If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Clinic for Ruminants with Ambulatory and Herd Health Services, Centre for Clinical Veterinary Medicine, Ludwig-Maximilians-University Munich, 85764 Oberschleißheim, GermanyImmunology Unit, University of Veterinary Medicine Hannover, 30559 Hannover, Germany
Clinic for Ruminants with Ambulatory and Herd Health Services, Centre for Clinical Veterinary Medicine, Ludwig-Maximilians-University Munich, 85764 Oberschleißheim, Germany
Clinic for Ruminants with Ambulatory and Herd Health Services, Centre for Clinical Veterinary Medicine, Ludwig-Maximilians-University Munich, 85764 Oberschleißheim, Germany
Technical University Braunschweig, Institute for Microbiology, 38023 Braunschweig, GermanyHelmholtz Centre for Infection Research, Microbial Proteomics, 38124 Braunschweig, Germany
Leibniz Institute for Farm Animal Biology (FBN), Institute of Genome Biology, 18196 Dummerstorf, GermanyAgricultural and Environmental Faculty, University Rostock, 18059 Rostock, Germany
Infection and inflammation of the mammary gland, and especially prevention of mastitis, are still major challenges for the dairy industry. Different approaches have been tried to reduce the incidence of mastitis. Genetic selection of cows with lower susceptibility to mastitis promises sustainable success in this regard. Bos taurus autosome (BTA) 18, particularly the region between 43 and 59 Mb, harbors quantitative trait loci (QTL) for somatic cell score, a surrogate trait for mastitis susceptibility. Scrutinizing the molecular bases hereof, we challenged udders from half-sib heifers having inherited either favorable paternal haplotypes for somatic cell score (Q) or unfavorable haplotypes (q) with the Staphylococcus aureus pathogen. RNA sequencing was used for an in-depth analysis of challenge-related alterations in the hepatic transcriptome. Liver exerts highly relevant immune functions aside from being the key metabolic organ. Hence, a holistic approach focusing on the liver enabled us to identify challenge-related and genotype-dependent differentially expressed genes and underlying regulatory networks. In response to the S. aureus challenge, we found that heifers with Q haplotypes displayed more activated immune genes and pathways after S. aureus challenge compared with their q half-sibs. Furthermore, we found a significant enrichment of differentially expressed loci in the genomic target region on BTA18, suggesting the existence of a regionally acting regulatory element with effects on a variety of genes in this region.
Rapid progress in molecular biology and genetics has made it possible to include information on genomic variance in modern animal breeding alternatively or complementary to the classical pedigree-based approach (
). Facing the challenges of rising antimicrobial resistances and an increasing awareness of animal welfare issues, this approach is particularly interesting for improving animal health (e.g., in dairy cattle). Genomic selection for reduced mastitis susceptibility is a sustainable option to reduce disease incidence, but it requires profound knowledge on the underlying genetic loci modulating the trait. In a previous study,
described that including genetic marker haplotype information associated with the SCS QTL on BTA18 improved selection for a favorable SCS compared with selection restricted to solely conventional pedigree information. Recently, differences between alternative paternally inherited BTA18 haplotypes in clinical performance were confirmed for first lactating heifers before and after challenge with mastitis pathogens at the early lactation stage (
). These results are in line with a large number of studies reporting a major QTL for functional traits (SCS, mastitis, longevity) in the respective genomic region on BTA18 (
). Hence, information on BTA18 haplotypes in Holstein cattle breeding could contribute to improved animal health and particularly to reduced susceptibility to mastitis in the German Holstein cattle population. However, haplotype effects need to be first characterized more precisely to exclude potential detrimental side effects. Furthermore, a deep phenotypic evaluation of potential BTA18 haplotype effects could also shed light on the precise causal molecular background of the QTL localized in this chromosomal region. The genetic variation underlying this major QTL is still unclear in spite of very powerful analyses and promising candidate genes (
The impact of intramammary Escherichia coli challenge on liver and mammary transcriptome and cross-talk in dairy cows during early lactation using RNAseq.
). Hence, we focused in this study on the hepatic transcriptome using RNA sequencing (RNAseq) to gain comprehensive information about immunological as well as metabolic differences between divergent haplotypes at the transcriptomic level. Therefore, we have looked at the effects of an intramammary challenge with a mastitis pathogen on the hepatic transcriptome of half-sib cows, which had inherited either favorable or unfavorable paternal haplotypes. The challenge was performed with Staphylococcus aureus strain 1027, commonly known as a causative pathogen for subclinical mastitis (
) for all 11,503 German Holstein AI sires born between 1999 and 2012 and recorded in the VIT genome database, we calculated SNP effects for divergent paternal haplotypes in the respective chromosomal regions (Q, meaning favorable for SCS, and q, meaning unfavorable for SCS). We selected those sires with differences of summarized SNP effects of at least 2 standard deviations larger than the mean with the assumption that those sires should be segregating for a Q or q haplotype. We excluded sires with extreme breeding values for SCS and milk performance traits to achieve similar performance levels within half-sib groups. The maternal grandsires were selected for their breeding values for SCS (Relativzuchtwert Somatischer Zellgehalt, above 112 for the Q cohort and below 100 for the q cohort). Moreover, we searched for cohorts with at least 3 potential Q and q half-sib sisters within one sire, respectively, and a maximal calving age of 36 mo. After all these filtering steps, a total of 282 heifers were genotyped with the 50k Illumina SNP chip (Illumina Inc., San Diego, CA) and haplotyped for their inherited paternal haplotypes (Q or q). Finally, 24 healthy, pregnant heifers, which originated from 6 sires, were selected. Within sire, each of the heifers was allocated to the Q or q group according to the inherited SNP haplotypes, which enabled monitoring of alternative paternally inherited haplotypes in the analysis. The selected heifers were purchased from conventional dairy farms across Germany and brought to the Clinic for Cattle at the University of Veterinary Medicine Hannover (
Twenty-four animals (12 Q, 12 q) were challenged in an infection model in the Clinic for Cattle at the University of Veterinary Medicine Hannover essentially as described in detail in
. Briefly, during the challenge experiment, the animals were kept in individual loose stall pens and received a component diet based on grass silage, corn silage, rapeseed extraction meal, soy extraction meal, concentrates, and minerals adjusted to milk performance. In the pre-challenge period, the animals were closely monitored for general health status and specific indicators of mastitis as described in
. The experiment was performed under the reference number 33.12-42502-04-15/2024 with approval by the Lower Saxony Federal State Office for Consumer Protection and Food Safety. Furthermore, this study was approved by the ethics committee of the University of Veterinary Medicine Hannover. All ethical evaluations were performed as required by the German Animal Care law (
Escherichia coli, but not Staphylococcus aureus triggers an early increased expression of factors contributing to the innate immune defense in the udder of the cow.
). Thirty-six ± 3 d after parturition, 24 healthy animals (12 Q, 12 q) were challenged with 10,000 cfu of S. aureus1027 each in both hind quarters of the mammary gland and killed 96 h later. A control udder quarter was infused with sterile sodium chloride solution. Heifers were closely monitored for clinical signs of mastitis [e.g., declining milk yield, local (pain, swelling, redness) or systemic (fever) signs of inflammation] during the postchallenge period as described by
. The animals were stunned with a penetrating captive bolt pistol, immediately followed by exsanguination via longitudinal section of the jugular veins and carotid arteries 96 h after the start of the challenge (
). The time point was selected because it was predicted to be the zenith of mammary gland inflammation according to experience from previous experiments (
Escherichia coli, but not Staphylococcus aureus triggers an early increased expression of factors contributing to the innate immune defense in the udder of the cow.
). During dissection of the animals, liver tissue was collected from the lobus caudatus, immediately shock frozen in liquid nitrogen, and subsequently stored at −80°C. About 6 mo after sampling, the tissue was used for transcriptome analysis as described below.
Transcriptome Analysis by RNA Sequencing
Frozen liver tissue samples (approximately 30 mg) were ground using the Precellys 24 tissue homogenizer with a lysing kit containing 1.4-mm ceramic beads (peQLab, Erlangen, Germany). Total RNA was extracted via an on-column purification following the protocol of the NucleoSpin RNA II kit (Macherey-Nagel, Düren, Germany), with an adapted DNase digestion step as described by
), and a second DNase digestion step was added if required. The RNA concentration and purity were quantified on a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA) and a Qubit 2.0 fluorometer (Thermo Fisher Scientific), and RNA integrity was evaluated on the Bioanalyzer 2100 (Agilent Technologies, Böblingen, Germany). Subsequently, a stranded library preparation protocol for RNA sequencing was applied (TruSeq Stranded mRNA LP, Illumina) with application of indices for multiplexing during cluster generation and polyA-selection to focus on polyadenylated RNA (in the majority mRNA). The RNAseq libraries were checked for quality on the Bioanalyzer 2100. Using the Illumina HiSeq 2500 system (Illumina), we performed paired-end sequencing (2 × 90 bp).
Bioinformatic Analysis
The CASAVA v1.8 (Illumina) software was used for demultiplexing of reads. Scripts written in SAMtools (
). The advantage of this approach is that a guided transcript assembly refers to the reference genome annotation and also enables the identification of transcripts, which have not yet been annotated. Using the StringTie (
) with a threshold for significance of adjusted P (Padj) < 0.05. Ingenuity Pathway Analysis (Qiagen, Hilden, Germany) was used to identify enriched biological pathways and predicted upstream regulators in response to a pathogen challenge.
During the selection process for Q and q heifers, the cholesterol deficiency (CD) defect was detected in the German Holstein population (
). In the end, 4 heterozygous CD carriers (CDC according to the World Holstein Friesian Federation, http://www.whff.info/documentation/genetictraits.php#go1) were included in the design, which were offspring of a single sire. The CDC heifers were distributed evenly across the Q and q groups. The CD carrier status was included as fixed effect in the differential expression data analysis.
RESULTS
RNA Sequencing Statistics
The transcriptome analysis by RNAseq generated 2.6 billion reads (on average 107 million reads per sample). A total of 98% of reads mapped at least once to the reference genome UMD3.1 (
). The data analysis revealed a total of 20,723 loci, which showed an expression level of at least 10 reads in at least 4 samples.
Differential Hepatic Transcriptome Expression in Animals Infected with S. aureus
In the expression analysis comparing S. aureus challenged animals that had either inherited the paternal haplotypes Q (n = 12) or q (n = 12), we found a total of 23 significantly (Padj < 0.05) differentially expressed (DE) loci (see Table 1 and Supplemental File S1; https://doi.org/10.3168/jds.2019-17612). Of these 23 significantly DE loci, 17 were annotated genes (74%), with 12 and 5 genes expressed at a higher and lower level, respectively, in Q compared with q animals.
Table 1All significantly (adjusted P-value, Padj < 0.05) differentially expressed annotated genes in the differential expression analysis between paternally inherited haplotypes Q (favorable for SCS) versus q (unfavorable for SCS) of animals challenged with Staphylococcus aureus
Vesicular trafficking through cortical actin during exocytosis is regulated by the Rab27a effector JFC1/Slp1 and the RhoA-GTPase-activating protein Gem-interacting protein.
Suppression of PLCbeta2 by endotoxin plays a role in the adenosine A(2A) receptor-mediated switch of macrophages from an inflammatory to an angiogenic phenotype.
Within the 17 annotated DE genes between paternally inherited haplotypes Q versus q of S. aureus challenged animals 10 genes are listed, which are involved in immune response [e.g., IGSF11 with known effects on human T cells (
)]. However, the top scorer of DE loci is a yet unannotated transcript on BTA18 located 2 Mb outside the genomic target region for our haplotype selection (see Supplemental File S1). Another locus localized on BTA18 at 48.9 Mb (according the UMD3.1) and still unannotated at the respective position in the current genome annotation ARS-UCD1.2 Ensembl 97 (https://www.ensembl.org/Bos_taurus/Info/Index) showed significant differential expression (see Supplemental File S1). The respective transcript does also not show any sequence homology to annotated genes in human or mouse genomes. Three DE (Padj < 0.05) loci are localized in the genomic region spanning 43 to 59 Mb containing the targeted chromosomal subregions for haplotype selection on BTA18. Within this region, a total of 525 loci are expressed in our data set. Thus, compared with the total number of 23 DE loci across all 20,724 expressed genomic loci in the entire genome there is a significant enrichment (P < 0.002) of DE loci in the BTA18 genomic region 43 to 59 Mb.
Pathway Analysis of DE Genes in Animals Infected with S. aureus
Ingenuity Pathway Analysis of those 17 DE loci, which are annotated in the bovine reference genome assembly (UMD3.1), provided 24 significantly (P < 0.05) enriched canonical pathways when comparing S. aureus challenged animals that had either inherited the paternal haplotypes Q or q (see Table 2 and Supplemental File S2; https://doi.org/10.3168/jds.2019-17612). Among them are the pathways integrin-linked protein kinase (ILK) signaling, IL-8 signaling, phospholipases, and phospholipase C signaling that are known to be directly related to immune response. The GPCR-mediated nutrient sensing in enteroendocrine cells, which was also significantly enriched between the paternal haplotypes Q or q, provides a link between the immune response and the dairy cow's metabolism (
Table 2All significantly (P < 0.05) canonical pathways enriched in the hepatic transcriptome of animals challenged with Staphylococcus aureus with haplotypes Q (favorable for SCS) versus q (unfavorable for SCS) using Ingenuity Pathway Analysis (Qiagen, Hilden, Germany)
Ingenuity canonical pathway
−log10 (P-value)
Phospholipases
2.94E00
Antioxidant action of vitamin C
2.55E00
GPCR-mediated nutrient sensing in enteroendocrine cells
2.52E00
p70S6K signaling
2.41E00
D-myo-inositol-5-phosphate metabolism
2.25E00
Gαq signaling
2.24E00
Dopamine-DARPP32 feedback in cAMP signaling
2.18E00
Heme degradation
2.12E00
ILK signaling
2.11E00
Synaptic long-term depression
2.09E00
Endothelin-1 signaling
2.07E00
IL-8 signaling
2.06E00
Breast cancer regulation by Stathmin1
2.05E00
mTOR signaling
2.01E00
Superpathway of inositol phosphate compounds
1.99E00
Phospholipase C signaling
1.84E00
Choline biosynthesis III
1.78E00
Xenobiotic metabolism signaling
1.77E00
G Protein signaling mediated by tubby
1.63E00
D-myo-inositol (1,4,5)-trisphosphate biosynthesis
1.62E00
Cell cycle regulation by BTG family proteins
1.60E00
Role of CHK proteins in cell cycle checkpoint control
The analysis of the predicted upstream regulators (see Table 3) confirmed modulation of the hepatic immune system in Q heifers compared with the q half-sibs within S. aureus challenged animals. In the list of the top 10 significantly enriched upstream regulators (Table 3), 2 important immune receptors (CXCR3, C-X-C chemokine receptor type 3, and TLR3, toll-like receptor 3) as well as mitogen-activated protein kinase kinase kinase 1 (MAP3K1), a kinase involved in the TNF-α and NFκB signal cascades (
Table 3Top 10 significantly (P < 0.05) enriched upstream regulators in Ingenuity Pathway Analysis (Qiagen, Hilden, Germany) of haplotype Q (favorable for SCS) versus q (unfavorable for SCS) within animals challenged with Staphylococcus aureus (exogenous chemical excluded)
. The comparison of heifers with Q and q haplotypes revealed that the q animals were more susceptible to early-lactation diseases. They had a higher number of udder quarters with very low cell count (<10,000 cells/mL) in the first weeks of lactation, but in the later course of lactation the weekly SCS was significantly lower for Q compared with q animals. The q animals had a higher SCS at 24 and 36 h after S. aureus challenge and a higher shedding of bacteria 12 h after challenge as well as a lower decline in milk yield after pathogen challenge (
). Most interestingly, in addition to differences in clinical parameters there were also significant differences in the endocrine and metabolic profiles of the animals postpartum: those cows with Q haplotypes displayed higher IGF-1, lower growth hormone, and BHB levels in plasma and serum, respectively, compared with q half-sibs (
The hepatic transcriptome is suitable to monitor potential metabolic as well as immunological effects of the BTA18 haplotypes, as the liver is at the center of the metabolic and immunological physiology of the dairy cow (
The impact of intramammary Escherichia coli challenge on liver and mammary transcriptome and cross-talk in dairy cows during early lactation using RNAseq.
). To follow up the differences in IGF-1, growth hormone, and BHB plasma/serum levels between Q and q cows we had observed before the challenge, we looked at potential related hepatic transcriptomic signatures in Q versus q after intramammary S. aureus challenge. However, we did not find significant differences in gene expression for IGF1 or other key genes directly involved in (short chain) fatty acid metabolism (Supplemental File S1; https://doi.org/10.3168/jds.2019-17612). Interestingly, the canonical pathway GPCR-mediated nutrient sensing in enteroendocrine cells was significantly enriched in the hepatic transcriptome of S. aureus challenged animals with the Q haplotype compared with animals with the q haplotype. G-protein-coupled receptors (GPCR) are membrane proteins, which detect their ligands (e.g., hormones, neurotransmitters, chemokines) in the extracellular matrix and can subsequently initiate the respective intracellular signaling cascades (
). Thus, they have a broad spectrum of influence, and the enrichment of the respective pathway in the hepatic transcriptome of S. aureus challenged animals with Q haplotypes compared with the half-sibs carrying a q haplotype indicates haplotype-specific differences regarding their metabolic and immune profiles.
After the intramammary S. aureus challenge, the livers of Q or q cows showed a different hepatic transcriptomic profile with respect to immune genes (see Table 1, Table 2).
The enrichment of the canonical pathways ILK signaling and IL-8 signaling indicates that immune pathways play a dominant role in the response of the divergent hepatic transcriptome response to intramammary S. aureus challenge. This is also indicated by the nominally significant enrichment of the Protein Information Resource category “Innate immunity” (P = 1.1 × 10−3) and also by the significantly DE genes themselves (see Table 1). DYSF, for example, which was originally observed for its role in muscle function, is involved in regulating cellular interactions, cell adhesion mechanisms, and has a function in inflammatory cells (
). The receptor encoded by GPBAR1 is implicated in the suppression of macrophage functions and regulation of energy homeostasis by bile acids and is also essential for the regulation of liver immunity (
). HMOX1 has been recognized as having major immunomodulatory and anti-inflammatory properties and has been regarded as an adaptive cellular response against inflammatory response and oxidative stress (reviewed by
). Furthermore, ICAM3 and IGSF11 are both members of the immunoglobulin super-family. ICAM3 has been described as chemotactic for macrophages and is therefore implicated in the phagocytosis of apoptotic cells (
). TNC, described to be upregulated in inflamed tissues, is an activator of innate immunity, which can stimulate the synthesis of inflammatory cytokines (
). The PLCB2 gene, also significantly differentially expressed, is part of the significantly enriched canonical pathways phospholipases and phospholipase C signaling. As a key regulator for macrophage function, PLCB2 is involved in the switch from an inflammatory (M1) to an angiogenic (M2-like) macrophage phenotype (
Suppression of PLCbeta2 by endotoxin plays a role in the adenosine A(2A) receptor-mediated switch of macrophages from an inflammatory to an angiogenic phenotype.
). As presented in Figure 1, the Ingenuity network analysis showed that DYSF, GPBAR1, HMOX1, FERMT2, ICAM3, TNC, and PLCB2 are interacting with each other in a network related to immune response, indicating an interrelationship between them.
Figure 1Highly enriched network in the differentially expressed gene analysis of Q (favorable paternal haplotypes for SCS) versus q (unfavorable haplotypes for SCS) in cows challenged with Staphylococcus aureus. Green = lower expression in Q cows compared with q cows, red = higher expression in Q compared with q cows (modified according to Ingenuity Pathway Analysis, Qiagen, Hilden, Germany).
As potential transcriptional regulators CXCR3, TLR3, and MAP3K1 could be predicted (Table 3). The CXCR3 is expressed by monocytes, T cells, natural killer cells, dendritic cells, and cancer cells (
). After activation by its selective ligands CXCL9, CXCL10, and CXCL11, it is involved in the recruitment and clustering of T cells and natural killer cells (
) and is therefore eminent for the appropriate response of the immune system to invasive pathogens. Thus, the upstream analysis confirmed alteration of the hepatic immune system in Q heifers compared with their q half-sibs within S. aureus-challenged animals.
Regarding a potential molecular background of the QTL for functional traits and udder health in particular, we found a significant enrichment of DE loci in the genomic target area of the different haplotypes and their close vicinity on BTA18. This suggests that a regionally acting regulatory element might be involved, which could modulate the expression of an array of genes located there. Regulatory variation in noncoding regions as background of the BTA18 telomeric functional QTL might also explain the lack of causal candidate variant evidence for this QTL. No causal mechanism for the QTL could be pinpointed despite genetic variants with predicted effects in coding and noncoding regions, as described in recent GWAS at the whole genome sequence level with powerful data sets (
). This also underlines the need to improve the current annotation of the bovine genome in terms of functional regulatory elements as addressed by the global Functional Annotation of Animal Genomes action (https://www.animalgenome.org/community/FAANG/;
Heifers with alternative paternal BTA18 haplotypes display a significantly different hepatic transcriptome upon intramammary S. aureus challenge. Animals with the favorable Q haplotypes showed a more activated immune system (as demonstrated by the higher activation of key pathways and divergent expression of relevant immune genes), which might be better able to defend the host during a bacterial challenge. Driver of these differences in the immune system might be different adaptations of metabolic pathways of the Q and q animals. These findings are in line with results from clinical data of the haplotypes, which showed Q cows to be more resistant to diseases, especially S. aureus mastitis (SCC, SCS, shedding of bacteria, milk yield) than their q half-sibs.
ACKNOWLEDGMENTS
The authors thank the laboratory staff at the FBN Dummerstorf as well as the PhD students of the ChronMast project at the University of Veterinary Medicine Hannover. We thankfully acknowledge Hans-Martin Seyfert (Dummerstorf, Germany) for very fruitful discussion. We also thank the Förderverein Bioökonomieforschung (Bonn, Germany) for continuous support. This work was supported by funds of the German Government's Special Purpose Fund held at Landwirtschaftliche Rentenbank (Frankfurt/Main, Germany). The authors have not stated any conflicts of interest.
Suppression of PLCbeta2 by endotoxin plays a role in the adenosine A(2A) receptor-mediated switch of macrophages from an inflammatory to an angiogenic phenotype.
Vesicular trafficking through cortical actin during exocytosis is regulated by the Rab27a effector JFC1/Slp1 and the RhoA-GTPase-activating protein Gem-interacting protein.
The impact of intramammary Escherichia coli challenge on liver and mammary transcriptome and cross-talk in dairy cows during early lactation using RNAseq.
Escherichia coli, but not Staphylococcus aureus triggers an early increased expression of factors contributing to the innate immune defense in the udder of the cow.
30274-5&id=gr1.jpg){kind=link}