The puzzling toxicity of superoxide

O - is a small, hard, non-polarizable anion whose thermodynamic stability increases with the ability of the solvent molecules to solvate the anion. It is highly solvated in aqueous medium. Consequently, the redox potential 0,[02/02-] is close to —0.6 V in aprotic solvents, versus —0.33 V in water [22], 

This means that aqueous 02 will not penetrate the hydrophobic compartment of a cell, and that 02 is not likely to be efficiently formed from the direct reduction of oxygen in most biological membranes where strong reducing agents are not in high steady-state concentrations. 

Salts of 02 can be solubilized as crown-ethers in aprotic solvents, where 02 then behaves as a powerful nucleophile [73], but this property is essentially not observed in water where its spontaneous dismutation to O2 and H2O2 proceeds efficiently around neutral pH. The pKa of H02/02 is only 4.8, but deprotonation of many organic substrates by 02 is driven to the right by the accelerated dismutation of (H02 + 02 )  

Besides the above hypothesis, there are only a few documented cases where superoxide toxicity does not involve H2O2 and/or metal ions. Accordingly, for the last decade, it was repeatedly stated that 02 served as a precursor of a much more reactive and oxidizing species such as OH. Because the Haber-Weiss reaction was extremely slow, it was suggested that 02 triggered OH production by recycling the reduced form of transition metal ions involved in Fenton reactions [42]  

This was based on the observation that not only SOD, but also catalase, some metal chelators and OH scavengers, often protected the systems studied from oxidative injury. A serious criticism comes to mind however, if one realizes that 

---