Inhibition of human CYP1A2 oxidation of 5,6-dimethyl-xanthenone-4-acetic acid by acridines: a molecular modelling study.

1. The aim of the present study was to investigate the structural requirements for the inhibition of 6-methyl-hydroxylation of the antitumour agent 5,6-dimethyl-xanthenone-4-acetic acid (DMXAA) by acridine analogues and use a CYP1A2 homology model to provide some insight into this interaction. 2. Concentrations causing 50% inhibition (IC50) of the 6-methylhydroxylation of DMXAA were determined in human liver microsomes in the presence of various acridines. Some of the acridines were also tested for their ability to inhibit the CYP1A2-mediated 7-ethoxyresorufin O-de-ethylation. The molecular modelling studies of human CYP1A2 used the crystal structure of rabbit CYP2C5 as a template based on protein sequence homology and an interactive docking procedure using a dynamic hydrogen bond feature. 3. The in vitro IC50 studies for the inhibition of 6-methylhydroxylation of DMXAA indicated: (i) the importance of the position of the carboxamide side-chain on the acridine nucleus (and, to a lesser extent, its composition); (ii) the addition of hydroxyl groups to the 5-, 6- and 7-position of the acridine nucleus diminished the inhibitory potency; and (iii) amsacrine (acridine nucleus with methansulphonanilide side-chain at the 9-position) had no significant inhibitory effect. Similar structural trends were observed for the inhibition of O-de-ethylation of 7-ethoxyresorufin by acridines, supporting the involvement of CYP1A2 in DMXAA 6-methyl hydroxylation. 4. The molecular modelling studies indicated: (i) both DMXAA and N-[2-(dimethylamino)-ethyl]acridine-4-carboxamide (DACA) form two hydrogen bonds plus putative pi-pi stacking interactions with the CYP1A2-binding domain, typical of CYP1A2 substrates and inhibitors; (ii) the DMXAA 6-methyl group is 4.0 A from the central iron atom of the heme moiety and ideal for oxidation; (iii) the known oxidation sites for DACA are orientated away from the heme iron, supporting the non-involvement of CYP1A2; and (iv) amsacrine did not fit the putative CYP1A2 site owing to the steric hindrance of the bulky methanesulphonanilide side-chain. 5. These results suggest that docking studies with this homology model may be useful in the design of further acridine anticancer agents, in particular to identify agents that do not interact either as substrates or inhibitors with the CYP1A2-binding domain.