In vitro studies were conducted to identify the hepatic cytochrome P450 (CYP) isoenzyme involved in the 6-methylhydroxylation of 5, 6-dimethylxanthenone-4-acetic acid (DMXAA) by using a human liver library (n = 14). The metabolite 6-hydroxymethyl-5-methylxanthenone-4-acetic acid (6-OH-MXAA) was determined by HPLC with fluorescence detection. The metabolite formed in human liver microsomes and by cDNA-expressed CYP isoform was identified by liquid chromatography mass spectrometry as 6-OH-MXAA. In human liver microsomes (n = 14), 6-methylhydroxylation of DMXAA followed monophasic Michaelis-Menten kinetics, with a mean apparent K(m) of 21 +/- 5 microM and V(max) of 0.043 +/- 0.019 nmol/min/mg. An approximate 10-fold interindividual variation in the intrinsic clearance (V(max)/K(m)) of DMXAA 6-methylhydroxylation in human liver microsomes was observed. The involvement of CYP1A2 in DMXAA metabolism by human livers was demonstrated by the following: 1) the potent inhibition of DMXAA metabolism by furafylline (k(inact) = 0.23 +/- 0.04 min(-1), K'(app) = 15.6 +/- 6.7 microM) and alpha-naphthoflavone (K(i) = 0.036 microM), but not by cimetidine, ketoconazole, tolbutamide, quinidine, chlorzoxazone, diethyldithiocarbamate, troleandomycin, and sulfaphenazole; 2) when incubated with human lymphoblastoid cell microsomes containing cDNA-expressed CYP isoenzymes, DMXAA was metabolized only by CYP1A2, with an apparent K(m) of 6.2 +/- 1.5 microM and V(max) of 0.014 +/- 0.001 nmol/min/mg, but not by CYP2A6, CYP2B6, CYP2C9 (Arg(144)), CYP2C19, CYP2D6 (Val(374)), CYP2E1, and CYP3A4; 3) a significant correlation (r = 0.90; P <.001) between 6-methylhydroxylation of DMXAA and 7-ethoxyresorufin O-deethylation; and 4) a significant correlation (r = 0.75; P <.01) between the CYP1A protein level determined by Western blots and DMXAA 6-methylhydroxylation.