Previous evidence has indicated that either purine starvation or incorporation into DNA may be the dominant biochemical effect of the antileukemic agent 6-thioguanine (TG), depending on exposure conditions. Furthermore, it has been suggested that the paradoxical decrease in TG-induced cytotoxicity at high drug concentrations may be due to an antagonistic interaction between these two mechanisms, in which purine starvation inhibits DNA synthesis and, therefore, incorporation of TG into DNA. In this report we test the hypothesis that by concurrent treatment of L1210 cells with TG and the purine precursor 4-amino-5-imidazolecarboxamide (AIC) it is possible to alleviate DNA synthesis inhibition caused by high concentrations of TG, thus enhancing TG incorporation into DNA and TG-induced cell kill. Both the cytotoxic and cytokinetic results presented support this hypothesis. However, gross incorporation of TG into DNA was not increased by AIC under conditions in which a significant enhancement of cytotoxicity (i.e., 1 log) was observed. These findings suggest that the potentiating effect of AIC may be most prominent on the subpopulation of cells that are resistant to treatment with TG alone, and they demonstrate that the cytotoxic effects of TG treatments are more accurately reflected by observing specific cytokinetic changes (delayed late S/G2 arrest) than by measuring the average extent of TG incorporation into DNA within a given population. Finally, we propose that it may be possible to select conditions for administration of TG that favor one or the other cytotoxic mechanism, depending on whether the clinical objective is induction of remission (where rapid cell lysis due to purine starvation would be desired) or eradication of subclinical disease during remission (where proliferation-dependent cytotoxicity due to DNA incorporation should be more effective.

A method is presented for the quantitative analysis of delayed cytokinetic effects resulting from the treatment of L1210 cells with 6-thioguanine (TG). By using dual-parameter (DNA/protein) flow cytometry, we could observe the accumulation of late S/G2/M cells with abnormally high green fluorescence (i.e., protein content), indicative of unbalanced growth. The use of mitotic cells from a pseudotetraploid line (HT29) as external markers for both red and green fluorescence facilitated highly reproducible measurement of the mean green fluorescence (GFLmean) of the arrested late S/G2/M population. We found that the dose dependence of the observed GFLmean values followed the same unusual biphasic pattern as did cytotoxicity in this cell line, indicating that this parameter might be a suitable means of predicting TG-induced toxicity in vivo. We propose that the low background expected for this kind of measurement would make it particularly appropriate for the analysis of clinical specimens (e.g., mononuclear bone marrow cells) from leukemic patients receiving thiopurines, to monitor (and, hopefully, predict) their response to treatment.