The
anthracycline doxorubicin (DOX) is an exceptionally good
antineoplastic agent, but its use is limited by formation of metabolites which induce acute and chronic
cardiac toxicities. Whereas the acute toxicity is mild, the chronic toxicity can produce a life-threatening
cardiomyopathy. Studies in laboratory animals are of limited value in predicting the structure and reactivity of toxic metabolites in humans; therefore, we used an ethically acceptable system which is suitable for exploring DOX metabolism in human myocardium. The system involves cytosolic fractions from myocardial samples obtained during aorto-coronary bypass grafting. After reconstitution with
NADPH and DOX, these fractions generate the alcohol metabolite
doxorubicinol (DOXol) as well as DOX deoxyaglycone and DOXol hydroxyaglycone, reflecting reduction of the side chain carbonyl group,
reductase-type deglycosidation of the
anthracycline, and
hydrolase-type deglycosidation followed by carbonyl reduction, respectively. The efficiency of each metabolic route has been evaluated at low and high DOX:
protein ratios, reproducing acute, single-dose and chronic, multiple-dose regimens, respectively. Low DOX:
protein ratios increase the efficiency of formation of DOX deoxyaglycone and DOXol hydroxyaglycone but decrease that of DOXol. Conversely, high DOX:
protein ratios facilitate the formation of DOXol but impair
reductase- or
hydrolase-type deglycosidation and uncouple hydrolysis from carbonyl reduction, making DOXol accumulate at levels higher than those of DOX deoxyaglycone and DOXol hydroxyaglycone. Structure-activity considerations have suggested that aglycones and DOXol may inflict cardiac damage by inducing oxidative stress or by perturbing
iron homeostasis, respectively. Having characterized the influence of DOX:
protein ratios on deglycosidation or carbonyl reduction, we propose that the benign acute toxicity should be attributed to the
oxidant activity of aglycones, whereas the life-threatening chronic toxicity should be attributed to alterations of
iron homeostasis by DOXol. This picture rationalizes the limited protective efficacy of
antioxidants against chronic
cardiomyopathy vis-à-vis the better protection offered by
iron chelators, and forms the basis for developing analogues which produce less DOXol.