Treatment of human HL-60 or KG1A leukemia cells with the topoisomerase II inhibitor etoposide resulted in extensive DNA degradation. When DNA integrity was analyzed by agarose gel electrophoresis, a nucleosomal ladder became evident 1.5–2 h after addition of etoposide to cells, increased in intensity over 6 h, and persisted at 24 h. Six h after addition of the drug, 94 ± 4% of the cells excluded trypan blue even though as much as 90% of the DNA had been degraded to oligosomal fragments. Exposure of cells to 10 µg/ml (17 µm) etoposide for as little as 45 min was sufficient to induce this DNA damage 4 h later. Preincubation with dinitrophenol abolished the effect of etoposide, suggesting that an energy-requiring step occurred prior to or during the endonucleolytic cleavage. In contrast, the effect of etoposide was not prevented by preincubation of HL-60 cells with the RNA synthesis inhibitor 5,6-dichloro-1-β-ribofuranosylbenzimidazole or the protein synthesis inhibitors cycloheximide or puromycin. On the contrary, high concentrations of 5,6-dichloro-1-β-ribofuranosylbenzimidazole, cycloheximide, or puromycin by themselves induced the same endonucleolytic cleavage, as did a variety of diverse cytotoxic agents, including camptothecin (0.1 µm), colcemid (0.1 µg/ml), cis-platinum (20 µm), methotrexate (1 µm), and 1-β-d-arabinofuranosylcytosine (3 µm). These results suggest that endonucleolytic DNA damage by a preexisting cellular enzyme occurs soon after treatment of HL-60 cells with any of a variety of cytotoxic agents. The observation that a variety of nuclear proteins [including poly(ADP-ribose) polymerase, lamin B, topoisomerase I, topoisomerase II, and histone H1] are degraded concomitant with the DNA fragmentation calls into question the selectivity of the degradative process for DNA. The implications of these results for (a) current theories which focus upon endonucleolytic damage of DNA as a critical early event during cell death, and (b) use of topoisomerase-directed drugs to map topoisomerase-binding sites in active chromatin are discussed.
This work was supported in part by grants from the American Cancer Society (Maryland Division), The Andrew Mellon Foundation, and by Grant CA 06973 from the NIH.