Amyotrophic lateral sclerosis (ALS) is a devastating neurodegenerative disease characterized by selective loss of CNS motor neurons, leading to rapidly progressing muscle weakness, wasting, paralysis, and ultimately death. Multiple cell death pathways have been implicated in ALS pathogenesis, but the causal event remains unknown. One hypothesis is that metabolic dysfunction underlies pathogenesis, since alterations in energy metabolism and mitochondrial function occur in patients. In addition, expression of mutant Cu,Zn-superoxide dismutase (mSOD1), associated with approximately 25% of patients with familial ALS (fALS), can induce mitochondrial abnormalities. To determine whether metabolic defects contribute to disease onset in vivo, we examined the association between energetic defects and the onset of symptoms and pathologic events in fALS transgenic mice over-expressing the G93A SOD1 mutation. We measured glucose use rates in 49 brain regions and 9 spinal cord regions in conscious 60 and 120 day-old G93A mice (n=6/group) and wild-type littermates (n=9/group) by quantitative [14C]-2-deoxyglucose autoradiography. Glucose utilization rates were impaired in multiple brain components of the motor system in G93A mice as early as 60 days of age. This precedes the first detectable pathologic changes in G93A mice (in spinal motor neuron mitochondria at 70–80 days), and symptom onset (hind-limb weakness at 90–100 days; mice die at 130–150 days). At 60 days, glucose use was reduced in components of the corticospinal projection, notably primary motor cortex (Fr1) layers I-III (innervation sites) and V (projection zones) (−19%, p<0.005, Student's unpaired t-test). A pattern of hypometabolism also emerged in several areas synaptically associated with Fr1, including the pontine nuclei (−25%, p<0.05) and the pontine reticular formation (−17%, p<0.05) of the bulbospinal pathway, and in several thalamic relay nuclei. In contrast, within the rubrospinal pathway glucose use was significantly reduced in the red nucleus only at 120 days (−28%, p<0.005), and sensorimotor cortical regions showed no alterations. In the spinal cord, generally regarded as the crucial site of neurodegeneration in ALS, glucose metabolism remained normal at 60 days, but was markedly impaired in cervical and thoracic grey matter by 120 days. In an additional experiment, 21 month-old mice overexpressing human wildtype SOD1 showed no alterations in cerebral or spinal cord glucose use with age, implying that the changes detected in G93A mice are due to the SOD1 mutation rather than SOD1 overexpression. We also examined metabolite levels in G93A brain and spinal cord. HPLC revealed depletions in ATP levels in the cerebral cortex of G93A mice concomitant with glucose use changes, which was partially rescued by administration of creatine. Further, cortical ATP levels were reduced by >40% as early as 30 days of age, implying that reduced neuronal energy generation is an extremely early consequence of mSOD1 expression. In conclusion, these studies demonstrate that energetic defects occur earlier than any other pathogenic processes reported to date in G93A mice, and suggest that dysfunction within the corticospinal projection may precede alterations in spinal neurons in this ALS model. Overall, results support a critical role for metabolic dysfunction in the pathogenesis of ALS.
