A Drosophila model for amyotrophic lateral sclerosis: Dissecting the contributions of glia and motor neurons

Mutations in Cu,Zn-superoxide dismutase (SOD1) are found in 20% of cases of familial amyotrophic lateral sclerosis (fALS). These mutations confer an undefined toxic property on SOD1 that causes motor neuron disease. To provide a genetic system to study mechanisms of SOD-linked fALS, we developed a model in Drosophila melanogaster. We tested effects of wild type (wt) or familial mutant (G85R, A4V) SOD1 in several tissues, including the retina, motor neurons and glia. These studies revealed negative effects of both wt and mutant human SOD1 in the fruitfly, most of which are not seen when expressing the fly\u27s own SOD1, indicating that wt human SOD1 is harmful in a foreign environment. Upon expression in the retina, G85R accumulated along the eye perimeter by immunohistochemistry, and exhibited reduced solubility and higher molecular weight complexes by biochemical analysis, but without evidence of physical degeneration. When expressed in the motor neurons, both wt and mutant forms were deleterious, causing a climbing deficit in adult flies (10--20% reduction in climbing index at 21d) and a decrease in survival probability (from 45d onwards). Electrophysiological recordings in the dorsal longitudinal muscles revealed defective high-frequency neurotransmission in old flies expressing either wt or G85R (55d), with the cut-off frequency reduced to 90 Hz from the normal cut-off frequency of 140 Hz. Both wt and mutant SOD1 formed foci within motor neuron somata and neurites that were observed more frequently as the flies aged. Despite SOD1 accumulation, we detected no chaperone induction in motor neurons themselves. However, nearby glia showed increased immunoreactivity for the chaperone heat shock cognate 70, which we suggest is a non-cell autonomous effect of motor neuron SOD1. When expressed in glia alone, or together in glia and motor neurons, mutant but not wt SOD1 caused behavioral deficits, shortened lifespans, and electrophysiological defects. Taken together, these studies indicate that expression of human SOD1 in motor neurons and/or glia has behavioral and functional effects, displays aberrant protein properties, and induces a glial stress response. These phenotypes comprise a genetic model for fALS in Drosophila that will facilitate further analysis of SOD1-linked motor neuron dysfunction