32P-postlabelling analysis of the DNA adducts formed by aristolochic acid I and II

We report the quantitation of DNA adducts in target and nontarget organs of male Wistar rats treated orally with five daily doses (10 mg/kg body wt) aristolochic acid I (AAI) or aristolochic acid II (AAII), the major components of the herbal drug aristolochic acid, a forestomach carcinogen in the rat. DNA adducts were detected and analysed using the nuclease P1-enhanced variation of the Randerath 32P-postlabelling assay. The highest level of DNA adducts formed was by AAI in the target organ, forestomach (330 +/- 30 adducts/10(8) nucleotides), but high levels were also observed in a non-target tissue, the glandular stomach (180 +/- 15). Lower amounts of adducts were detected in liver, kidney and urinary bladder epithelium. With AAII the binding levels were generally lower than the AAI, the highest level of adducts being detected in kidney (80 +/- 20 adducts/10(8) nucleotides) and lower levels in liver, stomach and urinary bladder epithelia. Adduct patterns similar to those in vivo were observed in two new in vitro assays. Rat faecal bacteria were shown to be able to activate AAI and AAII to reactive species, which were trapped with exogenous calf thymus DNA and analysed by postlabelling. Incubation of AAI and AAII in explanted rat stomach held in short-term organ culture resulted in DNA adduct formation in the epithelia of both forestomach and glandular stomach. To assign the recently characterized in vitro nucleoside adducts of AAI to the bisphosphate derivatives, a new ion-pair HPLC procedure on a reversed-phase column was developed. By monitoring Cerenkov radiation on-line, a good separation of AAI adducts was observed, demonstrating that adducts formed in vivo were chromatographically indistinguishable with those formed in vitro, and previously characterized as an aristolactam I moiety bound covalently to the exocyclic amino groups of deoxyadenosine and deoxyguanosine.