OBJECTIVES: Ingestion of aristolochic acid (AA) is associated with development of urothelial tumors linked with aristolochic acid nephropathy, and is implicated in the development of Balkan endemic nephropathy-associated urothelial tumors. Aristolochic acid I (AAI), the major toxic component of AA, is more toxic than its demethoxylated derivate AAII. A different enzymatic conversion of both carcinogens might be one of the reasons explaining this feature. Therefore, the present study has been designed to compare efficiency of human NAD(P)H:quinone oxidoreductase (NQO1) and phase II enzymes such as sulfotransferases (SULTs) and N,O-acetyltransferases (NATs) to activate AAI and AAII in vitro. In addition, to investigate the molecular mechanisms of AAI and AAII reduction by human NQO1, molecular modeling was used to compare interactions of AAI and AAII with the active site of this enzyme. METHODS: DNA adduct formation by AAI and AAII was investigated by the nuclease P1 version of the 32P-postlabeling method. In silico docking, employing soft-soft (flexible) docking procedure, was used to study the interactions of AAI and AAII with the active site of human NQO1. RESULTS: Human NQO1 activated AAI and AAII, generating DNA adduct patterns reproducing those found in several species including human exposed to these compounds. These results demonstrate that NQO1 is capable of reducing both AAs to reactive species binding to DNA. However, concentrations required for half-maximum DNA binding mediated by NQO1 were higher for AAII (158 µM) than for AAI (17 µM). One of the reasons causing this phenomenon is a lower efficiency of NQO1 to reduce AAII than AAI we found in this work; although both AAI and AAII are bound with similar binding affinities to the NQO1 active site, the binding orientation of AAII in the active site of NQO1 does not favor the effective reduction of its nitro group. Because reduced nitro-aromatics are often further activated by SULTs or NATs, their roles in AAI and AAII activation were investigated. Our results indicate that phase II reactions do not stimulate the bioactivation of AAs; neither enzymes present in human hepatic cytosols nor human SULT1A1, 1A2, 1A3, 1E, or 2A nor NAT1 or NAT2 further enhanced DNA adduct formation by AAs. In contrast, human SULT1A1, 1A2 and 1A3 as well as NAT1 and NAT2 enzymes even inhibited NQO1-mediated bioactivation of AAII. Therefore, under the in vitro conditions used, DNA adducts arise by enzymatic reduction of AAs through the formation of N-hydroxyaristolactams that are spontaneously decomposed to the reactive species forming DNA adducts. CONCLUSION: The results found in this study emphasize the importance of NQO1 in the metabolic activation of AAI and AAII and provide the evidence that initial nitroreduction is the rate limiting step in their activation. This enzyme is more effective in activation of AAI relative to AAII, which might contribute to its lower binding to DNA found both in vitro and in vivo, Moreover, inhibition effects of conjugation reactions on AAII activation might further contribute to its decreased capability of forming DNA adducts and its lower toxicity comparing with AAI.

Department of Biochemistry Faculty of Science Charles University Prague Czech Republic