Differential vulnerability of oculomotor, facial, and hypoglossal nuclei in G86R superoxide dismutase transgenic mice.

In recent years, several mouse models of amyotrophic lateral sclerosis (ALS) have been developed. One, caused by a G86R mutation in the superoxide dismutase-1 (SOD-1) gene associated with familial ALS, has been subjected to extensive quantitative analyses in the spinal cord. However, the human form of ALS includes pathology elsewhere in the nervous system. In the present study, analyses were extended to three motor nuclei in the brainstem. Mutant mice and control littermates were evaluated daily, and mutants, along with their littermate controls, were killed when they were severely affected. Brains were removed after perfusion and processed for Nissl staining, the samples were randomized, and the investigators were blinded to their genetic status. Stereologic methods were used to estimate the number of neurons, mean neuronal volumes, and nuclear volume in three brainstem motor nuclei known to be differentially involved in the human form of the disease, the oculomotor, facial, and hypoglossal nuclei. In the facial nucleus, neuron number consistently declined (48%), an effect that was correlated with disease severity. The nuclear volume of the facial nucleus was smaller in the SOD-1 mutant mice (45.7% difference from control mice) and correlated significantly with neuron number. The oculomotor and hypoglossal nuclei showed less extreme involvement (<10% neuronal loss overall), with a trend toward fewer neurons in the hypoglossal nucleus of animals with severe facial nucleus involvement. In the oculomotor nucleus, neuronal loss was seen only once in five mice, associated with very severe disease. There was no significant change in the volume of individual neurons in any of these three nuclei in any transgenic mouse. These results suggest that different brainstem motor nuclei are differentially affected in this SOD-1 mutant model of ALS. The relatively moderate and late involvement of the hypoglossal nucleus indicates that, although the general patterns of neuronal pathology match closely those seen in ALS patients, some differences exist in this transgenic model compared with the progression of the disease in humans. However, these patterns of cellular vulnerability may provide clues for understanding the differential susceptibility of neural structures in ALS and other neurodegenerative diseases. Copyright 2000 Wiley-Liss, Inc.