Quantitative analysis of five toxic alkaloids in Aconitum pendulum using ultra-performance convergence chromatography (UPC²) coupled with mass spectrometry

Abstract

A rapid and efficient ultra-performance convergence chromatography (UPC²) method coupled with electrospray ionization single quadrupole mass spectrometry (ESI-MS) was developed and validated for the simultaneous quantification of five diester diterpenoid alkaloid constituents (3-acetylaconitine, hypaconitine, deoxyaconitine, mesaconitine, aconitine) in Aconitum pendulum. Optimum separation was achieved on a BEH 2-EP C₁₈ column (2.1 × 150 mm i.d., 1.7 μm particle) with a gradient elution of a mixture of A (supercritical CO₂) and B (methanol containing 10 mmol L⁻¹ ammonium acetate) and at a flow rate of 0.8 mL min⁻¹ within 3 minutes. Quantification was performed using mass spectrometry in a positive ion ionization mode and selected ion recording (SIR) mode. The influences of column, modifier, additive, column temperature, and back pressure were investigated. The five alkaloids were identified and quantified using a comparison of retention time, ultraviolet spectrum, molecular ion peak (obtained from a selective ion recording mode) and peak areas with the reference compounds. The method was validated through linearity, limits of detection, limits of quantification, precision, stability, repeatability, and accuracy. The validated method was applied to analyze A. pendulum, which provided a reference for the quality evaluation of A. pendulum.

---

1. Introduction

Tibetan medicines, which have a history of more than 2500 years, have been attracting increasingly wide attention for their significant curative effects. However, a considerable number of them not only possess significant biological activities but also exhibit considerable toxicity, such as Aconitum. Aconitum is a genus of about 400 diversified species of herbaceous plants belonging to the family of Ranunculaceae, native to temperate regions of the northern hemisphere.¹ There are 211 species in China, of which 166 are endemic.^2–4 Although most Aconitum species possess excellent analgesic, anti-rheumatic and anti-arrhythmic effects, their extremely high toxicity is the principal obstacle against their extensive medical use.¹ Natural pharmaceutical chemistry studies have revealed that diterpenoid alkaloids are the main constituent accumulated in many plants of the Aconitum species, which is responsible for both their biological activity and high toxicity.⁵

Aconitum pendulum Busch, known by the name Xueshang Yizhihao in Chinese, is a valuable Tibetan medicine among the Aconitum species due to its analgesic, anti-inflammatory and antibacterial activities, and its therapeutic effects of invigorating blood circulation and dispelling rheumatism. A. pendulum is widely distributed in the mountain grassy slopes and forest margins of the Qinghai–Tibet plateau, Yunnan province, Sichuan province, Gansu province and Shanxi province in China, at an altitude range of 2300–4500 m.⁶ In the previous phytochemical studies, a number of alkaloids, such as aconitine, deoxyaconitine, 3-acetylaconitine, hypaconitine, mesaconitine, 15a-hydroxyneoline, 8-O-acetyl-15a-hydroxyneoline, 14-benzoyl-8-O-methylaconine, neoline, benzoylaconine, polyschistine A, polyschistine D, N-deethyl-3-acetylaconitine, N-deethyldeoxyaconitine, secoaconitine, benzoyldeoxyaconitine, aconine, dehydrolucidusculline and dehydronapelline, have been isolated from A. pendulum.^7–10 Among these alkaloids, the diester diterpenoid alkaloids (DDAs) have captured great attention for their high toxicity and wide range of bioactivities.^11–18 For example, aconitine, an extremely toxic ingredient of A. pendulum, possessing a narrow therapeutic index, has striking pharmacological effects such as anti-inflammatory and antinociceptive properties.^19,20 The poisonous dose of aconitine for humans is estimated to be 0.2 mg, and the lethal dose is 1–2 mg. Several fatal accidents have been reported for the administration of the raw material. In general, the raw material of A. pendulum is required for a series of processing steps, such as boiling, to reduce its toxicity prior to being used in clinical practice. Therefore, good quality control of this plant is needed to evaluate its toxicological risk and to guarantee its safe use. The related research, however, is still rather limited. Considering the vital role in quality control system played by DDAs, it is a requisite to develop a sensitive and reliable analytical method to quantify DDAs in A. pendulum.

For the analysis of DDAs, advanced chromatographic techniques, including gas chromatography (GC), high performance liquid chromatography (HPLC), ultra performance liquid chromatography (UPLC), capillary electrophoresis (CE) and certain hyphenated instrumental techniques have been utilized.^11,21–25 In light of the characteristics of high molecular weight and low volatility for DDAs, GC is undesirable. Thanks to the technological integration of supercritical fluid chromatography (SFC) and UPLC, ultra-performance convergence chromatography (UPC²) has provided a new choice for the analysis of DDAs.²⁶ In contrast with GC and liquid chromatography (LC), the separation performance of UPC² depends not only on the interaction between mobile phase and stationary phase, but also on the density of carbon dioxide (CO₂) which relies on temperature and pressure.²⁷ Moreover, coupled with mass spectrometry, UPC² can provide higher sensitivity and selectivity for the detection of targets. Nowadays, UPC² has been applied in the areas of foods^26,28,29 and drug safety,³⁰ but its application in Tibetan medicines is still limited.

In this work, a sensitive and reliable UPC² coupled with electrospray ionization mass spectrometry method was established for the simultaneous quantification of five DDAs, including 3-acetylaconitine, hypaconitine, deoxyaconitine, mesaconitine and aconitine in the roots of A. pendulum using a single run. The proposed method was validated and applied to determine five batches of A. pendulum collected from different regions. In addition, the proposed method was compared with the reported HPLC-UV, UPLC-UV and UPLC-MS methods in the literature, since these methods are popular in the field of medicinal analysis.

2. Materials and methods

2.1. Chemicals and reagents

Aconitine was purchased from Beijing H&Q Chemical Institute and Beijing Aoke Biological technology Co., Ltd. (Beijing, China). 3-Acetylaconitine was purchased from Beijing Beina Chuanglian Biotechnology Research Institute (Beijing, China). Mesaconitine, hypaconitine and deoxyaconitine were purchased from Chengdu Herb purify Co., Ltd. (Chengdu, China). The purities of the above five standards were all above 98% and their chemical structures are shown in Fig. 1. Five batches of A. pendulum were obtained from Gansu province (batches S1, S2, S3) and Qinghai province (batches S4, S5).

Fig. 1 Chemical structures of the five aconitum alkaloids.

Chromatographic-grade methanol and acetonitrile were purchased from Merck Co. (Darmstadt, Germany). Chromatographic-grade isopropanol and other chemicals of analytical grade were purchased from Tianjin Chemical Reagent Co. (Tianjin, China). Carbon dioxide (99.999% purity) was purchased from Zhongke Kaite Industry and Trade Co., Ltd. (Lanzhou, China). Ultra-pure water was prepared using an OKPVRE water ultrapure system (Shanghai, China).

2.2. Apparatus and UPC²-MS conditions

UPC²-MS analysis was performed on a Waters ACQUITY ultra-performance convergence chromatography (UPC²) system (Milford, MA, USA) using a SQ Detector 2 tandem mass spectrometer (Waters, USA). The UPC² system was equipped with a binary solvent manager, fixed loop sample manager, column manager, auxiliary manager, convergence manager which controls back pressure, and a photodiode array detector. The UPC² analysis was conducted on a Waters Acquity UPC²™ BEH 2-EP C₁₈ column (2.1 × 150 mm i.d., 1.7 μm particle), using a linear gradient elution of (A) supercritical CO₂ and (B) methanol with 10 mmol L⁻¹ ammonium acetate at a flow rate of 0.8 mL min⁻¹. The gradient elution program was as follows: 0–3 min, A 93–87%; 3–4 min, A 87–93%. The system was re-equilibrated with 93% A for 2 min before the next sample run. The back pressure was set at 2100 psi. The temperatures of the column and sample manager room were maintained at 55 °C and 18 °C, respectively. The injection volume was 1 μL, and partial loop with needle overfill was applied for sample injection. Methanol and methanol/isopropanol (1/1, v/v) were used as the strong and weak needle wash, respectively. The absorption spectra of the compounds were recorded in the range of 200–400 nm, and the detection wavelength was set at 225 nm with compensation from 350 to 400 nm. The mass spectrometer was equipped with an electrospray ionization (ESI) source, and the MS analysis was performed in a positive ion ionization mode of selected ion recording (SIR). Quantification of the analytical compounds was performed by employing the SIR mode. The effluent from the PDA cell outlet was split to MS and convergence manager by a splitter so that the pressure of CO₂ and modifier could be maintained. The MS analysis conditions were optimized as follows. The source temperature and the desolvation temperature were maintained at 150 °C and 350 °C, respectively. The capillary voltage and cone voltage were fixed at 2.3 kV and 70 V, respectively. The flow rates of desolvation gas and cone gas (nitrogen was used) were 600 L h⁻¹ and 50 L h⁻¹, respectively. Instrument control, data acquisition and processing were performed using a Masslynx 4.1 workstation (Waters, USA).

2.3. Preparation of standard solutions

Stock standard solutions were prepared in methanol at a concentration of 0.5 mg mL⁻¹. A mixed standard solution was prepared by mixing the five stock standard solutions at a concentration of 60 μg mL⁻¹. Calibration standard working solutions were freshly prepared using serial dilutions of the mixed standard solution to obtain final concentrations of 0.1, 1, 10, 50, 100 and 150 ng mL⁻¹. All the standard solutions prepared above were stored at 0–4 °C prior to analysis.

2.4. Preparation of sample solutions

After being air-dried and crushed into powder, 0.1 g of the A. pendulum roots was accurately weighed and placed in a 50 mL Erlenmeyer flask with a stopper. 2.0 mL ammonia solution and 30 mL diethyl ether were then successively added. The mixture was sonicated for 30 min (80 kHz), and then left at room temperature for 8 hours. The supernatant was collected and the residues were sonicated for another 30 min with 20 mL of diethyl ether. The supernatant was collected and the residue was washed with 10 mL diethyl ether three times. The extracts were combined and then concentrated to dry. The solid residue was re-dissolved with 60 mL of methanol, then the insoluble substances were removed by filtration and the filtrate was concentrated to a final volume of 50 mL. All the sample solutions were diluted 10 times and the dilution solutions were filtered through 0.2 μm membranes before being injected into the UPC²-MS system. All the solutions were stored at 0–4 °C prior to analysis.

2.5. Evaluation of the method

The method was validated for linearity, limits of detection (LOD), limits of quantification (LOQ), precision, stability, repeatability, and accuracy according to the guidance for method validation for traditional Chinese medicines in Chinese Pharmacopoeia.³¹

3. Results and discussion

3.1. Optimization of UPC² conditions

In order to obtain a good resolution within a reasonable analysis time, optimization of chromatographic parameters was performed by investigating the influence of column, mobile phase, flow rate, column temperature, back pressure, and injection volume.

In this study, three columns were examined to perform the experiments including Waters Acquity UPC² BEH 2-EP C₁₈ (2.1 mm × 150 mm, 1.7 μm), BEH C₁₈ (3.0 mm × 100 mm, 1.7 μm), and CSH Fluoro-Phenyl (2.1 mm × 150 mm, 1.7 μm) column. The BEH 2-EP C₁₈ column resulted in better resolution and peak shape within a short analysis time (within 3 minutes). Using a BEH C₁₈ column, the compounds of aconitine and hypaconitine could not be baseline separated, and the analysis time was longer than the others. Although five compounds are eluted with a good baseline resolution and peak shape within a short analysis time (within 5 minutes) using the CSH Fluoro-Phenyl column, the system pressure was much too high (close to 6000 psi). There is no doubt that the system carries a considerable burden. So, the BEH 2-EP column was selected in the subsequent experiments.

In order to improve the separation and the shape of the peaks, different modifiers (including methanol, methanol : acetonitrile, and methanol : isopropanol) and additives (including ammonium formate, ammonium acetate, and formic acid) were investigated. The results revealed that when the mixed modifiers methanol : acetonitrile and methanol : isopropanol were used, compounds deoxyaconitine and mesaconitine, and hypaconitine and deoxyaconitine could not be baseline separated. The five compounds, however, could be baseline separated by applying methanol as a modifier.

Furthermore, the influence of additive on peak shape was compared. When 10 mmol L⁻¹ ammonium formate was added into methanol as the additive, the compounds of mesaconitine and hypaconitine could not be separated. When 0.1% formic acid was used as the additive, no chromatographic peaks were found clearly. It seems that the chemical structures of aconitum alkaloids are unstable under acidic conditions in methanol. In other words, the chromatographic separation depends on the pH of the solution. The 10 mmol L⁻¹ ammonium acetate additive showed the best results in terms of retention time and resolution and thus was selected as the additive. So, the mixed solution of CO₂/methanol with 10 mmol L⁻¹ ammonium acetate was used as the mobile phase for UPC² analysis. The most suitable injection volume and flow rate were set as 1 μL and 0.8 mL min⁻¹, respectively.

Increasing column temperature has a certain influence on the separation selectivity by decreasing the viscosity of the methanol.³² In our work, column temperature was examined from 35 to 70 °C. As shown in Fig. 2a, with an increase in column temperature, the five aconitum alkaloids had longer retention times and broadened peaks. In the study, an optimal temperature of 55 °C was selected for the supercritical fluid chromatography analysis.

Fig. 2 Effects of (a) column temperature and (b) back pressure on the retention time of the analytes. Symbol marking: (1) 3-acetylaconitine; (2) hypaconitine; (3) deoxyaconitine; (4) mesaconitine; (5) aconitine.

Back pressure is a very vital factor which could influence the interaction between the analytes and mobile phase by changing the density of supercritical carbon dioxide in UPC². In our work, back pressure was examined from 1500 to 2300 psi. As shown in Fig. 2b, with increasing back pressure, the retention times of the five aconitum alkaloids decreased and the peaks become sharper. Meanwhile, the system pressure will increase. Combined with the effect of column temperature, the optimum back pressure chosen was 2100 psi.

3.2. MS analysis of five aconitum alkaloids

The reference compounds of five aconitum alkaloids were used to optimize the MS parameters, and to identify the corresponding compounds present in this medicinal plant. The effects of desolvation temperature, source temperature, desolvation gas flow, capillary voltage, cone voltage, cone gas flow and ionization mode were separately examined. The source temperature and cone gas (nitrogen) flow were fixed at 150 °C and 50 L h⁻¹, respectively. Other parameters were varied as follows: desolvation temperature (300, 350, 400, and 450 °C), desolvation gas (nitrogen) flow (600, 700, 800, 900, and 1000 L h⁻¹), capillary voltage (2.3, 2.5, 2.8, and 3.0 kV), and cone voltage (30, 40, 50, 60, 65, and 70 V). The trials showed that the positive ion mode fit the detection of the five aconitum alkaloids better than the negative ion mode, and that cone voltage influenced the ionization significantly. The optimum MS conditions were obtained after several trials.

In a positive ion mode, [M + H]⁺ ions were observed to be the most abundant ions, and the molecular weights were determined based on the formation of [M + H]⁺ ions. Herein, five compounds were identified and quantized by comparing the retention time, ultraviolet spectrum and mass data with those of the reference compounds. The UPC² chromatogram and total ion chromatogram of a mixed standard solution are shown in Fig. 3a and b, respectively.

Fig. 3 UPC² chromatograms at 225 nm and total ion chromatograms obtained from (a and b) a mixed standard solution and (c and d) a real sample solution. Peak numbering is the same as for Fig. 2.

3.3. Method validation

3.3.1. Calibration plots, LODs and LOQs. Under the optimal conditions, calibration standard working solutions at six concentrations (0.1, 1, 10, 50, 100 and 150 ng mL⁻¹) were analyzed using the UPC²-MS method in SIR mode, and three duplicate analyses were performed for each concentration level. The calibration curves were established by plotting the peak areas against the concentrations of analytes with linear regression analysis. The regression equations were expressed as y = ax + b, where y is peak area and x is concentration (ng mL⁻¹) of the analytes. LOD and LOQ were determined as the concentration at the signal to noise ratios of 3 and 10 times, respectively. The regression equations, correlation coefficients (r), linear ranges, LODs, and LOQs are listed in Table 1.

Table 1 Calibration plots, LODs and LOQs

Analyte	Regression equation	r	Linear range (ng mL^−1)	LOD (ng mL^−1)	LOQ (ng mL^−1)
3-Acetylaconitine	y = 52 314x + 13 174	0.9994	0.1–150	0.013	0.027
Hypaconitine	y = 60 021x + 57 111	0.9991	0.1–150	0.016	0.044
Deoxyaconitine	y = 83 512x + 14 836	0.9995	0.1–150	0.034	0.051
Mesaconitine	y = 57 768x + 59 934	0.9997	0.1–150	0.011	0.042
Aconitine	y = 60 341x + 10 002	0.9997	0.1–150	0.029	0.077

---

3.3.2. Precision, stability and repeatability. A real sample solution (S2), in which the concentrations of 3-acetylaconitine, hypaconitine, deoxyaconitine, mesaconitine and aconitine were 2.0, 0.6, 0.9, 4.6 and 32.1 ng mL⁻¹, respectively, was used for precision and stability studies. The precision was evaluated by performing intra-day and inter-day variation with consecutive injection of the sample solution. Intra-day variation was estimated using five successive injections within a day, and inter-day variation was measured on five consecutive days. For the stability test, the sample solutions were analyzed at 2 h intervals during storage for 12 h at room temperature. The sample solution can be regarded as stable within 12 h because the RSD% values of both retention times and peak areas were <2%. The specific data are listed in Table 2.

Table 2 The RSD% values for precision, repeatability and stability

Analyte	Intra-day precision		Inter-day precision		Stability		Repeatability
	RT^a	PA^b	RT	PA	RT	PA	Content
a RT: retention time.b PA: peak area.
3-Acetylaconitine	0.0	1.2	0.4	3.2	0.0	1.9	0.6
Hypaconitine	0.2	1.5	0.2	3.2	0.2	1.7	1.9
Deoxyaconitine	0.2	2.6	0.3	1.7	0.4	0.9	1.4
Mesaconitine	0.2	3.9	0.2	4.7	0.0	1.9	0.7
Aconitine	0.1	4.3	0.1	4.9	0.2	1.4	0.1

---

Repeatability was tested using injections of six different sample solutions, which were prepared in parallel with the same batch sample (S2) according to the procedure described above, and obtained from the RSD% of the component content. The specific data are listed in Table 2.

3.3.3. Accuracy. The accuracy of the method was investigated using a spike recovery test. Different amounts of the standards at three levels (low, medium and high level) were separately spiked into an originally analyzed real sample (S2), for which the contents of the compounds of interest were already known. Three sets of spiked samples were then treated according to the procedure described in the “preparation of sample solution” section, and analyzed using UPC²-MS in the SIR mode. The accuracy was expressed by the recovery, which was calculated using the following formula: recovery (%) = (found amount − original amount) × 100%/added amount. The results of the recovery test shown in Table 3 indicate that the proposed method enables highly accurate simultaneous analysis of the five analytes in A. pendulum roots.

Table 3 Recovery studies for the determination of five alkaloids

Alkaloids	Original amount (μg)	Added amount (μg)	Found amount (μg)	Recovery (%)	Average recovery (%)	RSD (%)
3-Acetylaconitine	10.1200	7.83	17.6838	96.6	98.2	1.4
	10.0110	9.78	19.6639	98.7
	10.0230	12.30	22.2246	99.2
Hypaconitine	2.8336	2.32	5.0956	97.5	96.9	1.9
	2.8031	2.81	5.5681	98.4
	2.8064	3.36	5.9917	94.8
Deoxyaconitine	4.5540	3.66	8.1042	97.0	95.3	2.6
	4.5050	4.54	8.8816	96.4
	4.5104	5.42	9.5239	92.5
Mesaconitine	21.9604	17.40	39.4300	100.4	97.6	4.0
	21.7239	21.82	43.3912	99.3
	21.7499	26.12	46.0676	93.1
Aconitine	156.2528	123.86	278.2549	98.5	97.3	4.7
	154.5698	154.38	310.8024	101.2
	154.7551	186.44	326.8392	92.3

---

3.4. Comparison of analytical methods

In order to evaluate the analytical performance of the proposed UPC²-MS method, the HPLC and UPLC methods reported in the literature were brought as references for comparison, and the results are summarized in the ESI (Table S1†). Furthermore, to compare the differences between UPC² and UPLC more objectively, an additional experiment was carried out, in which UPLC coupled with a MS/MS method in a positive ion ionization mode of multiple reaction monitoring (MRM) was developed and applied to the determination of the same samples. The UPLC-MS/MS conditions, along with the UPLC chromatogram, total ion chromatogram and method validation parameters are listed in the ESI (SF1, Fig. S1 and Table S1†). For the HPLC-UV and UPLC-UV methods, the LOD and LOQ are much higher than that of the UPLC-MS and UPC²-MS methods. Thus, a highly concentrated sample is required for the quantitative analysis of DDAs using the UV detection method. However, this kind of sample can cause chromatographic column overload and pollution, therefore, UV detection is not suitable for analyzing DDAs at such a low concentration in complex sample matrices. Both UPLC and UPC² exhibit major advantages over conventional HPLC, such as increased peak capacity, shorter retention time and less solvent consumption, owing to the utilization of sub-2 μm particles as the stationary phase. It is worth mentioning that UPC² integrates supercritical fluid chromatography (SFC) and UPLC technologies, and thus shows some superiorities compared with UPLC. Supercritical CO₂, the main mobile phase in UPC², offers superior solubility for the analytes and induces strong non-polar interactions between the analytes and the mobile phase, also allowing a large flow rate and thereby reducing the retention time remarkably. As shown in Table S1,† the five DDAs were separated within a very short time using the presented method. In addition, UPC² can deliver a reduction in waste generation and disposal compared with UPLC. Furthermore, UPC², based on the principles of normal-phase LC, with the ease-of-use of reversed-phase LC, is suitable for separating compounds in a wider range of polarities. Despite the superiorities described above however, UPC² has not been a preferential and popular technique in the field of analysis at present, which might be attributed to the expensive instruments required, the limitations in instrument manufacturing and the limited recognition of this technology.

4. Sample analysis

The developed method was applied to analyze five DDAs in five batches of A. pendulum root samples collected from the Qinghai and Gansu provinces of China. The UPC² chromatogram and total ion chromatogram obtained from the real sample solutions are shown in Fig. 3c and d. Their contents are presented in Table 4. The total content of the DDAs in the five batches of samples ranged from 71.5 to 798.2 μg g⁻¹. The content of aconitine was higher than the other alkaloids, except in the batch of sample S4. The results also revealed that the contents of the five alkaloids fluctuated largely for different batches. This might be accounted for by the variations in growing environments, growing years, collection regions, harvest seasons, and storage conditions for the crude herbs. The obtained results further demonstrated the importance and necessity for monitoring of DDAs in A. pendulum. In addition, the RSD values of the contents of the five alkaloids in the same batches of samples determined using the UPC² and UPLC methods were less than 15%, which could further prove the applicability of the developed method in real samples.

Table 4 Contents of the five alkaloids in five different batches of A. pendulum root samples^a

Content (μg g^−1)	Sample batches
	S1	S2	S3	S4	S5
a ND: undetected.
3-Acetylaconitine	8.8	10.0	11.8	10.1	0.6
Hypaconitine	3.8	2.8	6.0	230.0	3.7
Deoxyaconitine	7.7	4.5	4.6	5.1	17.9
Mesaconitine	6.6	21.7	3.0	434.7	ND
Aconitine	133.9	154.4	147.3	113.3	48.3
Total	162.8	196.4	176.7	798.2	71.5

---

5. Conclusion

Monitoring the contents of DDAs in A. pendulum is required to evaluate its toxicological risk and to guarantee its safe use. In this work, a UPC² method coupled with mass spectrometry in positive ionization mode possessing high linearity, precision, stability, repeatability, and accuracy was developed for the simultaneous determination of five aconitum alkaloids in the roots of A. pendulum. The developed method involved the use of [M + H]⁺ ions in the positive ion mode with selective ion recording (SIR). The five aconitum alkaloid constituents were authenticated based on a comparison of their retention times, ultraviolet spectra and molecular weights with a reference substance. The established method was successfully applied to five batches of A. pendulum and the results exhibited a substantial fluctuation in the contents of these specific components. This work provides a method for evaluating the quality of A. pendulum.

Acknowledgements

The financial support of the National Natural Science Foundation of China (No. 21375136 and 21575150) and the Scholar Program of West Light Project, Chinese Academy of Sciences, are acknowledged.

References

1. P. G. Xiao, F. P. Wang, F. Gao, L. P. Yan, D. L. Chen and Y. Liu, Acta Phytotaxon. Sin., 2006, 44, 1–46.
2. L. Q. Li and Y. Kadota, Actonitum in Flora of China, ed. Z. Y. Wu and P. Raven, Science Press, and Missouri Botanical Garden Press, Beijing, China, St. Louis, MO, USA, 2001, vol. 6, pp. 149–222.
3. Y. L. Zhu and R. H. Zhu, Heterocycles, 1982, 17, 607–614.
4. S. H. Jiang, P. M. Yang and D. Y. Zhu, Planta Med., 2002, 68, 1147–1149.
5. S. Yang, X. D. Yang, J. F. Zhao, Y. Jin, H. B. Zhang and L. Li, Chem. Nat. Compd., 2008, 44, 334–336.
6. W. C. Wang, Aconitum L., in Flora Reipublicae Popularis Sinicae, Beijing, Science Press, 1979, vol. 27, pp. 113–362.
7. L. M. Liu, H. C. Wang and Y. L. Zhu, Acta Pharmacol. Sin., 1983, 18, 39–44.
8. Y. J. Wang, C. J. Zeng, Z. Yao, J. Zhang, Y. Zhang and F. Zhang, Chin. Tradit. Herb. Drugs, 2010, 41, 347–351.
9. S. M. Zhang, L. Q. Tan and Q. Y. Ou, Chin. Chem. Lett., 1997, 8, 967–970.
10. Y. J. Wang, J. Zhang, C. J. Zeng, Z. Yao and Y. Zhang, Phytochem. Lett., 2011, 4, 166–169.
11. G. H. Lu, Z. Q. Dong, Q. Wang, G. S. Qian, W. H. Huang, Z. H. Jiang, K. S. Leung and Z. Z. Zhao, Planta Med., 2010, 76, 825–830.
12. A. Ameri, Eur. J. Pharmacol., 1998, 342, 183–191.
13. H. K. Desai, B. P. Hart, R. W. Caldwell, J. Z. Huang and S. W. Pelletier, J. Nat. Prod., 1998, 61, 743–748.
14. Y. Suzuki, T. Oyama, A. Ishige, T. Isono, A. Asami, Y. Ikeda, M. Noguchi and Y. Omiya, Planta Med., 1994, 60, 391–394.
15. X. C. Tang, Z. G. Lin, W. Cai, N. Chen and L. Shen, Acta Pharmacol. Sin., 1984, 5, 85–89.
16. Y. Komoda, S. Nosaka and M. Takenoshita, J. Anesth., 2003, 17, 241–245.
17. U. T. Gutser, J. Friese, J. F. Heubach, T. Matthiesen, N. Selve, B. Wilffert and J. Gleitz, Naunyn-Schmiedeberg’s Archives of Pharmacology, 1997, 357, 39–48.
18. Y. Ohno, Toxin Rev., 1998, 17, 1–11.
19. J. H. Chen, C. Y. Lee, B. C. Liau, M. R. Lee, T. T. Jong and S. T. Chiang, J. Pharm. Biomed. Anal., 2008, 48, 1105–1111.
20. L. Tang, Y. Gong, C. Lv, L. Ye, L. Liu and Z. Q. Liu, J. Ethnopharmacol., 2012, 141, 736–741.
21. J. M. Zhang, W. Liao, Y. X. He, Y. He, D. Yan and C. M. Fu, J. Ethnopharmacol., 2013, 147, 128–135.
22. H. Yue, Z. F. Pi, H. L. Li, F. R. Song, Z. Q. Liu and S. Y. Liu, Phytochem. Anal., 2008, 19, 141–147.
23. L. Qiao, J. M. Jia, Z. Q. Wang, G. J. Wang, H. M. Zhu and Z. H. Meng, Chin. Tradit. Herb. Drugs, 2013, 44, 3452–3456.
24. Y. Bao, F. Yang and X. R. Yang, Electrophoresis, 2011, 32, 1515–1521.
25. B. Sun, L. Li, S. M. Wu, Q. Zhang, H. J. Li, H. B. Chen, F. M. Li, F. T. Dong and X. Z. Yan, Anal. Biochem., 2009, 395, 125–133.
26. X. Gong, N. L. Qi, X. X. Wang, L. J. Lin and J. H. Li, Food Chem., 2014, 162, 172–175.
27. D. P. Poe, D. Veit, M. Ranger, K. Kaczmarski, A. Tarafder and G. Guiochon, J. Chromatogr. A, 2014, 1323, 143–156.
28. Q. Zhou, B. Y. Gao, X. Zhang, Y. W. Xu, H. M. Shi and L. L. Yu, Food Chem., 2014, 143, 199–204.
29. Y. F. Tao, R. Paula, A. A. M. Stolker, D. M. Chen and Z. H. Yuan, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2015, 988, 143–148.
30. V. N. R. Ganipisetty, D. Jalandhar, G. Gnanadev, R. Bandari, P. Manoj, D. Dhananjay and R. V. Nadh, Anal. Methods, 2013, 5, 7008–7013.
31. Chinese Pharmacopoeia Commission, Chinese Pharmacopoeia, China Medical Science Press, Beijing, 2010, appendix XVIII A.
32. L. Toribio, J. L. Bernal, M. T. Martín, J. Bernal and M. J. del Nozal, Biomed. Chromatogr., 2014, 28, 152–158.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra21233b

---