Involvement of cytochromes P450 (CYP) and other haem associated enzymes in the bioreduction of AQ4N, an antitumour prodrug
The anthraquinone di-N-oxide AQ4N is a prodrug designed to be excluded from cell nuclei until metabolised in hypoxic tumour regions to AQ4, a DNA binder and potent inhibitor of topoisomerase II. The antitumour effects of AQ4N in rodent neoplasms are well characterised but the identity of enzymes responsible for the metabolism are unknown. The aims of the present work were to identify Cytochrome P450 (CYP) enzymes responsible for AQ4N metabolism in rat and human tissue and to conduct a preliminary investigation into the in vivo metabolism of AQ4N in tumour bearing rodents. AQ4N was found to undergo a two electron reduction to the mono-Noxide AQM followed by a subsequent two electron reduction to cytotoxic AQ4. The process occurred in the microsomes of rat and human liver, was cofactor dependent and was inhibited by air. In rats, CYPs 2B and 2E were found to anaerobically metabolise both AQ4N and AQM. Kinetically, AQ4N metabolism conformed to a Michaelis-Menten model whereas the metabolism of AQM was better described by a sigmoidal relationship. In addition, both semi purified Cytochrome P450 reductase (CPR) and purified Nitric oxide synthase (NOS) were both able to anaerobically metabolise AQ4N. Both enzymes required NADPH and CPR mediated metabolism was dependent on the presence of exogenous haem. In humans, the anaerobic metabolism of both AQ4N and AQM correlated with CYP 3A activity and not with the activities of CYP 1 AI 2C and 2D. AQM metabolism correlated also with the activity of CYP 2A. The involvement of CYP 3A was confirmed by the use of CYP specific inhibitors and by the use of cDNA transfected cell microsomes. Human kidney and colonic tumours were found to anaerobically metabolise AQ4N and tumour metabolism was inhibited by the CYP inhibitor carbon monoxide (CO). Finally, the in vivo metabolism of AQ4N was studied in C3H tumour bearing mice. Metabolites of AQ4N were found in all tissues studied but the AQ4! AQ4N ratio was highest in the tumours. Collectively, these findings have identified the enzymes responsible for the metabolism of AQ4N and its mono-N-oxide. Differences exist between the CYP isoforms responsible for metabolism in rodents and in man, in humans, CYP 3A enzymes predominantly metabolise AQ4N and this subfamily of CYP are known to be well expressed in a broad spectrum of human cancers. With this in mind, AQ4N based therapy should be considered as a rational treatment regime for patients bearing solid tumour burdens.