Decreased glutathione levels cause overt motor neuron degeneration in hSOD1WT over-expressing mice

Abstract

Mutations in Cu/Zn-superoxide dismutase (SOD1) cause familial forms of amyotrophic lateral sclerosis (ALS), a fatal disorder characterized by the progressive loss of motor neurons. Several lines of evidence have shown that SOD1 mutations cause ALS through a gain of a toxic function that remains to be fully characterized. A significant share of our understanding of the mechanisms underlying the neurodegenerative process in ALS comes from the study of rodents over-expressing ALS-linked mutant hSOD1. These mutant hSOD1 models develop an ALS-like phenotype. On the other hand, hemizygous mice over-expressing wild-type hSOD1 at moderate levels (hSOD1^WT, originally described as line N1029) do not develop paralysis or shortened life-span. To investigate if a decrease in antioxidant defenses could lead to the development of an ALS-like phenotype in hSOD1^WT mice, we used knockout mice for the glutamate-cysteine ligase modifier subunit [GCLM(-/-)]. GCLM(-/-) mice are viable and fertile but display a 70-80% reduction in total glutathione levels. GCLM(-/-)/hSOD1^WT mice developed overt motor symptoms (e.g. tremor, loss of extension reflex in hind-limbs, decreased grip strength and paralysis) characteristic of mice models over-expressing ALS-linked mutant hSOD1. In addition, GCLM(-/-)/hSOD1^WT animals displayed shortened life span. An accelerated decrease in the number of large neurons in the ventral horn of the spinal cord and degeneration of spinal root axons was observed in symptomatic GCLM(-/-)/hSOD1^WT mice when compared to age-matched GCLM(+/+)/hSOD1^WT mice. Our results show that under conditions of chronic decrease in glutathione, moderate over-expression of wild-type SOD1 leads to overt motor neuron degeneration, which is similar to that induced by ALS-linked mutant hSOD1 over-expression.

Keywords: Amyotrophic lateral sclerosis; GCLM; Glutathione; SOD1.

Figure 1

Overt clinical features and decreased life span in GCLM(−/−)/hSOD1WT+ mice. A) Total glutathione (GSH+GSSG) content in different tissues from 48–52 weeks old GCLM(+/+) and GCLM(−/−) mice in the presence or absence of hSOD1^WT. CX, brain cortex; BS, brainstem; SC, spinal cord; Gastroc, gastrocnemius muscle. Each bar represents the mean±SD, n=3–4 animals. *Significantly different from its respective GCLM(+/+) tissue (p<0.05). B) Hind-limb grip strength starting at 32 weeks of age. At all ages, the values recorded for GCLM(−/−)/hSOD1WT+ mice are significantly different from GCLM(+/+)/hSOD1WT+ mice (p<0.05, at least n=7 in each group). # Significantly different from 32 weeks old GCLM(−/−)/hSOD1WT+ mice (p<0.05). C) Loss of hind limb extension reflex in a 52 weeks old GCLM(−/−)/hSOD1WT+ mice as compared with an aged-matched GCLM(+/+)/hSOD1WT+ mice. D) Lack of GCLM decreased the life span of hSOD1WT mice to a median survival of 477 days (aprox. 68 weeks) in GCLM(−/−)/hSOD1WT+ animals. Survival curves are significantly different p<0.0001. (At least n=7 in each group, see material and methods for details). E and F) hSOD1 protein expression in spinal cord extracts from 30 days old (E) and 48–52 weeks old (F) non-transgenic (NonTG), GCLM(+/+)/hSOD1WT+ and GCLM(−/−)/hSOD1WT+ mice. No difference was observed in hSOD1^WT expression levels between GCLM(+/+)/hSOD1WT+ and GCLM(−/−)/hSOD1WT+ mice when quantified and corrected by actin levels.

Figure 2

Motor neuron loss and astrogliosis in GCLM(−/−)/hSOD1WT+. A) Representative images from the lumbar ventral horn of 60 weeks old GCLM(−/−)/hSOD1WT+ mice and age-matched controls stained with cresyl violet. Higher magnification images of the indicated areas are shown in the lower panel. Scale bars: 50 μm. B) Number of large neurons in the ventral horn of the lumbar spinal cord from 60 weeks old mice of the indicated genotype. Data are presented as percentage of GCLM(+/+)/hSOD1WT− animals (mean±SD; n=4–5. * p<0.05). C) Luxol fast blue staining in cross sections of lumbar spinal cord roots from 60 weeks old GCLM(−/−)/hSOD1WT+ and age-matched GCLM(+/+)/hSOD1WT+ mice. Scale bar: 50 μm. D) Immunofluorescence against GFAP (red) in the anterior horn of lumbar spinal cord sections from 60 weeks old mice of the indicated transgenic genotypes. Nuclei were counterstained with DAPI. Higher magnification images of the indicated areas are shown in the lower panel. Scale bars: 20 μm. E) Relative optical density (O. D.) quantification of GFAP immunoreactivity in the grey matter of the anterior horn of lumbar spinal cord sections from the indicated genotypes. Data are presented as percentage of GCLM(+/+)/hSOD1WT− sections (mean±SD; * p<0.05)

Figure 3

Accelerated muscle denervation and wasting in GCLM(−/−)/hSOD1WT+ mice. Expression of the acetylcholine receptor gamma subunit (cholinergic receptor nicotinic gamma subunit, Chrng) (A) and acetylcholine receptor alpha1 subunit (cholinergic receptor nicotinic alpha 1 subunit, Chrna1) (B) in the gastrocnemius muscle of 60 weeks old animals from the indicated genotypes. For A and B data are expressed as percentage of GCLM(+/+)/hSOD1WT-animals (mean±SD; n=3–5). * Significantly different from GCLM(+/+)/hSOD1WT− (p<0.05). # Significantly different from GCLM(+/+)/hSOD1WT+ (p<0.05). C) Gastrocnemius muscle weight from 60 weeks old GCLM(+/+)/hSOD1WT+ and GCLM(−/−)/hSOD1WT+ mice (*p<0.05). D) Representative image of the gastrocnemius muscle from 60 weeks old GCLM(+/+)/hSOD1WT+ and GCLM(−/−)/hSOD1WT+ mice. Scale bar: 1 cm. E) Hematoxylin and eosin staining in gastrocnemius muscle sections from GCLM(+/+)/hSOD1WT+ and GCLM(−/−)/hSOD1WT+ mice. Scale bar: 200 μm.