Dicoumarol relieves serum withdrawal-induced G0/1 blockade in HL-60 cells through a superoxide-dependent mechanism

This work was set to study how dicoumarol affects the cell cycle in human myeloid leukemia HL-60 cells. Cells were accumulated in G0/1 after serum deprivation. However, when cells were treated with 5 microM dicoumarol in serum-free medium, a significant increment in the number of cells in S-phase was observed. Inhibition of G0/1 blockade was confirmed by the increase of thymidine incorporation, the phosphorylation of retinoblastoma protein, and the promotion of cell growth in long-term treatments in the absence of serum. Dicoumarol treatment increased superoxide levels, but did not affect peroxide. Increase of cellular superoxide was essential for inhibition of G0/1 blockade, since scavenging this reactive species with a cell-permeable form of SOD and the SOD mimetics 2-amino-3,5-dibromo-N-[trans-4-hydroxycyclohexyl]benzylamine (ambroxol, 100 microM) and copper[II]diisopropyl salicylate (CuDIPS, 10 microM) completely abolished the effect of dicoumarol. However, N-acetyl-cysteine, overexpression of Bcl-2 or a cell-permeable form of catalase were not effective. 5-Methoxy-1,2-dimethyl-3-[(4-nitrophenol)methyl]-indole-4,7-dione (ES936), a mechanism-based irreversible inhibitor of NAD(P)H:quinone oxidoreductase 1 (NQO1), did not promote S phase entry, and dicoumarol still inhibited G0/1 blockade in the presence of ES936. We demonstrate that dicoumarol inhibits the normal blockade in G0/1 in HL-60 cells through a mechanism involving superoxide, but this effect is not dependent solely on the inhibition of the NQO1 catalytic activity. Our results send a precautionary message about use of dicoumarol to elucidate cellular processes involving oxidoreductases.