Sulforaphane is not an effective antagonist of the human pregnane X-receptor in vivo

详细信息    查看全文

• 作者：Emma Jane Poulton^a ; ^b ; Lisa Levy^c ; Johanna W. Lampe^a ; ^c ; Danny D. Shen^a ; ^e ; Julia Tracy^a ; ^b ; Margaret C. Shuhart^d ; Kenneth E. Thummel^a ; ^e ; David L. Eaton^a ; ^b ; ^e ; ^{deaton@uw.edu}
• 关键词：Sulforaphane ; Cytochrome P450 ; CYP3A4 ; Midazolam ; Pregnane X-receptor ; NR1I2 ; Phase I clinical trial ; Humanized PXR mice ; Isolated hepatocytes
• 刊名：Toxicology and Applied Pharmacology
• 出版年：2013
• 出版时间：1 January, 2013
• 年：2013
• 卷：266
• 期：1
• 页码：122-131
• 全文大小：602 K

文摘

Sulforaphane (SFN), is an effective in vitro antagonist of ligand activation of the human pregnane and xenobiotic receptor (PXR). PXR mediated CYP3A4 up-regulation is implicated in adverse drug-drug interactions making identification of small molecule antagonists a desirable therapeutic goal. SFN is not an antagonist to mouse or rat PXR in vitro; thus, normal rodent species are not suitable as in vivo models for human response. To evaluate whether SFN can effectively antagonize ligand activation of human PXR in vivo, a three-armed, randomized, crossover trial was conducted with 24 healthy adults. The potent PXR ligand ¡ª rifampicin (300 mg/d) was given alone for 7 days in arm 1, or in daily combination with 450 ¦Ìmol SFN (Broccoli Sprout extract) in arm 2; SFN was given alone in arm 3. Midazolam as an in vivo phenotype marker of CYP3A was administered before and after each treatment arm. Rifampicin alone decreased midazolam AUC by 70 % , indicative of the expected increase in CYP3A4 activity. Co-treatment with SFN did not reduce CYP3A4 induction. Treatment with SFN alone also did not affect CYP3A4 activity in the cohort as a whole, although in the subset with the highest basal CYP3A4 activity there was a statistically significant increase in midazolam AUC (i.e., decrease in CYP3A4 activity). A parallel study in humanized PXR mice yielded similar results. The parallel effects of SFN between humanized PXR mice and human subjects demonstrate the predictive value of humanized mouse models in situations where species differences in ligand-receptor interactions preclude the use of a native mouse model for studying human ligand-receptor pharmacology.