Advertisement

Major quantitative trait loci influencing milk production and conformation traits in Guernsey dairy cattle detected on Bos taurus autosome 19

Open ArchivePublished:November 11, 2020DOI:https://doi.org/10.3168/jds.2020-18766

      ABSTRACT

      The goal of this study was to identify potential quantitative trait loci (QTL) for 27 production, fitness, and conformation traits of Guernsey cattle through genome-wide association (GWA) analyses, with extra emphasis on BTA19, where major QTL were observed for several traits. Animals' de-regressed predicted transmitting abilities (PTA) from the December 2018 traditional US evaluation were used as phenotypes. All of the Guernsey cattle included in the QTL analyses were predictor animals in the reference population, ranging from 1,077 to 1,685 animals for different traits. Single-trait GWA analyses were carried out by a mixed-model approach for all 27 traits using imputed high-density genotypes. A major QTL was detected on BTA19, influencing several milk production traits, conformation traits, and livability of Guernsey cattle, and the most significant SNP lie in the region of 26.2 to 28.3 Mb. The myosin heavy chain 10 (MYH10) gene residing within this region was found to be highly associated with milk production and body conformation traits of dairy cattle. After the initial GWA analyses, which suggested that many significant SNP are in linkage with one another, conditional analyses were used for fine mapping. The top significant SNP on BTA19 were fixed as covariables in the model, one at a time, until no more significant SNP were detected on BTA19. After this fine-mapping approach was applied, only 1 significant SNP was detected on BTA19 for most traits, but multiple, independent significant SNP were found for protein yield, dairy form, and stature. In addition, the haplotype that hosts the major QTL on BTA19 was traced to a US Guernsey born in 1954. The haplotype is common in the breed, indicating a long-term influence of this QTL on the US Guernsey population.

      Key words

      INTRODUCTION

      Guernsey cattle, originally from the Isle of Guernsey in the English Channel off the coast of France, have been developed over 2 centuries and are renowned for a unique golden-colored milk high in protein and butterfat (
      • World Guernsey Cattle Ferderation
      Guernsey Breed History.
      ). In the United States, traditional genetic evaluations for Guernsey have been available for decades, and official genomic evaluations for Guernsey started in 2016 (
      • Cooper T.A.
      • Eaglen S.A.E.
      • Wiggans G.R.
      • Jenko J.
      • Huson H.J.
      • Morrice D.R.
      • Bichard M.
      • Luff W.G.L.
      • Woolliams J.A.
      Genomic evaluation, breed identification, and population structure of Guernsey cattle in North America, Great Britain, and the Isle of Guernsey.
      ). Genomic evaluations for Guernsey were initially based on 2,376 genotyped Guernsey bulls and cows from the United States, Canada, the United Kingdom, and the World Guernsey Cattle Federation (Isle of Guernsey;
      • Cooper T.A.
      • Eaglen S.A.E.
      • Wiggans G.R.
      • Jenko J.
      • Huson H.J.
      • Morrice D.R.
      • Bichard M.
      • Luff W.G.L.
      • Woolliams J.A.
      Genomic evaluation, breed identification, and population structure of Guernsey cattle in North America, Great Britain, and the Isle of Guernsey.
      ). A potential QTL on BTA19 influencing several traits [e.g., milk, productive life, SCS, daughter pregnancy rate (DPR), cow conception rate, size, rump, udder, and teat length] was briefly mentioned by
      • Cooper T.A.
      • Eaglen S.A.E.
      • Wiggans G.R.
      • Jenko J.
      • Huson H.J.
      • Morrice D.R.
      • Bichard M.
      • Luff W.G.L.
      • Woolliams J.A.
      Genomic evaluation, breed identification, and population structure of Guernsey cattle in North America, Great Britain, and the Isle of Guernsey.
      , using a 60,671 SNP panel. However, due to limitations of data size and marker density, the precise location of the QTL is unknown, and it is unclear whether multiple QTL on BTA19 underlie associated traits.
      Currently in the United States, the number of genotyped animals for Guernsey evaluation has almost doubled compared with the initial data. Prediction reliabilities for Guernsey evaluation have increased since 2016, and more traits have been analyzed (net merit, 5 yield traits, 7 functional traits, and 14 conformation traits). A new reference assembly of the bovine genome, ARS-UCD 1.2, was applied to official genomic evaluation, and the new map has improved performances in marker locations, sequence alignment, and genotype imputation compared with the previous UMD 3.1 reference assembly (
      • Null D.J.
      • VanRaden P.M.
      • Rosen B.D.
      • O'Connell J.R.
      • Bickhart D.M.
      Using the ARS-UCD1.2 reference genome in U. S. evaluations.
      ). Meanwhile, high-density (HD) genotypes became available for a larger number of Guernsey animals, making it possible to obtain imputed HD genotypes for all Guernsey in the national database.
      Given larger amounts of information available, the aim of this study was to locate potential QTL for 27 production, fitness, and conformation traits of Guernsey cattle through genome-wide association (GWA) analyses using HD genotypes. Extra emphasis was placed on BTA19, where a major QTL was observed for several traits. A conditional analysis was further carried out to test for multiple QTL on BTA19 for the traits showing major QTL on BTA19.

      MATERIALS AND METHODS

       Data

      Twenty-seven traits of Guernsey cattle were studied, including net merit, 5 yield traits, 7 functional traits, and 14 conformation traits (Table 1). Animals' de-regressed PTA from the December 2018 traditional evaluation were used as the phenotype for each trait. Although yield deviations and daughter yield deviations may be more accurate than de-regressed PTA, these were not available for all traits for all animals. The animals used in this study include both bulls and cows from the reference population used to compute the US national genomic evaluations (
      • Wiggans G.R.
      • Cole J.B.
      • Hubbard S.M.
      • Sonstegard T.S.
      Genomic selection in dairy cattle: The USDA experience.
      ), which is defined as those animals with PTA reliability higher than the reliability of their parent average (PA); as a result, the set of animals varies slightly from trait to trait. All animals were genotyped and imputed up to HD genotypes of 643,059 SNP variants. The 643,059 variants were derived from 777K Illumina BovineHD Genotyping BeadChip (Illumina Inc., San Diego, CA) and 80K (79,276 SNP) markers used in the official US dairy cattle evaluations (
      • VanRaden P.M.
      • Tooker M.E.
      • O'Connell J.R.
      • Cole J.B.
      • Bickhart D.M.
      Selecting sequence variants to improve genomic predictions for dairy cattle.
      ;
      • Wiggans G.R.
      • VanRaden P.M.
      • Null D.J.
      • Cole J.B.
      Genomic predictions using more markers and gene tests.
      ). Properties of the chips used to genotype the animals in this study and imputation accuracy for each are shown in Table 2. Chromosome locations of the imputed HD markers were from the ARS-UCD 1.2 assembly of the Bos taurus genome (
      • Rosen B.D.
      • Bickhart D.M.
      • Schnabel R.D.
      • Koren S.
      • Elsik C.G.
      • Tseng E.
      • Rowan T.N.
      • Low W.Y.
      • Zimin A.
      • Couldrey C.
      • Hall R.
      • Li W.
      • Rhie A.
      • Ghurye J.
      • McKay S.D.
      • Thibaud-Nissen F.
      • Hoffman J.
      • Murdoch B.M.
      • Snelling W.M.
      • McDaneld T.G.
      • Hammond J.A.
      • Schwartz J.C.
      • Nandolo W.
      • Hagen D.E.
      • Dreischer C.
      • Schultheiss S.J.
      • Schroeder S.G.
      • Phillippy A.M.
      • Cole J.B.
      • Van Tassell C.P.
      • Liu G.
      • Smith T.P.L.
      • Medrano J.F.
      De novo assembly of the cattle reference genome with single-molecule sequencing.
      ).
      Table 1Number of animals for each trait included in genome-wide association studies
      TraitNumber of animals
      TotalBullsCows
      Yield traits
      Yield traits include milk yield, fat yield, protein yield, fat percentage, and protein percentage.
      1,6855031,182
      Net merit1,377484893
      Productive life1,358483875
      SCS1,355481874
      Daughter pregnancy rate1,354484870
      Heifer conception rate1,077248829
      Cow conception rate1,305420885
      Livability1,376483893
      Gestation length1,5034871,016
      Final score1,4954201,075
      Stature, strength, dairy form1,5464711,075
      Foot angle, rear legs (side view)1,5354711,064
      Rump angle, front teat placement1,5494741,075
      Rump width, rear udder height, udder depth1,5504741,076
      Teat length, fore udder attachment1,5524741,078
      Udder cleft1,5454741,071
      1 Yield traits include milk yield, fat yield, protein yield, fat percentage, and protein percentage.
      Table 2The SNP count, number of genotyped bulls and cows, and imputation accuracy for the chips used in this study
      Chip name
      Illumina Inc., San Diego, CA; GeneSeek, Inc., Lincoln, NE; Zoetis, Inc., Parsippany, NJ; Irish Cattle Breeding Association, Bandon, County Cork, Ireland.
      SNP countBulls genotypedCows genotypedImputation accuracy
      Mean imputation accuracy to high density (n = 643,059 SNP), measured as the squared correlations of imputed with true genotypes averaged across loci for each animal.
      Illumina BovineSNP50 BeadChip Version 258,336152180.7615
      Illumina Bovine3K BeadChip2,900030.7194
      Illumina BovineHD BeadChip777,962109250.9966
      GeneSeek Genomic Profiler8,762100.7405
      GeneSeek Genomic Profiler-HD77,0681723230.7744
      GeneSeek Genomic Profiler-Super LD19,80911560.7470
      Zoetis Low Density11,410010.7432
      GeneSeek Genomic Profiler LD Version 326,15151,2990.7506
      Zoetis Medium Density Version 260,91420170.7640
      GeneSeek Genomic Profiler Bovine 150K139,914183190.9982
      GeneSeek Genomic Profiler Bovine 7K7,083010.7380
      GeneSeek Genomic Profiler LD Version 430,1131155040.7526
      GeneSeek Genomic Profiler Bovine 50K47,8501145380.7618
      Zoetis Low Density Version 527,7801366870.7531
      Irish Beef and Dairy Chip Version 353,450010.7612
      Summary
      Total number of bulls and cows genotyped, and the weighted average imputation accuracy over all chips used.
      1,0073,5920.7740
      1 Illumina Inc., San Diego, CA; GeneSeek, Inc., Lincoln, NE; Zoetis, Inc., Parsippany, NJ; Irish Cattle Breeding Association, Bandon, County Cork, Ireland.
      2 Mean imputation accuracy to high density (n = 643,059 SNP), measured as the squared correlations of imputed with true genotypes averaged across loci for each animal.
      3 Total number of bulls and cows genotyped, and the weighted average imputation accuracy over all chips used.
      In the data filtering and quality control processes, SNP with call rates <0.90, SNP with minor allele frequency <0.01, and animals with parent-progeny Mendelian conflicts were omitted from the data set, following the method described by
      • Wiggans G.R.
      • VanRaden P.M.
      • Bacheller L.R.
      • Tooker M.E.
      • Hutchison J.L.
      • Cooper T.A.
      • Sonstegard T.S.
      Selection and management of DNA markers for use in genomic evaluation.
      . After quality control, the number of markers remaining in the final data set for GWA analyses ranged from 574,262 to 579,581 for different traits, and the number of genotyped animals included in the following GWA analyses is shown in Table 1. Mean imputation accuracy, estimated as the squared correlations of imputed with true genotypes averaged across loci for each animal from a series of simulaton studies (
      • VanRaden P.M.
      • Null D.J.
      • Sargolzaei M.
      • Wiggans G.R.
      • Tooker M.E.
      • Cole J.B.
      • Sonstegard T.S.
      • Connor E.E.
      • Winters M.
      • van Kaam J.B. C. H.M.
      • Valentini A.
      • Van Doormaal B.J.
      • Faust M.A.
      • Doak G.A.
      Genomic imputation and evaluation using high-density Holstein genotypes.
      ), was 87.4 ± 2.3%.

       Genome-Wide Association Analyses

      Single-trait GWA analyses were carried out by a mixed-model approach using MMAP version 2019_05_10_13_08.intel (
      • O'Connell J.R.
      MMAP: Mixed Model Analysis for Pedigrees and Populations.
      ). The genetic model used for GWA analyses was
      y = 1μ + Isex + Xb + g + e,
      [1]


      where y is the vector of de-regressed PTA for a single trait, calculated as y=PTA/RelPTA2 (
      • Garrick D.J.
      • Taylor J.F.
      • Fernando R.L.
      Deregressing estimated breeding values and weighting information for genomic regression analyses.
      ); μ is the global mean, and 1 is a vector of ones; sex is a vector of fixed effects of animal sex, and I is an identity matrix assigning individuals to their sex, which was found to be statistically significant (P < 0.05); X is a matrix of genotypes (coded as 0, 1, or 2 copies of the minor allele) for the animals with observations in y, and b is a vector of marker effects; g is a vector of polygenic effects accounting for population structure with Var(g) ~N(0,Gσg2) where the genomic relationship matrix (G) was built using the HD markers and σg2 is the genetic variance; and e is a vector of random residual errors with Var(e) ~N(0,Rσe2), where, σe2 is residual error variance and R is a diagonal matrix that adjusts σe2 to account for the different amount of information available for bulls versus cows.
      Preliminary analyses including only bulls, only cows, and bulls and cows together suggested that weighting of the residual variances could reduce stratification and stabilize solutions. The weights included in R were a function of reliabilities of traditional PTA (RelPTA) and reliabilities of traditional PA (RelPA), as
      WeightAnimal=RelAnimal1-RelAnimal,


      where reliabilties for each animal (RelAnimal) were computed as a function of daughter equivalents (DE):
      RelAnimal=DEPTA-DEPADEPTA-DEPA+1


      and daughter equivalents for PTA and PA were calculated as
      DEPTA=RelPTA1-RelPTA


      and
      DEPA=RelPA1-RelPA.


      Sex was also included in all models, because both bulls and cows were included in each analysis, which resulted in an approximate 3-fold increase in genotyped animals. Double-counting of information from both parents and their progeny was avoided by using a multiple-stage genomic BLUP mode that includes a 3-component seletion index of PA, genomic information, and a pedigree index based on relationships among genotyped animals (
      • VanRaden P.M.
      • Van Tassell C.P.
      • Wiggans G.R.
      • Sonstegard T.S.
      • Schnabel R.D.
      • Taylor J.F.
      • Schenkel F.S.
      Invited review: Reliability of genomic predictions for North American Holstein bulls.
      ). Subtraction of the pedigree index term avoids double-counting of the genomic information also included in the PA. This method may not avoid all double-counting but, in practice, seems to work reasonably well. A Bonferroni correction was applied to account for multiple testing in single-trait GWAS. Genome-wide significance thresholds corresponding to a nominal 5% type I error were applied and were 0.05/N, where N was the number of markers. The genome-wide significance threshold ranged from 8.6 × 10−8 to 8.7 × 10−8 for different traits, due to the different reference population sizes causing slightly different numbers of tested markers.

       Conditional Analyses on BTA19

      For the traits that showed major QTL on BTA19 from GWA analyses, conditional analyses were applied to detect independent SNP significantly associated with the traits on BTA19, considering many significant SNP from GWA were in high linkage disequilibrium (LD). In the next round, the most significant SNP on BTA19 observed from previous GWA analyses was fitted as a covariate in model [1], and the remaining SNP were scanned to detect the presence of other genome-wide significant SNP on BTA19. If the result of that round detected any other significant SNP on BTA19, the most significant SNP from that round was fixed as another covariate in the model, and the remaining SNP were scanned in the subsequent round. This same procedure was iterated until no additional SNP were detected as significant on BTA19. The list of significant SNP that were fixed as covariates were collected for each trait as the independent SNP on BTA19 significantly influencing the trait, inferring single or multiple QTL for the trait on BTA19.

       Haplotype Analyses for QTL on BTA19

      Based on the GWA results, the haplotype that hosts the major QTL on BTA19 was located using findhap.f90 version 3 (
      • VanRaden P.M.
      Findhap.F90: Find Haplotypes and Impute Genotypes Using Multiple Chip Sets and Sequence Data.
      ). The frequency of the haplotype in the Guernsey population was detected over years, and the haplotype was traced back to detect the founder animal of this haplotype in the Guernsey population. In addition, the haplotype was also traced back in Holstein, Jersey, Brown Swiss, and Ayrshire populations in the United States, to detect the appearance of this haplotype in other dairy breeds.

      RESULTS AND DISCUSSION

       Major QTL Detected on BTA19

      A large QTL was detected on BTA19, influencing several milk production traits, conformation traits, and livability of Guernsey cattle, where the top SNP of the QTL lie in the region of 26.2 to 28.3 Mb on BTA19 (Table 3). We used a mixed-model approach that incorporated reliability variation across individual animals. The average genomic inflation factor of our GWA analyses was 0.92 (ranging from 0.83 to 0.99) for all analyzed traits.
      Table 3Single nucleotide polymorphisms with large effects on BTA19 in Guernsey cattle: SNP names, positions (Pos), minor allele frequencies (MAF), P-values of the SNP effect, closest gene to the top SNP, direction of the SNP effect (+ = positive, − = negative),
      The sign is based on the effect of the minor allele and is relative to the trait definition (e.g., a positive SNP effect for stature is associated with taller cows).
      and size of the effect in additive genetic SD
      TraitPos (bp)SNP nameSNP
      rs = NCBI RefSeq ID.
      (rs)
      MAFP-valueGeneDirectionEffect
      Milk28,099,297BovineHD1900008426rs1338561410.465.44E-13MYH100.2307
      Protein27,838,415ARS-BFGL-NGS-95155rs1105027300.442.79E-09AURKB0.1860
      Productive life26,298,780BovineHD1900008028rs1351022990.428.26E-17RABEP10.3530
      Protein percent
      The top SNP for protein percentage, fat percentage, livability, and SCS were chromosome-wide significant but not genome-wide significant.
      28,096,864BovineHD1900008425rs1377689210.467.39E-07MYH100.1548
      Fat percent
      The top SNP for protein percentage, fat percentage, livability, and SCS were chromosome-wide significant but not genome-wide significant.
      27,838,415ARS-BFGL-NGS-95155rs1105027300.442.32E-07AURKB+0.1434
      Rump angle28,339,070BovineHD1900008517rs1361745730.497.09E-12PIK3R6+0.2112
      Rump width28,096,864BovineHD1900008425rs1377689210.462.21E-17MYH10+0.1295
      Stature28,096,864BovineHD1900008425rs1377689210.463.95E-65MYH10+0.4291
      Strength27,838,415ARS-BFGL-NGS-95155rs1105027300.432.52E-33AURKB0.3529
      Dairy form28,099,297BovineHD1900008426rs1338561410.462.82E-60MYH10+0.4683
      Teat length28,099,297BovineHD1900008426rs1338561410.464.14E-15MYH10+0.1973
      Udder depth25,859,348BovineHD1900007870rs1328569660.484.60E-21WSCD1+0.3984
      Livability
      The top SNP for protein percentage, fat percentage, livability, and SCS were chromosome-wide significant but not genome-wide significant.
      26,818,990BovineHD4100014094rs1098881620.402.82E-06+0.2456
      Net merit35,216,837BovineHD1900010409rs1364213320.011.30E-08TMEM110.1696
      SCS
      The top SNP for protein percentage, fat percentage, livability, and SCS were chromosome-wide significant but not genome-wide significant.
      36,221,449BovineHD1900010689rs419109150.011.39E-060.1322
      DPR
      Daughter pregnancy rate.
      35,283,135Hapmap42945-BTA-45128rs416435720.019.72E-10MAP2K30.1632
      HCR
      Heifer conception rate.
      37,705,042BovineHD1900011119rs1101413990.012.49E-09CALCOCO2+0.3468
      1 The sign is based on the effect of the minor allele and is relative to the trait definition (e.g., a positive SNP effect for stature is associated with taller cows).
      2 rs = NCBI RefSeq ID.
      3 The top SNP for protein percentage, fat percentage, livability, and SCS were chromosome-wide significant but not genome-wide significant.
      4 Daughter pregnancy rate.
      5 Heifer conception rate.
      The detected QTL on BTA19 in the region of 26.2 to 28.3 Mb has a large effect on milk yield, protein yield, and productive life, with the effects reaching genome-wide significance (Figure 1). For fat yield, the effect of the QTL on BTA19 was not shown to be significant, and the largest effects flanked the diacylglycerol O-acyltransferase 1 (DGAT1) gene on BTA14 (Figure 1). For fat percentage and protein percentage, effects of this QTL on BTA19 reached chromosome-wide significance level, but the effect sizes were smaller than the effect of DGAT1 (Figure 2). Meanwhile, the detected QTL at 26.2 to 28.3 Mb on BTA19 had a highly significant effect on several conformation traits, including stature, strength, dairy form, teat length, udder depth, rump angle, and rump width of Guernsey cattle (Figure 3; Table 3). In addition, apart from the candidate QTL region of 26.2 to 28.3 Mb on BTA19 for production and conformation traits, another region at 35.2 to 37.7 Mb on BTA19 showed significant effects on net merit, DPR, heifer conception rate, and SCS (Figure 4). The top associated SNP in this region had relatively low minor allele frequencies (Table 3). The low heritability of fertility traits and SCS may have contributed to noise in estimating SNP effects for these traits.
      Figure thumbnail gr1
      Figure 1Manhattan plots for milk yield, fat yield, protein yield, and productive life for Guernsey cattle. The horizontal line indicates the genome-wide significance threshold.
      Figure thumbnail gr2
      Figure 2Manhattan plots for milk fat percentage and protein percentage for Guernsey cattle. The horizontal line indicates the genome-wide significance threshold.
      Figure thumbnail gr3
      Figure 3Manhattan plots for stature, dairy form, teat length, and rump width for Guernsey cattle. The horizontal line indicates the genome-wide significance threshold.
      Figure thumbnail gr4
      Figure 4Manhattan plots for daughter pregnancy rate (DPR), heifer conception rate (HCR), and net merit (NM) for Guernsey cattle. The horizontal line indicates the genome-wide significance threshold.
      Selection over many generations has not resulted in notable allele frequency changes for the QTL, despite its large effect on several traits. This is probably because it favorably affects milk and stature but unfavorably affects health and fertility traits (Table 3). Genetic trends were favorable for milk yield and unfavorable for DPR (Figure 5), a pattern similar to the well-known correlation of production gains with fertility reductions in US Holstein cattle (e.g.,
      • Lucy M.C.
      Reproductive loss in high-producing dairy cattle: Where will it end?.
      ). It is also important to note that the SNP with the largest effect on lifetime net merit is located in a different region (~35.2 Mbp) than the QTL currently being studied, so selection for lifetime net merit would not necessarily be expected to change the frequency of the QTL. The total merit indices designed for commercial dairy farmers in the United States have changed over time (Table 4), with substantial selection emphasis on milk, fat, and protein until 1994. They did not include any emphasis on health until 1994, when productive life and SCS were added to the index, and DPR was not included until 2004.
      Figure thumbnail gr5
      Figure 5Genetic trend for milk yield (left) and daughter pregnancy rate (DPR, right) of Guernsey bulls (solid red line) and cows (broken blue line). BV = breeding value. Data were taken from the Council on Dairy Cattle Breeding's Genetic and Phenotypic Trend report (https://queries.uscdcb.com/eval/summary/trend.cfm).
      Table 4Traits included in USDA selection indices
      PD$ = predicted difference dollars, MFP$ = milk-fat-protein dollars, CY$ = cheese yield dollars, and NM$ = lifetime net merit.
      and the relative emphasis (%) placed on each for the years 1971 to 2018
      Trait
      PL = productive life, UC = udder composite, FLC = foot and leg composite, BWC = body weight composite, DPR = daughter pregnancy rate, SCE = sire calving ease, DCE = daughter calving ease, CA$ = calving ability dollars, HCR = heifer conception rate, CCR = cow conception rate, LIV = cow livability, and HTH$ = health dollars.
      PD$ (1971)MFP$ (1976)CY$ (1984)NM$ (1994)NM$ (2000)NM$ (2003)NM$ (2006)NM$ (2010)NM$ (2014)NM$ (2017)NM$ (2018)
      Milk5227−265000−1−1−1
      Fat4846452521222319222427
      Protein27534336332316201817
      PL2014111722191312
      SCS−6−9−9−9−10−7−7−4
      UC7767877
      FLC4434333
      BWC−4−3−4−6−5−6−5
      DPR7911777
      SCE−2
      DCE−2
      CA$65555
      HCR111
      CCR222
      LIV77
      HTH$2
      1 PD$ = predicted difference dollars, MFP$ = milk-fat-protein dollars, CY$ = cheese yield dollars, and NM$ = lifetime net merit.
      2 PL = productive life, UC = udder composite, FLC = foot and leg composite, BWC = body weight composite, DPR = daughter pregnancy rate, SCE = sire calving ease, DCE = daughter calving ease, CA$ = calving ability dollars, HCR = heifer conception rate, CCR = cow conception rate, LIV = cow livability, and HTH$ = health dollars.
      For production and conformation traits, the top SNP of the major QTL region in 26.2 to 28.3 Mb on BTA19 are common variants (Table 3). The closest genes to these top SNP include MYH10, AURKB, RABEP1, PIK3R6, and WSCD1 (Table 3). Among the genes flagged, MYH10 (myosin heavy chain 10) had effects on several production traits (milk yield, protein percentage) and conformation traits (stature, rump width, dairy form, teat length) of Guernsey cattle (Table 3). The MYH10 gene is a member of the myosin superfamily of protein-coding genes in cattle, humans, mice, chickens, sheep, and several other species. The MYH10 gene is associated with skeletal muscle in human (
      • Mascarello F.
      • Toniolo L.
      • Cancellara P.
      • Reggiani C.
      • Maccatrozzo L.
      Expression and identification of 10 sarcomeric MyHC isoforms in human skeletal muscles of different embryological origin. Diversity and similarity in mammalian species.
      ), muscle characteristics in beef cattle (
      • Yang Z.
      • Zhao X.
      • Xiong X.
      • Bao L.
      • Pan K.
      • Zhou S.
      • Wen L.
      • Xu L.
      • Qu M.
      Uncovering the mechanism whereby dietary nicotinic acid increases the intramuscular fat content in finishing steers by RNA sequencing analysis.
      ) and pigs (
      • Óvilo C.
      • Benítez R.
      • Fernández A.
      • Núñez Y.
      • Ayuso M.
      • Fernández A.I.
      • Rodríguez C.
      • Isabel B.
      • Rey A.I.
      • López-Bote C.
      • Silió L.
      Longissimus dorsi transcriptome analysis of purebred and crossbred Iberian pigs differing in muscle characteristics.
      ), and body size in sheep (
      • Kominakis A.
      • Hager-Theodorides A.L.
      • Zoidis E.
      • Saridaki A.
      • Antonakos G.
      • Tsiamis G.
      Combined GWAS and 'guilt by association’-based prioritization analysis identifies functional candidate genes for body size in sheep.
      ;
      • Signer-Hasler H.
      • Burren A.
      • Ammann P.
      • Drögemuller C.
      • Flury C.
      Runs of homozygosity and signatures of selection: A comparison among eight local Swiss sheep breeds.
      ). In Holstein cattle, the MYH10 gene is associated with milk production (
      • Do D.N.
      • Dudemaine P.-L.
      • Li R.
      • Ibeagha-Awemu E.M.
      Co-expression network and pathway analyses reveal important modules of miRNAs regulating milk yield and component traits.
      ) and body conformation traits (
      • Abo-Ismail M.K.
      • Brito L.F.
      • Miller S.P.
      • Sargolzaei M.
      • Grossi D.A.
      • Moore S.S.
      • Plastow G.
      • Stothard P.
      • Nayeri S.
      • Schenkel F.S.
      Genome-wide association studies and genomic prediction of breeding values for calving performance and body conformation traits in Holstein cattle.
      ;
      • Jiang J.
      • Cole J.B.
      • Freebern E.
      • Da Y.
      • VanRaden P.M.
      • Ma L.
      Functional annotation and Bayesian fine-mapping reveals candidate genes for important agronomic traits in Holstein bulls.
      ), which is consistent with our findings. The role of MYH10 in controlling body conformation traits is similar between Guernsey and Holstein cattle. However, in contrast with Guernsey cattle, the largest QTL in Holstein cattle influencing several conformation traits and calving traits is on BTA18 (
      • Cole J.B.
      • VanRaden P.M.
      • O'Connell J.R.
      • van Tassell C.P.
      • Sonstegard T.S.
      • Schnabel R.D.
      • Taylor J.F.
      • Wiggans G.R.
      Distribution and location of genetic effects for dairy traits.
      ;
      • Abo-Ismail M.K.
      • Brito L.F.
      • Miller S.P.
      • Sargolzaei M.
      • Grossi D.A.
      • Moore S.S.
      • Plastow G.
      • Stothard P.
      • Nayeri S.
      • Schenkel F.S.
      Genome-wide association studies and genomic prediction of breeding values for calving performance and body conformation traits in Holstein cattle.
      ). Another MYH gene, MYH14 (myosin heavy chain 14), is a candidate gene associated with the BTA18 QTL in Holsteins (
      • Abo-Ismail M.K.
      • Brito L.F.
      • Miller S.P.
      • Sargolzaei M.
      • Grossi D.A.
      • Moore S.S.
      • Plastow G.
      • Stothard P.
      • Nayeri S.
      • Schenkel F.S.
      Genome-wide association studies and genomic prediction of breeding values for calving performance and body conformation traits in Holstein cattle.
      ). These findings indicate a general role for the MYH gene family (e.g., MYH10, MYH14) in controlling body conformation traits of dairy cattle, which could be explained by the important role of expression of the MYH gene family in mammalian muscle physiology (
      • Weiss A.
      • Leinwand L.A.
      The mammalian myosin heavy chain gene family.
      ).
      Chromosome 19 was selected for detailed analysis based on an examination of the allele substitution effects that are computed as part of routine genome evaluations. Those preliminary results showed a large QTL on BTA19 affecting many traits, but few QTL on other chromosomes. These data are not shown because they are proprietary information belonging to the Council on Dairy Cattle Breeding (Bowie, MD). Access to those data may be requested from João Dürr, chief executive officer of the Council on Dairy Cattle Breeding ( joao.durr@uscdcb.com ).
      Although power calculations are not straightforward for GWAS studies (e.g.,
      • Ohashi J.
      • Tokunaga K.
      The power of genome-wide association studies of complex disease genes: Statistical limitations of indirect approaches using SNP markers.
      ;
      • Delongchamp R.
      • Faramawi M.F.
      • Feingold E.
      • Chung D.
      • Abouelenein S.
      The association between SNPs and a quantitative trait: Power calculation.
      ), it is clear that the sample size available in this study limits its ability to detect QTL of moderate or small size for most traits, and even large QTL may not be detectable for traits with low heritability (e.g., fertility). Our results should be interpreted as a lower bound, in a sense, of the number of true QTL in the Guernsey population.

       Testing for Multiple QTL on BTA19

      For most traits, no additional genome-wide significant SNP were detected on BTA19 after fixing the most significant SNP as covariates in the model, indicating a single independent SNP associated with the related traits on BTA19 (Table 5). An example is shown in Figure 6 for teat length, where no additional significant SNP were detected after the top SNP from GWAS was fixed in the model. However, for protein, dairy form, and stature, multiple independent SNP were detected on BTA19 that are significantly associated with these traits (Table 5). The multiple, independent, significant SNP on BTA19 for these 3 traits might indicate multiple QTL associated with these traits on BTA19. Alternatively, the multiple independent SNP might indicate that these markers are partially correlated with one QTL, are in weak LD with each other, or had differences in imputation accuracy.
      Table 5Independent SNP significantly associated with the putative QTL on BTA19 in Guernsey cattle, and direction of the SNP effect (+ = positive, − = negative)
      The sign is based on the effect of the minor allele and is relative to the trait definition (e.g., a positive SNP effect for stature is associated with taller cows).
      TraitPosition (bp)SNP nameSNP
      rs = NCBI RefSeq ID.
      (rs)
      Direction
      Milk28,099,297BovineHD1900008426rs133856141+
      Protein27,838,415ARS-BFGL-NGS-95155rs110502730
      31,759,563ARS-BFGL-NGS-6071rs109214986
      Productive life26,298,780BovineHD1900008028rs135102299
      Stature28,096,864BovineHD1900008425rs137768921+
      26,692,730ARS-BFGL-NGS-117046rs41907795
      Strength27,838,415ARS-BFGL-NGS-95155rs110502730
      Dairy form28,099,297BovineHD1900008426rs133856141+
      36,245,752BovineHD1900010701rs41911375+
      Rump angle28,339,070BovineHD1900008517rs136174573+
      Rump width28,096,864BovineHD1900008425rs137768921+
      Teat length28,099,297BovineHD1900008426rs133856141+
      Udder depth25,859,348BovineHD1900007870rs132856966+
      Net merit35,216,837BovineHD1900010409rs136421332
      DPR
      Daughter pregnancy rate.
      35,283,135Hapmap42945-BTA-45128rs41643572
      HCR
      Heifer conception rate.
      37,705,042BovineHD1900011119rs110141399+
      1 The sign is based on the effect of the minor allele and is relative to the trait definition (e.g., a positive SNP effect for stature is associated with taller cows).
      2 rs = NCBI RefSeq ID.
      3 Daughter pregnancy rate.
      4 Heifer conception rate.
      Figure thumbnail gr6
      Figure 6Conditional analyses to test for multiple QTL on BTA19 for teat length of Guernsey cattle. The −log (P-value) of SNP effects on BTA19 before (A) and after (B) correcting for the most significant SNP from GWA analyses. The red and blue horizontal lines indicate the genome-wide significance threshold and chromosome-wide significance threshold, respectively.
      The marker ARS-BFGL-NGS-95155 (19:27,838,415), which has an unfavorable effect on protein yield, and the SNP BovineHD1900008426 (19:28,099,297), which has a favorable effect on milk yield, are located relatively close to one another (260,882 bp). The measure of LD (r2) of 0.89 between them indicates fairly tight linkage, which suggests that these effects may be due to a polymorphism with pleitropic effects, rather than 2 different causative loci in the same small region. Complete pairwise LD measures computed using Haploview version 4.2 (
      • Barrett J.C.
      • Fry B.
      • Maller J.
      • Daly M.J.
      Haploview: Analysis and visualization of LD and haplotype maps.
      ) are provided in a text format in Supplemental File S1 (https://doi.org/10.3168/jds.2020-18766).

       Haplotype Status of the QTL on BTA19

      The haplotype carrying the QTL on BTA19, which spans 27.1 to 28.7 Mbp, can be traced back to a US Guernsey (GUUSA000000512974, HENSLEE FARMS V FAME) born in 1954. The haplotype is common at about 50% in the current population (Figure 7), and the frequencies of the haplotype indicate a potential long-term influence of this QTL on US Guernseys. This haplotype is not segregating in the genotyped Ayrshire, Brown Swiss, Holstein, or Jersey animals in the US national database.
      Figure thumbnail gr7
      Figure 7The frequency over time of the haplotype that contains the major QTL on BTA19 influencing milk production, conformation, and cow livability of Guernsey cattle. Due to sparsity of data from genotyped bulls born before 1980, frequencies from 1954 to 1980 were averaged into a single observation.

      CONCLUSIONS

      A major QTL influencing milk production, body conformation, and livability of Guernsey cattle was detected on BTA19. The SNP associated with the QTL lie in the region of 26.2 to 28.3 Mb on BTA19, which includes the MYH10 gene that is strongly associated with milk production and body conformation traits. For the 13 traits that showed genome-wide significance of QTL effects on BTA19, conditional analyses detected only 3 traits with additional, independent SNP with significant influence on BTA19. Those traits with multiple independent SNP were protein, dairy form, and stature of Guernsey. The haplotype carrying the QTL is common in the US Guernsey population, indicating a potential long-term influence on the breed, but was not detected in other breeds with genotypes in the US database.

      ACKNOWLEDGMENTS

      The Council on Dairy Cattle Breeding (CDCB; Bowie, MD) and its industry collaborators are thanked for providing phenotypic, pedigree, and genomic information used in this analysis. The Agriculture and Horticulture Development Board (Kenilworth, UK) and World Guernsey Cattle Federation (Castel, Guernsey, Channel Islands, UK) are gratefully acknowledged for providing genotypes included in this study as part of an international data exchange with CDCB. Two anonymous reviewers are thanked for their thoughtful feedback, which improved the quality of this paper. Cole, Null, and VanRaden were supported by appropriated project 8042-31000-002-00-D, “Improving Dairy Animals by Increasing Accuracy of Genomic Prediction, Evaluating New Traits, and Redefining Selection Goals.” B. Li was supported in part by an appointment to the Agriculture Research Service (ARS) Research Participation Program, administered by the Oak Ridge Institute for Science and Education (Oak Ridge, TN) through an interagency agreement between the U.S. Department of Energy and ARS (Washington, DC). Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. The USDA is an equal opportunity provider and employer. The authors have not stated any conflicts of interest.

      Supplementary Material

      REFERENCES

        • Abo-Ismail M.K.
        • Brito L.F.
        • Miller S.P.
        • Sargolzaei M.
        • Grossi D.A.
        • Moore S.S.
        • Plastow G.
        • Stothard P.
        • Nayeri S.
        • Schenkel F.S.
        Genome-wide association studies and genomic prediction of breeding values for calving performance and body conformation traits in Holstein cattle.
        Genet. Sel. Evol. 2017; 49 (29115939): 82-89
        • Barrett J.C.
        • Fry B.
        • Maller J.
        • Daly M.J.
        Haploview: Analysis and visualization of LD and haplotype maps.
        Bioinformatics. 2005; 15: 263-265
        • Cole J.B.
        • VanRaden P.M.
        • O'Connell J.R.
        • van Tassell C.P.
        • Sonstegard T.S.
        • Schnabel R.D.
        • Taylor J.F.
        • Wiggans G.R.
        Distribution and location of genetic effects for dairy traits.
        J. Dairy Sci. 2009; 92 (19448026): 2931-2946
        • Cooper T.A.
        • Eaglen S.A.E.
        • Wiggans G.R.
        • Jenko J.
        • Huson H.J.
        • Morrice D.R.
        • Bichard M.
        • Luff W.G.L.
        • Woolliams J.A.
        Genomic evaluation, breed identification, and population structure of Guernsey cattle in North America, Great Britain, and the Isle of Guernsey.
        J. Dairy Sci. 2016; 99 (27179857): 5508-5515
        • Delongchamp R.
        • Faramawi M.F.
        • Feingold E.
        • Chung D.
        • Abouelenein S.
        The association between SNPs and a quantitative trait: Power calculation.
        Eur. J. Environ. Public Health. 2018; 2 (31259311)
        • Do D.N.
        • Dudemaine P.-L.
        • Li R.
        • Ibeagha-Awemu E.M.
        Co-expression network and pathway analyses reveal important modules of miRNAs regulating milk yield and component traits.
        Int. J. Mol. Sci. 2017; 18 (28718798)1560
        • Garrick D.J.
        • Taylor J.F.
        • Fernando R.L.
        Deregressing estimated breeding values and weighting information for genomic regression analyses.
        Genet. Sel. Evol. 2009; 41 (20043827): 55
        • Jiang J.
        • Cole J.B.
        • Freebern E.
        • Da Y.
        • VanRaden P.M.
        • Ma L.
        Functional annotation and Bayesian fine-mapping reveals candidate genes for important agronomic traits in Holstein bulls.
        Commun. Biol. 2019; 2 (31240250): 212
        • Kominakis A.
        • Hager-Theodorides A.L.
        • Zoidis E.
        • Saridaki A.
        • Antonakos G.
        • Tsiamis G.
        Combined GWAS and 'guilt by association’-based prioritization analysis identifies functional candidate genes for body size in sheep.
        Genet. Sel. Evol. 2017; 49 (28454565): 41
        • Lucy M.C.
        Reproductive loss in high-producing dairy cattle: Where will it end?.
        J. Dairy Sci. 2001; 84 (11417685): 1277-1293
        • Mascarello F.
        • Toniolo L.
        • Cancellara P.
        • Reggiani C.
        • Maccatrozzo L.
        Expression and identification of 10 sarcomeric MyHC isoforms in human skeletal muscles of different embryological origin. Diversity and similarity in mammalian species.
        Ann. Anat. 2016; 207 (26970499): 9-20
        • Null D.J.
        • VanRaden P.M.
        • Rosen B.D.
        • O'Connell J.R.
        • Bickhart D.M.
        Using the ARS-UCD1.2 reference genome in U. S. evaluations.
        Interbull Bull. 2019; 55: 30-34
        • O'Connell J.R.
        MMAP: Mixed Model Analysis for Pedigrees and Populations.
        https://mmap.github.io/
        Date: 2020
        Date accessed: September 3, 2020
        • Ohashi J.
        • Tokunaga K.
        The power of genome-wide association studies of complex disease genes: Statistical limitations of indirect approaches using SNP markers.
        J. Hum. Genet. 2001; 46 (11501946): 478-482
        • Óvilo C.
        • Benítez R.
        • Fernández A.
        • Núñez Y.
        • Ayuso M.
        • Fernández A.I.
        • Rodríguez C.
        • Isabel B.
        • Rey A.I.
        • López-Bote C.
        • Silió L.
        Longissimus dorsi transcriptome analysis of purebred and crossbred Iberian pigs differing in muscle characteristics.
        BMC Genomics. 2014; 15 (24885501): 413
        • Rosen B.D.
        • Bickhart D.M.
        • Schnabel R.D.
        • Koren S.
        • Elsik C.G.
        • Tseng E.
        • Rowan T.N.
        • Low W.Y.
        • Zimin A.
        • Couldrey C.
        • Hall R.
        • Li W.
        • Rhie A.
        • Ghurye J.
        • McKay S.D.
        • Thibaud-Nissen F.
        • Hoffman J.
        • Murdoch B.M.
        • Snelling W.M.
        • McDaneld T.G.
        • Hammond J.A.
        • Schwartz J.C.
        • Nandolo W.
        • Hagen D.E.
        • Dreischer C.
        • Schultheiss S.J.
        • Schroeder S.G.
        • Phillippy A.M.
        • Cole J.B.
        • Van Tassell C.P.
        • Liu G.
        • Smith T.P.L.
        • Medrano J.F.
        De novo assembly of the cattle reference genome with single-molecule sequencing.
        Gigascience. 2020; 9 (32191811)giaa021
        • Signer-Hasler H.
        • Burren A.
        • Ammann P.
        • Drögemuller C.
        • Flury C.
        Runs of homozygosity and signatures of selection: A comparison among eight local Swiss sheep breeds.
        Anim. Genet. 2019; 50 (31365135): 512-525
        • VanRaden P.M.
        Findhap.F90: Find Haplotypes and Impute Genotypes Using Multiple Chip Sets and Sequence Data.
        https://aipl.arsusda.gov/software/findhap/
        Date: 2016
        Date accessed: April 22, 2016
        • VanRaden P.M.
        • Null D.J.
        • Sargolzaei M.
        • Wiggans G.R.
        • Tooker M.E.
        • Cole J.B.
        • Sonstegard T.S.
        • Connor E.E.
        • Winters M.
        • van Kaam J.B. C. H.M.
        • Valentini A.
        • Van Doormaal B.J.
        • Faust M.A.
        • Doak G.A.
        Genomic imputation and evaluation using high-density Holstein genotypes.
        J. Dairy Sci. 2013; 96 (23063157): 668-678
        • VanRaden P.M.
        • Tooker M.E.
        • O'Connell J.R.
        • Cole J.B.
        • Bickhart D.M.
        Selecting sequence variants to improve genomic predictions for dairy cattle.
        Genet. Sel. Evol. 2017; 49 (28270096): 32
        • VanRaden P.M.
        • Van Tassell C.P.
        • Wiggans G.R.
        • Sonstegard T.S.
        • Schnabel R.D.
        • Taylor J.F.
        • Schenkel F.S.
        Invited review: Reliability of genomic predictions for North American Holstein bulls.
        J. Dairy Sci. 2009; 92 (19109259): 16-24
        • Weiss A.
        • Leinwand L.A.
        The mammalian myosin heavy chain gene family.
        Annu. Rev. Cell Dev. Biol. 1996; 12 (8970733): 417-439
        • Wiggans G.R.
        • Cole J.B.
        • Hubbard S.M.
        • Sonstegard T.S.
        Genomic selection in dairy cattle: The USDA experience.
        Annu. Rev. Anim. Biosci. 2017; 5 (27860491): 309-327
        • Wiggans G.R.
        • VanRaden P.M.
        • Bacheller L.R.
        • Tooker M.E.
        • Hutchison J.L.
        • Cooper T.A.
        • Sonstegard T.S.
        Selection and management of DNA markers for use in genomic evaluation.
        J. Dairy Sci. 2010; 93 (20412945): 2287-2292
        • Wiggans G.R.
        • VanRaden P.M.
        • Null D.J.
        • Cole J.B.
        Genomic predictions using more markers and gene tests.
        J. Dairy Sci. 2019; 102 (Abstr.): 397
        • World Guernsey Cattle Ferderation
        Guernsey Breed History.
        https://www.worldguernseys.com/history-of-the-guernsey
        Date: 2004
        Date accessed: January 17, 2020
        • Yang Z.
        • Zhao X.
        • Xiong X.
        • Bao L.
        • Pan K.
        • Zhou S.
        • Wen L.
        • Xu L.
        • Qu M.
        Uncovering the mechanism whereby dietary nicotinic acid increases the intramuscular fat content in finishing steers by RNA sequencing analysis.
        Anim. Prod. Sci. 2019; 59: 1620-1630