The amyloid beta peptide (Abeta) is a product of the sequential gamma- and beta-secretase cleavage of amyloid precursor protein. Inhibitors of secretase enzymes have been proposed as a potential therapeutic strategy in the treatment of Alzheimer's disease. Here, we investigate the effect of inhibiting these key enzymes on the viability of a range of cell types. Treatment of rat cortical neurons for 24 hr with secretase inhibitors or an antibody that binds Abeta resulted in a marked reduction in cell viability, as measured by MTT reduction. Incubation with secretase inhibitors caused similar effects on other neuronal cell types (rat cerebellar granule neurons and the human SH-SY5Y cell line). Interestingly, rat astrocytes and a number of non-neuronal cell lines investigated (HEK293, DDT1-FM2, and human teratorhabdoid tumor cells) were unaffected by incubation with secretase inhibitors. The coincubation of Abeta1-40 prevented the toxicity of secretase inhibitors in neuronal cells. Abeta1-40 was protective in a concentration-dependent manner, and its effects were significant at concentrations as low at 10 pm. Importantly, the protective effects of Abeta were Abeta size-form specific, with the Abeta1-42 size form affording limited protection and the Abeta25-35 size form having very little protective effect. The present study demonstrates that inhibition of beta-or gamma-secretase activity induces death in neuronal cells. Importantly, this toxicity, which our data suggest is a consequence of a decline in neuronal Abeta levels, was absent in non-neuronal cells. This study further supports a key physiological role for the enigmatic Abeta peptide.