EXPERIMENTAL NEUROLOGY 131. 203-210 (19%) Motor Neuron-Astrocyte Interactions and Levels of Cu,Zn Superoxide Dismutase in Sporadic Amyotrophic Lateral Sclerosis S. A. O’REILLY,’ J. ROEDEA,* D. NAGY,* R. A. HALLEWELL,~ K. ALDERSON,$ S. L. ~JAF~KLUND,$ J. KUBY,* AND P. D. KUSHNER*,* *Departmen of Biology, San Francisco State University. San Francisco, California, TDepartment ofMolecular Biology. Imperial College, London, England, SDepartment of Neurology, University of Utah School ofMediclne, Salt Lake City, Utah; and IDepartment of Clinical Chemistry, Umea Uniuersitv. Umea. Sweden Copper, zinc superoxide dismutase (SODl) is in- volved in neutralizing free radicals within cells, and mutant forms of the enzyme have recently been shown to occur in about 20% of familial cases of amyotrophic lateral sclerosis (ALS). To explore the mechanism of SOD1 involvement in ALS, we have analyzed SOD1 in sporadic ALS using activity assays and immunocyto- chemistry. Analyses of SOD1 activity in washed eryth- rocytes revealed no difference between 13 ALS cases and 4 controls. Spinal cord sections from 6 ALS cases, 1 primary lateral sclerosis (PLS) case, and 1 control case were stained using three different antibodies to SODl. Since astrocytes are closely associated with motor neurons, antibodies to glial fibrillary acidic protein (GFAP) and vimentin were used as independent moni- tors of astrocytes. The principal findings from localiza- tions are: (1) normal motor neurons do not have higher levels of SOD1 than other neurons, (2) there was no detectable difference in SOD1 levels in motor neurons of ALS cases and controls, (3) ALS spinal cord dis- played a reduction or absence of SODl-reactive astro- cytes compared to the control and PLS cases, and (4) examination of GFAP-stained sections and morphom- etry showed that the normal close association between astrocytic processes and motor neuron somata was decreased in the ALS and PLS cases. These results indicate the disease mechanism in sporadic ALS may involve alterations in spinal cord astrocytes. c 1995 Academic Press. Inc. INTRODUCTION Point mutations in the gene encoding superoxide dismutase (SODl) are associated with some of the familial cases of amyotrophic lateral sclerosis (ALS) (5, 6, 26, 27, 34). These dominant mutations cause some reduction in enzyme activity (3, 5, 331, and it has been ’ To whom correspondence should be addressed at the Department of Anatomy, University of the Pacific, 2155 Webster Street, San Francisco, CA 94115. proposed that they all result in reduced enzyme stabil- ity, implying that a unifying property of the mutant enzymes may be an increased turnover and hence reduced activity in susceptible cells, such as motor neurons (5). Reduced activity of SOD1 may also exist in cases of sporadic ALS (4, 12-141, although this result has not been verified in other studies (3, 22, 33). There are currently two hypotheses for a superoxide-mediated mechanism by which SOD1 mutations could cause neuronal damage in ALS. McNamara and Fridovich (20) suggest that glutamate acts as an excitotoxin on motor neurons, releasing calcium and activating xanthine oxi- dase to produce superoxide, which cannot be adequately detoxified due to reduced SOD1 enzyme levels. Beckman et al. (1) propose that a reduction in SOD1 activity increases superoxide levels, which react with nitric oxide to form increased levels of toxic peroxynitrite; mutant SOD1 proteins may then react with peroxynitrate, forming a product that damages cells by increasing the rate of nitration of key cellular proteins. Recently, it has been shown that a transgenic mouse model for the disease can be made by over-expressing one of the SOD1 mutants associated with familial ALS (9). Interestingly, only mice expressing a large amount of the mutant enzyme, which was active in brain homogenates, devel- oped ALS-like symptoms. This finding suggests that the mechanism by which the mutant SOD1 proteins cause ALS may involve a property gained by the mutant enzyme, such as that proposed by Beckman et al. (1). Because of the excess human enzyme produced in the mouse model, most endogenous mouse SOD1 will be in mouse-human heterodimers, and it is possible that the turnover of both molecules is specifically increased in motor neurons, resulting in decreased SOD1 enzyme activity. Whatever disease mechanism occurs in the mouse model and in SODl-mutant humans, it will be important to compare their disease processes with the sporadic human disease. In this paper we have initiated this comparison by investigating the SOD1 status and the morphology of motor neurons and astrocytes from sporadic ALS patients and controls. 203 0014-4886/95$6.00 Copyright < 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.