A practical and novel “standard addition” strategy to screen pharmacodynamic components in traditional Chinese medicine using Heishunpian as an example

Abstract

The study of pharmacodynamic components in traditional Chinese medicine (TCM) is very important for future drug development and quality control of TCM. The present study established a new method for screening pharmacodynamic components (PCs) in TCM based on the strategy of standard addition (SA). The novel strategy was then applied to screen anti-inflammatory PCs in Heishunpian (HSP), the processed product of Aconitum carmichaeli Debx. We initially screened target components (TCs, possible PCs) by analyzing the correlation between UPLC-Q-TOF-MS fingerprints of chemical components and the anti-inflammatory efficacies of different HSP extracts. On the basis of spectrum-effect relationship (SER) analysis and TC determination, we added TC quantitatively into HSP extracts, evaluated the pharmacodynamic contribution ratios of TCs related to the original TCM using a mouse ear edema model, and compared them with the contribution ratio of the TC standard. The anti-inflammatory PCs of HSP were then defined. Results showed that hypaconitine, deoxyaconitine and chasmanine were anti-inflammatory PCs in HSP with positive relations with HSP efficacy. Thus, SA was used to systematically evaluate the effect of chemical ingredients in TCM. The proposed method presents simple operation, strong feasibility and reliability, and provides a new approach for screening PCs in TCM in a manner that highlights the complexity and multi-component effects of TCM.

---

1 Introduction

Traditional Chinese medicine (TCM) is an important area of study, especially in Asian countries, and has received increased attention and wider acceptance in recent decades because of its efficacy and minimal side effects.^1–3 The chemical components in TCM that are responsible for its clinical effects are called pharmacodynamic compounds (PCs). In contrast to single-component chemical drugs, PCs of TCM comprise a complex system with multiple components. Therefore, identifying the PCs of TCM can ensure the safety and efficacy of herbs and their derivatives. However, considering the large number of components of TCM and their various mechanisms of action, studies on PCs in TCM have been relatively few, which has greatly hindered TCM modernization, industrialization, and globalization.^4–6

Current strategies for studying PCs in TCM include separation of the individual chemical components and evaluation of each component's activity.^7,8 This strategy, however, disregards the entirety and systematic action of TCM and fails to reflect the characteristics of multi-components and multi-targets in TCM. An alternate strategy involves pharmacological activity-oriented tracking,⁷ in which the efficacy of each component obtained from each stage of separation is evaluated. Unfortunately, this method is labor-intensive and ignores the wholeness and systematic mode of action of TCM. Another strategy involves computer-aided virtual screening,⁹ which predicts PCs using computer-based algorithms. This approach, however, relies on programmed models and specialized databases, and the reliability and accuracy of prediction results require further validation. Finally, spectrum-effect analysis^10,11 combines pharmacodynamic activity and chemical fingerprints with a mathematical model to screen PCs. However, this method only reflects the relationship between ingredients and pharmacodynamics and also requires further validation.

Taking reference to the mode of “gene knock-out” or “gene deletion”, Li and Xiao recently established a target constituent knock-out/knock-in strategy wherein the target component (TC) in the herbs was eliminated by a specific technology.^12,13 PCs were screened by comparing pharmacological activities among TCs, negative samples, and whole herbs. This approach offers higher accuracy and takes the complex environment of TCMs into account.^14,15 However, its requires a rapid separation technology with good extraction efficiency, high yield, and no residues to be eliminated.¹⁶ Hence, this strategy requires a higher technology condition. Therefore, there is an urgent need for a strategy with general applicability based on the characteristics of TCM to be established.

The recovery test, which involves adding TC standards directly to samples, is used to examine the accuracy of component determination.¹⁷ Using this principle, the present article established a “standard addition (SA)” strategy to screen PCs that takes into account the entirety of components in TCM. Research on anti-inflammatory PCs in Heishunpian (HSP, the processed product of Aconitum carmichaeli Debx) was taken as an example. First, the correlation between the fingerprints and efficacy of different HSP extracts was analyzed. In this step, TCs were preliminarily identified. Next, the contents of TCs in different extracts were measured, and the content range was calculated. Finally, the TC standard was added to TCM extracts, and the contribution ratios of TCs were assessed. The ingredient that could significantly change the effect of TCM and contributed greatly to the effect of the whole TCM was considered to be a PC. This strategy allows accurate PC screening and embodies the systematic nature of TCM. Thus, this paper introduces a novel method for investigating the PCs of TCM.

2 Experiments

2.1 Chemicals and materials

10 batches of Heishunpian (HSP, the processed product of Aconitum carmichaeli Debx) were purchased from specialized companies and medicine markets, and authenticated by Dr Lu Zhang at Tianjin University of Traditional Chinese Medicine (Tianjin, China). These samples were sealed and deposited at the specimen laboratory in the School of Traditional Chinese Materia Medica, Tianjin University of Traditional Chinese Medicine, China, at room temperature. Keep dry indoor to avoid samples becoming damp. Because HSP is a sort of toxic herb, we store it separately from others so as not to interfere with them. After identifying and analyzing the relative content of chemical components in each batches of HSP, we selected one batch to use in subsequent experiments at last. The batch selected can stand for the majority of market HSP and was processed by specialized company of Chinese herbal medicine. It was bought from Hua Miao Engineering Technology Development Center of Traditional Chinese Medicine (Beijing, China), (no. 305231). The basis of selecting is shown in ESI.† Four standards, namely, chasmanine, mesaconitine, hypaconitine, and deoxyaconine, were purchased from Chengdu Herbpurify Co., Ltd. (Chengdu, China). Purity of all the standards was more than 98% as determined by HPLC. HPLC-grade acetonitrile, formic acid, and alcohol for UPLC-Q-TOF-MS analysis were purchased from Sigma-Aldrich (St. Louis, MO, USA). Other analytical grade chemicals, such as methanol and absolute ethyl alcohol, were obtained from Concord Science and Technology Co., Ltd. (Tianjin, China). Ultrapure water was produced using a Milli-Q system (Millipore, Bedford, MA, USA).

2.2 Animals

Male Kunming mice (SPF, 20 ± 2 g) were provided by the Military Academy of Medical Sciences Laboratory Animal Center (Beijing, China) and raised in the Institute of Radiation Medicine Chinese Academy of Medical Sciences (Tianjin, China). All mice were fed with standard laboratory food and water ad libitum for at least 5 days prior to the experiments in an environmentally controlled room (12 h/12 h dark/light cycle; 25 ± 1 °C, 50 ± 5% humidity). The study was approved by the Animal Ethics Committee of Tianjin University of Traditional Chinese Medicine under permit number TCM-2013-12-A06. All procedures were conducted in accordance with the Chinese national legislation and local guidelines.

2.3 Instruments and UPLC-Q-TOF-MS conditions

A Waters Acquity UPLC Class I series (Premier from Waters) setup equipped with a quat pump, an autosampler, a DAD detector, and a column compartment was used for analysis. Samples were separated on a Waters ACQUITY UPLC BEH C₁₈ column (2.1 mm × 100 mm, 1.7 µm; Waters, USA). The mobile phases contained water with 0.1% formic acid (eluent A) and acetonitrile with 0.1% formic acid (eluent B) at a flow rate of 0.3 mL min⁻¹. The following linear gradient was applied: 0–2 min, 1% B; 2–5 min, 1–7% B; 5–15 min, 7–10% B; 15–20 min, 10–20% B; 20–31 min, 20–30% B; 31–36 min, 30–50% B; and 36–38 min, 50–99% B. The column temperature was maintained at 40 °C, and the sample injection volume was 2 µL.

A Waters Xevo G2 Q-Tof setup was connected to the Waters Acquity UPLC Class I system via an electrospray ionization interface. Ultra-high purity helium (He) was used as the collision gas, and high-purity nitrogen (N₂) was used as the nebulizing gas. Data were collected in positive ion mode via full-scan mass analysis from m/z 100 to m/z 1000. Other operating parameters were as follows: capillary voltage, 3.0 kV; drying gas temperature, 350 °C; flow rate, 800 L h⁻¹; and nebulizing gas pressure, 35 psi. The Leu-enkephalin ion at m/z 556.2771 was used to calibrate mass accuracy.

2.4 Spectrum-effect relationship analysis for initial screening of target components

2.4.1 Preparation of sample solutions. Six dried powder samples of HSP (100 g) were refluxed twice (1 and 0.5 h) with water, 25% ethanol, 50% ethanol, 75% ethanol, 95% ethanol and methanol (1000 and 800 mL), respectively. Each extract was dried into powders. Finally, six sample solutions of different extracts were prepared by dissolving them in appropriate solvent. After centrifugation (5000 rpm, 10 min), the supernatants were transferred to an autosampler vial for UPLC-Q-TOF-MS analysis.

2.4.2 Pharmacodynamic inspection by xylene-induced ear edema in mice. A total of 70 male Kunming mice were randomly divided into 7 groups, of which one was a control group and the remaining six were test groups for anti-inflammatory tests. Water, 25% ethanol, 50% ethanol, 75% ethanol, 95% ethanol and methanol extracts (8 g kg⁻¹) were orally administered to different groups of mice; the control group was given equivalent amount of water. After 30 min, 20 µL of xylene was applied to the left ears of all mice; their right ears served as controls. The mice were sacrificed by cervical dislocation 1 h after xylene application, and 7.00 mm diameter disks were punched out from the ears and weighed. The degree of edema was calculated as the weight difference between the left and right ear disks of the same mouse.^18,19 The inhibitory ratio was calculated as follows:

I = (D_c − D_t)/D_c × 100%

where D_c and D_t represent the degrees of swelling of the control and test groups, respectively.

2.4.3 UPLC-Q-TOF-MS fingerprint analysis of different extracts. The fingerprint of the sample solution prepared in Section 2.4.1 was established using the method described in Section 2.3. The precision, reproducibility, and stability of the analysis method were subsequently verified. Data analysis of the characteristic peaks in the fingerprints was performed using MassLynx v4.1 (Waters Corporation, MA, USA).

2.4.4 Partial least squares regression analysis. The spectrum-effect relationship (SER) between the fingerprint and anti-inflammatory action of HSP was constructed by partial least squares (PLS) regression. PLS regression mode was built by taking the anti-inflammatory efficacy of different HSP extracts as the variable Y and the peak area of ingredients in different HSP extracts as the variable X. Some variable X with larger variable importance plot (VIP) value made larger contribution to variable Y; they were considered target components (TCs, possible PCs). SPSS 17.0 (SPSS Inc, Chicago, USA) was used to conduct statistical analysis, and differences were evaluated by t-test.

2.5 Quantitative analysis of target components in different extracts

2.5.1 Preparation of standard and sample solutions. Standard stock solution was prepared by dissolving each reference standard (chasmanine, mesaconitine, hypaconitine, and deoxyaconine) in 70% ethanol at a concentration of 1 mg mL⁻¹. Working standard solution was prepared by diluting the standard stock solution. All standard stock and working solutions were stored at −20 °C. Sample solutions were prepared as described in Section 2.4.1.

2.5.2 Validation of quantitative analysis. The method was validated for linearity, limits of detection and quantification (LODs and LOQs), precision, reproducibility, stability, and accuracy by referring to several previous reports on determination analysis.^17,20

2.5.3 Quantitative analysis of target components. Sample solutions prepared in Section 2.4.1 were analyzed as described above, and the content of each TC in different extracts were calculated according to their respective calibration curves.

2.6 Screening of pharmacodynamic components by standard addition

2.6.1 Test solution preparation.
Background drug. According to the results of quantitative analysis of TCs in different extracts, we selected methanol extract as the background drug.
TC standard. Each standard (chasmanine, mesaconitine, hypaconitine and deoxyaconitine) was dissolved to 1 mg mL⁻¹ and further diluted to the appropriate concentration with water at the required time.
SA drug. We ensured that the content of TCs was maintained within the usage range of the SER model after addition. Referring to the dose escalation scheme of previous pharmacological experiments,²¹ we added twice the amount of the TC standard to background drug, respectively.

2.6.2 Re-evaluation of the efficacy of target components. One hundred male Kunming mice were randomly divided into a blank control group, a background drug group, four SA groups (HSP + chasmanine, HSP + mesaconitine, HSP + hypaconitine and HSP + deoxyaconitine), and four standard control groups (chasmanine, mesaconitine, hypaconitine and deoxyaconitine). The background drug group was given the methanol extract of HSP at a dose of 8 g kg⁻¹; SA groups were given the SA drug as described in Section 2.6.1; the standard control groups were given diluted standard solutions at a dose equivalent to that of the methanol extract added; and the blank control group was given equivalent amount of water. The remaining experiments were conducted as described in Section 2.4.2. This experiment was repeated 3 times. And take the mean of 3 results as the finally result.

3 Result and discussion

3.1 Spectrum-effect relationship analysis to screen target components

Different inherent components of TCM contribute different clinical effects. However, the relationship between the efficacy and chemical compositions of TCM is still vague. SER defines the link between chemical fingerprints and specific efficacy data by using multivariate statistical tools.²² This method combines both chemical analysis and pharmacological studies, and can be used to screen the PCs of TCM.²³ SER has recently been widely used for screening PCs of TCM.^10,11,23,24 In this study, we chose to investigate the inflammatory TCs of HSP by the SER method.

3.1.1 Pharmacodynamic inspection by xylene-induced ear edema in mice. The classic xylene-induced ear edema model was selected to evaluate the anti-inflammatory efficacy of six different extracts of HSP, and the inhibition ratio was used as the index of efficacy. As shown in Fig. 1, different extracts of HSP exhibited different anti-inflammatory effects compared with the control.

Fig. 1 Inhibitory ratio of different extracts of Heishunpian (HSP). *p < 0.05, compared with control (n = 10), (SPSS 17.0, independent t-test).

3.1.2 UPLC-Q-TOF-MS fingerprint analysis of different extracts. The efficacy of TCM relies in its complex components. The chemical fingerprint of TCM can fully reflect the different types of chemical components, which can be used to evaluate integral quality.^25,26 Fig. 2 shows the base peak intensity (BPI) chromatogram of different HSP extracts. We found significant difference among them in the amounts and contents of each component. The identity of each component is shown in Table 1. The relative standard deviation (RSD) value of the relative peak area (RPA) was less than 5%, which demonstrated the stability, reproducibility, and precision of our fingerprint analysis.

Fig. 2 The base peak intensity (BPI) chromatogram of different HSP extracts (A: water extract; B: 25% ethanol extract; C: 50% ethanol extract; D: 75% ethanol extract; E: 95% ethanol extract; F: methanol extract).

Table 1 Identification of the components

No.	RT (min)	[M + H]^+	Identification
1	5.33	394.2593	Columbianine
2	6.95	486.2701	Mesaeonine
3	6.52	424.2687	Senbusine B
4	7.47	378.2643	Carmichaeline
5	7.68	408.2741	Isotalatizidine
6	8.11	358.238	Songorine
7	8.71	330.2066	Hetisine
8	10.73	454.2803	Fuziline
9	11.91	438.2854	Neoline
10	12.08	342.1692	Fuzitine
11	13.93	344.2586	Bullatine A
12	15.29	422.2894	Talatizamine
13	18.06	452.3002	Chasmanine
14	19.66	464.3007	14-Acetyltalatizamine
15	19.86	606.2925	10-OH-Benzoylmesaeonine
16	22.57	590.2958	Benzoylmesaconine
17	23.86	604.3152	Benzoyaeonine
18	24.9	574.3009	Benzoylhypaconine
19	26.43	588.3168	Dehydration-10-OH-benzoylmesaeonine
20	28.87	632.3055	Mesaconitine
21	29.17	662.3168	10-OH-Aeonitine
22	31.15	616.3127	Hypaconitine
23	31.35	572.3218	Dehydrated benzoylmesaconine
24	32.99	602.3331	Dehydrated-10-OH-aeonitine
25	33.37	630.3278	Deoxyaconitine

---

3.1.3 Partial least squares regression analysis. In SER, multivariate statistics is utilized to determine the efficacies of various chemical components. It is very important to choose an appropriate statistical method. PLS regression is one method that predicts multiple output variables from multiple input variables. PLS regression can derive its usefulness by analyzing data with mass, noise and collinearity.^27,28 It summarizes input variables (X) to extract the most predictive information for the response variable (Y) and maximizes the covariance between X and Y.²⁹ PLS is a simple but powerful process for analyzing data of complex problems. Notably, it has previously been used to analyze the active ingredients of TCM.^11,30 In this study, the anti-inflammatory activity of different HSP extracts and peak area were used as variable Y and as variable X, respectively, and the PLS regression model was established by SMICAP-11.5 (Umetrics AB, Sweden). In PLS regression mode, VIP value reflects the contribution of each input variable to overall PLS model. It summarizes all components and weights according to the Y variation accounted for by each component.³¹ The VIP cutoff value of 1.0 is a good starting point, and a value higher than 1.0 reveals high correlation.³² In general, variables with VIP > 1 are considered makers.^32,33 In the present study, chemical components with VIP > 1 were considered as TCs.

By conducting PLS regression analysis, we established a relationship between HSP fingerprints and its inhibitory effect against xylene-induced ear edema in mice, and identified the TCs. Depending on VIP value, chasmanine, mesaconitine, hypaconitine, deoxyaconitine, columbianine, and mesaconine were screened as TCs of HSP.

Many components, including several high content components, have not been screened by SER to date. Our results indicated that the change of efficacy might not have been caused by change of these components. Many studies have found that diester-diterpenoid alkaloids (DDAs) in HSP display anti-inflammatory activity, although the contents of such compounds are usually lower. However, very few reports exist on monoester-diterpenoid alkaloids (MDAs), which are more abundant.^34,35 Using paw swelling in mice and other inflammatory models, Nesterova et al. also showed that mesaconitine, hypaconitine, and other ingredients displayed evident anti-inflammatory actions.^36–38

Therefore, the results of SER can be believed to some extent. The anti-inflammatory effects of these higher content components will be investigated in future experiments. However, multivariate statistical methods are artificially defined algorithms with fixed principles and formulas. The chemical components and pharmacological effects of TCM are very complex. Multivariate statistical techniques may not be comprehensive enough to describe the relationship between the chemical components and pharmacological effects of TCM. That is to say, due to the limitations of statistical techniques and complication in chemical compounds and action mechanism of TCM, SER does have some disadvantages. For this reason, we only selected SER as a way to screen TCs initially, and proposed the SA technique to further confirm SER results.

3.2 Quantitative analysis of target components in different extracts

As we know, the action of an ingredient is related to its content. Therefore, it is necessary to clear the content range of TCs. Moreover, determining the content of TCs in HSP is the foundation of an SA strategy. Here, we determined four TCs with large VIP values and defined the usage range of the model established by SER analysis.

3.2.1 Validation of quantitative analysis method. To ensure the accuracy of the determination method, the linearity, LOD and LOQ, precision, reproducibility, and recovery tests were conducted. The results indicated good correlation between peak area and component concentration (r² > 0.998) in the linear range of 0.01–0.16 µg mL⁻¹ for chasmanine, mesaconitine and deoxyaconitine, and 0.1–1.6 µg mL⁻¹ for hypaconitine (Table S1†). The RSDs of precision, repeatability, and stability were less than 5%, which meets the demands of quantitative analysis. The mean recoveries of chasmanine, mesaconitine, deoxyaconitine and hypaconitine were 98.7%, 98.6%, 98.1%, and 99.0%, respectively (RSD < 5%).

3.2.2 Quantitative analysis of target components. The results above indicated that the method was reliable and accurate for determining PCs. The contents of chasmanine, mesaconitine, hypaconitine, and deoxyaconitine in different extracts were in the range of 1.805–14.31, 0–15.37, 30.49–206.0, and 1.894–15.40 µg g⁻¹, respectively (Fig. 3).

Fig. 3 The contents of target components (TCs) in different extracts; TCs content was within the useful range of the spectrum-effect relationship model after adding standards into methanol extracts of Heishunpian.

3.3 Screening of pharmacodynamic components by standard addition

PCs of TCM are important in clinical treatment and form the basis of many TCM studies. Whether a TC is an anti-inflammatory PC of HSP must be verified. In order to confirm the TCs screened by SER, we added each TC standard to the original extract of HSP, and then evaluated the activity of the original extract and TC with extract simultaneously. After clarifying the influence of each TC to overall HSP, we confirmed the PCs of HSP.

3.3.1 Determination of contribution ratio. Based on the results of Section 3.2, the methanol extract was selected as the background drug, and the content of chasmanine, mesaconitine, hypaconitine, and deoxyaconitine in the extract was determined as 2.1, 5.04, 60.38, and 4.47 µg g⁻¹, respectively. Following the rule of dose escalation of pharmacological experiments, two fold amounts of the TC standard (4.2, 10.08, 120.76, and 8.94 µg g⁻¹) were added to the background drug. The TC content was within the range used for the SER model, as show in Fig. 3. In the present study, the contribution ratio of TCs before or after addition was calculated using the following formulas:

C_a = (I_a − I)/I/n

C_s = (I_s − I)/I/n

where C_a and I_a are the contribution and inhibitory ratios, respectively, of SA groups while C_s and I_s are the contribution and inhibitory ratios, respectively, of standard control groups; I is the inhibitory ratio of the background drug group, and n is the content multiple of addition. The contribution ratio of TC was calculated using its content and efficacy. It represents the effect of a single component in the overall TCM, and can show their correction. Therefore, it can be used for screening the active ingredients in TCM.¹²

3.3.2 Confirmation of pharmacodynamic components by standard addition strategy. PCs were screened based on the following criteria: (1) if a single TC was effective and TCM efficacy was significantly enhanced after addition, the TC was one of the PCs. (2) If a single TC was effective but TCM efficacy decreased in comparison with the original medicine after addition, the TC was not one of the PCs. (3) If the single TC was not effective or had slight efficacy and TCM efficacy increased after addition, the TC and TCM efficacy are linked, and the TC can be regarded as a PC. (4) If a single TC was not effective or had slight efficacy and TCM efficacy did not change after addition, the TC was not a PC.

The efficacy of TCs is shown in Table 2. The consequences were the mean of 3 repeated experiments, which errors were less than 15%. Mesaconitine, hypaconitine, and deoxyaconitine in the standard solution group could significantly inhibit xylene-induced ear edema in mice (P < 0.05); by contrast, chasmanine showed no such inhibition. Each SA group showed significant inhibition of ear edema (P < 0.05). Mesaconitine, hypaconitine, and deoxyaconitine alone exhibited positive contributions to the anti-inflammatory action, whereas chasmanine alone showed no such contribution. After addition, hypaconitine, deoxyaconitine, and chasmanine showed positive contributions to the anti-inflammatory action, whereas mesaconitine showed negative contribution.

Table 2 Effect of target component standards with HSP extracts on xylene-induced ear swelling in mice^a,b,c,d

Group no.	Dose	Weight difference (g)	Inhibition ratio (%)	Contribution ratio (%)
a Values are expressed as mean ± S.E.M. There are 10 animals in each group, and this experiment was repeated 3 times. *P < 0.05, compared with control, **P < 0.01, compared with control; ^#P < 0.05, compared with HSP (SPSS 17.0, independent t-test).b HSP: Heishunpian, the processed product of Aconitum carmichaeli Debx.c Inhibitory ratio (I), I = (Dc − Dt)/Dc × 100%, where Dc and Dt represent the degrees of swelling of the control and test groups, respectively.d Contribution ratio (C) of TCs: Ca = (Ia − I)/I/n, Cs = (Is − I)/I/n, where Ca and Ia are the contribution and inhibitory ratios, respectively, of SA groups (Group no. 7, 8, 9, 10), while Cs and Is are the contribution and inhibitory ratios, respectively, of standard control groups (Group no. 3, 4, 5, 6); I is the inhibitory ratio of the background drug group (Group no. 2), and n is the content multiple of addition.
1	Control	0.01320 ± 0.001564	—	—
2	HSP	0.009158 ± 0.001983**	30.62	—
3	Chasmanine	0.01357 ± 0.001345	−2.78	−4.54
4	Mesaconitine	0.009880 ± 0.001911**	25.15	41.07
5	Hypaconitine	0.01083 ± 0.001009*	17.97	29.33
6	Deoxyaconitine	0.01085 ± 0.001791*	17.80	29.07
7	HSP + chasmanine	0.008200 ± 0.001121**	37.88	11.85
8	HSP + mesaconitine	0.01020 ± 0.001441*	22.73	−12.89
9	HSP + hypaconitine	0.006643 ± 0.002535**^#	49.68	31.11
10	HSP + deoxyaconitine	0.008750 ± 0.002203**	33.71	5.05

---

Hypaconitine alone showed significant anti-inflammatory effects, and addition of hypaconitine to the original HSP extract significantly enhanced its efficacy (P < 0.05) and contribution ratio in the entire TCM environment. Hypaconitine can thus be regarded as one of the anti-inflammatory PCs of HSP. Chasmanine alone did not significantly inhibit mice ear swelling, whereas deoxyaconitine alone did. However, when both components were added directly, HSP efficacy increased. This result revealed a significant relation of these components with efficacy, and the two components may also be regarded as anti-inflammatory PCs. Mesaconitine alone could inhibit mice ear swelling significantly, but the inhibition ratio was reduced when the component was quantitatively added to the extract. This result showed that mesaconitine was negatively correlated with anti-inflammatory effects and could not be an anti-inflammatory PC of HSP.

Thus, the results showed that the overall complex environment of TCM greatly impacted the action of individual components. Overall efficacy changes may be caused by synergism or antagonism among the complex chemical components of TCM during absorption, distribution, metabolism, and excretion. The SA strategy employed in this work revealed that hypaconitine, deoxyaconitine, and chasmanine can be identified as anti-inflammatory PCs in HSP but mesaconitine can not.

In our work, the TCs were screened by SER, and the SA method was used to verify the results. We found that some TCs, such as mesaconitine, may not be PCs. The SA results obtained in our experiments may be more reliable. We also compared the SA and the approach that separated individual chemical components and evaluated each component activity. As we know, the references were obtained from the separation of special factories or research institutes. After evaluating their activity, we found that individual chemical components cannot represent the whole extracts of TCM. Therefore, research on TCM should not ignore its wholeness and system.

Li et al. proposed “target constituent knock-out” strategy to screen PCs in TCM. This approach completely considers the complex environment of TCM and interaction of the bioactive components in TCM.^12,13 To confirm our results, we also tried to isolate mesaconitine from the HSP as much as possible and compared the anti-inflammatory efficacy of mesaconitine sample, negative sample (lacking mesaconitine) and overall HSP. However, we encountered some problems when isolating mesaconitine using the existing separation or analysis techniques. We found that some chemical components of HSP were unsteady, and their transformation, hydrolysis and pyrolysis took place easily. According to the literature, the chemical components of HSP can be divided into several types, namely, DDAs, MDAs, alkylolamine-diterpenoid alkaloid (ADAs), lipo-alkaloids (LDAs) and other type of alkaloids. DDAs are unsteady and can be easily turned into other compounds; while mesaconitine belonged to DDAs.^38–41 Removing some of the unstable compounds was difficult. Thus, it appears that the “target components knock-out strategy” is more suitable for studying Chinese medicines whose chemical compositions are relatively stable and simple. It may not be suitable for TCMs containing complex unstable ingredients. The SA strategy that we proposed for screening PCs also embodies the wholeness and system of TCM and effectively avoids the difficulty in removing TCs. To a certain degree, the SA strategy makes up for the “target components knock-out” strategy; it can be used to screen PCs of TCM with complex unstable compounds.

In order to systematically screen PCs of HSP, we will try to use other methods to isolate mesaconitine in future studies. To avoid mesaconitine turning into other compounds, the operational condition, such as the temperature and polarity of the solution, must be strictly controlled during the experiment. Because the purity of removed samples and the completeness of negative samples must be ensured, accurate identification of mesaconitine and the negative sample must be done.

4 Conclusions

We established a practical and novel SA strategy to study PCs in TCM. SA can systematically evaluate the effect of chemical ingredients in the whole TCM and represent the complexity and multi-component effects of TCM. In addition, SA features simple operation, strong feasibility and reliability, and applies to most TCM or its preparation. PC determination in TCM has great significance for future drug development and quality control of TCM. The SA strategy not only makes up for the deficiencies of other methods but also solves critical issues of PC identification. Thus, the present work provides a novel method for studying PCs in TCM.

Acknowledgements

This project was supported by the National Basic Research Program of China (973 Program) (2011CB505300, 2011CB505302) and the National Undergraduate Training Programs for Innovation and Entrepreneurship of China.

Notes and references

1. T. H. Xue and R. Roy, Science, 2003, 300, 740.
2. P. Li, L. W. Qi, E. H. Liu, J. L. Zhou and X. D. Wen, Trends Anal. Chem., 2008, 27, 66.
3. D. Normile, Science, 2003, 299, 188.
4. J. C. Tilburt and T. J. Kaptchuk, Bull. W. H. O., 2008, 86, 594.
5. R. Stone, Science, 2008, 319, 709.
6. D. A. Guo, A. P. Lu and L. Liu, J. Ethnopharmacol., 2012, 141, 547.
7. X. D. Huang, L. Kong, X. Li, X. G. Chen, M. Guo and H. F. Zou, J. Chromatogr. B, 2004, 812, 71.
8. Y. Jiang, B. David, P. F. Tu and Y. Barbin, Anal. Chim. Acta, 2010, 657, 9.
9. B. K. Shoichet, Nature, 2004, 432, 862.
10. W. J. Kong, Y. L. Zhao, X. H. Xiao, J. B. Wang, H. B. Li, Z. L. Li, C. Jin and Y. Liu, Anal. Chim. Acta, 2009, 634, 279.
11. C. Tistaert, B. Dejaegher, N. Nguyen Hoai, G. Chataigné, C. Rivière, V. Nguyen Thi Hong, M. Chau Van, J. Quetin-Leclercq and Y. Vander Heyden, Anal. Chim. Acta, 2009, 649, 24.
12. Y. Liu, J. L. Zhou, P. Liu, S. Sun and P. Li, J. Chromatogr. A, 2010, 1217, 5239.
13. W. J. Kong, J. B. Wang, Q. C. Zang, C. Jin, Z. W. Wang, X. Y. Xing, Y. Y. Wu, Y. L. Zhao, M. H. Yang and X. H. Xiao, J. Chromatogr. B, 2011, 879, 3565.
14. L. Y. Dong, Y. Luo, B. F. Cheng, Y. S. Zhang, N. Zhang, Y. Y. Hou, M. Jiang, G. A. Luo and G. Bai, Biomed. Chromatogr., 2013, 27, 960.
15. T. Uto, I. Tuvshintogtokh and Y. Shoyama, Curr. Drug Discovery Technol., 2011, 8, 16.
16. D. Yan, J. X. Li, Y. Xiong, C. G. Zhang, J. Y. Luo, Y. M. Han, R. L. Wang, C. Jin, H. Qian, J. G. Li, L. L. Qiu, C. Peng, Y. L. Lin, X. A. Song and X. H. Xiao, Sci. Rep., 2014, 4, 3668,  DOI:10.1038/srep03668.
17. H. Wei, L. N. Sun, Z. G. Tai, S. H. Gao, W. Xu and W. S. Chen, Anal. Chim. Acta, 2010, 662, 97.
18. J. H. Shang, X. H. Cai, T. Feng, Y. L. Zhao, J. K. Wang, L. Y. Zhang, M. Yan and X. D. Luo, J. Ethnopharmacol., 2010, 129, 174.
19. H. Hosseinzadeh, M. Khoshdel and M. Ghorbani, J. Acupunct. Meridian Stud., 2011, 4, 242.
20. D. Z. Yang, Y. Q. An, X. L. Jiang, D. Q. Tang, Y. Y. Gao, H. T. Zhao and X. W. Wu, Talanta, 2011, 85, 885.
21. S. Y. Xu, R. L. Bian and X. Chen, The method of pharmacology experiment, People's medical publishing house, Beijing, China, 2001.
22. G. L. Xu, M. Xie, X. Y. Yang, Y. Song, C. Yan, Y. Yang, X. Zhang, Z. Z. Liu, Y. X. Tian, Y. Wang, R. Jiang, W. R. Liu, X. H. Wang and G. M. She, Molecules, 2014, 19, 17897.
23. J. X. Wang, X. Tong, P. B. Li, M. H. Liu, W. Peng, H. Cao and W. W. Su, J. Ethnopharmacol., 2014, 155, 405.
24. J. Wang, H. W. Kong, Z. M. Yuan, P. Gao, W. D. Dai, C. X. Hu, X. Lu and G. W. Xu, J. Pharm. Biomed. Anal., 2013, 83, 57.
25. P. S. Xie, S. B. Chen, Y. Z. Liang, X. H. Wang, R. T. Tian and R. Upton, J. Chromatogr. A, 2006, 1112, 171.
26. W. J. Kong, Y. L. Zhao, X. H. Xiao, C. Jin and Z. L. Li, Phytomedicine, 2009, 16, 950.
27. S. Wold, M. Sjöström and L. Eriksson, Chemom. Intell. Lab. Syst., 2001, 58, 109.
28. B. S. Elif Seyma Bayrak, B. S. Kamuran Turksoy, P. D. Ali Cinar, P. D. R. N. Lauretta Quinn, M. D. Elizabeth Littlejohn and P. D. Derrick Rollins, J. Diabetes Sci. Technol., 2013, 7, 206.
29. Y. Akimoto, K. Yugi, S. Uda, T. Kudo, Y. Komori, H. Kubota and S. Kuroda, PLoS One, 2013, 8,  DOI:10.1371/journal.pone.0072780.
30. S. Thiangthuma, B. Dejaegher, M. Goodarzi, C. Tistaert, A. Y. Gordien, N. Nguyen Hoai, M. Chau Van, J. Quetin-Leclercq, L. Suntornsuk and Y. Vander Heydena, J. Chromatogr. B, 2012, 910, 114.
31. T. Mehmood, K. H. Liland, L. Snipen and S. Sæbø, Chemom. Intell. Lab. Syst., 2012, 118, 62.
32. R. Gosselin, D. Rodrigue and C. Duchesne, Chemom. Intell. Lab. Syst., 2010, 100, 12.
33. I. G. Chong and C. H. Jun, Chemom. Intell. Lab. Syst., 2005, 78, 103.
34. J. S. Wang, R. Heijden, S. Spruit, T. Hankermeier, K. Chan, J. Greef, G. W. Xu and M. Wang, J. Ethnopharmacol., 2009, 126, 31.
35. J. Singhuber, M. Zhu, S. Prinz and B. Kopp, J. Ethnopharmacol., 2009, 126, 18.
36. H. Hikino, H. Takata, M. Fujiwara, C. Konno and K. Ohuchi, Eur. J. Pharmacol., 1982, 82, 65.
37. Y. V. Nesterova, T. N. Povetieva, N. I. Suslov, G. N. Zyuz'kov, S. G. Aksinenko, S. V. Pushkarskii and A. V. Krapivin, Bull. Exp. Biol. Med., 2014, 156, 665.
38. M. Zhang, Y. Long, Y. Sun, Y. K. Wang, Q. Li, H. J. Wu, Z. J. Guo, Y. H. Li, Y. B. Niu, C. Li, L. Liu and Q. B. Mei, Eur. J. Pharmacol., 2011, 651, 187.
39. Y. Yang, X. J. Yin, H. M. Guo, R. L. Wang, R. Song, Y. Tian and Z. J. Zhang, Chin. J. Nat. Med., 2014, 12, 542.
40. H. Yue, Z. F. Pi, H. L. Li, F. R. Song, Z. Q. Liu and S. Y. Liu, Phytochem. Anal., 2008, 19, 141.
41. H. Sun, B. Ni, A. H. Zhang, M. Wang, H. Dong and X. J. Wang, Analyst, 2012, 137, 170.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra00461f

---