Final Report Summary - PIGENDEF (Towards marker assisted selection against genetic defects in pigs)
In January 2011, the PIGENDEF project was launched. The aim of the project was to increase our knowledge on the genetic background of congenital genetic defects in pigs. Congenital diseases are quite common in swine and they cover a wide range of undesirable conditions, such as hernias (umbilical and inguinal or scrotal hernias), cryptorchidism and splay leg, and to a lesser extent intersexuality, anal atresia and progressive ataxia and spastic paresis. As congenital defects lead to poor animal welfare and economic losses, they pose a significant problem for pig producers and breeding companies.
PIGENDEF is the acronym of the project 'Towards marker assisted selection against genetic defects in pigs'. It is funded by the Seventh Framework Program of the European Community for Research, Technological Development and Demonstration Activities within the framework of the Research for the benefit of small and medium enterprises (SME). The project consortium consists of 3 research partners and 5 SME partners located in 5 different countries.
In the PIGENDEF project, the three most common congenital genetic defects are researched: scrotal/inguinal and umbilical hernia, and cryptorchidism. The PIGENDEF project focuses on the rapid identification of genes underlying these congenital disorders in pigs for effective marker assisted selection. In order to achieve this goal, pigs are genotyped using the 60K Porcine Illumina SNP chip and analyzed using the newest, tailor-made genome wide association analysis software.
The PIGENDEF project is coordinated by KU Leuven (Belgium) which is a research partner together with CIGENE, University of Life Sciences (Norway) and University of Liège (Belgium). They provide a successful combination of expert groups in animal genetics, genomics and breeding, bioinformatics, statistical analysis and high throughput genotyping. This provides a solid scientific and applied basis for the PIGENDEF project from the first stages of its design, through performance of its core part, up to the final stages of knowledge and the SME partners’ implementation and dissemination.
The 5 SME partners of the PIGENDEF consortium are companies active in pig breeding. These companies with experience in pig genetic research ensure the availability of resource populations and guarantee the exploitation of the results in favor of the pig breeding community. The two core SME partners, Rattlerow-Seghers N.V. (Belgium) and Norsvin (Norway), are providing samples and datasets of affected piglets and/or unaffected siblings and their parents within their lines. These partners acquire the intellectual property and organize the implementation of the results. The other SME partners, Santa Fosca (Italy), Nordic Genetics (Sweden) and Selección Batallé S.A. (Spain), are producing crossbreds of the pure lines provided by the core SME partners which ensures confirmation and immediate valorization of the results.
Within the framework of the project, a Tag SNP panel was created for screening for the susceptibility to umbilical hernia in pigs. Although this tag SNP Panel is currently restricted to the use in some specific lines, research is ongoing in order to verify this SNP Panel in different lines and, possibly, locate the causative SNP underlying this region. For scrotal hernia and cryptorchidism, several interesting regions were located, however no easy-to-implement results were acquired for these defects.
Project Context and Objectives:
The PIGENDEF project focuses on the rapid identification of genes underlying congenital disorders in pigs for effective marker assisted selection against genetic defects. In the PIGENDEF project, the focus lies on the research for hernias and cryptorchidism, as they are by far the most important congenital defects in pigs.
For both hernias and cryptorchidism, there are strong indications for a genetic component and different research groups all over Europe have made efforts to find the underlying causative genetic mechanisms. However, the mode of inheritance seems complex and at the advent of this project, no genes or mutations had been identified yet. Furthermore, besides a genetic compound there is a strong influence of environmental factors. Environmental factors such as abnormal stretching of the umbilical cord (during farrowing or placing naval clips too close to the skin) or infection of the umbilical stump could contribute to failure of the umbilical cord opening to close and leading to the occurrence of umbilical hernia. This only adds to the complexity of the disorders studied in this project.
As stated before, the focus of this project lies on the research for genetic markers and genes underlying hernias and cryptorchidism in pigs. In general, two types of hernias are distinguished: umbilical hernia and inguinal/scrotal hernia. An umbilical hernia is phenotypically characterized as a part of the intestine that protrudes through an opening in the abdominal muscles at the site of the umbilicus, the protruding intestine covered by skin and subcutaneous tissue. The other type of hernias are hernia inguinalis and hernia scrotalis which occur within the pig population at frequencies from 1.7% to 6.7% (Thaller et al., 1996). These types of hernia are phenotypically characterized as an uncontrolled prolapse of the small intestine into the scrotum (scrotal hernia) or into the inguinal canal (inguinal hernia).
Several studies estimate the heritability for these types of hernia from 0.21 (Althoff et al., 1988) to 0.86 (Mikami and Freeden, 1979, Thaller et al., 1996). These numbers indicate that, in addition to environmental factors, also genetic factors, probably with incomplete penetrance, are involved in the development of inguinal and scrotal hernia (Cook et al., 2000; Koskimies et al., 2003). However the mode of inheritance has not yet been clarified (Beck et al,. 2006). Based on the high segregation ratio of hernia inguinalis in humans, autosomal dominant inheritance with incomplete penetrance and effect of gender was suggested. The authors of this study also described preferential paternal inheritance of the defect allele, which suggests a possible role for genomic imprinting in the prevalence of inguinal hernias in humans (Gong et al., 1994). Although such results were not (yet) discovered in pigs, it is probable that the incidence of inguinal or scrotal hernia is due to a small number of genes, among which might be some major genes (Deeb et al., 2004).
The second disorder, cryptorchidism, is a congenital sex-limited abnormality that occurs in approximately 1% to 1.4% of all male piglets (Knap et al., 1986; Rothschild et al., 1988). A pig is cryptorchid if one or both testicles are not descended correctly in the scrotum. Rothschild et al. (1988) propose for its genetic control a minimum of two loci. Heritability estimates between 0.06 and 0.33 were reported (Knap et al., 1986; Althoff et al., 1988). Also, in Pietrains low additive genetic correlations between hernia scrotalis and cryptorchidism exist (Beißner et al., 2003). This might be due to the fact that hernia scrotalis and cryptorchidism are biologically related.
In total, congenital genetic disorders affect on average 3% of pigs in commercial populations worldwide. As congenital diseases are widespread amongst pig populations, the occurrence of congenital hereditary disorders is a significant problem for pig producers and breeding companies and leads to poor animal welfare and severe economic loss.
In Europe the economic loss due to congenital defects is estimated at 200 million euros per year, with hernias and cryptorchidism accounting for more than half of this amount. However, besides the economic loss, these defects also have a serious impact on animal welfare and health. Some of the defects cause direct piglet mortality, while others lead to culling of piglets. For every single congenital disorder, it is possible to calculate an approximate loss for the pig industry. However, due to the specific physiology of the different defects, the costs and burdens associated with the diseases differ a lot. Therefore, in what follows, economic losses and impact on animal welfare and sustainable agriculture are described for inguinal and scrotal hernias, since this is one of the main focuses of this project. In order to calculate the losses due to umbilical hernia and cryptorchidism, the same approach was followed, but only the approximate monetary losses are given.
Hernia inguinalis and hernia scrotalis occur within the pig population at frequencies from 1.7% to 6.7% (Thaller et al., 1996). However, these numbers are averages and in practice percentages up to 10% are recorded in offspring of particular boars. On top of that, breeding organisations select against inguinal/scrotal hernia within their own lines, independent from each other. Due to this selection these organisations are able to keep the prevalence of hernias below 2%. However, despite years of selection, hernias remain present. Although the percentage of hernias can be fairly well controlled within lines, it shows in practice that certain combinations, over breeding programmes, can be disastrous with incidences of over 10%.
The economic losses associated with the prevalence of scrotal/inguinal hernias can be situated in 3 areas:
Firstly, animals that develop scrotal/inguinal hernia show an increased incidence of cryptorchidism and are difficult to castrate. This leads to an economic cost for the pig producer since cryptorchids and boars are fined in the slaughter house. Also porc producers are not willing to buy affected animals, which means the piglet producers have to raise them theirselves or sell them as slaughter piglet. Therefore, the average loss per pig can be estimated at 50 euro.
Secondly, pigs that develop scrotal/inguinal hernia need more medical attention compared to healthy animals. The extra labor and drugs invoke extra costs for the pig producer. On top of that, higher mortality rates are observed in animals affected with scrotal/inguinal hernia as a result of the strangulation of the small intestine in the scrotum. It is estimated that only 50% of the affected animals reaches the slaughter stage. The consequences of disease and mortality are estimated at 10-20 euro/pig.
Thirdly, pigs that develop inguinal/scrotal hernias display inferior zootechnical performance parameters. They cannot feed efficiently and their growth is affected. This leads to higher feed costs, slower throughput, lack of product uniformity and the consequent loss of income (Charagu, 2005).
Based on these factors, the economic losses due to scrotal/inguinal hernias in the European pig population are estimated to be nearly 50 million euros/year. These estimations were based on a prevalence of 2%, of which approximately half of the pigs will survive and will be fined in the slaughter house. The other half of the pigs will cost the pig producer approximately 20 euros per affected piglet due to inferior zootechnical performances, drugs and mortality during the growing period.
As mentioned before, in addition to economic losses, the incidence of congenital defects in the pig population also leads to a serious burden on sustainable agriculture. For inguinal/scrotal hernia, the negative effects with relation to sustainable technological development are twofold.
Firstly, pigs that develop scrotal/inguinal hernias display a very unfavorable feed conversion ratio. Therefore, during their growing period, herniated pigs will have a higher ecological cost compared to healthy animals. On top of that, approximately 50% of the affected pigs dies before they reach slaughter age. These animals have had a certain ecological cost but they did not end up as a slaughter animal. In this view, a solution for scrotal/inguinal hernia can decrease the nutrient emission in the pig sector.
Secondly, the prevalence of hernias leads to poor health and a low welfare status of the affected animals. In a social environment where animal welfare is a very important issue, this poses a significant problem. Therefore, a decrease in the incidence of hernias would be of great surplus value. Finally, selection for livestock improvement is widely accepted by the public so the development of a selection strategy against the prevalence of scrotal and inguinal hernias will receive a positive social evaluation.
For umbilical hernia and cryptorchidism, the same approach was followed to estimate the economic losses and social or ecological burdens.
• Yearly costs for umbilical hernia, estimations based on a prevalence of 1.5%: 46 million Euro
• Yearly costs for cryptorchidism, estimations based on a prevalence of 1.4%: 27 million Euro
With an economic loss, due to congenital defects, of 200 million annually, in Europe a genetic test that decreases the incidence of a genetic defect with 25 to 50% leads to a substantial surplus value for the pig sector.
However, as mentioned before, the mode of inheritance of the congenital genetic defects studied seems complex and at the advent of this project no genes or mutations had been identified yet. Therefore, this project bundles the efforts of different research groups that are currently working on inherited genetic defects in pigs and through this collaboration a faster identification of the genes and suitable gene markers for use in marker assisted selection against defects in pigs will be obtained. These markers can then be implemented in the breeding programs of the different SME partners and can be commercialized, which will improve the competitive position of the SME partners.
In order to locate markers that can be used for marker assisted selection against congenital genetic defects in pigs, the following scientific and technological objectives were formulated for the project:
1. Collection of phenotypical data and biological material from affected animals, litter mates and their parents: two of the SME participants, Norsvin and Rattlerow-Seghers N.V. are in the possession of resource populations for all congenital defects studied and already collected data and biological samples. This work is continued or initiated in all partner SMEs, which will result in an even higher number of samples at the start of the project. The availability of the datasets and blood samples is a valuable asset for the project, since the assembly of such a resource population is a time-consuming and costly work. It also considerably reduces the risk of failure of the project. In case the datasets available are not sufficient and more samples and data need to be collected, this should also not pose a problem. From the available dataset, it can be calculated that the incidence of the different congenital disorders in the different SME partners is still between 0.5 – 1.0 %. This means that data and samples could be collected on a short notice.
2. Identification of genetic markers associated with the sensitivity for the studied congenital disorders: Until sometime ago, the identification of genes causing inherited disorders was either performed using genome-wide pedigree-based linkage analyses, or by focusing on candidate genes using population-based association studies. However, due to the availability of the 60K porcine SNP chip (Illumina) these two approaches can be merged into a single, generic approach referred to as genome-wide association studies (GWA). GWA have been shown to be more powerful than linkage based methods (e.g. Risch and Merikangas, 1996). This has been spectacularly demonstrated by the identification of several genes underlying a range of common human diseases which were extremely difficult to study prior to the advent of GWA (e.g. Crohn disease, type II diabetes and obesity). Results obtained in cattle (Charlier et al., 2008) and dogs (Karlsson et al. 2007) clearly demonstrate that the approach is even more powerful in domestic animals. Since the 60K porcine SNP chip (Illumina) is commercially available since December 2008 and this project consortium has the privilege to incorporate two Illumina genotyping core facilities, the aim of this project is to apply the same approach to congenital defects in pigs.
In this project, samples from Norway and samples from Belgium for the disorders studied will be examined separately and together. The possibility of performing a meta-analysis is a great asset for this project, as markers identified in this analysis have a higher probability of being population independent and therefore causal.
3. Validation of the genetic markers found in crossbred lines: One of the main limitations of many livestock breeding programs is that selection occurs in pure breeds housed in high-health environments but the aim is to improve crossbred performance under field conditions. Differences in crossbred populations compared to purebred populations are that the effects of the SNP may be breed specific and linkage disequilibrium may not be restricted to markers that are tightly linked to the QTL (Ibanez-Escriche et al., 2009). Hence, the SNPs found to be associated with the susceptibility of the congenital genetic diseases studied in the purebred lines need to be validated in crossbreds.
4. Functional analysis of the uncovered genetic markers for congenital diseases: In this part of the project, it is researched whether the discovered genetic markers are causal mutations or markers perfectly correlated to the incidence of congenital defects in the lines studied. Significant regions will be subjected to resequencing and functional analysis. Since for the study of the different congenital disorders, lines and crossbred lines from two different pig selection companies (Norsvin and RaSe) were used, the odds that the genetic markers that are shared in both populations are causal are quite high. The RTD partners involved in this project already proved that the genetic analysis of a QTL can lead to the identification of the causative mutation (Van Laere et al., 2003) and that the availability of genome-wide, high-density SNP panels, combined with the typical structure of livestock populations, markedly accelerates the positional identification of genes and mutations that cause inherited defects (Charlier et al., 2008).
The overall scientific and technological objective of this project is to map the chromosomal areas that have an effect on congenital genetic defects and ideally identify the causal mutations that underlie these congenital hereditary diseases.
Although congenital diseases are quite common in swine and cover a range of conditions, in this project only three of the most important disorders were studied: inguinal/scrotal hernia, umbilical hernia and cryptorchidism. For both hernias and cryptorchidism, there are strong indications for a genetic component and different research groups all over Europe have made efforts to find the underlying causative genetic mechanisms. However, the mode of inheritance seems complex and at the advent of this project, no genes or mutations had been identified yet. Therefore, this project bundles the efforts of different research groups that are currently working on inherited genetic defects in pigs and through this collaboration a faster identification of the genes and suitable gene markers for use in marker assisted selection against defects in pigs will be obtained. These markers can then be implemented in the breeding programs of the different SME partners and can be commercialized, which will improve the competitive position of the SME partners.
In total, three RTD partners are involved in the project, namely KU Leuven (Belgium), which is also the coordinator, the university of Liège (Ulg, Belgium) and Cigene (Norway). The SME partners involved are Rattlerow-Seghers (Belgium), Norsvin (Norway), Santa Fosca (Italy), Selección Batallé (Spain) and Nordic Genetics (Sweden). Rattlerow-Seghers and Norsvin are the two main SME partners also providing samples and phenotype datasets.
SCIENTIFIC AND TECHNOLOGICAL RESULTS
1. Selection of phenotypical data and biological material from affected animals, litter mates and their parents
At the advent of this project, the two main SME partners, Rattlerow-Seghers and Norsvin, each brought in a substantial dataset and sample bank of animals affected with hernias or cryptorchidism, their parents and, sometimes, littermates. These datasets took several years to set up (7 to 10 years) and contain thousands of samples. The availability of these datasets were a major asset for the project. In addition to the samples from the datasets, within the framework of the PIGENDEF project additional samples were gathered. For this, a standardized protocol for sample collection was set up and circulated amongst the PIGENDEF participants.
Furthermore, also a standardized protocol for sample preparation, shipment and retrieval of results was set up and circulated amongst the PIGENDEF participants.
From the entire data and sample collection, suitable samples for genotyping were selected. In Norway (Norsvin), a sib pair analysis with incorporation of parental data was chosen. For this set-up, animals bearing one of the congenital disorders (cases), littermates (controls) and their parents were selected. For scrotal hernia, 144 cases, 60 controls and most of their parents were selected. For umbilical hernia, 349 cases, 170 controls and most of their parents were selected and for cryptorchidism 84 affected animals and their parents were selected.
In the Belgian dataset gathered by Rattlerow-Seghers, unaffected littermates were available for the old samples, however for the newer samples, only blood or DNA from cases and their parents were available. So in order to maintain a balanced dataset, no controls were selected from the Belgian data, only cases and both their parents. For scrotal hernia, 136 affected animals and their parents were selected, for cryptorchidism 196 affected animals and their parents were selected and for umbilical hernia 56 affected trios (case/mother/father) were picked. As mentioned before, during the project additional samples were gathered following the standardized protocols. Later in the project, these samples (scrotal hernia and cryptorchidism) were also selected and genotyped. However, since this took place at the very end of the project, the results from these samples are not yet available at the time of this report. However, it is envisioned that the results will be available in one of the following months.
The amounts of samples selected enabled us to perform separate and joint analysis for both scrotal hernia and cryptorchidism. However, due to the low number of cases for umbilical hernia in Belgium only the Norwegian data were processed in a first analysis of this trait, later the Belgian data were incorporated strictly for the joint analysis.
2. Statistical and functional analysis of the data
All samples described in the previous paragraph were genotyped using the PorcineSNP60 chip that features probes for 62163 different single nucleotide polymorphisms, offering more than sufficient SNP density for whole-genome association studies (WGAS) in pigs. After removal of markers with low call rate and markers in Hardy-Weinberg disequilibrium, a sufficient amount of SNP remained for the WGAS. Additionally, animals with low call rate and animals with sex problems were removed.
Dependent on the disorder studied and the samples used, different statistical analyses were performed on the remaining animals, using the remaining SNP. Both Plink (DFAM, APERM, MODEL, ASSOC, FISHER, TDT and PARENT) and GenABEL (qtscore, mmscore, ccfast, rassoc) were used to perform the analysis, but also a user-friendly software package, designed for the PIGENDEF project among others, was used. The design and development of this user-friendly genome-wide analysis package is part of previous and ongoing research activities in the Unit of Animal Genomics of the University of Liège. The package has been developed in consideration of the requirements for proper genetic analysis of pig data.
After a first analysis of the genotyping data, it was witnessed that the results for cryptorchidism and scrotal hernia were very similar. Based on the physiology of both disorders, this makes sense. The fact is that in normal circumstances, after testicular descent through the processus vaginalis, the processus vaginalis closes which prevents the jejenum and ileum to pass through it and descend into the inguinal canal or scrotum. However, when the processus vaginalis remains open, jejenum and ileum can pass through it, resulting in scrotal hernia. Often, the processus vaginalis remains open when the testis does not descend properly, since the processus vaginalis does not normally close unless the testes have reached the scrotum. Therefore, scrotal hernia and cryptorchidism are often related and it was decided to join the data for cryptorchidism and scrotal hernia. This leaves us with two major analyses: the analysis of the scrotal hernia/cryptorchidism data and the analysis of the umbilical hernia data. In what follows, both results from both analysis will be described separately.
2.1. Cryptorchidism/scrotal hernia:
For the cryptorchidism and scrotal hernia samples, a transmission disequilibrium test (TDT) was performed on the data of both countries separately and jointly. For this analysis, only those animals that have a father and a mother genotyped were used. A TDT is a family-based association test based on the rate of transmission of a certain allele from heterozygous parents to affected offspring. The TDT will detect genetic linkage only in the presence of genetic association. While genetic association can be caused by population substructure, genetic linkage cannot. This makes the TDT very robust to the presence of population substructure and very much suited for our analysis. The results obtained are described below:
a. The X-chromosome
In the Norwegian data, the Belgian data and the joint analysis, a strong signal was coming from the boundaries of the pseudo-autosomal region when analyzing the data of both disorders. However, since the two diseases under study only affect male offspring, we recalculated the TDT to capture the signal coming from maternal transmission only. This weakened the signal coming from the pseudo-autosomal region but nevertheless is was still significant. When the region was examined into more detail, it was found that the significant markers were located in/near a very important positional candidate gene. In some animal species, this gene is known to have an effect on reproduction traits in both males and females. As this would indeed also be true for pigs, it would explain why it is so difficult to select against the disorders. In order to find a possible causal mutation within this gene, we decided to perform three kinds of experiments/analysis simultaneously:
• Allelic discrimination assay
Firstly, an allelic discrimination assay was set up for one of the significant markers in the QTL region. This marker was genotyped in all sows from a previous project, performed by the University of Liège (Ulg) and Rattlerow-Seghers. This project aimed the identification of the causative genes for phenotypical differences in prolificacy, through the identification of quantitative trait loci (QTL) that influence reproductive characteristics in pigs. The project finished in December 2007, but the samples were conserved at the University of Liège. As a mutation in our positional candidate gene was expected to have an influence on prolificacy in sows, a group of sows with low prolificacy and a group of sows with high prolificacy were genotyped and allele frequencies were compared.
Linear regressions for a group of fertility quantitative traits was carried out using plink software and for checking the significance of the results, an adaptive permutation was performed. Results showed that there was no significant association between the marker genotyped and any of fertility traits tested.
Secondly, phasing was done using Beagle software (Browning & Browning, 2009). After getting the phased data from Beagle, the TDT haplotype test was performed. The significance level was determined by simulations. The simulations were done by just choosing a random haplotype from the heterozygote mothers and transmitting it to their offspring. However, performing this analysis no significant results could be obtained.
Thirdly, an immunohistochemistry experiment in mice was set up in order to locate the regions of expression of our positional candidate gene. As we envisioned a real-time experiment after the haplotype analysis, it is very important to know the location of the expression. All measures were taken to start the analysis, however after the non-significant results from the haplotype analysis, the experiment was stopped.
Based on the results we obtained from our experiments, and particularly the results we obtained from the haplotype analysis, we considered the significant signal that we observed before the fine mapping as an artifact signal. The reason of this artifact is because the region of interest was very near to the boundaries of the pseudo-autosomal region which could create such signals.
b. The autosomes
After exclusion of the pseudo-autosomal region on the X-chromosome, the results on the autosomes for cryptorchidism and scrotal hernia were reconsidered.
• Belgian data cryptorchidism
A suggestive peak for cryptorchidism in the Belgian data was found on Sus scrofa chromosome (SSC) 8.
• Belgian data cryptorchidism and scrotal hernia combined
Suggestive peaks for scrotal hernia and cryptorchidism in the Belgian data were found on SSC5 and SSC15.
• Norwegian data cryptorchidism
A suggestive peak for cryptorchidism in the Norwegian data was found on SSC18.
• Norwegian data scrotal hernia
Suggestive peaks for scrotal hernia in the Norwegian data were found on SSC14, SSC15 and SSC2.
• All data combined: Norwegian and Belgian data, Scrotal hernia and cryptorchidism
A suggestive peak for cryptorchidism and scrotal hernia in the Norwegian and Belgian data combined was found on SSC14.
In conclusion, it can be said that there are no markers for cryptorchidism and scrotal hernia that significantly pass the genome wide threshold. However, we still have quite some candidate signals from different chromosomes, even when we combine all data. The only problem is that we cannot confirm them with the data set at hand. For this, we need to increase our data set either by selecting more animals for the study or add other cohorts from different research groups around the world in order to perform a meta-analysis. In order to do this, we already started selecting and genotyping more animals from the populations used in our project. As we only excluded the X-chromosome rather late in the project, the genotyping of these samples is not finalized yet and therefore we cannot incorporate results from these analysis in this report. However, we envision that the genotyping and the analysis will be finished within a few months. Furthermore, we are currently trying to establish new cooperations in order to be able to perform an even larger meta-analysis.
2.2. Umbilical hernia
The genotyping data from the animals with umbilical hernia were analysed using both the program package PLINK (Purcell et al. 2007) and GenABEL (Aulchenko et al. 2007). The options used in PLINK were DFAM, APERM, MODEL, ASSOC, FISCHER, TDT and parent. In GenABEL the options qtscore, mmscore, ccfast and Rassoc were used.
As there were too little Belgian trios, in a first analysis only the Norwegian data were analyzed. In this analysis, assuming a corrected significance level of P<0.001 129 SNP markers reached the significant level for the Norwegian data. Three single SNP associations were obtained on SSC1 at the Porcine Build10.2 genome sequence (http://www.ensembl.org/index.html) on SSC8, and on SSC17, respectively. All the other 126 SNPs were located on SSC14. Looking at significance level of P<0.0001 62 SNP markers were significant and all of them were located to SSC14 at Build10.2.
In a later analysis running GenABEL, the Belgian material alone revealed no significance (p < 0.001). However, the most significant marker in this analysis (p=0.002) was in the middle of the Norwegian SSC14 QTL. Still, due to the low number of animals in this analysis, the data was merged with the Norwegian data for the most promising QTL regions on SSC14, SSC5 and SSC2. The two subpopulations occurring were accounted for by the population stratification option in GenABEL. Results from the merged analysis showed that both SSC14 and SSC5 were significant QTL regions (p < 0.001). As the region on chromosome 14 was clearly the most interesting region, it was decided to continue analysis on this region.
For the haplotype analysis of the significant region on SSC14 with an effect on the prevalence of umbilical hernia, data was phased using the Beagle software (Browning & Browning, 2009) and analyzed using the Haploview software (Barrett et al., 2005), where pair-wise linkage disequilibrium (LD) measures (r2) are calculated between the different SNPs. Haplotype blocks were defined using the default confidence interval method in Haploview by Gabriel et al., 2001.
After haplotype analysis, it was found that the Norwegian data showed practically complete LD between the more than 70 genotyped SNP in this ~3 Mb region on SSC14, while the combined Belgian data show almost no LD. Based on the porcine reference genome build 10.2 (Groenen et a. 2012) three candidate genes were observed within the SSC14 QTL region. The three genes were sequenced using the ABI 3730 sequencer and SNPs were detected in two of those genes. Association studies were performed in the Norwegian data and highly significant single SNP- and haplotype associations were obtained.
However, it was found that the significant SNP in the Belgian data was not located close to this candidate mutation. This may indicate that either the mutation present in the locus on SSC14 is not present in the Belgian data, or that the mutation found is not the causative mutation, or that there are several mutations underlying this QTL. At the moment, the region surrounding the (nearly) significant SNP in the Belgian data is screened for interesting candidate genes.
Furthermore, the Norwegian data was dominated by the presence of two different haplotypes in this large LD block of ~ 3 Mb with frequencies of 0.54 and 0.33 in the population. The most common haplotype was associated with the defect phenotype (5 times more frequent in the defect animals) and Tag SNP within this defect haplotype can be implemented as genetic markers for MAS in the Norwegian population.
In the Belgian data this haplotype was not found. Due to the absence of LD in the Belgian data, this ~ 3 Mb region contained numerous less frequent haplotypes. The question arose whether the fact that there is no common haplotype is due to the lack of LD in the Belgian data or due to a population specific character of the QTL.
At the moment, the tag SNP panel that can be implemented as genetic markers for MAS in the Norwegian purebred lines is tested on crossbreds from the Norwegian lines. Also, a new experiment is set up in order to verify the Tag SNP panel in different lines. Currently, these samples are also genotyped for the Tag SNP panel. Lastly, allele frequencies for those markers are also tested in the lines of other SME partners as well (purebred and crossbred lines). After finishing these analyses, we will be able to say whether the Tag SNP panel is population specific or can be used in a broader perspective.
3.1. Cryptorchidism/scrotal hernia:
A very interesting region associated with cryptorchidism and scrotal hernia was found on the porcine X chromosome. However, after further analysis, this signal turned out to be an artifact due to the sex-specific character of both traits.
When the autosomes are screened for interesting regions affecting scrotal hernia and/or cryptorchidism, it becomes apparent that no SNP reach the significance level, although several SNP are highly suggestive. The only problem is that we cannot confirm them with the data set at hand. For this, we need to increase our data set. At the moment, we are doing just that by selecting more animals for the study and add other cohorts from different research groups around the world in order to perform a meta-analysis.
During this project, we already started selecting and genotyping more animals from the populations used in our project. However, these analyses are not finalized yet.
In the meanwhile, new cooperations are being established in order to be able to perform a meta-analysis
3.2. Umbilical hernia
A Tag SNP panel is available for selection against the prevalence of umbilical hernia. For the Norwegian lines, the tag SNP for umbilical hernia will be implemented in their breeding programs. At the moment, the Tag SNP Panel is tested in other lines (crossbreds and purebreds). When results are good, these tag SNP are also going to be implemented in their breeding programs. Furthermore, significant SNP in the two positional candidate genes are evaluated for possible functionality and the region surrounding the nearly significant SNP in the Belgian data is screened for possible interesting candidate genes. This way we hope to find a causative mutation for the susceptibility for umbilical hernia in pigs.
1. Socio-economic and scientific impact
1.1. Socio-economic impact
The overall socio-economic impact of PIGENDEF is the improvement of the sustainability of pork production and pig health and welfare by providing the tools for higher efficiency in pig breeding. Through genetic test for the susceptibility to congenital genetic diseases, the project will improve the efficiency and profitability of animal production, and therefore the competitiveness of the SME partners, within the overall framework of European policies on sustainability and on animal health and welfare.
The agriculture – and especially pork - sector is vitally important to the European economy as a whole. The pork industry has to provide the highest amount of meat within the EU, corresponding with at close to 35 kg/person/year or about 47 % of the total meat consumption Europe.
The steady lowering of European financial support mechanisms for the pork sector is influencing the margins in animal production and, subsequently, influences the pork business operators themselves. A competitive and innovative European pork production system is, however, the prerequisite of a functioning social and cultural context in Europe, and the pork products are an asset of the European food production system.
A major problem and a heavy economic burden for pig producers are congenital genetic diseases. In Europe the economic loss due to these defects is estimated at 200 million euros per year. Congenital genetic defects are quite common in swine and cover a range of conditions. The most important defects that occur in piglets are hernias (umbilical hernia and inguinal or scrotal hernias), cryptorchidism and splay legs, and to a lesser extent intersexuality, anal atresia and progressive ataxia and spastic paresis. All together they affect on average 3% of commercial pigs worldwide. Besides the economic loss, these defects also have a serious impact on animal welfare and health. Some of the defects cause direct piglet mortality, while others lead to culling of piglets.
With an economic loss of 200 million annually in Europe, a genetic test that decreases the incidence of a genetic defect with 25 to 50% leads to a substantial surplus value and broadening of the margins for the pig sector. As the existing genetic tests for the mutations in for instance Ryanodine receptor 1 and Insulin-like growth factor 2 are very successful, we can say that the pig sector is very open towards marker assisted selection. Therefore we expect that a tests for markers associated with the incidence of genetic defects will be very successful too and that breeding animals screened with these tests will be considerable. So, by valorization of the markers, found in this project, the SME partners can adapt to the needs of the sector and therefore create a surplus value for themselves and for the European pig production as a whole.
The surplus value for the SME partners can be situated in two different areas. In the first two years after development of a test, the test will exclusively be used in the breeding programs of the SME partners. This gives them the opportunity to profile themselves as a company that incorporates the latest technologies and provides a competitive advantage towards other pig producers and pig breeding companies. It is expected that herewith, 10% market share can be gained. After two years, additional income can be generated by commercialization of the test and selling the test to other, competing pig selection companies. This way, the project will improve the competitiveness of the SMEs involved on different levels, both economically and socially, but it will also improve pig health and sustainability of pig production and broaden the economical margins for pig producers.
1.1.1. Economic surplus
At the moment, a Tag SNP Panel is available to be used in the breeding program of our Norwegian partner. It concerns markers in linkage disequilibrium with the prevalence of umbilical hernia. In order to examine whether these markers are also applicable in other lines or populations, animals of three different populations are currently screened with this Tag SNP panel. It concerns crossbreds of the lines of the Norwegian partner, animals from the same breed as used by the Norwegian partner, but from a different population and Belgian purebred lines. Until this experiment is finished, the valorisation of the marker is realized by the screening of the purebred lines of Norsvin. The added value of purebred sires screened for these defects is estimated at 50 euro per animal while the added value of a sow is estimated at 1.5 euro. Depending on the initial frequency of the unfavorable alleles and the number of lines in which the technology can be applied this will result in a significant added value for this breeding company per year. However, when the same relationship exists between the tag SNP panel and the prevalence of umbilical hernia in the lines that are currently being screened, these tag SNP are also going to be implemented in the breeding programs of the other industrial partners, which will significantly increase the added value.
As we are still continuing the project as we speak, it is expected to find one or more causative mutations for umbilical hernia in the near future. This is feasible as significant SNP in two positional candidate genes are evaluated for possible functionality and the region surrounding the nearly significant SNP in the Belgian data is screened for possible interesting candidate genes. When a causative mutation is found, it will be patented. The patent costs are estimated at 7500 euro the first two years and 20000 euro for the year thereafter (examination procedures national stages). During the first two years the use of the genetic test will be limited to the partners of the project. From the third year on, the tests will be licensed to commercial labs and sold as a package or as separate tests. The same strategy has been applied for other genetic tests in pigs such as the Ryanodine receptor 1 gene, the RN gene and the IGF2 gene. As recessive diseases are very difficult to select out of the population, and the availability of the genetic test will enable to select out carrier animals, it is very likely that other breeding companies and /or artificial insemination centers will screen for these genes. If different causative mutations are found, the tests can be sold as a package for selection against umbilical hernia, resulting in a higher royalty per individual animal tested. A genetic test based on the causative mutation will also be applied in the own breeding animals of the SME partners.
In addition to mainly selling the Tag SNP panel or the causative mutations found, the markers in LD or the causative markers can also be used to incorporate in genomic selection in swine. Currently, so-called VIP SNP are implemented in genomic selection in order to improve the efficiency of this technique. VIP SNP are mainly causative mutations or polymorphisms known to be in strong LD with a trait of interest. In this perspective, the use of our Tag SNP Panel or a causative mutation for the prevalence of umbilical hernia in pigs can be marketed for use in genomic selection in different pig breeding companies.
The reason that this project had to be organized at the European level is that the companies involved cannot afford to develop and use this technology on their own. However, in the very competitive world of the pig breeding industries the competitiveness of the breeding companies can only be ensured if they are capable to use the newest technologies, as the larger companies do. Therefore, by organizing this large scale, European research we can increase in competitiveness of the SME partners. This will encourage the economic growth of the SMEs, allow them to envisage new markets (e.g. through licenses to commercial labs) and to dedicate additional staff involved in the technicity of marker assisted selection and sales persons.
1.1.2. Social surplus
In addition to economic losses, the incidence of congenital defects in the pig population also leads to a serious burden on sustainable agriculture. As it is impossible to describe the problems of all congenital defects, we will take the negative effects with relation to sustainable technological development for hernias as an example. These negative effects are twofold.
Firstly, there is an ecological downside to hernias. Pigs that develop hernias display a very unfavorable feed conversion ratio. Therefore, during their growing period, herniated pigs will have a higher ecological cost compared to healthy animals. On top of that, approximately 50% of the affected pigs dies before they reach slaughter age. These animals have had a certain ecological cost but they did not end up as a slaughter animal. In this view, a marker or tag SNP panel for umbilical hernia can decrease the nutrient emission in the pig sector.
Secondly, there are also some social issues to scrotal/inguinal hernias. The prevalence of hernias leads to poor health and a low welfare status of the affected animals. In a social environment where animal welfare is a very important issue, this poses a significant problem. Therefore, a decrease in the incidence of hernias would be a great surplus value. As selection for livestock improvement is widely accepted by the public, the development of a selection strategy against the prevalence of hernias will receive a positive social evaluation.
1.2. Scientific impact
The scientific impact of the PIGENDEF project can be situated in two domains:
1.2.1. Genetic factors underlying hernias and cryptorchidism in pigs
At the moment, three scientific papers on the results of the GWAS for umbilical and scrotal hernia and cryptorchidism are being prepared for publication in renowned peer-reviewed journals. During the course of this project, a large amount of data was processed and several new insights on the genetics of the three congenital genetic defects studied were discovered. These insights are of great value to the pig breeding and pig scientific community as they can serve the purpose of further unraveling the genetic basis of those defects. Furthermore, the candidate genes sequenced and the mutations found in those candidate genes can lead to a better biological and genomic understanding, both of the occurrence of congenital genetic defects and in a more general perspective.
1.2.2. The possibility of false positive results on the sex chromosomes when analyzing sex-specific traits
For scrotal hernia and cryptorchidism, a strong signal was coming from the boundaries of the pseudo-autosomal region from Belgian and Norwegian populations. Because the two diseases under study only affect male offspring the TDT test was recalculated to capture the signal coming from maternal transmission only. The signal coming from the pseudo-autosomal region was weakened after looking at maternal transmission only but nevertheless it was still significant. In order to fine map the region and be able to perform targeted and focused resequencing and functional analysis, three different experiments were set up simultaneously, amongst which a phasing experiment. For this, the pseudo-autosomal region was phased with Beagle (Browning & Browning, 2009) in order to find the associated haplotype. However, after performing the TDT test on the phased data, no significant results were obtained. Therefore, we had to classify the observed signal an artifact signal. The reason for this artifact is the fact that the region of interest was very near to the boundaries of the pseudo-autosomal region which could create such signals. As not much is known on the genetic/genomic analysis of the sex-chromosomes, this is very relevant information for the genomic scientific community, both in animals and humans.
2. Dissemination activities
The main dissemination activities can be subdivided in three classes: popular dissemination activities, contributions to (inter)national scientific congresses and publications in scientific peer-reviewed journals. Over the entire PIGENDEF project, the main dissemination activities have been:
2.1. Popular dissemination activities
2.1.1. A PIGENDEF website
In the very beginning of the project, a PIGENDEF public website with clear and concise messages was designed and implemented. On this PIGENDEF website (www.PIGENDEF.org) information is given on news from the PIGENDEF project, the project aims, project consortium, project structure and work packages included in the PIGENDEF project. Furthermore, information about the PIGENDEF partners and the research performed and articles published within the framework of the PIGENDEF project is given. The website is updated regularly. As we encountered some problems with the web hosting, we created a second address (www.PIGENDEF.eu). Both addresses are active. The website has been promoted on several (inter)national scientific congresses.
Now the project is finished, the PIGENDEF website will remain active and will still be updated when necessary. This means that when new results arise or when a test is commercially available, this will be posted on our website. This way, our website will remain the premium dissemination tool for the PIGENDEF project. Furthermore, links to the PIGENDEF website will remain on the websites of the different RTD and SME partners. This way, the public, stakeholders and non-participating SME can easily find access to the PIGENDEF website.
2.1.2. Contributions to the popular press
• Press releases by KATHOLIEKE UNIVERSITEIT LEUVEN: Towards marker assisted selection against genetic defects in pigs (24/09/2012)
• Press releases by NORSVIN FORENING: Genetisk seleksjon for reduksjon av medfødte defekter hos gris (18/10/2012)
• Articles published in the popular press by RATTLEROW SEGHERS NV: SELECTIE TEGEN ERFELIJKE GEBREKEN (13/11/2012)
• Press releases by NORDIC GENETICS AB: Genetisk selection för reducering av medfödda missbildningar hos grisar (14/11/2012)
• Web sites/Applications by RATTLEROW-SEGHERS NV: www.wattagnet.com- Pigendef project aims at selection against genetic defects in pigs (14/02/2013) USA Industry - Medias All
• Flyers by RATTLEROW-SEGHERS NV: Selectie tegen erfelijke gebreken (14/02/2013)
• Flyers by RATTLEROW-SEGHERS NV: Selection against genetic defects in pigs (14/02/2013)
• Flyers by RATTLEROW-SEGHERS NV: Селекция срещу наследствени дефекти (18/02/2013)
• Flyers by RATTLEROW-SEGHERS NV: SELECTIE IMPOTRIVA DEFECTELOR GENETICE LA PORCINE (05/03/2013)
• Press releases by SELECCION BATALLE SA: Hacia una selección asistida por (05/03/2013)
2.2. Contributions to (inter)national scientific congresses
Within the framework of the project, contributions were made to renowned (inter)national congresses.
• A whole genome association study using the 60K porcine SNP beadchip reveals candidate susceptibility (27/08/2011) Stinckens A., Janssens S., Spincemaille G. & Buys N., 62nd annual meeting of the European Federation of Animal Science, Stavanger, Norway, poster
• Genome-wide association study for inguinal and umbilical hernias using the 60K Porcine SNParray (28/08/2011) Grindflek E., Hansen M.H.S. Hamland H. & Lien S, 62nd annual meeting of the European Federation of Animal Science, Stavanger, Norway, oral presentation
• Aangeboren genetische afwijkingen bij biggen (26/10/2011) Stinckens A., 3e Vlaamse Fokkerijdag, Leuven, Belgium, oral presentation
• Promising loci for the susceptibility to both scrotal hernia and cryptorchidism in pigs (19/07/2012) Stinckens A. & Buys N., 33rd Conference of the International Society of Animal Genetics, Cairns, Australia, oral presentation
• Promising loci for the susceptibility to both scrotal hernia and cryptorchidism in pigs (19/07/2012) Stinckens A. & Buys N., 33rd Conference of the International Society of Animal Genetics, Cairns, Australia, poster
2.3. Scientific publications
At the moment, 4 scientific papers for peer-reviewed journals are in preparation and are envisioned to be finished soon:
• A calculation on the risk of false positive signals on the sex-chromosomes when performing a genome wide association test on a sex-related trait
• The description of the results obtained on the research for the susceptibility for umbilical hernia in pigs (2 papers)
• The description of the results obtained on the research for the susceptibility for scrotal hernia and cryptorchidism in pigs
3. Exploitation of results
The exploitation of the results of the PIGENDEF project was highly dependent on the results found. In the beginning of the project, three different scenarios were foreseen for the exploitation: one scenario where no genetic markers were found, one where at least one marker in linkage disequilibrium with one of the studied genetic defects was found and the last where at least one causative mutation affecting one of the congenital defects studied can be found.
At the moment, the second scenario has to be applied, since a tag SNP panel for umbilical hernia was designed for the Norwegian lines. However, as several interesting SNP are being studied at the moment, the possibility of finding a causative mutation within a limited time period cannot be excluded. Therefore, here, both the current exploitation activities and possible future exploitation activities are cited.
3.1. A Tag SNP panel in linkage disequilibrium with the susceptibility to umbilical hernia in pigs
A the moment, a Tag SNP panel in linkage disequilibrium with the prevalence of umbilical hernia is available for use in the Norwegian lines. Currently, the panel can be used in the Norsvin selection program, however for other lines and populations the use and commercialisation will comprise two steps:
3.1.1. Step 1:
For the Belgian purebred lines, crossbred lines (Norwegian and Belgian) and the animals involved in the newly set-up experiment, it will be checked whether these markers also segregate and if the same relationship exists between the tag SNP panel and the prevalence of umbilical hernia in these lines.
3.1.2. Step 2:
If the located markers show the same effect in the newly tested lines as in the Norwegian lines, these markers can also be used in marker assisted selection against the prevalence of umbilical hernia in these lines. However, a patent can only be filed if the mutation is shown to be the causative mutation. Therefore, the valorisation perspectives of this strategy are mainly situated at the level of the competitive benefit of the industrial partners in this project towards other pig selection companies. In order to make the implementation of the markers into the breeding programmes of the different industrial partners as easy as possible, a training session was organised by the PIGENDEF consortium. During this training session, representatives of all SME partners were present and scientist of CIGENE, Ulg and KU Leuven presented the results obtained during the project and revealed possible MAS strategies with the markers found in this project.
As genomic selection becomes increasingly important in pig breeding, it is also important to mention that the Tag SNP panel can be incorporated in the genomic selection strategy when they segregate in the line of interest. This approach is called genomic selection with use of VIP (very important) SNP. This way, the tag SNP panel will not only serve the purpose of the settled technique of Marker assisted selection strategies but it can also be part of a very progressive technique like genomic selection.
For the purebred Norsvin lines, the tag SNP panel for umbilical hernia will be implemented in their breeding programs. Before the other partners can implement the tag SNP panel in their breeding program, results of the segregation and association analysis have to be awaited. When results are good however, these tag SNP are also going to be implemented in the breeding programs of the other industrial partners.
3.2. A causative mutation affecting the susceptibility to umbilical hernia in pigs
Although until now no causative mutations were found for any of the genetic disorders examined, given the current results, it is envisioned that in a rather limited timeframe a causative mutation for the susceptibility for umbilical hernia can be located. At the moment, significant SNP in the two positional candidate genes are evaluated for possible functionality and the region surrounding the nearly significant SNP in the Belgian data is screened for possible interesting candidate genes. This way we hope to find a causative mutation for the susceptibility for umbilical hernia in pigs. The plan to evaluate the joint umbilical hernia data into more detail after the end of the project in order to locate a possible causative mutation on SSC14 was proposed at the training session in Leuven and these plans were approved by the SME partners.
Based on literature, chances that a causative mutation can be found for umbilical hernia are relatively large. From literature, it was found that, using the bovine SNP chip, causative mutations were found for three recessive monogenic defects in cattle (Charlier et al. 2008). Although it is generally accepted that the localisation of causative mutations for monogenic traits, like the traits researched in cattle, is simpler compared to the research in polygenic traits, like congenital defects, it should be stated that the prevalence of umbilical hernia is probably due to a relative small number of genes, amongst which some major genes. This is also confirmed by the GWAS results from this project, in which only a limited number of significant signals can be found for umbilical hernia. This largely increases the chances of unravelling (a part of) the genetic background of umbilical hernia.
If one or more causative mutations for the incidence of the studied congenital disorders can be found, a unique test for the prevalence of these defects can be developed. This test can then be used in every line in which the marker segregates and in every selection program. Also, a patent will be filed for this mutation. The owners of the patent rights will be Rattlerow-Seghers and Norsvin. The inventors will be the same, but also including the RTD partners. The other industrial partners will be given a license to use the markers in their own breeding programs. Publication of results will be postponed until after the patent is filed. The first two years after filing the patent, the test will only be used in the selection programs of the SME partners. This provides them the chance to profile themselves as a company that implements the newest molecular techniques and also provides a competitive benefit with regard to other pig selection companies/pig breeding companies. Again, in order to make the implementation of the markers into the breeding programmes of the different SME partners as easy as possible, training sessions will be held. After two years, sublicenses will be sold to commercial laboratories which will place the developed test at the disposal of interested third parties like competing pig selection firms. At the moment, the causative mutations for the three recessive monogenic disorders in cattle, described by Charlier et al. (2008), are also commercialised in a similar manner. Also, in pigs, several markers including the mutations in Ryanodine receptor 1 and IGF2, are successfully commercialised in the same way.
Furthermore, as also mentioned with the Tag SNP panel, the causative mutation found can be used in genomic selection as a VIP SNP.
List of Websites: