Abstract

The regio- and stereoselectivity of warfarin metabolism have been used to assess structure-function relationships of human P4502C subfamily members. Metabolism was investigated using a yeast cDNA expression system in which full length cDNAs for P4502C8, -2C9 (alleles Arg144 Tyr358 Ile359 Gly417 and Arg144 Tyr358 Leu359 Gly417), -2C18 (alleles Thr385 and Met385), and -2C19 were expressed. Additionally, two mutations reported in other P4502C9/2C10 alleles were individually introduced into P4502C9 by site-directed mutagenesis, to yield Cys144 Tyr358 Ile359 Gly417, Arg144 Tyr358 Ile359 Asp417, and Arg144 Cys358 Ile359 Gly417, which were expressed in yeast; their ability to metabolize warfarin was then studied. Warfarin metabolism by purified preparations of P4502C9 allele Arg144 Tyr358 Ile359 Gly417 and its Leu359 mutant was also investigated in reconstituted systems. Both alleles of P4502C18 were regioselective for 4'-hydroxywarfarin, without any significant stereoselectivity. Both also metabolized warfarin at the 6-position, but to a lesser extent, and metabolism at this site was stereoselective for (R)-warfarin. P4502C8 metabolized warfarin at the 7-position and was stereospecific for (R)-warfarin. It also metabolized warfarin to a lesser extent at the 4'-position, and metabolism at this site was stereoselective for (R)-warfarin. P4502C19 was regioselective for 6- and 8-hydroxywarfarin and was stereoselective for (R)-warfarin. The highly conservative mutation of Ile359 to Leu359 in P4502C9 profoundly altered the regio- and stereoselectivity of warfarin metabolism, from regioselective for 7-hydroxywarfarin, with stereospecificity for (S)-warfarin, to regioselective for 4'-hydroxywarfarin, with stereoselectivity for (R)-warfarin, which was confirmed in a reconstituted system using purified recombinant enzymes. In contrast, individual mutations of P4502C9 of Arg144 to Cys, Tyr358 to Cys, and Gly417 to Asp did not markedly affect the regio- or stereoselectivity of warfarin metabolism, although the overall rates of warfarin metabolism were apparently increased by these changes. We conclude that residue 359 is at the substrate binding site of P4502C9, whereas residues 144, 358, and 417, and residue 385 of P4502C18, are not.