, Volume 8, Issue 1, pp 99–108 | Cite as

Profiling of the charged metabolites of traditional herbal medicines using capillary electrophoresis time-of-flight mass spectrometry

  • Keiko Iino
  • Masahiro Sugimoto
  • Tomoyoshi Soga
  • Masaru Tomita
Original Article


The quantification of a small number of bioactive components in herbal medicines is often inadequate when attempting to elucidate a medicine’s biological effects. Despite rapid advances in analytical technologies, obtaining comprehensive metabolomic profiles of herbal medicines remains difficult, due to the complexity of natural product mixtures. Toki-Shakuyaku-San is a Chinese medicine used widely to treat gynecological and obstetric disorders, such as infertility, dysmenorrhea, toxemia during pregnancy and neural dysfunction. It consists of Angelica acutiloba Radix (Toki), Cnidium officinale Rhizoma (Senkyu), Paeonia lactiflora Radix (Shakuyaku), Atractylodes lancea Rhizoma (Sojutsu), Alisma orientale Rhizoma (Takusha) and Poria cocos Hoelen (Bukuryo). To elucidate the composition of these herbal medicines individually, we conducted non-targeted profiling analyses of extracts of these herbs using capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS), which allows the simultaneous quantification of hundreds of charged metabolites. In total, 737 ± 183.1 (average ± SD) metabolite-derived features were observed, and of these, 119 metabolites were identified. Score plots of principal component analysis (PCA) showed a clear cluster including Shakuyaku, Bukuryo, and Sojutsu, while the other three herbs were distributed over PCA spaces. Loading plots revealed that amino acids and shikimate-derived alkaloids were the predominant metabolite constituents. Hierarchical clustering analysis revealed that few clusters overlapped in the herbal medicines tested. This report is the first demonstration of the characterization of a herbal medicine using large-scale metabolomic analysis, which is complementary to traditional quality control methods.


Capillary electrophoresis time-of-flight mass spectrometry Herbal medicine Charged metabolite Metabolomic profiling 

1 Introduction

Herbal medicines have been used as therapeutic agents for several thousands of years in Asian countries (Wang and Ren 2002; Wang et al. 2009). The evaluation of a herbal medicine’s safety and efficacy is required before its clinical use as a complementary and alternative medicine (CAM) in western countries (Miller et al. 2004; Murray and Rubel 1992; Calixto 2000; Zhang et al. 2010). The conventional pharmacological approach to characterize herbal medicines focuses on the identification and quantification of a single or several bioactive components (Murray and Rubel 1992; Liu et al. 2008), however, the complex properties of herbal medicines often render such an approach inadequate (Wang and Ren 2002; Lao et al. 2009; Liang et al. 2009; Chan 2003). The quantitative analysis of the complete metabolite profile of herbal medicines would therefore be of significant value (Liu et al. 2008).

Recent advances in the development of mass spectrometry (MS)-based profiling techniques have made a significant contribution to the study and quantification of the metabolome, the complete range of low molecular weight compounds of a natural sample, which is a field known as metabolomics. MS is typically used in combination with a separation system, such as liquid chromatography (LC), gas chromatography (GC) or capillary electrophoresis (CE). The selection of the separation system to be used is based on the chemical properties of the target compounds, since there is no single analytical methodology currently available capable of profiling complete metabolome. GC–MS and LC–MS are used routinely for targeted analysis (Liang et al. 2009) and are considered to be mature techniques (Weng and Jin 2002). GC–MS is limited to profiling volatile molecules, such as essential oils (Li et al. 2003; Gong et al. 2001), while LC–MS is suitable for separating and detecting more diverse molecules, such as flavonoids, glycosides, organic acids, saponins and lipids (Lao et al. 2009; Ma et al. 2007; Ren et al. 2008). However, the separation conditions used for LC should be optimized for individual molecular classes, which limits the number of metabolites simultaneously detected. CE coupled MS, in particular CE time-of-flight MS (CE-TOFMS), has been demonstrated to have high separation capability and sensitivity for the profiling of charged metabolites, including primary metabolites such as amino acids, amines, organic acids and nucleic acids (Soga et al. 2003, 2006; Monton and Soga 2007). CE alone has been used for the targeted analysis of alkaloids and flavonoids (Ganzera 2008; Lao et al. 2009; Hurtado-Fernandez et al. 2010) and used for quality control (QC) of herbal medicines (Ganzera 2008). LC-MS-based profiling showed variance attributed to harvesting region and processing protocol (Xie et al. 2008; Chan et al. 2007). Our CE–MS-based profiling method is also important for QC of herbal medicines, since charged metabolites are the dominant species dissolved in water extracts (Lao et al. 2009).

Toki-Shakuyaku-San, or Danggui-Shaoyao-San, is a Chinese herbal medicine used widely for the therapy of hematopoiesis and menstrual disorders, and has several functions, including smooth muscle relaxation and biofluid control (Akase et al. 2004). Toki-Shakuyaku-San consists of six individual herbal medicines, with each providing different pharmacological effects. These herbal medicines are Toki (Angelica acutiloba Radix), Senkyu (Cnidium officinale Rhizoma), Shakuyaku (Paeonia lactiflora Radix), Sojutsu (Atractylodes lancea Rhizoma), Takusha (Alisma orientale Rhizoma), and Bukuryo (Poria cocos Hoelen). Toki and Senkyu are used for the treatment of gynecological diseases (Lu et al. 2005; Kim et al. 2006; Yi et al. 2007; Bohrmann et al. 1967), while Takusha, Bukuryo and Sojutsu are used to improve water metabolism (Wang et al. 2004, 2008; Kitajima et al. 2003; Zhao et al. 2008), and Shakuyaku provides an anti-inflammatory effect (He et al. 2010; Ohta et al. 1993). While some of the active components, such as l-ornithine, gallic acid, paeoniflorin, ferulic acid and benzoic acid in this mixture, and their effects, have been well studied, the relationships between the complete compound profile and the pharmacological effects is not yet well understood (Hatip-Al-Khatib et al. 2004; Chen et al. 2009).

This study aimed to reveal the variance of the charged metabolites found in the six herbal medicines that constitute Toki-Shakuyaku-San, using CE-TOFMS. We conducted multivariate analysis to examine variations in the primary metabolites of the herbal medicine and to understand the relationships between the charged metabolites and the pharmacological effects. Most studies focus on the importance of secondary metabolites to the pharmacological effects of a herbal medicine, however we show that the profile of charged metabolites also varies among herbal medicines.

2 Materials and methods

2.1 Sample preparation

Six herbal medicines, Toki, Bukuryo, Sojutsu, Takusha, Shakuyaku and Senkyu, were obtained from Yatsume Pharmaceutical Co., Ltd. (Tokyo, Japan). Each herb was homogenized without solvent using a multi-bead shocker (Yasui Kikai Co., Osaka, Japan) at 2,500 rpm for 300 s. The crushed material (1.00 g) was dissolved in 3 ml of 50% MeOH containing 2 internal standards (200 μM each of methionine sulfone and 3-aminopyrrolidine). This solution was centrifuged at 3,000 rpm for 10 min at 4°C, and 0.2 ml of the supernatant fluid was transferred to a Millipore 5 kDa cutoff filter tube for centrifugal centrifugation (10,000 rpm × 2 h at 4°C). The filtrate was lyophilized and dissolved in 100 μl of Milli-Q (Millipore, Bedford, MA, USA) water prior to CE-TOFMS analysis.

2.2 Standard chemical compounds

Cysteine-glutathione disulphide was purchased from Apollo Scientific Ltd (Tokyo, Japan). 5-Methyl-2′-deoxycytidine was purchased from MP Biomedicals LLC. (Tokyo, Japan). γ-Glu-2AB and cystathionine were purchased from Toray Research Center (Tokyo, Japan). N-α-Dimethylhistidine was purchased from Bachem AG (Bubendorf, Switzerland). Isobutylamine and azetidine-2-carboxylate were purchased from Chem Service Inc. (West Chester, PA, USA). N8-Acetylspermidine, 5-aminovalerate, O-acetylcarnitine, betaine, glucosaminate, 7-methylguanine, 5-methoxy-3-indoleacetate and pyridoxamine were purchased from Fluka (Buchs, Switzerland). Isonicotinamide was purchased from Tokyo Chemical Industry (Tokyo, Japan). All other compounds were purchased from Sigma–Aldrich (St. Louis, MO, USA) or Wako (Osaka, Japan). All chemical standards were dissolved in Milli-Q water, 0.1 M HCl or 0.1 M NaOH to give 10 mM or 100 mM stock solutions. A standard mixture was prepared by diluting the stock solutions with Milli-Q water just prior to injection into the CE-TOFMS. The chemicals used were of analytical or reagent grade.

2.3 Instrument parameters

The instrumentation and measurement conditions used for CE-TOFMS were as described elsewhere (Sugimoto et al. 2010a). All CE-TOFMS experiments were performed using an Agilent CE capillary electrophoresis system (Agilent Technologies, Waldbronn, Germany), an Agilent G3250AA LS/MSD TOF system (Agilent Technologies, Palo Alto, CA), an Agilent 1100 series binary HPLC pump which delivers sheath liquid, a G1603A Agilent CE–MS adapter kit, and a G1607A Agilent CE–ESI–MS sprayer kit (Agilent Technologies, Waldbronn, Germany). The CE–MS adapter kit used includes a capillary cassette that facilitates thermostatting of the capillary. The CE–ESI–MS sprayer kit simplifies the coupling of the CE system with the MS system, which was equipped with an electrospray ionization source. For system control and data acquisition, G2201AA Agilent ChemStation software for CE and Analyst QS software for TOFMS were used.

2.4 Measurement conditions for CE-TOFMS

Samples were separated in fused silica capillaries (50 μm i.d. × 100 cm total length) filled with 1 M formic acid (pH 1.8) as the background electrolyte. The sample solutions were injected at 50 mbar for 3 s and a voltage of 30 kV was applied. The capillary temperature was maintained at 20°C and the temperature of the sample tray was maintained below 5°C using the external coolant system. The sheath liquid, comprising of methanol/water (50% v/v) and 0.1 μM hexakis-(2,2-difluorothoxy)-phosphazene (Hexakis), was delivered at 10 μl/min. ESI–TOFMS was conducted in positive ion mode. The capillary voltage was set at 4 kV and the nitrogen gas flow rate (heater temperature 300°C) was 10 psig. In TOFMS, the fragmentor, skimmer and OCT RFV voltage were set at 75, 50 and 125 V, respectively. Automatic recalibration of each acquired spectrum was performed using the reference masses of reference standards (13C isotopic ion of protonated methanol dimer [2MeOH + H]+, m/z 66.06371, and [Hexakis + H]+, m/z 622.02896). Mass spectra were acquired at a rate of 1.5 cycles/s over a 50–1,000 m/z range.

2.5 Data processing for CE-TOFMS

Raw data were analyzed with our proprietary software, MasterHands (Sugimoto et al. 2010b). Initially noise-filtering, baseline correction, peak detection and peak area integration from sliced electropherograms (m/z 0.02 width) were carried out. Subsequently, the accurate m/z value for each peak detected was calculated with Gaussian curve-fitting. A migration time normalization function was established using dynamic programming and the simplex optimization method (Soga et al. 2006). Peaks with small differences in m/z value (<20 ppm) and normalized migration time (<1.0 min) were treated as features. Subsequently, neutral compounds, salt ions of Na+ and K+, and redundant features such as fragments, adducts, isotopes, dimers and trimers were eliminated on the basis of established m/z differences (Brown et al. 2009). For the remaining features, metabolite identities were assigned by matching their m/z values and migration times with those of standard compounds. To quantify the detected metabolites, the injected volume for CE and sensitivity of MS were corrected using internal standards, then all annotated metabolites were further corrected with the same chemicals in a standard mixture to overcome different ionization patterns.

2.6 Statistical analysis

The concentration of each metabolite was divided by its average to scale the data, clustered on the basis of Euclidean distance, and visualized as a heat map representation using Mev TM4 software, version 4.6 (Dana-Farber Cancer Institute, Boston, MA) (Saeed et al. 2006). We conducted principal component analysis (PCA) using JMP software (version 8.0.2, SAS, Cary, NC). Metabolite data were mapped on the pathway map using Pathway Projector (Kono et al. 2009), a web-based pathway map application using KEGG Atlas (Kanehisa et al. 2010).

3 Results

3.1 PCA analysis of charged metabolites in herbal medicines and pharmacological effects

After removing peaks determined as non-metabolite, 737 ± 183.1 (average ± SD) peaks remained. Of these, 119 were assigned based on matched m/z and normalized migration times with compounds in our standard library. We conducted PCA using these annotated metabolites. Score plots (Fig. 1a) showed that, at the first principal component (PC), only Toki and Takusha were separated, showing highly positive scores (+9.5 and +7.9), while the others showed negative scores (<−0.2). However, at the second PC, Toki and Takusha were separated, showing strongly negative (−7.7) and positive scores (+8.6) respectively, while the others congregated around zero. At the third PC, only Senkyu was separated, showing a highly positive score (+8.9). Overall, loading plots (Fig. 1b; Supplementary Information Fig. 1) showed that most of the metabolites had similar loadings and no prominent features were observed. Standard amino acids, with the exception of aspartate, glutamate, and tyrosine (indicated with red circles), and shikimate-derived alkaloids, with the exception of mesalamine and dopamine (blue squares), were located furthest from the origin.
Fig. 1

PCA results of 119 annotated metabolites measured in six herbal medicines. a Score plots generated using all annotated metabolites. The cumulative proportions of the first, second and third PC were 43.0%, 65.9% and 83.8%, respectively. b Loading plots for the first and second PCs. Red circles, blue squares, and black crosses, represent standard amino acids, shikimate derived metabolites and other annotated metabolites, respectively. c Score plots for the first and second PCs. The plots colored red have the pharmaceutical effects of (1) tranquilization, (2) painkiller, menstruation problems (dysmenorrhea, oligomenorrhea), (3) removing blood stasis, tonic, and anemia, (4) antidiarrheal effect, (5) anti-convulsion effect (antispasmodic effect), (6) diuretic effect, and (7) oliguria, dizziness, retention of water in the stomach

In the obtained profiles, 39 metabolites showed higher loading score values at the first PC (>0.8). Of these, 20 metabolites were amino acids and amino acid derived metabolites; threonine, valine, serine, tyrosine, glycine, phenylalanine, isoleucine, alanine, lysine, asparagine, arginine, citrulline, homoserine, N-ε-acetyllysine, N-acetylvaline, octopine, ornithine, N-acetylornithine, N-γ-ethylglutamine and N-methylalanine. Along the second PC axis, seven metabolites, 5-aminovalerate, acetylcholine, isoamylamine, O-acetyl carnitine, tryptamine, tyramine and anthranilate, showed high loading score values. Along the third PC axis, 2′-deoxyguanosine, 5′-deoxyadenosine, 5-methyl-2′-deoxycytidine, 5-methylthioadenosine, adenosine, ethanolamine phosphate, N6-methyl-2′-deoxyadenosine and phosphorylcholine showed high loading score values. Of these, four compounds, 5′-deoxyadenosine, 5-methylthioadenosine, adenosine and N6-methyl-2′-deoxyadenosine, were adenosine derived metabolites. In short, these three PCs characterized amino acids, amines and adenosine-derived metabolites, respectively. The pharmacological effects of the herbal medicines used in this study are summarized in Supplementary Information Table 1. To evaluate the relationships between these efficacies and the observed metabolite profiles of the herbal medicines, the efficacies shared by multiple herbal medicines were mapped on the score plot of PCA results (Fig. 1c). One distinct result found was the shared diuretic effect of Bukuryo and Sojutsu, which were closely located on the score plot, indicating a correlation between the observed metabolite profiles and this effect (Fig. 1c). Both Bukuryo and Sojutsu showed minus values in the first PC on score plots (Fig. 1a). The metabolites located at large minus values at the first PC on loading plots, such as β-alanine, isobutylamine, cytosine, cystathionine, nicotine and nicotinamide, may be characteristic of the PC scores of Bukuryo and Sojutsu (Supplementary Information Fig. 1). Senkyu and Toki are used for purposes including tranquilization, alleviation of pain, menstruation problems, removing blood stasis, as a tonic and for anemia, while they are located at a significant distance in PC score plots (Fig. 1c). Although, the PC values of Senkyu and Toki were not similar, the concentrations of several individual metabolites were clearly higher than in the other herbs; 3-methylhistidine, spermidine, adenosine, pipecolate, tryptophan, thiamine and uridine were high in Senkyu and Toki, while agmatine was high in Takusha, Bukuryo and Sojutsu (Supplementary Information Fig. 2).

3.2 Profiling of charged metabolites in herbal medicines

We conducted cluster analysis to assess the similarity of the charged metabolite profiles among the six herbs (Fig. 2). Those metabolites highlighted in yellow were present in one of the herbs at a concentration twofold higher than the average concentration of all the herbs tested. Metabolite concentrations in most clusters (labeled 1–3, and 6–8) were high only in one herb, with few clusters (4 and 5) showing high levels in multiple herb samples. For example, metabolites in clusters 1, 2, 3 and 6 were present at high concentrations in Bukuryo, Senkyu, Toki and Sojutsu, respectively, and the metabolites in clusters 7 and 8 were present at high concentrations in Takusha. The latter clusters contain three xanthine analogs, hypoxanthine, guanine and xanthosine. The amounts of these three metabolites are shown in Supplementary Information Fig. 3a. Metabolites in cluster 4 were present in significantly higher levels in Toki and Sojutsu than the other herbs tested. Metabolites in cluster 5 were present in higher levels in Toki, Takusha and Senkyu than in Shakuyaku, Bukuryo and Sojutsu. A blue-orange heat map shows that the metabolites in cluster 3 were present at high loadings only in the first PC, while metabolites in clusters 7 and 8 were present at high loadings in both the first and second PC. Interestingly, metabolites in cluster 2, including five nucleic acids, ethanolaminephosphate and phosphorylcholine, showed high loading values in the third PC.
Fig. 2

Heatmap showing metabolite profiles and loading scores. Green–red heat map shows the quantified metabolite profiles of the six extracted herbal medicines. Each metabolite concentration shown in the heat map was divided by its average. Blue–orange heat map shows the loading score of the first three PCs. Metabolites highlighted in yellow were present in concentrations twofold higher in one herbal medicine compared with the average. See the text for numerical labels

3.3 Standard amino acid concentration

Figure 3 depicts the quantification of all detected metabolites, standard amino acids, and the sum of the standard amino acids derived from the shikimate pathway. The total metabolite concentration was especially high (>1.5 × 103 μM) in Toki, Takusha and Senkyu, while the lowest concentration (1.9 × 102 μM) was observed in Bukuryo (Fig. 3a). The sum of the standard amino acids was dominant (65–85%) except for Bukuryo (35%). The sum of tryptophan, tyrosine and phenylalanine, the shikimate derived amino acids, was notably high (>5.9 × 102 μM) in Takusha, Toki and Senkyu (Fig. 3a). Arginine was the most dominant amino acid in Toki, Shakuyaku and Takusha (42%, 39% and 41%, respectively), and proline was present in high levels in Toki (30%) and Sojutsu (38%) (Fig. 3b).
Fig. 3

Comparison of the compositions of the herbal medicines tested. a Total concentration of the 119 annotated metabolites (black bar), total concentration of standard amino acids (white-solid bar), and summed concentration of tryptophan, tyrosine and phenylalanine (white-dashed bar) are shown. b Standard amino acid composition of each of the six herbal medicines

4 Discussion

4.1 Overview of charged metabolite profiles and pharmacological effects

CE-TOFMS-based metabolomics has been used to successfully annotate mainly primary metabolites in herbal medicines, while many studies based on LC–MS or GC–MS give profiles of the secondary metabolites contained. The observed metabolite profiles of six herbs studied here showed large variances. PCA revealed that there was a particularly large variance in the herbs Toki, Takusha and Senkyu when compared with the other herbs tested (Fig. 1a). Several metabolites observed in this study have known pharmaceutical efficacies. Tryptophan is present at higher concentrations in Toki and Senkyu (Supplementary Information Fig. 2a, which have several common activities, including the improvement of blood stasis and anemia. It is a precursor of kynurenine, which plays a role in arterial vessel relaxation (Wang et al. 2010). Agmatine was present at high concentrations in Takusha, Bukuryo and Sojutsu, and is known to have a diuretic effect (Smyth and Penner 1995). Bukuryo, Sojutsu and Shakuyaku showed negative scores at the first PC and were closely plotted on the score plot (Fig. 1a) and β-alanine, isobutylamine, cytosine, cystathionine, nicotine and nicotinamide showed higher negative loading score values at the first PC (Supplementary Information Fig. 1). β-Alanine is the rate-limiting precursor of carnosine, which has an anti-inflammatory effect (Zhu et al. 2007), and single oral doses of isobutylamine have a sedative effect (Cheever et al. 1982). These metabolites therefore correlate with the clinical application of these herbs.

Clustering results, shown using a heat map, indicated that there was little similarity among the herbs tested (Fig. 2). Thus, the concentrations of many metabolites vary widely between the herbs. Izzettin Hatip-Al-Khatib et al. have shown previously that only ornithine in Toki extract is responsible for improving memory impairment (Hatip-Al-Khatib et al. 2004). Our profile revealed high concentrations of ornithine in Toki, which was consistent with their study, and in addition, other amines were also found to be present at high concentrations in Takusha. Such findings were obtained using the non-targeted profiling analysis demonstrated. Here, we discuss prominent features observed for amino acids, shikimate derived metabolites, and xanthine analogs.

4.2 Profile of amino acids

Amino acids have been reported to be the main components of Takusha extracts (Namba and Tsuda 1998), and our results are consistent with these findings; 74% of the total annotated metabolite composition was made up of standard amino acids in Takusha (Fig. 3a). However, particularly high amino acid content was not restricted to Takusha, with amino acids making up 64–85% of the composition of the other herbs tested. The only exception to this trend was Bukuryo, whose extract was composed of only 34% standard amino acids (Fig. 3a). Arginine was found to be the most dominant amino acid measured in all of the herbs tested (21–42%), again, with the exception of Bukuryo (9.4%) (Fig. 3b). Proline was found to be abundant in Toki (30%) and Sojutsu (38%), and alanine was abundant in Bukuryo (38%) (Fig. 3b). Indeed, PCA results showed high loading scores for all standard amino acids except for glutamine, asparagine and tyrosine, indicating that these amino acids may be a prominent factor useful for the characterization of these herbs (Fig. 1b).

4.3 Profile of shikimate derived metabolites

The shikimate-derived metabolites observed in our profiles are known to be bioactive, for example dopamine and kynurenine act as signal molecules in the central and peripheral nervous systems. Metabolites contained in the Toki extract are known to bind to a receptor of the central nervous system (Liao et al. 1995) and a liquid extract of Toki containing alkaloids improved retention memory in rats (Hatip-Al-Khatib et al. 2004). Our profiles showed the presence of seven shikimate-derived metabolites and three precursor amino acids (tryptophan, tyrosine and phenylalanine). The total concentration of these three amino acids was higher in Toki, Senkyu and Takusha (Fig. 3a). The concentrations of the shikimate-derived metabolites detected, including kynurenine, harman, tryptamine, tyramine and anthranilate, were also high in Toki, Senkyu and Takusha (Supplementary Information Fig. 3b). These results suggest a correlation between the concentrations of shikimate-derived metabolites and those of the precursor amino acids.

Toki and Senkyu have been reported to contain ferulic acid, which is derived from tyrosine or phenylalanine (Lu et al. 2005; Yi et al. 2007). Although, ferulic acid was not observed in our profiles, the high concentrations of the precursor amino acids of ferulic acid suggest that ferulic acid may be abundant in Toki and Senkyu. In conclusion, we have found some degree of quantitative interaction between bioactive compounds and their precursor amino acids, though there were some shikimate-derived metabolites, such as dopamine, 3-hydroxyanthranilate, which did not follow this trend.

4.4 Profiles of xanthine analogs

Xanthine analogs are known to have a wide range of biological targets. Examples include the adenosine receptor, where xanthines act as antagonists, and calcium release ryanodine-sensitive channels, where they act to sensitize the channels to calcium activation (Daly 2007). Xanthines are known to have several therapeutic benefits, including analgesic and diuretic effects (Daly 2007). Takusha was found to contain the xanthine analogs guanine (EC, hypoxanthine (EC and xanthosine (EC or EC at levels more than ten, four and two times higher than the other herbs tested, respectively. Xanthine analogs may be expected to participate in the analgesic and diuretic effects of Takusha.

4.5 Limitation of this study

In the profiles we obtained, the total concentration of all annotated metabolites was higher in Toki, Takusha, and Senkyu than in the other herbs tested (Fig. 3a). The complex metabolome of these three herbal medicines, including shikimate-derived metabolites and xanthine analogs, may be responsible for the therapeutic functions associated with the herbal medicines, such as analgesia and sedation, since structurally similar compounds are known to have similar functions by receptor binding (Wambach and Casals-Stenzel 1983; Monteith et al. 1996). Further analyses are necessary to compare metabolite extraction methods, with methanol useful to achieve high recovery rates of metabolites (Ren et al. 2008; Sato et al. 2004) and water useful when considering dosage (Liang et al. 2009). More intensive studies on the individual metabolites may give insight into the effect of chemical structure on these therapeutic effects. Further study is needed to elucidate the relationship between the many polar low molecular weight compounds found in these herbal medicines and the herb’s therapeutic functions, particularly their interactions with the nervous system. The quantified profiles obtained should also be validated using different samples of these herbal medicines, for example samples grown at different production sites, harvested at different times, and different batches, in order to establish the natural variability of the metabolites. Profiles of negatively charged metabolites and those obtained using other “omics” platforms should also be integrated to allow for more comprehensive analyses.

5 Concluding remarks

In this study, we have conducted a non-targeted analysis of the polar metabolite profiles of the six herbal medicines which compose Toki-Shakuyaku-San. Overall, there were few common characteristics observed among the herbs tested, and while the total amino acids content was consistently high among the herbs, individual amino acid content showed large variation. Notably, high concentrations of shikimate-derived metabolites were observed in Toki, Takusha and Senkyu, which helps to explain the herb’s biological effects, such as sedation, analgesia and diuretic effects. Multivariable analysis revealed a high correlation between the concentrations of shikimate-derived metabolites and the concentrations of their precursor amino acids, tryptophan, tyrosine and phenylalanine, in the herbs tested. The xanthine analogs guanine, hypoxanthine and xanthosine also showed a high correlation. Analysis of the multiple pharmacological effects of these herbal medicines and comparison with the quantitative compound datasets described here would be valuable in the evaluation of the quality and efficacy of herbal medicines.



This work was supported by research funds from the Yamagata Prefectural Government and the city of Tsuruoka. We thank Dr. Kazuko Otomo for technical assistance, and Wanjun Kong and Guo Jing for fruitful discussions.

Supplementary material

11306_2011_290_MOESM1_ESM.doc (1.2 mb)
Supplementary material 1 (DOC 1,259 kb)


  1. Akase, T., Onodera, S., Matsushita, R., & Tashiro, S. (2004). A comparative study of laboratory parameters and symptoms effected by Toki-Shakuyaku-San and an iron preparation in rats with iron-deficiency anemia. Biological and Pharmaceutical Bulletin, 27, 871–878.PubMedCrossRefGoogle Scholar
  2. Bohrmann, H., Stahl, E., & Mitsuhashi, H. (1967). Studies of the constituents of umbelliferae plants. 8. Chromatographic studies on the constituents of Cnidium officinale Makino. Chemical and Pharmaceutical Bulletin, 15, 1606–1608.CrossRefGoogle Scholar
  3. Brown, M., Dunn, W. B., Dobson, P., et al. (2009). Mass spectrometry tools and metabolite-specific databases for molecular identification in metabolomics. Analyst, 134, 1322–1332.PubMedCrossRefGoogle Scholar
  4. Calixto, J. B. (2000). Efficacy, safety, quality control, marketing and regulatory guidelines for herbal medicines (phytotherapeutic agents). Brazilian Journal of Medical and Biological Research, 33, 179–189.PubMedCrossRefGoogle Scholar
  5. Chan, K. (2003). Some aspects of toxic contaminants in herbal medicines. Chemosphere, 52, 1361–1371.PubMedCrossRefGoogle Scholar
  6. Chan, E. C., Yap, S. L., Lau, A. J., Leow, P. C., Toh, D. F., & Koh, H. L. (2007). Ultra-performance liquid chromatography/time-of-flight mass spectrometry based metabolomics of raw and steamed Panax notoginseng. Rapid Communications in Mass Spectrometry, 21, 519–528.PubMedCrossRefGoogle Scholar
  7. Cheever, K. L., Richards, D. E., & Plotnick, H. B. (1982). The acute oral toxicity of isomeric monobutylamines in the adult male and female rat. Toxicology and Applied Pharmacology, 63, 150–152.PubMedCrossRefGoogle Scholar
  8. Chen, L., Qi, J., Chang, Y. X., Zhu, D., & Yu, B. (2009). Identification and determination of the major constituents in Traditional Chinese Medicinal formula Danggui-Shaoyao-San by HPLC-DAD-ESI-MS/MS. Journal of Pharmaceutical and Biomedical Analysis, 50, 127–137.PubMedCrossRefGoogle Scholar
  9. Daly, J. W. (2007). Caffeine analogs: Biomedical impact. Cellular and Molecular Life Sciences, 64, 2153–2169.PubMedCrossRefGoogle Scholar
  10. Ganzera, M. (2008). Quality control of herbal medicines by capillary electrophoresis: Potential, requirements and applications. Electrophoresis, 29, 3489–3503.PubMedCrossRefGoogle Scholar
  11. Gong, F., Liang, Y. Z., Cui, H., Chau, F. T., & Chan, B. T. (2001). Determination of volatile components in peptic powder by gas chromatography-mass spectrometry and chemometric resolution. Journal of Chromatography A, 909, 237–247.PubMedCrossRefGoogle Scholar
  12. Hatip-Al-Khatib, I., Egashira, N., Mishima, K., et al. (2004). Determination of the effectiveness of components of the herbal medicine Toki-Shakuyaku-San and fractions of Angelica acutiloba in improving the scopolamine-induced impairment of rat’s spatial cognition in eight-armed radial maze test. Journal of Pharmacological Sciences, 96, 33–41.PubMedCrossRefGoogle Scholar
  13. He, C. N., Peng, Y., Xu, L. J., et al. (2010). Three new oligostilbenes from the seeds of Paeonia suffruticosa. Chemical and Pharmaceutical Bulletin, 58, 843–847.CrossRefGoogle Scholar
  14. Hurtado-Fernandez, E., Gomez-Romero, M., Carrasco-Pancorbo, A., & Fernandez-Gutierrez, A. (2010). Application and potential of capillary electroseparation methods to determine antioxidant phenolic compounds from plant food material. Journal of Pharmaceutical and Biomedical Analysis, 53, 1130–1160.PubMedCrossRefGoogle Scholar
  15. Kanehisa, M., Goto, S., Furumichi, M., Tanabe, M., & Hirakawa, M. (2010). KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Research, 38, D355–D360.PubMedCrossRefGoogle Scholar
  16. Kim, M. R., Abd El-Aty, A. M., Choi, J. H., Lee, K. B., & Shim, J. H. (2006). Identification of volatile components in Angelica species using supercritical-CO2 fluid extraction and solid phase microextraction coupled to gas chromatography-mass spectrometry. Biomedical Chromatography, 20, 1267–1273.PubMedCrossRefGoogle Scholar
  17. Kitajima, J., Kamoshita, A., Ishikawa, T., et al. (2003). Glycosides of Atractylodes lancea. Chemical and Pharmaceutical Bulletin, 51, 673–678.CrossRefGoogle Scholar
  18. Kono, N., Arakawa, K., Ogawa, R., et al. (2009). Pathway projector: Web-based zoomable pathway browser using KEGG atlas and Google Maps API. PLoS One, 4, e7710.PubMedCrossRefGoogle Scholar
  19. Lao, Y. M., Jiang, J. G., & Yan, L. (2009). Application of metabonomic analytical techniques in the modernization and toxicology research of traditional Chinese medicine. British Journal of Pharmacology, 157, 1128–1141.PubMedCrossRefGoogle Scholar
  20. Li, X. N., Cui, H., Song, Y. Q., Liang, Y. Z., & Chau, F. T. (2003). Analysis of volatile fractions of Schisandra chinensis (Turcz.) Baill. using GC–MS and chemometric resolution. Phytochemical Analysis, 14, 23–33.PubMedCrossRefGoogle Scholar
  21. Liang, X. M., Jin, Y., Wang, Y. P., Jin, G. W., Fu, Q., & Xiao, Y. S. (2009). Qualitative and quantitative analysis in quality control of traditional Chinese medicines. Journal of Chromatography A, 1216, 2033–2044.PubMedCrossRefGoogle Scholar
  22. Liao, J. F., Jan, Y. M., Huang, S. Y., Wang, H. H., Yu, L. L., & Chen, C. F. (1995). Evaluation with receptor binding assay on the water extracts of ten CNS-active Chinese herbal drugs. Proceedings of the National Science Council, Republic of China. Part B, Life Sciences, 19, 151–158.PubMedGoogle Scholar
  23. Liu, S., Yi, L. Z., & Liang, Y. Z. (2008). Traditional Chinese medicine and separation science. Journal of Separation Science, 31, 2113–2137.PubMedCrossRefGoogle Scholar
  24. Lu, G. H., Chan, K., Liang, Y. Z., et al. (2005). Development of high-performance liquid chromatographic fingerprints for distinguishing Chinese Angelica from related umbelliferae herbs. Journal of Chromatography A, 1073, 383–392.PubMedCrossRefGoogle Scholar
  25. Ma, C. M., Winsor, L., & Daneshtalab, M. (2007). Quantification of spiroether isomers and herniarin of different parts of Matricaria matricarioides and flowers of Chamaemelum nobile. Phytochemical Analysis, 18, 42–49.PubMedCrossRefGoogle Scholar
  26. Miller, F. G., Emanuel, E. J., Rosenstein, D. L., & Straus, S. E. (2004). Ethical issues concerning research in complementary and alternative medicine. JAMA, 291, 599–604.PubMedCrossRefGoogle Scholar
  27. Monteith, D. K., Emmerling, M. R., Garvin, J., & Theiss, J. C. (1996). Cytotoxicity study of tacrine, structurally and pharmacologically related compounds using rat hepatocytes. Drug and Chemical Toxicology, 19, 71–84.PubMedCrossRefGoogle Scholar
  28. Monton, M. R., & Soga, T. (2007). Metabolome analysis by capillary electrophoresis–mass spectrometry. Journal of Chromatography A, 1168, 237–246. discussion 236.PubMedCrossRefGoogle Scholar
  29. Murray, R. H., & Rubel, A. J. (1992). Physicians and healers—Unwitting partners in health care. New England Journal of Medicine, 326, 61–64.PubMedCrossRefGoogle Scholar
  30. Namba, T., & Tsuda, Y. (1998). Shoyakugakugairon (3rd ed.). Tokyo: Nankodo.Google Scholar
  31. Ohta, H., Ni, J. W., Matsumoto, K., Watanabe, H., & Shimizu, M. (1993). Peony and its major constituent, paeoniflorin, improve radial maze performance impaired by scopolamine in rats. Pharmacology, Biochemistry and Behavior, 45, 719–723.CrossRefGoogle Scholar
  32. Ren, M. T., Chen, J., Song, Y., Sheng, L. S., Li, P., & Qi, L. W. (2008). Identification and quantification of 32 bioactive compounds in Lonicera species by high performance liquid chromatography coupled with time-of-flight mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis, 48, 1351–1360.PubMedCrossRefGoogle Scholar
  33. Saeed, A. I., Bhagabati, N. K., Braisted, J. C., et al. (2006). TM4 microarray software suite. Methods in Enzymology, 411, 134–193.PubMedCrossRefGoogle Scholar
  34. Sato, S., Soga, T., Nishioka, T., & Tomita, M. (2004). Simultaneous determination of the main metabolites in rice leaves using capillary electrophoresis mass spectrometry and capillary electrophoresis diode array detection. The Plant Journal, 40, 151–163.PubMedCrossRefGoogle Scholar
  35. Smyth, D. D., & Penner, S. B. (1995). Renal I1-imidazoline receptor-selective compounds mediate natriuresis in the rat. Journal of Cardiovascular Pharmacology, 26(Suppl. 2), S63–S67.PubMedGoogle Scholar
  36. Soga, T., Baran, R., Suematsu, M., et al. (2006). Differential metabolomics reveals ophthalmic acid as an oxidative stress biomarker indicating hepatic glutathione consumption. The Journal of Biological Chemistry, 281, 16768–16776.PubMedCrossRefGoogle Scholar
  37. Soga, T., Ohashi, Y., Ueno, Y., Naraoka, H., Tomita, M., & Nishioka, T. (2003). Quantitative metabolome analysis using capillary electrophoresis mass spectrometry. Journal of Proteome Research, 2, 488–494.PubMedCrossRefGoogle Scholar
  38. Sugimoto, M., Koseki, T., Hirayama, A., et al. (2010a). Correlation between sensory evaluation scores of Japanese sake and metabolome profiles. Journal of Agriculture and Food Chemistry, 58, 374–383.CrossRefGoogle Scholar
  39. Sugimoto, M., Wong, D. T., Hirayama, A., Soga, T., & Tomita, M. (2010b). Capillary electrophoresis mass spectrometry-based saliva metabolomics identified oral, breast and pancreatic cancer-specific profiles. Metabolomics, 6, 78–95.PubMedCrossRefGoogle Scholar
  40. Wambach, G., & Casals-Stenzel, J. (1983). Structure-activity relationship of new steroidal aldosterone antagonists. Comparison of the affinity for mineralocorticoid receptors in vitro and the antialdosterone activity in vivo. Biochemical Pharmacology, 32, 1479–1485.PubMedCrossRefGoogle Scholar
  41. Wang, H. X., Liu, C. M., Liu, Q., & Gao, K. (2008). Three types of sesquiterpenes from rhizomes of Atractylodes lancea. Phytochemistry, 69, 2088–2094.PubMedCrossRefGoogle Scholar
  42. Wang, Y., Liu, H., Mckenzie, G., et al. (2010). Kynurenine is an endothelium-derived relaxing factor produced during inflammation. Nature Medicine, 16, 279–285.PubMedCrossRefGoogle Scholar
  43. Wang, Z. G., & Ren, J. (2002). Current status and future direction of Chinese herbal medicine. Trends in Pharmacological Sciences, 23, 347–348.PubMedCrossRefGoogle Scholar
  44. Wang, J., Van Der Heijden, R., Spruit, S., et al. (2009). Quality and safety of Chinese herbal medicines guided by a systems biology perspective. Journal of Ethnopharmacology, 126, 31–41.PubMedCrossRefGoogle Scholar
  45. Wang, Y., Zhang, M., Ruan, D., et al. (2004). Chemical components and molecular mass of six polysaccharides isolated from the sclerotium of Poria cocos. Carbohydrate Research, 339, 327–334.PubMedCrossRefGoogle Scholar
  46. Weng, Q., & Jin, W. (2002). Carbon fiber bundle–Au–Hg dual-electrode detection for capillary electrophoresis. Journal of Chromatography A, 971, 217–223.PubMedCrossRefGoogle Scholar
  47. Xie, G., Plumb, R., Su, M., et al. (2008). Ultra-performance LC/TOF MS analysis of medicinal Panax herbs for metabolomic research. Journal of Separation Science, 31, 1015–1026.PubMedCrossRefGoogle Scholar
  48. Yi, T., Leung, K. S., Lu, G. H., & Zhang, H. (2007). Comparative analysis of Ligusticum chuanxiong and related umbelliferous medicinal plants by high performance liquid chromatography–electrospray ionization mass spectrometry. Planta Medica, 73, 392–398.PubMedCrossRefGoogle Scholar
  49. Zhang, A., Sun, H., Wang, Z., Sun, W., Wang, P., & Wang, X. (2010). Metabolomics: Towards understanding traditional Chinese medicine. Planta Medica, 76, 2026–2035.PubMedCrossRefGoogle Scholar
  50. Zhao, M., Xu, L. J., & Che, C. T. (2008). Alisolide, alisols O and P from the rhizome of Alisma orientale. Phytochemistry, 69, 527–532.PubMedCrossRefGoogle Scholar
  51. Zhu, Y. Y., Zhu-Ge, Z. B., Wu, D. C., et al. (2007). Carnosine inhibits pentylenetetrazol-induced seizures by histaminergic mechanisms in histidine decarboxylase knock-out mice. Neuroscience Letters, 416, 211–216.PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2011

Authors and Affiliations

  • Keiko Iino
    • 1
    • 2
  • Masahiro Sugimoto
    • 1
    • 3
  • Tomoyoshi Soga
    • 1
    • 2
    • 3
  • Masaru Tomita
    • 1
    • 2
    • 3
  1. 1.Institute for Advanced BiosciencesKeio UniversityTsuruokaJapan
  2. 2.Department of Environment and Information StudiesKeio UniversityFujisawaJapan
  3. 3.Systems Biology Program, Graduate School of Media and GovernanceKeio UniversityFujisawaJapan

Personalised recommendations