Neurodegeneration associated with Alzheimer's disease is believed to involve toxicity to beta-amyloid (A beta) and related peptides. Treatment of cultured rat hippocampal neurons with A beta 1-40 (1 microM) or the active fragment A beta 25-35 (1 microM) for 5 days led to a approximately 40-50% decrease in neuronal viability. The hydrophilic antioxidant ascorbic acid (300 microM) and the lipophilic antioxidant 2-mercaptoethanol (10 microM) both protected significantly against A beta neurotoxicity. Despite the protective effects of these antioxidants, both acute and chronic treatments with A beta 25-35 did not increase production of superoxide anions, as monitored with the fluorescent probe hydroethidine. Similarly, overexpression of Cu/Zn-superoxide dismutase using adenovirus-mediated gene transfer did not protect against A beta neurotoxicity. A beta neurotoxicity, however, was prevented in cultures infected with a recombinant, replication-defective adenovirus overexpressing the Ca2+ binding protein calbindin D28k. Transforming growth factor-beta 1 (TGF-beta 1) has been shown to protect neurons against both Ca(2+)- and free radical-mediated neuronal degeneration. We found that A beta neurotoxicity was significantly attenuated by single treatments with TGF-beta 1 (0.1-10 ng/ml) and prevented by repetitive treatments (10 ng/ml/day). The protective effects of TGF-beta 1 were associated with a preservation of mitochondrial potential and function, as determined with rhodamine-123-based microfluorimetry. Because both increased oxidative stress and pathophysiological Ca2+ fluxes can impair mitochondrial function, preservation of mitochondrial potential by TGF-beta 1 could be directly associated with its protection against A beta neurotoxicity. The ability of TGF-beta 1 to increase the expression of the anti-apoptotic proteins Bcl-2 and Bcl-XL is discussed in this context.