Canine research is a growing and dynamic field, which not only informs veterinary practice and breeding but also provides insights into human disease. Addressing the breadth of findings this field has to offer, BioMed Central launched the journal Canine Genetics and Epidemiology with the Kennel Club, the UK’s largest organisation dedicated to the health and welfare of dogs. We spoke to members of the journal Editorial board, Cathryn Mellersh from the Animal Health Trust, UK, and David Sargan from the University of Cambridge, UK, about how canine research can advance our knowledge, treatment and prevention of both canine and human diseases.
Pedigree dog breeding has been criticised for impairing the health and welfare of dogs. What are the most common health problems associated with dog breeding and how can genetics insights into the canine genome improve this situation?
CM: The evolution and selection of the modern purebred dog population has been driven not by economic pressures, as has the development of other agricultural species where undesirable traits are quickly eliminated, but rather by our desire to shape dogs to look or behave a certain way. Most breeds of dog are closed populations in which intense selective pressure has been applied to a small number of founding animals. Additional bottlenecks have occurred during the histories of many breeds that have reduced effective breeding stocks to very few animals. The limitation of any gene pool by these means has the effect of reducing genetic diversity within the population and increasing the incidence of inherited disease. This effect is further confounded in dog populations by the extensive use of popular males (the so-called popular sire effect). Furthermore, to develop and maintain desired traits, both physical and behavioural, dog breeders tend to mate animals that share common ancestors thus increasing the likelihood that their offspring will inherit multiple copies of identical disease mutations.
The most efficient method of reducing the prevalence of any inherited disease from an affected breed is to identify the mutation(s) responsible, and then to screen all breeding dogs to identify those that carry the mutation(s), and to subsequently limit breeding to pairs of dogs that cannot produce clinically affected offspring. The progressively sophisticated tools available with which to interrogate the canine genome are making the development of DNA tests increasingly straightforward, time- and cost-effective.
DS: The most common health problems associated with dog breeding are defects associated with exaggerations of conformation or appearance within a given breed or group of breeds. Thus the attempt to give the companion or lap dog a cute babyish appearance has led to both neoteny and a broadened, flattened face accompanying a brachycephalic skull shape. The latter has a host of disease associated consequences: many (and in some breeds virtually all) dogs with this skull shape have difficulties breathing caused by excessive soft tissues impinging on a very reduced airway. We are all familiar with the wheezing sounds made by many pugs and bulldogs. At its least severe this contributes to exercise intolerance, but this is accompanied by an inability to cool oneself, and these two traits may tip over into cyanosis and collapse, and to heat stroke. These dogs suffer a wide spectrum of secondary problems including sleep apnoea, cardiac problems, difficulties in suckling and eating or regurgitation, whilst excess skin and shortened eye sockets make conjunctival damage and corneal ulceration more common in these breeds. The large head of neonates makes natural birthing impossible so that most pups are born by Ceasarian section.
I have given a fairly full account of the consequences of that particular exaggeration. But many other groups of dogs also have problems: companion dogs with a small and inappropriately upright skull case often show Chiari malformation and may suffer syringomyelia. Exaggeratedly small and exaggeratedly large dogs, dogs with very deep chests, shortened legs or exaggerated gaits or postures have their own sets of diseases (for example joint diseases, osteosarcoma and hypertrophic cardiomyopathy in the giant breeds). Excess skin causes disease problems in all breeds in which it occurs, not just brachycephalic breeds – and so on.
Other very common disease problems are those that are associated with genetic bottlenecks in a breed, often at breed foundation, but sometimes due to popular sire or strong co-selective effects. One example of this is an SCL2A9 gene mutation in Dalmatian dogs, which causes kidney and bladder stones (and gout in humans). This mutation was until a recent outcross present in all Dalmatians. The old idea that this was co-selected with spotting now seems unlikely, and the mutation is more probably a founder or popular sire effect.
Insights that improve the situation come from two directions. Old fashioned population genetics has shown us the small effective population sizes of some breeds and emphasised the potential usefulness of outbreeding, whilst tools like mate select in the UK help owners prevent further reductions in genetic diversity. More recently molecular tools have been helping at several levels, but perhaps first and foremost in the production of DNA based diagnostic tools for disease carriage, and in DNA directed selection to remove disease or disadvantageous traits. In these instances though care must be taken to advise the breeder of the meaning of test results, and not to cause unnecessary reductions in overall genetic diversity.
How can a better understanding of canine genetics, genomics and epidemiology, advance our knowledge and treatment of human diseases?
CM: There are a variety of reasons why results derived from the investigation of disease in dogs are likely to be extrapolative to human medicine. A compelling argument for the dog being the most appropriate model organism for studying diseases that afflict both dogs and people is that in the dog (as in man) the conditions are naturally occurring, and are similar biologically, histologically, and in clinical course. The dog also exhibits a physiology more suited to gross comparison with the human than many traditional model organisms.
Epidemiological study of the risk factors associated with complex human diseases may also be approachable via the study of canine cohorts. As dogs share a common environment with man, it is possible that the aetiology of canine diseases is similar to those of their human counterparts.
The genetic basis of inherited disorders can be investigated considerably more easily in dogs than humans because the genetic structure of canine populations enable disease-associated regions to be identified using far fewer cases than equivalent studies with humans. In addition, complex diseases that affect specific breeds may well be genetically ‘more simple’ than the same clinical condition in human populations because dogs of a particular breed are likely to be far more closely related to one another than individuals within a human disease cohort.
DS: Some 500 different inherited diseases have been described in dogs. Many of these occur in several breeds and although some of these originate from single mutations that occurred before separation of the affected breeds, others have already been shown to be genetically distinct in different breeds. In total there may well be twice the number of genetically distinct diseases, by a long way the largest number of described naturally occurring inherited diseases in any non-human mammalian species. These diseases, occurring in a large natural population under reasonably close veterinary medical surveillance represent an enormous potential resource for the modelling of pathogenesis and testing of therapy for autologous human diseases.
These resources are especially valuable thanks to our increasing understanding of the structure of the canine population as a set of genetically isolated sub-populations together with a larger interbreeding group, and our rapidly increasing knowledge of the canine genome. In the context of comparative medicine, this population structure allows a simplified analysis in particular because within each breed there is relative homozygosity within coding DNA, so that within a single breed, polygenic diseases are relatively simplified into a smaller number of polymorphic genes of larger effect. This, together with long linkage disequilibrium (LD) within breeds allows mapping of effector loci with relatively small numbers of individuals, although the downside of long LD is that loci are large and may contain many genes, so that actual causative mutations are sometimes hard to recognise.
What human conditions are particularly suited to investigation in a canine model and why?
CM: Canine genetics studies have already identified novel genes associated with a variety of conditions that affect both humans and dogs, including ichthyosis, inherited retinal disorders and various neurological disorders.
DS: It should be remembered that dogs don’t simply offer a model of human ‘conditions’ but also a great model to allow a fuller understanding of the normal human condition. A great deal of really excellent genetics and genomics in dogs has been refining our understanding of the loci that regulate canine skeletal development, conformation, pigmentation, and even canine metabolism. Many of the new discoveries in these areas have implications for the human genome.
In terms of genetic disorders, the canine genome is particularly suited for mapping many types of polygenic disease and for the discovery of genes that have small (but significant) effects in human but more measurable effects in the less genetically noisy genome of a single breed of dog. The low effective population sizes of individual breeds also reduces the heterogeneity of alleles in each system and increases their frequency in the population, potentiating discovery.
Furthermore, the dog is a long-lived and large mammal that, in features of its development and some aspects of its metabolism, is closer to Homo sapiens than is the mouse. This means that disease features in genetic orthologs may often mimic those of humans more closely than do the murine equivalents, as is the case for Duchenne muscular dystrophy and cystic fibrosis. All these features provide benefits as an intermediate model between mouse and man.
Among the disorders to which the canine genome can make, and is making, particular contributions are: developmental diseases of the skeleton and soft tissues that are exaggerated in dog models, cancer predispositions, and also endocrine, exocrine and other metabolic and neurological problems where canine studies have revealed roles for the major histocompatibility complex or other elements of the immune system. The dog’s large reservoir of simplex genetic diseases, size and lifespan as well as the similarity to human orthologous diseases have made it a productive test-bed for gene therapy research with tissue targets ranging from eye, liver and bone marrow, to the haematopoietic system and lungs.
What do you think are the most interesting recent advances in canine genetics and genomics, whether applied to human disease, veterinary research and/or breeding?
CM: New next-generation sequencing technologies are revolutionising the fields of both human and canine genetics and genomics. The ability to generate vast amounts of DNA sequence quickly and cheaply means the pace at which genetic variants associated with important inherited traits are being discovered is unprecedented.
DS: It is difficult to pick a single area, but one of the most interesting recent advances has come from studies of the genome of the canine transmissible veneral tumour (CTVT). It should not be forgotten that this tumour is of major welfare importance in areas with large stray or feral dog populations. But it gets my vote for its biology and the recent linkage of aspects of this to its genetic make up. Its survival in multiple hosts and ability to remain hidden from immune responses in many are fascinating behaviours, displayed by only one other mammalian tumour, the Tasmanian Devil Facial Tumour. Unlike that tumour, in which the genome is not very extensively rearranged and which exists in a host that shows only limited major histocompatibility complex diversity, the CTVT is ancient, extensively rearranged, has lost the functions of several hundred genes, and is able to colonise hosts of many and perhaps all canine breeds as well as feral dogs, reflecting a wide diversity of dog leukocyte antigen and other polymorphisms. Recent sequencing of genomes of CTVT examples from two hosts suggest an origin more than 11000 years ago, and that the genome has now become comparatively stable. This work will provide insights both into mechanisms that allow evasion of the host immune response, and the set of functions that must be maintained as well as those that can be disposed of by such a parasitic cell line.
In the recently launched journal, Canine Genetics & Epidemiology, every published article includes a lay summary and key ‘take home messages’. Why do you think it is particularly important to communicate research in this field to a wider audience?
CM: It is hugely important to communicate research in the field of canine genetics and epidemiology to non-specialists because, on the whole, the dogs that contribute to this field of research are owned by members of the general public, be they dogs owners or breeders. Now that sequencing technologies are so sophisticated sample collection is often the rate-limiting step of many research studies, and the better able the dog-owning public are to understand, embrace and contribute to the studies being undertaken, the quicker the studies will proceed, and the quicker dogs and their human friends will benefit from the research findings.
DS: Many scientists are less aware than they should be of the power of the canine model and raising awareness amongst this group by using a simple message can only be good for the exploitation of the model. But in particular, the lay summary offers the opportunity to communicate with veterinarians, with dog owners and dog lovers and with the curious public of all ages. In a highly linked-up age there is the possibility to interest and enthuse the public, to explain illnesses to dog owners and to let veterinarians know both what is new, and perhaps how they can help to push science forward.
Read more about the launch of Canine Genetics and Epidemiology in this BioMed Central blog.
More about the Editor(s)
Cathryn Mellersh is the Head of Canine Genetics at the Animal Health Trust, UK. She obtained her PhD in developmental genetics at Leicester University, UK, where she went on to pursue her postdoctoral training before joining the Fred Hutchinson Cancer Research Center, USA. On returning to the UK, Mellersh joined the Animal Health Trust. Here research aims to identify mutations responsible for inherited disorders that affect different breeds of domestic dog, with a focus on ocular and neurological disorders. Through collaborative efforts, research in the Mellersh lab also seeks to apply these findings to understand similar inherited conditions in humans. Mellersh is also a member of the Advisory Council on the Welfare Issues of Dog Breeding, the Kennel Club Dog Health Group and the Kennel Club Health Screening & Genetics Sub-Group.
David Sargan is a senior lecturer in molecular pathology in the Department of Veterinary Medicine at the University of Cambridge, UK. His current research interests focus on the comparative genetics of disease, with a particular emphasis on genetic diseases of dogs. Specific areas of investigation include the pathogenesis of retinal and other ocular inherited diseases, canine breed predispostion to cancer, comparative genomic mapping in companion animal species, and genetic aspects of welfare issues in breeding. Sargan is also a council member of the Advisory Council on Welfare Issues of Dog Breeding and a trustee of the Universities Federation for Animal Welfare.