Simultaneous Evaluation of the Influence of Panax ginseng on the Pharmacokinetics of Three Diester Alkaloids after Oral Administration of Aconiti Lateralis Radix in Rats Using UHPLC/QQQ-MS/MS

Objectives To investigate whether Panax ginseng (P. ginseng) could affect the metabolism of Diester Alkaloids (DAs) derived from Aconiti Lateralis Radix in vivo. Methods and Results 24 male Sprague-Dawley rats were randomized for 7-day treatment with P. ginseng (low, middle, and high), or vehicle. Aconiti Lateralis Radix was administered orally to each group on the 8th day. Plasma samples were collected, and Xevo TQ-S was used to detect the concentration of aconitine, mesaconitine, and hypaconitine in plasma. We describe a fast and reproducible method to detect the concentration of aconitine, mesaconitine, and hypaconitine in plasma. Compared to the control group, the AUC(0-t) of three DAs increased in both the middle and high dosing groups. The Vz/F of three DAs in these groups as well as the CLz/F of aconitine in all P. ginseng groups and the CLz/F of mesaconitine and hypaconitine in P. ginseng middle and high groups were decreased compared to the control group. Conclusion Orally administrated P. ginseng potentially inhibits the metabolism of DAs from Aconiti Lateralis Radix in rats.


Introduction
Panax ginseng C. A. Mey (referred as P. ginseng in this study) and Aconiti Lateralis Radix are two traditional Chinese medicines that have a long history of successful application in the clinic [1][2][3].
P. ginseng has been used for the treatment of various diseases such as liver dysfunction [4], hypertension [5], and cerebrovascular disease [6] in China, Japan, and Korea. Various ginsenosides such as Rg1, Rb1, Rc, Rd, Rg3, Rh1, and Rf have all been found to be important constituents in P. ginseng extracts [7][8][9][10]. Aconiti Lateralis Radix is traditional Chinese medicine famed for its toxicity and contains a range of diverse alkaloids such as aconitine, mesaconitine, hypaconitine, and benzoylaconitine among many others [11]. Aconitine, mesaconitine, and hypaconitine, which are classed as DAs, have been shown to be the major toxic components in Aconiti Lateralis Radix and have been known to exert both potent cardiac toxicity and neurotoxicity in several studies [12,13].
Routes of administration vary when a combination of P. ginseng with Aconiti Lateralis Radix is required, including the two traditional routes, by injection and by mouth [1,2]. These routes are commonly used for a range of disorders, most notable in the treatment of heart failure and rescuing hemorrhagic shock [14][15][16][17][18][19]. Although this treatment is often successful, overdose is common among patients. New measures to reduce this risk are now in place including more strictly controlled usage to limit cardiac toxicity and neurotoxicity. To date, the mode of action of this toxicity is unknown; however, several hypotheses have been developed. It was found that toxicity was reduced in doses that had increased levels of monoester alkaloids, while toxicity was directly correlated to an increase in DAs [20,21].
In our previous work, we reasoned that P. ginseng could have a direct effect on the metabolism of these 2 Evidence-Based Complementary and Alternative Medicine DAs (aconitine, mesaconitine, and hypaconitine) in vitro. Furthermore, we speculated that metabolism stemmed from oxidative transformations mediated by CYP3A4, which was confirmed to be the primary sub-cytochrome P450s (CYPs) for phase 1 metabolism of DAs. And researchers also found that biological activities and pharmacokinetics difference of three representative alkaloids: aconitine, benzoylaconine, and aconine were observed after oral administration in rats [22]. Though aconitine, mesaconitine, and hypaconitine were all belonging to DAs, the structural involvement of sub-CYPs in the metabolism and toxicity of them [23] was not entirely consistent. Zhang Q et al. investigated the pharmacokinetic difference of alkaloids in rats after oral administration of decoction of Aconiti Lateralis Radix. They found that different preparative methods resulted in significant difference on exposure and pharmacokinetic characteristics of alkaloids [24]. Therefore, we wanted to further evaluate whether P. ginseng directly influences the metabolism of DAs in vivo, which to the best of our knowledge has yet to be studied.
The present study therefore aims to investigate the pharmacokinetic difference between aconitine, mesaconitine, and hypaconitine after exposure to P. ginseng in rats.

The Preparation of P. ginseng and Aconiti Lateralis Radix
Extract. P. ginseng (300 g) was soaked in 2.4 L doubly distilled water and boiled for 1 hour. The filtrate was collected and previous step was repeated. The combined filtrates were concentrated by rotary evaporation to afford extracts of 0.5 g/ml (according to the concentration of the crude drugs). The same protocol was used to obtain the extract of Aconiti Lateralis Radix. Extracts were centrifuged at 12,000g for 10 min, the supernatant transferred to a clean centrifuge tube and stored at -20 ∘ C.

The Preparation of Reference Solution.
Stock solutions were prepared by dissolving accurately weighed reference compounds in methanol at concentrations 1mg/ml individually. Then working solutions were diluted with 50% methanol respectively at concentration of 10 g/ml, and centrifuged at 12,000 g for 10 min. Then samples were filtered through a 0.22 m micropore membrane. A 5 L volume of supernatant was injected into LC-MS system for analysis.

The Qualitative
Analysis of Herb Extracts. UPLC/Q-TOF-MS (Waters, USA) was used to detect the target components of herb extracts. The extracts were centrifuged at 12,000g for 10 min, diluted to a specific concentration, and filtered through a 0.22 m Micropore membrane. A 5 l injection was used for subsequent analysis.
Electrospray ionization (ESI) and positive ion V mode were used to determine the components of Aconiti Lateralis Radix extract. Negative ion V mode was used to determine the components of P. ginseng extract. The main parameters for mass spectrum were set as follows: capillary voltage: 3.0 kv; sample cone: 40 v; source temperature: 100 ∘ C; desolvation temperature: 350 ∘ C; cone gas flow: 450 L/h; desolvation gas flow: 900 L/h; injection volume: 5 l. The composition identification was done by combining reference substance and molecular weight deviation.

Animals and Treatments
. 24 male Sprague-Dawley (SD) rats (180-220 g) were obtained from the animal experiment center of the Academy of Military Medical Science of People's license (Beijing, China): SCXK (army)-2012-0004. The environment of raising animals was controlled (room temperature: 22 ± 2 ∘ C, humidity: 50 ± 5%, 12 h dark-light cycle). All of the animal experiments were performed in accordance with the guidelines of European community, approved by Institution of Animal Care and Use Committee, Academy of Military Medical Science (IACUC-AMMS). All animals were kept in these environments for 5 days before experiments.
Rats were randomly divided into 4 groups: control group, P. ginseng (low, middle, and high) group. 7 days of oral administration of P. ginseng extract was administered at a dosage of 0.1, 1, 10 g/kg for P. ginseng group (low, middle, and high), respectively, once daily. Distilled water was administered at a dosage of 2 ml/kg for control group, simultaneously. On the 8th day, Aconiti Lateralis Radix extracts (1 g/kg) were orally administered to all groups. 0.5 ml plasma samples were collected in 1.5 ml centrifuge tubes (containing heparin sodium) at 10, 30, 60, 90, 120, 240, 480, 720, 1440, and 2880 min. All plasma samples were centrifuged for 15 min at 4000g. The supernatant was separated to another 1.5 ml centrifuge tubes and stored at -20 ∘ C. precipitation (50% acetonitrile plus 50% methanol) methods were compared to determine the best method to prepare the plasma sample. Liquid-liquid extraction method was chosen, due to reproducibly high recovery rates. The details of the extraction are as follows: 10 L carbamazepine (IS) was added to 90 L plasma, spiked with 400 L ethyl acetate and the mixture vortexed for 15 seconds. After centrifugation at 12,000g for 10 min, the supernatant was transferred to another 1.5 mL centrifuge tube and dried under a stream of nitrogen at 37 ∘ C. The residue was redissolved in 100 L 50% acetonitrile and centrifuged at 12,000g for 10 min and filtered through a 0.22 m Micropore membrane. A 5 L volume of supernatant was injected into LC-MS system for analysis. Plasma from rats without any processing was used to evaluate the detection methods. 0.1, 0.25, 0.5, 1, 2.5, 5, 10 ng/ml were chosen to calculate standards curve. 0.1, 0.4, 4, 8 ng/mL were selected as the QC sample (n = 4) with repeated operations used to calculate accuracy and precision of the detection method. The data was acquired after keeping the samples at either +4 ∘ C 24 hours or room temperature (about 25 ∘ C) for 24 hours as well as using freeze-thaw methods (three times) to evaluate stability. Recovery of the developed methods was calculated by the following formula: recovery = measured value/ theoretical value × 100%.
Electrospray ionization (ESI) and positive ion V mode were used to determine the samples. The main parameters for mass spectrum were set as follows: capillary voltage: 3.5 kv; sample cone: 60 v; source offset: 50 v; source temperature: 150 ∘ C; desolvation temperature: 350 ∘ C; cone gas flow: 150 L/h; desolvation gas flow: 650 L/h; injection volume: 5 L.

Statistical
Analysis. The data was acquired and analyzed by XEVO-QT-S (Waters). The pharmacokinetic parameters were analyzed by DAS pharmacokinetic program (Version 2.0). SAS statistical software was used to determine the difference between each group. All data was presented as mean ± SD, and P<0.05 was considered statistically significant. The figures were processed and displayed by GraphPad prism software (Version 5.01).

The Qualitative Analysis of Herb Extracts.
The main components in P. ginseng included ginsenoside Rf, Rg2, Rg3, Rb3, Rh1, Rd, Rg6, F2, notoginsenoside R2, and chikusetsusaponin IVa ( Figure 1 and Table 1). The positive ion peak chromatograms of aconitine, mesaconitine, and hypaconitine in Aconiti Lateralis Radix by UPLC-Q/TOF-MS were shown in Figure 2 and clearly show that aconitine, mesaconitine, and hypaconitine were successfully extracted from Aconiti Lateralis Radix. The assigned identity, retention time, molecular formula, theoretical mass/Da, mean measured mass/Da, and mass accuracy (1.0 × 10 −6 ) of herb extract were shown in Tables 1 and 2.  Figure 4, with these peaks not observed for the blank control sample (Figure 3). The limit of detection (LOD) and the lowest limit of quantification (LLOQ) were defined by signal-to-noise ratio method. LOD should be three times of the noise level (S/N >3). LLOQ should be ten times of the noise level (S/N >10). The LOD and LLOQ of three DAs in this experiment were 0.025 ng/ml and 0.1 ng/ml, respectively (Table 3).   concentration range 0.1-10 ng/mL are shown in Table 3 and Figure 5. Three standard curves for aconitine, mesaconitine, and hypaconitine were y = 0.3651x + 0.021, y = 0.7952x + 0.0218, and y = 0.718x + 0.0605, respectively. The correlation coefficient was 0.9997 (aconitine), 0.9987 (mesaconitine), and 0.9970 (hypaconitine).

Precision and Accuracy.
The precision and accuracy of the detection of three compounds in both intra-and interday at the linear range are shown in Table 4. Four QC concentrations (0.1, 0.4, 4, 8 ng/mL) were analyzed for each compound. The average precision and accuracy of intraday samples ranged from 2.20% to 7.06% and from 91.18 ± 5.49% to 111.64 ± 2.45%, respectively. The average precision and accuracy of interday ranged from 1.98% to 15.88% and from 83.09 ± 6.75% to 105.48 ± 3.64%, respectively, which proved that the methods had a good accuracy and precision.

Stability and Recovery.
The stability data of detected methods were shown in Table 5. The stability of three compounds in 4 ∘ C and room temperature (about 25 ∘ C) within 24 hours are shown in Table 5(a), with the accuracies ranging from 97.02 ± 9.18% to 101.57 ± 1.29% and from 93.68 ± 2.77% to 100.64 ± 3.60%, respectively. The stability of freezethawing the samples is shown in Table 5(b). The RSDs of the accuracy were less than 10.10%. The data displayed that the three compounds had good overall stability.
The recovery of the developed methods was shown in Table 6. The extraction rate for each component at three QC concentrations ranged from 97.74% to 105.02%, and the RSDs ranged from 2.17% to 9.67%.

The Effect of P. ginseng on the Metabolism of Aconitine, Mesaconitine, and Hypaconitine in Rats.
The blood concentration-time curves of aconitine, mesaconitine, and hypaconitine are shown in Figure 6. The pharmacokinetic parameters of each compound in the four groups were shown in Tables 7-9. The AUC (0-t) of the three DAs were increased in P. ginseng middle group (61.05 ± 12.82 g/L * h for aconitine, 82.45 ± 18.57 g/L * h for mesaconitine, and 70.71 ± 7.55 g/L * h for hypaconitine) and high group (162.98 ± 48.50 g/L * h for aconitine, 126.07 ± 44.93 g/L * h for mesaconitine, and 71.75 ± 22.2 g/L * h for hypaconitine) when compared to the AUC (0-t) of the control group (40.77 ± 9.71 g/L * h for aconitine, 51.24 ± 21.48 g/L * h for mesaconitine, and 48.12 ± 24.59 g/L * h for hypaconitine).          The C max of aconitine (11.53 ± 3.49 g/L) and mesaconitine (8.94 ± 2.33 g/L) increased in P. ginseng high groups compared to control (3.88 ± 0.92 g/L for aconitine, 5.48 ± 2.61 g/L for mesaconitine), whereas the C max of hypaconitine did not change in all groups. The t 1/2 of three alkaloids of all groups had no significant difference.
Aconiti Lateralis Radix is the root of the Aconite and works to excite central nervous system as well as improve heart function in cases of heart failure [12,31]. Reports have suggested that Shenfu injection could treat advanced Non-Small-Cell Lung Cancer (NSCLC) [32].
It has also been found that the method of extraction used can produce different concentrations of the active DAs in vitro [33]. Due to the inherent toxicity of DAs, the dosages found in marketed drugs which also contain Aconiti Lateralis Radix are strictly limited according to the 2015 edition of Chinese Pharmacopoeia. It has also been observed that a small change in the concentration of DAs can have a significant influence on the toxicity and efficiency of Aconiti Lateralis Radix [34]. CYP3A4 is the main metabolizing enzymes of CYP450 in human and is responsible for the metabolism of almost 60% endogenous and exogenous substances [35,   36]. In previous research, many ginsenosides, the active component of P. ginseng, downregulated CYP3A4 expression through inactivation of human pregnant X receptor (PXR) [4,[37][38][39]. However, the effect of P. ginseng on the metabolism of DAs in vivo has not yet been reported.
Herein, we disclose a refined method to detect DAs in rat blood which contained aconitine, mesaconitine, and hypaconitine. The quantitative concentration range of each component was 0.1-10 ng/mL. The chromatographic behavior of each substance corresponding to signal of MS/MS allowed the simple quantitative and qualitative analysis of each component within the sample. Compared to the previous methods, this method has the advantages of time-saving (the retention time was less than 4.0 min) and low detection concentration, and the sample pretreatment process is simple [11,24,40].
In this study, we suggest that P. ginseng can inhibit the clearance of aconitine, mesaconitine, and hypaconitine in rats. This result in the accumulation of aconitine, mesaconitine, and hypaconitine within the body may be the basis of DDI between P. ginseng and Aconiti Lateralis Radix.
As the key toxic and effective components in Aconiti Lateralis Radix, the concentration of DAs is strictly controlled [41]. Pharmacokinetic changes for DAs when there is combined utilization of P. ginseng may lead to a double-sided result. Slowing down the rate of elimination can increase the efficiency of aconitine. However, metabolic inhibition from P. ginseng at DAs toxic dose may delay the detoxification, and the concentration range so far is hard to define. Consequently, it is necessary to study the pharmacokinetic changes and concentration range of aconitine, mesaconitine, and hypaconitine for the combined utilization of P. ginseng and Aconiti Lateralis Radix. In the next work, we would pay more attention to define the safety of the concentration range of DAs when treating with Aconiti Lateralis Radix. In our previous work, we suggested that P. ginseng could downregulate CYP3A4 activities through a combination method. According to the results obtained in this study, we surmise that P. ginseng may inhibit the metabolism of small quantities of aconitine, mesaconitine, and hypaconitine in Aconiti Lateralis Radix thus exerting a dramatic change in calcium ion levels which plays a vital role in excitationcontraction coupling in cardiac myocytes.
Researchers have also found that [3] SFI significantly decreased serum CK, LDH, and TNNI3 levels in myocardial ischemia/reperfusion injury (MIRI) rats, while it significantly increased the level of left ventricular systolic pressure (LVSP), left ventricular diastolic pressure (LVDP), maximal rate of the increase of left ventricular pressure (+dp/dtmax), maximal rate of the decrease of left ventricular pressure (-dp/dtmax), left ventricle ejection fraction percentage (EF), and stroke volume (SV).
Aconitine-type alkaloids associated with Aconiti Lateralis Radix efficacy have lower intake and slower elimination in CHF rats, indicating a noninterdependent relationship between its efficacy and toxicity [42]. Coupled with P. ginseng, aconitine, mesaconitine, and hypaconitine in Aconiti Lateralis Radix may account for the lower intake and slower elimination in the CHF patient; however, this requires further work to be confirmed.

Conclusion
A rapid and accurate method to detect the concentration of aconitine, mesaconitine, and hypaconitine in biological samples has been developed and successfully used. From the data obtained in this study, we suggest that P. ginseng could inhibit the metabolism of DAs in vivo. Our findings have important implications for P. ginseng therapy with medicine containing DAs in clinics.

DAs:
Diester Alkaloids P. ginseng: Panax ginseng UPLC-Q/TOF-MS: Ultra Performance Liquid Chromatography time-off-flight mass spectrometer ESI: Electro spray ionization DDI: Drug-drug interaction SD: Sprague-Dawley ACN: Acetonitrile MIRI: Myocardial ischemia/reperfusion injury NGF: Nerve growth factor LVSP: Left ventricular systolic pressure LVDP: Left ventricular diastolic pressure +dp/dtmax: Maximal rate of the increase of left ventricular pressure -dp/dtmax: Maximal rate of the decrease of left ventricular pressure EF: Left ventricle ejection fraction percentage SFI: Shenfu injection LOD: The limit of detection LLOQ: The lowest limit of quantification.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper with any people or organizations.