Final Report Summary - HORSEGENE (GENOMIC TOOLS FOR BREEDING AGAINST HERITABLE DISEASES IN HORSES)
Executive Summary:
The HORSEGENE project was initiated by SME’s and breeding organizations in order to find solutions to breed against 3 common diseases in horses. Four research groups in Europe were included in the consortium and state-of-the-art genomic technologies were applied for the identification of genetic markers for those heritable diseases in horses. The consortium of 16 partners conducted research on Insect Bite Hypersensitivity, Osteochondrosis and Chronic Progressive Lympoedema in different populations of horses (mainly on Exmoor and Shetland ponies, Icelandic horse, Friesian horse, Belgian Warmblood and Belgian Draught horse).
The occurrence of osteochondrosis (OC), insect bite hypersensitivity (IBH) and chronic progressive lymphedema (CPL) is a significant problem for horse owners and breeders and leads to poor animal welfare and severe economic losses. For each disease, a genetic component to disease susceptibility has been confirmed and several research groups in Europe have independently carried out small to medium-sized studies in order to find the causative genetic mechanisms underlying the disorder. However, the mode of inheritance for these diseases appears to be complex, and no genes or mutations were identified until then.
In the initial phase, the recording of the traits was as much as possible harmonized by agreeing on a single common protocol for data collection. In that way, phenotypes were made uniform across breeds. The same effort was made with respect to DNA extraction and quality control after genotyping.
An important deliverable of the project is the implementation of a web-based database that holds both the phenotypes and genotypes generated within HORSEGENE but also serves as a platform for other equine genetic research.
In the project many Genome Wide Association-studies were performed both within and also across populations using the latest Axiom 670 K SNP chip. These studies lead to the general conclusion that possibly the breed specificity in the expression of disease sensitivity is more important than previously thought off. Using the SNP approach, association were mostly breed specific and across breed association was inconclusive.
Analysis also pointed at associated regions in the horse genome containing genes that may be implicated in one of the diseases. This information is further explored and may support preventive and curative medicine because biological pathways become apparent.
For CPL and IBH, marker panels were proposed which explain 2.9% and 20% of the phenotypic variance respectively and it was theoretically shown that these panels could improve genetic progress when selecting against disease susceptibility.
In the project, a new kind of genetic variation was explored, namely Copy Number Variation. An association was found between IBH and Cop Number Variants warranting this type of analysis in equine genetics.
Updated information can be obtained from www.horsegene.eu .
Project Context and Objectives:
The main objective of the HORSEGENE project is to use reliable state-of-the-art genomic technologies for the rapid identification of genetic markers for heritable diseases in horses and so facilitate effective genomics based selection against disease susceptibility. In order to achieve this goal, a project consortium was set up comprising four RTD partners, 10 horse related partners (studbooks, breeders) and one certified laboratory specialized in DNA technologies.
The diseases of particular focus to the HORSEGENE project are osteochondrosis (OC), Insect bite hypersensitivity (IBH) and chronic progressive lymphedema (CPL). OC is a disturbance in the physiological process of endochondral ossification that occurs in young, growing horses. The resulting irregular ossification leads to formation of primary lesions which develop as retained cores of cartilage adjacent to, and extending into, the subchondral bone. Primary lesions may progress to form focal necrotic areas, cartilage flaps and eventually loose fragments. This progression appears to be associated with biomechanical trauma.
IBH is an allergic reaction of horses to the bite of certain Culicoides species and sometimes black flies or stable flies. Worldwide, it is the most common allergic skin disease in horses and it is characterized by a severe itch and dermatitis affecting primarily the mane and tail, but in severe cases it’s also seen on the entire body of the horse. Affected horses often develop, mainly in summer, severe skin lesions due to self-mutilation in an attempt to alleviate the itch. Severely affected horses are not suitable for riding or showing purposes and are sometimes euthanized.
Lastly, CPL is a disorder described in three related draught horse breeds: the Shire, the Clydesdale and the Belgian draught horse. In the latter, it affects over 70% of the population. The main cause is a systemic failure of the lymph system and the skin elastic network. Cutaneous changes and deformations of the lower limbs occur as a result of reduced lymphatic drainage. Clinical symptoms mostly aggravate through life, often ending with severe disability which frequently justifies euthanasia. Horses that suffer from this condition show progressive swelling of the lower limbs, associated with the development of thick skin folds, marked dermatitis, hyperkeratosis, dermal diffuse and nodular fibrosis.
The occurrence of these diseases is a significant problem for horse owners and breeders and leads to poor animal welfare and severe economic losses. The initiative of this project is mainly based on the strong wish from the studbooks to obtain a genetic alternative to the more traditional breeding methods which are often not feasible in the smaller breeds.
For each disease, a genetic component to disease susceptibility has been confirmed and several research groups in Europe have independently carried out small to medium-sized studies in order to find the causative genetic mechanisms underlying the disorder. However, the mode of inheritance for these diseases appears to be complex, and no genes or mutations have been identified yet, suggesting that analysis of much larger sample sizes is required or across datasets.
This project brings together the efforts of European research groups that are currently working on the genetics of disease susceptibility in horses. This collaboration will enable the groups to pool data for meta-analysis leading to more rapid progress in the identification of the genes and suitable gene markers for use in genomic selection against disease susceptibility. The SME associations, in close collaboration with the RTD performers, will develop breeding strategies that are optimized to their specific situation. If genetic markers are found, these markers will be implemented in the breeding programs of the different SME associations and can be commercialized. This will provide competitive advantage to the SME partners represented by the associations.
Project Results:
The HORSEGENE project was initiated by SME's and breeding organizations in order to find solutions to breed against 3 common diseases in horses. Four research groups in Europe were included in the consortium and applied state-of-the-art genomic technologies for the identification of genetic markers for those heritable diseases in horses. The consortium of 16 partners conducted research on Insect Bite Hypersensitivity (IBH), Osteochondrosis (OC) and Chronic Progressive Lymphedema (CPL) in different populations of horses (mainly Exmoor and Shetland ponies, Icelandic horse, Friesian horse, Belgian Warmblood and Belgian Draught horse).
The occurrence of OC, IBH and CPL is a significant problem for horse owners and breeders and leads to poor animal welfare and severe economic losses. For each disease, a genetic component to disease susceptibility has been confirmed and several research groups in Europe have independently carried out small to medium-sized studies in order to find causative genetic mechanisms underlying the disorder. However, the mode of inheritance appears to be complex, and no genes or mutations were identified until then.
In the initial phase, the recording of the traits was as much as possible harmonized by agreeing on a common protocol for data collection. Such common protocols were agreed upon for the 3 diseases (OC and IBH, CPL only one breed). Based on these protocols, data and biological samples were selected for genotyping purposes.
A common protocol was established also for DNA-extraction and genotyping using the Affymetrix 670 K High-density SNP-chip. The 670 K genotypes that were used both for GWAS and other genomic studies could be used for imputation of samples for which only a 50 K or a 70 K genotype was present. A database was conceived consisting of 2 parts (http://horsegene.interbull.org/ ). The first part contains an overview of research items for which collaboration and/or samples are available or requested. This part of the HORSEGENE database can be considered as a portal that will promote future equine research. The second part contains currently phenotypes and genotypes that were actually generated in HORSEGENE and that will be shared upon request. Access to this web-based database can be gained by any interested person.
In order to be able to efficiently analyze the genotyping data, a strategy was designed to optimally and accurately perform GWAS analysis in horses. For this, a workshop was organized for all members of the HorseGene consortium in November 2014 in Wageningen, The Netherlands.
In the project Genome Wide association studies were performed both within and across populations using the latest Axiom 670 K SNP Chip.
Genome-wide association (GWA) analysis was performed using OCD as a case/control phenotype and yielded a number of significant SNP’s but confirmatory research is currently underway so that a conclusive GWA for OC could not be reached by the end of the project.
GWAS performed on CPL as a binary trait revealed a number of significant SNP that were selected to be included in a panel explaining 2.9% of the variance.
IBH was recorded in several of the horse breeds in the consortium with different phenotypes. Most common was a binary or categorical (severity) scale which was used in case/control designs and subsequently in GWA studies. Additionally, IgE titers were measured in a number of horse breeds, after a challenge with allergens (derived from Culicoïdes or recombinant). It can be stated that the genomic association with IBH seemed strongly breed dependent which reduces the changes to develop a genetic test that is applicable over breeds. It also stresses the importance of within breed studies with good resolution. Results from the breed-specific genome-wide association study in Icelandic horses and Shetland ponies (only on 50k resolution) and a meta-analysis were not convincing and yielded no associated SNP’s. Probably the limited number of individuals in these two breeds and/or the lower density of SNPs were insufficient to yield significant associations.
The outcomes for the 3 diseases confirmed that a) a large number of genomic regions was involved (case of IBH) or b) only suggestive genomic variants could be detected using the current SNP chips, even at relatively high resolution which was the case of OC and to a lesser extent CPL. Both outcomes made a study of causal variants and fine mapping of regions quite inappropriate. With low % of variance explained by single or multiple SNP’s, detection of causal variants is technically very difficult and requires a lot of data. Therefore the consortium decided to investigate if other types of genetic variation might be involved in these diseases. Using the 670k chip genotypes, it is possible to study structural variation caused by Copy Number variation (CNV) of genes or genomic regions. CNV seemed to be involved in IBH sensitivity.
At the end of the project, 2 sets of SNP markers were included in 2 panels: one to select against CPL in draught horses and one to reduce IBH in Friesian horses. For CPL and IBH, marker panels explained approximately 2.9 % and 20% of the phenotypic variance and it was theoretically shown (by simulation) that these panels could improve genetic progress when selecting against diseases susceptibility compared to the current selection methods. A commercial development was however not realized. Main difficulty was the low % of phenotypic variance explained which would result in low accuracy of the tool and thus of the EBVs. The trade-off between accurate EBVs relatively late in life (based on offspring records) vs less accurate EBVs but available earlier in life, is an ongoing debate in practical horse breeding and has been revisited with the results of this project. As a result breeding associations agree that the topics should be taken up by their members. They fear that selection using an imprecise test applied early in life would not be accepted by breeders of horses.
At the start of the project, the odds of finding causative mutations were fair. All diseases showed a genetic component, present in more than one breed. This are usually indicators for a common genetic cause. During research, the breed specificity became more pronounced and only a limited number of common regions could be identified. This indicates that breed specific studies might be preferred and/or the possibilities of genomic selection should be explored.
The consortium used the 650K SNP-array that was only recently operational by that time. There was hardly any experience with this array. As we started the project it was believed that a genomic evaluation across-breeds could identify regions that harbor genes being involved in the diseases. In particular with high-density SNP-arrays and the involved breeds being a blend of highly inbred and less inbred breeds and populations. Results indicated too low LD to gain from an across-breed analysis in identifying genomic regions. This is in accordance with recent research in other species and it is generally believed that even higher density is required or even full genome sequencing. The consortium decided to engage in a novel type of analysis (Copy number variation) which, at the moment of writing, gave new scientific insights but applications are still ahead.
Nevertheless, diseases remain a major concern. Especially high prevalent diseases are a heavy burden, such as the three diseases in this project. For OC clinical prevalence is from 5 to 25%, but radiographic signs of abnormal development can be 70%, with large financial implications. In the Netherlands, the annual loss by OC in Warmbloods is estimated at 10 million euro (3000 affected foals and a loss of 3300 euro/foal). Moreover, there is an indirect cost of the disease, as horses with OC are not approved, which excludes a large part of the gene pool. The financial repercussions of IBH and CPL are not that severe, but they have a clear financial and animal welfare impact by premature culling of the horses.
The project has created awareness by affecting the viewpoints of SME partners towards genomics and genomic tools. They believe in investing in this technology and follow-up-projects are being prepared. Furthermore, they are aware of the need for uniform and extended phenotyping. The same mind-switch was observed in government officials during the final workshop. This is reassuring because the horse sector has been more hesitating to pick up new technologies compared to the typical livestock species.
Dissemination of knowledge has been focused mostly on the horse sector and not so much towards the general public. Nevertheless, the video-clip is available for wide distribution.
The scientific findings of the project are visible as peer reviewed articles and conference contributions. The cooperation between industry and RTD partners has created a platform that offers a unique resource (samples, phenotypes and genotypes) that is exploited by the members of the consortium but allows for new contributions.
The project has a significant leverage in terms of research as at least 3 proposals are or will be submitted nationally and further international cooperation is under discussion. This will ensure further support and development of the important horse industry in the EU.
Updated information can be obtained from www.horsegene.eu
Potential Impact:
The primary objective of the consortium have been largely reached. At the start of the project, the odds of being able to find causative mutations for one (or more) of the diseases were fair. All three diseases (or disease susceptibilities) showed a genetic component, present in more than one breed. However, as research progressed, the breed specificity become more pronounced and a single (or limited number of), common region(s) for any of the diseases could not be identified.
This breed specificity is an important result because it may indicate that breed specific studies might be the preferred route. Or, alternatively, genomic selection should be explored more.
The consortium made use of the latest state of the art technique and used therefore the 650K SNP-array that was only recently operational by that time. At that time there was hardly any experience with this array. As we started the project it was generally believed in the animal genetic community that a genomic evaluation across-breeds would be a very appropriate strategy to identify regions at the genome that harbours genes that are involved in the genetics of the diseases under study. In particular when high-density SNP-arrays were applied and when the breeds involved are a blend of highly inbred breeds (to identify number and location of involved regions) and less inbred populations (to narrow down the size of the identified regions) and therefore we were eager on using the latest 650K SNP array. The results of the GWA-studies on the 650K SNP-array indicated however that LD was too low, even using the 650K, to gain from an across-breed analysis in identifying genomic regions involved. This finding is in accordance with research in other species and it is now generally believed that much higher density is required or even sequencing to perform across breed analysis. The consortium decided therefor to engage in a novel type of analysis of the genotypes (Copy number variation) which, at the moment of writing, gave new scientific insights but applications are still ahead.
At the end of the project, 2 sets of SNP markers were selected by a commercial lab to be included in 2 panels: one to select against Chronic Progressive Lympoedema in draught horses and one to reduce the prevalence of IBH in Friesian horses. After discussions with breeding associations a commercial development was not realized. The main difficulty was the relatively low % of phenotypic variance explained by the tests which would result in low accuracy of the tool and thus low accuracy of the estimated breeding values. The trade-off between accurate breeding values relative late in life vs less accurate breeding values but available earlier in life, is a debate that is going on for a longer time in practical horse breeding and has been revisited with the results of this project. As a result of the discussions in this project some of the breeding associations agree that the topics should be taken up by their members. Breeding organisations feared that selection of breeding animals using an imprecise test would not be accepted by horse breeders.
Nevertheless, diseases remain one of the main concerns in horse breeding and keeping. Especially diseases that have a high prevalence in some or most horse breeds are a heavy burden. The three diseases studied in this project fall under this category. For OC for instance, clinical prevalence is usually between 5 and 25% in a given horse population, but radiographic signs of abnormal development can be as high as 70% in certain groups. Prevalence and incidence of this disease are severe with large financial implications for owners and breeders as a consequence. An example: In the Netherlands, the annual loss caused by OC in Warmbloods was estimated at 10 million euro (3000 affected foals and a loss of 3300 euro/foal). Moreover, apart from the direct economic loss, there is an even larger indirect cost of the disease, as many studbooks will not approve horses with evidence of OC, which implies the exclusion of a large part of the gene pool.
Although the financial repercussions of the other two conditions (IBH and CPL) are not that severe, these diseases also have a clear financial impact and largely affect animal welfare as they sometimes lead to a premature culling of the horses through euthanasia.
In terms of creating awareness, the project has strongly affected the viewpoints of the SME partners in the consortium towards genomics and genomic tools. The belief has grown in most studbooks that they should further invest in this technology and follow-projects are being prepared. Furthermore, SME-partners are now definitely well aware of the need for uniform and extended phenotyping. The same mind-switch was observed in government officials during the workshop at the end of the project.
This observation is re-assuring because the horse breeding sector has been more hesitating to pick up new technologies compared to the typical agricultural livestock species (dairy cattle, swine, poultry, ...).
Dissemination of knowledge has been focused mostly on horse breeding or horse owner-organisations and not so much towards the general public. Nevertheless, the video-clip is available for wide distribution.
The scientific findings of the project are ample and visible as peer reviewed articles and contributions to conferences. The cooperaton between the industry and RTD partners has created a very usefull platform that offers a unique resource (samples, phenotypes and genotypes) that is exploited further by the members of the consortium but allows for new contributions.
The project has a significant leverage in terms of research as at least 3 proposals are or will be submitted nationally (e.g. in Belgium, Sweden, Netherlands) and further international cooperation is under discussion. This will ensure further support and development of the important horse industry in the EU.
List of Websites:
http://www.horsegene.eu
The HORSEGENE project was initiated by SME’s and breeding organizations in order to find solutions to breed against 3 common diseases in horses. Four research groups in Europe were included in the consortium and state-of-the-art genomic technologies were applied for the identification of genetic markers for those heritable diseases in horses. The consortium of 16 partners conducted research on Insect Bite Hypersensitivity, Osteochondrosis and Chronic Progressive Lympoedema in different populations of horses (mainly on Exmoor and Shetland ponies, Icelandic horse, Friesian horse, Belgian Warmblood and Belgian Draught horse).
The occurrence of osteochondrosis (OC), insect bite hypersensitivity (IBH) and chronic progressive lymphedema (CPL) is a significant problem for horse owners and breeders and leads to poor animal welfare and severe economic losses. For each disease, a genetic component to disease susceptibility has been confirmed and several research groups in Europe have independently carried out small to medium-sized studies in order to find the causative genetic mechanisms underlying the disorder. However, the mode of inheritance for these diseases appears to be complex, and no genes or mutations were identified until then.
In the initial phase, the recording of the traits was as much as possible harmonized by agreeing on a single common protocol for data collection. In that way, phenotypes were made uniform across breeds. The same effort was made with respect to DNA extraction and quality control after genotyping.
An important deliverable of the project is the implementation of a web-based database that holds both the phenotypes and genotypes generated within HORSEGENE but also serves as a platform for other equine genetic research.
In the project many Genome Wide Association-studies were performed both within and also across populations using the latest Axiom 670 K SNP chip. These studies lead to the general conclusion that possibly the breed specificity in the expression of disease sensitivity is more important than previously thought off. Using the SNP approach, association were mostly breed specific and across breed association was inconclusive.
Analysis also pointed at associated regions in the horse genome containing genes that may be implicated in one of the diseases. This information is further explored and may support preventive and curative medicine because biological pathways become apparent.
For CPL and IBH, marker panels were proposed which explain 2.9% and 20% of the phenotypic variance respectively and it was theoretically shown that these panels could improve genetic progress when selecting against disease susceptibility.
In the project, a new kind of genetic variation was explored, namely Copy Number Variation. An association was found between IBH and Cop Number Variants warranting this type of analysis in equine genetics.
Updated information can be obtained from www.horsegene.eu .
Project Context and Objectives:
The main objective of the HORSEGENE project is to use reliable state-of-the-art genomic technologies for the rapid identification of genetic markers for heritable diseases in horses and so facilitate effective genomics based selection against disease susceptibility. In order to achieve this goal, a project consortium was set up comprising four RTD partners, 10 horse related partners (studbooks, breeders) and one certified laboratory specialized in DNA technologies.
The diseases of particular focus to the HORSEGENE project are osteochondrosis (OC), Insect bite hypersensitivity (IBH) and chronic progressive lymphedema (CPL). OC is a disturbance in the physiological process of endochondral ossification that occurs in young, growing horses. The resulting irregular ossification leads to formation of primary lesions which develop as retained cores of cartilage adjacent to, and extending into, the subchondral bone. Primary lesions may progress to form focal necrotic areas, cartilage flaps and eventually loose fragments. This progression appears to be associated with biomechanical trauma.
IBH is an allergic reaction of horses to the bite of certain Culicoides species and sometimes black flies or stable flies. Worldwide, it is the most common allergic skin disease in horses and it is characterized by a severe itch and dermatitis affecting primarily the mane and tail, but in severe cases it’s also seen on the entire body of the horse. Affected horses often develop, mainly in summer, severe skin lesions due to self-mutilation in an attempt to alleviate the itch. Severely affected horses are not suitable for riding or showing purposes and are sometimes euthanized.
Lastly, CPL is a disorder described in three related draught horse breeds: the Shire, the Clydesdale and the Belgian draught horse. In the latter, it affects over 70% of the population. The main cause is a systemic failure of the lymph system and the skin elastic network. Cutaneous changes and deformations of the lower limbs occur as a result of reduced lymphatic drainage. Clinical symptoms mostly aggravate through life, often ending with severe disability which frequently justifies euthanasia. Horses that suffer from this condition show progressive swelling of the lower limbs, associated with the development of thick skin folds, marked dermatitis, hyperkeratosis, dermal diffuse and nodular fibrosis.
The occurrence of these diseases is a significant problem for horse owners and breeders and leads to poor animal welfare and severe economic losses. The initiative of this project is mainly based on the strong wish from the studbooks to obtain a genetic alternative to the more traditional breeding methods which are often not feasible in the smaller breeds.
For each disease, a genetic component to disease susceptibility has been confirmed and several research groups in Europe have independently carried out small to medium-sized studies in order to find the causative genetic mechanisms underlying the disorder. However, the mode of inheritance for these diseases appears to be complex, and no genes or mutations have been identified yet, suggesting that analysis of much larger sample sizes is required or across datasets.
This project brings together the efforts of European research groups that are currently working on the genetics of disease susceptibility in horses. This collaboration will enable the groups to pool data for meta-analysis leading to more rapid progress in the identification of the genes and suitable gene markers for use in genomic selection against disease susceptibility. The SME associations, in close collaboration with the RTD performers, will develop breeding strategies that are optimized to their specific situation. If genetic markers are found, these markers will be implemented in the breeding programs of the different SME associations and can be commercialized. This will provide competitive advantage to the SME partners represented by the associations.
Project Results:
The HORSEGENE project was initiated by SME's and breeding organizations in order to find solutions to breed against 3 common diseases in horses. Four research groups in Europe were included in the consortium and applied state-of-the-art genomic technologies for the identification of genetic markers for those heritable diseases in horses. The consortium of 16 partners conducted research on Insect Bite Hypersensitivity (IBH), Osteochondrosis (OC) and Chronic Progressive Lymphedema (CPL) in different populations of horses (mainly Exmoor and Shetland ponies, Icelandic horse, Friesian horse, Belgian Warmblood and Belgian Draught horse).
The occurrence of OC, IBH and CPL is a significant problem for horse owners and breeders and leads to poor animal welfare and severe economic losses. For each disease, a genetic component to disease susceptibility has been confirmed and several research groups in Europe have independently carried out small to medium-sized studies in order to find causative genetic mechanisms underlying the disorder. However, the mode of inheritance appears to be complex, and no genes or mutations were identified until then.
In the initial phase, the recording of the traits was as much as possible harmonized by agreeing on a common protocol for data collection. Such common protocols were agreed upon for the 3 diseases (OC and IBH, CPL only one breed). Based on these protocols, data and biological samples were selected for genotyping purposes.
A common protocol was established also for DNA-extraction and genotyping using the Affymetrix 670 K High-density SNP-chip. The 670 K genotypes that were used both for GWAS and other genomic studies could be used for imputation of samples for which only a 50 K or a 70 K genotype was present. A database was conceived consisting of 2 parts (http://horsegene.interbull.org/ ). The first part contains an overview of research items for which collaboration and/or samples are available or requested. This part of the HORSEGENE database can be considered as a portal that will promote future equine research. The second part contains currently phenotypes and genotypes that were actually generated in HORSEGENE and that will be shared upon request. Access to this web-based database can be gained by any interested person.
In order to be able to efficiently analyze the genotyping data, a strategy was designed to optimally and accurately perform GWAS analysis in horses. For this, a workshop was organized for all members of the HorseGene consortium in November 2014 in Wageningen, The Netherlands.
In the project Genome Wide association studies were performed both within and across populations using the latest Axiom 670 K SNP Chip.
Genome-wide association (GWA) analysis was performed using OCD as a case/control phenotype and yielded a number of significant SNP’s but confirmatory research is currently underway so that a conclusive GWA for OC could not be reached by the end of the project.
GWAS performed on CPL as a binary trait revealed a number of significant SNP that were selected to be included in a panel explaining 2.9% of the variance.
IBH was recorded in several of the horse breeds in the consortium with different phenotypes. Most common was a binary or categorical (severity) scale which was used in case/control designs and subsequently in GWA studies. Additionally, IgE titers were measured in a number of horse breeds, after a challenge with allergens (derived from Culicoïdes or recombinant). It can be stated that the genomic association with IBH seemed strongly breed dependent which reduces the changes to develop a genetic test that is applicable over breeds. It also stresses the importance of within breed studies with good resolution. Results from the breed-specific genome-wide association study in Icelandic horses and Shetland ponies (only on 50k resolution) and a meta-analysis were not convincing and yielded no associated SNP’s. Probably the limited number of individuals in these two breeds and/or the lower density of SNPs were insufficient to yield significant associations.
The outcomes for the 3 diseases confirmed that a) a large number of genomic regions was involved (case of IBH) or b) only suggestive genomic variants could be detected using the current SNP chips, even at relatively high resolution which was the case of OC and to a lesser extent CPL. Both outcomes made a study of causal variants and fine mapping of regions quite inappropriate. With low % of variance explained by single or multiple SNP’s, detection of causal variants is technically very difficult and requires a lot of data. Therefore the consortium decided to investigate if other types of genetic variation might be involved in these diseases. Using the 670k chip genotypes, it is possible to study structural variation caused by Copy Number variation (CNV) of genes or genomic regions. CNV seemed to be involved in IBH sensitivity.
At the end of the project, 2 sets of SNP markers were included in 2 panels: one to select against CPL in draught horses and one to reduce IBH in Friesian horses. For CPL and IBH, marker panels explained approximately 2.9 % and 20% of the phenotypic variance and it was theoretically shown (by simulation) that these panels could improve genetic progress when selecting against diseases susceptibility compared to the current selection methods. A commercial development was however not realized. Main difficulty was the low % of phenotypic variance explained which would result in low accuracy of the tool and thus of the EBVs. The trade-off between accurate EBVs relatively late in life (based on offspring records) vs less accurate EBVs but available earlier in life, is an ongoing debate in practical horse breeding and has been revisited with the results of this project. As a result breeding associations agree that the topics should be taken up by their members. They fear that selection using an imprecise test applied early in life would not be accepted by breeders of horses.
At the start of the project, the odds of finding causative mutations were fair. All diseases showed a genetic component, present in more than one breed. This are usually indicators for a common genetic cause. During research, the breed specificity became more pronounced and only a limited number of common regions could be identified. This indicates that breed specific studies might be preferred and/or the possibilities of genomic selection should be explored.
The consortium used the 650K SNP-array that was only recently operational by that time. There was hardly any experience with this array. As we started the project it was believed that a genomic evaluation across-breeds could identify regions that harbor genes being involved in the diseases. In particular with high-density SNP-arrays and the involved breeds being a blend of highly inbred and less inbred breeds and populations. Results indicated too low LD to gain from an across-breed analysis in identifying genomic regions. This is in accordance with recent research in other species and it is generally believed that even higher density is required or even full genome sequencing. The consortium decided to engage in a novel type of analysis (Copy number variation) which, at the moment of writing, gave new scientific insights but applications are still ahead.
Nevertheless, diseases remain a major concern. Especially high prevalent diseases are a heavy burden, such as the three diseases in this project. For OC clinical prevalence is from 5 to 25%, but radiographic signs of abnormal development can be 70%, with large financial implications. In the Netherlands, the annual loss by OC in Warmbloods is estimated at 10 million euro (3000 affected foals and a loss of 3300 euro/foal). Moreover, there is an indirect cost of the disease, as horses with OC are not approved, which excludes a large part of the gene pool. The financial repercussions of IBH and CPL are not that severe, but they have a clear financial and animal welfare impact by premature culling of the horses.
The project has created awareness by affecting the viewpoints of SME partners towards genomics and genomic tools. They believe in investing in this technology and follow-up-projects are being prepared. Furthermore, they are aware of the need for uniform and extended phenotyping. The same mind-switch was observed in government officials during the final workshop. This is reassuring because the horse sector has been more hesitating to pick up new technologies compared to the typical livestock species.
Dissemination of knowledge has been focused mostly on the horse sector and not so much towards the general public. Nevertheless, the video-clip is available for wide distribution.
The scientific findings of the project are visible as peer reviewed articles and conference contributions. The cooperation between industry and RTD partners has created a platform that offers a unique resource (samples, phenotypes and genotypes) that is exploited by the members of the consortium but allows for new contributions.
The project has a significant leverage in terms of research as at least 3 proposals are or will be submitted nationally and further international cooperation is under discussion. This will ensure further support and development of the important horse industry in the EU.
Updated information can be obtained from www.horsegene.eu
Potential Impact:
The primary objective of the consortium have been largely reached. At the start of the project, the odds of being able to find causative mutations for one (or more) of the diseases were fair. All three diseases (or disease susceptibilities) showed a genetic component, present in more than one breed. However, as research progressed, the breed specificity become more pronounced and a single (or limited number of), common region(s) for any of the diseases could not be identified.
This breed specificity is an important result because it may indicate that breed specific studies might be the preferred route. Or, alternatively, genomic selection should be explored more.
The consortium made use of the latest state of the art technique and used therefore the 650K SNP-array that was only recently operational by that time. At that time there was hardly any experience with this array. As we started the project it was generally believed in the animal genetic community that a genomic evaluation across-breeds would be a very appropriate strategy to identify regions at the genome that harbours genes that are involved in the genetics of the diseases under study. In particular when high-density SNP-arrays were applied and when the breeds involved are a blend of highly inbred breeds (to identify number and location of involved regions) and less inbred populations (to narrow down the size of the identified regions) and therefore we were eager on using the latest 650K SNP array. The results of the GWA-studies on the 650K SNP-array indicated however that LD was too low, even using the 650K, to gain from an across-breed analysis in identifying genomic regions involved. This finding is in accordance with research in other species and it is now generally believed that much higher density is required or even sequencing to perform across breed analysis. The consortium decided therefor to engage in a novel type of analysis of the genotypes (Copy number variation) which, at the moment of writing, gave new scientific insights but applications are still ahead.
At the end of the project, 2 sets of SNP markers were selected by a commercial lab to be included in 2 panels: one to select against Chronic Progressive Lympoedema in draught horses and one to reduce the prevalence of IBH in Friesian horses. After discussions with breeding associations a commercial development was not realized. The main difficulty was the relatively low % of phenotypic variance explained by the tests which would result in low accuracy of the tool and thus low accuracy of the estimated breeding values. The trade-off between accurate breeding values relative late in life vs less accurate breeding values but available earlier in life, is a debate that is going on for a longer time in practical horse breeding and has been revisited with the results of this project. As a result of the discussions in this project some of the breeding associations agree that the topics should be taken up by their members. Breeding organisations feared that selection of breeding animals using an imprecise test would not be accepted by horse breeders.
Nevertheless, diseases remain one of the main concerns in horse breeding and keeping. Especially diseases that have a high prevalence in some or most horse breeds are a heavy burden. The three diseases studied in this project fall under this category. For OC for instance, clinical prevalence is usually between 5 and 25% in a given horse population, but radiographic signs of abnormal development can be as high as 70% in certain groups. Prevalence and incidence of this disease are severe with large financial implications for owners and breeders as a consequence. An example: In the Netherlands, the annual loss caused by OC in Warmbloods was estimated at 10 million euro (3000 affected foals and a loss of 3300 euro/foal). Moreover, apart from the direct economic loss, there is an even larger indirect cost of the disease, as many studbooks will not approve horses with evidence of OC, which implies the exclusion of a large part of the gene pool.
Although the financial repercussions of the other two conditions (IBH and CPL) are not that severe, these diseases also have a clear financial impact and largely affect animal welfare as they sometimes lead to a premature culling of the horses through euthanasia.
In terms of creating awareness, the project has strongly affected the viewpoints of the SME partners in the consortium towards genomics and genomic tools. The belief has grown in most studbooks that they should further invest in this technology and follow-projects are being prepared. Furthermore, SME-partners are now definitely well aware of the need for uniform and extended phenotyping. The same mind-switch was observed in government officials during the workshop at the end of the project.
This observation is re-assuring because the horse breeding sector has been more hesitating to pick up new technologies compared to the typical agricultural livestock species (dairy cattle, swine, poultry, ...).
Dissemination of knowledge has been focused mostly on horse breeding or horse owner-organisations and not so much towards the general public. Nevertheless, the video-clip is available for wide distribution.
The scientific findings of the project are ample and visible as peer reviewed articles and contributions to conferences. The cooperaton between the industry and RTD partners has created a very usefull platform that offers a unique resource (samples, phenotypes and genotypes) that is exploited further by the members of the consortium but allows for new contributions.
The project has a significant leverage in terms of research as at least 3 proposals are or will be submitted nationally (e.g. in Belgium, Sweden, Netherlands) and further international cooperation is under discussion. This will ensure further support and development of the important horse industry in the EU.
List of Websites:
http://www.horsegene.eu