Department of Reproduction, Obstetrics and Herd Health, Ghent University, Merelbeke 9820, Belgium; Department of Animal Science, Shiraz University, Shiraz 71441-65186, Iran.
The Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK.
J Dairy Sci. 2020 Jul;103(7):6392-6406. doi: 10.3168/jds.2019-17369. Epub 2020 Apr 22.
The aim of this study was to detect the genomic region or regions associated with metabolic clusters in early-lactation Holstein cows. This study was carried out in 2 experiments. In experiment I, which was carried out on 105 multiparous Holstein cows, animals were classified through k-means clustering on log-transformed and standardized concentrations of blood glucose, insulin-like growth factor I, free fatty acids, and β-hydroxybutyrate at 14 and 35 d in milk (DIM), into metabolic clusters, either balanced (BAL) or other (OTR). Forty percent of the animals were categorized in the BAL group, and the remainder were categorized as OTR. The cows were genotyped for a total of 777,962 SNP. A genome-wide association study was performed, using a case-control approach through the GEMMA software, accounting for population structure. We found 8 SNP (BTA11, BTA23, and BTAX) associated with the predicted metabolic clusters. In experiment II, carried out on 4,267 second-parity Holstein cows, milk samples collected starting from the first week until 50 DIM were used to determine Fourier-transform mid-infrared (FT-MIR) spectra and subsequently to classify the animals into the same metabolic clusters (BAL vs. OTR). Twenty-eight percent of the animals were categorized in the BAL group, and the remainder were classified in the OTR category. Although daily milk yield was lower in BAL cows, we found no difference in daily fat- and protein-corrected milk yield in cows from the BAL metabolic cluster compared with those in the OTR metabolic cluster. In the next step, a single-step genomic BLUP was used to identify the genomic region(s) associated with the predicted metabolic clusters. The results revealed that prediction of metabolic clusters is a highly polygenic trait regulated by many small-sized effects. The region of 36,258 to 36,295 kb on BTA27 was the highly associated region for the predicted metabolic clusters, with the closest genes to this region (ANK1 and miR-486) being related to hematopoiesis, erythropoiesis, and mammary gland development. The heritability for metabolic clustering was 0.17 (SD 0.03), indicating that the use of FT-MIR spectra in milk to predict metabolic clusters in early-lactation across a large number of cows has satisfactory potential to be included in genetic selection programs for modern dairy cows.
本研究旨在检测与泌乳早期荷斯坦奶牛代谢群相关的基因组区域或区域。本研究共进行了两项实验。在实验 I 中，对 105 头经产荷斯坦奶牛进行了研究，通过对 14 和 35 天产奶（DIM）时血液葡萄糖、胰岛素样生长因子 I、游离脂肪酸和β-羟丁酸的对数转换和标准化浓度进行 k-均值聚类，将动物分为代谢群，即平衡（BAL）或其他（OTR）。40%的动物被归类为 BAL 组，其余的动物被归类为 OTR。对这些奶牛进行了总计 777962 个 SNP 的基因分型。通过 GEMMA 软件，采用病例对照方法进行全基因组关联研究，同时考虑了群体结构。我们发现 8 个 SNP（BTA11、BTA23 和 BTAX）与预测的代谢群相关。在实验 II 中，对 4267 头二胎荷斯坦奶牛进行了研究，从第一周开始采集牛奶样本，直至 50 DIM，用于测定傅里叶变换中红外（FT-MIR）光谱，并随后将动物分为相同的代谢群（BAL 与 OTR）。28%的动物被归类为 BAL 组，其余的动物被归类为 OTR 组。尽管 BAL 奶牛的日产奶量较低，但我们发现 BAL 代谢群奶牛的日产脂肪校正奶和蛋白校正奶产量与 OTR 代谢群奶牛没有差异。在下一步中，使用单步基因组 BLUP 来识别与预测代谢群相关的基因组区域。结果表明，预测代谢群是一个受许多小效应调节的高度多基因性状。BTA27 上 36258 至 36295 kb 的区域是与预测代谢群高度相关的区域，该区域最接近的基因（ANK1 和 miR-486）与造血、红细胞生成和乳腺发育有关。代谢群聚类的遗传力为 0.17（SD 0.03），这表明使用牛奶的 FT-MIR 光谱来预测大量奶牛泌乳早期的代谢群具有良好的潜力，可以纳入现代奶牛的遗传选择计划。
