Research article Received: 15 August 2013,
Revised: 6 January 2014,
Accepted: 17 January 2014
Published online in Wiley Online Library: 12 March 2014
(wileyonlinelibrary.com) DOI 10.1002/bmc.3165
The pharmacokinetic study of sinomenine, paeoniflorin and paeonol in rats after oral administration of a herbal product Qingfu Guanjiesu capsule by HPLC Ying Xiea, Zhi-Hong Jianga, Hua Zhoua, Wen-Zhe Maa, Yuen-Fan Wongb, Zhong-Qiu Liuc and Liang Liua* ABSTRACT: An accurate and reliable high-performance liquid chromatography–diode array detector (HPLC-DAD) method was developed and validated for determination of sinomenine (SI), paeoniflorin (PF) and paeonol (PA), which was further applied to assess the pharmacokinetics of SI, PF and PA in an anti-arthritic herbal product, Qingfu Guanjieshu (QFGJS) capsule, in rats. Successful separation was achieved with a C18 column and a mobile phase composed of acetonitrile and aqueous phase (containing 0.1% formic acid, adjusted with triethylamine to pH 3.5 ± 0.2). The method was validated with excellent precision, accuracy, recovery and stability in calibration ranges from 0.06 to 11.62 μg/mL for SI, from 0.09 to 35.70 μg/mL for PF, and from 0.15 to 4.53 μg/mL for PA (with r2 > 0.999 for all three compounds). Our results showed that absorption of PF after administration of QFGJS was similar to that after oral administration of PF alone; the absorption of SI was decreased while the absorption of PA was increased after giving QFGJS orally compared with pure compounds. We may conclude that pharmacokinetic studies of complex herbal products are not only necessary but also feasible by using representative bioactive chemicals as indicators of establishing quality control standards and of determining pharmacokinetic behavior of herbal medicines. Copyright © 2014 John Wiley & Sons, Ltd. Keywords: HPLC-DAD; pharmacokinetics; sinomenine; paeoniflorin; paeonol; herbal product
Introduction
1294
Although herbal medicinal products (HMPs) are used extensively throughout the world, especially in the treatment of chronic diseases, the relationship between clinical efficacy and the bioactive compounds contained in many HMPs remains unknown. Since knowledge of pharmacokinetics can help us to understand and predict a variety of events related to the efficacy and toxicity of HMPs, it is important to perform pharmacokinetic studies of their ingredients in the plasma when the complex mixtures are administered. Qingfu Guanjieshu (QFGJS) capsule, an anti-arthritic standardized herbal product, is formulated from five herbs – Caulis Sinomenii, Radix Aconiti Lateralis Preparata, Rhizoma Curcumae Longae, Radix Paeoniae Alba and Cortex Moutan – all of which have well-established records of usage for treatment of arthritic diseases in China, Japan and other Asian countries (Chen et al., 2004). Distinguishing it from most currently marked herbal products in China and even in the Western world, systemic quality control system has been established for QFGJS (Xie et al., 2007). Together with the optimal manufacturing process, intensive comparative studies on the extraction conditions, contents of the bioactive constituents and the effects of anti-inflammationa, anti-arthritis and antinociception in rats have warranted the safety and effectiveness of QFG (Xie et al., 2007; Cai et al., 2005; Zhou et al., 2006; Liang et al., 1990). It is currently in the pipeline of research and development to be approved as a novel new drug for arthritis treatment in China.
Biomed. Chromatogr. 2014; 28: 1294–1302
Such a herbal formula is quite chemically complex, with hundreds or even thousands of constituents, which makes pharmacokinetic study of the product very difficult. Nevertheless, even though the exact chemical profile and the nature and even the interactions of its constituents remain unknown, certain potent bioactive chemicals contained in QFGJS have been previously identified. Among these are: sinomenine (SI) in Caulis Sinomenii; paeoniflorin (PF) in Radix Paeoniae Alba; and paeonol (PA) in Cortex Moutan (Fig. 1). Previous pharmacological studies have shown that these compounds exhibit marked anti-inflammatory, analgesic, anti-arthritic and immunosuppressive activities (Zhuang
* Correspondence to: L. Liu, State Key Laboratory for Quality Research of Chinese Medicines, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau. Email: [email protected] a
State Key Laboratory for Quality Research of Chinese Medicines, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau
b
School of Chinese Medicine, Hong Kong Baptist University, Kowloon Tong, Hong Kong
c
School of Pharmacy, Southern Medical University, Guangzhou, China Abbreviations used: ACN, acetonitrile; AUC, area under the concentration–time curve; Cmax, maximum plasma concentration; CV, coefficient of variation; HMPs, herbal medicinal products; HPLC-DAD, high-performance liquid chromatography coupled with diode array detector; LLOQ, lower limit of quantification; PA, paeonol; PF, paeoniflorin; QFGJS, Qingfu Guanjieshu capsule; SI, sinomenine; Tmax, maximum plasma concentration
Copyright © 2014 John Wiley & Sons, Ltd.
PK study of sinomenine, paeoniflorin and paeonol in a herbal product and Biological Products, Beijing, China. Pentoxifylline, the internal standard (IS), was purchased from Sigma (St Louis, MO, USA). The identity and purity of these chemicals were further validated by LC-MS in our laboratory.
HPLC conditions
Figure 1. Chemical structures of sinomenine, paeoniflorin, paeonol, and pentoxifylline (internal standard).
et al., 2013; Liang et al., 1990; Ostrem et al., 2013; Wisnovsky Simon et al., 2013; Tsai et al., 2001). Thus, these three compounds were treated as the representative constituents of QFGJS for the primary pharmacokinetic study. Our previous studies showed that the pharmacokinetic (PK) parameters, particularly the oral bioavailability of PF (after a single dosage of 150 mg/kg), could be markedly enhanced by co-administration of SI at a dosage of 90 mg/kg in rats (Liu et al., 2005a). Moreover, many other reports have demonstrated significant variation of the pharmacokinetic parameters of PF and PA after oral administration, either when purified compounds were given or when a herbal product containing PF or PA was given (Sheng et al., 2004; Wang et al., 2006; Yang et al., 2006). This indicates that multiple chemicals in HMPs could alter the pharmacokinetics of each other and thus may influence safety and therapeutic efficacy. Therefore, it is important to evaluate the pharmacokinetic profiles of HMPs by monitoring the plasma concentrations of representative bioactive components. However, previous reported chromatographic methods for determination of a single representative compound cannot be used to determine all of them in bio-samples at the same time. In the present study, a new HPLC-DAD method was developed for simultaneous determination of SI, PF and PA in plasma and different sample preparation methods were compared with consideration of a small volume of plasma and efficiency of determination. The method was validated according to the guidance of the US Food and Drug Administration (2001). Using the developed method, we described the PK profile of SI, PF and PA after giving QFGJS orally as part of the preclinical evaluation of this drug, and studied the variation of between the PK parameters of SI, PF and PA in QFGJS and that of pure compounds. Although different bio-assay methods were used for studying PK behavior of PA, similar parameters were observed to the previous report (Xie et al., 2008), and the increased absorption of PA in QFGJS was further confirmed in the current study.
Experimental Chemicals and reagents
Biomed. Chromatogr. 2014; 28: 1294–1302
Animals and surgery Male Sprague–Dawley rats weighing 200–250 g were purchased from the Laboratory Animal Services Center of the Chinese University of Hong Kong. Animals were housed four per cage with food and water provided ad libitum and acclimated in the laboratory for at least one week prior to the experiment. Rats were anesthetized by intraperitoneal administration of 0.7% chloral hydrate saline solution at a dosage of 5 mL/kg body weight (Huang et al., 2004). A longitudinal skin incision was made over the area where the right external jugular vein passes dorsal to the pectoral major muscle. The catheter (Silastic catalog no. 602-155; Dow Corning, Midland, MI, USA), filled with 20 units/mL heparinized saline solution, was put into the right jugular vein and then advanced into the sinus venosus. The catheter was inserted up to the first silicone stopper and anchored in place by suturing the stopper to muscle. The free end of the catheter was passed under the skin of the dorsum of the neck just caudal to the ears and attached to the skin, together with a metal spring, which was covered with PVC tubing for protection of the outer part of the catheter. Finally, the catheter was filled with 500 units/mL heparinized saline solution, and a plug was inserted in the free end of the catheter. After surgery, rats were housed individually in metabolite cages for 5 days of recovery and underwent pharmacokinetic treatment according to the jugular-catheterized rat model (Thrivikraman et al., 2002). All procedures involving animals and their care were approved and performed under the regulations of the Committee on Use of Human and Animal Subjects in Teaching and Research of Hong Kong Baptist University and the Department of the Health Department of Hong Kong Special Administration Region (permit number 09-35 in DH/HA&P/8/2/6 Part 1).
Drug administration and blood sampling QFGJS capsules were prepared from five herbs with pharmaceutical methods described in our previous reports (Zhou et al., 2006). The chemical consistency and stability of the QFGJS product were examined by quantitative determination of five marker compounds (SI, PF, PA, curcumin and hypaconitine) and by qualitative fingerprint analysis, which has been previously described (Xie et al., 2007). Contents of SI, PF, and PA in QFGJS capsules were determined to be 5.9, 25.3 and 19.0 mg/g, respectively. The rats were randomly divided into two groups with eight rats in each group. The oral administration of QFGJS was divided into two dosages of 3.89 and 0.97 g/20 mL/kg body weight prepared by dissolving the contents of capsules (with starches and other excipients) in 0.3% carboxymethylcellulose sodium solution, which was same as the previous
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/bmc
1295
Acetonitrile was of HPLC grade (International Laboratory, USA). Triethylamine, formic acid (International Laboratory, USA) and perchloric acid (MERCK, Germany) were of GR grade. Deionized water was prepared using a Millipore water purification system (Billerica, MA, USA). Reference chemical standards of sinomenine, paeoniflorin, and paeonol were purchased from the National Institute for the Control of Pharmaceutical
An Agilent 1100 series LC system (Hewlett Packard, CA, USA) consisting of a G1311A Quaternary Pump, a G1322A degasser, a G1315A diode-array detector and a G1313A autosampler was employed. The chromatographic analysis for determination of SI, PF and PA was carried out on a Phenomenex ODS (250 × 4.6 mm i.d.; particle size 5 μm; Phenomenex Inc., USA) protected by a Security Guard Cartridge (C18, 4 × 3.0 mm i.d.; Phenomenex Inc., USA). The mobile phase was acetonitrile (A) and aqueous solution (B) (containing 0.1% formic acid, adjusted with triethylamine to pH 3.5 ± 0.2). The conditions of the solvent gradient elution were 8–20% (A) in 0–25 min, 20–40% (A) in 25–30 min, 40–70% (A) in 30–55 min, 70–90% (A) in 55–60 min at a flow-rate of 1.0 mL/min. Detection was conducted with different wavelengths of 240 nm for PF, 270 nm for SI and PA at room temperature (20°C). All injection volumes of sample solutions were 50 μL.
Y. Xie et al. pharmacological studies (Cai et al., 2005). Before experiments the animals were fasted for 24 h with water ad libitum, and maintained at a temperature of 21°C with 60% relative humidity and a 12 h light/dark cycle. After oral administration of QFGJS, jugular vein blood samples were collected (0.2 mL) into heparinized 1.5 mL microcentrifuge tubes at the following time intervals: 0, 5, 15, 30, 45, 60, 90, 120, 240, 360, 540 and 720 min. After each blood sampling, the catheter was gently flushed with the same volume of heparinized saline solution (20 units/mL) to replace the lost volume of blood; a plug was then inserted at the end of the catheter. The blood samples were immediately centrifuged at 12,000 rpm for 5min at room temperature (20°C), and the resulting plasma was collected and stored at 20°C until analysis.
Preparation of plasma samples In the current studies, two methods were used to prepare plasma samples for HPLC analysis. Method A. The resulting plasma (50 μL) was mixed with 20 μL of 7.0% aqueous perchloric acid solution and 20 μL IS (30.0μg/mL) by vortexing for 1 min. The mixture was then centrifuged at 12,000 rpm for 10 min to separate precipitated protein at room temperature (20°C). A 50 μL volume of the supernatant was injected into the HPLC system for analysis. This method was used for measurement of plasma concentration of SI and PF. Method B. The resulting plasma (50 μL) was mixed with 100 μL acetonitrile by vortex for 1 min. The mixture was centrifuged at 12,000 rpm for 10 min. A 50 μL volume of the supernatant was injected into the HPLC system for analysis. This method was employed for determining plasma concentration of PA.
Method validation Specificity. Six different blank rat plasma samples were analyzed to detect potential interferences co-eluting on the analytes and IS Chromatographic peaks of analytes and IS were identified on the basis of their retention times. Calibration curve. The stock solutions of SI, PF and the IS were prepared separately in deionised water to yield final concentrations of 116.2, 357.0 and 30.0 μg/mL, respectively. The stock solution of PA was prepared by dissolving PA with 50% acetonitrile, which was further diluted with water to obtain a final concentration of 251.5 μg/mL. All stock solutions were stored at 4°C, at which temperature they were found to remain stable for at least one month. Calibration curves were prepared by spiking the appropriate aliquots of SI, PF and PA stock solutions into drug-free plasma to give final concentrations of 0.06, 0.12, 0.23, 0.46, 1.16, 2.32, 4.65 and 11.62 μg/mL for SI, 0.09, 0.18, 0.36, 0.71, 1.43, 3.57, 7.14, 14.28 and 35.70 μg/mL for PF, and 0.15, 0.30, 0.76, 1.51, 2.26, 3.02 and 4.53 μg/mL for PA. The quality control (QC) samples were prepared in the same way as calibration standards with three concentrations of each analyte and used for the validation procedures and stability tests. The QC samples were divided into 100 μL aliquots and stored in 1.5 mL microcentrifuge tubes at 20°C until analysis. The concentrations of low, middle and high QC samples were for 0.23, 1.16 and 4.65 μg/mL for SI, 0.71, 3.57 and 14.27 μg/mL for PF, and 0.30, 1.51 and 3.02 μg/mL for PA. The linearity of the method was evaluated by calibration curves of SI, PF and PA, including the lower limit of quantification (LLOQ). A leastsquares linear regression analysis was performed to determine slope, 2 intercept and coefficient of correlation (r ). The limit of detection (LOD) was determined as the concentration giving a signal-to-noise ratio (S/N) of 3:1.
1296
Accuracy, precision and recovery. The intra-day accuracy and precision were determined by replicate analysis of six sets of QC samples at four concentration levels (LLOQ, low, middle, and upper concentrations) on the same day. For the inter-day variation, six replicates of QC samples at
wileyonlinelibrary.com/journal/bmc
three concentration levels were analyzed along with a standard curve on three different days. For acceptable intra-day and inter-day values, accuracy should be within 85–115% and the coefficient of variation (CV), which serves as a measure of precision, should be below 15%, except for the LLOQ, where the accuracy should be between 80 and 120%, while CV should not exceed 20%. The recovery of each analyte from plasma was evaluated using the QC samples at three different concentrations by comparing the peak areas of the extracted samples with those of unprocessed standard solutions containing the corresponding concentrations that represent 100% recovery. The recovery of IS from plasma was determined at a concentration of 30.0 μg/mL by the same method. Stability. Stability of each analyte in rat plasma was tested using QC samples at three different concentrations for three freeze–thaws, longterm, short-term and post-preparative stabilities. In each freeze–thaw cycle, the QC samples were frozen for about 24 h at 20°C and then thawed at room temperature (20°C). The long-term stability was evaluated after keeping the QC samples frozen at 20°C for 1 month. For the short-term stability, thawed plasma samples were kept at room temperature for 5 h before sample preparation. The post-preparative stability was tested after keeping the processed samples in HPLC autosampler vials at room temperature for 24 h. The stabilities of the samples of the three freeze–thaw cycles, the long-term, the short-term and the postpreparation were evaluated by comparing the contents of three analytes in the samples with the contents in the freshly prepared QC samples.
Pharmacokinetic analysis To quantify the different compounds in plasma samples, calibration curves and three QC samples were used with every set of 20 unknown samples. The pharmacokinetic parameters for each analyte were evaluated by analyzing the data of the time–plasma concentration profile, which was calculated by the pharmacokinetic software, PK Solutions 2.0 (Summit Co., USA) with noncompartment analysis. The following noncompartmental pharmacokinetic parameters were calculated based on the moment method: half-life (T1/2), mean residence time (MRT), volume of distribution (Vd), oral clearance (CL/F), and area under the concentration–time curve (AUC).
Results Optimization of the chromatographic conditions Each analyte and the internal standard were well separated with sharp peaks under a gradient elution program. By comparing both the retention times and the UV spectra of the reference standards, four compounds (SI, PF, PA and IS) in rat plasma in the pharmacokinetic studies of QFGJS were satisfactorily identified. The peak purity was confirmed by studying the DAD data with all peaks of interests, in which no indication of impurity of peaks was found. Furthermore, no interfering peak was observed around the retention times of those target compounds in the drug-free plasma samples. The HPLC chromatograms of the blank plasma, the plasma spiked with SI, PF, PA and IS, as well as the plasma obtained at 15 min after oral administration of QFGJS capsules (at dosage of 3.89 g/kg), are shown in Fig. 2. Calibration curve Linear regression analysis was performed by plotting the peak area ratios vs concentrations in the range of 0.06–11.62 μg/mL for SI, and 0.09–35.70μg/mL for PF with method A of the sample preparation. The regression equations of these calibration curves and their correlation coefficients (r2) were calculated as follows:
Copyright © 2014 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 2014; 28: 1294–1302
PK study of sinomenine, paeoniflorin and paeonol in a herbal product
Figure 2. Typical HPLC chromatograms of the determination of sinomenine, paeoniflorin and paeonol in rat plasma: (A, B) chromatograms of a blank plasma sample; (C, D) chromatograms of the plasma sample spiked with sinomenine, paeoniflorin, paeonol and pentoxifylline (internal standard); (E, F) chromatograms of the plasma sample from a rat 15 min after oral administration of Qingfu Guanjieshu (QFGJS) capsule. The retention time was 10.90 min for sinomenine, 20.05 min for paeoniflorin, 26.50 min for IS and 37.80 min for paeonol. (A, C, E) were obtained under the wavelength of 240 nm, while (B, D, F) were obtained under the wavelength of 270 nm.
SI, y = 3.7948x + 0.1221, r2 = 0.9999; PF, y = 4.3529x + 1.8456, r2 = 0.9996. Linearity for PA (method B of the sample preparation) was obtained over a range of 0.15–4.53 μg/mL with the calibration curve: y = 231.82x 0.0325, r2 = 0.9999 (where y is the peak area and x is the concentration of PA).
Accuracy and precision
Biomed. Chromatogr. 2014; 28: 1294–1302
Recovery The recoveries for determination of SI, PF and PA from rat plasma are shown in Table 2. The mean recovery values of each analyte using sample preparation method A were 102.34–107.75% for
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/bmc
1297
The LLOQ in rat plasma was 0.06 μg/mL for SI, 0.09 μg/mL for PF (method A), and 0.15 μg/mL for PA (method B), as these were the lowest concentrations assessed; the accuracies and precisions (CV) were 106.4 and 4.60% for SI, 86.3 and 8.56% for PF (method A), and 100.5 and 3.38% for PA (method B). The lower limits of detection were 0.03, 0.04 and 0.03 μg/mL for SI, PF, and PA, respectively, at a signal-to-noise ratio of 3:1.
The precision and accuracy of the intra-day and inter-day assay variations in plasma are shown in Table 1. The accuracy of the method was determined by calculating the percentage deviations observed in the analysis of QC samples. The results show that the intra-day accuracy varied between 96.80 and 109.18%, while the CV of the precision was
Revised: 6 January 2014,
Accepted: 17 January 2014
Published online in Wiley Online Library: 12 March 2014
(wileyonlinelibrary.com) DOI 10.1002/bmc.3165
The pharmacokinetic study of sinomenine, paeoniflorin and paeonol in rats after oral administration of a herbal product Qingfu Guanjiesu capsule by HPLC Ying Xiea, Zhi-Hong Jianga, Hua Zhoua, Wen-Zhe Maa, Yuen-Fan Wongb, Zhong-Qiu Liuc and Liang Liua* ABSTRACT: An accurate and reliable high-performance liquid chromatography–diode array detector (HPLC-DAD) method was developed and validated for determination of sinomenine (SI), paeoniflorin (PF) and paeonol (PA), which was further applied to assess the pharmacokinetics of SI, PF and PA in an anti-arthritic herbal product, Qingfu Guanjieshu (QFGJS) capsule, in rats. Successful separation was achieved with a C18 column and a mobile phase composed of acetonitrile and aqueous phase (containing 0.1% formic acid, adjusted with triethylamine to pH 3.5 ± 0.2). The method was validated with excellent precision, accuracy, recovery and stability in calibration ranges from 0.06 to 11.62 μg/mL for SI, from 0.09 to 35.70 μg/mL for PF, and from 0.15 to 4.53 μg/mL for PA (with r2 > 0.999 for all three compounds). Our results showed that absorption of PF after administration of QFGJS was similar to that after oral administration of PF alone; the absorption of SI was decreased while the absorption of PA was increased after giving QFGJS orally compared with pure compounds. We may conclude that pharmacokinetic studies of complex herbal products are not only necessary but also feasible by using representative bioactive chemicals as indicators of establishing quality control standards and of determining pharmacokinetic behavior of herbal medicines. Copyright © 2014 John Wiley & Sons, Ltd. Keywords: HPLC-DAD; pharmacokinetics; sinomenine; paeoniflorin; paeonol; herbal product
Introduction
1294
Although herbal medicinal products (HMPs) are used extensively throughout the world, especially in the treatment of chronic diseases, the relationship between clinical efficacy and the bioactive compounds contained in many HMPs remains unknown. Since knowledge of pharmacokinetics can help us to understand and predict a variety of events related to the efficacy and toxicity of HMPs, it is important to perform pharmacokinetic studies of their ingredients in the plasma when the complex mixtures are administered. Qingfu Guanjieshu (QFGJS) capsule, an anti-arthritic standardized herbal product, is formulated from five herbs – Caulis Sinomenii, Radix Aconiti Lateralis Preparata, Rhizoma Curcumae Longae, Radix Paeoniae Alba and Cortex Moutan – all of which have well-established records of usage for treatment of arthritic diseases in China, Japan and other Asian countries (Chen et al., 2004). Distinguishing it from most currently marked herbal products in China and even in the Western world, systemic quality control system has been established for QFGJS (Xie et al., 2007). Together with the optimal manufacturing process, intensive comparative studies on the extraction conditions, contents of the bioactive constituents and the effects of anti-inflammationa, anti-arthritis and antinociception in rats have warranted the safety and effectiveness of QFG (Xie et al., 2007; Cai et al., 2005; Zhou et al., 2006; Liang et al., 1990). It is currently in the pipeline of research and development to be approved as a novel new drug for arthritis treatment in China.
Biomed. Chromatogr. 2014; 28: 1294–1302
Such a herbal formula is quite chemically complex, with hundreds or even thousands of constituents, which makes pharmacokinetic study of the product very difficult. Nevertheless, even though the exact chemical profile and the nature and even the interactions of its constituents remain unknown, certain potent bioactive chemicals contained in QFGJS have been previously identified. Among these are: sinomenine (SI) in Caulis Sinomenii; paeoniflorin (PF) in Radix Paeoniae Alba; and paeonol (PA) in Cortex Moutan (Fig. 1). Previous pharmacological studies have shown that these compounds exhibit marked anti-inflammatory, analgesic, anti-arthritic and immunosuppressive activities (Zhuang
* Correspondence to: L. Liu, State Key Laboratory for Quality Research of Chinese Medicines, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau. Email: [email protected] a
State Key Laboratory for Quality Research of Chinese Medicines, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau
b
School of Chinese Medicine, Hong Kong Baptist University, Kowloon Tong, Hong Kong
c
School of Pharmacy, Southern Medical University, Guangzhou, China Abbreviations used: ACN, acetonitrile; AUC, area under the concentration–time curve; Cmax, maximum plasma concentration; CV, coefficient of variation; HMPs, herbal medicinal products; HPLC-DAD, high-performance liquid chromatography coupled with diode array detector; LLOQ, lower limit of quantification; PA, paeonol; PF, paeoniflorin; QFGJS, Qingfu Guanjieshu capsule; SI, sinomenine; Tmax, maximum plasma concentration
Copyright © 2014 John Wiley & Sons, Ltd.
PK study of sinomenine, paeoniflorin and paeonol in a herbal product and Biological Products, Beijing, China. Pentoxifylline, the internal standard (IS), was purchased from Sigma (St Louis, MO, USA). The identity and purity of these chemicals were further validated by LC-MS in our laboratory.
HPLC conditions
Figure 1. Chemical structures of sinomenine, paeoniflorin, paeonol, and pentoxifylline (internal standard).
et al., 2013; Liang et al., 1990; Ostrem et al., 2013; Wisnovsky Simon et al., 2013; Tsai et al., 2001). Thus, these three compounds were treated as the representative constituents of QFGJS for the primary pharmacokinetic study. Our previous studies showed that the pharmacokinetic (PK) parameters, particularly the oral bioavailability of PF (after a single dosage of 150 mg/kg), could be markedly enhanced by co-administration of SI at a dosage of 90 mg/kg in rats (Liu et al., 2005a). Moreover, many other reports have demonstrated significant variation of the pharmacokinetic parameters of PF and PA after oral administration, either when purified compounds were given or when a herbal product containing PF or PA was given (Sheng et al., 2004; Wang et al., 2006; Yang et al., 2006). This indicates that multiple chemicals in HMPs could alter the pharmacokinetics of each other and thus may influence safety and therapeutic efficacy. Therefore, it is important to evaluate the pharmacokinetic profiles of HMPs by monitoring the plasma concentrations of representative bioactive components. However, previous reported chromatographic methods for determination of a single representative compound cannot be used to determine all of them in bio-samples at the same time. In the present study, a new HPLC-DAD method was developed for simultaneous determination of SI, PF and PA in plasma and different sample preparation methods were compared with consideration of a small volume of plasma and efficiency of determination. The method was validated according to the guidance of the US Food and Drug Administration (2001). Using the developed method, we described the PK profile of SI, PF and PA after giving QFGJS orally as part of the preclinical evaluation of this drug, and studied the variation of between the PK parameters of SI, PF and PA in QFGJS and that of pure compounds. Although different bio-assay methods were used for studying PK behavior of PA, similar parameters were observed to the previous report (Xie et al., 2008), and the increased absorption of PA in QFGJS was further confirmed in the current study.
Experimental Chemicals and reagents
Biomed. Chromatogr. 2014; 28: 1294–1302
Animals and surgery Male Sprague–Dawley rats weighing 200–250 g were purchased from the Laboratory Animal Services Center of the Chinese University of Hong Kong. Animals were housed four per cage with food and water provided ad libitum and acclimated in the laboratory for at least one week prior to the experiment. Rats were anesthetized by intraperitoneal administration of 0.7% chloral hydrate saline solution at a dosage of 5 mL/kg body weight (Huang et al., 2004). A longitudinal skin incision was made over the area where the right external jugular vein passes dorsal to the pectoral major muscle. The catheter (Silastic catalog no. 602-155; Dow Corning, Midland, MI, USA), filled with 20 units/mL heparinized saline solution, was put into the right jugular vein and then advanced into the sinus venosus. The catheter was inserted up to the first silicone stopper and anchored in place by suturing the stopper to muscle. The free end of the catheter was passed under the skin of the dorsum of the neck just caudal to the ears and attached to the skin, together with a metal spring, which was covered with PVC tubing for protection of the outer part of the catheter. Finally, the catheter was filled with 500 units/mL heparinized saline solution, and a plug was inserted in the free end of the catheter. After surgery, rats were housed individually in metabolite cages for 5 days of recovery and underwent pharmacokinetic treatment according to the jugular-catheterized rat model (Thrivikraman et al., 2002). All procedures involving animals and their care were approved and performed under the regulations of the Committee on Use of Human and Animal Subjects in Teaching and Research of Hong Kong Baptist University and the Department of the Health Department of Hong Kong Special Administration Region (permit number 09-35 in DH/HA&P/8/2/6 Part 1).
Drug administration and blood sampling QFGJS capsules were prepared from five herbs with pharmaceutical methods described in our previous reports (Zhou et al., 2006). The chemical consistency and stability of the QFGJS product were examined by quantitative determination of five marker compounds (SI, PF, PA, curcumin and hypaconitine) and by qualitative fingerprint analysis, which has been previously described (Xie et al., 2007). Contents of SI, PF, and PA in QFGJS capsules were determined to be 5.9, 25.3 and 19.0 mg/g, respectively. The rats were randomly divided into two groups with eight rats in each group. The oral administration of QFGJS was divided into two dosages of 3.89 and 0.97 g/20 mL/kg body weight prepared by dissolving the contents of capsules (with starches and other excipients) in 0.3% carboxymethylcellulose sodium solution, which was same as the previous
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/bmc
1295
Acetonitrile was of HPLC grade (International Laboratory, USA). Triethylamine, formic acid (International Laboratory, USA) and perchloric acid (MERCK, Germany) were of GR grade. Deionized water was prepared using a Millipore water purification system (Billerica, MA, USA). Reference chemical standards of sinomenine, paeoniflorin, and paeonol were purchased from the National Institute for the Control of Pharmaceutical
An Agilent 1100 series LC system (Hewlett Packard, CA, USA) consisting of a G1311A Quaternary Pump, a G1322A degasser, a G1315A diode-array detector and a G1313A autosampler was employed. The chromatographic analysis for determination of SI, PF and PA was carried out on a Phenomenex ODS (250 × 4.6 mm i.d.; particle size 5 μm; Phenomenex Inc., USA) protected by a Security Guard Cartridge (C18, 4 × 3.0 mm i.d.; Phenomenex Inc., USA). The mobile phase was acetonitrile (A) and aqueous solution (B) (containing 0.1% formic acid, adjusted with triethylamine to pH 3.5 ± 0.2). The conditions of the solvent gradient elution were 8–20% (A) in 0–25 min, 20–40% (A) in 25–30 min, 40–70% (A) in 30–55 min, 70–90% (A) in 55–60 min at a flow-rate of 1.0 mL/min. Detection was conducted with different wavelengths of 240 nm for PF, 270 nm for SI and PA at room temperature (20°C). All injection volumes of sample solutions were 50 μL.
Y. Xie et al. pharmacological studies (Cai et al., 2005). Before experiments the animals were fasted for 24 h with water ad libitum, and maintained at a temperature of 21°C with 60% relative humidity and a 12 h light/dark cycle. After oral administration of QFGJS, jugular vein blood samples were collected (0.2 mL) into heparinized 1.5 mL microcentrifuge tubes at the following time intervals: 0, 5, 15, 30, 45, 60, 90, 120, 240, 360, 540 and 720 min. After each blood sampling, the catheter was gently flushed with the same volume of heparinized saline solution (20 units/mL) to replace the lost volume of blood; a plug was then inserted at the end of the catheter. The blood samples were immediately centrifuged at 12,000 rpm for 5min at room temperature (20°C), and the resulting plasma was collected and stored at 20°C until analysis.
Preparation of plasma samples In the current studies, two methods were used to prepare plasma samples for HPLC analysis. Method A. The resulting plasma (50 μL) was mixed with 20 μL of 7.0% aqueous perchloric acid solution and 20 μL IS (30.0μg/mL) by vortexing for 1 min. The mixture was then centrifuged at 12,000 rpm for 10 min to separate precipitated protein at room temperature (20°C). A 50 μL volume of the supernatant was injected into the HPLC system for analysis. This method was used for measurement of plasma concentration of SI and PF. Method B. The resulting plasma (50 μL) was mixed with 100 μL acetonitrile by vortex for 1 min. The mixture was centrifuged at 12,000 rpm for 10 min. A 50 μL volume of the supernatant was injected into the HPLC system for analysis. This method was employed for determining plasma concentration of PA.
Method validation Specificity. Six different blank rat plasma samples were analyzed to detect potential interferences co-eluting on the analytes and IS Chromatographic peaks of analytes and IS were identified on the basis of their retention times. Calibration curve. The stock solutions of SI, PF and the IS were prepared separately in deionised water to yield final concentrations of 116.2, 357.0 and 30.0 μg/mL, respectively. The stock solution of PA was prepared by dissolving PA with 50% acetonitrile, which was further diluted with water to obtain a final concentration of 251.5 μg/mL. All stock solutions were stored at 4°C, at which temperature they were found to remain stable for at least one month. Calibration curves were prepared by spiking the appropriate aliquots of SI, PF and PA stock solutions into drug-free plasma to give final concentrations of 0.06, 0.12, 0.23, 0.46, 1.16, 2.32, 4.65 and 11.62 μg/mL for SI, 0.09, 0.18, 0.36, 0.71, 1.43, 3.57, 7.14, 14.28 and 35.70 μg/mL for PF, and 0.15, 0.30, 0.76, 1.51, 2.26, 3.02 and 4.53 μg/mL for PA. The quality control (QC) samples were prepared in the same way as calibration standards with three concentrations of each analyte and used for the validation procedures and stability tests. The QC samples were divided into 100 μL aliquots and stored in 1.5 mL microcentrifuge tubes at 20°C until analysis. The concentrations of low, middle and high QC samples were for 0.23, 1.16 and 4.65 μg/mL for SI, 0.71, 3.57 and 14.27 μg/mL for PF, and 0.30, 1.51 and 3.02 μg/mL for PA. The linearity of the method was evaluated by calibration curves of SI, PF and PA, including the lower limit of quantification (LLOQ). A leastsquares linear regression analysis was performed to determine slope, 2 intercept and coefficient of correlation (r ). The limit of detection (LOD) was determined as the concentration giving a signal-to-noise ratio (S/N) of 3:1.
1296
Accuracy, precision and recovery. The intra-day accuracy and precision were determined by replicate analysis of six sets of QC samples at four concentration levels (LLOQ, low, middle, and upper concentrations) on the same day. For the inter-day variation, six replicates of QC samples at
wileyonlinelibrary.com/journal/bmc
three concentration levels were analyzed along with a standard curve on three different days. For acceptable intra-day and inter-day values, accuracy should be within 85–115% and the coefficient of variation (CV), which serves as a measure of precision, should be below 15%, except for the LLOQ, where the accuracy should be between 80 and 120%, while CV should not exceed 20%. The recovery of each analyte from plasma was evaluated using the QC samples at three different concentrations by comparing the peak areas of the extracted samples with those of unprocessed standard solutions containing the corresponding concentrations that represent 100% recovery. The recovery of IS from plasma was determined at a concentration of 30.0 μg/mL by the same method. Stability. Stability of each analyte in rat plasma was tested using QC samples at three different concentrations for three freeze–thaws, longterm, short-term and post-preparative stabilities. In each freeze–thaw cycle, the QC samples were frozen for about 24 h at 20°C and then thawed at room temperature (20°C). The long-term stability was evaluated after keeping the QC samples frozen at 20°C for 1 month. For the short-term stability, thawed plasma samples were kept at room temperature for 5 h before sample preparation. The post-preparative stability was tested after keeping the processed samples in HPLC autosampler vials at room temperature for 24 h. The stabilities of the samples of the three freeze–thaw cycles, the long-term, the short-term and the postpreparation were evaluated by comparing the contents of three analytes in the samples with the contents in the freshly prepared QC samples.
Pharmacokinetic analysis To quantify the different compounds in plasma samples, calibration curves and three QC samples were used with every set of 20 unknown samples. The pharmacokinetic parameters for each analyte were evaluated by analyzing the data of the time–plasma concentration profile, which was calculated by the pharmacokinetic software, PK Solutions 2.0 (Summit Co., USA) with noncompartment analysis. The following noncompartmental pharmacokinetic parameters were calculated based on the moment method: half-life (T1/2), mean residence time (MRT), volume of distribution (Vd), oral clearance (CL/F), and area under the concentration–time curve (AUC).
Results Optimization of the chromatographic conditions Each analyte and the internal standard were well separated with sharp peaks under a gradient elution program. By comparing both the retention times and the UV spectra of the reference standards, four compounds (SI, PF, PA and IS) in rat plasma in the pharmacokinetic studies of QFGJS were satisfactorily identified. The peak purity was confirmed by studying the DAD data with all peaks of interests, in which no indication of impurity of peaks was found. Furthermore, no interfering peak was observed around the retention times of those target compounds in the drug-free plasma samples. The HPLC chromatograms of the blank plasma, the plasma spiked with SI, PF, PA and IS, as well as the plasma obtained at 15 min after oral administration of QFGJS capsules (at dosage of 3.89 g/kg), are shown in Fig. 2. Calibration curve Linear regression analysis was performed by plotting the peak area ratios vs concentrations in the range of 0.06–11.62 μg/mL for SI, and 0.09–35.70μg/mL for PF with method A of the sample preparation. The regression equations of these calibration curves and their correlation coefficients (r2) were calculated as follows:
Copyright © 2014 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 2014; 28: 1294–1302
PK study of sinomenine, paeoniflorin and paeonol in a herbal product
Figure 2. Typical HPLC chromatograms of the determination of sinomenine, paeoniflorin and paeonol in rat plasma: (A, B) chromatograms of a blank plasma sample; (C, D) chromatograms of the plasma sample spiked with sinomenine, paeoniflorin, paeonol and pentoxifylline (internal standard); (E, F) chromatograms of the plasma sample from a rat 15 min after oral administration of Qingfu Guanjieshu (QFGJS) capsule. The retention time was 10.90 min for sinomenine, 20.05 min for paeoniflorin, 26.50 min for IS and 37.80 min for paeonol. (A, C, E) were obtained under the wavelength of 240 nm, while (B, D, F) were obtained under the wavelength of 270 nm.
SI, y = 3.7948x + 0.1221, r2 = 0.9999; PF, y = 4.3529x + 1.8456, r2 = 0.9996. Linearity for PA (method B of the sample preparation) was obtained over a range of 0.15–4.53 μg/mL with the calibration curve: y = 231.82x 0.0325, r2 = 0.9999 (where y is the peak area and x is the concentration of PA).
Accuracy and precision
Biomed. Chromatogr. 2014; 28: 1294–1302
Recovery The recoveries for determination of SI, PF and PA from rat plasma are shown in Table 2. The mean recovery values of each analyte using sample preparation method A were 102.34–107.75% for
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/bmc
1297
The LLOQ in rat plasma was 0.06 μg/mL for SI, 0.09 μg/mL for PF (method A), and 0.15 μg/mL for PA (method B), as these were the lowest concentrations assessed; the accuracies and precisions (CV) were 106.4 and 4.60% for SI, 86.3 and 8.56% for PF (method A), and 100.5 and 3.38% for PA (method B). The lower limits of detection were 0.03, 0.04 and 0.03 μg/mL for SI, PF, and PA, respectively, at a signal-to-noise ratio of 3:1.
The precision and accuracy of the intra-day and inter-day assay variations in plasma are shown in Table 1. The accuracy of the method was determined by calculating the percentage deviations observed in the analysis of QC samples. The results show that the intra-day accuracy varied between 96.80 and 109.18%, while the CV of the precision was
