Molecular characterization of rare forms of canine neurological diseases as potential models for similar human diseases

Canis lupus familiaris, the domestic dog, possesses a huge variability in traits such as size, conformation, coat color, or character, which reflects the generations of targeted human selection after the dog's domestication thousands of years ago. The phenotypic differences naturally reflect the underlying, often breed-specific, genetic variation. While heterogeneity between breeds is large, at the same time, the individuals within one breed are usually very homogeneous. This specific population structure of modern dog breeds with low genetic diversity favors the propagation of spontaneous occurrences of genetic mutations that might lead to the development of diseases, and thus makes the dog a valuable animal model. Inherited neurological disorders in animals as well as in human patients are incurable, often severe, and result in progressively worsening quality of life; early diagnosis is therefore beneficial for managing the disease development. DNA-based precision medicine using state-of-the-art methods, e.g. whole-genome sequencing (WGS) has been successfully utilized in recent years for routine diagnosis of rare diseases in human as well as in veterinary medicine. Identification of disease-causing variants allows dog breeders to avoid the spread of such variants in the affected dog breed, ultimately improving the health of the whole population through better breeding management, as well as to advance the understanding of the molecular mechanisms involved in corresponding human disease, and may be useful for the development of novel therapeutic strategies. In this thesis, I took part in the analysis of seven specific canine neurological phenotypes applying different genetic mapping methods, candidate gene analysis, and WGS. I also generated and analyzed extensive genealogical and genomic data on the worldwide Leonberger dog population in regards to its diversity and disease prevalence. Despite its increasing size in recent years, the population lost considerable genetic diversity due to a bottleneck in the last century. The heavy use of popular sires led to high relatedness among the breeding dogs and thus to high inbreeding rates. This facilitated undesirable genetic traits to spread within the gene pool of the Leonberger breed. A private homozygous frameshift variant in the GJA9 gene was identified in Leonbergers with an adult-onset form of polyneuropathy using genome-wide association study (GWAS) and WGS. The GJA9 gene encodes a connexin gap junction family protein, which are important components of peripheral myelinated nerve fibers; this discovery for the first time adds GJA9 to the list of candidate genes for similar human conditions. During a study of additional forms of polyneuropathy and/or laryngeal paralysis, I found a missense variant in the CNTNAP1 gene in Leonbergers and Saint Bernards showing early signs of laryngeal paralysis. CNTNAP1 encodes a contactin-associated protein important for the organization of myelinated axons and has been implicated in various forms of human neurological diseases. Interestingly, this variant was seen in several other unrelated dog breeds and most likely predates modern breed establishment. A similar approach revealed two independent variants in the NAPEPLD gene in Leonberger and Rottweiler dogs affected by leukoencephalomyelopathy providing evidence for allelic heterogeneity of this disorder and the first description of NAPEPLD-associated inherited defects in the endocannabinoid system associated with myelin disorders. In another study, a form of canine neuroaxonal dystrophy occurred in young adult Rottweiler dogs. WGS data of two cases revealed a homozygous missense variant in the VPS11 gene, encoding a member of VPS class C complex, a key factor of the endosome-autophagosome-lysosome pathway, previously associated with an infantile-onset neurological syndrome in humans. In a family of Alpine dachsbracke dogs, I used linkage analysis and homozygosity mapping to discover an autosomal recessive variant in the puppies affected by spinocerebellar ataxia, which affects the SCN8A gene. The gene encodes a subunit of a channel important for sodium ion transport to neurons in the central nervous system and was previously implicated in human neurogenetic conditions. By in-depth pedigree analysis, I found a common ancestor of two geographically separated families of Saluki dogs in which puppies suffering from succinic semialdehyde dehydrogenase deficiency causing neurological abnormalities were observed. GWAS and subsequent filtering of WGS data of two affected Saluki cases identified a causative variant in the ALDH5A1 gene encoding an essential enzyme of the gamma-aminobutyric acid neurotransmitter metabolic pathway. Finally, the underlying genetics of a previously described Leigh-like subacute necrotizing encephalopathy in Yorkshire terriers was solved by discovering a perfectly associated loss-of-function indel variant in the SLC19A3 gene encoding thiamine transporter 2, which is important in brain development, and its disruption was previously seen in similar human neurometabolic disease. In conclusion, the discovery of the herein described likely pathogenic DNA variants enabled systematic genetic testing of breeding dogs, and selection against the corresponding disorders to improve the health and welfare of the respective populations. This thesis provides molecular descriptions of several canine neurological conditions and presents additional physiologically relevant models of corresponding human diseases. Apparently, species- and site-specific differences in pathological phenotypes for mutations within the same gene exist as seen, e.g. in canine VPS11-related neuroaxonal dystrophy. All these spontaneous canine models closely resemble rare human syndromes and provide physiologically relevant models to better understand poorly characterized gene functions, e.g. in defects of the endocannabinoid system related to NAPEPLD; and provide potential new candidate genes for corresponding human forms of diseases with yet unsolved genetic etiology, e.g. GJA9-associated polyneuropathy. Therefore, this thesis demonstrates that genomic studies of domestic animal species such as the dog improve the understanding of rare complex and heterogeneous groups of neurodegenerative disorders