Relationship Between Flavonoid Structure And Phase-II Metabolism

Objective: The overall objective is to develop structure-metabolism relationships (SMRs) between UGTs and flavonoids for predicting glucuronidation of flavonoids. The goals of this research project were to: 1) identify the major UGT isoform(s) contributing to the glucuronidation of flavonoids and predicting the major organ of metabolism; 2) establish the substrate-selectivity and regiospecificity of these major UGT isoform(s); 3) develop the in silico prediction models for UGT 1A8 and UGT1A9 using pharmacophore and 2-D/3-D QSAR modeling techniques; 4) study the effect of change in backbone on the disposition of flavonoids in Caco-2 cells; 5) study the rate-limiting role of efflux transporters in the disposition of flavonols in Caco-2 cells; and 6) study the regiospecific disposition of flavones in Caco-2 cells. Method: For objectives 1 and 2, in vitro recombinant human UGT isoforms glucuronidation model was used. For objective 3, in silico pharmacophore and 2-D/3-D QSAR modeling was used along with in vitro glucuronidation intrinsic clearance values in recombinant human UGT isoforms. For objectives 4, 5 and 6, intact Caco-2 cell monolayers was used as the transport model, and Caco-2 cell lysate was used for measuring glucuronide formation rates. Results: 1) To identify the major UGT isoform(s) contributing to the glucuronidation of flavonoids and predicting the major organ of metabolism, we found that flavonoids were mainly glucuronidated by UGT1A1, 1A8 and 1A9 at the substrate concentration of 2.5, 10 and 35μM. 2) To establish the substrate-selectivity and regiospecificity of these major UGT isoform, we found that UGT1A1 showed no regiospecificity for glucuronidating any position, whereas, UGT1A8 and UGT1A9 showed either dominant, preferred or weak regiospecificity for 3-O or 7-O position, depending on the structure of the compound. In general, the addition of hydroxyl group at C-4' reduced, whereas the addition of hydroxyl group at C-5 and/or C-7 improved the rates of glucuronidation of flavonoids by UGT1A8 and 1A9. On the other hand, the rates of glucuronidation by UGT1A1 reduced as number of hydroxyl group in the structure increased. 3) To develop the in silico prediction models for UGT1A8 and UGT1A9 using pharmacophore and 2-D/3-D QSAR modeling techniques, we found that pharmacophore-based semi-quantitative SMR models for UGT1A9 with >75% predictive ability could be developed. But neither semi-quantitative SMR models for UGT1A8 nor the quantitative SMR models for UGT1A8 and UGT1A9 could be successfully developed. 4) To study the effect of change in backbone on the disposition of flavonoids in Caco-2 cells, we found that the change in backbone impacts the excretion of flavonoid sulfates more significantly than the excretion of their glucuronides except for genistein. 5) To study the rate-limiting role of efflux transporters in the disposition of flavonols in Caco-2 cells, we found that excretion of flavonol glucuronides in Caco-2 cells were not limited by efflux transporters and glucuronides of flavonols showed basolateral preference in their excretion. 6) To study the regiospecific conjugation of flavones in Caco-2 cells, we found that both glucuronidation and sulfation of flavones mainly happened at hydroxyl group at C-7 position. Conclusion: UGT1A9, UGT1A8 and UGT1A1 are the most important isoforms that can glucuronidate vast majority of tested flavonoids. Based on published UGT isoform expression pattern in human liver and intestine, they should serve as the major first-pass metabolism organs for flavonoids. UGT1A8 and UGT1A9 showed regiospecificity for 3-O or 7-O position, depending on the structure of the compound, whereas UGT1A1 showed no regiospecificity. Also, the addition of hydroxyl group at C-4' reduced, whereas the addition of hydroxyl group at C-5 and/or C-7 improved the rates of glucuronidation of flavonoids by UGT1A8 and 1A9, with rare exceptions. In contrast, the rates of glucuronidation by UGT1A1 reduced as number of hydroxyl group in the structure increased. Isoform-specific semi-quantitative Pharmacophore-based 3-D SMR prediction models could be developed for UGT1A9 with the predictive ability of more than 75%, but more efforts are needed to develop better quantitative models of prediction. We also probed the SMR experimentally using the Caco-2 model, and the results showed that the excretion of glucuronides was impacted more by the change in number and position of hydroxyl group in the flavonoid structure than changes in backbone. The excretion of glucuronides of flavones but not flavonols is rate-limited by efflux transporters. Future SMR research will incorporate more experimentally derived information to develop better models to predict glucuronidation of flavonoids in humans.Pharmacological and Pharmaceutical Sciences, Department o