The pathogenesis underlying the selective degeneration of dopaminergic neurons in Parkinson's disease (PD) 1 is not fully defined, but multiple lines of evidence implicate mitochondrial dysfunction. Thus, defects in complex I of the mitochondrial respiratory chain in the substantia nigra and evidence of oxidative stress have been noted in the brains of PD patients (1, 2). PD is also associated with exposure to pesticides (3, 4), many of which are either oxidants or mitochondrial toxins. Rapid onset of Parkinsonism in young adults following injection of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (5) occurs through its active metabolite, 1-methyl-4-phenylpyridinium, an inhibitor of mitochondrial complex I (6), potentially by depleting energy stores and inducing oxidative stress (7,8).Complex I is specifically and irreversibly inhibited by the neurotoxin rotenone, a hydrophobic pesticide epidemiologically linked to PD. Continuous, systemic, in vivo infusion of 2-3 mg/kg/day rotenone in rats over weeks produces relatively selective nigrostriatal dopaminergic neurodegeneration, with formation of ubiquitin-and ␣-synuclein-positive inclusions (9, 10). This low dose paradigm results in about 75% inhibition of complex I activity, and corresponds to a brain concentration of 20 -30 nM (9, 10). However, although these studies establish the principle that complex I dysfunction leads to degeneration in dopaminergic neurons, consistent levels of functional deficits related to complex I dysfunction have been difficult to reproduce (11). Accordingly, this in vivo model has not lent itself easily to the study of the mechanism of rotenone toxicity (12). Similarly, the use of primary cultures of dopaminergic neurons to study mechanisms of rotenone-induced neurodegeneration has been limited: short term, nanomolar rotenone administration in primary cultures of dopaminergic neurons does not produce selective dopaminergic damage (13). In addition, primary cultures contain a mixture of dopaminergic and nondopaminergic neurons and therefore do not lend themselves to molecular and biochemical analyses. The use of existing cell lines to study the effects of chronic rotenone exposure (14) also has shortcomings. Most cell lines do not exhibit robust dopaminergic phenotypes. In addition, the high rate of proliferation that characterize these cell lines confound the assessment of cellular susceptibility to rotenone, because proliferating cells may replace those that have succumbed to rotenone, and chronic exposure may lead to the selection of cells that become resistant to rotenone toxicity.To study the cellular mechanisms of rotenone-induced dopaminergic neuronal degeneration, we therefore developed a novel cellular model derived from immortalized midbrain MN9D cells (15), a dopaminergic neuronal cell line that has been extensively characterized as a model of dopaminergic neurons (16 -19). Using sodium butyrate (NaBu) (20) to differentiate MN9D cells and cells from a related, immortalized, non-dopaminergic midbrain neuronal cell line (MN9X ...