Pathogenesis

Until recently the pathogenesis of canine degenerative myelopathy had remained unknown. Griffiths and Duncan originally hypothesized DM to be a "dying-back disease" confined to the CNS suggesting a toxic etiology. Recently, studies of the brain of DM affected GSDs showed neuronal degeneration and loss in some brainstem nuclei. Johnston et al. suggested that a defect in the neuron, itself that may lead to abnormal axonal transport and degeneration in the distal axon.

An immunologic role in the pathogenesis of GSD DM has been proposed based upon observations of depressed responses to thymus-dependent mitogens and increased concentrations of circulating immune complexes. Although immune-related degenerative disease is a plausible theory, immunosuppressive therapies have shown no long-term benefits in halting the progression of DM. Based on the immunologic hypothesis, Clemmons et al. (FASEB, 2006) claimed that there is a point mutation in the hypervariable region 2 of DLA-DR� and termed allele *1101J and reported homozygosity in DM affected GSDs. A more recent study showed that data did not provide evidence for involvement of DLA-DRBI in DM and indicated that the test offered by the University of Florida was not predictive for DM.

We used the canine SNP chip (Affymetrix v2) to genotype DNA samples from affected and healthy dogs. Genome-wide association mapping analysis revealed a peak with strongest association on canine chromosome 31 which also showed markers that contained the canine the superoxide dismutase (SOD1) gene. Resequencing of SOD1 in normal and affected dogs revealed a G to A transition, resulting in an E40K missense mutation. Homozygosity for the A allele was associated with DM in five common dog breeds (German Shepherd dog, Boxer, Rhodesian Ridgeback, Welsh Corgi, and Chesapeake Bay Retriever). The frequency of the A allele in a separate "other breeds" control group, consisting of samples from dog breeds in which DM was rarely diagnosed, was significantly lower than that for the controls from the affected breeds. Not all animals that are homozygous for this mutation will get the disease. Thus, penetance amongst the mutant homozygotes is incomplete. To summarize, we have discovered a risk factor for dogs to develop DM. We believe that there are other genetic modifiers and environmental factors that could influence whether these dogs at risk actually get DM.

Mutations in the SOD1 gene are known to cause amyotrophic lateral sclerosis (ALS) in humans; also known as Lou Gehrig's disease. The disease derives its name from the combined degeneration of upper and lower motor neurons projecting from the brain and spinal cord. The Greek derivation of amyotrophy means, "muscles without nourishment." Lateral is the location within the spinal cord of axonal disease and sclerosis refers to diseased axons being replaced by sclerosis or "scar" tissue. The disease affects people in their fourth or fifth decade of life. Most people die within 5 years after disease onset.

There are several different forms of ALS in people that lead to progressive paralysis and subsequent respiratory failure. The major hallmarks of criteria for a diagnosis of ALS include signs of upper (brain or spinal cord) and lower (nerve and muscle) motor neuron degeneration, and progressive spread of these signs within a region or to other regions. Degenerative myelopathy affected dogs initially manifest UMN signs, similar to UMN-onset ALS, that progress to also involve the thoracic limbs and later manifest LMN signs. DM affected dogs also have an absence of electrophysiological or neuroimaging evidence of other disease processes that could explain the combination of progressive upper and lower motor neuron dysfunction. Because this is a spontaneously occurring disease and because we consider the dog nervous system more like human than either the rat or mouse, we hope that this model of ALS may be more predictive of outcome in human ALS clinical trials. Mouse and rat ALS models have thus far been disappointing since many drugs that were successful in the mouse/rat models subsequently failed in human ALS. Furthermore, the SOD1 mutation in dogs may offer significant advantages for some studies of ALS pathogenesis.