The mSOD1 mouse, harboring a mutated human superoxide dismutase (SOD1) gene, has long held the spotlight as the model of choice for amyotrophic lateral sclerosis (ALS). It has yielded scores of papers on disease progression and potential treatments, but it is far from a perfect model. At the Society for Neuroscience annual meeting, held 15-19 November in Washington, DC, veterinary researchers from the University of Missouri in Columbia reported that canine degenerative myelopathy, an inherited condition with ALS-like symptoms, is linked to a mutation in the canine superoxide dismutase gene, as are 20 percent of familial ALS cases. Neurons from afflicted dogs showed SOD1 aggregates reminiscent of ALS pathology. Study authors Joan Coates and Dennis O’Brien of the University of Missouri in Columbia, along with other collaborators at the University of Missouri and at the Broad Institute of MIT/Harvard, suggested that the dogs could serve as a valuable model for ALS.

There is only one drug for ALS, and no treatment exists for canine degenerative myelopathy (DM). “We want to help the dogs and also help people,” O’Brien said. Dogs have served as valuable testing grounds for cancer therapies, the authors noted, and these animals might serve the same purpose in testing ALS treatments. The researchers first described the link between DM and SOD at the 19th International Symposium on ALS/MND, held 3-5 November in Birmingham, U.K.

Canine degenerative myelopathy and ALS share the late onset of a progressive motor neuron disease and the SOD1 mutation and pathology. The main difference between the conditions is that while the vast majority of human SOD1 mutations associated with ALS are dominant, the dog mutation is recessive, apparently with incomplete penetrance. Dogs have a shorter lifespan than people; disease onset is nine to 11 years and the canine disease progression is faster than in humans, from between six and 18 months depending on breed. “This means that the results of any treatment can be assessed much faster in dogs than in people,” wrote William Thomas, a veterinary neurologist at the University of Tennessee in Knoxville, in an e-mail to ARF. Thomas was not involved with the study.

Mutant SOD1 mice, in contrast to the dogs, exhibit an artificial disease. Scientists engineered the animals to express unnaturally high levels of mutant human SOD1. “In these mice…SOD1 has 10-fold the activity as normal,” Coates said. “You may not see the true clinical picture there.” The mice are convenient models because they reliably develop disease and progress to end-stage within months. However, the mSOD1 mouse has been criticized because several treatments that appeared promising in mice failed to benefit humans (Schnabel, 2008). “This [dog] model may more closely represent the clinical situation in the human population,” Coates said.

“I think it’s a neat potential tool, but there’s a lot more that has to be done,” said Jeffrey Rothstein, a neurologist at Johns Hopkins University in Baltimore, Maryland, whose research includes ALS therapeutics. Scientists have not yet defined the details of DM progression in dogs, and a deep clinical understanding of the disease is necessary to produce significant results in a therapy trial. Even in genetically identical mice, Rothstein said, it took a decade to work out the disease progression needed to make the animals experimentally useful.

Degenerative myelopathy, or DM, was first recognized in German shepherds in the 1970s, and was long believed to be specific to the breed (Averill, 1973). It was also thought to only affect the spinal cord. Because owners often choose to euthanize a pet once it becomes partially paralyzed, it was not until recently that Coates and others were able to characterize the later stages of disease, particularly in Pembroke Welsh Corgis, another breed that suffers from DM. Being much smaller than German shepherds, Corgis are more easily nursed through the later stages of disease, providing researchers with further data on progression. Scientists found that in later-stage animals, motor neurons were affected as well. Sclerosis occurs in the lateral and dorsal spinal cord funiculi, peripheral nerves have axonal loss, and muscles atrophy. Similarly, the degeneration of motor neurons and axonal loss in ALS causes muscle atrophy. “The disease progression and clinical spectrum are similar to upper motor neuron ALS,” Coates said. Boxers, Rhodesian ridgebacks, and Chesapeake Bay retrievers are also susceptible to DM.

To trace the genetic cause of DM, Coates and O’Brien collaborated with Gary Johnson, also at the University of Missouri, and Kersten Lindblad-Toh and Claire Wade at the Broad Institute in Cambridge, Massachusetts. With support of the American Kennel Club-Canine Health Foundation, they carried out genomewide association mapping. This pointed to a potential DM locus on canine chromosome 31, which includes the dog SOD1 gene. “There were some excited phone calls about that,” O’Brien said. Resequencing of SOD1 revealed a missense mutation that changes canine SOD1 glutamate 40 to a lysine. Of 100 dogs with DM, 96 were homozygous for the SOD1 E40K mutation, suggesting the allele is recessive. (The other four animals, which were homozygous wild-type, likely had a different condition that mimicked DM symptoms, Coates said.) A codon for amino acid 40 of human SOD1 lies within a cluster of missense mutations that are associated with human ALS, including a glutamate-to-glycine point mutation at the position homologous to the canine E40K mutation.

Among healthy animals from the five breeds known to be susceptible to DM, there was an even distribution of wild-type homozygotes, mutant homozygotes, and heterozygotes. This suggests the allele is common in those five breeds, and that it shows incomplete penetrance. However, since many of the homozygous mutant animals were young, it is possible they might develop the disease later.

Mutant human SOD1 forms aggregates, and in neurons from ALS patients the protein is found in Lewy body like hyaline inclusions. To look for similar pathology, collaborator Martin Katz of the University of Missouri stained lower spinal cord sections from 19 dogs, covering all three genotypes, with rabbit anti-SOD1 antibody. In contrast to the diffuse SOD1 staining in neurons from normal dogs, all animals with DM had small, punctate SOD1-positive inclusions in their neurons. The heterozygote neurons exhibited an intermediate phenotype, inclusions that lightly stained for SOD1. While the dog SOD1 inclusions were small, not exactly like the large aggregates found in humans, the authors noted that the dogs were euthanized and may not represent end-stage disease as postmortem human tissues do.

The potential value of DM as a model for ALS lies in the quicker progression of the disease and the less stringent regulations for testing veterinary therapeutics. It would probably not be practical, O’Brien said, to breed mSOD1 dogs for research purposes. Since the age of onset is nine to 11 years, the costs for such a colony would be prohibitive. However, Coates thinks there are enough dogs within the general population to make a study sample; the overall disease prevalence of DM among all dogs is 0.19 percent, according to The Veterinary Medical Database, and the prevalence in certain breeds is likely to be much higher.

“This research represents a big advance in the understanding of degenerative myelopathy,” Thomas wrote “It provides useful clues regarding the underlying cause, it allows the first diagnostic test for the disease, and over the long term, it should help breeders plan their breeding strategy to hopefully decrease the incidence of this disease.” However, the prevalence of the mSOD1 allele means that breeders may not be able to completely eliminate it from the gene pool.

It’s not certain if DM is becoming more common or simply more commonly diagnosed. Purebred animals are particularly vulnerable to inherited disorders, and top show performers can cause a “popular sire effect” when bred extensively, leading to an increase in disease prevalence. “I really think over the years it has become more of a problem,” Coates said. Early symptoms are spastic paraparesis and ataxia; owners often notice their dogs stumbling and dragging their nails on the ground. The dogs also lose general proprioception in their limbs. If the researchers turn the animal's paw over, dogs with DM are not as quick to return it to normal position as a healthy dog.

The study authors are working on further pathological analysis of disease, such as examining cell body loss in the ventral horn and associated nerve loss, and developing tools to assess uniformity of disease progression, such as a rating scale for gait and electrophysiologic techniques. Since purebred animals are more alike genetically than people, it’s possible that the disease progression will be fairly uniform within each breed. If not, Rothstein said, “You need a lot of animals to get an answer. If you don't have stable numbers, you can never do therapeutics.” The exact value of dogs with DM as an ALS model, then, has yet to be determined.—Amber Dance.


  1. This work illustrates two frequently under-emphasized points about animal models of disease. First, although mice have proven fantastically useful and easy to manipulate experimentally, they rarely provide perfect models of any human disease. Second, genetic manipulations in mice often produce complex phenotypes that are more closely related to the function of the transgene than to the human disease that they are aiming to model. Our high rate of failure in getting therapeutically useful compounds from preclinical mouse models to the target human population is certainly related to both of these points; more work on complementary models (canines, primates, etc.) is essential.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Neuroscience: Standard model. Nature. 2008 Aug 7;454(7205):682-5. PubMed.
  2. . Degenerative myelopathy in the aging German Shepherd dog: clinical and pathologic findings. J Am Vet Med Assoc. 1973 Jun 15;162(12):1045-51. PubMed.

Further Reading