DNA damage by reactive species is associated with susceptibility to chronic human degenerative disorders. Anthocyanins are naturally occurring antioxidants, that may prevent or reverse such damage. There is considerable interest in anthocyanic food plants as good dietary sources, with the potential for reducing susceptibility to chronic disease. While structure-activity relationships have provided guidelines on molecular structure in relation to free hydroxyl-radical scavenging, this may not cover the situation in food plants where the anthocyanins are part of a complex mixture, and may be part of complex structures, including anthocyanic vacuolar inclusions (AVIs). Additionally, new analytical methods have revealed new structures in previously-studied materials. We have compared the antioxidant activities of extracts from six anthocyanin-rich edible plants (red cabbage, red lettuce, blueberries, pansies, purple sweetpotato skin, purple sweetpotato flesh and Maori potato flesh) using three chemical assays (DPPH, TRAP and ORAC), and the in vitro Comet assay. Extracts from the flowering plant, lisianthus, were used for comparison. The extracts showed differential effects in the chemical assays, suggesting that closely related structures have different affinities to scavenge different reactive species. Integration of anthocyanins to an AVI led to more sustained radical scavenging activity as compared with the free anthocyanin. All but the red lettuce extract could reduce endogenous DNA damage in HT-29 colon cancer cells. However, while extracts from purple sweetpotato
OPEN ACCESSInt. J. Mol. Sci. 2009, 10 1082 skin and flesh, Maori potato and pansies, protected cells against subsequent challenge by hydrogen peroxide at 0 o C, red cabbage extracts were pro-oxidant, while other extracts had no effect. When the peroxide challenge was at 37 o C, all of the extracts appeared prooxidant. Maori potato extract, consistently the weakest antioxidant in all the chemical assays, was more effective in the Comet assays. These results highlight the dangers of generalising to potential health benefits, based solely on identification of high anthocyanic content in plants, results of a single antioxidant assay and traditional approaches to structure activity relationships. Subsequent studies might usefully consider complex mixtures and a battery of assays.