Sir

We wish to comment on the recent Review SOD Mimetics Coming of Age, which was published in your journal1. The article focuses on Mn ii pentaazamacrocyclic ligand-based compounds. Other compounds with superoxide dismutase (SOD) mimetic activity, Mn iii metalloporphyrins and Mn iii salen complexes, are not thoroughly reviewed. The brief discussion of Mn iii–salen complexes mistakenly implies instability in aqueous media, citing only a piece of published information2 that EDTA inhibits a complicated coupled enzymatic activity assay conducted with one analogue (now known in the literature as EUK-8), without mentioning that, in the same publication, EDTA had no effect on the activity of a different analogue. (Indeed, in genuine stability studies, in which intact complex is directly measured, even EUK-8, essentially an early prototype, is stable for years in aqueous media and highly resistant to metal removal by EDTA.)

In discussion of these other two classes of mimetics, their ability to neutralize both superoxide ions as well as other reactive oxygen species is described, oddly, as a “general limitation”. This gives no credence to the far more widely held view that these other reactive oxygen species, most notably hydrogen peroxide, are more damaging and longer-lived than superoxide and, further, that by producing hydrogen peroxide, SOD and, presumably, selective SOD mimetics could themselves cause tissue damage. Evidence that excess SOD activity can be deleterious does not appear in this Review. The likely response of the authors to this comment would be that the only 'limitation' they intended to imply is that these other mimetics, unlike the Mn ii pentaazamacrocyclic ligand-based compounds, could not serve as selective probes for the role of superoxide in disease processes. However, this Review is focused on the therapeutic applications of such compounds. In this context, the term general limitation implies much more than that, and is therefore unwarranted in the face of evidence favouring broader selectivity for reactive oxygen species as being, in fact, an asset.

As a final point, a selective agent against superoxide, and, in particular, a biologically active one with SOD activity described as at least equivalent to that of the native enzyme, might be expected to show extraordinary activity in an in vivo model specifically designed to enhance superoxide-associated oxidative stress. Such a model — mice with Sod2 genetically deleted — has been described. In fact, the authors discuss this Sod2 knock-out mouse to illustrate the lethal consequences of removal of Sod2, the mitochondrial defence against superoxide. However, two pertinent papers, not cited in this Review, show that a Mn iii metalloporphyrin3 and three Mn iii–salen complexes4 are highly protective in Sod2 knock-out mice, rescuing lethal oxidative phenotypes and extending lifespan up to threefold. The most simple interpretation, supported by the protection of a mitochondrial enzyme from oxidative damage in treated animals4, is that these other two classes of Mn complexes can functionally replace Sod2 in these mice. Establishing efficacy in this model seems to be a logical thing to attempt, and it would be interesting to know how Mn ii pentaazamacrocyclic ligand-based compounds fare. But, even if such data are not yet available, it is important to recognize the protective effects of other SOD mimetics in an experimental system that so pointedly addresses the damage caused by superoxide in vivo.