Recent data from several groups suggest that the primary mechanism of ß-amyloid neurotoxicity may be mediated by reactive oxygen species. To evaluate this hypothesis, we first compared the efficacy of antioxidant agents in preventing toxicity caused by oxidative insults (iron, hydrogen peroxide, and tert-butyl hydroperoxide) and ß-amyloid peptides in cultured rat hippocampal neurons. Tested antioxidants (propyl gallate, Trolox, probucol, and promethazine) generally provided significant protection against oxidative insults but not ß-amyloid peptides. Next, we examined whether ß-amyloid causes oxidative stress, by comparing levels of lipid peroxidation after exposure to either iron or ß-amyloid. In a cell-free system, iron but not ß-amyloid generated lipid peroxidation. In culture, both insults caused rapid increases in lipid peroxidation, with iron inducing higher levels at later time points. Pretreatment with the antioxidant probucol significantly reduced lipid peroxidation caused by both insults but only attenuated iron toxicity, suggesting that lipid peroxidation does not contribute directly to cell death induced by ß-amyloid. Finally, we observed that increasing basal levels of oxidative stress by pretreating cultures with subtoxic doses of iron significantly increased neuronal vulnerability to ß-amyloid. The ability of ß-amyloid to induce oxidative stress and the demonstration that oxidative stress potentiatesß-amyloid toxicity support the clinical use of antioxidants for AD. However, these data do not support the theory that the primary mechanism of ß-amyloid toxicity involves oxidative pathways, indicating a continued need to identify additional cellular responses to ß-amyloid that underlie its neurodegenerative actions. Key Words: ß-AmyloidAlzheimer's disease-Antioxidants-Free radicalsLipid peroxidation-Oxidative stress. J. Neurochem. 69,1601Neurochem. 69, -1611Neurochem. 69, (1997.sociated with neuritic dystrophy and reactive glia (Wisniewski and Terry, 1973;Probst et al., 1987;Rozemuller et al., 1989). The Aß hypothesis is further supported by our recent demonstration that the extent of Aß deposition in entorhinal cortex correlates with cognitive decline in AD patients (r '-.~-O.9) better than other pathological indices (Cummings et al., 1996). Causal contributions of Aß to AD neurodegeneration are also implied by analyses of genetic mutations linked to familial forms of AD. Specifically, several rare autosomal dominant mutations in the Aß precursor protein gene on chromosome 21 and relatively more prevalent mutations in the presenilin-1 and presenilin-2 genes on chromosomes 14 and 1, respectively, appear to precipitate AD at least in part by causing significant increases in either the overall production of Aß or the levels of particularly amyloidogenic X-42/43 Aß species (for review, see Selkoe, 1996).An important contributory role of Aß in AD neurodegeneration is also supported by in vitro and in vivo experimental paradigms that have demonstrated neurodegenerative actions of Aß peptides (for...