Retigabine (D-23129), an N-2-amino-4-(4-fluorobenzylamino)phenylcarbamine acid ethyl ester, is a novel antiepileptic drug which is currently in phase II clinical development. This drug undergoes N-glucuronidation. We aimed to identify the principal enzymes involved in the N-glucuronidation pathway of retigabine and compared our findings with those obtained from human liver (a pool of 30 donors) and kidney microsomes (a pool of 3 donors) and with results from a human absorption, distribution, metabolism, and excretion study upon administration of 200 microCi of [(14)C]-D-23129. Essentially, microsomal assays with UGT1A1 produced only one of the 2 N-glucuronides, whereas UGT1A9 is capable of forming both N-glucuronides. The rates of metabolism for UGT1A9, human liver microsomes, and UGT1A1 were 200, 100, and 100 pmol N-glucuronide per minute per milligram of protein, respectively. At the 50 micromol/L uridine diphosphate glucoronic acid (UDPGA) concentration, UGT1A4 also catalyzed the N-glucuronidation of retigabine, the rates being approximately 5 and 6 pmol/(min.mg protein). With UGT1A9, the production of metabolites 1 and 2 proceeded at a K(m) of 38+/-25 and 45+/-15 micromol/L, whereas the K(m) for retigabine N-glucuronidation by human liver microsomal fractions was 145+/-39 micromol/L. Furthermore, a V(max) of 1.2+/-0.3 (nmol/[min.mg protein]) was estimated for human liver microsomes (4 individual donors). We investigated the potential for drug-drug interaction using the antiepileptic drugs valproic acid, lamotrigine, the tricyclic antidepressant imipramine, and the anesthetic propofol. These are commonly used medications and are extensively glucuronidated. No potential for drug-drug interactions was found at clinically relevant concentrations (when assayed with human liver microsomes or UGT1A9 enzyme preparations). Notably, the biosynthesis of retigabine-N-glucuronides was not inhibited in human liver microsomal assays in the presence of 330 micromol/L bilirubin, and glucuronidation of retigabine was also observed with microsomal preparations from human kidney and Crigler-Najjar type II liver. This suggests that lack of a particular UDP-glucuronosyltransferase (UGT) isoform (eg, UGT1A1 in kidney) or functional loss of an entire UGT1A gene does not completely abolish disposal of the drug. Finally, chromatographic separations of extracts from microsomal assays and human urine of volunteers receiving a single dose of (14)C-retigabine provided clear evidence for the presence of the 2 N-glucuronides known to be produced by UGT1A9. We therefore suggest N-glucuronidation of retigabine to be of importance in the metabolic clearance of this drug.