The effect of superoxide dismutase, catalase and l-cysteine on spontaneous and on mitomycin C induced chromosomal breakage in Fanconi's anemia and normal fibroblasts as measured by the micronucleus method

Abstract

Fanconi's anemia (FA) is characterized by high frequencies of chromosome aberrations and predisposition to cancer. There are two suggestions concerning the nature of the biochemical defect in FA: (a) that the condition is the result of a defect in DNA repair specifically of DNA crosslinks and (b) that the condition is associated with a deficiency in superoxide dismutase. Since it is not evident how either of these biochemical alterations would lead to the other, these hypotheses may be regarded as mutually exclusive. If a deficiency in superoxide dismutase is the primary biochemical defect in FA, then treatment with this enzyme should have a much greater effect in Fanconi than in normal cells. Furthermore, superoxide dismutase should be more effective than catalase and l-cysteine. On the repair-deficiency hypothesis, any effect of superoxide dismustase should be proportionately the same in Fanconi and normal cells, and need not be more effective than the catalase or l-cysteine. To test these predictions we have measured chromosomal breakage in 3 Fanconi fibroblast strains and 1 normal strain treated with superoxide dismutase, catalase, and l-cysteine. All 3 of the Fanconi strains were more sensitive to a DNA crosslinking agent, mitomycin C (MMC), than the normal strain and all had a high spontaneous level of chromosomal damage. Treatment by superoxide dismutase reduced both the spontaneous and MMC induced chromosomal breakage in all strains. There was no systematic difference among the Fanconi and the normal strains in the extent of the reductions. Treatment by catalase was even more effective than treatment by superoxide dismutase. We conclude that the effect of superoxide dismustase and of catalase and l-cysteine is not specific, at least in Fanconi cell strains that are MMC-sensitive. These results are in accord with the repair-deficiency hypothesis, but not with the superoxide dismutase deficiency hypothesis.