Profiling the Change of Key Chemical Ingredients in Combination of Aconitum carmichaeli Debx. and Bletilla striata (Thunb.)Reichb.f. by UPLC-QTOF/MS with Multivariate Statistical Analysis

Wng Cho (王 超), Wng Yugung (王宇光),Go Yue (高 月)

a. Anhui Academy of Medical Science, Hefei 230061, China;

b. Department of Pharmacology and Toxicology, Beijing Institute of Radiation Medicine, Beijing, 100850, China.

ABSTRACT In the present study, an ultra performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry (UPLC-QTOF/MS) based chemical profiling approach to rapidly evaluate chemical diversity after codecocting of the combination of Aconitum carmichaeli Debx. (wu-tou in Chinese, WT) and Bletilla striata (Thunb.)Reichb.f. (bai-ji in Chinese, BJ) incompatible pair. Two different kinds of decoctions, namely WT-BJ mixed decoction:mixed water extract of each individual herbs, and WT-BJ co-decoction: water extract of mixed two constituent herbs,were prepared. Batches of these two kinds of decoction samples were subjected to UPLC-QTOF/MS analysis, the datasets of tR-m/z pairs, ion intensities and sample codes were processed with supervised orthogonal partial least squared discriminant analysis (OPLS-DA) to holistically compare the difference between these two kinds of decoction samples.Once a clear classification trend was found in score plot, extended statistical analysis was performed to generate S-plot,in which the variables (tR-m/z pair) contributing most to the difference were clearly depicted as points at the two ends of"S", and the components that correlate to these ions were regarded as the most changed components during co-decocting of the incompatible pair. The identities of the changed components can be identified by comparing the retention times and mass spectra with those of reference compounds and/or tentatively assigned by matching empirical molecular formulae with those of the known compounds published in the literatures. Using the proposed approach, global chemical difference was found between mixed decoction and co-decoction, and hypaconitine, mesaconitine, deoxyaconitine, aconitine,10-OH-mesaconitine, 10-OH-aconitine and deoxyhypaconitine were identified as the most changed toxic components of the combination of WT-BJ incompatible pair during co-decocting. It is suggested that this newly established approach could be used to practically reveal the possible toxic components changed/increased of the herbal combination taboos, e.g.the Eighteen Incompatible Medications (Shi Ba Fan), in traditional Chinese medicines.

KEYWORDS: Eighteen incompatible medications (Shi Ba Fan); UPLC-QTOF/MS; Aconitum carmichaeli Debx.(Wutou); Bletilla striata (Thunb.) Reichb.f. (Baiji); Complex sample profiling; Multivariate statistical analysis.

INTRODUCTION

For oriental medicines such as traditional Chinese medicines (TCM) in China and Kampo in Japan,medicinal herbs are often prescribed in the form of combinatorial formulae based on the traditional medicine philosophy. Chinese herbs are commonly used in a combination of two or more herbs, called formula, to mixed together and boiled with water to get decoction.

According to TCM theory, mutual incompatibility,firstly recorded in the earliest Chinese medicinal classic work, Shennong's Materia Medica, which dates back to a period around 200 B.C. to 200 A.D.,refers to the combination of two herbs that causes toxic or adverse side effects and should not be used simultaneously. Ancient medical literature summarized the incompatibility into two lists called the Eighteen Incompatible Medications (Shi Ba Fan) and the Nineteen Medications of Mutual Antagonism (Shi Jiu Wei). These traditional incompatibility lists has been lasting for at least one thousand years. In the Eighteen Incompatible Medications, it is acknowledged that Aconitum species are incompatible with Rhizoma Pinelliae (Banxia),Fructus Trichosanthes (Gualou), Bulbus Fritillariae(Beimu), Radix Ampelosis (Bailian), and Rhizoma Bletillae (Baiji). These incompatible taboos are also included in the Chinese Pharmacopoeia 2015 (CP 2015)[1],and are bound to be "mandatory" in clinical practice when TCM practitioners prescribe their herbal formulas. So far many of these taboos have not yet been investigated for their incompatibility.

Aconitum species have been used in China as an essential drug in TCM for two thousand years. The tubers and roots of Aconitum (Ranunculaceae) are commonly applied for various diseases, such as collapse, syncope,rheumatic fever, painful joints, gastroenteritis, diarrhea,oedema, bronchial asthma, various tumors, and some endocrinal disorders like irregular menstruation[2].Aconitum tubers.(wu-tou in Chinese, WT) is the mother root of Aconitum Carmichaeli Debx. (Ranunculaceae)and mainly distributed in northern hemisphere of China. As an herbal drug, it has been reputed to be an important remedy. Have analgetic, diuretics, antiinflammatory and cardiotonic actions[3]. At the same time, they also present enormous toxicity and there is a narrow margin of safety between a therapeutic and toxic dose. The toxicity of WT mainly derives from the diester diterpene alkaloids (DDAs), including aconitine (AC),mesaconitine (MA) and hypaconitine (HA), the three well known toxic alkaloids, and the LD50values in mice were 1.8, 1.9 and 5.8 mg/kg p.o., respectively[4]. They can be decomposed into less or non-toxic derivatives through Chinese traditional processing methods (in Chinese Paozhi), mostly decocting, which play an essential role in detoxification[5]. But many monographs of TCM indicated that WT cannot be co-used with Bletilla striata(Thunb.) Reichb.f. (bai-ji in Chinese, BJ), as WT-BJ is a typical "eighteen incompatible" medicaments, which will increase the toxicity of WT while decocting them together (co-decocting). These incompatible pairs in herbal formulations mostly have not yet been validated scientifically for their incompatibility. Investigating these taboos may generally help to understand why some herbal drugs must not be combined.

Ultra performance liquid chromatography coupled with quadrupole time-of-flight mass spectrometry (UPLCQTOF/MS) is a newly developed hyphenated technique.By employing sub-2μm particles column, the enhanced retention time reproducibility, high chromatographic resolution, improved sensitivity and increased operation speed of UPLC[6]as well as the accurate mass values and fragment ions of QTOF/MS have made UPLC-QTOF/MS a powerful tool for the complex sample analysis,especially in the complex sample analysis of TCM[7].

In this study, using WT-BJ incompatible pair as an example, an approach using UPLC-QTOF/MS with multivariate statistical analysis was developed to rapidly find potential chemical markers for the toxicity increase of the incompatible pair during co-decocting. Batches of mixed decoction and co-decoction samples of WT-BJ were determined by UPLC-QTOF/MS. The datasets of tR-m/z pair, ion intensity and sample code were subjected to multivariate statistical analysis using MarkerLynx software to holistically compare the difference between mixed decoction and co-decoction samples. Once a clear cluster was found, extended statistical analysis was performed to find the changed ions that contribute most to the difference. By comparing the retention time and mass spectra with those of the reference compounds and/or tentatively assigned by matching the empirical molecular formula with that of the published compounds, the most changed chemical markers can be identified.

EXPERIMENTAL

Chemicals, solvents and herbal materials

Standards of aconitine (110720-200410),hypaconitine (110799-200404), and mesaconitine(110798-200404) were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). The purity was determined to be higher than 98% by HPLC–DAD (MS).

Aconitum carmichaeli Debx. and Bletilla striata(Thunb.) Reichb.f. were purchased from AnGuo, Hebei Province, and were authenticated according to the standards documented in Chinese Pharmacopoeia by Dr.Bai-Ping Ma, a pharmacognosist in our team.

HPLC-MS grade acetonitrile were purchased from Fisher (USA), and MS grade formic acid from Merck(Darmstadt, Germany). Purified water was prepared using a Millipore water purification system (Bedford, MA, USA).

Liquid chromatography

UPLC was performed using a Waters Acquity UPLC? system (Waters, Milford, MA, USA), equipped with a binary solvent delivery system, an autosampler,and a photodiode-array detection (PDA) system. The chromatography was performed on a Waters Acquity BEH C18column (2.1mm×100mm, 1.7 μm, Waters,Milford, MA, USA). The mobile phase consisted of (A)water containing 0.1% formic acid and (B) acetonitrile.The UPLC eluting conditions were optimized as follows:linear gradient from 1% to 3% B (0–1.5 min), 3% to 32%B (1.5–13.5 min), 32% to 60% B (13.5–18.5 min), 60% to 99% B (18.5–20.5 min), 99% to 1% B (20.5–22 min). The flow rate was 0.4 ml/min. The column and autosampler were maintained at 45 ℃ and 4 ℃, respectively. The injection volume is 5 μl.

Mass spectrometry

Mass spectrometry was performed using a Waters Q-TOF Premier? (Micromass MS Technologies,Manchester, UK) equipped with an electrospray ionization (ESI) source operating in positive ion mode. The desolvation gas flow was set to 900 l/h at temperature of 300 ℃, the cone gas set to 50 l/h, and the source temperature set to 120 ℃. The capillary voltage and sampling cone voltage were set to 3000 V and 35 V, respectively. The collision energy was set at 20eV and 45eV to 70eV for the MS/MS analyses. The TOF acquisition rate was set to 0.15 s, with a 0.02 s inter-scan delay.

Accurate mass measurement

All MS data were acquired using the LockSpray?to ensure mass accuracy and reproducibility. The [M+H]+ions of Leucine-enkephalin at m/z 556.2771 were used as the lock mass in positive electrospray ionization modes.The concentration of Leucine-enkephalin was 50 pg/μl and the infusion flow rate was 10 μl/min. Centroided data were acquired for each sample from 100 to 900 Da,and dynamic range enhancement (DRE?) was applied throughout the MS experiment to ensure accurate mass measurement over a wide dynamic range.

Sample preparation

For co-decoction, sliced samples of WT(approximately 10 g) and BJ (approximately 10 g) were accurately weighed and mixed, the mixed herbs were put into a flask containing 200 ml distilled water, and boiled for 60 min. This procedure was then repeated by adding 100 ml water, and boiled for 30 min. The two extracts were combined and filtered by three-tiered gauze, and then centrifuged at 3000 rpm for 10 min. The supernatants were regarded as the sample of co-decoction, then filtered through a 0.22 μm PTFE syringe filter, and an aliquot (1 ml) of each filtrate was subjected to UPLC/Q-TOFMS analysis.

For mixed decoction, sliced samples of WT(approximately 10 g) was accurately weighed and put into flask containing 100 ml distilled water, and boiled for 60 min. This procedure was then repeated by adding 50 ml water, and boiled for 30 min. The two extracts were combined and filtered by three-tiered gauze, and then centrifuged at 3000 rpm for 10 min. The supernatants were filtered through a 0.22 μm PTFE syringe filter.Using this extraction protocol, the decoction of BJ(approximately 10 g) was prepared. The filtrate of the two decoction were mixed (1:1, v/v) and were regarded as the sample of mixed decoction, then an aliquot (1 ml) of each sample was subjected to UPLC-QTOF/MS analysis.

Establishment of in-house library and peak assignment

By searching databases, such as PubMed of the U.S. National Library Medicine and the National Institutes of Health, Chinese National Knowledge Infrastructure (CNKI) of Tsinghua University and Traditional Chinese Medicines: Molecular Structures,Natural Sources and Applications, all components reported in the literatures of Aconitum and Bletilla species were respectively summarized in Microsoft Office Excel tables to establish the in-house library,which includes the name, molecular formulae, relative molecular weight, chemical structures and references of each published known compound. The "Find" function of Microsoft Office Excel was used to match the empirical molecular formulae with those of published known compounds in the library.

Multivariate statistical analysis

The UPLC-QTOF/MS data of all determined samples were analyzed by MarkerLynx within MassLynx software (version 4.1, Waters Corporation, Milford, USA)to reveal any possible changed components in decoctions of the combination of WT and BJ. For data collection, the parameters were set as follows: retention time ranging from 3 to 17 min, mass ranging from 100 to 900 Da,retention time tolerance at 0.01 min, mass tolerance at 0.01 Da. For peak integration, peak width at 5% of the height was 1 s, peak-to-peak baseline noise was 0.1, peak intensity threshold was 10. No specific mass or adduct was excluded.

The resulting three-dimensional data comprising of peak number (tR–m/s pair), sample name and ion intensity were analyzed by supervised orthogonal partial least squared discriminant analysis (OPLS–DA) using the MarkerLynx software.

RESULTS AND DISCUSSION

Chromatographic conditions and QTOF/MS method development

In our preliminary test, two columns, ACQUITY BEH C18(100 mm×2.1 mm, 1.7 μm) and ACQUITY HSS T3 (100 mm×2.1 mm, 1.8 μm), were tested. It was found that the ACQUITY BEH C18column has better resolution of major components, more peak capacity and stronger retention ability, thus ACQUITY BEH C18(100 mm×2.1 mm, 1.7 μm) column was chosen for this study. The sensitivities of the components in decoction samples were found to be higher in positive ion mode.The representative base peak ion (BPI) chromatograms of mixed decoction and co-decoction samples were illustrated in Fig. 1. From Fig. 1 it was found that the peak no. from 13 to 19 were increased notably during codecocting of WT-BJ incompatible pair.

Fig. 1. Representative BPI chromatograms of(A) mixed decoction and (B) co-decoction of WT-BJ incompatible pair.

Multivariate statistical analysis and potential chemical markers exploring

To compare the difference between mixed decoction and co-decoction of WT-BJ incompatible pair, supervised orthogonal partial squared discriminant analysis (OPLS-DA) were performed. After Pareto scaling with mean centering, the data were displayed as a scores plot in Fig. 2. The scores plot shows that the decoction samples clearly clustered into two groups, indicating that the co-decocting procedures caused changes in the composition and/or content of components in the resulting decoctions.

Fig. 2. OPLS-DA/Scores plot of mixed decoction and co-decoction of WT-BJ incompatible pair obtained using Pareto scaling with mean centering.

To find the potential chemical markers for the discrimination between mixed decoction and co-decoction of WT-BJ incompatible pair, the extended statistical analysis was performed to generate S-plot in Fig. 3. In the S-plot, each point represents an ion tR-m/z pair, the X-axis represents variable contribution, where the farther distance the ion tR-m/z pair points from zero, the more the ion contributes to the difference between two groups; the Y-axis represents variable confidence, where the farther distance the ion tR-m/z pair points from zero, the higher the confidence level of the ion to the difference between two groups. So, the tR-m/z pair points at the two ends of "S" represent characteristic markers with the most confidence to each group.

The first nine ions (a-i) at the top right corner of"S" are the ions of co-decoction contributing most to difference between mixed decoction and co-decoction of WT-BJ incompatible pair. The ion intensity trends of these ions in analyzed samples were shown in Fig. 4(a-i). It was found that ion a (tR15.2 min, m/z 616.3106),b (tR14.27 min, m/z 632.3076); c (tR15.97 min,m/z 630.3267); d (tR15.17 min, m/z 646.3214);e (tR13.24 min, m/z 648.3014); f (tR6.7 min, m/z 436.2338); g (tR12.28 min, m/z 572.2879); h (tR14.45 min,m/z 662.3184); i (tR15.43 min, m/z 600.3174) were detected with higher intensity in all co-decoction samples, but were undetectable and/or lower in all mixed decoction samples.

Fig. 3. OPLS-DA/S-Plot of mixed decoction and co-decoction of WT-BJ incompatible pair obtained using Pareto scaling with mean centering.

Similarly, the first seven ions (j-p) at the bottom left corner of "S" are the ions of mixed decoction contributing most to difference between mixed decoction and co-decoction of WT-BJ incompatible pair. The ion intensity trends of these ions in analyzed samples were shown in Fig. 4 (j-p). It was found that ion j (tR5.29 min,m/z 486.2710); k (tR7.07 min, m/z 438.2844); l (tR7.93 min,m/z 422.2880); m (tR9.52 min, m/z 464.2999);n (tR12.63 min, m/z 574.2999); o (tR6.72 min, m/z 454.2797); p (tR5.75 min, m/z 408.2727) were detected with higher intensity in all mixed decoction samples, but were lower in all co-decoction samples.

Identity assignment and confirmation of chemical markers

The potential chemical markers (a-p) in two types of decoctions were identified by comparison the retention times and mass spectra of reference compounds and/or tentatively assigned by matching the empirical molecular formula with that of the published compounds of Aconitum or Bletilla species[8,9,10]. Mesaconitine(peak 14), aconitine (peak 16) and hypaconitine(peak 17) were confirmed by comparing with the reference compounds. The details of the identified components were summarized in Table 1, including the retention time (tR), formulae and variable importance in the projection (VIP) value of the alkaloids, and the mass accuracy for all assigned components was less than 10 ppm.

Fig. 4. Selected ion intensity trend plots. a (tR 15.2 min, m/z 616.3106); b (tR 14.27 min, m/z 632.3076);c (tR 15.97 min, m/z 630.3267); d (tR 15.17 min, m/z 646.3214); e (tR 13.24 min, m/z 648.3014); f (tR 6.7 min, m/z 436.2338); g (tR 12.28 min, m/z 572.2879); h (tR 14.45 min, m/z 662.3184); i (tR 15.43 min, m/z 600.3174);j (tR 5.29 min, m/z 486.2710); k (tR 7.07 min, m/z 438.2844); l (tR 7.93 min, m/z 422.2880); m (tR 9.52 min, m/z 464.2999); n (tR 12.63 min, m/z 574.2999); o (tR 6.72 min, m/z 454.2797); p (tR 5.75 min, m/z 408.2727).

Fig. 5. Representative mass spectra of most changed toxic components during co-decocting of WT-BJ incompatible pair. (A, B) hypaconitine; (C, D) mesaconitine; (E, F) aconitine.

Fig. 6. Conversion among aconitine, benzoylaconine and aconine.

Fig. 5 showed the representative mass spectra of most changed toxic components. In the mass spectra, the compounds with m/z 616.3106 (tR15.2), m/z 632.3076(tR14.27) and m/z 646.3214 (tR15.17) corresponded to HA, MA and AC, respectively, which were confirmed by tRand comparison with the MS/MS data obtained from the analysis of standards. The loss of acetic acid is the dominant fragmentation. The fragment ions of m/z 556.2850, m/z 572.2846 and m/z 586.3016 were found respectively, which indicated the neutral losses of acetic acid (60 Da.) from deprotonated molecular ions[M+H]+of HA, MA and AC, respectively, in the ESI MS/MS analysis.

Chemical difference analysis.

As shown in Table 1, the most changed alkaloids components derived mainly from Aconitum species. Among the 19 alkaloids, seven were diester-diterpenoid alkaloids (DDAs) including hypaconitine, mesaconitine, deoxyaconitine, aconitine,10-OH-mesaconitine, 10-OH-aconitine and deoxyhypaconitine, and five were monoester-diterpenoid alkaloids (MDAs) including mesaconine, cammaconine,pengshenine A, dehydrated benzoylmesaconine,benzoylhypaconine, while the other four were aminediterpenoid alkaloids (ADAs) including fuziline, neoline,talatizamine, acetyltalatizamine. It was reported that the LD50 value of intravenous injection of aconitine,mesaconitine and hypaconitine in mice was 0.12, 0.10 and 0.47 mg/kg, respectively, that of benzoylaconine,benzoylmesaconine and benzoylhypaconine was 23,21 and 23 mg/kg, respectively, and that of aconine was 120 mg/kg, indicating the toxicity of the three types of aconitum alkaloids in descending order[11]. As shown in Fig. 6, a carboxyl group of DDAs was taken off and converted to MDAs, while a phenyl group of MDAs was taken off and transformed to ADAs after decocting. It is interesting to find that DDAs were the most increased components in WT-BJ co-decoction, indicating that seven DDAs were the major toxic components changed during the co-decocting. Meanwhile most of these toxic alkaloids converted to the monoester and amine form and greatly reduced toxicity in the mix decoction which without decocting together. The results suggested that normal hydrolysis of three types aconitum alkaloids can be reversed in the process of co-decocting. Based on the fact that significant chemical difference exist between co-decoction and mix decoction, whether these chemical difference is associated with the increased toxicity in vivo depends on the further comparative pharmacological and toxicological investigations.

CONCULSION

For the first time, an UPLC-QTOF/MS based chemical profiling method was proposed and validated to rapidly evaluate the chemical diversity after co-decocting of the combination of WT-BJ incompatible pair, and to reveal possible changed toxic components of mixed decoction and co-decoction during co-decocting.The advantages of this newly proposed approach over the conventional chromatographic fingerprints are that the present method can rapidly and globally compare mixed decoction and co-decoction, and with the help of multivariate statistical analysis, to reveal the possible changed toxic components, the identities of which can be extensively determined with online MS information.The most changed toxic components hypaconitine,mesaconitine, deoxyaconitine, aconitine, 10-OH-mesaconitine, 10-OH-aconitine and deoxyhypaconitine of the combination of WT-BJ incompatible pair during co-decocting was revealed using this newly established method suggested that this method could be used to practically reveal the possible toxic components changed/increased of the herbal combination taboos, e.g. the Eighteen Incompatible Medications (Shi Ba Fan), in TCM.

Table 1. The most changed components identified during co-decocting of WT-BJ incompatible pair.

This work is supported by the National Basic Research Program of China ("973 Program")(No2011CB505304) and the Youth Scientific Research Project of Anhui Academy of Medical Science(YKY2018003).

REFERENCE

1. Chinese Pharmacopoeia Commission. Pharmacopoeia of the People's Republic of China [M]. Beijing: China Medical Science Press, 2015, 39-40.

2. Judith S, Ming Z, Sonja P, et al. Aconitum in traditional Chinese medicine-A valuable drug or an unpredictable risk?Journal of Ethnopharmacology. 2009, 126:18-30.

3. Desai HK, Hart BP, Caldwell RW, et al. Certain norditerpenoid alkaloids and their cardiovascular action. Journal of Natural Products. 1998, 61:743-748.

4. Wada K, Nihira M, Hayakawa H, et al. Effects of longterm administrations of aconitine on electrocardiogram and tissue concentrations of aconitine and its metabolites in mice.Forensic Science International. 2005,148:21-29.

5. Tong P, Wu C, Wang X, et al. Development and assessment of a complete-detoxication strategy for Fuzi (lateral root of Aconitum carmichaeli) and its application in rheumatoid arthritis therapy, Journal of Ethnopharmacology. 2013,146:562-571.

6. Churchwell MI, Twaddle NC, Meeker LR, et al.Improving LC–MS sensitivity through increases in chromatographic performance: Comparisons of UPLC-ES/MS/MS to HPLC–ES/MS/MS, Journal of Chromatography B. 2005, 825:134-143.

7. Yan G, Sun H, Sun W, et al. Rapid and global detection and characterization of aconitum alkaloids in Yin Chen Si Ni Tang,a traditional Chinese medical formula, by ultra performance liquid chromatography–high resolution mass spectrometry and automated data analysis. Journal of Pharmaceutical and Biomedical Analysis. 2010, 53:421-431.

8. Yue H, Pi ZF, Song FR, et al. Studies on the aconitine-type alkaloids in the roots of Aconitum Carmichaeli Debx by HPLC/ESIMS/MSn. Talanta, 2009, 77:1800-1807.

9. Zheng XT, Shi PY, Cheng YY, et al. Rapid analysis of a Chinese herbal prescription by liquid chromatography-time-offlight tandem mass spectrometry. Journal of Chromatography A. 2008, 1206: 140-146.

10. Ma XJ, Cui BS, Han SW, et al. Chemical constituents from tuber of Bletilla striata. Zhongguo Zhong Yao Za Zhi, 2017,42(8):1578-1584.

11. Liu M, Cao Y, Lv D, et al. Effect of processing on the alkaloids in Aconitum tubers by HPLC-TOF/MS. Journal of

12. Pharmaceutical Analysis. 2017, 7(3):170-175.