“There is a need for large animal species models for ALS,” said Amelie Gubitz, Program Director for ALS research at the National Institute of Neurological Disorders and Stroke in Bethesda, Maryland. “They fill an important gap,” she said, between the tiny brains and spinal cords of mice and the human-sized nervous system. Gubitz, who organized a satellite meeting to discuss ALS models on 12 November, noted that a golden retriever model has proved useful in several studies of Duchenne’s muscular dystrophy (Banks and Chamberlain, 2008).

At the satellite meeting, Bryden Stanley of Michigan State University in East Lansing presented a poster on dogs with ILP. These animals suffer problems in swallowing that mirror the bulbar onset in one-quarter of people who get ALS. There is no real animal model for bulbar-onset ALS, said Bob Brown of the University of Massachusetts in Worcester, although some mice may exhibit symptoms in the neck area. Thus far, Stanley has been unable to find a genetic cause for the ILP.

In another poster session during the SfN general meeting, Brandie Morgan of the University of Missouri in Columbia discussed her data on axon counts in dogs with DM. Morgan works with Missouri researcher Joan Coates, who presented her dog model two years ago at the SfN meeting in Washington, DC (see ARF related news story and ARF related news story on Awano et al., 2009; reviewed in Coates and Wininger, 2010). These animals have a missense mutation in superoxide dismutase 1 (SOD1); SOD1 mutations are responsible for some cases of inherited ALS.

The two models might turn out to complement each other, Gubitz said, with Coates’s model mimicking familial ALS and Stanley’s standing in for bulbar-onset, and possibly sporadic, forms.

Dog models offer many advantages. “The dog and human share an incredibly close environment,” Stanley told ARF in an interview. Thus, they may be exposed to the same toxins that some researchers suspect contribute to ALS. Because owners choose to euthanize their pets at different stages, dogs offer the opportunity to examine disease pathology before end-stage, which is impossible in people.

Dogs, given their inbred genomes, are also convenient for genomewide association studies (GWAS), said Dennis O’Brien, the director of the veterinary comparative neurology program at the University of Missouri. Dogs have less well-shuffled genomes than do people, he said, so their linkage disequilibrium groups are large, and scientists can find gene associations with a relatively small sample. “That is probably the biggest advantage that the dogs have,” O’Brien said. “The GWAS is so powerful.”

Bulbar Barks
Stanley originally set out to study not ALS but ILP, a common disorder in elderly dogs, particularly Labrador retrievers. Degeneration of the laryngeal nerves leads to paralysis of the laryngeal muscles, and symptoms include gagging, throat clearing, and a raspy or hoarse-sounding bark. Dogs tend to get sick around 10 years of age, and generally last for one to three years longer, Stanley told ARF.

Stanley and colleagues followed dogs with ILP for a year and performed neurological tests, looking at gait, muscle tone, and reflexes (see Stanley et al., 2010). Of 32 dogs with ILP, 10 had neurologic problems when they enrolled in the study. Half of the dogs had them by six months, and all had neurological signs within a year of enrollment. Eventually, the dogs with ILP became paralyzed. At autopsy, the dogs with ILP evinced muscle atrophy and loss of axons in the lumbar spinal cord’s ventral roots. In comparison, none of the 34 control dogs showed neurological symptoms.

Stanley found herself thinking that “There must be something in humans like this.” She started talking to neurologists, who immediately saw the similarities between ILP and bulbar ALS, which starts with problems in speaking and swallowing. “Their jaws dropped,” Stanley told ARF. “They were just taken away with the similarity.”

It is too early to consider ILP a definite ALS model, Stanley said, but it is promising enough to warrant further study. “It is going to be a good model for something,” O’Brien told ARF. “I do not know if we know enough yet to see if it is going to be a good model for ALS.”

Stanley is working to further characterize ILP, and recruited Michigan State neuropathologist Howard Chang to help analyze tissue samples. “Clearly, we have a problem of neuropathy,” he said. “The question is, where does the neuropathy come from?” He said he has only seen a few samples so far, and needs more control tissues from healthy animals before forming any conclusions. Stanley is searching for a genetic cause of ILP, and has tested several of the usual ALS suspects: SOD1, angiogenin, FUS, TDP-43, and FIG4. So far, she has had no hits, and plans to continue the search with a GWAS. Brown told ARF he hopes these studies will identify new ALS genes.

Dog Genes
The University of Missouri researchers do have a genetic link for degenerative myelopathy. This is an advantage, Gubitz said, because research colonies can be bred. Indeed, the scientists have already begun a proof-of-concept study of ALS treatment in bred-for-research animals, Coates told ARF, although she said it is too early to reveal any details.

Coates and colleagues first studied DM, which has its onset in the hind limbs, in Pembroke Welsh corgies and boxers. “We are continuing to find other breeds or dogs that are affected with this disease,” Coates told ARF in an interview. At least 17 breeds can get the disease, and more than two dozen breeds carry the missense SOD1 allele at a frequency greater than 20 percent. Most of the dogs have the same missense mutation—E40K—although the scientists may have found a second SOD1 mutation in one Bernese mountain dog, Coates said.

In people, a single faulty SOD1 gene amounts to a 100 percent certainty of getting ALS. That is not true in dogs, where few heterozygotes get sick and some homozygotes escape disease. This incomplete penetrance, Coates suggested, may be because some dogs die before the disease has time to develop. If that is true, she thinks there must be genetic factors that enhance or reduce risk. The researchers are conducting a GWAS to find those risk factors.

Coates and colleagues are also interested in finding biomarkers that would allow them to follow disease progression. At the ALS International Conference to be held in Orlando, Florida, later this month, Coates will present her work with a method she calls MUNE, for motor unit number estimation. Using electronic stimulation and recording, MUNE tells the scientists how many neuromuscular junctions a particular muscle has.

While much of the work in Coates’s group has focused on upper motor neuron degeneration, Morgan wanted to know whether axon fiber numbers drop in lower motor neurons. She counted axon fibers in cross-sections from T8 vertebra motor roots. Two healthy boxers had an average of 5,788 axons. But in seven boxers with DM, Morgan discovered an average of 4,603 axons. This reduction in fiber count is similar to ALS pathology.

Given the heterogeneity of ALS in people, it is unlikely that any single model will be the only one used in research, Gubitz said. Instead, she envisions a panel of models that researchers will use in parallel. Dogs just might be on the list.—Amber Dance.