Repaglinide metabolism by human hepatocytes reveals CYP2C8 dominance and misassignment of CYP3A4 by liver subcellular fractions and recombinant enzymes.

Abstract

Repaglinide is an antidiabetic agent and Food and Drug Administration-endorsed clinical CYP2C8 probe with a well characterized human disposition profile in which a dicarboxylic acid, M2, is the main metabolite accounting for 66% of the administered dose. The remainder is distributed among M1 (4%), unchanged repaglinide (2%), M4 (<0.5%), several minor metabolites (each <1%), and some unidentified metabolites. Although substantial in vivo evidence supports repaglinide as a CYP2C8 probe, in vitro studies indicate a greater CYP3A4 contribution to its overall metabolism. Previous research using chemical inhibitors in human liver subcellular fractions and recombinant enzymes consistently assigned M1 and M4 formation to CYP3A4 and CYP2C8, respectively, with M2 predominantly formed via a CYP3A4-mediated first-step and subsequent completion by a cytosolic enzyme. However, corresponding experiments using chemical inhibitors were not conducted in human hepatocytes (HH). Motivated by our recent data indicating potential misassignment of CYP3A4 involvement by human liver microsomes and recombinant enzymes, this study compared these matrices with HH. Here, human liver microsomes, liver S9, and recombinant enzymes also assigned M1 and M4 formation to CYP3A4 and CYP2C8, respectively. In contrast, slightly over half of M1 formed in HH was attributed to CYP2C8, a surprising result given the enzyme's lower hepatic expression and minimal contribution in subcellular fractions. M4 formation remained predominantly CYP2C8-dependent in HH. Crucially, and contrary to results in subcellular fractions, HH showed minimal CYP3A4 involvement in the formation of the major in vivo metabolite M2, firmly establishing CYP2C8 as the primary enzyme. These findings provide robust data for rationalizing repaglinide pharmacokinetics and demonstrate a real risk of CYP3A4 misassignment when relying solely on liver subcellular fractions or recombinant enzymes. SIGNIFICANCE STATEMENT: The author and colleagues recently discovered that human liver microsomes and recombinant enzymes can misidentify CYP3A4 as the main savolitinib-metabolizing enzyme. It is, therefore, crucial for researchers and regulatory authorities to understand the risk of misassignment. This study shows that the same phenomenon underlies the inconsistent in vitro-in vivo relative contributions of CYP3A4 and CYP2C8 to repaglinide metabolism that have been difficult to explain and demonstrates that liver subcellular fractions and recombinant enzymes pose a real risk for enzyme misidentification.