Cytochrome P450-mediated herb-drug interaction (HDI) of Polygonum multiflorum Thunb. based on pharmacokinetic studies and in vitro inhibition assays

Abstract

BackgroundPolygonum multiflorum Thunb. (PM) is well known both in China and other countries of the world for its tonic properties, however, it has lost its former glory due to liver toxicity incidents in recent years. PurposeThe purpose of this study is to determine whether the occurrence of herb-drug interaction (HDI) caused by PM is associated with cytochrome P450 (CYP450) based on pharmacokinetic studies and in vitro inhibition assays. The objective was to provide a reference for the rational and safe use of drugs in clinical practice. MethodsIn this study, raw PM (R), together with its two processed products which included PM by Chinese Pharmacopoeia (M) and PM by “nine cycles of steaming and sunning (NCSS)” (“9”), were prepared as the main research objects. A method based on fluorescence technology was used to evaluate the inhibition levels of raw and processed PMs, as well as corresponding characteristic compounds on seven recombinant human cytochrome P450s (rhCYP450s). The pharmacokinetics of sulindac (a representative of commonly used nonsteroidal anti-inflammatory drugs) and psoralen (a major compound of Psoralea in combination with PM) in rat plasma were studied when combined with raw and different processed products of PM. ResultsThe inhibitory level order of the three extracts on major different subtypes of CYP450 (CYP1A2, CYP2B6, CYP2C8, CYP2C19, CYP2D6, and CYP3A4) was: R > M > “9”. However, the inhibition level of R and “9” is higher than that of M on CYP2C9. Further studies showed that trans-THSG and emodin could selectively inhibit CYP3A4 and CYP1A2, respectively. Epicatechin gallate mainly inhibited CYP3A4 and CYP1A2, followed by CYP2C8 and CYP2C9. Genistein mainly inhibited CYP3A4, followed by CYP2C9 and CYP2C8. CYP3A4 and CYP2C9 were also inhibited by daidzein. The inhibitory effects of all the PM extracts were associated with their characteristic compounds. The results of HDI showed that R increased sulindac exposure to rat blood, and R and M increased psoralen exposure to rat blood, which were consistent with corresponding metabolic enzymes. Overall, the in vitro and in vivo results indicated that PM, especially R, would be at high risk to cause toxicity and drug interactions via CYP450 inhibition. ConclusionThis study not only elucidates the scientific connotation of “efficiency enhancement and toxicity reduction” of PM by NCSS from the perspective of metabolic inhibition but also contributes to HDI prediction and appropriate clinical medication of PM.