Oxidative Stress and the CNS

Bernhard H.J. Juurlink
juurlink@duke.usask.ca

     Another example of a positive feedback loop is exemplified by amyotrophic lateral sclerosis (ALS).  Dr. Hugon discusses the evidence that excitotoxicity and oxidative stress is involved in the progression of ALS.  Excessive amounts of glutamate, or prolonged action of glutamate, at receptors result in excessive Ca2+ influx, via voltage-dependent Ca2+ channels or via glutamate receptor-linked channels that allow Ca2+ influx such as the N-methyl-D-aspartate channel, resulting in prolonged periods of elevated intracellular Ca2+.  Such elevated Ca2+ can activate Ca2+-dependent enzymes such as phospholipase A2 that can release arachidonic acid whose metabolism generates superoxide anion.  Mitochondria Ca2+ cycling also results in mitochondrial damage increasing the production of superoxide anion.  The elevated Ca2+ also stimulates release of more glutamate.  Hence, we have established a vicious interactive spiral involving glutamate-mediated excitotoxicity, prolonged rises in intracellular Ca2+, increase in oxidative stress and depletion of ATP.  That oxidative stress which overwhelms the antioxidant capacity of neurons plays a major role in the progression of ALS is supported by the findings of a small proportion of familial ALS where the cause has been demonstrated to be a mutation in the Cu,Zn-SOD.  Dr. Yim and colleagues have demonstrated that such mutations generally do not affect the dismutase activity of the enzyme and that a gain-of-function is involved in the deleterious effect of the mutation.  This gain of function appears to be the slightly increased affinity of the enzyme for hydrogen peroxide (Km of wildtype is 44 mM while that of mutants range from 13 to 25 mM) such that the Cu of the Cu,Zn-SOD can convert the hydrogen peroxide to the hydroxyl radical.  So a mutation that slightly increases the chances of the Cu being available to interact with hydrogen peroxide and, therefore, slightly increases the oxidative stress of the neuron ultimately causes the death of the individual.  Two aspects intrigue me: One is why are motoneurons preferentially affected, and the Second is why does this mutation manifest itself clinically not until after one enters the fifth decade of life.  It is likely that motoneurons are preferentially affected because they have the highest rate of oxidative metabolism and, therefore, the highest production of hydrogen peroxide.  Their high rate of oxidative metabolism is likely due to the fact that they have huge cellualr machinery to maintain - very long axons (in certain individuals more than a metre in length for each of upper and lower motoneurons) and complex dendritic tree receiving ~10,000 synaptic inputs resulting in a large expenditure of ATP to maintain membrane potentials.  Why does the problem not manifest itself until the fifth decade or later?  I think that this is likely due to the fundamental reason that over time, our cellular machinery undergoes the wear and tear of oxidative stress with more mitochondrial DNA problems, less efficient production of ATP.  There is evidence from rats that with age: i) superoxide anion production in the CNS increases, likely due to increasing damage of mitochondria; ii) reduced-glutathione (GSH) decreases; iii) glutathione reductase activity decreases, indicative of increasing damage to thiol-containing proteins.  

     The decrease in GSH with age is significant since GSH plays a central role in peroxide scavenging as the electron donor for glutathione peroxidase.  Hence, one possible way of delaying the onset of ALS, or other neurodegenerative diseases, may be by elevating neuronal GSH.  We have shown that in cells with a low GSH content, a relatively small increase in GSH can markedly increase the ability of the cells to cope with oxidative stress.  Dr. Henderson and colleagues have demonstrated that the administration of the cysteine (limiting amino acid precursor for GSH synthesis) prodrug, N-acetylcysteine, delays the onset of motoneuron disease in a mouse model of motoneuron disease.  This is likely due to an increase in motoneuron GSH content.  As I have mentioned elswhere, I think that if one could increase neuronal GSH by a relatively modest amount, it seems very likely that one could delay the onset (perhaps indefinitely) of neurodegenerative diseases such as Parkinsonism, Alzheimer’s disease and ALS.  These diseases appear when one ages and are fueled by the positive feedback loops of oxidative stress.  I think that it is significant that the brain has ~25% of the GSH content of liver and that, as shown by Dr. Paterson, a relatively short time of removal of sulfur amino acids from the diet causes a 10-15% decrease in brain GSH, making the brain even more vulnerable to perturbations.

     A few things puzzle me in this symposium.  One is the apparent absence of Mn-SOD from the mitochondria of oligodendroglia as shown by Drs. Perraut and Tholey.  Neurons do have detectable Mn-SOD as demosntrated by Dr. Lindenau and colleagues.  Our laboratory has shown that oligodendroglia have a high rate of oxidative metabolism.  This is demonstrated by high rate of CO2 production from pyruvate in comparison to astrocytes (unpublished observations).  We have also demonstrated that oligodendroglia produce 6 times as much strong oxidants under normal culture conditions as do astrocytes (Thorburne and Juurlink, 1996) and have more peroxidized lipids than astrocytes (Husain and Juurlink, 1995).  In addition, we have demonstrated that oligodendroglial mitochondria have a much stronger immunocytochemical signal for Hsp60 (Juurlink, 1997) suggesting a higher turnover rate for mitochondrial proteins, likely due to more oxidative damage.  It seems most peculiar that cells which are so oxidatively active should have no detectable Mn-SOD.  My question to Drs. Perraut and Tholey is whether they have isolated oligodendroglial mitochondria and measured superoxide dismutase activity?  Is it possible that the Mn-SOD in mitochondria is post-translationally modified such that your antibody does not recognize it?

     The other thing that puzzles me is why neurons which have a poor ability to reduce oxidized-ascorbate should have an order of magnitude more ascorbate than astrocytes which have a very good ability to reduce oxidized-ascorbate.  Note, I do not question the interpretation of the data by Dr. Rice since I think that it is correct.  Is there any evidence for ascorbate metabolic coupling between neurons and astrocytes with oxidized-ascorbate transferred from neurons to astrocytes and reduced-ascorbate transferred from astrocytes to neuron?  And why should there be so much ascorbate in neurons?  Does it have a function other than reducing oxidized-tocopherol?  What would this be?