Studies were performed to elucidate the mechanism responsible for the reduction in K m values of UDP-glucuronosyltransferase 2B7 (UGT2B7) substrates observed for incubations conducted in the presence of albumin. Addition of bovine serum albumin (BSA) and fatty acid-free human serum albumin (HSA-FAF), but not "crude" HSA, resulted in an approximate 90% reduction in the K m values for the glucuronidation of zidovudine (AZT) by human liver microsomes (HLM) and UGT2B7 and a 50 to 75% reduction in the S 50 for 4-methylumbelliferone (4MU) glucuronidation by UGT2B7, without affecting V max . Oleic, linoleic, and arachidonic acids were shown to be the most abundant unsaturated long-chain fatty acids present in crude HSA and in the membranes of HLM and human embryonic kidney (HEK)293 cells, and it was demonstrated that these and other unsaturated long-chain fatty acids were UGT2B7 substrates.Glucuronides with R f (retention factor) values corresponding to the glucuronides of linoleic and arachidonic acid were detected when HLM and HEK293 cell lysates were incubated with radiolabeled cofactor, and the intensity of the bands was modulated by the presence of crude HSA (increased) and BSA or HSA-FAF (decreased). Oleic, linoleic, and arachidonic acid inhibited AZT and 4MU glucuronidation by HLM and/or UGT2B7, due to an increase in K m /S 50 without a change in V max . Addition of BSA and HSA-FAF reversed the inhibition. Likewise, coexpression of UGT2B7 and HSA in HEK293 cells reduced the K m /S 50 values of these substrates. It is postulated that BSA and HSA-FAF sequester inhibitory fatty acids released during incubations, and the apparent high K m values observed for UGT2B7 substrates arise from the presence of these endogenous inhibitors.

Major advances in the characterization of uridine diphosphate (UDP)-glucuronosyltransferase (UGT) enzyme substrate and inhibitor selectivities and the development of experimental paradigms to investigate xenobiotic glucuronidation in vitro now permit the prediction of a range of drug-glucuronidation parameters in humans. In particular, the availability of substrate and inhibitor "probes" for the major hepatic drug metabolizing UGTs together with batteries of recombinant enzymes allow the reaction phenotyping of drug glucuronidation reactions. Additionally, in vitro experimental approaches and scaling strategies have been successfully applied to the quantitative prediction of in vivo clearance via glucuronidation and drug-drug interaction potential.

ABSTRACT:Bovine serum albumin (BSA) and fatty acid-free human serum albumin (HSAFAF) reduce the K m values for UGT2B7 substrates by sequestering inhibitory long-chain fatty acids released by incubations of human liver microsomes (HLM) and HEK293 cells expressing this enzyme. However, the scope of the "albumin effect " is unknown. In this investigation we characterized the effects of albumin on the kinetics of 4-methylumbelliferone (4MU) glucuronidation by UDP-glucuronosyltransferase (UGT) 1A1, 1A6, and 1A9, and propofol (PRO) glucuronidation by UGT1A9 and HLM. BSA and HSAFAF, but not human serum albumin, reduced the K m values for 4MU and PRO glucuronidation by UGT1A9. For example, HSAFAF (2%) reduced the K m values for 4MU and PRO glucuronidation from 13.4 to 2.9 and 41 to 7.2 M, respectively. Similarly, HSAFAF (2%) reduced the K m for PRO glucuronidation by HLM from 127 to 10.6 M. Arachidonic, linoleic, and oleic acids and a mixture of these decreased the rates of 4MU and PRO glucuronidation by UGT1A9. K m values for these reactions were increased 3-to 6-fold by the fatty acid mixture. Inhibition was reversed by the addition of BSA (2%). Extrapolation of kinetic constants for PRO glucuronidation by HLM in the presence of HSAFAF predicted in vivo hepatic clearance within 15%. Fatty acids had no effect on 4MU glucuronidation by UGT1A1 and UGT1A6 but, paradoxically, all forms of albumin altered the kinetic model for 4MU glucuronidation by UGT1A6 (from Michaelis-Menten to two-site). Only BSA caused a similar effect on 4MU glucuronidation by UGT1A1. It is concluded that BSA and HSAFAF reduce the K m values of only those enzymes inhibited by long-chain unsaturated fatty acids.

Although knowledge of human renal cytochrome P450 (CYP) and UDP‐glucuronosyltransferase (UGT) enzymes and their role in xenobiotic and endobiotic metabolism is limited compared with hepatic drug and chemical metabolism, accumulating evidence indicates that human kidney has significant metabolic capacity. Of the drug metabolizing P450s in families 1 to 3, there is definitive evidence for only CYP 2B6 and 3A5 expression in human kidney. CYP 1A1, 1A2, 1B1, 2A6, 2C19, 2D6 and 2E1 are not expressed in human kidney, while data for CYP 2C8, 2C9 and 3A4 expression are equivocal. It is further known that several P450 enzymes involved in the metabolism of arachidonic acid and eicosanoids are expressed in human kidney, CYP 4A11, 4F2, 4F8, 4F11 and 4F12. With the current limited evidence of drug substrates for human renal P450s drug–endobiotic interactions arising from inhibition of renal P450s, particularly effects on arachidonic acid metabolism, appear unlikely. With respect to the UGTs, 1A5, 1A6, 1A7, 1A9, 2B4, 2B7 and 2B17 are expressed in human kidney, whereas UGT 1A1, 1A3, 1A4, 1A8, 1A10, 2B10, 2B11 and 2B15 are not. The most abundantly expressed renal UGTs are 1A9 and 2B7, which play a significant role in the glucuronidation of drugs, arachidonic acid, prostaglandins, leukotrienes and P450 derived arachidonic acid metabolites. Modulation by drug substrates (e.g. NSAIDs) of the intrarenal activity of UGT1A9 and UGT2B7 has the potential to perturb the metabolism of renal mediators including aldosterone, prostaglandins and 20‐hydroxyeicosatetraenoic acid, thus disrupting renal homeostasis.

No abstract