Oxidative damage to DNA has been reported to occur in a wide variety of disease states. The most widely used "marker" for oxidative DNA damage is 8-hydroxyguanine. However, the use of only one marker has limitations. Exposure of calf thymus DNA to an .OH-generating system (CuCl(2), ascorbate, H(2)O(2)) or to hypochlorous acid (HOCl), led to the extensive production of multiple oxidized or chlorinated DNA base products, as measured by gas chromatography-mass spectrometry. The addition of peroxynitrite (ONOO(-)) (<200 microM) or SIN-1 (1mM) to oxidized DNA led to the extensive loss of 8-hydroxyguanine, 5-hydroxycytosine, 2,6-diamino-4-hydroxy-5-formamidopyrimidine, 2-hydroxyadenine, 8-hydroxyadenine, and 4,6-diamino-5-formamidopyrimidine were lost at higher ONOO(-) concentrations (>200 microM). Exposure of DNA to HOCl led to the generation of 5-Cl uracil and 8-Cl adenine and addition of ONOO(-) (<200 microM) or SIN-1 (1mM) led to an extensive loss of 8-Cl adenine and a small loss of 5-Cl uracil at higher concentrations (>500 microM). An .OH-generating system (CuCl(2)/ascorbate/H(2)O(2)) could also destroy these chlorinated species. Treatment of oxidized or chlorinated DNA with acidified nitrite (NO(2)(-), pH 3) led to substantial loss of various base lesions, in particular 8-OH guanine, 5-OH cytosine, thymine glycol, and 8-Cl adenine. Our data indicate the possibility that when ONOO(-), nitrite in regions of low pH or .OH are produced at sites of inflammation, levels of certain damaged DNA bases could represent an underestimate of ongoing DNA damage. This study emphasizes the need to examine more than one modified DNA base when assessing the role of reactive species in human disease.