Phenotypic relationships, genetic parameters, genome-wide associations, and identification of potential candidate genes for ketosis and fat-to-protein ratio in German Holstein cows

Abstract

Energy demand for milk production in early lactation exceeds energy intake, especially in high-yielding Holstein cows. Energy deficiency causes increasing susceptibility to metabolic disorders. In addition to several blood parameters, the fat-to-protein ratio (FPR) is suggested as an indicator for ketosis, because a FPR >1.5 refers to high lipolysis. The aim of this study was to analyze phenotypic, quantitative genetic, and genomic associations between FPR and ketosis. In this regard, 8,912 first-lactation Holstein cows were phenotyped for ketosis according to a veterinarian diagnosis key. Ketosis was diagnosed if the cow showed an abnormal carbohydrate metabolism with increased content of ketone bodies in the blood or urine. At least one entry for ketosis in the first 6 wk after calving implied a score = 1 (diseased); otherwise, a score = 0 (healthy) was assigned. The FPR from the first test-day was defined as a Gaussian distributed trait (FPRgauss), and also as a binary response trait (FPRbin), considering a threshold of FPR = 1.5. After imputation and quality controls, 45,613 SNP markers from the 8,912 genotyped cows were used for genomic studies. Phenotypically, an increasing ketosis incidence was associated with significantly higher FPR, and vice versa. Hence, from a practical trait recording perspective, first test-day FPR is suggested as an indicator for ketosis. The ketosis heritability was slightly larger when modeling the pedigree-based relationship matrix (pedigree-based: 0.17; SNP-based: 0.11). For FPRbin, heritabilities were larger when modeling the genomic relationship matrix (pedigree-based: 0.09; SNP-based: 0.15). For FPRgauss, heritabilities were almost identical for both pedigree and genomic relationship matrices (pedigree-based: 0.14; SNP-based: 0.15). Genetic correlations between ketosis with FPRbin and FPRgauss using either pedigree-or genomic-based relationship matrices were in a moderate range from 0.39 to 0.71. Applying genome-wide association studies, we identified the specific SNP rs109896020 (BTA 5, position: 115,456,438 bp) significantly contributing to ketosis. The identified potential candidate gene PARVB in close chromosomal distance is associated with nonalcoholic fatty liver disease in humans. The most important SNP contributing to FPRbin was located within the DGAT1 gene. Different SNP significantly contributed to ketosis and FPRbin, indicating different mechanisms for both traits genomically.

Symposium review: Big data, big predictions: Utilizing milk Fourier-transform infrared and genomics to improve hyperketonemia management

“…Investigations into genetic selection for HYK susceptibility have predominately relied on voluntary records provided by dairy producers. Threshold model heritability estimates have ranged from 0.02 to 0.17 in Holstein cows (Kadarmideen et al, 2000;Zwald et al, 2004;Parker Gaddis et al, 2014;Klein et al, 2019). Presently, Zoetis Genetics (Kalamazoo, MI) is the only provider of genomic evaluations for HYK susceptibility, which is marketed as ketosis resistance, based on voluntary records in the United States (Vukasinovic et al, 2017); however, misclassification of HYK cases as controls (or as nonevents) is a problem with voluntary records, especially when considering subclinical cases and the variation in HYK diagnosis criteria (Oetzel, 2004;Pryce et al, 2016).…”

“…Therefore, genome-wide association studies (GWAS) for HYK may indicate a subset of SNP worthy of increased weighting in genomic evaluation of HYK susceptibility [e.g., weighted genomic best linear unbiased prediction (GBLUP)]. Previous GWAS for HYK based on voluntary records have not identified SNP with genome-wide significant associations (Parker Gaddis et al, 2018;Klein et al, 2019). We believe the present lack of genome-wide associations is due to voluntary record HYK phenotypes having variation in trait definition across herds (i.e., subclinical vs. clinical, urine vs. blood ketone body metrics) and having high misclassification rates caused by insufficient assessment frequency (i.e., only 1 blood BHB measurement within the first 21 DIM).…”

Survival analyses in Holstein cows considering direct disease diagnoses and specific SNP marker effects

“…Starting in 2013 in the framework of the German cow training set project (e.g., Klein et al, 2019), all first-calving cows from the so-called contract herds were considered for genotyping. Hence, the genotyped 17,362 cows represent a random sample of the herd, especially for the birth years 2011 and later.…”

Genome-Wide Association Study Using Whole-Genome Sequence Data for Fertility, Health Indicator, and Endoparasite Infection Traits in German Black Pied Cattle

“…Although well adapted to grazing systems, DSN are exposed to metabolic stress due to heat stress impact on pastures and associated mobilization of body reserve in response to restricted feed intake. The fat-to-protein ratio (FPR) is a valuable indicator for metabolic health (e.g., ketosis) and for metabolic stability in this regard [26,27]. Moreover, FPR might be a novel indicator trait for robustness in DSN, especially in the context of challenging heat stress environments in outdoor systems [28].…”