DNA platination by cisplatin (CDDP) was investigated in peripheral blood mononuclear cells and ovarian cancer cells using atomic absorption spectroscopy. Plots showing the amount of platinum (Pt) bound to DNA versus the molar concentration of cisplatin in the incubation medium ([CDDP]) were nonlinear. For [CDDP] < about 5 microM, the amount of Pt bound to DNA increased slowly with added drug. However, for larger [CDDP], the slope of the plot increased significantly. To study the role of thiols in affecting cisplatin binding to DNA, cells were treated with N-ethylmaleimide, which modifies thiol groups, rendering them incapable of binding cisplatin. Analysis using high-pressure liquid chromatography showed that approximately 99% of cellular glutathione was modified by N-ethylmaleimide. A plot of the amount of Pt bound to DNA versus [CDDP] for thiol-blocked cells is linear, with a slope similar to that of unblocked cells at high [CDDP]. Neither S-2-(3 aminopropylamino)ethanethiol (WR-1065) nor mesna, when added at clinically achievable concentrations (i.e., < approximately 300 microM), affected DNA platination. However, DNA platination was totally abolished by millimolar concentrations of the drug thiols (approximately 1.25 mM WR-1065 or approximately 5 mM mesna). Thus, the data show that endogenous thiols intercept cellular cisplatin, but this mechanism is less important at high [CDDP]. Moreover, therapeutic concentrations of drug thiols do not significantly affect DNA platination. A simple model that reproduces the experimental results of the amount of cisplatin binding to DNA as a function of [CDDP], time, and thiol content is proposed. The model takes into account passage of cisplatin and thiols through the cell membrane, binding of cisplatin to cellular thiols, and platination of DNA.