Oxidative damage to nuclear DNA is known to involve site-specific Fenton-type chemistry catalysed by redox-active iron or copper in the immediate vicinity of DNA. However, the presence of transition metals in the nucleus has not been shown convincingly. Recently, it was proposed that a major part of the cellular pool of loose iron is confined within the acidic vacuolar compartment [Yu, Persson, Eaton and Brunk (2003) Free Radical Biol. Med. 34, 1243-1252; Persson, Yu, Tirosh, Eaton and Brunk (2003) Free Radical Biol. Med. 34, 1295-1305]. Consequently, rupture of secondary lysosomes, as well as subsequent relocation of labile iron to the nucleus, could be an important intermediary step in the generation of oxidative damage to DNA. To test this concept we employed the potent iron chelator DFO (desferrioxamine) conjugated with starch to form an HMM-DFO (high-molecular-mass DFO complex). The HMM-DFO complex will enter cells only via fluid-phase endocytosis and remain within the acidic vacuolar compartment, thereby chelating redox-active iron exclusively inside the endosomal/lysosomal compartment. Both free DFO and HMM-DFO equally protected lysosomal-membrane integrity against H2O2-induced oxidative disruption. More importantly, both forms of DFO prevented H2O2-induced strand breaks in nuclear DNA, including telomeres. To exclude the possibility that lysosomal hydrolases, rather than iron, caused the observed DNA damage, limited lysosomal rupture was induced using the lysosomotropic detergent O-methyl-serine dodecylamine hydrochloride; subsequently, hardly any DNA damage was found. These observations suggest that rapid oxidative damage to cellular DNA is minimal in the absence of redox-active iron and that oxidant-mediated DNA damage, observed in normal cells, is mainly derived from intralysosomal iron translocated to the nucleus after lysosomal rupture.