The ability of mitoxantrone to form DNA adducts was investigated in a series of human tumor cell lines consisting of human cervical cancer (HeLa), human breast cancer (MCF-7), and human neuroblastoma (IMR-32) cells. The mitoxantrone-resistant human promyelocytic leukemia cell line HL60/MX2 was also compared to the parental cell line HL60 in terms of adduct formation in cellular DNA, RNA, and protein. DNA adduct formation detected using [14C]mitoxantrone as a single agent occurred at very low levels but addition of the formaldehyde-releasing prodrug AN-9 (pivaloyloxymethyl butyrate) increased adduct formation considerably in all cell lines tested. Adduct formation increased when increasing ratios of AN-9 were used, and were observed at maximal levels when AN-9 addition was 4 h after the addition of mitoxantrone. However, low levels of adducts were observed when AN-9 addition was 16 h prior to mitoxantrone. The ability of [14C]mitoxantrone to form adducts with DNA, RNA, and protein was assessed in HL60 cells, and DNA was found to be the major substrate for adduct formation. RNA was also shown to be a good substrate while protein adduct levels were consistently very low. In mitoxantrone-resistant HL60/MX2 cells, DNA adduct levels were approximately fourfold lower. To establish the influence of DNA methylation on the ability of mitoxantrone to form adducts in cells, decitabine was used to reduce DNA methylation levels in cells prior to mitoxantrone treatment. This was clearly shown to influence adduct formation, with increasing decitabine levels leading to a decrease in the level of adducts observed in both IMR-32 and MCF-7 cell lines. Collectively, these results suggest that two major factors that influence the extent of mitoxantrone adduct formation in cells are the availability of formaldehyde and the extent of genomic DNA methylation.