N-[(2'-Dimethylamino)ethyl]acridine-4-carboxamide (DACA) is a new anticancer agent currently undergoing clinical trials. The metabolism of DACA to acridone metabolites by aldehyde oxidase (AO) (EC 1.2.3.1) appears to play a major role in its elimination in human patients and rodents. The aim of this study was to compare the ability of human, guinea pig, and rat AO preparations to metabolise DACA, and to determine if either animal model was appropriate for predicting AO-mediated DACA-drug interactions in humans. Both human and rodent liver samples were homogenised in buffer before sequential centrifugation to produce the cytosol fraction. Human supernatant underwent an additional ammonium sulphate precipitation procedure, which produced a 2-fold increase in enzyme activity per milligram of protein. After incubations with DACA (range, 0-200 microM), DACA-9(10H)-acridone formation was determined by HPLC analysis. Michaelis-Menten parameters, Km and Vmax, were determined from the best fit curves by nonlinear regression. Three of the four human liver preparations had similar DACA intrinsic clearance values (Vmax/Km) ranging from 0.27 to 0.35 mL/min/mg protein, whereas both the rat and guinea pig had approximately 7- and 160-fold greater intrinsic clearances, due to lower Km values in rats (4.5 +/- 0.7 microM) and guinea pigs (0.15 +/- 0.1 microM) compared with humans (28.3 +/- 8.3 microM, N = 4). Amsacrine, menadione, and 7-hydroxy-DACA were potent inhibitors of DACA metabolism in all three species, but 10-fold differences in IC50 values were apparent between species. In addition, SKF-525A was a potent inhibitor of the metabolism of DACA in rat cytosol but caused minimal inhibition in the guinea pig or human preparations. These results suggest that neither rat nor guinea pig AO preparations are suitable for predicting AO-mediated DACA-drug interactions in humans.