequence. Using this software, SNP were detected based on the nucleotide sequences and attached standard chromatography files. The SNP that lead to a change in the sequence of AA were also detected.

The third step was to determine the position of SNP within the gene sequence. Briefly, the whole gene sequence was acquired from NCBI in FASTA format. Then it was aligned with the reference (cDNA) and the EST using Sequencher software. The flanking sequences of the SNP were obtained for genotyping purposes. In total, 39 SNP from 8 genes were selected from the in silico results. Eighteen of the SNP have not been previously reported in the public domain (Table 2) whereas 21 SNP have been reported in the SNP database in NCBI (Table 3).


View Full Table | Close Full ViewTable 2.

Gene name, chromosome number (Bos taurus autosome (BTA), GenBank Entrez Gene Identifier, SNP name, SNP position, nucleotide change, and functional consequences for SNP discovered in silico

 
Gene BTA: gene ID SNP name SNP position1 AA change2 SNP
AMY2B 3:539383 AMY2B1 39936868 Ser/Asn G/A
AMY2B2 39935769 Arg/Arg C/T
AMY2B4 39891301 Thr/Ala A/G
LOC780933 4:780933 CT1 106362449 Ala/Ala C/T
CT2 106361029 Pro/Pro C/T
CT3 106360880 Ser/Cys C/G
CT4 106359896 Ser/Phe C/T
CT5 106359889 Ser/Ser T/C
CT6 106359796 Ala/Ala C/T
PRSS2 4: 282603 TRYP81 106888934 Ala/Ala A/G
PTI 13:404172 PTI2 74947682 Val/Ile C/T
PTI3 74947651 Ala/Val G/A
PTI4 74944796 Pro/Ser C/T
PTI6 74944705 Arg/Lys G/A
UCP2 15:281562 UCP22 54197781 Ala/Ala T/C
UCP23 54197685 Ala/Ala A/G
UCP24 54197451 Tyr/Tyr C/T
1The SNP position is based on Bos taurus UMD 3.1, Genome Build 37.3 (Zimin et al., 2009).
2The effect of mutation (SNP) on the AA sequence.

View Full Table | Close Full ViewTable 3.

Gene name, GenBank Entrez Gene Identifier, chromosome number, SNP name, SNP position, nucleotide change, and functional consequences for SNP reported in the National Center for Biotechnology Information

 
Gene BTA and gene bank ID1 SNP name Accession number SNP position2 AA change3 SNP
AMY2B 3:539383 AMY2B6 rs42312301 39931232 Asp/Asn G/A
PRSS2 4: 282603 TRYP82 rs41256900 106888943 Ser/Ser C/T
TRYP83 rs41256901 106890553 Ser/Phe T/C
ATP6V1B2 8:338082 ATPase1 rs43563470 67810773 Asp/Asp C/T
ATPase2 rs43562811 67823611 3′ UTR C/T
ATPase3 rs43562810 67823802 3′ UTR C/T
ATPase4 rs43562809 67824091 3′ UTR A/G
PTI 13:404172 PTI1 rs43024409 74947703 Met/Leu A/T
PTI5 rs41257167 74944767 Ile/Met T/G
PTI8 rs430243454 74943702 3′ near gene T/C
CCKBR 15:281665 CCKBR1 rs42670351 47386394 Arg/Arg A/C
CCKBR2 rs42670352 47385604 Ala/Ala G/T
CCKBR3 rs42670353 47385334 Phe/Phe C/T
UCP2 15:281562 UCP21 rs41255549 54199080 Ala/Ala G/T
UCP25 rs41774217 54196971 Cys/Cys A/G
PC 29:338471 PC1 rs42194938 45602034 Intronic A/G
PC2 rs42194937 45601239 Intronic G/T
PC3 rs42195008 45529368 Ile/Ile A/G
PC4 rs42197374 45510553 Val/Ile A/G
PC5 rs42197375 45510113 Tyr/Tyr C/T
PC6 rs42197376 45508443 3′ UTR A/G
1BTA = Bos taurus autosome.
2The SNP position is from Bos taurus UMD_3.1, Genome Build 37.3 (Zimin et al., 2009).
3The effect of mutation (SNP) on AA sequence. UTR = the untranslated region.
4This SNP was merged to rs41257168 SNP.

Animal Genotyping.

Genomic DNA was extracted from tissue or blood samples (Sambrook et al., 1989; Rudi et al., 1997; Caldarelli-Stefano et al., 1999). Prepared DNA samples were sent to Merial Ltd. (Lincoln, NE) for genotyping using a commercial platform for high-throughput SNP genotyping and an allele-specific primer extension on a microarray (Pastinen et al., 2000; Makridakis and Reichardt, 2001). In total, 993 animals were genotyped for 39 SNP for the discovery population.

Statistical Analysis.

Allele frequencies were calculated for each SNP on all genotyped animals. The Hardy-Weinberg equilibrium (HWE) was tested using the likelihood ratio test (G-test), described by Lynch and Walsh (1998):in which Nij and are the observed and the expected number of genotype gij. The extent of linkage disequilibrium (LD) between pairs of SNP was calculated using the haploxt program from the Graphical Overview of Linkage Disequilibrium (GOLD) package (Abecasis and Cookson, 2000). The file of marker haplotypes was prepared using fastPHASE 1.1 (Scheet and Stephens, 2006). Haploview software (Barrett et al., 2005) was used to graphically view the extent of LD, assign the haplotype blocks (i.e., SNP with high LD, D′ > 0.77) based on the 4-gamete rule (Wang et al., 2002), and identify the haplotype-tagging SNP using the TAGGER algorithm (De Bakker et al., 2005).

Genotype Analysis.

Associations of the genotypes for each SNP at a time with the traits were evaluated by genetic analysis using ASReml (Gilmour et al., 2009). An animal model was fitted as follows:in which yjklm is the trait measured in the mth animal of the kth sex and the lth treatment trial-year group, µ is the overall mean for the trait, Gj is the fixed effect of the jth genotype for the SNP considered, Sexk is the fixed effect of the kth sex of mth animal, TTYl is the fixed effect of the lth treatment trial-year group, β1 is the regression coefficient of the linear regression on age at the end of test period (AET) of the mth animal, β2 to β7 are the regression coefficients of the linear regressions on the proportion of AN, CH, LIM, SM, PI, and GV breeds in the mth animal, β8 is the regression coefficient of the linear regression on the percent of heterozygosity (HET) of mth animal, am is the random additive genetic (polygenic) effect of the mth animal, and ejklm is the residual random effect associated with the mth animal record. Assumptions for this model are am: a ∼ N (0, Aσ2a) in which A is the relationship matrix and σ2a is the additive genetic variance and ejklm: e ∼ (0, Iσ2e) in which I is the identity matrix and σ2e is the error variance. The expectations are E(am) = 0 and E(ejklm) = 0 and the variances are Var(am) = σ2a and Var(ejklm) = σ2e. The Aσ2a is the covariance matrix of the vector of animal additive genetic effects and the relationship matrix (A) is assumed to be complete back to the base population.

For the significance level used to assess the results, an overall value of P < 0.05 (α) was used. A modified Bonferroni correction was used [α/(N)1/2; Mantel, 1980] to adjust for multihypotheses testing for controlling type I errors where N is the number of SNP multiplied by the number of traits. Therefore, the modified Bonferroni-corrected significance level in the discovery population is 0.0043 [0.005/(27 × 5)1/2] at α = 0.05.

Allele Substitution Effect Model.

This model included the same effects as the genotypic model except that the genotypic effect was replaced with an allele substitution effect, which is estimated by regressing the phenotype on the number of copies of a given allele (0, 1, or 2) using ASReml.

Validation Study

Tissue or blood samples from 1,032 animals born subsequent to the animals used in the discovery population were prepared and sent to Molecular Supercentre Laboratory Services, University of Guelph, Guelph, Canada, for genomic DNA extraction. Then prepared DNA samples were sent to GeneSeek, Inc., (Lincoln, NE) for genotyping using Illumina Infinium BeadChip with single-base extension assay (Steemers et al., 2006). A total of 1,032 animals were genotyped for validation using 17 out of 39 SNP from the discovery population. The reason for reducing the number of SNP (17) is related to cost and efficiency as some of these SNP were highly linked to each other.

A quality control (QC) procedure was conducted using the GenABEL package (Aulchenko et al., 2007) in R software. The SNP and animals with a low call rate (<90%) were excluded from the analysis. Animals with an estimated high frequency of SNP identical by state ≥ 0.95 were excluded. The SNP with a minor allele frequency <1% (e.g., UCP25, AMY2B6, and CT2 SNP) were excluded from the analysis. Animals with high autosomal heterozygosity 0.446 [false discovery rate < 0.05] were also excluded. Phenotypes not within the mean ± 3 SD were excluded. Contemporary groups (TTY levels) that had fewer than 3 animals were excluded. The QC procedure resulted in 14 SNP and 726 animals being used for further analyses.

The association analysis was performed with a univariate animal model fitting the allele substitution or genotypic effect using ASReml. The model included the same fixed systematic effects as previously stated as well as the fixed effect of herd by year of birth. The modified Bonferroni-corrected significance level in the validation population is 0.0058 [0.05/(15 × 5)1/2] at α = 0.05.


RESULTS

Discovery Population

In Silico Study.

The genotyping success rate ranged from 95 to 98% except for SNP rs42197374 and rs41256900, which had success rates of 0.7 and 0%, respectively. Genotyping results showed that 8 of the 18 SNP discovered using Sequencher have both alleles in the genotyped discovery population (Table 4) whereas the remaining putative SNP were fixed (i.e., only 1 allele was present in the genotyped population). The SNP genotyped for the validation study are summarized in Table 5.


View Full Table | Close Full ViewTable 4.

Genotypic and minor allele frequencies and the Hardy-Weinberg equilibrium for SNP in the discovery population

 
Gene SNP Genotype frequency MAF1 G2
AMY2B rs42312301 GG (0.000) AG (0.999) AA (0.001) 0.50 1,306.803
ATP6V1B2 rs43563470 CC (0.853) CT (0.139) TT (0.007) 0.08 0.33
rs43562811 TT (0.592) CT (0.356) CC (0.052) 0.23 0.02
rs43562810 TT (0.885) CT (0.113) CC (0.002) 0.06 0.61
rs43562809 AA (0.008) AG (0.144) GG (0.847) 0.08 0.54
CCKBR rs42670351 AA (0.598) AC (0.350) CC (0.051) 0.23 0
rs42670352 TT (0.957) GT (0.004) GG (0.039) 0.04 280.853
rs42670353 TT (0.297) CT (0.481) CC (0.221) 0.46 0.93
LOC780933 CT2 CC (0.002) CT (0.999) TT (0.002) 0.50 1,301.403
CT5 CC (0.133) CT (0.867) TT (0.001) 0.43 714.553
PC rs42194938 AA (0.522) AG (0.405) GG (0.073) 0.28 0.24
rs42194937 GG (0.944) GT (0.055) TT (0.001) 0.03 0.06
rs42195008 GG (0.942) AG (0.057) AA (0.001) 0.03 0.04
rs42197374 GG (0.571) GA (0.000) AA (0.429) 0.43 9.563
rs42197375 TT (0.733) CT (0.248) CC (0.019) 0.14 0.14
rs42197376 GG (0.943) AG (0.056) AA (0.001) 0.03 0.04
PTI rs43024409 AA (0.000) AT (0.531) TT (0.469) 0.27 188.043
PTI2 CC (0.187) CT (0.812) TT (0.001) 0.41 571.123
PTI3 GG (0.747) AG (0.253) AA (0.000) 0.13 34.863
rs41257167 GG (0.007) GT (0.993) TT (0.000) 0.50 1,246.633
PRSS2 TRYP81 GG (0.451) AG (0.462) AA (0.087) 0.32 4.033
rs41256901 CC (0.679) CT (0.321) TT (0.000) 0.16 58.413
UCP2 rs41255549 TT (0.497) GT (0.413) GG (0.089) 0.30 0.07
UCP22 TT (0.603) CT (0.352) CC (0.045) 0.22 0.48
UCP23 AA (0.600) AG (0.356) GG (0.044) 0.22 1.01
UCP24 CC (0.902) CT (0.097) TT (0.001) 0.05 1.06
rs41774217 GG (0.971) AG (0.029) AA (0.000) 0.01 0.41
1MAF = minor allele frequency.
2G = the G-test statistic.
3Not in agreement with the Hardy-Weinberg equilibrium

View Full Table | Close Full ViewTable 5.

Genotypic and minor allele frequencies (MAF) for SNP in the validation population

 
Gene ID1 SNP name Accession number Genotype frequency
MAF
539383 AMY2B6 rs42312301 AA (0.000) AG (0.004) GG (0.996) 0.002
338082 ATPase1 rs43563470 TT (0.004) TC (0.133) CC (0.863) 0.07
338082 ATPase2 rs43562811 TT (0.554) TC (0.382) CC (0.063) 0.255
338082 ATPase4 rs43562809 AA (0.006) AG (0.137) GG (0.858) 0.074
281665 CCKBR1 rs42670351 AA (0.624) AC (0.319) CC (0.056) 0.216
281665 CCKRB2 rs42670352 TT (0.629) TG (0.317) GG (0.055) 0.213
281665 CCKBR3 rs42670353 TT (0.329) TC (0.496) CC (0.176) 0.423
780933 CT2 in silico TT (0.999) TC (0.000) CC (0.001) 0.001
338471 PC1 rs42194938 AA (0.062) AG (0.347) GG (0.591) 0.235
338471 PC3 rs42195008 AA (0.002) AG (0.04) GG (0.958) 0.022
338471 PC4 rs42197374 AA (0.447) AG (0.435) GG (0.119) 0.336
338471 PC5 rs42197375 TT (0.775) TC (0.211) CC (0.014) 0.119
338471 PC6 rs42197376 AA (0.003) AG (0.038) GG (0.959) 0.022
404172 PTI1 rs43024409 TT (0.131) AT (0.457) AA (0.412) 0.36
282603 TRYP81 in silico AA (0.101) AG (0.437) GG (0.462) 0.32
282603 TRYP83 rs41256901 TT (0.000) TC (0.316) CC (0.684) 0.158
281562 UCP25 rs41774217 AA (0.001) AG (0.013) GG (0.986) 0.007
1Gene ID = Entrez Gene Identifier.

The population was tested for HWE using a G-test where the G value has a distribution that approximates to χ2 with df equal to the number of genotypes minus the number of alleles. The total number of genotyped animals, the allelic frequencies, and the G value for each SNP are reported in Table 4. The SNP rs42312301, rs42670352, CT2, CT5, rs42197374, rs43024409, PTI2, rs41257167, and rs41256901 were not in HWE.

The values of LD (r2) for each marker pair on a given chromosome are presented in Table 6. Values ranged from 0.0 to 0.99. The LD between SNP within CCKBR and UCP2 was less than 0.10 whereas r2 was 0.24 between gene LOC780933 (SNP CT2) and gene PRSS2 (SNP TRYP81). In addition, the extent of LD was high, ranging from 0.26 to 0.97, between the SNP pairs in ATP6V1B2. The extent of LD between SNP pairs is presented graphically in Fig. 1.


View Full Table | Close Full ViewTable 6.

The extent of linkage disequilibrium (r2) between pairs of SNP within the same chromosome (Chr) in the discovery population

 
BTA1 SNP12 SNP22 r2 BTA SNP1 SNP2 r2
15 rs42670353 rs42670352 0.048 8 rs43563470 rs43562811 0.294
15 rs42670353 rs42670351 0.307 8 rs43563470 rs43562810 0.662
15 rs42670353 rs41774217 0.009 8 rs43563470 rs43562809 0.965
15 rs42670353 UCP24 0.006 8 rs43562811 rs43562810 0.215
15 rs42670353 UCP23 0.065 8 rs43562811 rs43562809 0.304
15 rs42670353 UCP22 0.064 8 rs43562810 rs43562809 0.706
15 rs42670353 rs41255549 0.014 13 rs41257167 PTI3 0.129
15 rs42670352 rs42670351 0.145 13 rs41257167 PTI2 0.116
15 rs42670352 rs41774217 0.004 13 rs41257167 rs43024409 0.157
15 rs42670352 UCP24 0.002 13 PTI3 PTI2 0.191
15 rs42670352 UCP23 0 13 PTI3 rs43024409 0.004
15 rs42670352 UCP22 0 13 PTI2 rs43024409 0.087
15 rs42670352 rs41255549 0.003 4 CT2 CT5 0.761
15 rs42670351 rs41774217 0.034 4 CT2 TRYP81 0.243
15 rs42670351 UCP24 0 4 CT2 rs41256901 0
15 rs42670351 UCP23 0.032 4 CT5 TRYP81 0.185
15 rs42670351 UCP22 0.032 4 CT5 rs41256901 0
15 rs42670351 rs41255549 0.094 4 TRYP81 rs41256901 0.055
15 rs41774217 UCP24 0.001 29 rs42197376 rs42197375 0.186
15 rs41774217 UCP23 0.004 29 rs42197376 rs42195008 0.982
15 rs41774217 UCP22 0.004 29 rs42197376 rs42194937 0.964
15 rs41774217 rs41255549 0.002 29 rs42197376 rs42194938 0.077
15 UCP24 UCP23 0.182 29 rs42197375 rs42195008 0.182
15 UCP24 UCP22 0.183 29 rs42197375 rs42194937 0.179
15 UCP24 rs41255549 0.021 29 rs42197375 rs42194938 0.003
15 UCP23 UCP22 0.991 29 rs42195008 rs42194937 0.982
15 UCP23 rs41255549 0.115 29 rs42195008 rs42194938 0.079
15 UCP22 rs41255549 0.117 29 rs42194937 rs42194938 0.078
1BTA = Bos taurus autosome.
2SNP1 = single nucleotide polymorphism in locus 1; SNP2 = single nucleotide polymorphism in locus 2.
Figure 1.
Figure 1.

The extent of linkage disequilibrium (r2) between pairs of SNP and haplotypes block structure in the candidate genes using Haploview. Cationic trypsin (LOC780933), protease serine 2 (trypsin 2; PRSS2), pancreatic trypsin inhibitor (PTI), cholecystokinin B receptor (CCKBR), uncoupling protein 2 (UCP2), pyruvate carboxylase (PC), and ATP6V1B2 genes using Haploview. Linkage disequilibrium between each SNP pair is illustrated in a square where the number on the square represents the r2 value between the 2 SNP corresponding to the cell. The empty square refers to r2 = 1. Thick lines (black triangles) specify haplotype blocks where the size of the block is written in parentheses. See online version for figure in color.

 

Association Analysis

Protease Serine 2.

In gene PRSS2, substitution with the T allele of SNP rs41256901 was associated with a decrease of 0.184 kg in DMI (Table 7; P = 0.084), a decrease of 0.298 kg DMI/kg gain in FCR (Table 7; P < 0.001), and a decrease of 0.199 kg in RFI (Table 7; P < 0.01). The SNP rs41256901 was significantly associated with FCR where genotype CC had greater FCR (5.1%) than CT genotype (P < 0.001).


View Full Table | Close Full ViewTable 7.

Estimates of allele substitution effects and genotypic effects (least squares means) of SNP in gene protease serine 2 (trypsin 2; PRSS2) in the discovery population

 
Genotype as fixed effect
Allele substitution effect
LSM3 ± SE LSM ± SE LSM ± SE
Trait1 SNP name P-value RA2 Estimate ± SE P-value C/C (G/G) C/T (A/G) T/T (A/A)
ADG rs41256901 0.128 T 0.038 ± 0.02 0.128 1.659 ± 0.048 1.697 ± 0.048
TRYP81 0.595 A 0.01 ± 0.02 0.737 1.668 ± 0.048 1.668 ± 0.048 1.698 ± 0.048
DMI rs41256901 0.084 T –0.184 ± 0.11 0.084 9.596 ± 0.20 9.412 ± 0.20
TRYP81 0.189 A 0.101 ± 0.08 0.295 9.464 ± 0.20 9.621 ± 0.20 9.597 ± 0.20
FCR rs41256901 <0.001** T –0.293 ± 0.08 <0.001** 6.201 ± 0.16 5.909 ± 0.16
TRYP81 0.332 A 0.059 ± 0.06 0.421 6.104 ± 0.149 6.21 ± 0.149 6.163 ± 0.149
MMWT rs41256901 0.627 T –0.331 ± 0.69 0.627 102.4 ± 1.331 102.069 ± 1.33
TRYP81 0.279 A 0.549 ± 0.5 0.058 101.599 ± 1.25 103.071 ± 1.25 101.536 ± 1.25
RFI rs41256901 0.010* T –0.199 ± 0.08 0.010* –0.141 ± 0.146 –0.34 ± 0.146
TRYP81 0.478 A 0.039 ± 0.06 0.766 –0.216 ± 0.147 –0.186 ± 0.147 –0.126 ± 0.147
Tended to affect the trait before the modified Bonferroni adjustment for multiple testing (P < 0.10).
Significant after the modified Bonferroni adjustment for multiple testing (P < 0.05).
*Significant effect before the modified Bonferroni adjustment for multiple testing (P < 0.05).
**Significant effect before the modified Bonferroni adjustment for multiple testing (P < 0.01).
1ADG = average daily gain (kg d–1); DMI = daily dry matter intake (kg d–1); FCR = feed conversion ratio (kg gain kg–1 DM); MMWT = midpoint metabolic weight (kg); RFI = residual feed intake (kg d–1).
2RA = substitution allele.
3LSM = least squares mean.

Cholecystokinin B Receptor.

Genotypes in SNP rs42670351 were significantly associated with RFI (Table 8; P < 0.05). Substitution to the G allele of SNP rs42670352 tended to be associated with a 0.236 kg decrease in DMI (Table 8; P = 0.055) and a decrease of 0.175 kg in RFI (Table 8; P = 0.053). Genotypes in rs42670352 were significantly associated with DMI (Table 8; P = 0.033) and RFI (Table 8; P = 0.002). Substitution to the G allele of SNP rs42670353 was significantly associated with a 0.043 kg increase in ADG (Table 8; P = 0.006) and a 0.114 kg gain kg–1 DMI decrease in FCR (Table 8; P = 0.033).


View Full Table | Close Full ViewTable 8.

Estimates of allele substitution effects and genotypic effects (least squares means) of SNP in gene cholecystokinin B receptor (CCKBR) in the discovery population

 
Genotype as fixed effect
Allele substitution effect
LSM3 ± SE LSM ± SE LSM ± SE
Trait1 SNP name P-value RA2 Estimate ± SE P-value C/C (G/G) C/T (A/G) T/T (A/A)
ADG rs42670353 0.006** T –0.043 ± 0.02 0.019* 1.73 ± 0.05 1.67 ± 0.05 1.64 ± 0.05
rs42670351 0.135 C 0.03 ± 0.02 0.112 1.66 ± 0.05 1.70 ± 0.05 1.65 ± 0.05
rs42670352 0.427 G –0.024 ± 0.03 0.620 1.63 ± 0.05 1.56 ± 0.05 1.67 ± 0.05
DMI rs42670352 0.055 G –0.236 ± 0.12 0.033* 9.14 ± 0.20 8.28 ± 0.20 9.52 ± 0.20
rs42670351 0.467 C –0.06 ± 0.08 0.069 9.08 ± 0.20 10.00 ± 0.20 9.53 ± 0.20
rs42670353 0.563 T –0.038 ± 0.07 0.828 9.61 ± 0.20 9.55 ± 0.20 9.53 ± 0.20
FCR rs42670353 0.033* T 0.114 ± 0.05 0.055 5.92 ± 0.16 6.12 ± 0.16 6.16 ± 0.16
rs42670351 0.194 C –0.087 ± 0.07 0.405 6.00 ± 0.15 6.14 ± 0.151 6.21 ± 0.15
rs42670352 0.673 G –0.041 ± 0.1 0.727 6.12 ± 0.17 5.82 ± 0.17 6.17 ± 0.17
MMWT rs42670353 0.229 T –0.523 ± 0.43 0.306 103.44 ± 1.33 102.28 ± 1.33 102.3 ± 1.33
rs42670351 0.391 C –0.467 ± 0.54 0.689 101.42 ± 1.24 102.01 ± 1.24 102.45 ± 1.24
rs42670352 0.490 G –0.55 ± 0.8 0.396 100.33 ± 1.36 105.66 ± 1.36 101.82 ± 1.36
RFI rs42670352 0.053 G –0.175 ± 0.09 0.002 –0.42 ± 0.15 –1.61 ± 0.15 –0.18 ± 0.15
rs42670353 0.280 T 0.052 ± 0.05 0.434 –0.30 ± 0.15 –0.19 ± 0.15 –0.19 ± 0.15
rs42670351 0.355 C –0.056 ± 0.06 0.028* –0.60 ± 0.15 –0.16 ± 0.15 –0.21 ± 0.15
Tended to affect the trait before the modified Bonferroni adjustment for multiple testing (P < 0.10)
Significant after the modified Bonferroni adjustment for multiple testing (P < 0.05)
*Significant effect before the modified Bonferroni adjustment for multiple testing (P < 0.05)
**Significant effect before the modified Bonferroni adjustment for multiple testing (P < 0.01)
1ADG = average daily gain (kg d–1); DMI = daily dry matter intake (kg d–1); FCR = feed conversion ratio (kg gain kg–1 DM); MMWT = midpoint metabolic weight (kg); RFI = residual feed intake (kg d–1).
2RA = substitution allele.
3LSM = least squares mean.

Uncoupling Protein 2.

Substitution to the T allele of SNP UCP24 was slightly associated with a 0.076 kg decrease in ADG (Table 9; P = 0.054). The SNP UCP24 did not show a significant relationship with feed efficiency.


View Full Table | Close Full ViewTable 9.

Estimates of allele substitution effects on feed efficiency traits in the validation study

 
BTA: gene ID1 Trait2 SNP name Ref_SNP n3 MA4 MAF5 Estimate ± SE P-value6
8: 338082 MMWT ATPase4 rs43562809 726 A 0.073 1.292 ± 0.76 0.0897 d
8: 338082 FCR ATPase2 rs43562811 726 C 0.252 –0.139 ± 0.06 0.0218 d
8: 338082 ADG ATPase2 rs43562811 726 C 0.252 0.037 ± 0.02 0.0277
8: 338082 MMWT ATPase2 rs43562811 726 C 0.252 0.941 ± 0.45 0.0378
13: 404172 MMWT PTI1 rs43024409 698 T 0.352 0.973 ± 0.43 0.0234
13: 404172 ADG PTI1 rs43024409 698 T 0.352 0.026 ± 0.02 0.0970
15: 281665 DMI CCRB3 rs42670353 725 G 0.420 –0.251 ± 0.07 0.0008 d
15: 281665 RFI CCRB3 rs42670353 725 G 0.420 –0.159 ± 0.06 0.0135
15: 281665 FCR CCRB3 rs42670353 725 G 0.420 –0.125 ± 0.05 0.0168 sd*
15: 281665 DMI CCRB2 rs42670352 725 G 0.217 –0.222 ± 0.09 0.0116 sd*
15: 281665 RFI CCRB2 rs42670352 725 G 0.217 –0.139 ± 0.08 0.0658 sd*
15: 281665 FCR CCRB2 rs42670352 725 G 0.217 –0.099 ± 0.06 0.106 d
15: 281665 DMI CCRB1 rs42670351 700 C 0.221 –0.235 ± 0.09 0.0084 d
15: 281665 RFI CCRB1 rs42670351 700 C 0.221 –0.164 ± 0.08 0.0315 d*
15: 281665 FCR CCRB1 rs42670351 700 C 0.221 –0.117 ± 0.06 0.0589 d
Significant after the modified Bonferroni adjustment for multiple testing (P < 0.05) in the validation population.
1BTA = Bos taurus autosome; gene ID = Entrez Gene Identifier.
2ADG = average daily gain (kg d–1); DMI = daily dry matter intake (kg d–1); FCR = feed conversion ratio (kg gain kg–1 DM); MMWT = midpoint metabolic weight (kg); RFI = residual feed intake (kg d–1).
3n = the number of records used in the analyses.
4MA = minor allele.
5MAF = minor allele frequency.
6d = The same direction but was not significant in the discovery population; d = the same direction but showed a trend (P = 0.069) in the discovery population using the genotypic model; sd* = the same direction and significance was found for both the discovery and validation populations; d* = the same direction and significance was found for both the discovery and validation populations using the genotypic model.

Validation Study

The SNP within PRSS2 were not associated with feed efficiency and performance traits. Four SNP in CCKBR were evaluated in the validation data set. The C allele of SNP CCKBR1 (rs42670351) decreased DMI by 0.235 kg (Table 9; P = 0.00084), decreased RFI by 0.164 kg (Table 9; P = 0.0315), and decreased DMI by 0.117 (kg BW gain kg–1 DMI) in FCR (Table 9; P = 0.059). In addition, genotypes in rs42670351 were associated with DMI, RFI, and FCR (Table 10; P = 0.009, P = 0.014, and P = 0.085, respectively).


View Full Table | Close Full ViewTable 10.

The genotypic analysis for SNP affecting feed efficiency traits in the validation population

 
LSM7 ± SE LSM ± SE LSM ± SE
BTA1 Gene ID2 SNP name Ref. SNP3 MA4 Trait5 P-value MAF6 C/C (GG) C/T (A/G) T/T (A/A)
8 338082 ATPase2 rs43562811 C ADG 0.058 0.255 1.76 ± 0.03 1.70 ± 0.03 1.67 ± 0.03
8 338082 ATPase2 rs43562811 C FCR 0.067 0.255 6.59 ± 0.12 6.75 ± 0.12 6.89 ± 0.12
8 338082 ATPase2 rs43562811 C MMWT 0.037 0.255 106.33 ± 0.87 104.09 ± 0.87 103.53 ± 0.87
13 404172 PTI1 rs43024409 T MMWT 0.082 0.36 105.03 ± 0.88 104.41 ± 0.88 103.31 ± 0.88
15 281665 CCKBR1 rs42670351 C DMI 0.009 0.216 10.22 ± 0.16 10.69 ± 0.16 10.87 ± 0.16
15 281665 CCKBR1 rs42670351 C FCR 0.085 0.216 6.58 ± 0.12 6.66 ± 0.12 6.82 ± 0.12
15 281665 CCKBR1 rs42670351 C RFI 0.014 0.216 –0.18 ± 0.13 0.04 ± 0.13 0.25 ± 0.13
15 281665 CCKBR2 rs42670352 G DMI 0.011 0.213 10.22 ± 0.16 10.689 ± 0.16 10.85 ± 0.16
15 281665 CCKBR2 rs42670352 G FCR 0.107 0.213 6.63 ± 0.12 6.72 ± 0.12 6.86 ± 0.12
15 281665 CCKBR2 rs42670352 G RFI 0.024 0.213 –0.16 ± 0.13 0.085 ± 0.13 0.26 ± 0.13
15 281665 CCKBR3 rs42670353 G DMI 0.002 0.423 10.48 ± 0.16 10.70 ± 0.16 10.98 ± 0.16
15 281665 CCKBR3 rs42670353 G FCR 0.048 0.423 6.68 ± 0.1178 6.80 ± 0.1178 6.94 ± 0.1178
15 281665 CCKBR3 rs42670353 G RFI 0.005 0.423 0.0003 ± 0.13 0.13 ± 0.13 0.38 ± 0.13
29 338471 PC5 rs42197375 C MMWT 0.062 0.119 108.98 ± 0.87 103.986 ± 0.87 104.56 ± 0.87
1BTA = Bos taurus autosome.
2Gene ID = Entrez Gene Identifier.
3Ref. SNP = SNP reference number at the National Center for Biotechnology Information.
4MA = minor allele.
5ADG = average daily gain (kg d–1); DMI = daily dry matter intake (kg d–1); FCR = feed conversion ratio (kg gain kg–1 DM); MMWT = midpoint metabolic weight (kg); RFI = residual feed intake (kg d–1)
6MAF = minor allele frequency.
7LSM = least squares mean.

The G allele of rs42670352 was associated with a 0.222 kg decrease in DMI (Table 9; P = 0.0116), a 0.139 kg decrease in RFI (Table 9; P = 0.066), and a 0.099 kg BW gain kg–1 DMI decrease in FCR (Table 9; P = 0.106). Genotypes in rs42670352 were significantly associated with DMI and RFI (Table 10; P < 0.05).

The G allele of SNP CCKBR3 (rs42670353) was associated with a 0.251 kg decrease in DMI (Table 9; P = 0.0008), a 0.159 kg decrease in RFI (Table 9; P = 0.0135), and a 0.125 kg gain kg–1 DMI decrease in FCR (Table 9; P = 0.0168). Genotypes in rs42670353 were significantly associated with DMI, RFI, and FCR (Table 10; P = 0.002, P = 0.005, P = 0.048, respectively).

In gene PTI, substitution to the T allele of SNP PTI1 (rs43024409) was associated with a 0.973 kg75 increase in MMWT (Table 9; P = 0.023). The C allele of SNP ATPase2 (rs43562811) was associated with a 0.037 kg increase in ADG (Table 9; P = 0.028), a 0.941 kg75 increase in MMWT (Table 9; P = 0.038), and a 0.139 BW kg gain kg–1 DMI decrease in FCR (Table 9; P = 0.022). Genotypes in SNP ATPase2 (rs43562811) were significantly associated with MMWT (Table 10; P = 0.037).


DISCUSSION

Three SNP, AMY2B6 (rs42312301), TRYP83 (rs41256901), and UCP21 (rs41255549), were present in the GenBank. The GenBank had information for only SNP TRYP83 and UCP21. The remaining 22 SNP reported in the GenBank were identified using the Bos taurus Assembly SNP Discovery method where little information were available for these SNP. The SNP PTI8 (rs43024345) was reported in the GenBank and was not segregating in the current genotyped population. It is common that SNP in a database such as GenBank can be segregating in 1 particular population but not in another (Kitts and Sherry, 2007). The percentage of segregating SNP in the current population was 44.4% (8/18) based on SNP resulting from Sequencher with a minimum match percentage of 85% and minimum overlap of 20 bases. This proportion of segregated SNP was in agreement with Weckx et al. (2005) who estimated the false positives percentage (i.e., the percentage of SNP found fixed after genotyping) using different sequence-variation programs (PolyPhred (Nickerson et al., 1997), PolyBayes (Marth et al., 1999), and novoSNP (Weckx et al., 2005)). Weckx et al. (2005) reported that the percentages of false positives were 15.4, 51.5, and 86.2% for PolyPhred, PolyBayes, and novoSNP, respectively, at the greatest level of quality cutoff. However, they found the false positive rates at the lowest level of quality cutoff were greater at 94.9, 66.7, and 92.6% for PolyPhred, PolyBayes, and novoSNP, respectively. The results from using Sequencher, PolyPhred, PolyBayes, and novoSNP programs during SNP discovery indicated there was a high rate of false positives due to the direct relationship between the false positive rate and the quality of the sequence traces, particularly the background noise (Picoult-Newberg et al., 1999; Cox et al., 2001; Weckx et al., 2004). Nonetheless, the in silico approach provides cost-effective SNP detection in spite of the high rate of false positives, particularly with the advent of overwhelming results (millions of EST or reads) obtained from next generation sequencing stored in the public domain at Sequence Read Archive (at http://www.ncbi.nlm.nih.gov/Traces/sra/ from NCBI, http://www.ebi.ac.uk/ena/, or http://trace.ddbj.nig.ac.jp/dra/index_e.shtml; Shumway et al., 2010; Leinonen et al., 2011). The in silico approach provides a lower cost option with fewer lab resources required compared with direct sequencing of a particular gene from pooled DNA samples for SNP discovery. In addition, the probability of discovering SNP using the in silico approach may be greater than direct sequencing as a result of accumulation of new sequences over time in the public domain as well as these sequences might be from different populations increasing the possibility of finding new SNP.

Markers deviating from HWE indicate problems with genotyping or population stratification. Because of the missing class of genotypes within some SNP, the association analysis results must be viewed with caution. Nonetheless, some deviation from HWE indicates a potential association between a particular marker and the trait of interest (Wigginton et al., 2005). The functional mutation might have a rare allele that can be missing in some breeds or in populations within a breed (Goddard, 2009). Furthermore, the genetic markers that are linked to the QTL with large effects within a particular breed contribute to the composite or cross (Piyasatian et al., 2006). Bansal et al. (2010) discussed many reasons for considering rare variants as a source of variation. Therefore, in the current study, associations were tested for rare variants or for SNP that are not in HWE (minor allele frequency of less than 10%) as these might be informative or provide promising results that could be considered in crossbred or multibreed populations. However, the obtained significant associations should be validated in other populations.

The extent of LD in the current study was measured using r2 as it is less dependent on allele frequencies or affected by small sample size (Ardlie et al., 2002; McRae et al., 2002; Khatkar et al., 2008). Generally, the magnitude of LD between SNP pairs was greater in some genes than in others as it is due to SNP density (i.e., the relationship between distance and r2 was high; Khatkar et al., 2008). The magnitude of LD present between SNP pairs in genes CCKBR and UCP2 was less than 0.1, which is expected as they were up to 7.1 Mbp apart (Sargolzaei et al., 2008). The LD between SNP CT2 in gene LOC780933 and SNP TRYP81 in gene PRSS2 was 0.243 as the distance between these 2 SNP is 0.75 Mbp, indicating they may capture some of the same effects. The SNP ATPase4 (rs43562809) is sufficient to capture 100% of the genetic variation explained by SNP ATPase1 in gene ATP6V1B2 using haplotype tagging from the Haploview analysis. In addition, SNP PC2 can capture 100% of the variance explained by SNP PC3 and PC6. Consequently, not all SNP would be selected for genotyping in the validation population to reduce costs.

Gene PRSS2 (SNP rs41256901) was significantly associated with RFI and FCR and suggestively associated with DMI. The identified significant associations were not found in the validation study. These associations are in agreement with the significant relationship between feed efficiency and digestive function reported by Richardson and Herd (2004) where the digestibility of feed accounted for 10 to 14% of the variation in feed efficiency. Pancreatic enzymes may be partially responsible for the variation in digestive efficiency between animals (Swanson et al., 2004). Conversely, there was no significant relationship between either performance or feed efficiency and the concentration of the pancreatic trypsin enzyme in feedlot cattle (Mader et al., 2009). Trypsinogen can be an activator of proteinase-activated receptor 2 (PAR-2), which is highly expressed in digestive organs, such as the pancreas and intestine, and stimulates many biological processes, such as cell proliferation (Ossovskaya and Bunnett, 2004). In mice, downregulation of trypsinogen was associated with growth retardation in α1, 6-fucosyltransferase-knockout mice (Li et al., 2006).

Results from the current association analysis indicate that there were significant associations between gene CCKBR, represented in SNP rs42670351 and rs42670352, and RFI, DMI, ADG, and FCR. Substitution to the C allele in SNP rs42670351 was associated with decreasing DMI, RFI, and FCR in the validation population. Also, the G allele of SNP CCKBR2 (rs42670352) was validated to be associated with decreasing DMI and RFI. In addition, substitution to the G allele of SNP CCKBR3 (rs42670353) was validated to be associated with decreasing FCR and found to be significantly associated with decreasing DMI and RFI. The identified significant SNP in gene CCKBR were synonymous, so they might be in LD with the functional mutation. Recently, a missense mutation (rs133526822) was identified using a whole-genome sequencing method (Kawahara-Miki et al., 2011) where SNP rs133526822 is located between SNP rs42670351 and rs42670352. Therefore, further investigation of SNP rs133526822 is required as it might be the functional mutation responsible for the reported significant association. In pigs, Houston et al. (2006, 2008) reported an association between polymorphisms in the 5′-untranslated region of the porcine CCK type A receptor gene with feed intake and growth. Gene CCKBR is expressed in gastric parietal cells, the brain, and smooth muscle (Huppi et al., 1995; Wank, 1995). The significant associations with feed efficiency and performance found in the current study are consistent with the functions of gene CCKBR. Gene CCKBR is predominant in the hypothalamus and is also expressed in the vagus nerve stem complex, so it plays a very important role as a mediator in the satiety effect of CCK (Dufresne et al., 2006), affecting feed intake and efficiency. Gene CCKBR is the predominant CCK receptor subtype for the veal and weaned calves (Le Meuth et al., 1993). As in calves, CCKBR are predominant in the pancreas of pigs (Philippe et al., 1997). Pancreatic enzyme secretion was mediated by CCKBR under stimulation by the physiological levels of CCK and gastrin (Le Dréan et al., 1999). Also, pancreatic growth and secretion were regulated by CCKBR particularly after weaning (Le Meuth et al., 1993). Therefore, association between polymorphisms in gene CCKBR might be associated with pancreas growth or secretion, suggesting further study to test these biological relationships.


Conclusion

The in silico study was an effective method for SNP discovery in candidate genes. New SNP were reported in genes PRSS2 and CCKBR that have an association with feed efficiency and performance traits in these data. The SNP rs42670352 in CCKBR was significantly associated with RFI and DMI in the discovery and validation populations and had the same phase of associations. In addition, SNP rs42670353 in CCKBR was significantly associated with FCR in the discovery population with same phase of association in the validation populations. Investigating the biological mechanisms underpinning these discoveries by studying gene expression (RNA and protein abundance) will also increase our understanding of the underlying biology of these SNP.

APPENDIX


View Full Table | Close Full ViewAppendix 1.

Biological mechanisms, molecular function, and pathways associated with the candidate genes in the study

 
Category1 Gene ontology term P-value2 Genes
BP GO:0007586∼digestion 0.013025 282603, 780933
PATH bta04080:Neuroactive ligand-receptor interaction 0.021155 282603, 780933, 281665
MF GO:0004252∼serine-type endopeptidase activity 0.062294 282603, 780933
MF GO:0008236∼serine-type peptidase activity 0.070418 282603, 780933
MF GO:0017171∼serine hydrolase activity 0.070923 282603, 780933
MF GO:0004175∼endopeptidase activity 0.156439 282603, 780933
MF GO:0070011∼peptidase activity, acting on L-amino acid peptides 0.201141 282603, 780933
BP GO:0055085∼transmembrane transport 0.202684 281562, 338082
MF GO:0008233∼peptidase activity 0.207363 282603, 780933
MF GO:0016787∼hydrolase activity 0.212711 282603, 780933, 338082
MF GO:0005509∼calcium ion binding 0.252329 282603, 780933
BP GO:0006508∼proteolysis 0.304393 282603, 780933
MF GO:0003824∼catalytic activity 0.343324 282603, 780933, 338082, 338471
MF GO:0046872∼metal ion binding 0.378607 282603, 780933, 338471
MF GO:0043169∼cation binding 0.384751 282603, 780933, 338471
MF GO:0043167∼ion binding 0.390012 282603, 780933, 338471
MF GO:0005515∼protein binding 0.704959 282603, 780933, 281665
MF GO:0005488∼binding 0.773127 282603, 780933, 281562, 281665, 338471
MF GO:0004857∼enzyme inhibitor activity 1 404172
MF GO:0004866∼endopeptidase inhibitor activity 1 404172
MF GO:0004867∼serine-type endopeptidase inhibitor activity 1 404172
BP GO:0005996∼monosaccharide metabolic process 1 338471
BP GO:0006006∼glucose metabolic process 1 338471
BP GO:0006091∼generation of precursor metabolites and energy 1 338082
BP GO:0006094∼gluconeogenesis 1 338471
BP GO:0006119∼oxidative phosphorylation 1 338082
BP GO:0006163∼purine nucleotide metabolic process 1 338082
BP GO:0006754∼ATP biosynthetic process 1 338082
BP GO:0006793∼phosphorus metabolic process 1 338082
BP GO:0006811∼ion transport 1 338082
BP GO:0006812∼cation transport 1 338082
BP GO:0006818∼hydrogen transport 1 338082
BP GO:0006839∼mitochondrial transport 1 281562
BP GO:0006874∼cellular calcium ion homeostasis 1 281665
BP GO:0007166∼cell surface receptor linked signal transduction 1 281665
BP GO:0007186∼G-protein coupled receptor protein signaling pathway 1 281665
BP GO:0008284∼positive regulation of cell proliferation 1 281665
BP GO:0008610∼lipid biosynthetic process 1 338471
BP GO:0009057∼macromolecule catabolic process 1 282603
BP GO:0009141∼nucleoside triphosphate metabolic process 1 338082
BP GO:0009205∼purine ribonucleoside triphosphate metabolic process 1 338082
BP GO:0015985∼energy coupled proton transport, down electrochemical gradient 1 338082
BP GO:0015986∼ATP synthesis coupled proton transport 1 338082
BP GO:0015992∼proton transport 1 338082
BP GO:0016051∼carbohydrate biosynthetic process 1 338471
BP GO:0016310∼phosphorylation 1 338082
BP GO:0019318∼hexose metabolic process 1 338471
BP GO:0030003∼cellular cation homeostasis 1 281665
BP GO:0030163∼protein catabolic process 1 282603
MF GO:0030234∼enzyme regulator activity 1 404172
BP GO:0032963∼collagen metabolic process 1 282603
BP GO:0034220∼ion transmembrane transport 1 338082
BP GO:0034404∼nucleobase, nucleoside and nucleotide biosynthetic process 1 338082
BP GO:0034637∼cellular carbohydrate biosynthetic process 1 338471
BP GO:0034654∼nucleobase, nucleoside, nucleotide and nucleic acid biosynthetic process 1 338082
BP GO:0042127∼regulation of cell proliferation 1 281665
BP GO:0044257∼cellular protein catabolic process 1 282603
BP GO:0044259∼multicellular organismal macromolecule metabolic process 1 282603
BP GO:0044271∼nitrogen compound biosynthetic process 1 338082
BP GO:0046034∼ATP metabolic process 1 338082
BP GO:0046364∼monosaccharide biosynthetic process 1 338471
BP GO:0046907∼intracellular transport 1 281562
BP GO:0050801∼ion homeostasis 1 281665
BP GO:0051480∼cytosolic calcium ion homeostasis 1 281665
BP GO:0051603∼proteolysis involved in cellular protein catabolic process 1 282603
BP GO:0055065∼metal ion homeostasis 1 281665
BP GO:0055074∼calcium ion homeostasis 1 281665
PATH bta00620:Pyruvate metabolism 1 338471
PATH bta00500:Starch and sucrose metabolism 1 539383
PATH bta00190:Oxidative phosphorylation 1 338082
PATH bta00020:Citrate cycle (TCA cycle) 1 338471
PATH bta04020:Calcium signaling pathway 1 281665
1BP = biological process; PATH = KEGG biological pathway; MF = molecular function.
2P-value = the P-value produced by enrichment analysis using DAVID software (Huang et al., 2009).

View Full Table | Close Full ViewAppendix 2.

Reported QTL in the Animal QTL database overlapping with candidate genes

 
Gene name QTL name1 QTL start position (bp) QTL end position Trait
AMY2B QTL_10129 34002036 65739450 Milk fat yield (daughter deviation)
AMY2B QTL_10683 29956341 43937012 Height (mature)
AMY2B QTL_10684 23481347 43937012 BW (birth)
AMY2B QTL_10685 23481347 43937012 BW (weaning)
AMY2B QTL_10686 29956341 43937012 Height (yearling)
AMY2B QTL_10687 23481347 43937012 Carcass weight
AMY2B QTL_1326 17737279 42568866 BW (birth)
AMY2B QTL_1351 37151994 48988111 Marbling score
AMY2B QTL_2437 17033086 46894623 Milk protein yield
AMY2B QTL_2442 20288493 43373658 Milk fat percentage
AMY2B QTL_2443 18441472 63526272 Milk protein percentage
AMY2B QTL_2490 0 43937012 Somatic Cell Count
AMY2B QTL_2541 28000529 90939023 Marbling score
AMY2B QTL_2584 22194674 57994437 Milk yield
AMY2B QTL_2653 29300268 47891582 Milk yield
AMY2B QTL_2654 29300268 47891582 Milk protein percentage
AMY2B QTL_2657 29300268 48706434 Milk protein yield
ATP6V1B2 QTL_10831 62732685 70703039 Calving ease (direLOC780933)
ATP6V1B2 QTL_10832 62732685 83158666 BW (birth)
ATP6V1B2 QTL_11442 46113400 92708598 Dystocia (maternal)
ATP6V1B2 QTL_11443 46113400 92708598 Stillbirth (maternal)
ATP6V1B2 QTL_1683 62735433 83360187 Somatic Cell Count
ATP6V1B2 QTL_1684 62735433 83360187 StruLOC780933ural soundness (legs
ATP6V1B2 QTL_2497 43089984 83569900 Clinical mastitis
ATP6V1B2 QTL_2498 43089984 83569900 Somatic Cell Count
ATP6V1B2 QTL_3599 59682392 103895369 Foot angle
CCKBR QTL_10993 32063789 47949409 Height (mature)
CCKBR QTL_10994 32063789 47949409 Carcass weight
CCKBR QTL_10995 32063789 47949409 Longissimus muscle area
CCKBR QTL_12195 32881427 55518119 Liver percentage
CCKBR QTL_1335 21718187 54967641 Kidney
CCKBR QTL_1596 40000451 47942043 Udder attachment
CCKBR QTL_1598 40000451 47942043 Stature
CCKBR QTL_1601 40000451 47942043 Udder depth
CCKBR QTL_1699 15363886 47946463 Rump angle
CCKBR QTL_2678 40000451 47942043 Somatic cell score
LOC780933 QTL_10515 98051474 119280357 Parasites mean of natural logarithm
LOC780933 QTL_10716 95179745 107915671 BW (mature)
LOC780933 QTL_10717 81300315 107915671 Fat thickness at the 12th rib
LOC780933 QTL_10718 95179745 113949759 Scrotal circumference
LOC780933 QTL_4485 98051474 107905229 Postweaning average daily gain
LOC780933 QTL_5055 57599923 119280357 Milk fat percentage
PC QTL_11297 37089419 48660164 Carcass weight
PC QTL_11298 37089419 48660164 BW (yearling)
PC QTL_11299 37089419 51080436 BW (birth)
PC QTL_11301 37089419 48660164 BW (weaning)
PC QTL_1343 34055421 46031383 Retail produLOC780933 yield
PC QTL_1345 27433819 48199160 Tenderness score
PC QTL_1373 36423729 46550490 Tenderness score
PC QTL_1374 37521610 51800755 Tenderness score
PC QTL_1380 32188341 45594750 BW at castration
PC QTL_1664 35467866 46211463 Foot angle
PC QTL_1665 35467866 46211463 StruLOC780933ural soundness (legs
PC QTL_1717 26859725 48660164 Teat placement
PC QTL_1722 26859725 48660164 Twinning
PC QTL_2593 18168758 46211940 Milk yield
PC QTL_2612 18168758 46211940 Milk protein yield
PC QTL_4506 37089419 48660164 305 d milk yield
PC QTL_4651 35465705 46211940 Rump angle
PC QTL_4851 34170672 51606269 Juiciness
PC QTL_4852 34170672 51606269 Shear force
PC QTL_4853 34170672 51606269 Tenderness score
PC QTL_5371 1983481 45887319 Gestation length
PC QTL_7153 34170672 51606269 Flight from feeder
PC QTL_7154 34170672 51606269 Flight fm feeder
PC QTL_7154 34170672 51606269 Flight from feeder
PTI QTL_10946 50098085 77464923 Weaning weight-maternal milk
PTI QTL_10947 60604089 77464923 Marbling score (EBV)
PTI QTL_10948 60604089 77464923 Longissimus muscle area
PTI QTL_10949 71832245 77464923 BW (weaning)
PTI QTL_11446 69174400 81815003 Stillbirth (direLOC780933)
PTI QTL_1386 15506564 84433115 Teat length
PTI QTL_1584 59740226 79842537 Udder attachment
PTI QTL_1585 59740226 79842537 PTA type
PTI QTL_1586 59740226 79842537 Udder height
PTI QTL_1587 59740226 79842537 Udder width
PTI QTL_1588 59740226 79842537 Udder depth
PTI QTL_1589 59740226 79842537 Udder composite index
PTI QTL_2670 59740226 79842537 Milk yield
PTI QTL_2671 59740226 79842537 Milk protein yield
PTI QTL_2775 69333703 77472469 Somatic cell score
PTI QTL_3569 0 77464923 Heat intensity
PTI QTL_5011 71832245 77464923 Interval to first estrus after calving
PRSS2 QTL_10515 98051474 119280357 Parasites mean of natural logarithm
PRSS2 QTL_10716 95179745 107915671 BW (mature)
PRSS2 QTL_10717 81300315 107915671 Fat thickness at the 12th rib
PRSS2 QTL_10718 95179745 113949759 Scrotal circumference
PRSS2 QTL_4485 98051474 107905229 Postweaning average daily gain
PRSS2 QTL_5055 57599923 119280357 Milk fat percentage
PRSS2 QTL_10515 98051474 119280357 Parasites mean of natural logarithm
PRSS2 QTL_10716 95179745 107915671 BW (mature)
PRSS2 QTL_10717 81300315 107915671 Fat thickness at the 12th rib
PRSS2 QTL_10718 95179745 113949759 Scrotal circumference
PRSS2 QTL_4485 98051474 107905229 Postweaning average daily gain
PRSS2 QTL_5055 57599923 119280357 Milk fat percentage
UCP2 QTL_10996 47949409 60494505 Weaning weight-maternal milk
UCP2 QTL_10997 51606842 60494505 BW (mature)
UCP2 QTL_10998 47949409 60494505 BW (weaning)
UCP2 QTL_10999 51606842 60641830 Marbling score (EBV)
UCP2 QTL_11001 51606842 79392371 Height (mature)
UCP2 QTL_12195 32881427 55518119 Liver percentage
UCP2 QTL_1335 21718187 54967641 Kidney
UCP2 QTL_1594 47946463 57050270 Body form composite index
UCP2 QTL_1595 47946463 57050270 Teat placement
UCP2 QTL_1597 47946463 57050270 PTA type
UCP2 QTL_1599 47946463 57050270 Thurl width
UCP2 QTL_1600 47946463 57050270 Udder cleft
UCP2 QTL_1602 47946463 57050270 Udder composite index
UCP2 QTL_5122 50892329 55864326 Abomasum displacement
1QTL = Quantitative trait loci.
 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.