Development of polymorphic markers in the immune gene complex loci of cattle

Authors

K Bakshy, Dairy Forage Research Center
D Heimeier, The Pirbright Institute
J C. Schwartz, The Pirbright Institute
E J. Glass, The University of Edinburgh
S Wilkinson, The University of Edinburgh
R A. Skuce, Agri-Food and Biosciences Institute
A R. Allen, Agri-Food and Biosciences Institute
J Young, Dairy Forage Research Center
J C. McClure, Dairy Forage Research Center
J B. Cole, Animal Genomics and Improvement Laboratory
D J. Null, Animal Genomics and Improvement Laboratory
J A. Hammond, The Pirbright Institute
T P.L. Smith, Meat Animal Research Center, Clay Center
D M. Bickhart, Dairy Forage Research Center

ORCID IDs

E.J. Glass

A.R. Allen

J.C. McClure

D.J. Null

T.P.L. Smith

D.M. Bickhart

Date of this Version

1-18-2021

Document Type

Article

Citation

J. Dairy Sci. 104:6897–6908 https://doi.org/10.3168/jds.2020-19809

Comments

CC BY-NC-ND

Abstract

The addition of cattle health and immunity traits to genomic selection indices holds promise to increase individual animal longevity and productivity, and decrease economic losses from disease. However, highly variable genomic loci that contain multiple immune-related genes were poorly assembled in the first iterations of the cattle reference genome assembly and underrepresented during the development of most commercial genotyping platforms. As a consequence, there is a paucity of genetic markers within these loci that may track haplotypes related to disease susceptibility. By using hierarchical assembly of bacterial artificial chromosome inserts spanning 3 of these immune-related gene regions, we were able to assemble multiple full-length haplotypes of the major histocompatibility complex, the leukocyte receptor complex, and the natural killer cell complex. Using these new assemblies and the recently released ARS-UCD1.2 reference, we aligned whole-genome shotgun reads from 125 sequenced Holstein bulls to discover candidate variants for genetic marker development. We selected 124 SNPs, using heuristic and statistical models to develop a custom genotyping panel. In a proof-of-principle study, we used this custom panel to genotype 1,797 Holstein cows exposed to bovine tuberculosis (bTB) that were the subject of a previous GWAS study using the Illumina BovineHD array. Although we did not identify any significant association of bTB phenotypes with these new genetic markers, 2 markers exhibited substantial effects on bTB phenotypic prediction. The models and parameters trained in this study serve as a guide for future marker discovery surveys particularly in previously unassembled regions of the cattle genome.