Expensive canine molecular genetic proposals and why human geneticists dedicate the time and effort.
Jörg T. Epplen, Regina Kropatsch; Human Genetics, Ruhr-University, 44780 Bochum, Germany
Successes in molecular genetic research are celebrated every now and then, even reaching the layperson presses. Headlines such as ‘dog genome decoded’, ‘gene therapy successes’ and ‘genetic peep show into the canine personality’ trigger somewhat unrealistic hopes and expectations concerning fast research successes in the entire molecular genetic spectrum. In reality, practically all experimental data are accomplished by working tremendously hard. Apart from sound ideas, these successes require dedicated and tenacious scientists, time and of course, extensive funds. The initial, provisional results for the immediate use and benefit of the participating dog owners can be communicated first (see e.g. Katrin Streitberger et al. 2008), although the full-fledged scientific publication in an internationally recognized journal usually takes still more time.
Since Retinitis Pigmentosa (RP) is the homologous disease to progressive retinal atrophy (PRA) in man, the mutated genes causing RP are particularly interesting candidates for the mutation search for canine PRA. We discuss here two examples of canine research projects and point out respective aspects of the financial and time requirements.
For some years, our group has been concentrating on certain canine eye diseases, which among veterinarians and dog breeders is summarized as (generalized) progressive retinal atrophy (PRA). A basic condition for such research is the accurate clinical (differential) diagnosis of the illness by the veterinarian. Specialists like the Dortmund ophthalmological circle (DOK; Dr. Brahm) created very solid starting conditions by thoughtfully training a team of veterinarians for clinical PRA diagnostics. Yet for us to execute such research projects depends initially and wholly on the voluntary supply of DNA samples; hence the dedicated co-operation of the dog owners. Unfortunately, such selfless commitment by the owners exists only in a few breeds such as the Irish Glen of Imaal, as well as the Airedale Terrier, the Löwchen and similarly; the following recent example of PRA-affected Schapendoes.
The PRA mutation in Schapendoes - a long piece in several parts
After a careful search and selection of Schapendoes family lines with blind dogs, we began blood collection intensively in 2003. Unfortunately, mucosal swabs are not used for long-term genome-wide research projects since the isolated DNA quantity is comparatively small and the DNA quality decreases too fast during storing.
The final entry of the 57th blood sample determined the start of the project in March 2004, even though everything else had been ready for months and the co-workers were anxious to begin the experiments.
In order to ensure molecular genetic dog research projects are as economical as possible, we reduce costs primarily in personnel resources: instead of an experienced scientist, graduate students are employed during their scientific education. They receive only half of a scientist’s income, but they work particularly hard and for long hours to obtain their PhD after three years. A graduate student or a scientist requires at least 15 000 € per annum at pure expendable material costs of chemicals etc. without even calculating costs for laboratory space and the other overheads. All the necessary molecular biological reagents are obtained as economically as possible by integrating their purchase within larger purchase contracts (volume purchasing) which of course are not limited to the dog project.
By July, 2005 and after 16 months of intensive practical experimenting and evaluating, we defined a large chromosome section in the Schapendoes where the PRA mutation had to be located with certainty. After this break-through, the first and most promising candidate gene was quickly characterised in this region by Tanja Lippmann. Unfortunately however, the responsible PRA-mutation could not be identified in this gene.
Next in late February 2007, these intermediate results were published in the respected journal Molecular Vision. Comparatively only a small success, given the efforts of all those involved after almost five years since the start of the project.
But finally two and a half years later in early in 2008, the corks popped: the causative aberration was identified in the eleventh candidate gene examined in the respective chromosomal region. Information concerning the existence of this gene was based on the Whole Genome Sequencing Project of man, mouse and dog, but the responsible gene and its adequate protein product are completely new discoveries to date. Recently we also detected the derived protein in the retina, a further milestone on the long journey of necessary proof that we are on the right track. After understanding the gene’s function and that of the protein during the next few years, the final goal is to finally undertake experimental therapy trials.
In order to characterize the function(s) of the new protein, an independent research project proposal was submitted for funding consideration to the Deutsche Forschungsgesellschaft (DFG). Having made these discoveries, we then wanted to offer the Schapendoes friends the possibility of a direct mutation test. On the other hand, further research data will be necessary with additional methods in order to publish these results comprehensively and convincingly on the highest scientific level. Therefore, the accurate denotation of the PRA mutation for Schapendoes has not been officially communicated so far due to normal and necessary protocols. Yet more than 1 000 Schapendoes have been tested for the mutation, so that homozygous affected puppies could be avoided by selective matings from the tested breeding pairs. Genome screening in three additional dog breeds - a continuation
Similar to the PRA project in Schapendoes, mutations need to be identified for three more breeds: the Irish Glen of Imaal Terrier (GIT), the Airedale Terrier (AT) and the Löwchen (Lo). The authors have been working on this project since 2007. Yet a comparatively more difficult situation then that of the Schapendoes already existed in these three breeds. DNA material was obtained from fewer dogs based on even smaller pedigrees. The information concerning PRA in these populations is lesser only because the informational content of pedigrees depends on the number of affected dogs and their direct relatives. Yet this initial situation is not at all due to lack cooperation by the breeding clubs concerned. Rather, the limitations result from missing generations of parents and/or grand-parents which had already expired by the time we recruited the material from the populations. Given these constraints, we had to optimize the genome-wide screening for these three breeds thereby increasing the chances for success in identifying marker linkage with PRA. Therefore, substantially more microsatellite markers were established in addition to the existing marker set. Using more carefully selected markers, the distances between these are reduced and concomitantly recombination probability should be minimized. On the other hand, important retina-specific as well as RP-associated gene regions were included by investigating even more markers. Even with these measures, there is still no certain ‘land in sight’ after two years of most intensive work and despite a designed strategy, a residual risk remains that we may miss the region which includes a PRA-linked marker.
To date it has been assumed that there exist relatively large haplotype blocks of several mega bases (Mb) in the dog genome. A marker set consisting of 325 markers with an average distance from 9 Mb (~9 centi Morgen, cM) distributed over the whole canine genome is sufficient for a successful linkage analysis. Various simulation studies conformed to this hypothesis. Here, we now reduced average marker distance by additional markers to ~5 Mb and also minimized recombination probability; and still no marker could be identified as being linked to PRA. Therefore, haplotype blocks may be smaller in some chromosomal regions than assumed and recombinations may have taken place between the examined marker and the PRA mutation. Comparative analyses of the canine genome with that of man have shown that some sequence sections -so-called hot spot regions- also exhibit relatively high recombination frequencies. A high proportion of all recombinations take place in such locally limited hot spots between haplotype blocks. Hot spots possess an increased content of guanine and cytosine nucleotides. In the canine genome, they are probably localized predominantly at the telomeres (ends of chromosomes), perhaps also in other chromosomal regions. So far their exact localizations are elusive and further expansion of the marker set via additional markers appears to be not particularly rewarding.
PRA is a monogenetic disease in most dog breeds, which means that a mutation in only one gene is causative for PRA. In breeds such as the Miniature Schnauzer, Kuvasz, Golden Retriever and others, PRA causal mutations are already known. Yet the company Optigen could not confirm these mutations in all PRA-affected dogs in the above-mentioned breeds. Thus, other mutations and probably in different genes may be the cause for PRA in these breeds. Another potential explanation would be that a breed was not limited to the country of origin but also expanded in distinct areas (Germany, The Netherlands, England, and Finland). Depending on the time of geographical separation, country-specific pedigrees could be regarded as independent populations which developed independently and thus afforded different PRA-causing mutations. For the GIT, the AT and the Lo, such assertions appear inappropriate because these breeds are very closely bred and their creation is based on only a few forbearers. Another - rather more improbable - starting point could be falsely diagnosed PRA - despite well trained ophthalmologists and veterinarians. For example, if a dog suffers additionally from an accompanying eye disease like glaucoma, evaluation of the PRA diagnosis is possibly difficult and may lead to false diagnoses. Since we are convinced of the correctness of our PRA diagnoses, we would like to continue the search for the PRA causal mutation in the three breeds with an alternative method - modern SNP microarray technology. Substantial advantage of this technology in relation to the microsatellite-based genotyping method is the parallel analysis of several thousand SNP (single nucleotide polymorphism) markers in a small quantity of DNA. Such microarrays comprise ~27,000 SNPs distributed over the entire canine genome which corresponds to a genomic SNP coverage of 1 SNP per ~0,1Mb - thus a substantially higher marker density than our extended microsatellite marker set with in average 1 microsatellite per 5 Mb. This method offers a better chance in narrowing down the PRA causal chromosomal region. Because the new strategy of microarray technology requires the application of specific ”chips” which are only produced on request by means of a complex, photolithographic procedure, the PRA project is at least 25.000 € more expensive due to consumables costs.
Conclusion and perspectives
All modern research depends on the co-operation of the breeders and owners as well as their organisations including all the necessary information that must not be suppressed. Molecular genetic research supports modern canine breeding and it represents a solid scientific basis for empirically developed breeding strategies. Hereditary disorders, i.e. Mendelian or monogenic conditions respectively, can be eradicated in dog breeds with cooperative breeders – and without loss of genetic variation. Specific tests for Mendelian exterior characteristics like color etc., become increasingly available. Biobanking and genomic progress will offer more efficient investigative possibilities of multifactoral characteristics in the not too distant future for behavioural and performance characteristics as readily as illnesses, such as epilepsy or heart diseases. Certainly, the financial expenditures for this are by no means insignificant, however very meaningful investments for the progress of kynology in the long run and poises us that much closer to comparable human methodologies and possible discoveries.
Acknowledgements
We thank numerous owners supporting research by the necessary biological material of the dogs.
Selected references
T Lippmann, A Jonkisz, T Dobosz, E Petrasch-Parwez, JT Epplen, G Dekomien: Haplotype defined linkage region for gPRA in Schapendoes dogs. Molecular Vision 13: 174-180, 2007
K Streitberger, MS Fischer, O Distl, JT Epplen: Molecular-genetic investigation to the genetic variability in dog races. Our pedigree dog 7:30 - 33, 2008