Ultra-high performance liquid chromatography coupled with time-of-flight mass spectrometry screening and analysis of potential bioactive compounds from traditional chinese medicine Kai-Xin-San, using a multivariate data processing approach and the MetaboLynx tool

Abstract

Traditional Chinese Medicine (TCM) has been used in clinical practice for several thousand years, with an indispensable role in prevention and treatment of disease using multiple ingredients. Kai-Xin-San (KXS) is a TCM formula consisting of four herbs, Ginseng Radix, Polygalae Radix, Poria and Acori Tatarinowii Rhizoma, which has been used to treat Alzheimer's disease, depression and Parkinson's disease. However, the constituents absorbed into blood after oral administration of KXS remain unknown. Here, a sensitive and rapid method using ultra performance liquid chromatography/quadrupole time-of-flight mass spectrometry (UPLC/Q-TOF-MS/MS) combined with a multivariate data processing approach (Mdpa) and MetaboLynx software tool was used for analysis and identification of bioactive components and their metabolites in rat plasma following oral administration of KXS. A hyphenated MS/MS analyzer was used for determination of accurate fragment ions and mass spectrometric fragmentation mechanisms, and enhanced data acquisition. Metabolite elucidation was performed using UPLC/Q-TOF-MS/MS coupled with MetaboLynx tool. With the established method, a total of 49 peaks were tentatively characterized in vivo based on MS and MS/MS data, and comparison with available databases. Of the 49 compounds tentatively characterized in rat plasma, seven metabolites were detected and identified by comparing their fragmentation patterns using MetaboLynx software tool. On the basis of the chromatographic peak area, the glucuronidated conjugates were identified as major metabolites. This work demonstrated the potential of the UPLC/Q-TOF-MS/MS with Mdpa and MetaboLynx for rapid, simple, reliable and automated identification of metabolites of herbal medicine.

---

1. Introduction

Traditional Chinese Medicine (TCM) prescriptions have been used widely in antibacterial, antifungal, anticancer, antiviral, anti-inflammatory and other pharmacological applications.¹ The pharmacological effects of TCM prescription are expressed through multi-active chemical constituents.² The sources of these effective constituents are complicated, including the originally prescribed compounds and the products of processed decoction. For screening and analysis of effective compounds in TCM, conventional phytochemical approaches that isolate and identify individual components one by one can be used to search the lead compounds in herbal medicine, but this strategy is time-consuming, and labor intensive.^3,4 In view of this, there is a need for novel methods that could overcome all these limitations, to reveal the complex compounds in TCM prescription.

Recently, high-resolution ultra-high-performance liquid chromatography equipped with electrospray ionization quadrupole time-of-flight mass spectrometry (UPLC/Q-TOF-MS/MS) has been used to identify and analyze the chemical compounds in TCM. In particular, it provides accurate precursor and/or product ion information with a mass error of less than 5 ppm, which substantially enhances the constituent characterization reliability.⁵ A new method based on UPLC/Q-TOF-MS/MS combined with a multivariate data processing approach (Mdpa) has been introduced for screening and analysis of herbal medicine.^6–8 With the advantages of short analysis time as well as high accuracy of mass values, UPLC/Q-TOF-MS/MS with Mdpa provides an efficient strategy for rapid screening, and identification of the active components in herbal medicine.⁹ Substantial progress has been made using MetaboLynx software to further clarify the metabolic profile. With key parameters carefully set, MetaboLynx is able to show the presence of a wide range of metabolites with only a limited requirement for manual intervention and data interpretation time.

Kai-Xin-San (KXS), first recorded in the Chinese ancient medical prescription book Qian-Jin-Yao-Fang around 1300 years ago, is a TCM prescription consisting of four herbs, Ginseng Radix, Polygalae Radix, Poria and Acori Tatarinowii Rhizoma.¹⁰ In Asian countries, it has been used to treat Alzheimer's disease, Parkinson's disease, etc.^11–13 Recently, chemical analysis study of KXS identified polygalaxanthone III, ginsenoside Rb1, ginsenoside Rd, ginsenoside Re, and ginsenoside Rg1 in the plasma of rat after oral administration of KXS, using ultra-fast liquid chromatography with tandem mass spectrometry .¹⁴ However, to date, the systematic chemical profile of the effective ingredients in the KXS prescription is not fully known. In this paper, UPLC/Q-TOF-MS/MS was combined with Mdpa to provide a reliable method for carrying out comprehensive characterization of the major potential bioactive constituents and metabolites of KXS. To our best knowledge, this is the first systematic study screening the bioactive components in KXS.

2. Material and methods

2.1 Chemicals and materials

Methanol and acetonitrile (HPLC grade) were purchased from Merck (Darmstadt, Germany); leucine enkephalin was purchased from Sigma-Aldrich (MO, USA); deionized water (18.2 MΩ) was further purified using a Milli-Q system (Millipore, Billerica, USA); HPLC grade formic acid was purchased from Kermel Chemical Reagent Co., Ltd (Tianjin, China); formic acid was purchased from (DIKMA, USA). Ginseng Radix, Poria, Polygalae Radix and Acori Tatarinowii Rhizoma were purchased from Harbin Tongrentang Drug Store (Harbin, China), and authenticated by Prof. Xijun Wang, Department of Pharmacognosy of Heilongjiang University of Chinese Medicine. Voucher specimens were deposited at the authors' laboratory.

2.2 Preparation of KXS samples

The KXS was prepared as follows: Ginseng Radix, Poria, Polygalae Radix and Acori Tatarinowii Rhizoma in proportions 3 : 3 : 2 : 2, were ground into crude powders, mixed, and then reflux extracted in a rotary evaporator with six times volume of 70% ethanol for 2 h twice, then the filtrate was freeze-dried. The freeze-dried powder was dissolved in water, ultrasonic for 10 min, to make a solution of 0.08 g ml⁻¹ concentration.

2.3 Animal handling

Male Wistar rats (300 ± 10 g) were purchased from the Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The rats were housed in an animal room (24 ± 2 °C, 40% relative humidity). A 12 h dark/light cycle was set, and the rats were given water and fed normal food for 1 week before the experiment. All rats were randomly divided into two groups, a control group and a dosed group, of three rats each. Rats were orally administered KXS extract at a dose of 1 ml per 100 g body weight. The control group was orally administrated with an equivalent volume of distilled water. After 60 min, the rats were anaesthetized by intraperitoneal injection of 3% pentobarbital sodium (0.3 ml per 100 g body weight).

2.4 Preparation of plasma samples in vivo

Blood samples were collected from hepatic portal vein at 1 h after administration and the rats were sacrificed. Then, the plasma was separated immediately by centrifuging at 13 000 rpm for 10 min at 4 °C. All samples were stored at −80 °C until analysis. 8 ml methanol was added to 2.0 ml blood samples and then vortexed for 30 s, ultrasound and centrifuged at 13 000 rpm for 10 min at 4 °C. The supernatant was then dried under a stream of nitrogen gas at 45 °C. Each dried sample was reconstituted in 200 μl methanol prior to analyses, centrifuged at 13 000 rpm for 10 min at 4 °C. The sample was filtered through a 0.22 μm membrane, and a 5 μl aliquot was injected for UPLC/MS analysis.

2.5 Ultra-high performance liquid chromatography condition

Chromatographic separation for samples was performed using a Waters Acquity ultra performance LC system controlled by MassLynx software (Version 4.1). Separation was performed with an acquity UPLC HSS C₁₈ column (100 mm × 2.1 mm i.d., 1.8 μm particle size, Waters Corporation, Milford, MA, USA) maintained at 40 °C. A gradient with eluent A (HCOOH : H₂O = 0.1 : 100, v/v) and B (HCOOH : CH₃CN = 0.1 : 100, v/v) was used at a flow rate of 0.3 ml min⁻¹. The linear elution gradient program was used as follows: 0–1.0 min, 2–20% A; 1.0–4.0 min, 20–21% A; 4.0–6.5 min, 21–30% A; 6.5–9.5 min, 30–35% A; 9.5–14.0 min, 35–50% A; 14.0–16.0 min, 50–52% A; 16.0–16.5 min, 52–82% A; 16.5–19.0 min; 82–83% A; 19.0–19.5 min, 83–100% A; 19.5–20.0 min, 100–2% A; 20.0–21.0 min, 2% A. The sample-tray temperature was kept at 4 °C.

2.6 Mass spectrometry condition

The MS instrument was a Waters Synapt QTOF/MS (Waters Corp., Milford, MA, USA). Ionization was performed in both positive and negative ion modes. The MS source temperature was set at 110 °C, and the desolvation temperature was set at 300 °C with desolvation gas flow at 600 L h⁻¹. The capillary voltage was 3 kV. The mass spectra were recorded across the range of m/z 100 to 2700 Da, with accurate mass measurement of all mass peaks. Ar gas was used as collision gas at a pressure of 0.2 Mpa. The collision energy was set as 10–30 eV for low-energy scans, and 30–50 eV for high-energy scans. The MS/MS experiment was performed for major metabolites to obtain additional information from product ions. Leucine-enkephalin was used as the lock mass generating a reference ion for accurate mass acquisition. The instrument was controlled by MassLynx 4.1 (Waters Corp.).

2.7 Multivariate data processing approach

The raw data of all tested samples were analyzed by MarkerLynx and EZINFO software (Waters Corp., USA) for preliminary phytochemical screening. The three-dimensional data comprising peak number, sample name and ion intensity were analyzed by principal component analysis (PCA) and orthogonal partial least-squared discriminant analysis (OPLS-DA) using EZINFO 2.0 software. To further confirm the structure and the source of the metabolites, all data were put into the MassFragment software tool. The ions present in the dosed group and absent in the control group were extracted with the help of the VIP plot of OPLS-DA, and these ions were identified with a combination of elemental composition tool and MS/MS fragment mass spectra.

2.8 MetaboLynx processing approach for metabolites analysis

Expected and unexpected metabolites were found using the MetaboLynx software (Waters Corporation, Milford, USA) by automatically comparing MS(E) data from the sample and control. The metabolites and their different fragmentation pathways were reliably characterized by accurate MS/MS spectra, with an extensive list of potential biotransformation reactions consisting of deglycosylation, ring cleavage, dehydroxylation, decarbonylation, methylation, sulfation, hydrogenation, and hydroxylation. MetaboLynx in combination with the elemental composition of the substrate molecules generated a series of extracted ion chromatograms. The extracted ion chromatograms which present in the analyte were evaluated relative to the control sample by comparing their retention times and peak areas. The method parameters for data processing were set as follows: retention time range 0.1–20.0 min, mass range 50–1500 Da, retention time tolerance 0.1 min, mass tolerance 0.05 Da, noise elimination level 5, and peak intensity threshold 50. From any different peaks found in the analyte compared with the control sample, only those peaks with area ratio of analyte to control sample above 3 were considered to be the compound-related metabolites.

3. Results and discussion

3.1 Optimization of chromatographic and mass spectrometry conditions

UPLC and MS conditions were optimized for better detection. UPLC chromatograms with good separation, different mobile-phase compositions were screened (data not shown), and it was found that acetonitrile and formic acid was the most suitable eluting solvent system. The mobile phase played an important role in achieving good chromatographic behavior and appropriate ionization. A mixture of 0.1% v/v aqueous formic acid and acetonitrile was finally chosen as the preferred mobile phase because it produced the desired separation and acceptable tailing factors within the 20 min run time. In the course of optimizing separation conditions, mobile phase, gradient program, column temperature, and detection wavelength were investigated. The final results showed that best resolution, shortest analysis time, and lowest pressure variations were achieved with a gradient elution mode composed of acetonitrile (containing 0.1% formic acid, eluent A) and water (containing 0.1% formic acid, eluent B) programmed as follows: 0–1.0 min, 2–20% A; 1.0–4.0 min, 20–21% A; 4.0–6.5 min, 21–30% A; 6.5–9.5 min, 30–35% A; 9.5–14.0 min, 35–50% A; 14.0–16.0 min, 50–52% A; 16.0–16.5 min, 52–82% A; 16.5–19.0 min; 82–83% A; 19.0–19.5 min, 83–100% A; 19.5–20.0 min, 100–2% A; 20.0–21.0 min, 2% A. The most efficient way to obtain rapid analysis time, while generating high column efficiency and resolution, is by utilizing a small particle size column. Therefore, chromatographic analysis was executed through a 1.8 μm particle size HSS C₁₈ column. Considering sensitivity and resolution, the ultimate flow rate was optimized at 0.3 ml min⁻¹.

To reduce the column pressure resulting from a higher flow rate, the liquid chromatography column temperature was set at 40 °C. TOF/MS, which enables peak deconvolution of analytes of differing m/z after ionization, was utilized to achieve fast scanning characteristics for rapid chromatographic separations. For the MS conditions, a series of parameters including the ion spray voltage, turbo spray temperature, declustering potential, collision energy, nebulizer gas, heater gas and curtain gas were all optimized. Both positive and negative ion modes were employed to screen as many potential compound signals as possible. The negative ion mode provided higher signal intensity and the ability to detect more peak signals, perhaps because of the existence of some compounds easy to ionize in the negative mode. The negative ion mode also has increased sensitivity to the signals of the common constituents compared with the positive ion mode. Some ions were observed only in the negative ion mode, which is helpful for the structural determination. As a result, the negative ion mode was used. The base peak ion (BPI) chromatogram in rat plasma after oral administration of KXS analyzed by UPLC/Q-TOF-MS/MS in positive ion mode and negative ion mode is shown in Fig. 1.

Fig. 1 ESI base peak ion (BPI) chromatogram of the KXS analyzed by UPLC/QTOF MS. Positive ion mode (A, KXS serum; B, control serum); negative ion mode (C, KXS serum; D, control serum). (The peak numbers are displayed in Table S1†).

3.2 Multivariate statistical analysis

Mdpa converts the multidimensional data space into two matrices known as scores and loadings. In the PCA score plot (Fig. 2A and B), each coordinate represents a sample, and it was observed that the determined samples were clearly divided into two clusters. For analysis of the differences in chemical compositions between dosed rat plasma and blank plasma samples, OPLS-DA, a supervised multivariate analysis method, was performed. The ions of interest present only in the dosed group and absent in the control group were extracted easily by VIP plot of OPLS-DA (Fig. 2C and D). In the VIP plot, each point represents an ion RT-m/z pair; the X-axis represents variable contribution, and the further the ion RT-m/z pair point departs from zero, the more the ion contributes to the difference between the dosed rat plasma and blank plasma samples. Using UPLC/Q-TOF-MS/MS and database-matching techniques, a total of 42 compounds were characterized tentatively. As demonstrated above, 42 ions of interest in blood samples were extracted and identified in vivo, 21 of which were detected by pattern recognition methods in positive ion mode and others were detected in negative mode. Information regarding the 42 compounds, such as the t_R (min), identity, observed m/z values, mass error, molecular formula, and botanical source, MS/MS data is listed in Table S1.†

Fig. 2 Multivariate statistical analysis of constituents in serum dosed with KXS. PCA scores plot (A, positive ion mode; B, negative ion mode) of dosed rat serum and control rat serum; the VIP plot of OPLS-DA of dosed rat serum and control rat serum (C, positive ion mode; D, negative ion mode). bc: Controls.

3.3 Metabolites identified using MetaboLynx tool

Using MetaboLynx software, peaks present in the analyte were evaluated relative to the control samples by comparing their retention times with post-acquisition analyses, MS spectra and peak area. When a peak was included as a metabolite, the peak area in the analyte had to be at least five times greater than that of the control. After post-acquisition data processing of plasma samples using a MetaboLynx program, the results demonstrated the vast majority of the metabolites. The extracted ion chromatogram of plasma samples with MetaboLynx tool is presented in Fig. 3, and the products were well separated using the developed UPLC method. Through a comprehensive analysis of the peaks, we found seven peaks compared with blank samples: one parent constituent and three metabolites which indicated that glucuronide conjugation, decarbonylation, and deethylation were the major metabolic pathways of constituents in vivo. Table S2† lists the detailed information of these metabolites, including the retention times, proposed elemental compositions, and the characteristic fragment ions. Taking an example, ESI negative ion mode gave an [M − H]⁻ ion at m/z 567.1332, indicating the formula C₂₅H₂₈O₁₅ (calc. 568.1428). According to its retention time and MS data, the compound was identified as polygalaxanthone III (Fig. 4). Other compounds could be similarly characterized according to the above methods.

Fig. 3 The extracted ion chromatogram of serum treated accordingly to experimental schemes with MetaboLynx tool. Extracted ion chromatograms of metabolites with MetaboLynx in positive ion mode (A), and extracted ion chromatograms of metabolites with MetaboLynx in negative ion mode (B).

Fig. 4 The mass spectra of polygalaxanthone III in negative mode. MS^E spectra and fragment assignment at high collision energy (A); the mass spectrum at low collision (B); proposed fragmentation pathways of polygalaxanthone III (C).

Previous studies examined the uptake of ginsenosides Rg1, Re and Rb1 to the brain after oral administration of KXS preparation.¹⁵ The results suggested that the presence of a delivery agent in the KXS formula promoted initial absorption of ginsenosides Rg1 and Re in the gastrointestinal tract, but is unlikely to have affected the brain-to-plasma AUC ratios. A study showed that KXS aqueous extract significantly ameliorated depressive symptoms, including the reduced preference index and prolonged latency to novelty-suppressed feeding.¹⁶ Simultaneously, KXS exerts its antidepressant-like and nootropic effects by modulating the HPA axis, monoamine neurotransmitter and cholinergic systems. The mechanism of its antidepression action was preliminarily explored.¹⁷ The expressions of the molecular bio-markers relating to depression in rat brains were altered by treatment with KXS. It is suggested that the anti-depressant-like action of KXS might be mediated by an increase of neurotransmitters and expression of neurotrophic factors and corresponding receptors in the brain.

In this paper, integrated UPLC/Q-TOF-MS/MS with Mdpa and MetaboLynx software tools was performed to screen the bioactive compounds from KXS, one of the most well-known TCM prescriptions, which is clinically effective for Alzheimer's disease and Parkinson's disease. Although KXS has been used widely in clinics, the bioactive ingredients and metabolites of KXS are not fully known. The bioactive compounds in KXS could be identified by comparative analysis of the chemical profiles of control plasma and dosed plasma, and then identified based on their MS and MS/MS spectra. The data sets of retention time (RT)-m/z pairs, ion intensities and sample codes were further processed with Mdpa to generate a VIP plot. Furthermore, the established method allowed detection of low-abundance metabolites along with their structural elucidation. The Mdpa method indicated that 42 components of the KXS were absorbed into the rat's body. In the present study, metabolites in the in vivo system of Wistar rats were identified and elucidated by UPLC/Q-TOF-MS/MS. The automated data analysis (MetaboLynx) was developed and validated for simultaneous identification and determination of metabolites produced by the body in vitro. Expected and unexpected metabolites were detected by MetaboLynx software, which can automatically compare MS data from the sample and control.^18–20 In addition, seven components might be metabolites of other components in the KXS. From these results, it can be concluded that the proposed method could be used to simultaneously analyze and screen the multiple absorbed bioactive constituents and metabolites in TCM. The present method provides a quick and reliable analytical tool to investigate the prototype components absorbed and metabolite metabolism of TCM. To illustrate the dynamic metabolism profiles of absorbed constituents and metabolites in vivo, further research is required on pharmacokinetics to understand the absorption, distribution, and excretion of these components.

4. Conclusion

In this work, we demonstrated, for the first time, use of UPLC/Q-TOF-MS/MS coupled with automated Mdpa and MetaboLynx data analysis for structural characterization of global compounds in rat plasma following oral administration of KXS. After the automated analysis, a total of 49 compounds was characterized tentatively. The proposed method was appropriate for rapid screening and identification of absorbed and metabolic components of KXS. Additionally, we concluded that the 49 compounds including 42 components from KXS and seven metabolites were simultaneously determined, providing essential data for further pharmacological studies of KXS. This approach yielded a series of potential bioactive compounds in a high-throughput manner, which is useful for drug discovery of compounds from KXS. This method could be developed as an integrated approach for screening and identifying the active ingredients in TCM prescription and provides useful chemical information for further pharmacology and active mechanism research.

Conflict of interest

The authors declare no competing financial interests.

Acknowledgements

This work was supported by grants from the Key Program of Natural Science Foundation of State (Grant no. 90709019, 81173500, 81373930, 81302905, 81102556, 81202639), National Key Technology Research and Development Program of the Ministry of Science and Technology of China (Grant no. 2011BAI03B03, 2011BAI03B06, 2011BAI03B08), and National Key Subject of Drug Innovation (Grant no. 2009ZX09502-005).

References

1. A. Zhang, H. Sun and X. Wang, Potentiating therapeutic effects by enhancing synergism based on active constituents from traditional medicine, Phytother. Res., 2014, 28(4), 526–533.
2. X. Wang, A. Zhang, P. Wang, H. Sun, G. Wu, W. Sun, H. Lv, G. Jiao, H. Xu, Y. Yuan, L. Liu, D. Zou, Z. Wu, Y. Han, G. Yan, W. Dong, F. Wu, T. Dong, Y. Yu, S. Zhang, X. Wu, X. Tong and X. Meng, Metabolomics coupled with proteomics advancing drug discovery toward more agile development of targeted combination therapies, Mol. Cell. Proteomics, 2013, 12(5), 1226–1238.
3. H. Wang, H. Sun, A. Zhang, Y. Li, L. Wang, H. Shi, X. L. Dizou and X. Wang, Rapid identification and comparative analysis of the chemical constituents and metabolites of Phellodendri amurensis cortex and Zhibai dihuang pill by ultra-performance liquid chromatography with quadrupole TOF-MS, J. Sep. Sci., 2013, 36(24), 3874–3882.
4. A. Zhang, H. Sun and X. Wang, Recent advances in natural products from plants for treatment of liver diseases, Eur. J. Med. Chem., 2013, 63, 570–577.
5. A. Zhang, D. Zou, G. Yan, Y. Tan, H. Sun and X. Wang, Identification and characterization of the chemical constituents of Simiao Wan by ultra high performance liquid chromatography with mass spectrometry coupled to an automated multiple data processing method, J. Sep. Sci., 2014, 37(14), 1742–1747.
6. H. Cao, A. Zhang, F. M. Zhang, Q. Q. Wang, H. Zhang, Y. H. Song, Y. Zhou, H. Sun, G. L. Yan, Y. Han and X. Wang, Ultra-performance liquid chromatography tandem mass spectrometry combined with automated MetaboLynx analysis approach to screen the bioactive components and their metabolites in Wen-Xin-Formula, Biomed. Chromatogr., 2014, 28(12), 1774–1781,  DOI:10.1002/bmc.3220.
7. D. Zou, A. Zhang, G. Yan, Y. Tan, H. Sun and X. Wang, UPLC/MS coupled with a dynamic multiple data processing method for comprehensive detection of chemical constituents of herbal San-Miao-Wan formula, Anal. Methods, 2014, 6(9), 2848–2854.
8. H. Sun, W. Dong, A. Zhang, W. Wang and X. Wang, Pharmacokinetics study of multiple components absorbed in rat plasma after oral administration of Stemonae radix using ultra-performance liquid-chromatography/mass spectrometry with automated MetaboLynx software analysis, J. Sep. Sci., 2012, 35(24), 3477–3485.
9. G. L. Yan, A. H. Zhang, H. Sun, Y. Han, H. Shi, Y. Zhou and X. J. Wang, An effective method for determining the ingredients of Shuanghuanglian formula in blood samples using high-resolution LC-MS coupled with background subtraction and a multiple data processing approach, J. Sep. Sci., 2013, 36(19), 3191–3199.
10. B. Hu, H. Zhang, X. Meng, F. Wang and P. Wang, Aloe-emodin from KXS (Rheum rhabarbarum) inhibits lipopolysaccharide-induced inflammatory, J. Ethnopharmacol., 2014, 153(3), 846–853.
11. Y. Cao, Y. Hu, P. Liu, H. X. Zhao, X. J. Zhou and Y. M. Wei, Effects of a Chinese traditional formula Kai Xin San (KXS) on chronic fatigue syndrome mice induced by forced wheel running, J. Ethnopharmacol., 2012, 139(1), 19–25.
12. Q. Guan, S. Liang, Z. Wang, Y. Yang and S. Wang, ¹H NMR-based metabonomic analysis of the effect of optimized KXS aglycone on the plasma and urine metabolic fingerprints of focal cerebral ischemia-reperfusion rats, J. Ethnopharmacol., 2014, 154(1), 65–75.
13. J. Cao, Z. Wang, Y. Zhang, F. Qu, L. Guo, M. Zhong, S. Li, H. Zou, J. Chen and X. Wang, Identification and characterization of the related immune-enhancing proteins in crab Scylla paramamosain stimulated with KXS polysaccharides, Mol. Immunol., 2014, 57(2), 263–273.
14. S. Y. Wei, W. X. Yao, W. Y. Ji, J. Q. Wei and S. Q. Peng, Qualitative and quantitative analysis of anthraquinones in KXSs by high performance liquid chromatography with diode array detector and mass spectrometry, Food Chem., 2013, 141(3), 1710–1715.
15. W. Wang, Q. P. Liao, L. H. Quan, C. Y. Liu, Q. Chang, X. M. Liu and Y. H. Liao, The effect of Acorus gramineus on the bioavailabilities and brain concentrations of ginsenosides Rg1, Re and Rb1 after oral administration of Kai-Xin-San preparations in rats, J. Ethnopharmacol., 2010, 131(2), 313–320.
16. H. Dang, L. Sun, X. Liu, B. Peng, Q. Wang, W. Jia, Y. Chen, A. Pan and P. Xiao, Preventive action of Kai Xin San aqueous extract on depressive-like symptoms and cognition deficit induced by chronic mild stress, Exp. Biol. Med., 2009, 234(7), 785–793.
17. K. Y. Zhu, Q. Q. Mao, S. P. Ip, R. C. Choi, T. T. Dong, D. T. Lau and K. W. Tsim, A standardized Chinese herbal decoction, kai-xin-san, restores decreased levels of neurotransmitters and neurotrophic factors in the brain of chronic stress-induced depressive rats, J. Evidence-Based Complementary Altern. Med., 2012, 2012, 149256.
18. S. Yang, Y. Li, X. Cao, D. Hu, Z. Wang, Y. Wang, J. Shen and S. Zhang, Metabolic pathways of T-2 toxin in in vivo and in vitro systems of Wistar rats, J. Agric. Food Chem., 2013, 61(40), 9734–9743.
19. H. Wu, L. Li, J. Shen, Y. Wang, K. Liu and S. Zhang, In vitro metabolism of cyadox in rat, chicken and swine using ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry, J. Pharm. Biomed. Anal., 2012, 67–68, 175–185.
20. F. Hernández, S. Grimalt, O. J. Pozo and J. V. Sancho, Use of ultra-high-pressure liquid chromatography-quadrupole time-of-flight MS to discover the presence of pesticide metabolites in food samples, J. Sep. Sci., 2009, 32(13), 2245–2261.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra08992h

---