on microglia cells. Importantly, compared with classical mediators of microglia activation (e.g., TNFα, LPS), oxidative stress was the most potent. TNRα can render microglia sensitive to FasL apoptosis by inducing Fas expression and down-regulation of Bcl-2 and Bcl-xL. (Spanaus et al., 1998). Further, hypoxia followed by re-oxygenation resulted in increased FasL expression (Vogt et al., 1998). Because the temporal and spatial patterns of microglia activation in injuries such as hypoxia coincides with the onset of DNA degradation and apoptosis in regions of selective neuronal loss, it has been suggested that microglia play a possible role in apoptosis.
Other risk factors can also induce FasL. In the Alzheimer's-affected brain, there is a reduction in the glutamate transporter (Masliah et al., 1996) that may contribute to neurodegeneration due to additional activation of glutamate receptors. Activation of NMDA receptors may also cause an increase in FasL. After a single injection of NMDA, there is an increase in FasL that begins after about 10 days and persists for up to 5 months. This increase may participate in long-term degeneration as part of a mechanism in the balance between repair and synaptic turnover/remodeling (Shin et al., 1998).
In summary, the gradual age-dependent increase of cytokines and their receptors capable of inducing apoptosis in the central nervous system could place neurons at increased risk for degeneration. It is possible that microglia and other cells in the brain also participate in the production of death ligands and thereby enhance the risk. Thus either autocrine or paracrine mechanisms may become active. Figure B-3 summarizes a model of the possible mechanism. At present a detailed study on the aging and Alzheimer's-affected brain has not been conducted, nor in fact has a detailed study at the anatomical level been reported in animal models. The development of many new reagents and the continued articulation of the mechanisms of Fas/caspase regulation provide a window of opportunity for the pursuit of this research direction.
Oxidative stress is a candidate for causing neuron dysfunction through a molecular cascade (Figure B-4). Oxidative stress is problematic for a number of reasons. Oxidation of proteins and enzymes within cells can interfere with their normal function (Stadtmann, 1992). Free radicals can also damage DNA, and a 50 percent increase in DNA oxidative damage has been reported in the human brain (Lyras et al., 1997; Gabbita et al., 1998; Mecocci et al., 1994; Lovell et al., 1999). Indeed, extensive DNA damage appears to accumulate with age and is particularly prominent in the aged dog brain and in humans (Su et al., 1997; Anderson et al., in press). Lipids are also vulnerable to oxidative stress and the levels of lipid peroxidation are elevated in the human brain, which in turn may induce membrane disturbances and loss of homeostasis within cells (Balazs and Leon, 1994; Palmer and Burns, 1994). In fact,