In vitro evidence of baicalein’s inhibition of the metabolism of zidovudine (AZT)

Background: Herb-drug interaction (HDI) has been regarded as a key factor limiting the clinical application of herbs and drugs. Aims: Potential baicalein-zidovudine (AZT) interaction was predicted in the present study. Methods: In vitro evaluation of baicalein’s inhibition towards human liver microsomes (HLMs)-catalyzed metabolism of zidovudine (AZT) was performed. Dixon and Lineweaver-Burk plots were used to determine the inhibition kinetic type, and second plot with the slopes from Lineweaver-Burk plot versus the concentrations of baicalein was employed to calculate the inhibition parameter (K i ). In combination with the in vivo concentration of baicalein, in vitro - in vivo extrapolation (IVIVE) was carried out to predict in vivo baicalein-AZT interaction. Results: Competitive inhibition of baicalein towards AZT metabolism was demonstrated, and the K i value was calculated to be 101.2 μM. The value of AUC i /AUC was calculated to be 2. Conclusion: Potential baicalein-AZT interaction was indicated in the present study, indicating the need for monitoring when AZT is co-administrated with baicalein or baicalein-containing herbs.

weeks of ginseng administration, which was closely correlated with the reduced peak plasma level of warfarin 5 . Zidovudine (AZT) is a nucleoside analog reversetranscriptase inhibitor (NRTI) to treat HIV/AIDS , and has been approved and prescribed under the name Retrovir 6 . AZT has been demonstrated to undergo UDP-glucuronosyltransferase (UGT) 2B7-catalyzed glucuronidation elimination after its absorption into the serum 7 . To date, many herbal components have been demonstrated to exhibit the inhibition towards the metabolism of AZT, such as demethylzeylasteral and 20(S)-protopanaxatriol (ppt) 7, 8 . Baicalein is the hydrolyzed product and main effective substance basis of baicalin which is a flavonoid isolated from Scutellariae radix, and has been demonstrated to exhibit various biochemical and pharmacological activities, including anti-inflammatory, anti-tumor, anti-allergic, and anti-oxidation effects 9,10 . As a herbal component exhibiting many pharmacological activities, the co-exposure possibility between baicalein and AZT is very high. Therefore, the aim of the present study was to evaluate the inhibition potential of baicalein towards the metabolism of AZT. The inhibition kinetic type and parameter (K i ) were determined, and in vitro-in vivo extrapolation (IVIVE) was carried out. Corresponding author: Ling-Ting Kong, Yantaishan hospitalYantai,ShandongChina E-mail:konglingting123456@163.com ethicin and uridine 5'-diphosphoglucuronic acid (UD-PGA) (trisodium salt) were purchased from Sigma-Aldrich (St Louis, MO). Baicalein was purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Pooled human liver microsomes (HLMs) were purchased from BD Gentest (Woburn, MA). All other reagents were of HPLC grade or of the highest grade commercially available.

Inhibition evaluation of HLMs-mediated AZT glucuronidation by baicalein
Baicalein was dissolved in methanol at 20 mmol/L as stock solution. The incubation mixture (200 μL total volume) contained HLMs (final concentration=0.5 mg/mL), 5 mM UDPGA, 5 mM MgCl 2 , 50 mM Tris-HCl buffer (pH=7.4), 50 μg/mg protein alamethicin, and AZT (the concentration is corresponding to the apparent Km value). The incubation time was 30 min. The mixture was kept on ice until it was centrifuged at 20000 ×g for 10 min at 4 °C. Aliquots of supernatants were transferred for HPLC analysis. For the analysis AZT glucuronidation, the HPLC column was eluted at 1 mL/min with a mobile phase of acetonitrile aqueous (v/v = 12:88). The aqueous phase contained 0.4 mL concentrated H 3 PO 4 diluted to 1 L with water (pH 2.4). Ultraviolet detection was at 267 nm. Furthermore, the reaction was determined at multiple concentrations of AZT and baicalein. Dixon plot and Lineweaver-Burk plot were used to determine the inhibition kinetic type, and the second plot was further employed to calculate the inhibition kinetic parameter (K i ).

Prediction of in vivo baicalein-drug interaction magnitude
The baicalein-AZT interaction magnitude is affected by both in vitro inhibition kinetic parameters (K i ) and in vivo concentration of baicalein. The most common equation for drug-drug interaction prediction was employed to predict the AUC alteration of AZT caused by co-administration of baicalein. AUC i /AUC=1+[I] in vivo /K i The terms are defined as follows: AUC i /AUC is the predicted ratio of in vivo exposure of AZT with coadministration of baicalein vs. that in control situation. K i is the in vitro reversible inhibition constant, [I]in vivo is the in vivo concentration of baicalein. In this study, the maximum plasma concentration (C max ) of baicalein was used.

Fig. 2 Determination of inhibition kinetic type for baicalein's inhibition towards the metabolism of zidovudine (AZT). (A) Dixon plot for baicalein's inhibition towards the metabolism of zidovudine (AZT); (B) Lineweaver-
Burk plot for baicalein's inhibition towards the metabolism of zidovudine (AZT). The data point represents the mean of duplicate experiment.
The second plot with the slopes from Lineweaver-Burk plot versus baicalein's concentrations was performed to determine the inhibition parameter (Ki) to be 101.2 μM. The maximum concentration of baicalein in plasma was reported to be approximately 100 μM after i.v. administration of 2.2 mg/kg or 5.5 mg/kg of baicalein for rats 11 . Using this value in combination the K i value, the value of AUC i /AUC was calculated to be 2.

Discussion
Drug-metabolizing enzymes (DMEs)-based clinical drug-drug interaction (DDI) and herb-drug interaction (HDI) limited the utilization of many clinical drugs and R&D process of potential drugs. For example, the strong inhibition of noscapine towards cytochrome P450 2C9 and 3A4 strongly limited the utilization of noscapine as an efficient antitussive drug and the development of this compound as a potential anti-tumor drug 12 . The inhibition of herbal components liquiritigenin and isoliquiritigenin towards the metabolism of hypnotic agent propofol limited the co-utilization of these two herbal components and propofol 13 . The dose beyond the efficient concentration of AZT can induce the genotoxicity and minimum cell death 14 . AZT has a narrow therapeutic index, and slight alteration of AZT concentration will exceed the minimal toxicity concentration of AZT. The present study reported the AUC i /AUC value was 2 through in vitro-in vivo extrapolation (IVIVE) using the in vitro K i value and in vivo concentration of baicalein. This result can be affected by the following factors: 1) In vitro incubation factors: previous literature has reported that the addition of bovine serum albumin (BSA) can significantly cause the reduction of K i value for fluconazole inhibition of AZT glucuronidation by HLM 15 . Therefore, the K i value calculated in the present study might underestimate the baicalein-AZT interaction due to the lack of BSA in our present incubation system. 2) In vivo factors: The present study using the in vivo concentration after i.v. adminstration of single purified compound. When the administration route is different (e.g., oral gavage (p.o.), intraperitoneal injection (i.p.), etc.) or the herbs containing baicalein is given, the in vivo concentration of baicalein might be different which can further complicate the IVIVE results. In conclusion, the baicalein-AZT interaction was indicated in the present study using in vitro inhibition experiment, which challenges the co-utilization of AZT and baicalein or baicalein-containing herbs.