Journal of Chromatography B, 974 (2015) 131–137
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Measurement of hydroxysafflor yellow A in human urine by liquid chromatography–tandem mass spectrometry Chang-Yin Li 1 , Ji-Hong Chu 1 , Jun Zhang, Bing-Ting Sun, Guo-Liang Dai, Shi-Jia Liu, Wen-Zheng Ju ∗ Department of Clinical Pharmacology, Affiliated Hospital of Nanjing University of Chinese Medicine, No. 155 Hanzhong Road, Nanjing 210029, China
a r t i c l e
i n f o
Article history: Received 7 July 2014 Received in revised form 20 September 2014 Accepted 26 October 2014 Available online 4 November 2014 Keywords: Hydroxysafflor yellow A SPE LC–MS/MS Human urine Excretion Column switching
a b s t r a c t A rapid and specific high performance liquid chromatography–tandem mass spectrometry (LC–MS/MS) was developed for the quantification of hydroxysafflor yellow A (HSYA) in human urine with isorhamnetin-3-O-neohespeidoside as internal standard (IS). HSYA and IS were extracted from urine samples by simple solid-phase extraction and separated on an Agilent Zorbax SB C18 column (4.6 mm × 150 mm, 5 m) with the mobile phase of 0.2 mM ammonium acetate: methanol (30/70, v/v) at a flow rate of 0.4 mL/min. Polar endogenous interferences eluted in 0.1–2.5 min were switched into waste channel by the Valve Valco, to reduce the possible matrix effect for MS detection in each run. The MS detection of analytes was performed on a tandem mass spectrometer equipped with an electrospray ionization source in negative mode using multiple-reaction monitoring. The MS/MS ion transitions monitored were m/z 611.3 → 491.2 for HSYA and m/z 623.2 → 299.2 for IS. The method was fully validated for selectivity, sensitivity, linearity, precision, accuracy, recovery, matrix effect and stability, and then was applied to the urinary excretion study of injectable powder of pure HSYA in healthy Chinese volunteers for the first time. The results suggested that urine was the main excretion way of HSYA in healthy volunteers, further demonstrating the feasibility and necessity of our current method. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The safflower plant, namely Carthamus tinctorius L. is a kind of therophyte belonging to the Dicotyledoneae subclass, Compositae family. The dried flower of this plant, named as Honghua in traditional Chinese medicine, has been widely used to treat coronary heart disease, hypertension, and cerebrovascular disease in clinical practice of China [1]. Over the past decades, hydroxysafflor yellow A (HSYA, Fig. 1a), the main chemical component of Honghua, has attracted much attention because of its various attractive pharmacological activities such as cardiovascular protection [2], improvement of learning and memory [3], attenuation of hepatic fibrosis [4,5], anti-inflammation [6–8] and neuroprotection [9–14]. Accordingly, HSYA was chosen as an active marker component for controlling the quality of safflower in Chinese Pharmacopoeia [1]. In recent years, several HSYA-containing preparations have been
∗ Corresponding author. Tel.: +86 25 8661 7141x50523; fax: +86 25 8655 5033. E-mail addresses: [email protected] (C.-Y. Li), [email protected] (W.-Z. Ju). 1 The first two authors contributed equally to this work. http://dx.doi.org/10.1016/j.jchromb.2014.10.036 1570-0232/© 2014 Elsevier B.V. All rights reserved.
approved for clinical research by China Food and Drug Administration (CFDA). Considering that better understanding the in-vivo metabolism of HSYA is an indispensable part in clinical study of HSYA-related drugs, it is necessary to develop simple and reliable method for analysis and determination of HSYA in human biological matrices such as plasma, urine and feces. To date, both LC–UV [15,16] and LC–MS/MS [17] methods have been reported for measuring HSYA in human plasma. However, these methods were not very suitable for the urinary excretion study of HSYA owing to the following two reasons: first, urine is a biological matrix quite different from plasma, and it contains various polar endogenous compounds which may result in remarkable matrix effect; and second, drug concentration in urine is also very different from that in plasma. Considering that the urinary excretion study of HSYA is an indispensible step for the research and development of new drug, several LC–UV methods [18–20] has been applied for the determination of HSYA concentration in rat urine previously. However, the disadvantages of LC technology, such as long analysis time, complex sample preparation and unsatisfied sensitivity and specificity, made these LC methods not very suitable for human urinary excretion study of HSYA-containing preparations. Given that LC–MS/MS technique
132
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
Fig. 1. The chemical structure of HSYA (a) and internal standard SLS (b).
is currently considered as the method of first choice for supporting clinical drug metabolism studies, recently a UPLC–ESI–MS/MS method was employed for the determination of HSYA in human urine [21]. Compared with the above LC methods, this UPLC–MS/MS method offers a rapid analysis of HSYA in human urine with excellent sensitivity and specificity. Despite this, the absence of internal standard and evaluation for matrix effect [22], two indispensible parts for a complete LC–MS bioanalytical method, made this method less accurate and robust to some extent. Furthermore, the complex sample preparation procedures including evaporation to dryness and reconstitution steps are unavoidable, hindering high throughput analysis for large numbers of samples in clinical studies. Therefore, it is still in great need to develop a more convenient, sensitive, accurate and high-throughput LC–MS/MS method to quantify the concentration of HSYA in human urine. As an adaptation of the previously published LC–MS/MS method for the determination of HSYA in human plasma [17], the purpose of the present study was to develop and validate a simple, rapid, accurate and specific LC–MS/MS method for the quantification of HSYA in human urine. Column switching technique, simple sample preparation and stable internal standard were applied into our current LC–MS/MS method, to make it more suitable for the human urinary excretion study of HSYA.
Agilent ZORAX SB-C8 guard column (2.1 × 12.5 mm, 5 m). The system was eluted in isocratic mode with a mobile phase consisting of 0.2 mM ammonium acetate: methanol (30/70, v/v) at a flow rate of 0.4 mL/min. The column oven temperature was maintained at 30 ◦ C and the total LC run time was 5.0 min. Volumes of 10 L were injected using an autosampler thermostatted at 8 ◦ C. The eluant was switched into waste channel for 0.1–2.5 min and MS channel for 2.5–5.0 min by the Valve Valco. An API4000 tandem quadrupole mass spectrometer equipped with electrospray ionization (ESI) source (Applied Biosystems, USA) operating in the negative ion mode was used as a detector. The MS/MS system was operated at unit resolution in the multiplereaction monitoring (MRM) mode, using precursor to product ion transition of m/z 611.3 → 491.2 for HSYA, m/z 623.2 → 299.2 for SLS (IS). The major ESI source parameters were kept as follows: ion spray voltage (ISV), −4500 V; turbo heater temperature (TEM), 400 ◦ C; collision activation dissociation (CAD), 10 psi, curtain gas (CUR), 25 psi. While the main compound dependent parameters, declustering potential (DP) and collision energy (CE), were optimized at −125 V and −34 eV for HSYA, −150 V and −62 eV for IS, respectively. Dwell time set for both HSYA and IS was 150 ms. All the operations, acquisition and data analysis were controlled by Analyst software version 1.6.0.
2. Experimental
2.3. Preparation of stock solutions, calibration curve and quality control samples
2.1. Material and reagents Reference standards of HSYA (purity: 92.5%, lot 111637– 201106) and the internal standard (IS) isorhamnetin-3-Oneohespeidoside (SLS, purity: 93.2%, lot 111571–201205, Fig. 1b) were purchased from National Institutes for Food and Drug Control (Beijing, PR China). Injectable powder of pure HSYA (IPPH) was provided by Guangzhou Yuekang biopharmaceutical Co., Ltd. (Guangzhou, China). Methanol and ammonium acetate were of HPLC grade (Merck, USA). Other chemicals were of analytical reagent grade. Ultrapure water used for the LC–MS/MS was produced using the Milli-Q water purification system (Millipore, USA). Blank urine used in this study was obtained from healthy volunteers.
The standard stock solutions of HSYA were prepared in methanol at the concentration of 1.05 mg/mL for calibration curve, 1.07 mg/mL for quality control (QC). A series of HSYA working solutions for calibration curve (1050, 349, 116, 38.8, 12.9, 4.31, 1.44 and 0.48 g/mL) and QC (855, 8.55, 0.95 g/mL) were prepared by stepwise dilution of the stock solution with methanol. The IS working solution (1.99 g/mL) was obtained by diluting the stock solution of SLS (0.99 mg/mL) with methanol. All the above solutions were stored at 4 ◦ C until use. Blank urine was stored in the freezer and allowed to completely thaw before use. The calibration curve and quality control samples were prepared by spiking 20 L of working solutions into 100 L of blank human urine. The spiked urine samples were stored at −20 ◦ C for method validation.
2.2. Instrumentation and conditions
2.4. Sample extraction procedures
LC separation was performed on a series 1200 HPLC system (Agilent, USA) equipped with a binary pump (1260G1312A), a column Oven (1260G1316A), an auto sampler (1260G1367E) and a Valve Valco 2-Position. HSYA and IS were separated on an Agilent Zorbax SB C18 column (4.6 mm × 150 mm, 5 m), preceded with an
Solid-phase extraction (SPE) of HSYA from human urine was performed using Waters (Michigan, USA) Oasis® HLB 3cc (60 mg) Extraction Cartridges. Before use, the cartridges should be sequentially conditioned with 3 mL of methanol and followed by 3 mL of water.
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
133
After being diluted by water in the volume proportion of 1:9, a 100 L aliquot of the human dilution urine was mixed with 20 L of IS working solution and 400 L of ammonium acetate (0.2 mM) water solution. The mixture was vortexed for 30 s and subsequently centrifuged at 12,000 × g for 5 min at 4 ◦ C. A 480 L aliquot of the supernatant was then loaded onto the conditioned cartridges, which was followed by being washed with 2 mL of water. Analytes were finally eluted with 1 mL of 70% methanol/water (v/v) solution and a 5 L aliquot of the eluant was injected into the LC/MS/MS system for analysis.
Freeze–thaw stability was conducted by testing spiked samples following three freeze–thaw cycles (−70 to 25 ◦ C, at least 12 h time duration between cycles). All stability evaluations were based on back-calculated concentrations. Samples were considered to be stable if assay values were within the acceptable limits of accuracy (85–115% of nominal value) and precision (less than 15%). Besides, the stability of the methanol stock solutions of HSYA at 4 ◦ C was also evaluated for as long as 290 days using the concentration level of approximately 3.0 g/mL.
2.5. Method validation
2.6. Application to clinical samples
According to the Food Drug and Administration (FDA) guidance on bio-analytical method validation [23], the method was validated for selectivity and sensitivity, linearity, precision, accuracy, recovery, matrix effect, and stability. Selectivity was proved by comparing chromatograms of six independent sources of blank human urine with a urine sample spiked with HSYA and IS. Chromatographic peaks of analyte and IS were identified on the basis of their MRM responses and retention times, and there should be no significant interference at the retention time of analyte and IS in the chromatograms of blank urine samples. Sensitivity was determined from the signal-to-noise ratio (S/N) of the analyte response in six replicates of the lowest calibration standard sample, and the S/N ratio should be no less than 10 for the lower limit of quantification (LLOQ), with acceptable precision (less than 20%) and accuracy (80–120%). Linearity was evaluated over five working days in the concentration range of 0.10–210 g/mL, using eight calibration standard samples prepared as described in Section 2.3. Calibration curves were created by plotting the LC–MS/MS peak area ratio of HSYA to IS against the nominal concentration of HSYA in urine sample, with 1/x2 weighted regression. The coefficient of correlation (r2 ) should be more than 0.99, and the accuracy for each standard point had to be within 85–115% of the theoretical concentration, except for 80–120% at LLOQ point. The intra-batch and inter-batch precision and accuracy were checked on three consecutive days by measuring HSYA concentration in five replicates at three QC levels at the nominal concentrations of 0.19, 1.71, 171 g/mL. The precision at each concentration was reported as the relative standard deviation (RSD%), expressing the standard deviation (SD) as a percentage of the mean calculated concentration, while the accuracy was determined by depicting the mean calculated concentration as a percentage of the nominal concentration. The criteria for data acceptance are less than 15% for precision and within 85–115% for accuracy. The extraction recovery of HSYA was determined in five replicates, by comparing the peak area of the analyte extracted from the above three levels of QC urine samples, with the peak area of the extracted matrix added with the same amount of the analyte. Extraction recovery of IS was calculated by the same way at a concentration of 0.04 g/mL. The recovery was deemed acceptable if the RSD% value of peak area ratio was less than 15% of mean recoveries. Matrix effect was evaluated on six independent sources of blank human urine, by calculating the ratio of the analyte peak areas in blank matrix spiked post-SPE to that in methanol solution at three QC levels of HSYA (0.19, 1.71, 171 g/mL). If the peak area ratio is less than 85% or more than 115%, a matrix effect is implied and should not be overlooked. Fresh stock solutions of HSYA and IS were used for the urine stability studies. The stability of HSYA was tested under various conditions using three levels of QC samples. Bench-top (short-term) stability of spiked human urine was evaluated for 4 h at room temperature (25 ◦ C) and 24 h at 8 ◦ C. Long-term stability was examined by assaying spiked urine samples stored at −70 ◦ C for 13 days.
After being fully validated, this LC–MS/MS method was then applied to the human urinary excretion study of HSYA following drip intravenous infusion of IPPH at a single dose of 75 mg. Three healthy male Chinese volunteers, aged from 22 to 28, with Body Mass Index (BMI) from 19 to 24, were recruited for this study at the Clinical Pharmacology Laboratory, They were all informed of the aim and risks involved in the study and then signed an informed consent. The experiment protocols were approved by the Ethical Review Committee of the Affiliated Hospital of Nanjing University of Chinese Medicine, in accordance with the principles of the Declaration of Helsinki, and the recommendations of China Food and Drug Administration. After an overnight fasting, volunteers received a single drip intravenous infusion dose of 75 mg injectable HSYA powder over a period of approximately 1 h. Serial urine samples were collected prior to dosing and over the intervals of 0–4, 4–8, 8–12, 2–24, 24–48 h post-dosing. The exact total volume of urine samples was recorded and 15 mL of each sample was finally retained and stored at −70 ◦ C until use.
3. Results and discussion 3.1. Method development All steps of the method, including sample preparation, selection of IS, chromatographic conditions and mass parameters were systematically optimized in the process of method development. Protein precipitation, liquid–liquid extraction and SPE extraction were three frequently used methods for sample preparation, and the former two methods are relatively convenient to operate and therefore very suitable for high throughput analysis. However, our initial analysis indicated that liquid–liquid extraction of urine samples yielded poor extraction recovery for HSYA (as low as approximately 1%); protein precipitation by methanol and acetonitrile obtained better extraction recovery, but suffered from endogenous interference and irregular chromatographic peaks for HSYA and IS. While Using waters Oasis@ HLB 3cc (60 mg) Extraction Cartridges and 70% methanol/water solution (as the eluent), SPE extraction could achieve excellent extraction efficiency of HSYA and IS, with no remarkable endogenous interference. Therefore, SPE extraction was finally selected for sample preparation, which was consistent with the previous report for plasma samples [17]. Unlike the above previous method, before SPE extraction, 400 L aliquot of 0.2 M ammonium acetate should be added into the urine samples to improve the extraction recovery of HSYA. Furthermore, aided by the optimization of LC–MS conditions and SPE method, SPE extraction was further simplified by the omission of the steps of dryness and reconstitution. Puerarin, a flavanoid compound used as IS in the previous study [17], proved not to be stable under the current conditions during the whole analytical process. Instead, SLS, another flavanoid compound that possessed similar molecular weight, chromatographic behavior and ESI response with HSYA, was finally selected as the
134
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
Fig. 2. Typical MRM chromatograms of (A) blank human urine; (B) HSYA spiked urine sample (0.10 g/mL, marked concentration in urine sample) for LLOQ; (C) methanol solution of standard mixture for HSYA (0.017 g/mL, final concentration for analysis) and IS (0.04 g/mL); (D) HSYA spiked urine sample (1.71 g/mL) for quality control; (E) urine sample obtained from a volunteer at the interval of 24–48 h after drip intravenous infusion of HSYA injection containing 75 mg of HSYA.
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
135
Fig. 3. Product ion mass spectra of [M–H]− of HSYA (a) and internal standard SLS (b).
Table 1 Intra- and inter-batch precision and accuracy for determination of HSYA in human urine. (Data are based on assay of five replicates per day, on three consecutive days). Spiked concentration (g/mL)
0.19 1.71 171
Intra-batch (n = 5)
Inter-batch (n = 15)
Mean (g/mL)
SD
RSD (%)
Accuracy (%)
Mean (g/mL)
SD
RSD (%)
Accuracy (%)
0.18 1.73 164
0.01 0.04 8.68
7.95 2.23 5.29
96.56 101.4 96.11
0.19 1.7 164
0.02 0.09 9.11
8.87 5.18 5.55
101.1 99.28 96.13
internal standard in our study. The extraction recovery of SLS was satisfactory and it was stable during the whole analytical process. Chromatographic conditions, especially mobile phase compositions, types of analytical column, were evaluated and optimized to obtain good chromatographic separation and ionization. As shown in Fig. 2, the final LC condition as described in Section 2.2 could bring the symmetrical peak shapes, good resolution and short retention times (less than 5 min) for both HSYA and IS. It should be noted that the addition of 0.2 mM ammonium acetate in mobile phase was an indispensible step for keeping the regular peak shape of HSYA. The mass spectrometric conditions were also optimized for HSYA and IS to obtain the maximal sensitivity, and the typical product ion mass spectra of [M–H]− of HSYA and IS are shown in Fig. 3. As for LC–MS analysis of urine samples, it should be noted that the abundant salts and polar endogenous polar compounds in urine may bring significant ion suppression or enhancement for
mass detection, even though after SPE extraction, while column switching technique could effectively minimize the salt effects or matrix effects [22]. Therefore, in the current study, the eluant in 0.1–2.5 min for each LC run was switched into waste channel, not MS channel, by a Valve Valco 2-Position. 3.2. Method validation 3.2.1. Selectivity and sensitivity Representative MRM chromatograms of blank urine, methanol solution of standards, blank urine spiked with HSYA and IS and urine sample from a volunteer after intravenous administration are shown in Fig. 2. The retention times for HSYA and IS were approximately 2.8 and 4.1 min, respectively. There was no apparent interference in urine matrix at the retention times of HSYA and I.S in the corresponding ion channel, demonstrating that the analysis
Table 2 Stability of HSYA in human urine (n = 3). Spiked concentration (g/mL)
0.19
1.71
171
Short-term stability (4 h, room temperature)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 5.14 106.7
1.56 0.07 4.16 91.42
158 2.1 1.33 92.36
Freeze and thaw stability (cycle 1)
Mean (g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 5.52 105.6
1.81 0.13 7.4 105.9
165 1.48 0.9 96.71
Freeze and thaw stability (cycle 2)
Mean (g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 6.71 102.5
1.84 0.08 4.51 107.5
163 3.15 1.93 95.36
Freeze and thaw stability (cycle 3)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.19 0.01 6.98 102.3
1.66 0.06 3.54 97.05
168 6.95 4.13 98.51
Long-term stability (13 days, −70 ◦ C)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 3.11 106.6
1.64 0.12 7.32 96.13
167 4.36 2.61 97.66
Post-preparative stability (24 h, 8 ◦ C)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 5.97 106.1
1.68 0.04 2.3 98.33
168 6.07 3.61 98.37
136
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
Table 3 Concentration of HSYA in human urine samples for three male volunteers following drip intravenous infusion of injectable HSYA powder at a single dose of 75 mg. Time interval (h) Subject A
Subject B
Subject C
Concentration (g/mL) Volume (mL) Excretion amount (mg) Concentration (g/mL) Volume (mL) Excretion amount (mg) Concentration (g/mL) Volume (mL) Excretion amount (mg)
0–4
4–8
8–12
12–24
24–48
33.79 480 16.22 33.79 510 17.23 181.7 150 27.26
8.99 850 7.64 19.3 400 7.72 30.5 400 12.20
7.93 280 2.22 6.15 650 4.00 6.80 340 2.31
17.14 230 3.94 2.01 1100 2.21 8.74 790 6.90
1.33 1100 1.46 0.41 2550 1.04 0.31 2630 0.82
TU (mg)
TU%
31.48
41.97
32.19
42.92
49.49
65.98
TU: total urinary excretion amount of HSYA in 48 h; TU%: total urinary excretion proportion of HSYA in 48 h.
of HSYA and IS using the MRM function was highly selective in the current study. The LLOQ of HSYA was 0.10 g/mL in human urine, and six replicate analyses of LLOQ samples showed excellent accuracy of 99.80% and satisfied precision with RSD% value of 2.07%
3.2.2. Calibration linearity The calibration curves showed good linearity over five working days in the concentration range of 0.10–210 g/mL. The equation of linearity was y = (0.217 ± 0.011)x + (0.0129 ± 0.0048), with the coefficient of correlation (r2 ) ranging from 0.993 to 0.998, while the accuracy for all calibration concentrations was within 85–115% of the theoretical values, with RSD% ranging from 1.66 to 7.61.
3.2.3. Accuracy and precision The intra-batch and inter-batch precision and accuracy of HSYA are summarized in Table 1. The RSD% values were no more than 8.87% for both intra-batch and inter-batch precision, while the accuracy varied from 96.11% to 101.1%. All of the values well meet the criteria for data acceptance in the guidance [21], indicating that the present LC–MS/MS method was accurate, reliable and reproducible for the quantitative analysis of HSYA in human urine samples.
3.2.4. Extraction recovery and matrix effect The average extraction recovery for HSYA were 72.65%, 72.81% and 78.52% at low, medium and high QC levels respectively, with RSD% values of 4.08%, 2.51%, and 2.10%, while the average recovery of IS was 97.14%, with RSD% value of 4.10%. The results demonstrated that the extraction recovery of our method was acceptable and stable for bio-analytical analysis. The average matrix effect for HSYA varied from 102% to 110.1% at three QC levels, with RSD% values ranged from 0.50% to 6.74%. The average matrix effect value for IS was 87.96%. These data suggested that no significant ion suppression or enhancement was observed under our experimental conditions.
3.2.5. Stability The stability data of HSYA kept under various storage conditions were summarized in Table 2. Urine samples were proved to be stable for 4 h at room temperature, for 13 days when stored at −70 ◦ C and through three freeze–thaw cycles. Samples after treatment were also stable at 8 ◦ C in auto-sampler for 24 h. In addition, after 290 days’ storage at 4 ◦ C, the methanol solutions of HSYA kept stable, as the accuracy normalized to freshly prepared solution was 96.25% with the RSD value of 2.02% (n = 3). These results indicated that HSYA was stable under the different storage and temperature conditions and allowed stockpiling of urine samples obtained in the excretion studies for subsequent batch analysis.
3.3. Urinary excretion study The developed LC–MS/MS method was then applied to the human urinary excretion study of HSYA following drip intravenous infusion of IPPH at a single dose of 75 mg. Table 3 summarized Concentration of HSYA in human urine samples for three male volunteers following drip intravenous infusion of injectable HSYA powder. The results indicated that the linearity range of our method (0.10–210 g/mL) covered well with the concentrations of all tested human urine samples (0.31–182 g/mL), without any need for dilution or concentration. The total urinary excretion proportion of HSYA in 48 h varied from 41.97% to 65.98%, demonstrating that urine was the main excretion way of HSYA in human. 4. Conclusion A simple, rapid and sensitive LC–MS/MS method was developed and fully validated for determination of HSYA in human urine. This method offered advantages over previously reported assay of plasma samples in terms of convenient sample preparation (simple SPE extraction without dryness and reconstitution) and short analysis time. Moreover, owing to the switch of polar endogenous interference into waste channel, no significant matrix effect was observed for analysis of HSYA in human urine, and the LC–MS/MS method presented acceptable precision, accuracy, linearity, extraction and stability. Using the current method, urine was clarified as the main excretion way of HSYA in healthy volunteers for the first time. Acknowledgements This research was financially supported by A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, 2010, PAPD (TCM combined with western medicine); Leading Talents of scientific research in TCM of Jiangsu Province (No. LJ200906); the National Science and Technology Major Project ‘Creation of Major New Drugs’ (2012ZX09303009002) from China. References [1] Committee of National Pharmacopoeia, Chinese Pharmacopoeia, Chemical Industry Press, Beijing, China, 2010, pp. 141. [2] J. Wang, Q. Zhang, X. Mei, X. Zhang, Pharm. Biol. 52 (2014) 31. [3] N. Zhang, M. Xing, Y. Wang, H. Liang, Z. Yang, F. Shi, Y. Cheng, Neurosci. Bull. 30 (2014) 417. [4] C.Y. Wang, Q. Liu, Q.X. Huang, J.T. Liu, Y.H. He, J.J. Lu, X.Y. Bai, Phytomedicine 20 (2013) 592. [5] Y.B. Zhang, H.Y. Dong, X.M. Zhao, L. Fan, Y. Zou, C. Zhang, G. Li, J.C. Liu, Y.C. Niu, Am. J. Chin. Med. 40 (2012) 481. [6] L. Song, Y. Zhu, M. Jin, B. Zang, Fitoterapia 84 (2013) 107. [7] J. Li, S. Zhang, M. Lu, Z. Chen, C. Chen, L. Han, M. Zhang, Y. Xu, Neurosci. Lett. 535 (2013) 51. [8] Y. Wu, L. Wang, M. Jin, B.X. Zang, Biol. Pharm. Bull. 35 (2012) 515.
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137 [9] Y. Liu, Z. Lian, H. Zhu, Y. Wang, S. Yu, T. Chen, J. Qu, J. Li, S. Ma, X. Chen, Evid. Based Complement. Alternat. Med. 2013 (2013) 147362. [10] S.Z. Kong, Y.F. Xian, S.P. Ip, X.P. Lai, X.G. Shi, Z.X. Lin, Z.R. Su, Neurochem. Res. 38 (2013) 951. [11] B. Han, J. Hu, J. Shen, Y. Gao, Y. Lu, T. Wang, Eur. J. Pharmacol. 714 (2013) 83. [12] L. Chen, Y. Xiang, L. Kong, X. Zhang, B. Sun, X. Wei, H. Liu, Neurochem. Res. 38 (2013) 2268. [13] Y. Pan, D.Y. Zheng, S.M. Liu, Y. Meng, H.Y. Xu, Q. Zhang, J. Gong, Z.L. Xia, L.B. Chen, H.Y. Li, Phytother. Res. 26 (2012) 1500. [14] Q. Yang, Z.F. Yang, S.B. Liu, X.N. Zhang, Y. Hou, X.Q. Li, Y.M. Wu, A.D. Wen, M.G. Zhao, Neurochem. Res. 35 (2010) 1353. [15] J. Yang, F.Z. Yang, Y. Tian, Y.Y. Jia, F. Xiao, Y. Qiao, A.D. Wen, Pharmaceut. J. Chin. People’s Liberation Army 25 (2009) 68.
137
[16] Z.P. Zeng, S.Q. Wang, Central South Pharm. 8 (2010) 33. [17] A. Wen, J. Yang, Y. Jia, Z. Yang, Y. Tian, Y. Wu, Z. Wang, Z. He, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 876 (2008) 41. [18] J.R. Li, M.J. Sun, X.J. Chen, Q.N. Ping, Pharmaceut. Clin. Res. 14 (2009) 191. [19] Z.P. Zeng, S.Q. Wang, J. Hainan Med. Coll. 16 (2010) 10. [20] D. Chu, W. Liu, Z. Huang, S. Liu, X. Fu, K. Liu, Planta Med. 72 (2006) 418. [21] Y. Guo, Y. Wang, X. Huang, H. Lv, R. Fan, W. Huang, P. Gan, W. Liu, K. Yan, Z. Xia, J. Liu, Biomed. Chromatogr. 28 (2014) 1090. [22] A. Van Eeckhaut, K. Lanckmans, S. Sarre, I. Smolders, Y. Michotte, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 877 (2009) 2198. [23] Department of Health and Human Services, F.D.A., Guidance for Industry Bioanalytical Method Validation, 2001.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Measurement of hydroxysafflor yellow A in human urine by liquid chromatography–tandem mass spectrometry Chang-Yin Li 1 , Ji-Hong Chu 1 , Jun Zhang, Bing-Ting Sun, Guo-Liang Dai, Shi-Jia Liu, Wen-Zheng Ju ∗ Department of Clinical Pharmacology, Affiliated Hospital of Nanjing University of Chinese Medicine, No. 155 Hanzhong Road, Nanjing 210029, China
a r t i c l e
i n f o
Article history: Received 7 July 2014 Received in revised form 20 September 2014 Accepted 26 October 2014 Available online 4 November 2014 Keywords: Hydroxysafflor yellow A SPE LC–MS/MS Human urine Excretion Column switching
a b s t r a c t A rapid and specific high performance liquid chromatography–tandem mass spectrometry (LC–MS/MS) was developed for the quantification of hydroxysafflor yellow A (HSYA) in human urine with isorhamnetin-3-O-neohespeidoside as internal standard (IS). HSYA and IS were extracted from urine samples by simple solid-phase extraction and separated on an Agilent Zorbax SB C18 column (4.6 mm × 150 mm, 5 m) with the mobile phase of 0.2 mM ammonium acetate: methanol (30/70, v/v) at a flow rate of 0.4 mL/min. Polar endogenous interferences eluted in 0.1–2.5 min were switched into waste channel by the Valve Valco, to reduce the possible matrix effect for MS detection in each run. The MS detection of analytes was performed on a tandem mass spectrometer equipped with an electrospray ionization source in negative mode using multiple-reaction monitoring. The MS/MS ion transitions monitored were m/z 611.3 → 491.2 for HSYA and m/z 623.2 → 299.2 for IS. The method was fully validated for selectivity, sensitivity, linearity, precision, accuracy, recovery, matrix effect and stability, and then was applied to the urinary excretion study of injectable powder of pure HSYA in healthy Chinese volunteers for the first time. The results suggested that urine was the main excretion way of HSYA in healthy volunteers, further demonstrating the feasibility and necessity of our current method. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The safflower plant, namely Carthamus tinctorius L. is a kind of therophyte belonging to the Dicotyledoneae subclass, Compositae family. The dried flower of this plant, named as Honghua in traditional Chinese medicine, has been widely used to treat coronary heart disease, hypertension, and cerebrovascular disease in clinical practice of China [1]. Over the past decades, hydroxysafflor yellow A (HSYA, Fig. 1a), the main chemical component of Honghua, has attracted much attention because of its various attractive pharmacological activities such as cardiovascular protection [2], improvement of learning and memory [3], attenuation of hepatic fibrosis [4,5], anti-inflammation [6–8] and neuroprotection [9–14]. Accordingly, HSYA was chosen as an active marker component for controlling the quality of safflower in Chinese Pharmacopoeia [1]. In recent years, several HSYA-containing preparations have been
∗ Corresponding author. Tel.: +86 25 8661 7141x50523; fax: +86 25 8655 5033. E-mail addresses: [email protected] (C.-Y. Li), [email protected] (W.-Z. Ju). 1 The first two authors contributed equally to this work. http://dx.doi.org/10.1016/j.jchromb.2014.10.036 1570-0232/© 2014 Elsevier B.V. All rights reserved.
approved for clinical research by China Food and Drug Administration (CFDA). Considering that better understanding the in-vivo metabolism of HSYA is an indispensable part in clinical study of HSYA-related drugs, it is necessary to develop simple and reliable method for analysis and determination of HSYA in human biological matrices such as plasma, urine and feces. To date, both LC–UV [15,16] and LC–MS/MS [17] methods have been reported for measuring HSYA in human plasma. However, these methods were not very suitable for the urinary excretion study of HSYA owing to the following two reasons: first, urine is a biological matrix quite different from plasma, and it contains various polar endogenous compounds which may result in remarkable matrix effect; and second, drug concentration in urine is also very different from that in plasma. Considering that the urinary excretion study of HSYA is an indispensible step for the research and development of new drug, several LC–UV methods [18–20] has been applied for the determination of HSYA concentration in rat urine previously. However, the disadvantages of LC technology, such as long analysis time, complex sample preparation and unsatisfied sensitivity and specificity, made these LC methods not very suitable for human urinary excretion study of HSYA-containing preparations. Given that LC–MS/MS technique
132
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
Fig. 1. The chemical structure of HSYA (a) and internal standard SLS (b).
is currently considered as the method of first choice for supporting clinical drug metabolism studies, recently a UPLC–ESI–MS/MS method was employed for the determination of HSYA in human urine [21]. Compared with the above LC methods, this UPLC–MS/MS method offers a rapid analysis of HSYA in human urine with excellent sensitivity and specificity. Despite this, the absence of internal standard and evaluation for matrix effect [22], two indispensible parts for a complete LC–MS bioanalytical method, made this method less accurate and robust to some extent. Furthermore, the complex sample preparation procedures including evaporation to dryness and reconstitution steps are unavoidable, hindering high throughput analysis for large numbers of samples in clinical studies. Therefore, it is still in great need to develop a more convenient, sensitive, accurate and high-throughput LC–MS/MS method to quantify the concentration of HSYA in human urine. As an adaptation of the previously published LC–MS/MS method for the determination of HSYA in human plasma [17], the purpose of the present study was to develop and validate a simple, rapid, accurate and specific LC–MS/MS method for the quantification of HSYA in human urine. Column switching technique, simple sample preparation and stable internal standard were applied into our current LC–MS/MS method, to make it more suitable for the human urinary excretion study of HSYA.
Agilent ZORAX SB-C8 guard column (2.1 × 12.5 mm, 5 m). The system was eluted in isocratic mode with a mobile phase consisting of 0.2 mM ammonium acetate: methanol (30/70, v/v) at a flow rate of 0.4 mL/min. The column oven temperature was maintained at 30 ◦ C and the total LC run time was 5.0 min. Volumes of 10 L were injected using an autosampler thermostatted at 8 ◦ C. The eluant was switched into waste channel for 0.1–2.5 min and MS channel for 2.5–5.0 min by the Valve Valco. An API4000 tandem quadrupole mass spectrometer equipped with electrospray ionization (ESI) source (Applied Biosystems, USA) operating in the negative ion mode was used as a detector. The MS/MS system was operated at unit resolution in the multiplereaction monitoring (MRM) mode, using precursor to product ion transition of m/z 611.3 → 491.2 for HSYA, m/z 623.2 → 299.2 for SLS (IS). The major ESI source parameters were kept as follows: ion spray voltage (ISV), −4500 V; turbo heater temperature (TEM), 400 ◦ C; collision activation dissociation (CAD), 10 psi, curtain gas (CUR), 25 psi. While the main compound dependent parameters, declustering potential (DP) and collision energy (CE), were optimized at −125 V and −34 eV for HSYA, −150 V and −62 eV for IS, respectively. Dwell time set for both HSYA and IS was 150 ms. All the operations, acquisition and data analysis were controlled by Analyst software version 1.6.0.
2. Experimental
2.3. Preparation of stock solutions, calibration curve and quality control samples
2.1. Material and reagents Reference standards of HSYA (purity: 92.5%, lot 111637– 201106) and the internal standard (IS) isorhamnetin-3-Oneohespeidoside (SLS, purity: 93.2%, lot 111571–201205, Fig. 1b) were purchased from National Institutes for Food and Drug Control (Beijing, PR China). Injectable powder of pure HSYA (IPPH) was provided by Guangzhou Yuekang biopharmaceutical Co., Ltd. (Guangzhou, China). Methanol and ammonium acetate were of HPLC grade (Merck, USA). Other chemicals were of analytical reagent grade. Ultrapure water used for the LC–MS/MS was produced using the Milli-Q water purification system (Millipore, USA). Blank urine used in this study was obtained from healthy volunteers.
The standard stock solutions of HSYA were prepared in methanol at the concentration of 1.05 mg/mL for calibration curve, 1.07 mg/mL for quality control (QC). A series of HSYA working solutions for calibration curve (1050, 349, 116, 38.8, 12.9, 4.31, 1.44 and 0.48 g/mL) and QC (855, 8.55, 0.95 g/mL) were prepared by stepwise dilution of the stock solution with methanol. The IS working solution (1.99 g/mL) was obtained by diluting the stock solution of SLS (0.99 mg/mL) with methanol. All the above solutions were stored at 4 ◦ C until use. Blank urine was stored in the freezer and allowed to completely thaw before use. The calibration curve and quality control samples were prepared by spiking 20 L of working solutions into 100 L of blank human urine. The spiked urine samples were stored at −20 ◦ C for method validation.
2.2. Instrumentation and conditions
2.4. Sample extraction procedures
LC separation was performed on a series 1200 HPLC system (Agilent, USA) equipped with a binary pump (1260G1312A), a column Oven (1260G1316A), an auto sampler (1260G1367E) and a Valve Valco 2-Position. HSYA and IS were separated on an Agilent Zorbax SB C18 column (4.6 mm × 150 mm, 5 m), preceded with an
Solid-phase extraction (SPE) of HSYA from human urine was performed using Waters (Michigan, USA) Oasis® HLB 3cc (60 mg) Extraction Cartridges. Before use, the cartridges should be sequentially conditioned with 3 mL of methanol and followed by 3 mL of water.
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
133
After being diluted by water in the volume proportion of 1:9, a 100 L aliquot of the human dilution urine was mixed with 20 L of IS working solution and 400 L of ammonium acetate (0.2 mM) water solution. The mixture was vortexed for 30 s and subsequently centrifuged at 12,000 × g for 5 min at 4 ◦ C. A 480 L aliquot of the supernatant was then loaded onto the conditioned cartridges, which was followed by being washed with 2 mL of water. Analytes were finally eluted with 1 mL of 70% methanol/water (v/v) solution and a 5 L aliquot of the eluant was injected into the LC/MS/MS system for analysis.
Freeze–thaw stability was conducted by testing spiked samples following three freeze–thaw cycles (−70 to 25 ◦ C, at least 12 h time duration between cycles). All stability evaluations were based on back-calculated concentrations. Samples were considered to be stable if assay values were within the acceptable limits of accuracy (85–115% of nominal value) and precision (less than 15%). Besides, the stability of the methanol stock solutions of HSYA at 4 ◦ C was also evaluated for as long as 290 days using the concentration level of approximately 3.0 g/mL.
2.5. Method validation
2.6. Application to clinical samples
According to the Food Drug and Administration (FDA) guidance on bio-analytical method validation [23], the method was validated for selectivity and sensitivity, linearity, precision, accuracy, recovery, matrix effect, and stability. Selectivity was proved by comparing chromatograms of six independent sources of blank human urine with a urine sample spiked with HSYA and IS. Chromatographic peaks of analyte and IS were identified on the basis of their MRM responses and retention times, and there should be no significant interference at the retention time of analyte and IS in the chromatograms of blank urine samples. Sensitivity was determined from the signal-to-noise ratio (S/N) of the analyte response in six replicates of the lowest calibration standard sample, and the S/N ratio should be no less than 10 for the lower limit of quantification (LLOQ), with acceptable precision (less than 20%) and accuracy (80–120%). Linearity was evaluated over five working days in the concentration range of 0.10–210 g/mL, using eight calibration standard samples prepared as described in Section 2.3. Calibration curves were created by plotting the LC–MS/MS peak area ratio of HSYA to IS against the nominal concentration of HSYA in urine sample, with 1/x2 weighted regression. The coefficient of correlation (r2 ) should be more than 0.99, and the accuracy for each standard point had to be within 85–115% of the theoretical concentration, except for 80–120% at LLOQ point. The intra-batch and inter-batch precision and accuracy were checked on three consecutive days by measuring HSYA concentration in five replicates at three QC levels at the nominal concentrations of 0.19, 1.71, 171 g/mL. The precision at each concentration was reported as the relative standard deviation (RSD%), expressing the standard deviation (SD) as a percentage of the mean calculated concentration, while the accuracy was determined by depicting the mean calculated concentration as a percentage of the nominal concentration. The criteria for data acceptance are less than 15% for precision and within 85–115% for accuracy. The extraction recovery of HSYA was determined in five replicates, by comparing the peak area of the analyte extracted from the above three levels of QC urine samples, with the peak area of the extracted matrix added with the same amount of the analyte. Extraction recovery of IS was calculated by the same way at a concentration of 0.04 g/mL. The recovery was deemed acceptable if the RSD% value of peak area ratio was less than 15% of mean recoveries. Matrix effect was evaluated on six independent sources of blank human urine, by calculating the ratio of the analyte peak areas in blank matrix spiked post-SPE to that in methanol solution at three QC levels of HSYA (0.19, 1.71, 171 g/mL). If the peak area ratio is less than 85% or more than 115%, a matrix effect is implied and should not be overlooked. Fresh stock solutions of HSYA and IS were used for the urine stability studies. The stability of HSYA was tested under various conditions using three levels of QC samples. Bench-top (short-term) stability of spiked human urine was evaluated for 4 h at room temperature (25 ◦ C) and 24 h at 8 ◦ C. Long-term stability was examined by assaying spiked urine samples stored at −70 ◦ C for 13 days.
After being fully validated, this LC–MS/MS method was then applied to the human urinary excretion study of HSYA following drip intravenous infusion of IPPH at a single dose of 75 mg. Three healthy male Chinese volunteers, aged from 22 to 28, with Body Mass Index (BMI) from 19 to 24, were recruited for this study at the Clinical Pharmacology Laboratory, They were all informed of the aim and risks involved in the study and then signed an informed consent. The experiment protocols were approved by the Ethical Review Committee of the Affiliated Hospital of Nanjing University of Chinese Medicine, in accordance with the principles of the Declaration of Helsinki, and the recommendations of China Food and Drug Administration. After an overnight fasting, volunteers received a single drip intravenous infusion dose of 75 mg injectable HSYA powder over a period of approximately 1 h. Serial urine samples were collected prior to dosing and over the intervals of 0–4, 4–8, 8–12, 2–24, 24–48 h post-dosing. The exact total volume of urine samples was recorded and 15 mL of each sample was finally retained and stored at −70 ◦ C until use.
3. Results and discussion 3.1. Method development All steps of the method, including sample preparation, selection of IS, chromatographic conditions and mass parameters were systematically optimized in the process of method development. Protein precipitation, liquid–liquid extraction and SPE extraction were three frequently used methods for sample preparation, and the former two methods are relatively convenient to operate and therefore very suitable for high throughput analysis. However, our initial analysis indicated that liquid–liquid extraction of urine samples yielded poor extraction recovery for HSYA (as low as approximately 1%); protein precipitation by methanol and acetonitrile obtained better extraction recovery, but suffered from endogenous interference and irregular chromatographic peaks for HSYA and IS. While Using waters Oasis@ HLB 3cc (60 mg) Extraction Cartridges and 70% methanol/water solution (as the eluent), SPE extraction could achieve excellent extraction efficiency of HSYA and IS, with no remarkable endogenous interference. Therefore, SPE extraction was finally selected for sample preparation, which was consistent with the previous report for plasma samples [17]. Unlike the above previous method, before SPE extraction, 400 L aliquot of 0.2 M ammonium acetate should be added into the urine samples to improve the extraction recovery of HSYA. Furthermore, aided by the optimization of LC–MS conditions and SPE method, SPE extraction was further simplified by the omission of the steps of dryness and reconstitution. Puerarin, a flavanoid compound used as IS in the previous study [17], proved not to be stable under the current conditions during the whole analytical process. Instead, SLS, another flavanoid compound that possessed similar molecular weight, chromatographic behavior and ESI response with HSYA, was finally selected as the
134
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
Fig. 2. Typical MRM chromatograms of (A) blank human urine; (B) HSYA spiked urine sample (0.10 g/mL, marked concentration in urine sample) for LLOQ; (C) methanol solution of standard mixture for HSYA (0.017 g/mL, final concentration for analysis) and IS (0.04 g/mL); (D) HSYA spiked urine sample (1.71 g/mL) for quality control; (E) urine sample obtained from a volunteer at the interval of 24–48 h after drip intravenous infusion of HSYA injection containing 75 mg of HSYA.
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
135
Fig. 3. Product ion mass spectra of [M–H]− of HSYA (a) and internal standard SLS (b).
Table 1 Intra- and inter-batch precision and accuracy for determination of HSYA in human urine. (Data are based on assay of five replicates per day, on three consecutive days). Spiked concentration (g/mL)
0.19 1.71 171
Intra-batch (n = 5)
Inter-batch (n = 15)
Mean (g/mL)
SD
RSD (%)
Accuracy (%)
Mean (g/mL)
SD
RSD (%)
Accuracy (%)
0.18 1.73 164
0.01 0.04 8.68
7.95 2.23 5.29
96.56 101.4 96.11
0.19 1.7 164
0.02 0.09 9.11
8.87 5.18 5.55
101.1 99.28 96.13
internal standard in our study. The extraction recovery of SLS was satisfactory and it was stable during the whole analytical process. Chromatographic conditions, especially mobile phase compositions, types of analytical column, were evaluated and optimized to obtain good chromatographic separation and ionization. As shown in Fig. 2, the final LC condition as described in Section 2.2 could bring the symmetrical peak shapes, good resolution and short retention times (less than 5 min) for both HSYA and IS. It should be noted that the addition of 0.2 mM ammonium acetate in mobile phase was an indispensible step for keeping the regular peak shape of HSYA. The mass spectrometric conditions were also optimized for HSYA and IS to obtain the maximal sensitivity, and the typical product ion mass spectra of [M–H]− of HSYA and IS are shown in Fig. 3. As for LC–MS analysis of urine samples, it should be noted that the abundant salts and polar endogenous polar compounds in urine may bring significant ion suppression or enhancement for
mass detection, even though after SPE extraction, while column switching technique could effectively minimize the salt effects or matrix effects [22]. Therefore, in the current study, the eluant in 0.1–2.5 min for each LC run was switched into waste channel, not MS channel, by a Valve Valco 2-Position. 3.2. Method validation 3.2.1. Selectivity and sensitivity Representative MRM chromatograms of blank urine, methanol solution of standards, blank urine spiked with HSYA and IS and urine sample from a volunteer after intravenous administration are shown in Fig. 2. The retention times for HSYA and IS were approximately 2.8 and 4.1 min, respectively. There was no apparent interference in urine matrix at the retention times of HSYA and I.S in the corresponding ion channel, demonstrating that the analysis
Table 2 Stability of HSYA in human urine (n = 3). Spiked concentration (g/mL)
0.19
1.71
171
Short-term stability (4 h, room temperature)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 5.14 106.7
1.56 0.07 4.16 91.42
158 2.1 1.33 92.36
Freeze and thaw stability (cycle 1)
Mean (g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 5.52 105.6
1.81 0.13 7.4 105.9
165 1.48 0.9 96.71
Freeze and thaw stability (cycle 2)
Mean (g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 6.71 102.5
1.84 0.08 4.51 107.5
163 3.15 1.93 95.36
Freeze and thaw stability (cycle 3)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.19 0.01 6.98 102.3
1.66 0.06 3.54 97.05
168 6.95 4.13 98.51
Long-term stability (13 days, −70 ◦ C)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 3.11 106.6
1.64 0.12 7.32 96.13
167 4.36 2.61 97.66
Post-preparative stability (24 h, 8 ◦ C)
Mean(g/mL) SD RSD (%) Accuracy (%)
0.2 0.01 5.97 106.1
1.68 0.04 2.3 98.33
168 6.07 3.61 98.37
136
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137
Table 3 Concentration of HSYA in human urine samples for three male volunteers following drip intravenous infusion of injectable HSYA powder at a single dose of 75 mg. Time interval (h) Subject A
Subject B
Subject C
Concentration (g/mL) Volume (mL) Excretion amount (mg) Concentration (g/mL) Volume (mL) Excretion amount (mg) Concentration (g/mL) Volume (mL) Excretion amount (mg)
0–4
4–8
8–12
12–24
24–48
33.79 480 16.22 33.79 510 17.23 181.7 150 27.26
8.99 850 7.64 19.3 400 7.72 30.5 400 12.20
7.93 280 2.22 6.15 650 4.00 6.80 340 2.31
17.14 230 3.94 2.01 1100 2.21 8.74 790 6.90
1.33 1100 1.46 0.41 2550 1.04 0.31 2630 0.82
TU (mg)
TU%
31.48
41.97
32.19
42.92
49.49
65.98
TU: total urinary excretion amount of HSYA in 48 h; TU%: total urinary excretion proportion of HSYA in 48 h.
of HSYA and IS using the MRM function was highly selective in the current study. The LLOQ of HSYA was 0.10 g/mL in human urine, and six replicate analyses of LLOQ samples showed excellent accuracy of 99.80% and satisfied precision with RSD% value of 2.07%
3.2.2. Calibration linearity The calibration curves showed good linearity over five working days in the concentration range of 0.10–210 g/mL. The equation of linearity was y = (0.217 ± 0.011)x + (0.0129 ± 0.0048), with the coefficient of correlation (r2 ) ranging from 0.993 to 0.998, while the accuracy for all calibration concentrations was within 85–115% of the theoretical values, with RSD% ranging from 1.66 to 7.61.
3.2.3. Accuracy and precision The intra-batch and inter-batch precision and accuracy of HSYA are summarized in Table 1. The RSD% values were no more than 8.87% for both intra-batch and inter-batch precision, while the accuracy varied from 96.11% to 101.1%. All of the values well meet the criteria for data acceptance in the guidance [21], indicating that the present LC–MS/MS method was accurate, reliable and reproducible for the quantitative analysis of HSYA in human urine samples.
3.2.4. Extraction recovery and matrix effect The average extraction recovery for HSYA were 72.65%, 72.81% and 78.52% at low, medium and high QC levels respectively, with RSD% values of 4.08%, 2.51%, and 2.10%, while the average recovery of IS was 97.14%, with RSD% value of 4.10%. The results demonstrated that the extraction recovery of our method was acceptable and stable for bio-analytical analysis. The average matrix effect for HSYA varied from 102% to 110.1% at three QC levels, with RSD% values ranged from 0.50% to 6.74%. The average matrix effect value for IS was 87.96%. These data suggested that no significant ion suppression or enhancement was observed under our experimental conditions.
3.2.5. Stability The stability data of HSYA kept under various storage conditions were summarized in Table 2. Urine samples were proved to be stable for 4 h at room temperature, for 13 days when stored at −70 ◦ C and through three freeze–thaw cycles. Samples after treatment were also stable at 8 ◦ C in auto-sampler for 24 h. In addition, after 290 days’ storage at 4 ◦ C, the methanol solutions of HSYA kept stable, as the accuracy normalized to freshly prepared solution was 96.25% with the RSD value of 2.02% (n = 3). These results indicated that HSYA was stable under the different storage and temperature conditions and allowed stockpiling of urine samples obtained in the excretion studies for subsequent batch analysis.
3.3. Urinary excretion study The developed LC–MS/MS method was then applied to the human urinary excretion study of HSYA following drip intravenous infusion of IPPH at a single dose of 75 mg. Table 3 summarized Concentration of HSYA in human urine samples for three male volunteers following drip intravenous infusion of injectable HSYA powder. The results indicated that the linearity range of our method (0.10–210 g/mL) covered well with the concentrations of all tested human urine samples (0.31–182 g/mL), without any need for dilution or concentration. The total urinary excretion proportion of HSYA in 48 h varied from 41.97% to 65.98%, demonstrating that urine was the main excretion way of HSYA in human. 4. Conclusion A simple, rapid and sensitive LC–MS/MS method was developed and fully validated for determination of HSYA in human urine. This method offered advantages over previously reported assay of plasma samples in terms of convenient sample preparation (simple SPE extraction without dryness and reconstitution) and short analysis time. Moreover, owing to the switch of polar endogenous interference into waste channel, no significant matrix effect was observed for analysis of HSYA in human urine, and the LC–MS/MS method presented acceptable precision, accuracy, linearity, extraction and stability. Using the current method, urine was clarified as the main excretion way of HSYA in healthy volunteers for the first time. Acknowledgements This research was financially supported by A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, 2010, PAPD (TCM combined with western medicine); Leading Talents of scientific research in TCM of Jiangsu Province (No. LJ200906); the National Science and Technology Major Project ‘Creation of Major New Drugs’ (2012ZX09303009002) from China. References [1] Committee of National Pharmacopoeia, Chinese Pharmacopoeia, Chemical Industry Press, Beijing, China, 2010, pp. 141. [2] J. Wang, Q. Zhang, X. Mei, X. Zhang, Pharm. Biol. 52 (2014) 31. [3] N. Zhang, M. Xing, Y. Wang, H. Liang, Z. Yang, F. Shi, Y. Cheng, Neurosci. Bull. 30 (2014) 417. [4] C.Y. Wang, Q. Liu, Q.X. Huang, J.T. Liu, Y.H. He, J.J. Lu, X.Y. Bai, Phytomedicine 20 (2013) 592. [5] Y.B. Zhang, H.Y. Dong, X.M. Zhao, L. Fan, Y. Zou, C. Zhang, G. Li, J.C. Liu, Y.C. Niu, Am. J. Chin. Med. 40 (2012) 481. [6] L. Song, Y. Zhu, M. Jin, B. Zang, Fitoterapia 84 (2013) 107. [7] J. Li, S. Zhang, M. Lu, Z. Chen, C. Chen, L. Han, M. Zhang, Y. Xu, Neurosci. Lett. 535 (2013) 51. [8] Y. Wu, L. Wang, M. Jin, B.X. Zang, Biol. Pharm. Bull. 35 (2012) 515.
C.-Y. Li et al. / J. Chromatogr. B 974 (2015) 131–137 [9] Y. Liu, Z. Lian, H. Zhu, Y. Wang, S. Yu, T. Chen, J. Qu, J. Li, S. Ma, X. Chen, Evid. Based Complement. Alternat. Med. 2013 (2013) 147362. [10] S.Z. Kong, Y.F. Xian, S.P. Ip, X.P. Lai, X.G. Shi, Z.X. Lin, Z.R. Su, Neurochem. Res. 38 (2013) 951. [11] B. Han, J. Hu, J. Shen, Y. Gao, Y. Lu, T. Wang, Eur. J. Pharmacol. 714 (2013) 83. [12] L. Chen, Y. Xiang, L. Kong, X. Zhang, B. Sun, X. Wei, H. Liu, Neurochem. Res. 38 (2013) 2268. [13] Y. Pan, D.Y. Zheng, S.M. Liu, Y. Meng, H.Y. Xu, Q. Zhang, J. Gong, Z.L. Xia, L.B. Chen, H.Y. Li, Phytother. Res. 26 (2012) 1500. [14] Q. Yang, Z.F. Yang, S.B. Liu, X.N. Zhang, Y. Hou, X.Q. Li, Y.M. Wu, A.D. Wen, M.G. Zhao, Neurochem. Res. 35 (2010) 1353. [15] J. Yang, F.Z. Yang, Y. Tian, Y.Y. Jia, F. Xiao, Y. Qiao, A.D. Wen, Pharmaceut. J. Chin. People’s Liberation Army 25 (2009) 68.
137
[16] Z.P. Zeng, S.Q. Wang, Central South Pharm. 8 (2010) 33. [17] A. Wen, J. Yang, Y. Jia, Z. Yang, Y. Tian, Y. Wu, Z. Wang, Z. He, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 876 (2008) 41. [18] J.R. Li, M.J. Sun, X.J. Chen, Q.N. Ping, Pharmaceut. Clin. Res. 14 (2009) 191. [19] Z.P. Zeng, S.Q. Wang, J. Hainan Med. Coll. 16 (2010) 10. [20] D. Chu, W. Liu, Z. Huang, S. Liu, X. Fu, K. Liu, Planta Med. 72 (2006) 418. [21] Y. Guo, Y. Wang, X. Huang, H. Lv, R. Fan, W. Huang, P. Gan, W. Liu, K. Yan, Z. Xia, J. Liu, Biomed. Chromatogr. 28 (2014) 1090. [22] A. Van Eeckhaut, K. Lanckmans, S. Sarre, I. Smolders, Y. Michotte, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 877 (2009) 2198. [23] Department of Health and Human Services, F.D.A., Guidance for Industry Bioanalytical Method Validation, 2001.
