Prevention of the degeneration of human dopaminergic neurons in an astrocyte co-culture system allowing endogenous drug metabolism.

BACKGROUND AND PURPOSE: Few neuropharmacological model systems use human neurons. Moreover, available test systems rarely reflect functional roles of co-cultured glial cells. There is no human in vitro counterpart of the widely used 1-methyl-4-phenyl-tetrahydropyridine (MPTP) mouse model of Parkinson's disease EXPERIMENTAL APPROACH: We generated such a model by growing an intricate network of human dopaminergic neurons on a dense layer of astrocytes. In these co-cultures, MPTP was metabolized to 1-methyl-4-phenyl-pyridinium (MPP(+) ) by the glial cells, and the toxic metabolite was taken up through the dopamine transporter into neurons. Cell viability was measured biochemically and by quantitative neurite imaging, siRNA techniques were also used. KEY RESULTS: We initially characterized the activation of PARP. As in mouse models, MPTP exposure induced (poly-ADP-ribose) synthesis and neurodegeneration was blocked by PARP inhibitors. Several different putative neuroprotectants were then compared in mono-cultures and co-cultures. Rho kinase inhibitors worked in both models; CEP1347, ascorbic acid or a caspase inhibitor protected mono-cultures from MPP(+) toxicity, but did not protect co-cultures, when used alone or in combination. Application of GSSG prevented degeneration in co-cultures, but not in mono-cultures. The surprisingly different pharmacological profiles of the models suggest that the presence of glial cells, and the in situ generation of the toxic metabolite MPP(+) within the layered cultures played an important role in neuroprotection. CONCLUSIONS AND IMPLICATIONS: Our new model system is a closer model of human brain tissue than conventional cultures. Its use for screening of candidate neuroprotectants may increase the predictiveness of a test battery.