Genome-wide pleiotropy and shared biological pathways for resistance to bovine pathogens

M. Mahmoud; Y. Zeng; M. Shirali; T. Yin; K. Brügemann; S. König; C. Haley

doi:10.1371/journal.pone.0194374

Abstract

Host genetic architecture is a major factor in resistance to pathogens and parasites. The collection and analysis of sufficient data on both disease resistance and host genetics has, however, been a major obstacle to dissection the genetics of resistance to single or multiple pathogens. A severe challenge in the estimation of heritabilities and genetic correlations from pedigree-based studies has been the confounding effects of the common environment shared among relatives which are difficult to model in pedigree analyses, especially for health traits with low incidence rates. To circumvent this problem we used genome-wide single-nucleotide polymorphism data and implemented the Genomic-Restricted Maximum Likelihood (G-REML) method to estimate the heritabilities and genetic correlations for resistance to 23 different infectious pathogens in calves and cows in populations undergoing natural pathogen challenge. Furthermore, we conducted gene-based analysis and generalized gene-set analysis to understand the biological background of resistance to infectious diseases. The results showed relatively higher heritabilities of resistance in calves than in cows and significant pleiotropy (both positive and negative) among some calf and cow resistance traits. We also found significant pleiotropy between resistance and performance in both calves and cows. Finally, we confirmed the role of the B-lymphocyte pathway as one of the most important biological pathways associated with resistance to all pathogens. These results both illustrate the potential power of these approaches to illuminate the genetics of pathogen resistance in cattle and provide foundational information for future genomic selection aimed at improving the overall production fitness of cattle.

Citation: Mahmoud M, Zeng Y, Shirali M, Yin T, Brügemann K, König S, et al. (2018) Genome-wide pleiotropy and shared biological pathways for resistance to bovine pathogens. PLoS ONE 13(4): e0194374. https://doi.org/10.1371/journal.pone.0194374

Editor: Tosso Leeb, University of Bern, SWITZERLAND

Received: November 16, 2017; Accepted: March 1, 2018; Published: April 2, 2018

Copyright: © 2018 Mahmoud et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are available from Edinburgh Datashare (https://datashare.is.ed.ac.uk/handle/10283/3040) with DOI 10.7488/ds/2318.

Funding: This research funded by the German Federal Ministry of Education and Research (BMBF), the Förderverein Bioökonomieforschung e.V. (FBV) / German Holstein Association (DHV) and through the collaborative project “KMU-innovativ-10: Kuh-L – cow calibration groups for the implementation of selection strategies based on high-density genotyping in dairy cattle” (grant no. 031A416C). CSH was supported by the Biotechnology and Biological Sciences Research Council in the UK (Grant BB/P013759/1). CSH and YZ were supported by the Medical Research Council in the UK (Grant MC_PC_U127592696). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Infection is the colonization of the host by at least one domain of pathogens such as viruses, bacteria, fungi, and/or parasites (parasite infestation). Each type of pathogen is different in its invasion and replication in the host tissue, namely “infectivity” and/or the capacity to pass from one individual to another, namely “transmissibility” [1, 2]. Although variation in resistance to individual infectious agents (pathogens) is associated with levels of immunisation, disease treatment policies, diet and other environmental factors, previous studies of the resistance to various pathogens in animals [3] and humans [4] have also revealed a major role of host genetic factors in pathogen resistance and host survival.

Host genetic architecture in cattle has been found to be a major factor in resistance to multifactorial diseases and disorders such as infertility, metabolic disorder, claw disorder and mastitis [5]. Several studies have addressed the importance of the genetic contribution of host resistance/susceptibility to different domains, species, and subspecies of pathogens [6, 7]. The most studied pathogen subspecies in the field of dairy cattle breeding (particularly in Europe) are Mycobacterium avium subspecies Tuberculosis [8–14], and Mycobacterium avium subspecies Paratuberculosis [15–18]. Another important pathogen affecting calves is Salmonella typhimurium. Wray and Sojka [19] reported some phenotypic variation in resistance to Salmonella between Jersey calves and Friesian calves, which may be due to the genetic variation among breeds. Templeton et al., [20] noted that calves of sires with a high resistance to Brucellosis also show a high resistance to Salmonella, suggesting a genetic contribution to the resistance to this pathogen. In studies of the genetics of resistance to viral pathogens, resistance has mostly been measured indirectly as the symptoms of infection in animals, rather than directly measured as susceptibility to the pathogen itself. For instance, in 2008, Heringstad et al., [21] estimated the heritability of susceptibility to respiratory diseases to be 0.05±0.018. Estimation of genetic variation underlying the resistance/susceptibility for parasitic infestation has been well studied in small ruminants [22], and to a small extent in cattle [23]. In Canadian Holstein, the heritability for susceptibility to Neospora caninum was in a range between 0.08±0.02 and 0.12±0.04 [24]. In Dutch Holstein-Friesian, the heritability of eggs/larvae count in animal feces was estimated to be from 0.00±0.02 to 0.25±0.05 [25]. The resistance to ectoparasites in cattle was studied and reviewed by [26] who concluded that the heritability of resistance to ticks was 0.31, and 0.21 for resistance to Buffalo flies.

A number of studies have also detected quantitative trait loci (QTLs) associated with susceptibility/resistance to infection diseases in farm animals. Lee et al., [27] found 11 QTLs on three chromosomes (BTA15, BTA17, and BTA22), significantly associated with susceptibility to Foot-and-mouth disease in Holstein cattle. Casas et al., [28] searched for markers directly associated with the susceptibility to infection with Bovine viral diarrhea virus in feedlot cattle, and found a significant association on Chromosome BTA14.

Despite estimates of the heritability of resistance for some pathogens, genetic correlations (pleiotropy) among resistance between these domains of pathogens are still unknown, as is the level of pleiotropy among resistance to different species and subspecies within each domain of these pathogens. Several reasons lie behind the absence of the genetic correlations among resistance/susceptibility to various pathogens. First, the most popular method used in animal breeding for estimating genetic correlation has been the pedigree-base restricted maximum likelihood approach (Pedigree-REML). This method tests genetic overlap of traits between related individuals within pedigrees. To quantify the genetic correlation between traits in a family-based study, we may need to measure the traits in individuals with pedigree relationships [29]. Consequently, it will be challenging and costly to repeat measurements on all animals, in particular for these low-prevalence traits and/or for traits where slaughter of the animals to measure their resistance is needed (i.e. endo-parasite infestation). Moreover, some disease traits (i.e. Bovine respiratory diseases) may result in death of the animal (at a young age) before it is possible to measure other traits (such as milk production after first calving) for which we want to test the correlation.

The genomics era has provided a solution to the problem of estimating genetic correlations, by allowing the genetic correlation to be estimated using genomic variants (i.e. SNPs) instead of using pedigree information, providing more precise and accurate estimates for the narrow-sense heritability () in case-control studies and for quantitative traits as well as for the coheritability between such traits, which in particular does not need the measurement of multiple traits per animal. So far, several different methods have been developed for estimating the (co)heritability using genomic data: the first method uses the significantly associated SNPs found in large GWAS studies to estimate the causal relationships between risk factors and disease. This method is efficient only in case of the traits with many significant SNPs, which is usually not the case in resistance/susceptibility traits. For complex traits (especially for those recorded in case control studies), it is recommended to use genome-wide data (array or sequence data) instead of using only significant markers to estimate genetic correlation [30], and this approach has been implemented in two published methods; Genomic-restricted maximum likelihood (G-REML) [31, 32] and polygenic scores [33, 34]. One of the limitations for the application of the last two methods in complex traits is the availability of individual-level genotype data. Hence, Bulik-Sullivan and colleagues [29, 35] developed LD Score regression using the GWAS summary statistics instead of the individual-level genotype data for estimating heritabilities and genetic correlations.

In this study, we use individual-level genotypes for ~ 20000 animals from cow calibration groups (NB Cow calibration groups: is a programme initiated in Germany combining information for novel traits with high-density genetic markers based on ~20,000 genotyped cows, to offer a new perspective on breeding for improved disease resistance [36]). We applied the G-REML method to estimate the heritability of the 23 resistance traits and the genetic correlations between these traits, and the correlations between the resistance traits with calf performance and cow productivity. We also implemented a post-GWAS functional analysis to estimate the pleiotropy based on different scales; scale 1) pleiotropy among all 23 resistance traits based on gene analysis; scale 2) pleiotropy among all 23 traits based on gene-set analysis (biological pathways), to understand the biological background of the underlying resistance/susceptibility to infectious diseases.

Results

SNP heritabilities

Using a univariate G-REML model, we estimated the SNP-based heritabilities (h²_SNP) for all resistance, performance, and productivity traits (Table 1). SNP heritabilities of resistance to bacterial pathogens ranged from 0.03±0.01 to 0.21±0.01 in calves, and from 0.02±0.01 to 0.13±0.02 in cows. Higher SNP heritabilities were estimated for the resistance to viral pathogens, ranging from 0.16±0.03 to 0.22±0.03 in calves, and 0.10±0.01 in cows. For resistance to trichophyton (the only fungal pathogen) in calves, the heritability estimate was 0.17±0.03, and in cows it was 0.04±0.01. In parasitic infestation, the SNP heritability estimates in calves ranged from 0.04±0.01 to 0.14±0.03, while in cow’s estimates varied from 0.19±0.02 to 0.25±0.02. For calf performance traits, the SNP heritability explained 0.30±0.01 of the phenotypic variance in birthweight, while 0.08±0.02 of the phenotypic variance in average daily gain was explained. For cow productivity traits, SNPs explained 0.19±0.02 of the variance in average milk yield, and 0.25±0.02 of the variance for the fat to protein ratio trait.

Download:

Table 1. Number of genotyped samples and estimated SNP-based heritabilities (h²_SNP) for resistance and performance traits in calves; and resistance and productivity traits in cows.

https://doi.org/10.1371/journal.pone.0194374.t001

Pleiotropy among calf resistance traits

A bivariate analysis was used to estimate the level of pleiotropy among calf resistance traits using genome-wide SNPs for all 36 pairwise combinations of the nine resistance traits, shown in Fig 1 and S1 Table. The genetic correlation was significantly different from zero (according to FDR <1% and according to the Bonferroni threshold) between the resistance to Salmonella pathogen and Trichophyton pathogen (0.55±0.07), between Salmonella and Cryptosporidium (0.98±0.01), between Bovine respiratory syn. and Coccidia (0.46±0.06), between Bovine herpesvirus 1 and Bovicola bovis (0.74±0.18), between Cryptosporidium and Coccidia (0.52±0.06), and between Myiasis and Bovicola bovis (-0.54±0.06).

Download:

Fig 1. Genetic correlations among the 9 calf resistance traits analyzed by G-REML.

Blue, positive genetic correlation; red, negative genetic correlation. Larger squares correspond to more significant P values. Genetic correlations that are different from zero at a false discovery rate (FDR) of 1% are shown as full-sized squares. Genetic correlations that are significantly different from zero after Bonferroni correction for the 36 tests in this analysis are marked with a yellow asterisk. We show results that do not pass multiple-testing correction as smaller squares. All genetic correlations in this report can be found in tabular form in S1 Table.

https://doi.org/10.1371/journal.pone.0194374.g001

The pleiotropy among cow resistance traits

A second bivariate analysis was used to estimate the pleiotropy among cow resistance traits using genome-wide SNPs for all 105 pairwise combinations of the 14 resistance traits (all results are shown in Fig 2 and S2 Table). The estimated genetic correlation was significantly different from zero (according to FDR <1% and according to the Bonferroni threshold) between the resistance to Escherichia coli pathogen and Staph. Aureus pathogen (0.71±0.15), between Staph. Aureus and Strep. Uberis (0.56±0.10), between Staph. Haemolyticus and Strep. Dysgalactiae (1.00±0.34), between Staph. Haemolyticus and Chorioptic scabies (1.00±0.11), between Strep. Agalactiae and Strep. Dysgalactiae (1.00±0.28), between Strep. Agalactiae and Bovicola bovis (1.00±0.36), and between Rotavirus and Chorioptic scabies (-0.52±0.16). The estimated genetic correlation was significantly different from zero (with FDR <2%) between the resistance to Escherichia coli and Strep. Dysgalactiae (0.78±0.22), Strep. Agalactiae and Strep. Uberis (0.72±0.27), Strep. Agalactia and Trichophyton (0.96±0.33), and Rotavirus and Dictyocaulus viviparus (0.44±0.13).

Download:

Fig 2. Genetic correlations among the 14 cow resistance traits analyzed by G-REML.

Blue, positive genetic correlation; red, negative genetic correlation. Larger squares correspond to more significant P values. Genetic correlations that are different from zero at a false discovery rate (FDR) of 1% are shown as full-sized squares. Genetic correlations that are significantly different from zero after Bonferroni correction for the 105 tests in this analysis are marked with a yellow asterisk. We show results that do not pass multiple-testing correction as smaller squares. All genetic correlations in this report can be found in tabular form in S2 Table.

https://doi.org/10.1371/journal.pone.0194374.g002

The genetic correlation between calf and cow resistance traits was significantly different from zero for two different traits (S3 Table). The estimated genetic correlation between resistance to the Salmonella pathogen in calves and cows was -0.26±0.09 (p-value ≤ 0.001), and the genetic correlation between resistance to the Trichophyton pathogen in calves and cows was 0.18±0.12 (p-value ≤ 0.05).

Pathogen resistance and calf performance

The genetic correlations of birthweight with the resistance to viral, fungal, and parasitic (except for Cryptosporidium) pathogens were highly significantly negative for resistance to Bovine herpes virus 1 (P ≤ 0.01), and also significantly negative (P ≤ 0.05) for resistance to the Bovine respiratory syn., Trichophyton, Coccidia and Bovicola bovis pathogens. There were no significant positive correlations between birthweight and resistance, although that with Myiasis approached significance (P ≤ 0.1). The genetic correlation of average growth rate was very highly significantly (P ≤ 0.001) negative with resistance to Salmonella, highly significantly negative with resistance to the Escherichia coli and Bovine respiratory syn. Pathogens, and negative and approaching significance (P ≤ 0.1) with resistance to Bovine herpes virus 1 and Trichophyton pathogens. Positive correlations with average daily gain were only significant for Cryptosporidium (P ≤ 0.01) and approaching significance for Coccidia (P ≤ 0.1) (Fig 3). All correlations in calf performance traits had standard errors that ranged from 0.05 to 0.14 (S4 Table).

Download:

Fig 3. Estimated genetic correlations of birthweight (BW) and average daily gain (ADG) with all resistance traits in calves.

This plot compares the genetic correlation between BW and all calf resistance traits with the genetic correlation between ADG and all calf resistance traits obtained from G-REML. The horizontal axis indicates pairs of phenotypes (BW and ADG), and the vertical axis indicates genetic correlation. Error bars represent standard errors. ‘***’P ≤ 0.001; ‘**’P ≤ 0.01; ‘*’P ≤ 0.05; ‘·’P ≤ 0.1.

https://doi.org/10.1371/journal.pone.0194374.g003

Pathogen resistance and cow productivity

(Fig 4 and S5 Table). The resistance to bacterial pathogens were generally positively correlated with milk yield, the correlation being very highly significance positive (P ≤ 0.001) with Staph. Aureus, highly significantly positive (P ≤ 0.01) with Strep. Uberis and approaching significantly positive (P ≤ 0.1) with Staph. Haemolyticus and Strep. Agalactiae. Milk yield was also significantly positively correlated with resistance to Rotavirus. Resistance to Trichophyton was approaching significance for a negative correlation with milk yield. Fat to protein ratio was positively correlated with resistance Staph. Aureus (P ≤ 0.01), Rotavirus and Trichophyton (P ≤ 0.05). Fat to protein ratio was significantly negatively correlated with Salmonella, Mycobac. Paratuberculosis and Chorioptic scabies (P ≤ 0.05), and approaching significance for Staph. Haemolyticus (P ≤ 0.1).

Download:

Fig 4. Estimated genetic correlations of average milk yield (AMY) and fat to protein ratio (FTP) with all resistance traits in cows.

This plot compares the genetic correlation between AMY and all calf resistance traits with the genetic correlation between FTP and all calf resistance traits obtained from G-REML. The horizontal axis indicates pairs of phenotypes (AMY and FTP), and the vertical axis indicates genetic correlation. Error bars represent standard errors. ‘***’P ≤ 0.001; ‘**’P ≤ 0.01; ‘*’P ≤ 0.05; ‘·’P ≤ 0.1.

https://doi.org/10.1371/journal.pone.0194374.g004

Biological pathway analysis

All SNPs (~50K SNPs) were annotated to the nearest gene where possible using a gene boundary extended by 20kb distance outside the transcription start site or transcription end site of the gene for all the 23 resistance traits, using MAGMA (version 1.06) and bovine gene location (UMD3.1) through ensemble-biomaRt (www.ensembl.org/biomart). This annotation pipeline resulted in a total of 16,094 genes ready for the next step of the analysis.

After testing the association of the 16,094 genes across all the 23 resistance traits, using p-values of summary statistics of GWAS from GCTA, we selected the top (most significant) 20 genes based on average p-values. The lowest average p-value for the correlation between the genes and all resistance traits was the average p-value for the RRM2B gene (average p-value = 0.28) (Fig 5 and S6 Table). After estimating the p-values on the scale of genesets (Biological pathways), we selected the top 20 genesets based on average p-values (Fig 6 and S7 Table) out of 1083 pathways that were available from BIOCARTA, KEGG and REACTOME databases. We found that “Reactome pre-notch transcription and translation” and “Biocarta B-lymphocyte pathway” were the most associated pathways to all resistance traits according to their calculated p-values (the average p-values = 0.07 and 0.11 respectively) (Fig 6 and S7 Table). The final step was to illustrate the “Biocarta B-lymphocyte pathway” against the 19 genes that were associated with resistance at the gene-level. Fig 7 shows how the candidate genes (from gene analysis) were directly and indirectly connected to the “Biocarta B-lymphocyte pathway.”

Download:

Fig 5. P-values for the selected annotated candidate genes in all resistance traits in calves and cows.

Dark red color means very high p-value, dark blue color means very low (i.e. more significant) p-value.

https://doi.org/10.1371/journal.pone.0194374.g005

Download:

Fig 6. P-values for the selected 20 pathways tested across all resistance traits in calves and cows.

Dark red color means very high p-value, dark blue color means very low (i.e. more significant) p-value.

https://doi.org/10.1371/journal.pone.0194374.g006

Download:

Fig 7. Network of gene-interactions the candidate genes (outside the red circle) and the “Biocarta B-lymphocyte pathway” (inside of the red circle).

https://doi.org/10.1371/journal.pone.0194374.g007

Discussion

We implemented G-REML using genome-wide SNP data to explore the genetic etiology of resistance to pathogens in cattle. This study thus provides the first genomic overview of the genetics of resistance across a range of pathogens together with the genetic correlations between pathogens and with production traits for Holstein cattle, the world’s predominant dairy breed. We estimated the phenotypic variance explained by all SNPs () for resistance, performance and productivity traits, and calculated the genetic correlation (r_g) between them, using genome-wide SNPs in calves and cows. The average heritability of resistance traits at a young age (calf traits) was larger than the average heritability of the same population at a later age (cow traits), reflecting the increase of the environmental effect potentially including increased pathogen exposure after first calving in cattle. The SNP heritabilities estimated using GREML in this study were of a similar scale to those for which estimates for individual pathogens have been previously published using pedigree and genome-wide data [6, 7].

Pleiotropy between resistance to pathogens and calf performance and cow productivity

There have only been very limited previous estimates of genetic correlations among resistance to different pathogens with which our estimates can be compared (e.g. [37]). We thus report a number of new findings that would be difficult to obtain from pedigree studies, including some very high estimates of positive pleiotropy between the resistance to bacterial, fungal and parasite pathogens. Negative pleiotropy was found between the resistance to most of the bacterial, viral and fungal pathogens with both performance traits that we analysed in calves (birthweight and average daily gain). Results supports the hypothesis that some resistance genes may negatively impact performance traits in young calves, perhaps reflecting a balance between energy expenditure on disease resistance and growth. However, clear positive pleiotropy found between resistance to bacterial, viral and fungal pathogens with average milk yield in cows supports the hypothesis that cows that are genetically less vulnerable to infections can produce more milk. This provides economic in addition to welfare justification for increasing focus of breeding objectives on these disease resistance traits.

Shared biological pathways

To date, few studies have implemented biological pathway analysis in animal health [38, 39]. This is the first study post-GWAS for resistance to most of all infection pathogens in dairy cattle. We found a group of 20 genes shared effects across all resistance traits and showed that “Reactome pre-notch transcription and translation” and “Biocarta B-lymphocyte pathway” are the most consistently associated pathways with resistance to different pathogens. The genetic correlations that we observed in this study show a clear pleiotropy (by the means of similar resistance mechanism against the 23 pathogens). The combination of accurate recording of multiple diseases and associated pathogens combined with associated genomic data is unusual and possibly so far unique in a mammalian species including humans. However, given the abundance of genomic data in humans combined with GWAS for a number of individual infectious diseases analyses such as performed here should become increasingly possible for human populations.

The scale of this dataset facilitated by the thorough electronic data collection within the dairy cattle test-herd system in north-eastern Germany has enabled us to obtain unique insight into the genetics of resistance across a range of pathogens. The corollary of these currently unique data is that the overall multivariate pattern will need to await further data collection from this or other similar programs before it is possible to replicate our results as a whole. Nonetheless, all significant correlations have a low standard error (S3, S4 and S5 Tables), and our estimates of heritabilities and pairwise genetic correlations are generally consistent with others in the literature where these are available (e.g. [6, 7, 37, 40]).

The biological impact of the genetic pleiotropy on breeding and selection strategies

The estimated heritabilities and the pattern of genetic correlations between pathogens and with production traits provide valuable information allowing the further optimisation of cattle breeding programmes. For example, several low and non-significant genetic correlations were found among calf resistance traits as well as among cow resistance traits, questioning the traditional hypothesis of selection for mastitis resistance based on somatic cell count as a consequence of multiple pathogen infection in cattle [41] or in sheep [42]. Additionally, the highly significant negative correlation between resistance to the Salmonella pathogen in calves and cows implies that resistance traits in calves are not good indicators or early predictors for the resistance traits and genetic improvement of cow health after calving in first parity, supporting a previous pedigree and GWA study [40]. The estimated zero genetic correlation between resistance to the Trichophyton pathogen in calves and cows indicates that, for this pathogen, calf and cow resistance, are genetically distinct. Thus, overall it is clear that to design an efficient breeding program, we need to take into account these results and utilise a programme combining selection in both cows and calves and consider how to most effectively collect and incorporate information on resistance and susceptibility to multiple pathogens. Fortunately, the increasing availability of genomic data in cattle combined with collection of data such as analysed here ultimately will facilitate genomic selection programmes that meet these objectives.

In summary, despite some limitations of available health traits and genotyped data, the use of G-REML method to estimate the genetic correlation among health traits and between health trait and performance and productivity traits promises to be a very valuable tool in the genetic improvement of animal health. Biological pathway analysis appears to be a very useful tool also, but at present we have had to use information for other species as have no biological pathways specifically tested and verified for cattle and the development of such databases would provide an invaluable resource for future research.

Materials and methods

The resistance to infectious disease traits in Holstein cattle was measured through the framework of the dairy cattle test-herd system of northeast Germany, including the federal states of Mecklenburg-Westpommerania and Berlin-Brandenburg. Dairy cattle farmers and veterinarians used electronic recording systems, which were based on the diagnosis key as developed by Feucker and Staufenbiel [43]. This diagnosis key was also considered when developing the International Committee for Animal Recording (ICAR) [44].

Phenotypes

In calves and cows, four main domains of pathogens were used to classify 23 resistance traits: (a) bacterial pathogens, (b) viral pathogens, (c) fungal pathogens, and (d) parasitic pathogens. We distinguished between calf resistance traits and cow resistance traits [40]. For the traits recorded in calves, we defined a time window from birth to the age of 150 days. For the traits recorded in cows, the window was from 20 days before first calving up to 365 days after first calving (a 385-day period). At least one entry for the respective pathogen implied a score = 0 for infected (non-resistant); otherwise, score = 1 for non-infected (resistant). The infected animal with a given pathogen was the animal (calf/cow) that was recorded as infected with this pathogen during its calf/cow age. A non-infected (resistant) calf to a given pathogen was defined as a calf that was found to be healthy on a farm infected with the given pathogen and that was born after the first record of this infection on that farm; a resistant cow to a given pathogen is defined as a cow that was found to be healthy in a farm where the first record of infection with the given pathogen was at least 20 days before its first calving date. Note that in common with other studies of natural infection in livestock and other species, we use resistance to define animals that did not become infected in a herd that was undergoing a specific disease challenge. For a given pathogen, all herds that showed no occurrence of infection were excluded because we do not know whether these herds were challenged with the particular pathogen.

All veterinary diagnosis and infection pathogen recording was done according to the ICAR (S8 Table, also available online through: www.icar.org). The nine pathogen resistance traits that were recorded in calves were Salmonella, Escherichia coli, Bovine respiratory syn., Bovine herpesvirus 1, Trichophyton, Cryptosporidium, Coccidia, Myiasis and Bovicola bovis. The 14 pathogen resistance traits that were recorded in cows were Salmonella, Escherichia coli, Staphylococcus aureus (Staph. Aureus), Staphylococcus haemolyticus (Staph. Haemolyticus), Streptococcus agalactiae (Strep. Agalactiae), Streptococcus dysgalactiae (Strep. Dysgalactiae), Streptococcus uberis (Strep. Uberis), Clostridium perfringens (Clost. Perfringens), Mycobacterium avium paratuberculosis (Mycobac. Paratuberculosis), Rotavirus, Trichophyton, Dictyocaulus viviparous, Bovicola bovis and Chorioptic scabies. For testing the genetic correlation between the resistance and performance in calves we considered two performance traits in calves: birthweight (in kg) and average daily gain (in g/day) during the first 360 days of calf life. For testing the genetic correlation between the resistance and productivity in cows, we considered two productivity traits in cows: average milk yield (in kg) during the first lactation and fat to protein ratio (in %) during the first lactation.

Genotypes

Genotyping was performed using the Illumina Bovine 50K SNP-BeadChip V2 (Illumina Inc., San Diego, CA), and with the Illumina Bovine Eurogenomics 10K low-density chip. The low-density genotypes (10K) were imputed by Vereinigte Information system Tierhalung (Verden, Germany) to the 50K panel applying the algorithm by Segelke et al. [45]. In post-imputation SNP quality checks; animals with almost identical SNP genotypes (>95% congruency across all SNPs) were eliminated from the ongoing analyses; SNPs with minor allele frequency <0.01 and SNPs showing a significant (P < 10⁻⁵) deviation from Hardy-Weinberg equilibrium were discarded. All SNPs had a genotype call rate greater than 95%.

Genotyping was only undertaken for infected and resistant animals from populations where we could be sure that all animals genotyped have been challenged with the relevant pathogen (i.e. that they come from an infected population). Hence, the incidences from these genotyped samples do not reflect the actual incidences in the German Holstein population, but are likely to be higher as we have excluded data from herds where there is no evidence for a disease challenge for a particular pathogen. Further epidemic research, it is required to study the true incidence at the population scale. For a full statistical descriptive of the 23 resistance traits and the four quantitative traits in calves and cows in the genotyped sample, see Table 1.

Statistical models

Correction for population stratification in genome-wide data

The most common method for dealing with population stratification is principal component analysis (PCA) [46, 47]. Fitting the leading principal components in the model can correct for stratification for analyses such as estimating the proportion of variance explained by genome-wide SNPs and for genome-wide association studies (GWAS). Here, we applied principal components analysis to the genome-wide SNP data to infer continuous axes of genetic variation. Hence, the new axes will reduce the data dimensions (eigenvectors), describing as much variability as possible (eigenvalues): V⁻¹(c^tc)V = D, where V was the matrix of eigenvectors which diagonalizes the covariance matrix c^tc (covariance matrix of genotyped data), D was the diagonal matrix of eigenvalues of c^tc. Then, we adjusted the phenotypes by including the first five eigenvectors as covariates in the model when estimating the proportion of variance explained by all the SNPs, or in G-REML and GWAS (see model-1).

A univariate mixed linear model was used to estimate the phenotypic variance explained by all autosomal SNPs () by applying the genomic-restricted maximum likelihood analysis (G-REML), and using the GCTA software [48]. In matrix notation, the model was defined as: (1) where y refered to the vector of the quantitative trait (for performance and productivity traits) or of unobserved liabilities (for resistance traits); b was the vector of the fixed effects (herd, birth year and birth month for calf traits; herd, calving year, calving season, age at first calving for cow traits and the first 5 PCs), X was an incidence matrix for the fixed effects; g_G was the vector of aggregated effects of all autosomal SNPs with and A_G was the genomic relationship matrix (GRM). The heritability explained by all autosomal SNPs () was defined as where was the phenotypic variance.

A bivariate model was used to estimate the genetic correlation among resistance traits and between resistance and performance or productivity traits [48]. (2) The two equations are the same as in model 1, while the (co)variance matrix was: (3) where A is GRM, I is the identity matrix, is the genetic variance, is the residual variance and is the genetic covariance between the two traits. We assumed that all environmental correlations between pathogens were zero, as less than 10% of animals shared diagnoses for any pair of pathogens.

Significance thresholds

To determine the significance of the estimated pleiotropy among calf and cow resistance traits, two methods are commonly used to determine the significance threshold for genome-wide analysis: the false discovery rate and Bonferroni correction. The False discovery rate (FDR) correction was introduced by Benjamini and Hochberg [49]. The FDR method first ranks all p-values from the smallest to the largest, and then adjusts each p-value accordingly: For the pleiotropy among resistance traits, we used the FDR of 1%. Bonferroni correction bases on the number of independent tests performed in each scenario (36 tests in calf traits and 105 tests in cow traits). The Bonferroni threshold was used at α = 1%, and calculated as follows:

Biological pathway analysis

Biological pathway analysis is an approach where the association between a select set of genes (biological pathways) and a trait of interest (the resistance to different pathogens) was tested. This analysis can be used to test the cumulative genetic effects across multiple genes within a pathway.

Multi-marker Analysis of GenoMic Annotation (MAGMA)

MAGMA used in our pathway analysis according to Leeuw et al., [50] with the following three steps: First, an annotation step to map SNPs onto genes using the bovine gene location (UMD3.1), from ensemble-biomaRt (www.ensembl.org/biomart). Second: a gene-based analysis step to compute gene p-values, using MAGMA and the output of GWAS from GCTA (pre-estimation of summary statistics of GWAS for each trait, performed using the GCTA software [48]. Third: a gene-set (Biological pathways) analysis step, to compute biological pathway p-values, using MAGMA (with the “competitive gene-set analysis” function) with publicly available BIOCARTA, KEGG and REACTOME database. All analyses in MAGMA are structured as a linear regression model on gene-level data. (Model – 4) Where Z was the phenotype vector, Gene-sets G₁ were binary indicator variables, coded with “1” for genes in the gene-set, and with “0” otherwise. e was the residual vector. The intercept β₀ represents the mean, and β₁ the association specific to the gene-set 1. One last step was to illustrate the most significant gene rich network related to most of the resistance traits in calf and cows. This was done using the web-based software GeneMANIA [50], and then to illustrate the most significant related genes to all resistance traits (both in the same Fig) to see how these genes are related to this pathway. The post-GWAS functional analyses were performed using the MAGMA software [50] and the GWAS output from GCTA software [48].

Supporting information

S1 Table. (3 sheets).

correlations: Genetic correlations among the 9 calf traits analyzed by G-REML
p-value: p-values for genetic correlations among the 9 calf traits analyzed by G-REML
se: Standard error for genetic correlations among the 9 calf traits analyzed by G-REML.

https://doi.org/10.1371/journal.pone.0194374.s001

(XLSX)

S2 Table. (3 sheets).

correlations: Genetic correlations among the 14 cow traits analyzed by G-REML
p-value: p-values for Genetic correlations among the 14 cow traits analyzed by G-REML
se: Standard error for Genetic correlations among the 14 cow traits analyzed by G-REML.

https://doi.org/10.1371/journal.pone.0194374.s002

(XLSX)

S3 Table. Genetic correlations between resistance to the same pathogen in calves and cows analyzed by G-REML.

https://doi.org/10.1371/journal.pone.0194374.s003

(XLSX)

S4 Table. Genetic correlations between all calf resistance traits and either birth weight or average daily gain in calves.

https://doi.org/10.1371/journal.pone.0194374.s004

(XLSX)

S5 Table. Genetic correlations between all cow resistance traits and either milk yield or fat to protein ratio.

https://doi.org/10.1371/journal.pone.0194374.s005

(XLSX)

S6 Table. P-values for the candidate 20 genes and their average across all resistance traits in calves and cows.

https://doi.org/10.1371/journal.pone.0194374.s006

(XLSX)

S7 Table. P-values for the most important 20 pathways and their average across all resistance traits in calves and cows.

https://doi.org/10.1371/journal.pone.0194374.s007

(XLSX)

S8 Table. Summarizes the code of all pathogens.

https://doi.org/10.1371/journal.pone.0194374.s008

(XLSX)

Acknowledgments

We acknowledge the invaluable contributions of Pau Navarro, Carmen Amador and Charley Xia at the MRC Human Genetics Unit, University of Edinburgh and thank Andrea Wilson of The Roslin Institute, University of Edinburgh, for additional valuable comments on the manuscript. This research funded by the German Federal Ministry of Education and Research (BMBF), the Förderverein Bioökonomieforschung e.V. (FBV) / German Holstein Association (DHV) and through the collaborative project “KMU-innovativ-10: Kuh-L–cow calibration groups for the implementation of selection strategies based on high-density genotyping in dairy cattle” (grant no. 031A416C).

References

1. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Introduction to Pathogens. 2002.
2. Garrity G, Boone DR, Castenholz RW. Bergey's Manual of Systematic Bacteriology: Volume One: The Archaea and the Deeply Branching and Phototrophic Bacteria: Archaea and the Deeply … of Systematic Bacteriology. 2nd ed. 2001 edition ed. Edinburgh; New York: Springer; 2001 2001. 722 p.
3. Bishop SC, Woolliams JA. Genomics and disease resistance studies in livestock. Livest Sci. 2014;166(Supplement C):190–8.
- View Article
- Google Scholar
4. Frodsham AJ, Hill AVS. Genetics of infectious diseases. Hum Mol Genet. 2004;13(suppl_2):R187–R94.
- View Article
- Google Scholar
5. Gernand E, Rehbein P, von Borstel UU, König S. Incidences of and genetic parameters for mastitis, claw disorders, and common health traits recorded in dairy cattle contract herds. J Dairy Sci. 2012;95(4):2144–56. pmid:22459859
- View Article
- PubMed/NCBI
- Google Scholar
6. Berry DP, Bermingham ML, Good M, More SJ. Genetics of animal health and disease in cattle. Ir Vet J. 2011;64(1):5. pmid:21777492
- View Article
- PubMed/NCBI
- Google Scholar
7. Raszek MM, Guan LL, Plastow GS. Use of Genomic Tools to Improve Cattle Health in the Context of Infectious Diseases. Front Genet. 2016;7.
- View Article
- Google Scholar
8. Allen AR, Minozzi G, Glass EJ, Skuce RA, McDowell SWJ, Woolliams JA, et al. Bovine tuberculosis: the genetic basis of host susceptibility. Proceedings of the Royal Society B: Biological Sciences. 2010;277(1695):2737–45. pmid:20519223
- View Article
- PubMed/NCBI
- Google Scholar
9. Finlay EK, Berry DP, Wickham B, Gormley EP, Bradley DG. A Genome Wide Association Scan of Bovine Tuberculosis Susceptibility in Holstein-Friesian Dairy Cattle. PLoS ONE. 2012;7(2):e30545. pmid:22355315
- View Article
- PubMed/NCBI
- Google Scholar
10. Bermingham ML, Bishop SC, Woolliams JA, Pong-Wong R, Allen AR, McBride SH, et al. Genome-wide association study identifies novel loci associated with resistance to bovine tuberculosis. Heredity. 2014;112(5):543–51. pmid:24496092
- View Article
- PubMed/NCBI
- Google Scholar
11. Tsairidou S, Brotherstone S, Coffey M, Bishop SC, Woolliams JA. Quantitative genetic analysis of the bTB diagnostic single intradermal comparative cervical test (SICCT). Genet Sel Evol. 2016;48.
- View Article
- Google Scholar
12. Tsairidou S, Woolliams JA, Allen AR, Skuce RA, McBride SH, Wright DM, et al. Genomic Prediction for Tuberculosis Resistance in Dairy Cattle. PLOS ONE. 2014;9(5):e96728. pmid:24809715
- View Article
- PubMed/NCBI
- Google Scholar
13. Richardson IW, Berry DP, Wiencko HL, Higgins IM, More SJ, McClure J, et al. A genome-wide association study for genetic susceptibility to Mycobacterium bovis infection in dairy cattle identifies a susceptibility QTL on chromosome 23. Genet Select Evol. 2016;48:19.
- View Article
- Google Scholar
14. Kiser JN, White SN, Johnson KA, Hoff JL, Taylor JF, Neibergs HL. Identification of loci associated with susceptibility to Mycobacterium avium subspecies paratuberculosis (Map) tissue infection in cattle. J. Animal Sci. 2017.
- View Article
- Google Scholar
15. Settles M, Zanella R, McKay SD, Schnabel RD, Taylor JF, Whitlock R, et al. A whole genome association analysis identifies loci associated with Mycobacterium avium subsp. paratuberculosis infection status in US holstein cattle. Anim Genet. 2009;40(5):655–62. pmid:19422364
- View Article
- PubMed/NCBI
- Google Scholar
16. Zanella R, Settles ML, McKay SD, Schnabel R, Taylor J, Whitlock RH, et al. Identification of loci associated with tolerance to Johne's disease in Holstein cattle. Anim Genet. 2011;42(1):28–38. pmid:20477805
- View Article
- PubMed/NCBI
- Google Scholar
17. Alpay F, Zare Y, Kamalludin MH, Huang X, Shi X, Shook GE, et al. Genome-Wide Association Study of Susceptibility to Infection by Mycobacterium avium Subspecies paratuberculosis in Holstein Cattle. PLOS ONE. 2014;9(12):e111704. pmid:25473852
- View Article
- PubMed/NCBI
- Google Scholar
18. Zare Y, Shook GE, Collins MT, Kirkpatrick BW. Genome-Wide Association Analysis and Genomic Prediction of Mycobacterium avium Subspecies paratuberculosis Infection in US Jersey Cattle. PLOS ONE. 2014;9(2):e88380. pmid:24523889
- View Article
- PubMed/NCBI
- Google Scholar
19. Wray C, Sojka WJ. Experimental Salmonella typhimurium infection in calves. Res Vet Sci. 1978;25(2):139–43. pmid:364573
- View Article
- PubMed/NCBI
- Google Scholar
20. Templeton JW, Estes DM, Price RE, Smith R, III, Adams LG, editors. Immunogenetics of natural resistance to bovine brucellosis. Proceedings of the 4th World Congress on Genetics applied to Livestock Production, Edinburgh 23–27 July 1990.
21. Heringstad B, Chang YM, Gianola D, Østerås O. Short Communication: Genetic Analysis of Respiratory Disease in Norwegian Red Calves. J Dairy Sci. 2008;91(1):367–70. pmid:18096960
- View Article
- PubMed/NCBI
- Google Scholar
22. Bishop SC, Morris CA. Genetics of disease resistance in sheep and goats. Small Rumin Res. 2007;70:48–59.
- View Article
- Google Scholar
23. Stear MJ, Murray M. Genetic resistance to parasitic disease: particularly of resistance in ruminants to gastrointestinal nematodes. Vet. Parasitol. 1994;54:161–76. pmid:7846849
- View Article
- PubMed/NCBI
- Google Scholar
24. Pan Y, Jansen GB, Duffield TF, Hietala S, Kelton D, Lin CY, et al. Genetic susceptibility to Neospora caninum infection in Holstein cattle in Ontario. J Dairy Sci. 2004;87(11):3967–75. pmid:15483182
- View Article
- PubMed/NCBI
- Google Scholar
25. Coppieters W, Mes THM, Druet T, Farnir F, Tamma N, Schrooten C, et al. Mapping QTL influencing gastrointestinal nematode burden in Dutch Holstein-Friesian dairy cattle. BMC Genomics. 2009;10:96. pmid:19254385
- View Article
- PubMed/NCBI
- Google Scholar
26. Davis GP. Genetic parameters for tropical beef cattle in northern Australia: a review. Aust J Agric Res. 1993;44(2):179–98.
- View Article
- Google Scholar
27. Lee B-Y, Lee K-N, Lee T, Park J-H, Kim S-M, Lee H-S, et al. Bovine Genome-wide Association Study for Genetic Elements to Resist the Infection of Foot-and-mouth Disease in the Field. Asian-australas J Anim Sci. 2015;28(2):166–70. pmid:25557811
- View Article
- PubMed/NCBI
- Google Scholar
28. Casas E, Hessman BE, Keele JW, Ridpath JF. A genome-wide association study for the incidence of persistent bovine viral diarrhea virus infection in cattle. Anim Genet. 2015;46(1):8–15. pmid:25394207
- View Article
- PubMed/NCBI
- Google Scholar
29. Bulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh P-R, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47(11):1236–41. pmid:26414676
- View Article
- PubMed/NCBI
- Google Scholar
30. Solovieff N, Cotsapas C, Lee PH, Purcell SM, Smoller JW. Pleiotropy in complex traits: challenges and strategies. Nature Rev Genet. 2013;14(7):483–95. pmid:23752797
- View Article
- PubMed/NCBI
- Google Scholar
31. Yang J, Benyamin B, McEvoy BP, Gordon S, Henders AK, Nyholt DR, et al. Common SNPs explain a large proportion of the heritability for human height. Nat Genet. 2010;42(7):565–9. pmid:20562875
- View Article
- PubMed/NCBI
- Google Scholar
32. Chen G-B, Lee SH, Brion M-JA, Montgomery GW, Wray NR, Radford-Smith GL, et al. Estimation and partitioning of (co)heritability of inflammatory bowel disease from GWAS and immunochip data. Hum Mol Genet. 2014;23(17):4710–20. pmid:24728037
- View Article
- PubMed/NCBI
- Google Scholar
33. Purcell SM, Wray NR, Stone JL, Visscher PM, O'Donovan MC, Sullivan PF, et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009;460(7256):748–52. pmid:19571811
- View Article
- PubMed/NCBI
- Google Scholar
34. Dudbridge F. Power and Predictive Accuracy of Polygenic Risk Scores. PLoS Genet. 2013;9(3):e1003348. pmid:23555274
- View Article
- PubMed/NCBI
- Google Scholar
35. Bulik-Sullivan BK, Loh P-R, Finucane HK, Ripke S, Yang J, Schizophrenia Working Group of the Psychiatric Genomics C, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291–5. pmid:25642630
- View Article
- PubMed/NCBI
- Google Scholar
36. Swalve HH. German Projects for Genomic Based Breeding Programs in Holstein. Walsrode and Verden/Aller, Germany 2015.
37. Passafaro TL, Carrera JPB, Santos LLd, Raidan FSS, Santos DCCd, Cardoso EP, et al. Genetic analysis of resistance to ticks, gastrointestinal nematodes and Eimeria spp. in Nellore cattle. Vet. Parasitol. 2015;210(3–4):224–34. pmid:25899078
- View Article
- PubMed/NCBI
- Google Scholar
38. Thompson-Crispi KA, Sargolzaei M, Ventura R, Abo-Ismail M, Miglior F, Schenkel F, et al. A genome-wide association study of immune response traits in Canadian Holstein cattle. BMC genomics. 2014;15(1):559.
- View Article
- Google Scholar
39. Hamzić E, Buitenhuis B, Hérault F, Hawken R, Abrahamsen MS, Servin B, et al. Genome-wide association study and biological pathway analysis of the Eimeria maxima response in broilers. Genet Select Evol. 2015;47:91.
- View Article
- Google Scholar
40. Mahmoud M, Yin T, Brügemann K, König S. Phenotypic, genetic, and single nucleotide polymorphism marker associations between calf diseases and subsequent performance and disease occurrences of first-lactation German Holstein cows. J Dairy Sci. 2017;100(3):2017–31. pmid:28109590
- View Article
- PubMed/NCBI
- Google Scholar
41. Heringstad B, Klemetsdal G, Ruane J. Selection for mastitis resistance in dairy cattle: a review with focus on the situation in the Nordic countries. Livest Prod Sci. 2000;64(2):95–106.
- View Article
- Google Scholar
42. Rupp R, Bergonier D, Dion S, Hygonenq MC, Aurel MR, Robert-Granié C, et al. Response to somatic cell count-based selection for mastitis resistance in a divergent selection experiment in sheep. J Dairy Sci. 2009;92(3):1203–19. pmid:19233814
- View Article
- PubMed/NCBI
- Google Scholar
43. Feucker W, Staufenbiel R. Zentraler Diagnoseschlüssel. http://www.portal-rind.de/index.php?module=Downloads&func=prep_hand_out&lid=17. Access 16.08.2016.
44. Stock KF, Cole J, Pryce J, Gengler N, Bradley A, Andrews L, et al. Standardization of health data–ICAR guidelines including health key. ICAR Technical Series no 17. 2013:75.
45. Segelke D, Chen J, Liu Z, Reinhardt F, Thaller G, Reents R. Reliability of genomic prediction for German Holsteins using imputed genotypes from low-density chips. J Dairy Sci. 2012;95(9):5403–11. pmid:22916947
- View Article
- PubMed/NCBI
- Google Scholar
46. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38(8):904–9. pmid:16862161
- View Article
- PubMed/NCBI
- Google Scholar
47. Price AL, Zaitlen NA, Reich D, Patterson N. New approaches to population stratification in genome-wide association studies. Nature Rev Genet. 2010;11(7):459–63. pmid:20548291
- View Article
- PubMed/NCBI
- Google Scholar
48. Yang J, Lee SH, Goddard ME, Visscher PM. GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet. 2011;88(1):76–82. pmid:21167468
- View Article
- PubMed/NCBI
- Google Scholar
49. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Series B Stat Methodol. 1995;57(1):289–300.
- View Article
- Google Scholar
50. Leeuw Cad, Mooij JM, Heskes T, Posthuma D. MAGMA: Generalized Gene-Set Analysis of GWAS Data. PLOS Comput. Biol. 2015;11(4):e1004219. pmid:25885710
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Introduction to Pathogens. 2002.

[ref2] 2. Garrity G, Boone DR, Castenholz RW. Bergey's Manual of Systematic Bacteriology: Volume One: The Archaea and the Deeply Branching and Phototrophic Bacteria: Archaea and the Deeply … of Systematic Bacteriology. 2nd ed. 2001 edition ed. Edinburgh; New York: Springer; 2001 2001. 722 p.

[ref3] 3. Bishop SC, Woolliams JA. Genomics and disease resistance studies in livestock. Livest Sci. 2014;166(Supplement C):190–8.
View Article
Google Scholar

[4] View Article

[5] Google Scholar

[ref4] 4. Frodsham AJ, Hill AVS. Genetics of infectious diseases. Hum Mol Genet. 2004;13(suppl_2):R187–R94.
View Article
Google Scholar

[7] View Article

[8] Google Scholar

[ref5] 5. Gernand E, Rehbein P, von Borstel UU, König S. Incidences of and genetic parameters for mastitis, claw disorders, and common health traits recorded in dairy cattle contract herds. J Dairy Sci. 2012;95(4):2144–56. pmid:22459859
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref6] 6. Berry DP, Bermingham ML, Good M, More SJ. Genetics of animal health and disease in cattle. Ir Vet J. 2011;64(1):5. pmid:21777492
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref7] 7. Raszek MM, Guan LL, Plastow GS. Use of Genomic Tools to Improve Cattle Health in the Context of Infectious Diseases. Front Genet. 2016;7.
View Article
Google Scholar

[18] View Article

[19] Google Scholar

[ref8] 8. Allen AR, Minozzi G, Glass EJ, Skuce RA, McDowell SWJ, Woolliams JA, et al. Bovine tuberculosis: the genetic basis of host susceptibility. Proceedings of the Royal Society B: Biological Sciences. 2010;277(1695):2737–45. pmid:20519223
View Article
PubMed/NCBI
Google Scholar

[21] View Article

[22] PubMed/NCBI

[23] Google Scholar

[ref9] 9. Finlay EK, Berry DP, Wickham B, Gormley EP, Bradley DG. A Genome Wide Association Scan of Bovine Tuberculosis Susceptibility in Holstein-Friesian Dairy Cattle. PLoS ONE. 2012;7(2):e30545. pmid:22355315
View Article
PubMed/NCBI
Google Scholar

[25] View Article

[26] PubMed/NCBI

[27] Google Scholar

[ref10] 10. Bermingham ML, Bishop SC, Woolliams JA, Pong-Wong R, Allen AR, McBride SH, et al. Genome-wide association study identifies novel loci associated with resistance to bovine tuberculosis. Heredity. 2014;112(5):543–51. pmid:24496092
View Article
PubMed/NCBI
Google Scholar

[29] View Article

[30] PubMed/NCBI

[31] Google Scholar

[ref11] 11. Tsairidou S, Brotherstone S, Coffey M, Bishop SC, Woolliams JA. Quantitative genetic analysis of the bTB diagnostic single intradermal comparative cervical test (SICCT). Genet Sel Evol. 2016;48.
View Article
Google Scholar

[33] View Article

[34] Google Scholar

[ref12] 12. Tsairidou S, Woolliams JA, Allen AR, Skuce RA, McBride SH, Wright DM, et al. Genomic Prediction for Tuberculosis Resistance in Dairy Cattle. PLOS ONE. 2014;9(5):e96728. pmid:24809715
View Article
PubMed/NCBI
Google Scholar

[36] View Article

[37] PubMed/NCBI

[38] Google Scholar

[ref13] 13. Richardson IW, Berry DP, Wiencko HL, Higgins IM, More SJ, McClure J, et al. A genome-wide association study for genetic susceptibility to Mycobacterium bovis infection in dairy cattle identifies a susceptibility QTL on chromosome 23. Genet Select Evol. 2016;48:19.
View Article
Google Scholar

[40] View Article

[41] Google Scholar

[ref14] 14. Kiser JN, White SN, Johnson KA, Hoff JL, Taylor JF, Neibergs HL. Identification of loci associated with susceptibility to Mycobacterium avium subspecies paratuberculosis (Map) tissue infection in cattle. J. Animal Sci. 2017.
View Article
Google Scholar

[43] View Article

[44] Google Scholar

[ref15] 15. Settles M, Zanella R, McKay SD, Schnabel RD, Taylor JF, Whitlock R, et al. A whole genome association analysis identifies loci associated with Mycobacterium avium subsp. paratuberculosis infection status in US holstein cattle. Anim Genet. 2009;40(5):655–62. pmid:19422364
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref16] 16. Zanella R, Settles ML, McKay SD, Schnabel R, Taylor J, Whitlock RH, et al. Identification of loci associated with tolerance to Johne's disease in Holstein cattle. Anim Genet. 2011;42(1):28–38. pmid:20477805
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref17] 17. Alpay F, Zare Y, Kamalludin MH, Huang X, Shi X, Shook GE, et al. Genome-Wide Association Study of Susceptibility to Infection by Mycobacterium avium Subspecies paratuberculosis in Holstein Cattle. PLOS ONE. 2014;9(12):e111704. pmid:25473852
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref18] 18. Zare Y, Shook GE, Collins MT, Kirkpatrick BW. Genome-Wide Association Analysis and Genomic Prediction of Mycobacterium avium Subspecies paratuberculosis Infection in US Jersey Cattle. PLOS ONE. 2014;9(2):e88380. pmid:24523889
View Article
PubMed/NCBI
Google Scholar

[58] View Article

[59] PubMed/NCBI

[60] Google Scholar

[ref19] 19. Wray C, Sojka WJ. Experimental Salmonella typhimurium infection in calves. Res Vet Sci. 1978;25(2):139–43. pmid:364573
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref20] 20. Templeton JW, Estes DM, Price RE, Smith R, III, Adams LG, editors. Immunogenetics of natural resistance to bovine brucellosis. Proceedings of the 4th World Congress on Genetics applied to Livestock Production, Edinburgh 23–27 July 1990.

[ref21] 21. Heringstad B, Chang YM, Gianola D, Østerås O. Short Communication: Genetic Analysis of Respiratory Disease in Norwegian Red Calves. J Dairy Sci. 2008;91(1):367–70. pmid:18096960
View Article
PubMed/NCBI
Google Scholar

[67] View Article

[68] PubMed/NCBI

[69] Google Scholar

[ref22] 22. Bishop SC, Morris CA. Genetics of disease resistance in sheep and goats. Small Rumin Res. 2007;70:48–59.
View Article
Google Scholar

[71] View Article

[72] Google Scholar

[ref23] 23. Stear MJ, Murray M. Genetic resistance to parasitic disease: particularly of resistance in ruminants to gastrointestinal nematodes. Vet. Parasitol. 1994;54:161–76. pmid:7846849
View Article
PubMed/NCBI
Google Scholar

[74] View Article

[75] PubMed/NCBI

[76] Google Scholar

[ref24] 24. Pan Y, Jansen GB, Duffield TF, Hietala S, Kelton D, Lin CY, et al. Genetic susceptibility to Neospora caninum infection in Holstein cattle in Ontario. J Dairy Sci. 2004;87(11):3967–75. pmid:15483182
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref25] 25. Coppieters W, Mes THM, Druet T, Farnir F, Tamma N, Schrooten C, et al. Mapping QTL influencing gastrointestinal nematode burden in Dutch Holstein-Friesian dairy cattle. BMC Genomics. 2009;10:96. pmid:19254385
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref26] 26. Davis GP. Genetic parameters for tropical beef cattle in northern Australia: a review. Aust J Agric Res. 1993;44(2):179–98.
View Article
Google Scholar

[86] View Article

[87] Google Scholar

[ref27] 27. Lee B-Y, Lee K-N, Lee T, Park J-H, Kim S-M, Lee H-S, et al. Bovine Genome-wide Association Study for Genetic Elements to Resist the Infection of Foot-and-mouth Disease in the Field. Asian-australas J Anim Sci. 2015;28(2):166–70. pmid:25557811
View Article
PubMed/NCBI
Google Scholar

[89] View Article

[90] PubMed/NCBI

[91] Google Scholar

[ref28] 28. Casas E, Hessman BE, Keele JW, Ridpath JF. A genome-wide association study for the incidence of persistent bovine viral diarrhea virus infection in cattle. Anim Genet. 2015;46(1):8–15. pmid:25394207
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref29] 29. Bulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh P-R, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47(11):1236–41. pmid:26414676
View Article
PubMed/NCBI
Google Scholar

[97] View Article

[98] PubMed/NCBI

[99] Google Scholar

[ref30] 30. Solovieff N, Cotsapas C, Lee PH, Purcell SM, Smoller JW. Pleiotropy in complex traits: challenges and strategies. Nature Rev Genet. 2013;14(7):483–95. pmid:23752797
View Article
PubMed/NCBI
Google Scholar

[101] View Article

[102] PubMed/NCBI

[103] Google Scholar

[ref31] 31. Yang J, Benyamin B, McEvoy BP, Gordon S, Henders AK, Nyholt DR, et al. Common SNPs explain a large proportion of the heritability for human height. Nat Genet. 2010;42(7):565–9. pmid:20562875
View Article
PubMed/NCBI
Google Scholar

[105] View Article

[106] PubMed/NCBI

[107] Google Scholar

[ref32] 32. Chen G-B, Lee SH, Brion M-JA, Montgomery GW, Wray NR, Radford-Smith GL, et al. Estimation and partitioning of (co)heritability of inflammatory bowel disease from GWAS and immunochip data. Hum Mol Genet. 2014;23(17):4710–20. pmid:24728037
View Article
PubMed/NCBI
Google Scholar

[109] View Article

[110] PubMed/NCBI

[111] Google Scholar

[ref33] 33. Purcell SM, Wray NR, Stone JL, Visscher PM, O'Donovan MC, Sullivan PF, et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009;460(7256):748–52. pmid:19571811
View Article
PubMed/NCBI
Google Scholar

[113] View Article

[114] PubMed/NCBI

[115] Google Scholar

[ref34] 34. Dudbridge F. Power and Predictive Accuracy of Polygenic Risk Scores. PLoS Genet. 2013;9(3):e1003348. pmid:23555274
View Article
PubMed/NCBI
Google Scholar

[117] View Article

[118] PubMed/NCBI

[119] Google Scholar

[ref35] 35. Bulik-Sullivan BK, Loh P-R, Finucane HK, Ripke S, Yang J, Schizophrenia Working Group of the Psychiatric Genomics C, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291–5. pmid:25642630
View Article
PubMed/NCBI
Google Scholar

[121] View Article

[122] PubMed/NCBI

[123] Google Scholar

[ref36] 36. Swalve HH. German Projects for Genomic Based Breeding Programs in Holstein. Walsrode and Verden/Aller, Germany 2015.

[ref37] 37. Passafaro TL, Carrera JPB, Santos LLd, Raidan FSS, Santos DCCd, Cardoso EP, et al. Genetic analysis of resistance to ticks, gastrointestinal nematodes and Eimeria spp. in Nellore cattle. Vet. Parasitol. 2015;210(3–4):224–34. pmid:25899078
View Article
PubMed/NCBI
Google Scholar

[126] View Article

[127] PubMed/NCBI

[128] Google Scholar

[ref38] 38. Thompson-Crispi KA, Sargolzaei M, Ventura R, Abo-Ismail M, Miglior F, Schenkel F, et al. A genome-wide association study of immune response traits in Canadian Holstein cattle. BMC genomics. 2014;15(1):559.
View Article
Google Scholar

[130] View Article

[131] Google Scholar

[ref39] 39. Hamzić E, Buitenhuis B, Hérault F, Hawken R, Abrahamsen MS, Servin B, et al. Genome-wide association study and biological pathway analysis of the Eimeria maxima response in broilers. Genet Select Evol. 2015;47:91.
View Article
Google Scholar

[133] View Article

[134] Google Scholar

[ref40] 40. Mahmoud M, Yin T, Brügemann K, König S. Phenotypic, genetic, and single nucleotide polymorphism marker associations between calf diseases and subsequent performance and disease occurrences of first-lactation German Holstein cows. J Dairy Sci. 2017;100(3):2017–31. pmid:28109590
View Article
PubMed/NCBI
Google Scholar

[136] View Article

[137] PubMed/NCBI

[138] Google Scholar

[ref41] 41. Heringstad B, Klemetsdal G, Ruane J. Selection for mastitis resistance in dairy cattle: a review with focus on the situation in the Nordic countries. Livest Prod Sci. 2000;64(2):95–106.
View Article
Google Scholar

[140] View Article

[141] Google Scholar

[ref42] 42. Rupp R, Bergonier D, Dion S, Hygonenq MC, Aurel MR, Robert-Granié C, et al. Response to somatic cell count-based selection for mastitis resistance in a divergent selection experiment in sheep. J Dairy Sci. 2009;92(3):1203–19. pmid:19233814
View Article
PubMed/NCBI
Google Scholar

[143] View Article

[144] PubMed/NCBI

[145] Google Scholar

[ref43] 43. Feucker W, Staufenbiel R. Zentraler Diagnoseschlüssel. http://www.portal-rind.de/index.php?module=Downloads&func=prep_hand_out&lid=17. Access 16.08.2016.

[ref44] 44. Stock KF, Cole J, Pryce J, Gengler N, Bradley A, Andrews L, et al. Standardization of health data–ICAR guidelines including health key. ICAR Technical Series no 17. 2013:75.

[ref45] 45. Segelke D, Chen J, Liu Z, Reinhardt F, Thaller G, Reents R. Reliability of genomic prediction for German Holsteins using imputed genotypes from low-density chips. J Dairy Sci. 2012;95(9):5403–11. pmid:22916947
View Article
PubMed/NCBI
Google Scholar

[149] View Article

[150] PubMed/NCBI

[151] Google Scholar

[ref46] 46. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38(8):904–9. pmid:16862161
View Article
PubMed/NCBI
Google Scholar

[153] View Article

[154] PubMed/NCBI

[155] Google Scholar

[ref47] 47. Price AL, Zaitlen NA, Reich D, Patterson N. New approaches to population stratification in genome-wide association studies. Nature Rev Genet. 2010;11(7):459–63. pmid:20548291
View Article
PubMed/NCBI
Google Scholar

[157] View Article

[158] PubMed/NCBI

[159] Google Scholar

[ref48] 48. Yang J, Lee SH, Goddard ME, Visscher PM. GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet. 2011;88(1):76–82. pmid:21167468
View Article
PubMed/NCBI
Google Scholar

[161] View Article

[162] PubMed/NCBI

[163] Google Scholar

[ref49] 49. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Series B Stat Methodol. 1995;57(1):289–300.
View Article
Google Scholar

[165] View Article

[166] Google Scholar

[ref50] 50. Leeuw Cad, Mooij JM, Heskes T, Posthuma D. MAGMA: Generalized Gene-Set Analysis of GWAS Data. PLOS Comput. Biol. 2015;11(4):e1004219. pmid:25885710
View Article
PubMed/NCBI
Google Scholar

[168] View Article

[169] PubMed/NCBI

[170] Google Scholar

Figures

Abstract

Introduction

Results

SNP heritabilities

Pleiotropy among calf resistance traits

The pleiotropy among cow resistance traits

Pathogen resistance and calf performance

Pathogen resistance and cow productivity

Biological pathway analysis

Discussion

Pleiotropy between resistance to pathogens and calf performance and cow productivity

Shared biological pathways

The biological impact of the genetic pleiotropy on breeding and selection strategies

Materials and methods

Phenotypes

Genotypes

Statistical models

Correction for population stratification in genome-wide data

Significance thresholds

Biological pathway analysis

Multi-marker Analysis of GenoMic Annotation (MAGMA)

Supporting information

S1 Table. (3 sheets).

S2 Table. (3 sheets).

S3 Table. Genetic correlations between resistance to the same pathogen in calves and cows analyzed by G-REML.

S4 Table. Genetic correlations between all calf resistance traits and either birth weight or average daily gain in calves.

S5 Table. Genetic correlations between all cow resistance traits and either milk yield or fat to protein ratio.

S6 Table. P-values for the candidate 20 genes and their average across all resistance traits in calves and cows.

S7 Table. P-values for the most important 20 pathways and their average across all resistance traits in calves and cows.

S8 Table. Summarizes the code of all pathogens.

Acknowledgments

References