Whereas Cu,Zn-superoxide dismutase (Cu,Zn-SOD) is an essential antioxidant enzyme for superoxide dismutation in vivo(1 ), many reports have shown that an elevated level of Cu,Zn-SOD induces cell killing (2, 3), increases lipid peroxidation (4,5), and interferes with the transport of neurotransmitters (4, 6). All of these observations are archetypical results inducible by oxygen radicals. Cu,Zn-SOD (7) and FALS mutants (8) are known to develop neurodegeneration. These findings suggest that under certain conditions Cu,Zn-SOD may behave directly or indirectly as a pro-oxidant.In addition to the usual superoxide dismutation activity, Cu,Zn-SOD is known to exhibit anion binding capacity (9, 10), inactivation by its own reaction product H202 (11-13), and the purported peroxidase activity (14). Here we summarize our findings that Cu,Zn-SOD has free radical ('OH and scavenger-derived radicals)-generating activity (15,16). TheactiveCu,Zn-SOD can catalyze the generation of free 9 radicals from H202. In the presence of H202 and radical scavengers, especially with anionic character, both free 9 radicals and scavenger-derived radicals are produced (15, 16). We also found that the free radical-generating function of several FALS-associated Cu,Zn-SOD mutants is enhanced relative to that of the wild-type enzyme, while the dismutation activities remain unchanged (17,18).To monitor the production of free radicals, two spin traps, 5,5-dimethyl-l-pyrroline N-oxide (DMPO) and Ntert-butyl-c~-phenylnitrone (PBN), were used in our studies (15) to convert transient free radicals to stable free radicals adducts according to the following reactions:The nature of the trapped free radical was identified by EPR spectroscopy. When O2-" (dissolved KO 2 in dried dimethylsulfoxide) was added to the NaHCOJCO 2 buffer at pH 7.6 containing 100 mM DMPO, EPR OH adducts. DMPO-OH formation also required active Cu,Zn-SOD, because heat-inactivated Cu,Zn-SOD or Mn-SOD failed to produce these radical adducts. The direct addition of H202 in place of O2-" produced similar results. In the presence of .OH radical scavengers with anionic character such as HCO 2 or N3-, DMPO adducts of the scavenger-derived radicals, DMPO-CO2-" or DMPO-N3", were observed in addition to DMPO-OH. The concentration ratio between DMPO-OH and DMPO-scavenger-derived radical adducts observed in these experiments reflects competition reactions between DMPO and scavengers for free "OH radicals. Surprisingly, with neutral'OH radical scavengers, such as ethanol, the formation of DMPO-hydroxyethyl radical adducts was insignificant and the concentration of DMPO-OH formed was unaffected. In contrast, with a different spin trap, PBN, in place of DMPO, ethanol yields PBNhydroxyethyl radical adducts as expected in the competition reaction for free "OH radicals. These seemingly contradicting results may be caused by the different affinities of these spin traps for the positively charged active channel leading to the active site of Cu,Zn-SOD, where the'OH radicals are generated. To prove this hypo...