Reports about the effects of ascorbate (vitamin C) on cultured cells are confusing and conflicting. Some authors show inhibition of cell death by ascorbate, whereas others demonstrate that ascorbate is cytotoxic. In this report, using three different cell types and two different culture media (Dulbecco's modified Eagle's medium and RPMI 1640), we show that the toxicity of ascorbate is due to ascorbate-mediated production of H2O2, to an extent that varies with the medium used to culture the cells. For example, 1 mM ascorbate generates 161 +/- 39 microM H2O2 in Dulbecco's modified Eagle's medium and induces apoptosis in 50% of HL60 cells, whereas in RPMI 1640 only 83 +/- 17 microM H2O2 is produced and no apoptosis is detected. Apoptosis is prevented by catalase, and direct addition of H2O2 at the above concentration to the cells has similar effects to ascorbate. These results show that ascorbate itself is not toxic to the cell lines used and that effects of ascorbate in vivo cannot be predicted from studies on cultured cells. The ability of ascorbate to interact with different cell culture media to produce H2O2 at different rates could account for many or all of the conflicting results obtained using ascorbate in cultured cell assays.

Administration of L-DOPA is commonly used to treat Parkinson's disease, yet controversy continues as to whether the dopamine arising from it aggravates neuronal loss. Several authors have reported cytotoxic effects of L-DOPA and dopamine on cultured cells, but others have not. In this report using the rat pheochromocytoma cell line PC12 and the M14 human melanoma cell line we show that dopamine-mediated cell death is not specific for neuronal cells. Moreover, our data show that both L-DOPA and dopamine interact with commonly used cell culture media, undergoing oxidation to generate hydrogen peroxide and dopamine semiquinones/quinones. Catalase and reduced glutathione could protect against cytotoxicity. These results suggest that caution needs to be employed when using cell culture studies to predict effects of L-DOPA and/or dopamine in vivo because of the extracellular generation of reactive species in the culture media.