Journal of Chromatography B, 973 (2014) 120–125
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Determination of liquiritigenin by ultra high performance liquid chromatography coupled with triple quadrupole mass spectrometry: Application to a linear pharmacokinetic study of liquiritigenin in rat plasma Jie Gu a,b , Huan Li a,b,c,d , Ke Pei a,b , Hui Cai d , Kunming Qin a,b , Xinghai Zhang a,b , Lijuan Zheng a,b , Xiao Liu a,b , Yunqing Cai e , Baochang Cai a,b,∗ a
Pharmacy College of Nanjing University of TCM, Nanjing 210023, China Nanjing University of Chinese Medicine, Jiangsu Key Laboratory of Chinese Medicine Processing, Engineering Center of State Ministry of Education for Standardization of Chinese Medicine Processing, Nanjing 210023, China c Institute of Bioengineering and Nanotechnology, A* STAR, The Nanos, Singapore 138669, Singapore d School of Applied Science, Temasek Polytechnic, Singapore 529757, Singapore e Department of Nutrition and Food Hygiene, School of Public Health, Nanjing Medical University, Nanjing 210029, China b
a r t i c l e
i n f o
Article history: Received 10 February 2014 Accepted 7 September 2014 Available online 17 September 2014 Keywords: Liquiritigenin UHPLC–MS/MS Intravenous administration Linear pharmacokinetic study
a b s t r a c t A simple, sensitive and rapid ultra high performance liquid chromatography tandem mass spectrometry (UHPLC–MS/MS) method has been developed and validated for the quantification of liquiritigenin, a promising anti-tumor agent. Liquiritigenin and the internal standard were separated on an Agilent Extend C18 column and eluted with a gradient mobile phase system of acetonitrile and water. The analysis was performed on a negative ionization electrospray mass spectrometer via multiple reaction monitoring (MRM). Transitions of m/z 255.0 → 119.0 for liquiritigenin and m/z 269.0 → 117.0 for the IS were monitored. One-step protein precipitation with acetonitrile was used to remove impurities and extract the analytes from plasma. The method had a chromatographic run time of 4.5 min and a good linearity in the range of 1–1000 ng/mL. The precision (R.S.D.) of intra-day and inter-day ranged from 4.54 to 10.65% and 5.94 to 13.81%, respectively; while the accuracy of intra-day and inter-day ranged from 104.06 to 109.28% and 94.98 to 112.05%. The recovery and stability were also within the acceptable limits. The validated method was applied to a linear pharmacokinetic study of liquiritigenin in rat plasma for the first time. © 2014 Published by Elsevier B.V.
1. Introduction Liquiritigenin (7,4 -dyhydroxy flavanone) is a poly phenolic flavanone structure that exists in Radix glycyrrhizae. As an inhibitor of Akt protein kinase and selective estrogen receptor agonist [1], liquiritigenin has been found to have various biological activities, such as anti-inflammatory [2], cytoprotection [3], inhibition of acute hepatic injury [4], and antitumor activity. Liquiritigenin has been shown to possess significant antitumor activity. It has cytotoxic activity against five human cancer cell lines
∗ Corresponding author at: Nanjing University of Chinese Medicine, Jiangsu Key Laboratory of Chinese Medicine Processing, Engineering Center of State Ministry of Education for Standardization of Chinese Medicine Processing, Nanjing 210023, China. Tel.: +86 86798281; fax: +86 25 86798281. E-mail address: [email protected] (B. Cai). http://dx.doi.org/10.1016/j.jchromb.2014.09.002 1570-0232/© 2014 Published by Elsevier B.V.
in vitro [5]. In accordance with our previous studies, liquiritigenin inhibits the growth of SGC-7901 and Lovo cells. Time- and dosedependent inhibition and apoptosis induction were particularly exhibited on SMMC 7721 cells [6]. Liquiritigenin suppresses cancer cell proliferation and angiogenesis, and promotes apoptosis [7,8]. These findings suggested strong cancer preventive effects of liquiritigenin in vitro, yet the underlying molecular mechanisms remain unknown. Our previous study also showed that liquiritigenin significantly inhibited the growth of murine H22 hepatocarcinoma in vivo through a 15-day treatment after tumor inoculation and exhibited potent tumor therapeutic activity [9]. All these results suggested that liquiritigenin could be a promising antitumor agent for human cancers. To ensure safety and effectiveness of the medication, a novel UHPLC–MS/MS method, aiming at both simplicity and sensitivity, was established for the preclinical linear pharmacokinetic study. Several methods involving high-performance liquid
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
chromatography and mass spectrometry have been described previously to analyze liquiritigenin concentration in plasma after administration [10–19]. While these methods provide precise quantitation, the sample preparation and mobile phase composition of such methods are rather complex. A longer time is also required for analysis. In conventional HPLC analysis, liquiritigenin has a long retention time of approximately 9.8 min. By focusing on a single compound rather than multiple constituents in the plasma, shorter analysis time can be achieved by UHPLC–MS/MS. In the present work, we developed and validated a UHPLC–MS/MS method for the determination and quantitation of liquiritigenin in rat plasma that boasts higher specificity, increased sensitivity, simpler sample preparation, and shorter analysis time. The method has been applied to the pharmacokinetic study involving three-level single-dose intravenous administration of liquiritigenin in rats for the first time.
2. Materials and methods 2.1. Materials and reagents Liquiritigenin (purity > 98.0%) was provided by Professor Wei Li from Nanjing University of Chinese Medicine. The internal standard, apigenin was purchased from Shanghai Yuanye Biotechnology Co. Ltd. (Shanghai, China), lot number: 20120220. HPLC–MS-grade acetonitrile and water were purchased from ANPEL Scientific Instrument Co. Ltd. (Shanghai, China). HPLC-grade acetonitrile and methanol were purchased from Merck (Darmstadt, Germany).
2.2. Instruments The chromatography system used was composed of a Shimadzu LC-30A series chromatographic system (Shimadzu Corporation UFLC XR, Kyoto, Japan) with a LC-30AD binary pump, a DGU-20A 5R degasser, a SIL-30AC auto sampler and a CTO-30A column oven. The separation was performed using the C18 column (Agilent, USA), 2.1 × 100 mm, 1.8 m, at a temperature of 35 ◦ C. Mass spectrometric detection was performed by using 5500 triple quad tandem mass spectrometer equipped with electrospray ionization (ESI) source (AB SCIEX, Foster City, CA, USA). Peak Genius2 3030 instrument was used for generating nitrogen gas.
2.3. LC–MS/MS conditions The mobile phase containing water (A) and acetonitrile (B) was used to achieve separation from endogenous interferences. The gradient program was as follows: 0–1.0 min, 5–50% B, 1.0–3.0 min, 50–90% B, 3.0–4.0 min, 90% B, 4.0–4.1 min, 90–5% B, 4.1–4.5 min, 5% B. The flow rate was set at 0.3 mL/min. The autosampler was conditioned at 4 ◦ C and the injection volume was 2 L for analysis. Mass parameters were optimized by infusing neat solutions of liquiritigenin and internal standard (IS) separately into mass with an apparatus syringe pump. A turbo ion spray interface operating in negative ionization mode was used. Typical source conditions were as follows: curtain-gas (CUR), 35; ion source temperature, 500 ◦ C; ion source gas1 (GAS1), 35; ion source gas2 (GAS2), 35; ionspray voltage, 5500 V; declustering potential (DP), −60.73; collision energy (CE), −33.03; entrance potential (EP), −12.87; collision cell exit potential (CXP), −12.96. Data acquisition was performed with Analyst Software (AB SCIEX, Foster City, CA, USA).
121
2.4. Preparation of standard solutions, calibration standards and quality control samples (QC) Standard stock of liquiritigenin of 1.02 mg/mL and IS of 1.06 mg/mL solutions were prepared using 100% methanol. The concentration of working solution for internal standard was 106 ng/mL. All liquiritigenin and IS solutions were stored at 4 ◦ C until needed. Working solutions of calibration standards (10, 50, 100, 250, 500, 1000, 5000, 10,000 ng/mL) and quality control samples (25, 1000 and 8000 ng/mL) were prepared by diluting the stock solutions with methanol. Calibration standards were prepared by spiking 10 L of working stock to 90 L of blank rat plasma to obtain concentrations of 1, 5, 10, 25, 50, 100, 500, 1000 ng/mL. Quality control (QC) samples (of low, medium and high concentration) at 2.5, 100 and 800 ng/mL were prepared in the same way as the calibration standards and stored at −20 ◦ C. 2.5. Sample preparation 100 L of plasma sample was mixed with 20 L of the IS working solution (106 ng/mL) and vortexed for 30 s. Then, 400 L of acetonitrile was added and the mixture was vortexed for 3 min to precipitate protein. These samples were centrifuged at 12,000 rpm for 5 min. 300 L of the supernatant was then transferred to a new tube and centrifuged further at the same conditions. 2 L of the sample was injected into the HPLC for analysis. 2.6. Method validation 2.6.1. Specificity Plasma samples from six different rats were screened for the presence of endogenous components which might interfere with detection of liquiritigenin or the IS. The specificity was investigated by comparing the chromatograms of the blank plasma with plasma samples spiked with liquiritigenin and IS, as well as samples collected from treated rats. 2.6.2. Linearity and lower limit of quantification (LLOQ) Calibration standards of liquiritigenin at concentrations of 1, 5, 10, 25, 50, 100, 500, 1000 ng/mL were prepared using the standard plasma samples described in Section 2.5. Peak-area ratios of liquiritigenin to IS were utilized for construction of calibration curves, using weighted liner least square regression (weighting: 1/C2 ) of the plasma concentrations and the measured ratios. The linearity of the calibration curve was assessed by plotting the peak-area ratios versus the concentrations of liquiritigenin. The lower limit of quantification (LLOQ) was determined as the lowest concentration on the calibration curve at which accuracy (RE) within ±20% and a precision (R.S.D) below 20% can be obtained. The LOD was defined as the plasma concentration that produced a signal-to-noise ratio (S/N) at 3. 2.6.3. Precision and accuracy The precision and accuracy of the assay were determined at QC samples of low, medium and high concentrations at 2.5, 100 and 800 ng/mL of liquiritigenin, following the steps in Section 2.4. Intraday precision was determined by repeated analyses of each QC sample for six times during one day. Inter-day precision was determined by repeated analyses of six replicates of each QC sample on three consecutive days. The concentration of each sample was calculated by the calibration curve each day. Precision was assessed by calculating the relative standard deviation (R.S.D) for each concentration level. Accuracy was calculated by comparing the average measurements with the nominal values.
122
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
2.6.4. Recovery and matrix effect Extraction recovery was evaluated at three QC levels of 2.5, 100 and 800 ng/mL and IS at one concentration by comparing the peak areas obtained from plasma samples with the analytes spiked before and after extraction. Matrix effects were expressed as the ratio of the mean peak area of analyte spiked post-extraction to that of neat standard solution with mobile phase at equivalent concentrations (n = 5). 2.6.5. Stability To evaluate the stability of liquiritigenin in plasma, six replicates of QC samples at 2.5, 100 and 800 ng/mL were analyzed. Short-term room temperature stability of liquiritigenin in plasma was assessed by analyzing QC samples stored at room temperature for 4 h. For freeze-thaw stability, QC samples were subjected to three freezethaw cycles and analyzed. Long term plasma stability was assessed by analyzing QC samples stored at −20 ◦ C for 45 days. Stability was assessed by comparing the mean concentration of the stability QC samples with the mean concentration of freshly prepared calibration and QC samples. 2.6.6. Pharmacokinetic study Healthy male SD rats, weighting 250–300 g, were obtained from the Animal Center of Nanjing University of Chinese Medicine (Nanjing, China). All rats were housed at 20 ± 5 ◦ C with a relative humidity 30–70% under 12 h dark/light cycle. The rats were fasted for 12 h with free access to water prior to experiments. Liquiritigenin was dissolved with 40% propylene glycol (prepared with water for injection) yielding concentrations of 0.5, 1, 2 mg/mL. Rats were divided into three groups of low, medium and high concentration, dosed intravenously at 0.5, 1 and 2 mg/kg, respectively, through tail vein after ether anesthesia (n = 6). The dosage of administration was calculated by effective dosage of our previous study on mice, which indicated that 20 mg/kg might be the optimal dosage for inhibitory effects [9]. Venous blood samples of 0.3 mL were collected for different time points at zero (control), 1 (end of infusion), 3, 5, 7, 10, 20, 30, 45, 60, 180, 360 min after the start of the intravenous administration. Blood samples were placed into heparinized tubes, and immediately centrifuged at 12,000 rpm for 5 min at 4 ◦ C to harvest plasma. The plasma samples obtained were then stored at −20 ◦ C until analysis. 3. Results and discussion 3.1. Mass spectra analysis The selected internal standard, apigenin, has similar retention time with that of liquiritigenin and exhibits a good response and good peak shape. Furthermore, apigenin is not a metabolite of liquiritigenin in vivo, hence it is appropriate as an IS. When comparing one-step protein precipitation with liquid–liquid extraction, the former is faster, simpler, and has a lower limit of quantification. Since it allows rapid and simple sample preparation, high sensitivity (1 ng/mL) and high degree of specificity, one-step protein precipitation was chosen as the sample preparation method. When the neat solution of liquiritigenin was infused, the precursor ion m/z 255.0 of liquiritigenin [M−H]− (molecular formula: C15 H11 O4 − ) was observed in the negative ionization mode. The product ions fragments of m/z 119.0, 90.9, 135.0, m/z 119.0 (molecular formula: C8 H7 O− ) presented greatest intensity. Similarly, the precursor ion m/z 269.0 of the IS [M-H]− (molecular formula: C15 H9 O5 − ) was also observed in the negative ionization mode and produced fragments of m/z 117.0, 151.0, 106.9, with m/z 117.0 (molecular formula: C8 H5 O− ) showing highest intensity. The MS/MS spectrum of liquiritigenin and the IS were shown in Fig. 1. Two fragments have been monitored, one for quantitation,
Fig. 1. Chemical structures and MS/MS spectrum of liquiritigenin (A) and IS (B).
the ratio between this and the other fragment for confirmation. Precursor/product ion pairs monitored for quantitation were m/z 255.0 → 119.0 (liquiritigenin) and m/z 269.0 → 117.0 (IS), and another pairs for confirmation were m/z 255.0 → 135.0 (liquiritigenin) and m/z 269.0 → 151.0 (IS). Liquiritigenin and IS showed maximum and stable response with the mass parameter of CE, DP, EP and CXP. 3.2. Method validation 3.2.1. Specificity Due to high selectivity of the method, no endogenous interference was observed at retention time of liquiritigenin (2.32 min) and the IS (2.43 min). Representative chromatograms of blank plasma, spiked plasma sample and rat sample were shown in Fig. 2. 3.2.2. Linearity and lower limit of quantification (LLQQ) The method showed good linearity in the range of 1–1000 ng/mL. A typical equation of the calibration curve was y = 0.01381x + 0.01553 (r = 0.99710), where y was the peak area ratio of liquiritigenin to the IS and x was the concentration of liquiritigenin. The LLOQ and LOD for liquiritigenin in plasma were 1 ng/mL (lowest standard level) and 0.29 ng/mL, respectively. 3.2.3. Precision and accuracy The intra- and inter-day precision and accuracy of the method were evaluated using the above three QC samples, the results were listed in Table 1. The precision (R.S.D.) of intra-day and inter-day ranged from 4.54 to 10.65% and 5.94 to 13.81%, respectively; while the accuracy of intra-day and inter-day ranged from 104.06 to 109.28% and 94.98 to 112.05%, respectively. 3.2.4. Recovery and matrix effects The extraction recoveries of liquiritigenin using six replicates of QC samples at three concentrations (2.5, 100, 800 ng/mL) were
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
123
Fig. 2. Typical MRM chromatograms of blank plasma (A), blank plasma spiked with liquiritigenin (1 ng/mL) and IS (B), and rat plasma sample at 1 min after 0.5 mg/kg intravenous administration of liquiritigenin (C).
124
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
Table 1 The validation of intra- and inter-day precision and accuracy with liquiritigenin QC samples (n = 6). Concentration (ng/mL)
2.5 100 800
Precision (% R.S.D.)
Accuracy (%)
Intra-day
Inter-day
Intra-day
Inter-day
10.65 7.13 4.54
13.81 7.50 5.94
106.09 109.28 104.06
94.98 112.05 108.85
Table 2 Stability of liquiritigenin in rat plasma samples under various conditions (n = 6). Concentration added (ng/mL)
Concentration found (ng/mL) (mean ± S.D.)
2.5 At the beginning Short-term room temperature stability After 4 h at room temperature Bias (%) Freeze-thaw stability After three freeze-thaw cycles Bias (%) Long-term freezing storage stability After 45 days at −20 ◦ C Bias (%)
100.0
800.0
2.615 ± 0.078
107.847 ± 4.563
804.913 ± 17.304
2.577 ± 0.149 −1.45
105.179 ± 4.073 −2.47
810.799 ± 13.529 0.73
2.660 ± 0.088 1.72
105.824 ± 5.278 −1.88
813.638 ± 10.158 1.08
2.568 ± 0.140 −1.80
102.852 ± 3.782 −4.63
810.742 ± 8.560 0.72
Table 3 Pharmacokinetic parameters (mean ± S.D., n = 6) of liquiritigenin after a single intravenous injection at the dose of 0.5, 1 and 2 mg/kg, respectively. Parameters
0.5 mg/kg
Cmax (ng/mL) t1/2 (min) AUC0−t (min ng/mL) AUC0−∞ (min ng/mL) MRT (min) CL (mL/min)
2045.03 30.11 8972.60 8911.01 7.13 15.53
± ± ± ± ± ±
1 mg/kg 1037.77 21.48 3995.65 3926.74 3.39 5.87
3669.56 38.71 19,259.41 19,240.43 8.67 16.93
2 mg/kg ± ± ± ± ± ±
1010.50 34.60 9371.37 9358.95 7.03 8.30
4407.02 70.80 23,457.88 23,472.31 19.88 26.97
± ± ± ± ± ±
5983.93 40.53 10,565.63 10,523.24 10.03 10.33
88.02 ± 5.83%, 82.25 ± 6.51%, 80.92 ± 8.43%, respectively, whereas it was 87.17 ± 5.20% for the IS. Under the LC–MS/MS conditions, the matrix effects of the three QC samples were 90.96 ± 2.43%, 92.78 ± 9.66%, 89.77 ± 1.17%, respectively. In addition, the matrix effect of the IS was 93.41 ± 7.47%. No significant matrix effects for liquiritigenin and the IS were observed, indicating that no coeluting substance influenced the ionization of the analyte and the IS. 3.2.5. Stability The stability of QC samples at three concentrations (2.5, 100, 800 ng/mL) during short-term room temperature, freeze-thaw and long-term freezing storage was shown in Table 2. The stability was evaluated by bias (%) of the observed concentration from the actual concentration. Bias (%) = (observed concentration − actual concentration)/actual concentration. As the results have shown, liquiritigenin was stable for 4 h at room temperature, within three freeze-thaw cycles, and for 45 days during long-term freezing storage in plasma samples. 3.2.6. Application The validated method was applied to a pharmacokinetic study of liquiritigenin after intravenous injection to rats. Calibration levels were sufficient to quantitate plasma samples obtained in the pharmacokinetic study. Pharmacokinetic calculations were performed using WinNonlin software (ver. 5.2; Pharsight, Mountain View, CA, USA). Plasma concentration–time curves of liquiritigenin were shown in Fig. 3. The relevant pharmacokinetic parameters, including Cmax , t1/2 , AUC, MRT, CL, were listed in Table 3. As seen from Table 3, the plasma concentrations of liquiritigenin rapidly dropped within the first 30 min following intravenous
Fig. 3. Mean plasma concentration–time profiles of liquiritigenin in SD rats (n = 6) after a single intravenous injection of liquiritigenin at the dose of 0.5, 1, 2 mg/kg, respectively.
administration. After a single intravenous dose of 0.5, 1 and 2 mg/kg (n = 6), liquiritigenin exhibited elimination half-life (t1/2 ) of 30.11 ± 21.48 min, 38.71 ± 34.60 min, 70.80 ± 40.53 min, respectively. However, the t1/2 obtained had significant difference after Student’s t-test was used for comparison, which indicated a probable accumulation and saturation of liquiritigenin at higher dose. Cmax , AUC0−t and AUC0−∞ increased with the dose of liquiritigenin, but did not exhibit dose-dependent growth. The pharmacokinetic properties of liquiritigenin exhibited nonlinear trend. All the data obtained showed that liquiritigenin could be metabolized and eliminated rapidly from the body after administration. According to relevant literature, the rapid elimination of liquiritigenin may be
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
due to drug metabolism or direct excretion of prototype drug [20,21]. Further investigations on tissue distribution, excretion and metabolite of liquiritigenin are still necessary. 4. Conclusion A simple, sensitive and rapid LC–MS/MS method was developed and validated for the quantification of liquiritigenin in rat plasma. This method has been successfully applied to the pharmacokinetic analysis of liquiritigenin involving three-level single-dose intravenous administration of liquiritigenin in rats. Critical validation parameters including specificity, linearity, precision, recovery and stability were all within the acceptable limits. Therefore, this method could be of promising use for pharmacokinetic studies on liquiritigenin. As liquiritigenin is only one of the ingredients in Radix glycyrrhizae, the effects of the other compounds in Radix glycyrrhizae on the pharmacokinetics of liquiritigenin should be considered in future studies. References [1] J.E. Mersereau, N. Levy, R.E. Staub, S. Baggrtt, T. Zogric, S. Chow, W.A. Ricke, M. Tagliaferri, I. Cohen, L.F. Bjeldanes, D.C. Leitman, Mol. Cell Endocrinol. 238 (2008) 49–57. [2] Y.W. Kim, R.J. Zhao, S.J. Pank, J.R. Lee, I.J. Cho, C.H. Yang, S.G. Kim, S.C. Kim, Br. J. Pharmacol. 154 (2008) 165–173. [3] S.C. Kim, S.H. Byun, C.H. Yang, C.Y. Kim, J.W. Kim, S.G. Kim, Toxicology 197 (2004) 239–251.
125
[4] Y.W. Kim, S.H. Ki, J.R. Lee, C.W. Kim, S.C. Kim, S.G. Kim, Chem. Biol. Int. 161 (2006) 125–138. [5] M.J. Falcal, Y.B. Pouliquem, M.A. Lima, N.V. Gramosa, L.V. Costa-Lotufo, G.C. Militao, C. Pessoa, M.O. Moraes, E.R. Silveira, J. Nat. Prod. 68 (2005) 423–426. [6] S.P. Zhang, Y.J. Zhou, Y. Liu, Y.Q. Cai, Food Chem. Toxicol. 47 (2009) 693–701. [7] C.W. Liu, Y. Wang, S.R. Xie, Y.J. Zhou, X.M. Ren, X.T. Li, Y.Q. Cai, Phytother. Res. 25 (2011) 277. [8] S.R. Xie, Y. Wang, C.W. Liu, K. Luo, Y.Q. Cai, Phytother. Res. 26 (2011) 1133–1141. [9] M. Zhou, H. Higo, Y.Q. Cai, Phytother. Res. 24 (2010) 827. [10] H.E. Kang, Y.K. Cho, H.Y. Jung, K.Y. Choi, S.I. Sohn, S.R. Baek, M.G. Lee, Xenobiotica 39 (2009) 465. [11] H.E. Kang, H.Y. Jung, Y.K. Cho, S.H. Kim, S.I. Sohn, S.R. Baek, M.G. Lee, J. Pharm. Sci. 98 (2009) 4327. [12] H.E. Kang, H.J. Chung, H.S. Kim, J.W. Lee, M.G. Lee, Eur. J. Pharm. Sci. 39 (2010) 181. [13] H.E. Kang, Y.W. Kim, S.I. Sohn, S.R. Baek, J.W. Lee, S.G. Kim, I. Lee, M.G. Lee, Xenobiotica 40 (2010) 424. [14] H.E. Kang, S.I. Sohn, S.R. Baek, J.W. Lee, M.G. Lee, J. Pharm. Pharmacol. 63 (2011) 49. [15] Y.P. Wu, X.S. Meng, Y.R. Bao, S. Wang, J. Ethnopharmacol. 148 (2013) 266. [16] T.X. Li, Z.X. Yan, C. Zhou, J. Sun, C. Jiang, X.H. Yang, Biomed. Chromatogr. 27 (2013) 1041. [17] L. Li, S.P. Liang, F.F. Du, C. Li, J. Am. Soc. Mass Spectrom. 18 (2007) 778. [18] X. Qiao, M. Ye, C. Xiang, Q. Wang, C.F. Liu, W.J. Miao, D.A. Guo, J. Chromatogr. A 1258 (2012) 84–93. [19] Y.L. Wang, C.G. Ding, C.S. Wu, K.H. Du, J.L. Zhang, H.L. Qin, J.F. Hou, G.H. Du, Acta Pharm. Sin. B 2 (2012) 188–197. [20] H. Shimamura, S. Nakai, K. Yamamoto, N. Hitomi, E. Yumioka, Jpn. J. Pharmacogn. 44 (1990) 265–268. [21] H. Shimamura, H. Suzuki, M. Hanano, A. Suzuki, Y. Sugiyama, Biol. Pharmaceut. Bull. 16 (1993) 899–907.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Determination of liquiritigenin by ultra high performance liquid chromatography coupled with triple quadrupole mass spectrometry: Application to a linear pharmacokinetic study of liquiritigenin in rat plasma Jie Gu a,b , Huan Li a,b,c,d , Ke Pei a,b , Hui Cai d , Kunming Qin a,b , Xinghai Zhang a,b , Lijuan Zheng a,b , Xiao Liu a,b , Yunqing Cai e , Baochang Cai a,b,∗ a
Pharmacy College of Nanjing University of TCM, Nanjing 210023, China Nanjing University of Chinese Medicine, Jiangsu Key Laboratory of Chinese Medicine Processing, Engineering Center of State Ministry of Education for Standardization of Chinese Medicine Processing, Nanjing 210023, China c Institute of Bioengineering and Nanotechnology, A* STAR, The Nanos, Singapore 138669, Singapore d School of Applied Science, Temasek Polytechnic, Singapore 529757, Singapore e Department of Nutrition and Food Hygiene, School of Public Health, Nanjing Medical University, Nanjing 210029, China b
a r t i c l e
i n f o
Article history: Received 10 February 2014 Accepted 7 September 2014 Available online 17 September 2014 Keywords: Liquiritigenin UHPLC–MS/MS Intravenous administration Linear pharmacokinetic study
a b s t r a c t A simple, sensitive and rapid ultra high performance liquid chromatography tandem mass spectrometry (UHPLC–MS/MS) method has been developed and validated for the quantification of liquiritigenin, a promising anti-tumor agent. Liquiritigenin and the internal standard were separated on an Agilent Extend C18 column and eluted with a gradient mobile phase system of acetonitrile and water. The analysis was performed on a negative ionization electrospray mass spectrometer via multiple reaction monitoring (MRM). Transitions of m/z 255.0 → 119.0 for liquiritigenin and m/z 269.0 → 117.0 for the IS were monitored. One-step protein precipitation with acetonitrile was used to remove impurities and extract the analytes from plasma. The method had a chromatographic run time of 4.5 min and a good linearity in the range of 1–1000 ng/mL. The precision (R.S.D.) of intra-day and inter-day ranged from 4.54 to 10.65% and 5.94 to 13.81%, respectively; while the accuracy of intra-day and inter-day ranged from 104.06 to 109.28% and 94.98 to 112.05%. The recovery and stability were also within the acceptable limits. The validated method was applied to a linear pharmacokinetic study of liquiritigenin in rat plasma for the first time. © 2014 Published by Elsevier B.V.
1. Introduction Liquiritigenin (7,4 -dyhydroxy flavanone) is a poly phenolic flavanone structure that exists in Radix glycyrrhizae. As an inhibitor of Akt protein kinase and selective estrogen receptor agonist [1], liquiritigenin has been found to have various biological activities, such as anti-inflammatory [2], cytoprotection [3], inhibition of acute hepatic injury [4], and antitumor activity. Liquiritigenin has been shown to possess significant antitumor activity. It has cytotoxic activity against five human cancer cell lines
∗ Corresponding author at: Nanjing University of Chinese Medicine, Jiangsu Key Laboratory of Chinese Medicine Processing, Engineering Center of State Ministry of Education for Standardization of Chinese Medicine Processing, Nanjing 210023, China. Tel.: +86 86798281; fax: +86 25 86798281. E-mail address: [email protected] (B. Cai). http://dx.doi.org/10.1016/j.jchromb.2014.09.002 1570-0232/© 2014 Published by Elsevier B.V.
in vitro [5]. In accordance with our previous studies, liquiritigenin inhibits the growth of SGC-7901 and Lovo cells. Time- and dosedependent inhibition and apoptosis induction were particularly exhibited on SMMC 7721 cells [6]. Liquiritigenin suppresses cancer cell proliferation and angiogenesis, and promotes apoptosis [7,8]. These findings suggested strong cancer preventive effects of liquiritigenin in vitro, yet the underlying molecular mechanisms remain unknown. Our previous study also showed that liquiritigenin significantly inhibited the growth of murine H22 hepatocarcinoma in vivo through a 15-day treatment after tumor inoculation and exhibited potent tumor therapeutic activity [9]. All these results suggested that liquiritigenin could be a promising antitumor agent for human cancers. To ensure safety and effectiveness of the medication, a novel UHPLC–MS/MS method, aiming at both simplicity and sensitivity, was established for the preclinical linear pharmacokinetic study. Several methods involving high-performance liquid
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
chromatography and mass spectrometry have been described previously to analyze liquiritigenin concentration in plasma after administration [10–19]. While these methods provide precise quantitation, the sample preparation and mobile phase composition of such methods are rather complex. A longer time is also required for analysis. In conventional HPLC analysis, liquiritigenin has a long retention time of approximately 9.8 min. By focusing on a single compound rather than multiple constituents in the plasma, shorter analysis time can be achieved by UHPLC–MS/MS. In the present work, we developed and validated a UHPLC–MS/MS method for the determination and quantitation of liquiritigenin in rat plasma that boasts higher specificity, increased sensitivity, simpler sample preparation, and shorter analysis time. The method has been applied to the pharmacokinetic study involving three-level single-dose intravenous administration of liquiritigenin in rats for the first time.
2. Materials and methods 2.1. Materials and reagents Liquiritigenin (purity > 98.0%) was provided by Professor Wei Li from Nanjing University of Chinese Medicine. The internal standard, apigenin was purchased from Shanghai Yuanye Biotechnology Co. Ltd. (Shanghai, China), lot number: 20120220. HPLC–MS-grade acetonitrile and water were purchased from ANPEL Scientific Instrument Co. Ltd. (Shanghai, China). HPLC-grade acetonitrile and methanol were purchased from Merck (Darmstadt, Germany).
2.2. Instruments The chromatography system used was composed of a Shimadzu LC-30A series chromatographic system (Shimadzu Corporation UFLC XR, Kyoto, Japan) with a LC-30AD binary pump, a DGU-20A 5R degasser, a SIL-30AC auto sampler and a CTO-30A column oven. The separation was performed using the C18 column (Agilent, USA), 2.1 × 100 mm, 1.8 m, at a temperature of 35 ◦ C. Mass spectrometric detection was performed by using 5500 triple quad tandem mass spectrometer equipped with electrospray ionization (ESI) source (AB SCIEX, Foster City, CA, USA). Peak Genius2 3030 instrument was used for generating nitrogen gas.
2.3. LC–MS/MS conditions The mobile phase containing water (A) and acetonitrile (B) was used to achieve separation from endogenous interferences. The gradient program was as follows: 0–1.0 min, 5–50% B, 1.0–3.0 min, 50–90% B, 3.0–4.0 min, 90% B, 4.0–4.1 min, 90–5% B, 4.1–4.5 min, 5% B. The flow rate was set at 0.3 mL/min. The autosampler was conditioned at 4 ◦ C and the injection volume was 2 L for analysis. Mass parameters were optimized by infusing neat solutions of liquiritigenin and internal standard (IS) separately into mass with an apparatus syringe pump. A turbo ion spray interface operating in negative ionization mode was used. Typical source conditions were as follows: curtain-gas (CUR), 35; ion source temperature, 500 ◦ C; ion source gas1 (GAS1), 35; ion source gas2 (GAS2), 35; ionspray voltage, 5500 V; declustering potential (DP), −60.73; collision energy (CE), −33.03; entrance potential (EP), −12.87; collision cell exit potential (CXP), −12.96. Data acquisition was performed with Analyst Software (AB SCIEX, Foster City, CA, USA).
121
2.4. Preparation of standard solutions, calibration standards and quality control samples (QC) Standard stock of liquiritigenin of 1.02 mg/mL and IS of 1.06 mg/mL solutions were prepared using 100% methanol. The concentration of working solution for internal standard was 106 ng/mL. All liquiritigenin and IS solutions were stored at 4 ◦ C until needed. Working solutions of calibration standards (10, 50, 100, 250, 500, 1000, 5000, 10,000 ng/mL) and quality control samples (25, 1000 and 8000 ng/mL) were prepared by diluting the stock solutions with methanol. Calibration standards were prepared by spiking 10 L of working stock to 90 L of blank rat plasma to obtain concentrations of 1, 5, 10, 25, 50, 100, 500, 1000 ng/mL. Quality control (QC) samples (of low, medium and high concentration) at 2.5, 100 and 800 ng/mL were prepared in the same way as the calibration standards and stored at −20 ◦ C. 2.5. Sample preparation 100 L of plasma sample was mixed with 20 L of the IS working solution (106 ng/mL) and vortexed for 30 s. Then, 400 L of acetonitrile was added and the mixture was vortexed for 3 min to precipitate protein. These samples were centrifuged at 12,000 rpm for 5 min. 300 L of the supernatant was then transferred to a new tube and centrifuged further at the same conditions. 2 L of the sample was injected into the HPLC for analysis. 2.6. Method validation 2.6.1. Specificity Plasma samples from six different rats were screened for the presence of endogenous components which might interfere with detection of liquiritigenin or the IS. The specificity was investigated by comparing the chromatograms of the blank plasma with plasma samples spiked with liquiritigenin and IS, as well as samples collected from treated rats. 2.6.2. Linearity and lower limit of quantification (LLOQ) Calibration standards of liquiritigenin at concentrations of 1, 5, 10, 25, 50, 100, 500, 1000 ng/mL were prepared using the standard plasma samples described in Section 2.5. Peak-area ratios of liquiritigenin to IS were utilized for construction of calibration curves, using weighted liner least square regression (weighting: 1/C2 ) of the plasma concentrations and the measured ratios. The linearity of the calibration curve was assessed by plotting the peak-area ratios versus the concentrations of liquiritigenin. The lower limit of quantification (LLOQ) was determined as the lowest concentration on the calibration curve at which accuracy (RE) within ±20% and a precision (R.S.D) below 20% can be obtained. The LOD was defined as the plasma concentration that produced a signal-to-noise ratio (S/N) at 3. 2.6.3. Precision and accuracy The precision and accuracy of the assay were determined at QC samples of low, medium and high concentrations at 2.5, 100 and 800 ng/mL of liquiritigenin, following the steps in Section 2.4. Intraday precision was determined by repeated analyses of each QC sample for six times during one day. Inter-day precision was determined by repeated analyses of six replicates of each QC sample on three consecutive days. The concentration of each sample was calculated by the calibration curve each day. Precision was assessed by calculating the relative standard deviation (R.S.D) for each concentration level. Accuracy was calculated by comparing the average measurements with the nominal values.
122
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
2.6.4. Recovery and matrix effect Extraction recovery was evaluated at three QC levels of 2.5, 100 and 800 ng/mL and IS at one concentration by comparing the peak areas obtained from plasma samples with the analytes spiked before and after extraction. Matrix effects were expressed as the ratio of the mean peak area of analyte spiked post-extraction to that of neat standard solution with mobile phase at equivalent concentrations (n = 5). 2.6.5. Stability To evaluate the stability of liquiritigenin in plasma, six replicates of QC samples at 2.5, 100 and 800 ng/mL were analyzed. Short-term room temperature stability of liquiritigenin in plasma was assessed by analyzing QC samples stored at room temperature for 4 h. For freeze-thaw stability, QC samples were subjected to three freezethaw cycles and analyzed. Long term plasma stability was assessed by analyzing QC samples stored at −20 ◦ C for 45 days. Stability was assessed by comparing the mean concentration of the stability QC samples with the mean concentration of freshly prepared calibration and QC samples. 2.6.6. Pharmacokinetic study Healthy male SD rats, weighting 250–300 g, were obtained from the Animal Center of Nanjing University of Chinese Medicine (Nanjing, China). All rats were housed at 20 ± 5 ◦ C with a relative humidity 30–70% under 12 h dark/light cycle. The rats were fasted for 12 h with free access to water prior to experiments. Liquiritigenin was dissolved with 40% propylene glycol (prepared with water for injection) yielding concentrations of 0.5, 1, 2 mg/mL. Rats were divided into three groups of low, medium and high concentration, dosed intravenously at 0.5, 1 and 2 mg/kg, respectively, through tail vein after ether anesthesia (n = 6). The dosage of administration was calculated by effective dosage of our previous study on mice, which indicated that 20 mg/kg might be the optimal dosage for inhibitory effects [9]. Venous blood samples of 0.3 mL were collected for different time points at zero (control), 1 (end of infusion), 3, 5, 7, 10, 20, 30, 45, 60, 180, 360 min after the start of the intravenous administration. Blood samples were placed into heparinized tubes, and immediately centrifuged at 12,000 rpm for 5 min at 4 ◦ C to harvest plasma. The plasma samples obtained were then stored at −20 ◦ C until analysis. 3. Results and discussion 3.1. Mass spectra analysis The selected internal standard, apigenin, has similar retention time with that of liquiritigenin and exhibits a good response and good peak shape. Furthermore, apigenin is not a metabolite of liquiritigenin in vivo, hence it is appropriate as an IS. When comparing one-step protein precipitation with liquid–liquid extraction, the former is faster, simpler, and has a lower limit of quantification. Since it allows rapid and simple sample preparation, high sensitivity (1 ng/mL) and high degree of specificity, one-step protein precipitation was chosen as the sample preparation method. When the neat solution of liquiritigenin was infused, the precursor ion m/z 255.0 of liquiritigenin [M−H]− (molecular formula: C15 H11 O4 − ) was observed in the negative ionization mode. The product ions fragments of m/z 119.0, 90.9, 135.0, m/z 119.0 (molecular formula: C8 H7 O− ) presented greatest intensity. Similarly, the precursor ion m/z 269.0 of the IS [M-H]− (molecular formula: C15 H9 O5 − ) was also observed in the negative ionization mode and produced fragments of m/z 117.0, 151.0, 106.9, with m/z 117.0 (molecular formula: C8 H5 O− ) showing highest intensity. The MS/MS spectrum of liquiritigenin and the IS were shown in Fig. 1. Two fragments have been monitored, one for quantitation,
Fig. 1. Chemical structures and MS/MS spectrum of liquiritigenin (A) and IS (B).
the ratio between this and the other fragment for confirmation. Precursor/product ion pairs monitored for quantitation were m/z 255.0 → 119.0 (liquiritigenin) and m/z 269.0 → 117.0 (IS), and another pairs for confirmation were m/z 255.0 → 135.0 (liquiritigenin) and m/z 269.0 → 151.0 (IS). Liquiritigenin and IS showed maximum and stable response with the mass parameter of CE, DP, EP and CXP. 3.2. Method validation 3.2.1. Specificity Due to high selectivity of the method, no endogenous interference was observed at retention time of liquiritigenin (2.32 min) and the IS (2.43 min). Representative chromatograms of blank plasma, spiked plasma sample and rat sample were shown in Fig. 2. 3.2.2. Linearity and lower limit of quantification (LLQQ) The method showed good linearity in the range of 1–1000 ng/mL. A typical equation of the calibration curve was y = 0.01381x + 0.01553 (r = 0.99710), where y was the peak area ratio of liquiritigenin to the IS and x was the concentration of liquiritigenin. The LLOQ and LOD for liquiritigenin in plasma were 1 ng/mL (lowest standard level) and 0.29 ng/mL, respectively. 3.2.3. Precision and accuracy The intra- and inter-day precision and accuracy of the method were evaluated using the above three QC samples, the results were listed in Table 1. The precision (R.S.D.) of intra-day and inter-day ranged from 4.54 to 10.65% and 5.94 to 13.81%, respectively; while the accuracy of intra-day and inter-day ranged from 104.06 to 109.28% and 94.98 to 112.05%, respectively. 3.2.4. Recovery and matrix effects The extraction recoveries of liquiritigenin using six replicates of QC samples at three concentrations (2.5, 100, 800 ng/mL) were
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
123
Fig. 2. Typical MRM chromatograms of blank plasma (A), blank plasma spiked with liquiritigenin (1 ng/mL) and IS (B), and rat plasma sample at 1 min after 0.5 mg/kg intravenous administration of liquiritigenin (C).
124
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
Table 1 The validation of intra- and inter-day precision and accuracy with liquiritigenin QC samples (n = 6). Concentration (ng/mL)
2.5 100 800
Precision (% R.S.D.)
Accuracy (%)
Intra-day
Inter-day
Intra-day
Inter-day
10.65 7.13 4.54
13.81 7.50 5.94
106.09 109.28 104.06
94.98 112.05 108.85
Table 2 Stability of liquiritigenin in rat plasma samples under various conditions (n = 6). Concentration added (ng/mL)
Concentration found (ng/mL) (mean ± S.D.)
2.5 At the beginning Short-term room temperature stability After 4 h at room temperature Bias (%) Freeze-thaw stability After three freeze-thaw cycles Bias (%) Long-term freezing storage stability After 45 days at −20 ◦ C Bias (%)
100.0
800.0
2.615 ± 0.078
107.847 ± 4.563
804.913 ± 17.304
2.577 ± 0.149 −1.45
105.179 ± 4.073 −2.47
810.799 ± 13.529 0.73
2.660 ± 0.088 1.72
105.824 ± 5.278 −1.88
813.638 ± 10.158 1.08
2.568 ± 0.140 −1.80
102.852 ± 3.782 −4.63
810.742 ± 8.560 0.72
Table 3 Pharmacokinetic parameters (mean ± S.D., n = 6) of liquiritigenin after a single intravenous injection at the dose of 0.5, 1 and 2 mg/kg, respectively. Parameters
0.5 mg/kg
Cmax (ng/mL) t1/2 (min) AUC0−t (min ng/mL) AUC0−∞ (min ng/mL) MRT (min) CL (mL/min)
2045.03 30.11 8972.60 8911.01 7.13 15.53
± ± ± ± ± ±
1 mg/kg 1037.77 21.48 3995.65 3926.74 3.39 5.87
3669.56 38.71 19,259.41 19,240.43 8.67 16.93
2 mg/kg ± ± ± ± ± ±
1010.50 34.60 9371.37 9358.95 7.03 8.30
4407.02 70.80 23,457.88 23,472.31 19.88 26.97
± ± ± ± ± ±
5983.93 40.53 10,565.63 10,523.24 10.03 10.33
88.02 ± 5.83%, 82.25 ± 6.51%, 80.92 ± 8.43%, respectively, whereas it was 87.17 ± 5.20% for the IS. Under the LC–MS/MS conditions, the matrix effects of the three QC samples were 90.96 ± 2.43%, 92.78 ± 9.66%, 89.77 ± 1.17%, respectively. In addition, the matrix effect of the IS was 93.41 ± 7.47%. No significant matrix effects for liquiritigenin and the IS were observed, indicating that no coeluting substance influenced the ionization of the analyte and the IS. 3.2.5. Stability The stability of QC samples at three concentrations (2.5, 100, 800 ng/mL) during short-term room temperature, freeze-thaw and long-term freezing storage was shown in Table 2. The stability was evaluated by bias (%) of the observed concentration from the actual concentration. Bias (%) = (observed concentration − actual concentration)/actual concentration. As the results have shown, liquiritigenin was stable for 4 h at room temperature, within three freeze-thaw cycles, and for 45 days during long-term freezing storage in plasma samples. 3.2.6. Application The validated method was applied to a pharmacokinetic study of liquiritigenin after intravenous injection to rats. Calibration levels were sufficient to quantitate plasma samples obtained in the pharmacokinetic study. Pharmacokinetic calculations were performed using WinNonlin software (ver. 5.2; Pharsight, Mountain View, CA, USA). Plasma concentration–time curves of liquiritigenin were shown in Fig. 3. The relevant pharmacokinetic parameters, including Cmax , t1/2 , AUC, MRT, CL, were listed in Table 3. As seen from Table 3, the plasma concentrations of liquiritigenin rapidly dropped within the first 30 min following intravenous
Fig. 3. Mean plasma concentration–time profiles of liquiritigenin in SD rats (n = 6) after a single intravenous injection of liquiritigenin at the dose of 0.5, 1, 2 mg/kg, respectively.
administration. After a single intravenous dose of 0.5, 1 and 2 mg/kg (n = 6), liquiritigenin exhibited elimination half-life (t1/2 ) of 30.11 ± 21.48 min, 38.71 ± 34.60 min, 70.80 ± 40.53 min, respectively. However, the t1/2 obtained had significant difference after Student’s t-test was used for comparison, which indicated a probable accumulation and saturation of liquiritigenin at higher dose. Cmax , AUC0−t and AUC0−∞ increased with the dose of liquiritigenin, but did not exhibit dose-dependent growth. The pharmacokinetic properties of liquiritigenin exhibited nonlinear trend. All the data obtained showed that liquiritigenin could be metabolized and eliminated rapidly from the body after administration. According to relevant literature, the rapid elimination of liquiritigenin may be
J. Gu et al. / J. Chromatogr. B 973 (2014) 120–125
due to drug metabolism or direct excretion of prototype drug [20,21]. Further investigations on tissue distribution, excretion and metabolite of liquiritigenin are still necessary. 4. Conclusion A simple, sensitive and rapid LC–MS/MS method was developed and validated for the quantification of liquiritigenin in rat plasma. This method has been successfully applied to the pharmacokinetic analysis of liquiritigenin involving three-level single-dose intravenous administration of liquiritigenin in rats. Critical validation parameters including specificity, linearity, precision, recovery and stability were all within the acceptable limits. Therefore, this method could be of promising use for pharmacokinetic studies on liquiritigenin. As liquiritigenin is only one of the ingredients in Radix glycyrrhizae, the effects of the other compounds in Radix glycyrrhizae on the pharmacokinetics of liquiritigenin should be considered in future studies. References [1] J.E. Mersereau, N. Levy, R.E. Staub, S. Baggrtt, T. Zogric, S. Chow, W.A. Ricke, M. Tagliaferri, I. Cohen, L.F. Bjeldanes, D.C. Leitman, Mol. Cell Endocrinol. 238 (2008) 49–57. [2] Y.W. Kim, R.J. Zhao, S.J. Pank, J.R. Lee, I.J. Cho, C.H. Yang, S.G. Kim, S.C. Kim, Br. J. Pharmacol. 154 (2008) 165–173. [3] S.C. Kim, S.H. Byun, C.H. Yang, C.Y. Kim, J.W. Kim, S.G. Kim, Toxicology 197 (2004) 239–251.
125
[4] Y.W. Kim, S.H. Ki, J.R. Lee, C.W. Kim, S.C. Kim, S.G. Kim, Chem. Biol. Int. 161 (2006) 125–138. [5] M.J. Falcal, Y.B. Pouliquem, M.A. Lima, N.V. Gramosa, L.V. Costa-Lotufo, G.C. Militao, C. Pessoa, M.O. Moraes, E.R. Silveira, J. Nat. Prod. 68 (2005) 423–426. [6] S.P. Zhang, Y.J. Zhou, Y. Liu, Y.Q. Cai, Food Chem. Toxicol. 47 (2009) 693–701. [7] C.W. Liu, Y. Wang, S.R. Xie, Y.J. Zhou, X.M. Ren, X.T. Li, Y.Q. Cai, Phytother. Res. 25 (2011) 277. [8] S.R. Xie, Y. Wang, C.W. Liu, K. Luo, Y.Q. Cai, Phytother. Res. 26 (2011) 1133–1141. [9] M. Zhou, H. Higo, Y.Q. Cai, Phytother. Res. 24 (2010) 827. [10] H.E. Kang, Y.K. Cho, H.Y. Jung, K.Y. Choi, S.I. Sohn, S.R. Baek, M.G. Lee, Xenobiotica 39 (2009) 465. [11] H.E. Kang, H.Y. Jung, Y.K. Cho, S.H. Kim, S.I. Sohn, S.R. Baek, M.G. Lee, J. Pharm. Sci. 98 (2009) 4327. [12] H.E. Kang, H.J. Chung, H.S. Kim, J.W. Lee, M.G. Lee, Eur. J. Pharm. Sci. 39 (2010) 181. [13] H.E. Kang, Y.W. Kim, S.I. Sohn, S.R. Baek, J.W. Lee, S.G. Kim, I. Lee, M.G. Lee, Xenobiotica 40 (2010) 424. [14] H.E. Kang, S.I. Sohn, S.R. Baek, J.W. Lee, M.G. Lee, J. Pharm. Pharmacol. 63 (2011) 49. [15] Y.P. Wu, X.S. Meng, Y.R. Bao, S. Wang, J. Ethnopharmacol. 148 (2013) 266. [16] T.X. Li, Z.X. Yan, C. Zhou, J. Sun, C. Jiang, X.H. Yang, Biomed. Chromatogr. 27 (2013) 1041. [17] L. Li, S.P. Liang, F.F. Du, C. Li, J. Am. Soc. Mass Spectrom. 18 (2007) 778. [18] X. Qiao, M. Ye, C. Xiang, Q. Wang, C.F. Liu, W.J. Miao, D.A. Guo, J. Chromatogr. A 1258 (2012) 84–93. [19] Y.L. Wang, C.G. Ding, C.S. Wu, K.H. Du, J.L. Zhang, H.L. Qin, J.F. Hou, G.H. Du, Acta Pharm. Sin. B 2 (2012) 188–197. [20] H. Shimamura, S. Nakai, K. Yamamoto, N. Hitomi, E. Yumioka, Jpn. J. Pharmacogn. 44 (1990) 265–268. [21] H. Shimamura, H. Suzuki, M. Hanano, A. Suzuki, Y. Sugiyama, Biol. Pharmaceut. Bull. 16 (1993) 899–907.
