Tierärztliche Hochschule Hannover / Bibliothek – University of Veterinary Medicine Hannover – Foundation / Library


Stefanie Hedwig Lehner


Molecular genetic analyses

for left-sided displacement of the abomasum

in German Holstein cattle





Hannover, Tierärztliche Hochschule, Habilitationsschrift, 2014




Left-sided displacement of the abomasum (LDA) is a common and economically important disease in German Holstein cows. Pathogenesis of LDA is multifactorial, but nevertheless genetic components are important as demonstrated by the high heritability at up to 50%. Therefore, molecular genetic analyses are supposed to identify LDA-associated genomic regions and unravel single nucleotide polymorphisms (SNPs) responsible for LDA. For this purpose, linkage analysis and genome-wide association analyses (GWAS) using medium and high density SNP genotyping microarrays were performed to detect genomic regions for LDA. Subsequently, next generation sequencing (NGS) was used to target SNPs of those regions.

The linkage analysis revealed two quantitative trait loci (QTL) on BTA1 and 3 showing genome-wide and three further ones on BTA21, 23, and 24 showing chromosome-wide significant linkage with LDA. Eleven further regions showed linkage with LDA on a family-specific basis. To provide a better understanding of genetic traits co-segregating with LDA, the same data was analysed for milk performance traits and length of productive life. The results showed that LDA was especially correlated with the length of productive live, milk-fat yield, and milk-protein percentage. Unravelling those co-localized genomic regions was an important task to improve genomic selection methods, as the incautious use of production trait markers might lead to an increase of the LDA prevalence and in consequence to a decrease in productivity and longevity of dairy cows. Furthermore, polymorphisms located within the DGAT1 and ABCG2 genes currently used for selection of milk performance traits in cattle were analysed. As no co-segregation of polymorphisms within these genes has been found with LDA, a selection on genotypes of these genes is not likely to raise the incidence of LDA.

For association-mapping and to identify new regions for LDA, we employed the lllumina bovine 50K and the Illumina bovine high density beadchips. Using the Illumina bovine 50K beadchip, a total of 36 significantly associated SNPs on 17 bovine chromosomes were identified for LDA at –log10P>3.0. One of these SNPs was located within the QTL for LDA on BTA24 and two further SNPs were located within or closely adjacent to family-specific QTL of the previous linkage study. Two SNPs on BTA11 and 20 even showed genome-wide significant associations with LDA. The SNP on BTA11 was located within a functional candidate gene for LDA. Pathway analyses using the data of this association study revealed genetic pathways for insulin-dependent diabetes mellitus and calcium metabolism associated with LDA. This provides new insights into the genetic pathogenesis of LDA.

The SNPs associated with LDA using the Illumina bovine HD beadchip showed higher ‑log10P-values and explained larger proportions of the phenotypic variance than the LDA-associated SNPs using the 50K beadchip. Six regions showing highly significant associations with LDA were detected and further 27 locations also showed significant associations with LDA. The highest variance explained by single LDA-associated SNPs of each of the six main regions for LDA was at 6.1% - 7.3%. This confirms the predictive value of these SNPs for a genome-wide LDA-test. Of these six regions detected for LDA, the location on BTA8 was in complete concordance with the one of the 50K beadchip. The region associated with LDA on BTA24 was located within the QTL for LDA of the linkage analysis, and in the 50K beadchip analysis an adjacent region was detected. Therefore, these regions might contain genes with high impact on the affection risk of a cow for LDA. Functional candidate genes for LDA were detected on both chromosomes.

Differences between our linkage analysis and both GWAS as well as between both GWAS are explained by the small effects contributed by each region for LDA as well as the different groups of animals employed and the different marker density used in the studies. Therefore, implementing different study designs for LDA, we were able to detect a good portion of the regions containing common variants for LDA. Combining the polymorphisms of these regions for LDA, it should be possible to explain a bigger part of LDA heritability.

Based on the linkage analysis for LDA, a search for candidate genes for LDA was performed. The bovine motilin (MLN) gene was a functional candidate for LDA due to its known influence on gastrointestinal peristalsis in humans. Sequencing of this gene revealed a SNP (FN298674:g.90T>C) within the first, non-coding exon of MLN, which affected a NKX2-5 transcription factor binding site and showed significant associations with LDA. Furthermore, heterozygous and homozygous mutant genotypes of FN298674:g.90T>C were shown to significantly decrease expression of MLN. The SNP explains a proportion of 3.1% of the phenotypic variance and can be used for selection to reduce the LDA-affection risk in the German Holstein population.

Next generation sequencing (NGS) subsequently was performed to identify SNPs for LDA based on the locations of the previous studies. In total, six samples were sequenced and yielded 7,939,073 different SNPs and indels (insertions and deletions). The proportion of novel SNPs was at 47.9% and the proportion of novel indels was even higher. Analysing the SNPs associated with LDA among the sequenced samples, a high density of LDA associated polymorphisms was found for a region on BTA13, which was also detected on the Illumina bovine HD beadchip. The DCLRE1C gene in this region contains a potentially damaging SNP with unknown effect. In addition to this SNP, 27 further potentially damaging, coding SNPs were located within or closely neighboring to regions described in the previous studies for LDA. Of all genes containing these SNPs, EGF, SDCCAG8, and PHLPP1 can be regarded as functional candidates for LDA. This study therefore provides SNPs within the coding sequences of functional and positional candidate genes for LDA.

All causal and potentially causal mutations, which have been discovered for LDA, may provide deeper insights into the pathogenesis of LDA and illuminate an important part of the genetic background of this disease. It may also lead on a track to LDA prevention by drug discovery or selection based on a genome-wide LDA-test.



Labmagenverlagerung, Deutsche Holstein Rinder, molekulargenetische Analysen, displacement oft he abdomasum, German Holstein cattle, molecular genetic analyses