Simultaneous determination of thirteen flavonoids from Xiaobuxin-Tang extract using high-performance liquid chromatography coupled with electrospray ionization mass spectrometry

Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba

Simultaneous determination of thirteen flavonoids from Xiaobuxin-Tang extract using high-performance liquid chromatography coupled with electrospray ionization mass spectrometry Meifeng Cen a,1 , Jinxiu Ruan b,1 , Lihua Huang a , Zhenqing Zhang b , Nengjiang Yu b , Youzhi Zhang b , Xuange Cheng a , Xiaohong Xiong a , Guixiang Wang a , Linquan Zang a , Sujun Wang a,∗ a

School of Pharmacy, Guangdong Pharmaceutical University, Guangzhou 510006, People’s Republic of China Key Laboratory of Drug Metabolism and Pharmacokinetics, Beijing Institute of Pharmacology and Toxicology, 27 Taiping Road, Beijing 100850, People’s Republic of China b

a r t i c l e

i n f o

Article history: Received 5 May 2015 Received in revised form 16 July 2015 Accepted 17 July 2015 Available online 26 July 2015 Keywords: HPLC-ESI–MS Thirteen flavonoids Intestinal absorption

a b s t r a c t A simple and reliable high performance liquid chromatography coupled with electrospray ionization mass spectrometry (HPLC-ESI–MS) analysis method was established to simultaneously determine thirteen flavonoids of Xiaobuxing-Tang in intestine perfusate, namely onpordin, 3 -O-methylorobol, glycitein, patuletin, genistein, luteolin, quercetin, nepitrin, quercimeritrin, daidzin, patulitrin, quercetagitrin and 3-glucosylisorhamnetin. Detection was performed on a quadrupole mass spectrometer equipped with an electrospray ionization (ESI) source operating in negative ionization mode. Negative ion ESI was used to form deprotonated molecules at m/z 315 for onpordin, m/z 299 for 3 -O-methylorobol, m/z 283 for glycitein, m/z 331 for patuletin, m/z 269 for genistein, m/z 285 for luteolin, m/z 301 for quercetin, m/z 477 for nepitrin, m/z 463 for quercimeritrin, m/z 461 for daidzin, m/z 493 for patulitrin, m/z 479 for quercetagitrin, m/z 477 for 3-glucosylisorhamnetin and m/z 609.2 for rutin. The linearity, sensitivity, selectivity, repeatability, accuracy, precision, recovery and matrix effect of the assay were evaluated. The proposed method was successfully applied to simultaneous determination of these thirteen flavonoids, and using this method, the intestinal absorption profiles of thirteen flavonoids were preliminarily predicted. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Xiaobuxin-Tang (XBXT), a traditional Chinese herbal prescription used to remit depressive disorders, comprising Haematitum, Flos Inulae, Folium Phyllostachydis Henonis and Semen Sojae Preparatum four crude medicines, was originally recorded in the silk scroll manuscript of “Fuxingjue Zangfu Yongyao Fayao” [1]. We had found that the major chemical constituents of the extract (XBXT -2) from XBXT were flavones, flavonols, isoflavones and their glycosides [2]. Pharmacological investigations reveal that those compounds are responsible for significantly anti-depressant

∗ Corresponding author at: School of Pharmacy, Guangdong Pharmaceutical University, Guangzhou Higher Education Mega Center, 280 Wai Huan East Road, Guangzhou 510006, People’s Republic of China. E-mail address: [email protected] (S. Wang). 1 These authors contribute equally to this work. http://dx.doi.org/10.1016/j.jpba.2015.07.015 0731-7085/© 2015 Elsevier B.V. All rights reserved.

activity [3–5]. Onpordin, 3 -O-methylorobol, glycitein, patuletin, genistein, luteolin, quercetin, nepitrin, quercimeritrin, daidzin, patulitrin, quercetagitrin and 3-glucosylisorhamnetin are the thirteen main flavonoids in XBXT-2. However, quantification and pharmacokinetic study of these flavonoids of XBXT-2 have not been conducted. High performance liquid chromatography coupled with electrospray ionization mass spectrometry (HPLC-ESI–MS) has been applied for in vivo analysis of glycitein [6], genistein [6], daidzin [7] and quercetin [8] of other plants with properties of high mass accuracies and high selectivity, nevertheless, it has not been developed and reported along with in vivo studies of patulitrin, nepitrin, patuletin, luteolin, onpordin, quercimeritrin, quercetagitrin, 3 -O-methylorobol and 3-glucosylisorhamnetin. Furthermore, HPLC-ESI–MS method has not been reported to quantification of 3-glucosylisorhamnetin, onpordin, quercimeritrin and quercetagitrin these four flavonoids in some plant materials. Chinese herbal medicines (TCM) are usually used through oral administration, this renders intestinal absorption is significantly

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

crucial to evaluate the bioactivity of these flavonoids. Although investigations on intestinal absorption of flavonoids have been performed to a great extent with several methods included Caco-2 cell monolayer model [9], in situ single-pass intestinal perfusion [10], in vitro everted gut sac model [11], and in vivo pharmacokinetic study [12], the in situ vascularly perfused rat intestine preparation method available is scarce and has not been used to study these thirteen flavonoids as well. In this study, a sensitive, precise and convenient HPLC-ESI–MS method for simultaneous determination of thirteen flavonoids compounds in XBXT-2 was established and used to determine them in the intestinal perfusate for the first time. The in situ vascularly perfused rat intestine preparation was used to preliminarily investigate their intestinal absorption characteristic, and the obtained results would provid a meaningful basis for their in vivo pharmacokinetic studies. 2. Experimental 2.1. Chemicals, reagents and materials The extract (XBXT-2) from Xiaobuxing-Tang was provided by Professor Yimin Zhao (Laboratory of phytochemistry, Beijing Institute of Pharmacology and Toxicology, Beijing, China) and the method of extracting flavonoids was according to the literature previously described [13]. Reference substances of onpordin, 3 -O-methylorobol, glycitein, patuletin, genistein, luteolin, quercetin, nepitrin, quercimeritrin, daidzin, patulitrin, 3glucosylisorhamnetin, quercetagitrin and rutin (≥96.5%, purity) were purchased from Shanghai Forever Biotech Co., Ltd. (Shanghai, China). 5% Bovine serum albumin (≥98.0%, purity) was purchased from Shanghai Junruishengwu Biomart Co., Ltd. (Shanghai, China), 10% washed rat red blood cells were taken from the healthy partial adult male sprague-dawley rat blood withdrawn within one week. 3% dextran (T-40), 0.02% dexamethasone and 0.004% noradrenalin were obtained from National Institute for the Control of Pharmaceutical and Biological Products. The Krebs-Ringer solution (intestinal perfusate, PH 7.4) consisted of NaH2 PO4 (0.22 g), NaCl (7.8 g), CaCl2 (0.37 g), MgCl2 (0.22 g), KCl (0.35 g), NaHCO3 (1.37 g), glucose (3.0 g), 3% dextran (T-40), 5% bovine serum albumin, 10% washed rat erythrocyte, 0.02% hexadecadrol and 0.004% noradrenalin dissolved in 1.0 L of deionized water. Acetonitrile and methanol were of HPLC grade (Fisher Scientific, Fair Lawn, NJ, USA). All other chemicals were of analytical reagent grade. Distilled, deionized water was produced by a Milli-Q Reagent Water System (Millipore, MA, USA). 2.2. Instrument and chromatographic conditions The molecular weights of thirteen flavonoids were determined performing on Agilent LC–MSD quadrupole mass spectrometer, equipped with a series 1100HPLC system consisting of a binary pump, automatic solvent degasser and autosampler. Intersil C8 3 column (250 mm × 2.1 mm, 5 ␮m, DIKMA, Japan) was used and the column temperature was maintained at 25 ◦ C. The binary solvent system was consisted of A (acetonitrile) and B (water with 5 mM ammonium formate) using a gradient elution at a flow of 0.2 mL/min. The gradient conditions of the mobile phase were as follows: 0–10 min, 20% A, 10–15 min, 25% A, 15–20 min, 40% A, 20–25 min, 55% A, 25–40 min, 20% A. The Agilent MS system operated in negative ion electrospray ionization was used to form deprotonated molecules for all compounds. Nitrogen was used as nebulization gas and was set to 10 L/min at a temperature of 350 ◦ C. The capillary voltage was set at 4.0 kV and the nebulizer pressure was 40 psi.

215

2.3. In situ vascularly perfused rat intestine preparation The surgical procedure for the in situ vascularly perfused rat intestine preparation was similar to that for the in situ vascularly perfused rat intestine–liver preparation with modification [14]. Adult male SD rats (180–220 g) were anesthetized with intraperitoneal injection of sodium pentobarbital solution following overnight fasting. The abdomen was opened with a midline incision to expose the organs, pyloric vein, celiac artery, gastric artery, splenic blood vessels, and the lower aorta were tied before and after the common juncture of the superior mesenteric artery and right renal artery. The surgical cannulation for intestinal perfusion was illustrated in Fig. 1 and performed on peristaltic pump (HL-2, Luxi Analysis Instrument Co., Ltd., Shanghai, China). The intestine was perfused via the superior mesenteric artery, and its venous outflow into the portal vein, therefore, the diaphragm was cut to allow the outflow catheter to be in a straight line with the portal vein. Perfusion was initiated at a constant rate of 2–3 mL/min after the insertion of the cannula and was gradually increased to the desired value of 7.5 mL/min upon completion of the outflow circuit after the completion of surgery. The perfusate was oxygenated with carbogen (95% oxygen and 5% carbon dioxide) at 1.0 L/min. The surgical procedure for initial 15 min to wash the residual blood in the vessel and balance the whole perfusion path was carried out with perfusate that was not appended 3% dextran (T-40) and 10% washed rat erythrocyte, and then the perfused flow was increased to 10 mL/min before intraduodenal adminstration. The incisions were made on the duodenum to allow for insertion of the cannulas and on the ileocecal junction for collection of intestinal perfusate. Perfusion was started at a flow of 10 mL/min with the perfusate (PH 7.4) contained 3% dextran (T-40) and 10% washed rat erythrocyte, and the experimental time lasted for 2 h. Effluent perfusate samples were collected quantitatively from the reservoir before drug administration (zero time) and other 9 intervals (5, 15, 30, 40, 60, 75, 90, 105 and 120 min) after intraduodenal adminstration. The rats were placed on a flat plate under a heating pad to maintain the temperature at 37 ◦ C throughout the experiment and a cotton gauze moistened with isotonic saline solution was used to cover the surgical incision to prevent the intestine drying. 2.4. Animals and analysis sampling Adult male Sprague-Dawley (SD) rats, weighing 180–220 g, were purchased from the Experimental Animal Center of Guangzhou University of Chinese Medicine (Guangzhou, China). The rats were housed in the stainless steel metabolic cages equipped with an automated watering valve and cared for according to the regulations of the animal committeen under a constant temperature at (22 ± 1 ◦ C), humidity at (50 ± 20%), 12 h light/12 h dark cycle for one week. The experimental protocol was approved by the Ethics Committee of Guangdong Pharmaceutical University. The anaimls were fasted for 12 h prior to study and anesthetized with sodium pentobarbital solution via intraperitoneal injection. Each rat was given XBXT-2 at a single does of 0.2 g through intraduodenal administration, and perfused 2 h with the in situ vascularly perfused rat intestine preparation. The samples (500 ␮L) were withdrawn at different time intervals (0, 5, 15, 30, 40, 60, 75, 90, 105 and 120 min). 2.5. Sample preparation 2.5.1. Stock solutions, calibration standards and quality control (QC) Standard stock solutions of the thirteen flavonoids and rutin (internal standard, IS), were separately prepared by amounts

216

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

Fig. 1. Surgical cannulation for the in situ vascularly perfused rat intestine preparation. (a and b) respectively, the connections for inflow and outflow of the intestinal loop. (c) the inflow of perfusate into the superior mesenteric artery. (d) the outflow of perfusate from the venous return of the intestine at the portal vein.

of each component accurately weighed and dissolved in 5% Dimethyl sulfoxide (DMSO) and then stored at 4 ◦ C until analysis. The mixture standard stock solution was obtained by mixing thirteen flavonoids stock solutions above, and giving a final concentration of 10 ␮g/mL for each flavonoids. The mixed stock solution was serially diluted with acetonitrile to provid working standard solutions of desired concentrations. The samples for standard calibration curves were prepared by spiking the blank rat perfusate (190 ␮L) with 10 ␮L of the standard mixture working solutions to yield the following concentrations:10, 20, 50, 100, 200, 500, 1000 and 2000 ng/mL for onpordin, 3 -Omethylorobol, genistein and quercimeritrin, 50, 100, 200, 500, 1000, 2000 and 5000 ng/mL for glycitein, patuletin, nepitrin, daidzin, patulitrin, quercetagitrin and 3-glucosylisorhamnetin, and 20, 50, 100, 200, 500, 1000 and 2000 ng/mL for luteolin and quercetin. The IS working solution (250 ng/mL) was prepared by diluting the IS stock solution with acetonitrile. All solutions were stored at −20 ◦ C before analysis. Quality control (QC) samples were prepared by the same manner as calibration standards with blank perfusate. The concentrations were 20, 200 and 2000 ng/mL for onpordin, 3 -O-methylorobol, genistein, luteolin, quercetin and quercimeritrin, as well as 50, 500 and 5000 ng/mL for glycitein, patuletin, nepitrin, daidzin, patulitrin, quercetagitrin and 3-glucosylisorhamnetin. All samples were stored at −20 ◦ C until analysis.

2.5.2. Pretreatment of perfusate sample The perfusate samples (500 ␮L) withdrawn at different time intervals described above were at once centrifuged at 4000 × g for 10 min at 4 ◦ C, then a 0.1 mL volume of the resulting supernatant was added with 50 ␮L of the IS working solution and 100 ␮L of acetonitrile, and immediately vortexed for 3 min. The mixture was centrifuged again and the supernatant was transferred to an autosampler vial and an aliquot of 20 ␮L was injected for HPLC/ESI/MS system for analysis.

2.6. Validation of the method 2.6.1. Assay specificity and selectivity The specificity and selectivity were used to analyze to investigate the potential interferences from endogenous substance. Comparion study was conducted with chromatograms of blank rat perfusate, perfusate spiked with analytes and IS, and perfusate samples after intraduodenal administration of XBXT-2. 2.6.2. Linearity, limit of detection and the lower limit of quantification The linearity study was carried out by the calibration curves in the concentration rang 10–2000 ng/mL for onpordin, 3 O-methylorobol, genistein and quercimeritrin, 20–2000 ng/mL for luteolin and quercetin, and 50–5000 ng/mL for quercetagitrin, glycitein, patuletin, nepitrin, daidzin, patulitrin and 3-glucosylisorhamnetin. The calibration graphs were plotted after linear regression of the peak areas versus the corresponding concentrations. Weighted (1/conc2 ) least-squares linear regression analysis was used to determine the slope, intercept and correlation coefficient. Concentrations of these thirteen flavonoids in rat perfusate were determined from the peak-area ratios using the equations of linear regression obtained from the calibration curves. The limit of detection (LOD) and the lower limit of quantification (LLOQ) values obtained were determined when the signal-to-noise ratio of the testing peak of analyte were 3 and 10, respectively. 2.6.3. Recovery and matrix effect The extraction recovery (absolute recovery) of analytes from rat intestinal perfusate after the extraction procedure was assessed by comparing the analyte/IS peak area ratio of blank perfusate extracts spiked with standard solution (R1 ) with that of the standards prepared in mobile phase (R2 ). IS was spiked before extraction in both cases. QC samples at three concentrations were evaluated. The extraction recovery was expressed as (1.3 R1 /R2 ) × 100%. The matrix effect was measured by comparing the peak response of the

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

217

Fig. 2. (Continued)

for 3-glucosylisorhamnetin, glycitein, patuletin, nepitrin, daidzin, quercetagitrin and patulitrin.

Fig. 2. Mass spectra of thirteen compounds from Xiaobuxin-Tang extract.

post-extracted spiked samples with that of the unextracted samples (pure sample prepared in mobile phase) at concentration of 20, 200 and 2000 ng/mL for onpordin, 3 -O-methylorobol, genistein, luteolin, quercetin and quercimeritrin, of 50, 500 and 5000 ng/mL

2.6.4. Precision, accuracy and repeatablity The precision and accuracy were assessed by QC samples at three concentration levels. Intra-day and inter-day accuracy and precision (each, n = 5) were determined by repeatedly assaying QC samples on the same day and on three consecutive validation days, respectively. The concentration of each sample was determined using calibration standards prepared on the same day. Precision was expressed as a relative standard deviation

218

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

short-term stability tests were determined after the samples at room temperature for 12 h and long-term stability in perfusate was tested by assaying frozen QC samples after storage at −20 ◦ C for one month. The post-preparation stability was tested by determined of the extracted QC samples stored in the auto-sampler (4 ◦ C) for 24 h. Both thirteen flavonoids stock solutions and IS stock solution were tested after storing at room temperature for 12 h and −20 ◦ C for two weeks for the stability investigation. 3. Results and discussion 3.1. Optimisation of chromatographic conditions

Fig. 2. (Continued)

(RSD) and accuracy of the method was evaluated as relative error (RE), which was calculated by the equation: (mean concentration measured–concentration spiked/concentration spiked) × 100%. To confirm the repeatability, five different working solutions prepared from the calibration samples were analyzed, and the relative standard deviation (RSD) was taken as a measure of it. 2.6.5. Stability experiments The stability of the analytes in rat intestine perfusate was assessed by analyzing QC samples (low, medium and high concentration levels). Free-thaw stability tests were evaluated after three freeze-thaw cycles (−20 ◦ C to room temperature as one cycle),

Adequate chromatographic separation condition for thirteen flavonoids and IS from the co-eluted endogenous compounds was necessary to be established to avoid the matrix effect in the extracted perfusate. Consequently, the composition of the mobile phase and the buffer were screened in this study. The selection of eluent composition for HPLC–MS is usually compromise between HPLC separation and ionization efficiency, and the column diameter and solid-phase material exert significant impacts on the separation of the HPLC, therefore, for HPLC separation, different columns (Zoxbax Ecilipse plus-C18 column (100 mm × 2.1 mm, 5 ␮m), TskGEL Amide-80 column (100 mm × 2.1 mm, 5 ␮m) and Intersil C8 -3 column (250 mm × 2.1 mm, 5 ␮m)) and different mobile phases (acetonitrile and water containing 0.1% (v/v) formic aid, acetonitrile and water containing 5 mM ammonium formate, acetonitrile and water containing 10 mM ammonium formate as well as acetonitrile and water containing 5 mM ammonium acetate) were compared, it is worth noting that using acetonitrile and water with ammonium acetate, ammonium formate and formic acid can adjust the PH, which were optimized for each column in gradient elution. The trifluoroacetic acid (TFA) was not considered to be the additive used in this mobile phase because the strong volatile acids easily suppress the ESI signal [15]. Contrarily, the mobile phase containing a low concentration of ammonium formate or formic acid can increase analyte ESI response and control against matrix effects [16]. The result showed that 0.1% (v/v) formic acid added in water could lead to a singificant improvement in the separation, peak shape of the flavonoids and flavonoid glycosides, however, the limit of low sensitivity and stability was not satisfactory because formic acid as the aqueous component of the LC eluent was weak absorbance and unsuitability with negative ion detection of the flavonoids in contrast with ammonium salt [17]. The concentration of additives required in LC are important for ESI, which can tolerate only low additive concentration, and in practice, the optimization of additive concentrations is not exceed 10 mM in order to avoid suppression of ionization and reducing

Table 1 Statistical results of linear regression equation analysis in the determination of thirteen flavonoids. Investigated compound

Onpordin 3 -O-methylorobol Glycitein Patuletin Genistein Luteolin Quercetin Nepitrin Quercimeritrin Daidzin Patulitrin Quercetagitrin 3-Glucosylisorhamnetin

Regression equation Linear rang(ng/mL)

Slope (a)

Intercept (b)

10–2000 10–2000 50–5000 50–5000 10–2000 20–2000 20–2000 50–5000 10–2000 50–5000 50–5000 50–5000 50–5000

106606 107741 29380 32620 163861 77267 51582 39842 129337 24350 22274 20665 34975

23701 −2343 877 4698 20371 10652 8526 7858 20170 4823 −1087 913 8577

R2 (n = 5) 0.9971 0.9992 0.9979 0.9985 0.9951 0.9975 0.9981 0.9974 0.9966 0.9989 0.9989 0.9964 0.9980

LOD (ng/mL)

LLOQ (ng/mL)

3 3 15 15 3 5 5 15 3 15 15 15 15

9 9 45 45 9 15 15 45 9 45 45 45 45

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

219

Fig. 3. Selected ion monitoring mass chromatograms obtained during LC/MS with negative ion ESI of total ion chromatography (TIC) of blank rat intestinal perfusate sample (A), a blank intestinal perfusate sample spiked with standards of thirteen flavonoids and IS (B), and an actual intestinal perfusate sample of a rat taken 2 h after intraduodenal administration of XBXT-2 (C). (1) Onpordin, (2) 3 -O-methylorobol, (3) glycitein, (4) patuletin, (5) genistein, (6) luteolin, (7) quercetin, (8) nepitrin, (9) quercimeritrin, (10) daidzin, (11) patulitrin, (12) quercetagitrin, (13) 3-glucosylisorhamnetin, (14) rutin (IS).

sensitivity [15]. On the basis of the experimental evidence, the best chromatographic resolution was chosen to achieve by separating samples on Intersil C8 -3 column using a gradient elution of acetonitrile and water with 5 mM ammonium formate. The precursor and product ions of these thirteen flavonoids and IS were ascertained by infusing 200 ng/mL standard solution for analysis in selective ion monitoring (SIM) mode. In the precursor ion full-scan spectra, onpordin, 3 -O-methylorobol, glycitein, patuletin, genistein, lute-

olin, quercetin, nepitrin, quercimeritrin, patulitrin, quercetagitrin and 3-glucosylisorhamnetin exhibited higher response in negative ion mode and the most abundant ions were deprotonated molecules ion [M − H]− , while the ionization of daidzin of the negative deprotonated molecular ion [M + HCOO− ]− was detected, which suggested that the identification of the detected compounds was based on the search of the main molecular ious and also on some of the useful observed fragmentations. ESI /LC–MS was

220

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

Table 2 Intra-day and inter-day accuracy and precision of LC/MS determinaiton of thirteen flavonoids (n = 5). Concentration spiked(ng/ml)

Intra-day Measured concentration (ng/ml)

RSD (%)a

RE (%)b

Measured concentration (ng/ml)

Inter-day RSD (%)a

RE (%)b

Onpordin 20 200 2000

20.3 ± 1.3 181.5 ± 7.3 1821.0 ± 16.7

6.4 4.0 0.9

1.6 −9.3 −9.0

18.2 ± 1.4 186.9 ± 9.1 1858.3 ± 83.3

7.8 4.9 4.5

−9.0 −6.6 −7.1

3 -O-methylorobol 20 200 2000

19.3 ± 1.1 201.0 ± 13.4 1977.3 ± 87.8

5.6 6.7 4.4

−3.5 0.5 −1.1

18.3 ± 1.7 192.8 ± 14.0 1874.7 ± 151.6

9.3 7.3 8.1

−8.6 −3.6 −6.3

Glycitein 50 500 5000

46.9 ± 3.4 480.4 ± 18.1 4913.8 ± 83.0

7.2 3.8 1.7

−6.2 −3.9 −1.7

46.7 ± 4.3 503.8 ± 10.4 4700.6 ± 243.6

9.3 2.1 5.2

−6.6 0.8 −6.0

Patuletin 50 500 5000

47.7 ± 2.9 458.1 ± 13.9 4834.2 ± 174.3

6.0 2.9 3.6

−4.6 −3.0 −3.3

49.7 ± 4.3 482.0 ± 34.2 4882.0 ± 53.3

8.6 7.1 1.1

−0.5 −3.6 −2.4

Genistein 20 200 2000

19.4 ± 1.1 197.6 ± 10.6 1938.9 ± 69.7

5.6 5.4 3.6

−3.2 −1.2 −3.1

18.3 ± 1.7 192.8 ± 14.0 1934.7 ± 106.4

9.3 7.3 5.5

−8.6 −3.6 −3.3

Luteolin 20 200 2000

20.6 ± 1.7 180.4 ± 14.1 1957.1 ± 116.3

8.2 7.8 5.9

2.8 −9.8 −2.1

19.2 ± 1.4 180.9 ± 14.7 1967.2 ± 125.8

7.5 8.1 6.4

−3.9 −9.5 −1.6

Quercetin 20 200 2000

18.4 ± 1.0 201.8 ± 11.1 1910.5 ± 48.9

5.3 5.5 2.6

−8.1 0.9 −4.5

18.6 ± 1.8 180.9 ± 15.0 1997.5 ± 62.2

9.7 8.3 3.1

−7.2 −9.5 −0.1

Nepitrin 50 500 5000

49.4 ± 2.2 495.3 ± 13.6 4980.0 ± 53.7

4.5 2.7 1.1

−1.1 −0.9 −0.4

49.1 ± 3.9 459.8 ± 20.6 4911.5 ± 100.3

7.9 4.5 2.0

−1.9 −8.0 −1.8

Quercimeritrin 20 200 2000

20.4 ± 0.9 202.4 ± 11.6 1974.9 ± 71.2

4.3 5.7 3.6

1.9 1.2 −1.3

19.9 ± 1.5 190.5 ± 4.7 1965.6 ± 101.5

7.5 2.5 5.2

−0.7 −4.8 −1.7

Daidzin 50 500 5000

47.6 ± 3.3 485.6 ± 19.6 4939.7 ± 74.2

6.9 4.0 1.5

−4.8 −2.9 −1.2

50.7 ± 4.0 456.7 ± 32.4 4836.2 ± 193.8

7.9 7.1 4.0

1.4 −8.7 −3.3

Patulitrin 50 500 5000

45.6 ± 2.0 494.8 ± 11.8 4594.4 ± 209.4

4.3 2.4 4.6

−8.8 −1.0 −8.1

46.9 ± 3.4 468.5 ± 28.4 4552.8 ± 273.1

7.3 6.1 6.0

−6.3 −6.3 −8.9

Quercetagitrin 50 500 5000

48.9 ± 3.8 489.8 ± 23.6 4856.9 ± 326.9

7.8 4.8 6.7

−2.1 −2.0 −2.9

52.2 ± 5.1 491.3 ± 43.8 4885.1 ± 156.8

9.8 8.9 3.2

4.4 −1.7 −2.3

3-glucosylisorhamnetin 50 500 5000

46.1 ± 2.7 479.0 ± 29.0 4939.6 ± 169.1

6.0 6.1 3.4

−7.8 −4.2 −1.2

49.2 ± 4.2 466.3 ± 26.1 4649.9 ± 251.2

8.5 5.6 5.4

−1.7 −6.7 −7.0

performed using cone voltages with the higher cone voltage providing additional fragmentation date, and the valuable information was obtained concerning the presence and the nature of thirteen flavonoids. Consequently, the LC/MS analysis of these thirteen flavonoids was analysed by HPLC-ESI–MS, ionised in the negative mode in this study. A total ion chromatogram of the twelve flavonoids were restricted to those in which [M − H]− ions, and [M + HCOO− ]− ion was chosen to MS fragmentation analysis of daidzin. Mass spectra and the structures of thirteen flavonoids were presented in Fig. 2.

3.2. Validation of the method The specificity of the method after the optimisation of chromatographic conditions was tested by comparing the chromatograms with HPLC–MS of blank rat perfusate, blank perfusate spiked with thirteen flavonoids and IS, and actual perfusate samples after intraduodenal administration of XBXT-2. The representative chromatograms were shown in Fig. 3, thirteen flavonoids and IS were well separated from each other, and the retention time was approximately 8–30 min. There was no interference detected from

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

221

Table 3 Repeatability and recoveries of thirteen flavonoids in rat perfusate (mean ± SD, n = 5). Constituent

Spiked concentration(ng/mL)

Repeatability RSD (%)

Onpordin

20 200 2000 20 200 2000 50 500 5000 50 500 5000 20 200 2000 20 200 2000 20 200 2000 50 500 5000 20 200 2000 50 500 5000 50 500 5000 50 500 5000 50 500 5000

6.2 3.9 4.3 5.6 6.9 7.4 7.1 2.4 2.1 6.4 7.1 2.2 5.2 4.3 4.3 7.0 6.8 6.2 6.3 4.6 5.3 3.4 4.1 1.0 3.7 4.6 3.8 5.8 2.1 4.7 4.3 3.5 4.8 6.3 3.5 4.2 5.1 5.4 6.3

3 -O-methylorobol

Glycitein

Patuletin

Genistein

Luteolin

Quercetin

Nepitrin

Quercimeritrin

Daidzin

Patulitrin

Quercetagitrin

3-Glucosylisorhamnetin

endogenous substances under the established chromatographic condition. Table 1 listed linear equation and its correlation coefficient, linear range, LODs and LLOQs of these thirteen flavonoids determined. The linearity calibration curves were constructed by at least six assays of each reference compound. The regression equation was calculated in the form of y = ax + b, where y and x were the values of peak area and concentration of each reference compound, respectively. The high correlation coefficient values (r > 0.995) indicated good linearity between their peak areas (y) and investigated compound concentrations (x, ng/mL) in relatively wide concentration ranges. Besides, the limits of detection (LOD, S/N = 3) and the lower limit of quantification (LLOQ, S/N = 10) were in the rang of 3–15 ng/mL and 9–45 ng/mL, respectively. The results demonstrated a high sensitivity under the optimized chromatographic conditions. The analytical accuracy and precision data were shown in Table 2. Relative standard deviations (RSD) of repeatability, intraand inter-day precision were 1.0–7.4%, 0.9–8.2% and 1.1–9.8%, respectively. Assay accuracy, assessed by (RE), was found to rang from −9.8 to 4.4%. Those results indicated that the method provided adequate precision, accuracy and good repeatability. The extraction recoveries of thirteen flavonoids from sipked rat perfusate were determined at the concentrations of QC samples. A single step of liquid–liquid extraction with acetonitrile was proved to be simple, rapid and successful with a recovery ration over 85% for analytes at all tested concentrations (Table 3). The possibility of a matrix effect caused by ionization competition occuring between

Recovery (%) 89.2 85.3 92.1 90.1 86.9 91.6 87.5 89.4 90.2 87.1 88.6 89.7 89.5 85.8 89.8 89.6 86.2 88.9 90.5 86.3 89.7 90.1 88.7 90.2 89.3 87.2 90.3 88.5 89.3 89.5 89.1 90.3 90.2 89.5 90.5 90.9 88.9 89.6 90.6

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

2.1 1.2 3.3 2.8 3.9 4.8 4.2 2.7 1.7 2.9 3.5 3.2 1.0 2.8 2.0 3.7 5.3 4.7 7.4 10.1 2.4 11.1 10.2 6.9 6.1 3.3 2.5 4.1 3.5 2.1 3.7 2.6 3.6 4.2 2.8 4.3 5.3 3.5 2.2

RSD (%) 2.3 2.8 2.4 3.0 3.2 4.1 3.2 3.6 2.8 3.2 2.8 3.6 2.7 2.8 3.1 3.6 2.8 2.6 3.2 2.8 3.1 3.6 2.8 2.9 4.1 3.5 4.6 4.0 3.4 2.5 3.1 2.9 3.2 3.4 2.1 3.8 3.2 3.7 3.6

Fig. 4. Content of thirteen flavonoids in 0.2 g XBXT-2 and rat perfusate at 2 h after in situ vascularly perfused rat intestine preparation following intraduodenal administration of XBXT-2 at 0.2 g/rat.

222

Table 4 Stablility of thirteen flavonoids in rat perfusate (n = 5). Compound

Concentration (ng/mL)

Accuracy (%)

Room temperature (12 h) measured concentration (ng/mL)

Accuracy (%)

Stored at −20 ◦ C for 1 months measured concentration (ng/mL)

Accuracy (%)

Autosampler for 24 h

Accuracy(%)

19.3 ± 0.9 179.1 ± 6.9 1823 ± 16.3 18.3 ± 1.3 191.1 ± 13.2 1959 ± 80.4 45.6 ± 3.6 480.2 ± 19.1 4868 ± 77.3 48.2 ± 3.2 481.3 ± 12.9 4823.3 ± 162.7 19.2 ± 1.5 194.6 ± 11.8 1942.4 ± 63.6 19.2 ± 1.4 179.3 ± 13.7 1947.7 ± 120.3 18.4 ± 1.8 201.6 ± 12.1 1910.4 ± 57.7 50.4 ± 2.2 491.7 ± 14.1 4978 ± 57.9 20.1 ± 1.6 209.3 ± 10.8 1978.5 ± 71.3 47.1 ± 3.2 484.4 ± 18.7 4927.7 ± 90.6 45.5 ± 2.8 488.4 ± 17.2 4589.6 ± 256.5 48.6 ± 3.5 489.9 ± 26.7 4857.0 ± 319.3 45.8 ± 2.6 475.7 ± 2.5 4936.6 ± 162.9

96.5 89.6 91.2 91.5 95.6 102.1 91.2 96.0 97.4 96.4 96.3 96.5 96.0 97.3 97.1 96.0 89.7 97.4 92.0 100.8 95.5 100.8 98.3 99.6 100.5 104.7 98.9 94.2 96.9 98.6 91.0 97.7 91.8 97.2 98.0 97.1 91.6 95.1 98.7

19.0 ± 1.2 178.9 ± 7.7 1822 ± 16.6 18.1 ± 2.3 190.9 ± 12.9 1948 ± 92.6 45.5 ± 3.2 476.3 ± 17.5 4868 ± 77.2 47.5 ± 2.8 479.1 ± 11.9 4823.1 ± 161.2 18.9 ± 1.1 194.7 ± 12.1 1947.4 ± 67.6 19.9 ± 1.6 179.2 ± 14.7 1945.7 ± 113 17.9 ± 2.4 201.5 ± 16.1 1912.4 ± 54.7 49.6 ± 3.2 489.7 ± 14.7 4971 ± 58.9 19.61 ± 1.9 208.3 ± 11.1 1977.5 ± 71.6 47.6 ± 2.8 483.4 ± 20.4 4918.2 ± 89.9 44.5 ± 3.2 487.4 ± 17.2 4589.6 ± 266.5 47.9 ± 4.3 492.9 ± 23.7 4855.0 ± 309.3 45.9 ± 4.2 486.1 ± 3.2 4933.6 ± 161.9

95.0 89.5 91.1 90.5 95.5 97.4 91.0 95.3 97.4 95.0 95.8 97.2 94.5 97.4 97.4 99.5 89.6 97.3 89.5 100.8 95.6 99.2 97.9 99.4 98.5 104.2 98.9 95.2 96.7 98.4 89.0 97.5 91.8 95.8 98.6 97.1 91.8 97.2 98.7

18.9 ± 2.1 178.2 ± 11.6 1819 ± 16.3 17.9 ± 2.1 189.8 ± 12.7 1948 ± 87.4 44.8 ± 4.8 476.1 ± 10.2 4859 ± 78.1 47.2 ± 3.1 478.5 ± 11.3 4821.4 ± 159.9 18.8 ± 2.4 193.6 ± 11.1 1942.4 ± 65.6 19.1 ± 1.3 173.2 ± 14.7 1945.2 ± 115.2 17.3 ± 4.3 203.5 ± 16.1 1899.4 ± 52.7 49.2 ± 3.6 487.7 ± 14.2 4969 ± 57.9 18.9 ± 2.3 208.3 ± 11.3 1973.5 ± 70.6 46.6 ± 2.7 479.4 ± 19.6 4910.4 ± 89.5 43.5 ± 2.9 482.4 ± 19.2 4586.6 ± 255.5 48.4 ± 3.5 492.1 ± 24.7 4851.0 ± 312.3 46.9 ± 5.1 495.1 ± 2.0 4907.6 ± 172.1

94.5 89.1 90.9 89.5 94.9 97.4 89.6 95.2 97.8 94.4 95.7 96.4 94.0 96.8 97.1 95.5 86.6 97.3 86.5 101.8 95.0 98.4 97.5 96.4 94.5 104.1 98.7 93.2 95.9 98.2 87.0 96.48 91.7 96.8 98.4 97.0 93.8 99.0 98.1

18.3 ± 1.4 177.9 ± 7.2 1862 ± 16.6 17.6 ± 3.2 194.9 ± 11.9 1937 ± 82.6 44.2 ± 4.2 472.6 ± 16.5 4888 ± 72.2 46.6 ± 3.4 499.1 ± 10.9 4853.1 ± 151.2 18.2 ± 1.6 196.7 ± 10.1 1997.4 ± 59.6 20.9 ± 4.6 184.2 ± 20.7 1920.2 ± 123 18.5 ± 5.4 199.5 ± 14.3 1942.4 ± 44.7 50.6 ± 2.6 494.7 ± 13.9 4891 ± 64.4 19.78 ± 2.3 212.3 ± 20.2 1963.5 ± 67.4 49.6 ± 4.8 493.2 ± 17.8 4948.2 ± 79.9 45.8 ± 3.6 475.6 ± 16.9 4601.6 ± 206.5 48.8 ± 2.5 500.9 ± 20.7 4945.0 ± 299.3 44.9 ± 5.6 484.4 ± 4.2 4953.6 ± 168.7

91.5 89.0 96.8 88.0 97.45 97.2 88.4 94.5 97.8 93.2 99.8 97.1 91.0 98.35 97.3 104.5 92.1 96.2 92.5 99.75 98.4 101.2 98.9 97.8 98.9 106.2 95.9 99.2 98.6 99.0 91.6 95.1 92.0 97.6 100.2 98.9 89.8 96.9 99.1

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

20 200 2000 20 3 200 Omethylorobol 2000 50 Glycitein 500 5000 50 Patuletin 500 5000 20 Genistein 200 2000 20 Luteolin 200 2000 20 Quercetin 200 2000 50 Nepitrin 500 5000 Quercimeritrin 20 200 2000 50 Daidzin 500 5000 50 Patulitrin 500 5000 Quercetagitrin 50 500 5000 50 3500 Glucosylisorhamnetin 5000 Onpordin

Three freeze–thaw cycles Measured concentration (ng/mL)

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

223

Table 5 Contents and intestinal abosorption of thirteen flavonoids in XBXT-2 (n = 5).

a

Compound

Content in 0.2 g XBXT-2/mg

Onpordin 3 -O-methylorobol Glycitein Patuletin Genistein Luteolin Quercetin Nepitrin Quercimeritrin Daidzin Patulitrin Quercetagitrin 3-Glucosylisorhamnetin

0.960 0.608 0.113 0.050 0.052 1.456 1.276 0.896 0.274 0.562 1.060 0.266 0.340

± ± ± ± ± ± ± ± ± ± ± ± ±

0.092 0.048 0.010 0.006 0.003 0.131 0.082 0.057 0.022 0.029 0.059 0.020 0.023

Contentin perfusate/mg

Intestinal absorption /%

0.014 ± 0.012 0.013 ± 0.004 0.014 ± 0.007 0.022 ± 0.009 0.005 ± 0.002 0.004 ± 0.003 ND 0.065 ± 0.027 ND 0.063 ± 0.026 ND ND ND

1.5 ± 1.3 2.2 ± 0.6 12.2 ± 6.1 44.8 ± 17.8 9.2 ± 3.7 0.3 ± 0.2 ND 7.3 ± 3.0 ND 11.1 ± 4.8 ND ND ND

ND: not detected; Intestinal absorption% = content in perfusate/content in 0.2 g XBXT-2 × 100%.

the analytes and the endogenous co-eluents was evaluated at three concentrations, as illustrated above, in triplicate. Results from comparing the peak responses of the post-extraction spiked samples with that of the pure standards prepared in mobile phase intimated that a negligible matrix effect occurred in this method. The stability investigation of the thirteen flavonoids in the perfusate sample after storing in different conditions and processing procedures was evaluated by analysis of QC samples at high, medium and low concentration levels. The concentration variations found after three cycles of freezing and thawing, 12 h at the room temperature, −20 ◦ C for one month and 24 h in the auto-sampler (4 ◦ C) were with accuracy in rang of 86.5–106.2% (Table 4). The results also showed that thirteen flavonoids stock solutions and IS stock solution were stable under the room temperature for 12 h and −20 ◦ C for two weeks, the concentration deviated less than 10% from those in fresh preparation, indicating no significant substance loss during the storage. 3.3. Quantification and absorption Under the current chromatographic conditions described above, the selective ion monitoring was employed for the quantitative measurement, in which only the ion currents from selected masses was recorded in the instrument, contributed to achieving the quantification more specific and sensitive. The contents of thirteen flavonoids in XBXT-2 (0.2 g) and following intraduodenal adminstration of XBXT-2 at a single dose of 0.2 g/rat perfused 2 h with the in situ vascularly perfused rat intestine preparation were quantitatively determined by a internal standard method that was performed by preparing calibration curves for thirteen flavonoids standards using at least six calibration points, after establishing their linear response range in ESI–MS. As shown in Table 5 and Fig. 4,

the contents of these thirteen flavonoids in 0.2 g XBXT-2 were ranged from 0.050 to 1.456 mg. The in situ vascularly perfused rat intestine preparation used to investigate the intestinal absorption of flavonoids is seldom illustrated in the literature compared to Caco-2 cell monolayer model, in situ single-pass intestinal perfusion and in vitro everted gut sac model. Herbal complicated compounds are difficult to be detected due to lower concentration and interference from endogenous compounds in vivo, however, compared to in vivo pharmacokintic study, the experimental conditions of the in situ vascularly perfused rat intestine preparation, such as drug concentration, perfusate flow and the composition of perfusate, can be properly adjusted to survey the effect on the intestinal absorption of the tested drugs [14,18], it is an ideal approach for traditional Chinese medicines (TCM) to investigate the target constituents and enhance their detected concentration. Besides, the in situ vascularly perfused rat intestine preparation is advantageous of that the circulation and morphology of the tissues remain intact, which is precise assessment of the contribution of intestinal absorption of drug [14], and we had successfully investigated intestinal absorption properties of caffeic acid using this method [19]. In this study, we used this method to investigate intestinal absorption of these thirteen flavonoids of XBXT-2 and the in situ experiment date analysis for the intestinal absorption was calculated as follows: Intestinal absorption% = content in perfused/content in 0.2 g XBXT-2 × 100%. Onpordin, 3 -O-methylorobol, glycitein, patuletin, genistein, luteolin and quercetin were flavonoid aglycones, while nepitrin, 3-glucosylisorhamnetin, quercimeritrin, daidzin, patulitrin and quercetagitrin were glycosides. Eight flavonoids (onpordin, 3 -O-methylorobol, glycitein, patuletin, genistein, luteolin, nepitrin and daidzin) were detected in this study (Fig. 5), which indicated that aglycones were absorbed more efficiently than glycosides as the previously described [20]. The intestinal absorption of patuletin was about 44.8%, however, patulitrin was not detected in the perfusate. The in vitro studies show that flavonoid can be easily transformed into another one or more flavonoids, and patulitrin is glucopyranoside of patuletin, consequently, hypothesis that patulitrin was transformed into patuletin will be authenticated in further study. The contents of nepitrin and daidzin these two flavonoids in perfusate were 0.065 ± 0.027 mg and 0.063 ± 0.026 mg, respectively. Daidzin, as one of flavonoid glycosides, the intestinal absorption was 11.1%, which was similar to the previous document [21]. 5. Conclusion

Fig. 5. The intestinal absorption of eight flavonoids in XBXT-2 for 2 h after in situ vascularly perfused rat intestine preparation by intraduodenal administration of XBXT-2 at 0.2 g/rat (*Intestinal absorption (%) = content in perfusate/content in XBXT-2 × 100%).

In this study, HPLC-ESI–MS method showed performances of excellent sensitivity, good linearity of response and high precision for qualitative and quantitative analysis of these thirteen

224

M. Cen et al. / Journal of Pharmaceutical and Biomedical Analysis 115 (2015) 214–224

flavonoids. Additionally, the in situ vascularly perfused rat intestine preparation used was a very useful reference for the researches of intestinal absorption of complicated herbal compounds, and it provided a meaningful basis for better understanding the characteristic of intestinal absorption of thirteen flavonoids.The pharmacological activity of eight flavonoids detected after perfusion need to be intensively studied, especially for patuletin, nepitrin and daidzin these three flavonoids. The in vivo pharmacokinetics of XBXT-2 will be established as a topic of study in the future. Acknowledgements The work was financially supported by National Natural Science Foundation of China (81073141), Guangdong Provincial Natural Science Foundation (9152402301000007) and Guangdong Province “12-5”meical key subject. We are thankful to Professor Yimin Zhao, Yunfeng Li (Laboratory of phytochemistry, Beijing Institute of Pharmacology and Toxicology (Beijing, China)) for chemicals provided and technical assistances. References [1] C.Y. Cong, The paraphrase about the prescription in Fuxinjue Zangfu Yongyao Fayao, which are used in treating discorders of five internal organs, Mogao Caves Res. (China) 3 (2002) 93–99. [2] L. An, Y.Z. Zhang, N.J. Yu, X.M. Liu, N. Zhao, L. Yuan, Y.F. Li, Role for serotonin in the antidepressant-like effect of a flavonoid extract of Xiaobuxin-Tang, Pharmacol. Biochem. Be. 89 (2008) 572–580. [3] Y.Z. Zhang, Y.F. Li, N.J. Yu, L. Yuan, Y.M. Zhao, W.B. Xiao, Z.P. Luo, Antidepressant-like effects of the ethanolic extract of Xiaobuxin-Tang, a traditional Chinese herbal prescription in animal models of depression, Clin. Med. J. 120 (2007) 1792–1796. [4] Y.Z. Zhang, N.J. Yu, L. Yuan, L. An, Y.M. Zhao, Antidepressant effect of total flavonoids extracted from Xiaobuxing-Tang in forced swimming tests and learned helplessness in rats and mice, Chin. J. Pharmacol. Toxicol. 22 (2008) 1–8. [5] L. An, Y.Z. Zhang, N.J. Yu, H.X. Chen, Y.M. Zhao, W.B. Xiao, X.M. Liu, Y.F. Li, Antidepressant effect of total flavonoids extracted from Xiaobuxin-Tang on chronically mildly stressed model rats, Chin. J. Pharmacol. Toxicol. 22 (2008) 102–107. [6] E. Sepehr, G. Cooke, P. Robertson, G.S. Gilani, Bioavailability of soy isoflavones in rats Part I: application of accurate methodology for studying the effects of gender and source of isoflavones, Mol. Nutr. Food Res. 51 (2007) 799–812.

[7] E. Sepehr, G.M. Cooke, P. Robertson, G.S. Gilani, Effect of glycosidation of isoflavones on their bioavailability and pharmacokinetics in aged male rats, Mol. Nutr. Food Res. 53 (2009) S16–S26. [8] D. An, S. Wu, J. Wei, J. Yang, HPLC–MS analysis of quercetin and its metabolites in portal vein after intragastrical administration of quercetin in rats, Wei Sheng Yan Jiu 38 (2009) 417–419. [9] R. Mukai, Y. Fujikura, K. Murota, M. Uehara, S. Minekawa, N. Matsui, T. Kawamura, H. Nemoto, J. Terao, Prenylation enhances quercetin uptake and reduces efflux in Caco-2 cells and enhances tissue accumulation in mice fed long-term, J. Nutr. 143 (2013) 1558–1564. [10] B. Xia, Q. Zhou, Z. Zheng, L. Ye, M. Hu, Z. Liu, A novel local recycling mechanism that enchances enteric bioavailability of flavonoids and prolong their residence time in the gut, Mol. Pharm. 9 (2012) 3246–3258. [11] Y.A. Xu, G. Fan, S. Gao, Z. Hong, Assessment of intestinal absorption of vitexin-2-o-rhamnoside in hawthorn leaves flavonoids in rat using in situ and in vitro absorption models, Drug Dev. Ind. Pharm. 34 (2008) 164–170. [12] M.M. Skopec, A.K. Green, W.H. Karasow, Flavonoids have differential effects on glucose absorption in rats (Rattus norvegicus) and American robins (Turdis migratorius), J. Chem. Ecol. 36 (2010) 236–243. [13] A. Lei, Y.Z. Zhang, X.M. Liu, N.J. Yu, H.X. Chen, N. Zhao, L. Yuan, Y.F. Li, Total flavonoids extracted from xiaobuxing-tang on the hyperactivity of hypothalamic-pituitary-adrenal axis in chronically stressed rats, Evid. Based Complement. Alternat. Med. 2011 (2011) 367619. [14] K.S. Pang, W.F. Cherry, E.H. Ulm, Disposition of enalapril in the perfused rat intestine-liver preparation: absorption, metabolism and first-pass effect, J. Pharmacol. Exp. Ther. 233 (1985) 788–795. [15] R. Kostiainen, T.J. Kauppila, Effect of eluent on the ionization process in liquid chromatography-mass spectrometry, J. Chromatogr. A 1216 (2009) 685–699. [16] L. Wang, Y. Sun, F. Du, W. Niu, T. Lu, J. Kan, F. Xu, K. Yuan, T. Qin, C. Liu, C. Li, ‘LC-electrolyte effects’ improve the bioanalytical performance of liquid chromatography/tandem mass spectrometric assays in supporting pharmacokinetic study for drug discovery, Rapid Commun. Mass Spectrom. 21 (2007) 2573–2584. [17] E. de Rijke, P. Out, W.M. Niessen, F. Ariese, C. Gooijer, U.A. Brinkman, Analytical separation and detection methods for flavonoids, J. Chromatogr. A 1112 (2006) 31–63. [18] H. Hirayama, X. Xu, K.S. Pang, Viability of the vascularly perfused, recirculating rat intestine and intestine-liver preparations, Am. J. Physiol. 257 (1989) G249–258. [19] S.J. Wang, J. Zeng, B.K. Yang, Y.M. Zhong, Bioavailability of caffeic acid in rats and its absorption properties in the Caco-2 cell model, Pharm. Biol 52 (2014) 1150–1157. [20] T. Izumi, M.K. Piskula, S. Osawa, A. Obata, K. Tobe, M. Saito, S. Kataoka, Y. Kubota, M. Kikuchi, Soy isoflavone aglycones are absorbed faster and in higher amounts than their glucosides in humans, J. Nutr. 130 (2000) 1695–1699. [21] R.A. King, Daidzein conjugates are more bioavailable than genistein conjugates in rats, Am. J. Clin. Nutr. 68 (1998) 1496S–1499S.