gests that in the course of aging and age-related neurodegenerative disease, neurons are increasingly subjected to apoptosis inducers. In some cases, these factors may act synergistically. For example, neuronal apoptosis may be significantly potentiated by the addition of subthreshold doses of β-amyloid and either excitotoxic or oxidative insults (Dornan et al., 1993; Koh et al., 1990; Mattson et al., 1992; Pike et al., 1997). Finally, mitochondrial damage may contribute to apoptosis as an intracellular effector. Mitochondria are a major source of free radicals and the release of cytochrome c is a potent inducer of caspase activation. Indeed, this organelle may be a prime target of aging and thus a contributor to the apoptosis cascade.
Some genetic risk factors also increase the probability that cells will engage apoptotic mechanisms. Overexpression of presenilin (PS) 1 or 2 results in an increased susceptibility of cells to apoptotic insults. PS mutations sensitize neurons to apoptosis by trophic factor withdrawal, metabolic insults and β-amyloid (Deng et al., 1996; Wolozin et al., 1996; Kim et al., 1997). It has been suggested that PS mutations cause perturbed calcium release from the endoplasmic reticulum and increased levels of oxidative stress (Mattson et al., 1998). Indeed, introduction of PS-1 into oocytes results in enhanced release of intracellular calcium and this is further increased by the presence of a PS-1 mutation. The effect appears to be downstream from the inositol trisphosphate receptor, because inositol trisphosphate injected directly into the cell elicits the increased release. Thus, because calcium homeostasis contributes to apoptosis, these gene products increase the probably that neurons may degenerate via apoptosis. Thus, in patients carrying PS mutations, apoptosis is likely to be one of the mechanisms of neuronal degeneration. The amyloid precursor protein itself appears capable of initiating apoptosis. There is growing evidence that the amyloid precursor protein is a receptor resembling a polypeptide hormone receptor (Nishimoto et al., 1997). The cytoplasmic portion of the protein contains a G-protein activator sequence (H657–K676) and will bind and activate G0. It has been suggested that the mutations result in a constitutively active G0 and that this causes apoptosis (Nishimoto et al., 1997; Yamatsuji et al., 1996).
Clearly then, there is ample potential for the induction of apoptosis mechanisms in the aging and Alzheimer's-affected brain. In this context, it is essential to determine if such pathways are activated in the Alzheimer's-affected brain. Indeed, a growing body of evidence supports this hypothesis (see Cotman et al., 1999).
In general, it appears as if brain aging acute phase responses often become chronic and escape the local microenvironment. The same mechanisms that are normally adaptive can become dysfunctional. This is ''dysfunctional plasticity," in which the same adaptive mechanisms turn against the system as overcompensation evolves, safety margins decline, and redundancy is lost.
Multiple-level cascades can shift the balance between beneficial and