Serum Pharmacochemistry Analysis Using UPLC-Q-TOF/MS after Oral Administration to Rats of Shenfu Decoction

The purpose of this study was to study the serum pharmacochemistry of SFD as well as the material basis through analyzing the constituents absorbed in blood. The SFD was orally administrated to Wistar rats at 20 g·kg−1, and Ultra Performance Liquid Chromatography (UPLC) fingerprints of SFD were created. Serum samples were collected for analysis, and further data processing used MarkerLynx XS software. 19 ginsenosides and 16 alkaloids were detected in SFD. The absorption of alkaloids (mainly monoester diterpenoid alkaloids) increased when Aconitum carmichaeli Debx. was combined with Panax ginseng, while the ginsenosides remained stable. Diester diterpenoid alkaloids were not present in the serum samples. A suitable serum pharmacochemistry method was successfully established to study pharmacological effects and potential improvements in formulation. This may also be useful for toxicity reduction. We suspect that the increased absorption of the monoester diterpenoid alkaloids from the mixture of Panax and Radix, compared to the Panax only extract, may be the reason for the combination of the two herbs in popular medicine formulas in China.


Introduction
Shenfu Decoction (SFD) consisting of an equal ratio of ginseng root (radix ginseng, Renshen) and aconite root (Radix Aconiti Lateralis Preparata, Fuzi) is an example of a classic Chinese traditional herb-couple formulation, where two herbs are prescribed together to decrease toxicity and/or increase efficacy. For example, inclusion of Glycyrrhiza uralensis can prevent or destroy the toxicity of treated Fuzi [1]. The main components are ginsenosides and alkaloids. SFD was originally described in Ji Sheng Xu Fang (1253 in western calendar) written by Yonghe in the Song Dynasty. In many formulae of Traditional Chinese Medicine (TCM), Renshen is frequently prescribed in combination with other herbs to decrease toxicity and increase efficacy. Ginsenosides are the main bioactive constituents in the famous Chinese herb Renshen. They include protopanaxadiol, protopanaxatriol, octotillol, and oleanolic acid. Aconitines are the main constituents in Aconitum carmichaelii Debx. Aconitines include monoester diterpenoid alkaloids (MDAs) and diester diterpenoid alkaloids (DDAs) and have differences in esterification. Their chemical structures are shown in Figure 1. Due to the high toxicity of DDAs, Fuzi is combined with Renshen to decrease its toxicity.
SFD is used to treat cardiovascular diseases such as circulatory collapse, shock, thoracic obstruction, and acute thoracic pain. It has a therapeutic effect on heart failure and ischemia-reperfusion injury [2]. One study showed that blockage of the sodium channels in cardiac myocytes may be one of the important molecular mechanisms of its cardiac effect [3]. There are no detailed studies of its mechanism of action, and the bioactive compounds that account for its therapeutic effects remain unclear.
While they have the so-called active ingredients, there are no empirical data to prove its effectiveness. Oral administration is one of the primary modes for TCM. The bioactive compounds are absorbed in the blood and transferred to the  target. Thus, while there are multiple components in herbs, only those that are absorbed into circulation are effective [4]. Pharmacokinetics remains unresolved in the use of Chinese herbs. In our previous research, we focused on the influence of DDAs, drug metabolism enzymes, and transporters after the combination of Renshen and Fuzi [5][6][7]. Our previous study in vitro showed that there were significant differences between codecoction and mixed decoction of Renshen and Fuzi [5]. The content of toxic alkaloids was higher in mixed decoction than codecoction which was probably the attenuation basis. To confirm whether they behaved similarly in vivo is one of our goals. Here, we developed a rapid and sensitive Ultra Performance Liquid Chromatography/Quadrupole Time-of-Flight Mass Spectrometry Evidence-Based Complementary and Alternative Medicine 3 (UPLC/Q-TOF-MS) method to study the pharmacokinetics of SFD constituents.

Preparation of Decoction and for LC-MS Analysis.
Renshen (100 g) was soaked in 800 mL distilled water for 30 min and decocted for 1 h and then filtered. The residue was decocted in 500 mL distilled water for 1 h and filtered. The filtrate was then combined and labeled Renshen decoction. Fuzi (100 g) was decocted in 800 mL distilled water for 10 min and then filtered. The residue was decocted in 500 mL distilled water for 30 min and filtered and labeled Fuzi decoction. Then two different kinds of decoctions were prepared, namely, Renshen-Fuzi codecoction where the two herbs mixed together were extracted with water and Renshen-Fuzi mixed decoction where the individual herbs were extracted separately with water and the extracts then mixed together. The decoctions were evaporated to 1 g crude herb per mL. The decoctions were stored at 4 ∘ C, and the decoction samples were subjected to UPLC/Q-TOF-MS analysis, and the datasets were processed with MassLynx software. . There were 8 rats in each group. The rats were kept in an animal room with controlled environment (temperature: 22 ± 2 ∘ C, relative humidity: 55 ± 5%, and 12 h light-dark cycle) for 3 days before the experiment. All animals were free to access distilled water and standard food.

Animals Handling and Serum Samples Preparation.
Every group except the blank group received an oral administration of 20 g⋅kg −1 decoction for 3 days. A distilled water vehicle control was given to the blank group. Blood samples (500 L) were collected from the retroorbital sinus 1 h after oral administration on the 3rd day and centrifuged at 3000 rpm for 10 min. Then, 800 L methanol was added to the 200 L serum samples, vortexed, and then centrifuged at 13000 rpm for 10 min. The supernatant solution was transferred to another tube and dried with nitrogen gas. The residue was stored in 50% acetonitrile (200 L) and frozen at −80 ∘ C until analysis.  2.5. Data Analysis. All data were processed by MassLynx V4.1 software (Waters). The chromatographic peaks were integrated, aligned, and combined with accurate mass to charge ratios. A reference retention time was found for each expected compound. Using our previous work, the compounds in the decoction were identified [8][9][10]. The data were further processed by MarkerLynx XS software (Waters). The exported data list, partial least-squares-discriminant analysis (PLS-DA) as well as principal component analysis (PCA), was used to analyze the differences between the groups.

Chemical Analysis of Codecoction.
The ginsenosides and alkaloids in SFD were identified using UPLC combined with a TOF-MS detector. The total ion chromatogram of SFD in positive and negative ion modes is shown in Figure 2 and was processed with MassLynx V4.1 (Waters, USA). Comparing the retention time and mass data with reference compounds identified the target compounds. Both the MDAs and DDAs of alkaloids and 20(s)-protopanaxadiol type, 20(s)-protopanaxatriol type, and oleanolic acid saponin-type ginsenosides were detected in SFD. The chemical compositions are shown in Tables 1 and 2.

Differences in the Five Groups.
All five groups were processed by PLS-DA and were used to highlight variation among the five groups (see Figure 3). Exported PCA plots and loading plots showed that the five groups formed five clusters. This indicated that the components in the serum were different.

Alkaloids Difference between B-Group, F-Group, M-
Group, and C-Group. Comparing F-Group with B-Group, 14 kinds of alkaloids including DDAs, MDAs, and general alkaloids in aconitum plants were detected. Most of them were trace amount, which indicated the alkaloids poor absorption in vivo. The main reason was high efflux ratio reduced by P-glycoprotein [11][12][13]. These 14 kinds of alkaloids might be the chemical constituents involved in therapeutic efficacy of SFD. Some researchers thought that Fuzi's therapeutic activity (anti-inflammatory, analgesic, and cardiotonic activity) seems to have relevance to the presence of toxic alkaloids [14,15] but the other researchers thought that Fuzi's toxicity compounds (mainly aconitine, mesaconitine, and hypaconitine) are not essential for its efficacy [16] (see Figure 4).
After detecting these 14 alkaloids, we further processed the data. Comparing C-Group with F-Group, the absorption of MDAs (benzoylaconine, mesaconine, 10-OHbenzoylmesaconine, and dehydrated benzoylmesaconine) decreased and the general alkaloids in aconitum plants (cammaconine, neoline, and talatizamine) increased. Most alkaloids (cammaconine, carmichaeline, talatizamine, acetyltalatizamine, and dehydrated benzoylhypaconine) absorption in M-Group was increased rather than F-Group. The increased or decreased absorption of alkaloids may contribute to the attenuation and synergistic effects. The decreased absorption of toxic alkaloids leads to attenuation and increased nontoxic alkaloids contribute to synergistic effects mainly because of compatibility of Renshen. The phenomenon was found in ancient year by the Chinese and we make it clear by analyzing the chemical compounds (see Figures 5 and 6).
The M-Group data showed that all alkaloids were absorbed more than the C-Group. The relative intensity of the ion indicated that the toxic chemicals (mesaconine, 10-OH-benzoylmesaconine, dehydrated benzoylmesaconine, and dehydrated benzoylhypaconine) were less absorbed than most general alkaloids in Aconitum plants (cammaconine, carmichaeline, fuziline, talatizamine, acetyltalatizamine, and neoline). The content of toxic alkaloids was higher in mixed decoction than codecoction which was probably the attenuation basis in vitro, and the results in vivo showed a similar consequence which indicated that the Renshen attenuated the toxicity of Fuzi. And ester exchange or degradation reactions occurring during the processing of codecoction with Renshen were considered to be the key factor of attenuation [5] (see Figure 7).

Ginsenosides Differences in the Serum in Negative Mode.
The ginsenosides and alkaloids were detected in negative and positive mode, respectively. In our study, the codecoction and mixed decoction had small differences in the negative mode. This indicated that the ginsenosides behaved similarly, and the results presented the same tendency with an unpublished paper for the negative result in vitro. The same processing method used in positive mode was also applied to negative mode to analyze the ginsenoside differences. The PCA and PLS-DA/S-Plot showed no obvious distinction between the C-Group and the M-Group. However, 13 kinds of ginsenosides (GRg 1 , GRe, GRd, GRo, GRc, GRb 2 , GRb 3 , GRb 1 , GRa 2 , GRa 1 , MalonylGRc, MalonylGRb 1 , and MalonylGRb 2 ) were detected in the serum of the Panax ginseng group. The extracted ion chromatogram and mass spectrogram of the ginsenosides are shown in Figures 8, 9, and 10.

Discussion
Researchers have reported quantitative analysis of aconitum alkaloids of SFD in vitro as well as the pharmacokinetic behavior of Fuzi and drug-drug interaction mechanisms in herb pair decoctions [17][18][19]. Relatively few studies have been conducted on the material basis of SFD in vivo, and most researchers focused on Shenfu injection which is a kind of processed Shenfu formulation [20,21]. The material basis of toxicity reduction and pharmacological effect improvement in vivo remains unclear. In previous studies, the attenuation and synergistic effects of SFD including decoction time and herb ratio were studied in vitro. However, aconitum alkaloids serum levels were difficult to detect after oral administration, and only trace level was seen for its low bioavailability [18,22]. Only Shenfu injected powder showed detectable alkaloids in serum [23]. Multicomponents and multitarget herbs resulted in complex systems with difficult detection. Recently, UPLC coupled with TOF-MS has become a vitally important tool in studying Chinese herbs. It is fast with good selectivity, high resolution, and accurate mass measurements. This makes it attractive for the analysis of complex biological sample [24]. Shenfu formulation was a classical Chinese medicine for its obvious therapy efficiency on heart fail [25], but Fuzi in the formulation was a famous toxicity herb for its severe arrhythmia and neurological, cardiovascular, and gastrointestinal symptoms [26]. The herb pair was used for synergism and attenuation and the decoction method was thought to be the main reason for its chemical difference, but whether it behaves similarly in vivo was still a mystery, and here we approved an evidence to evaluate the Shenfu formulation in vivo. The concentration of hypaconitine and deoxyaconitine decreased, while benzoylmesaconine, benzoylhypaconine, and dehydrated benzoylmesaconine increased in the codecoction samples. The toxicity increased in single decocted or mixed decocted samples [5]. The codecoction inhibited toxic compounds, and the inhibited alkaloids dissolution resulted in a lower alkaloids content in the C-Group than the M-Group. They behaved similarly in vivo and in vitro.
Here, Renshen showed evidence of increasing the absorption of alkaloids. These alkaloids were measured in a relative concentration range and found to be in accordance with the Evidence-Based Complementary and Alternative Medicine tradition of TCM. We suspect that toxicity reduction is due to decoction in vitro; the pharmacological effect improvement is due to increased absorption and metabolism in vivo.
The pharmacological effect in vivo was affected by absorption and metabolism. In previous study, alkaloids were metabolized by Cytochrome P450 (CYP450) 3A [27]. A study on Shenfu injection indicated that it inhibited the enzyme activity of CYP3A at the mRNA level [6]. Mean Residence Time (MRT) increased when the enzyme activity was inhibited. Its absorption increase contributed to an increase in the Area under the Curve (AUC). The increased MRT and AUC indicated an increase in pharmacological effects.
Efflux transporters in the intestine play an important role in absorption. They also protect the body from toxin damage. The ATP-binding cassette transporters include Pglycoprotein (P-gp), multidrug resistance-associated protein isoform 2 (MRP2), and breast cancer resistance protein (BCRP). They modulate the absorption, distribution, metabolism, and excretion of medicine. These proteins are highly expressed in the apical membranes of intestinal epithelial, hepatic, and renal tubular cells. Transport studies show that the efflux ratios of aconitine, mesaconitine, and hypaconitine were significantly elevated due to P-gp and BCRP [11]. Panax ginseng and ginsenosides like ginsenoside F1, ginsenoside Re, and ginsenoside Rb 2 induced the function of P-gp [28]. An important clue is seen in that Fuzi's toxicity compounds (aconitine, mesaconitine, and hypaconitine) are not essential for its efficacy [16]. While aconitine, mesaconitine, and hypaconitine were not detected in serum, it could be because they were below the detection limit.
The difference in molecular structure between DDAs and MDAs is an acetyl group. This caused the different efflux ratios. The lost acetyl group decreased the toxicity of the DDAs and the efflux ratio [29]. In another TCM herb pair Fuzi-Ganjiang formula, Ganjiang (Rhizoma Zingiberis, derived from the dry rhizome of Zingiber officinale Rosc.) was combined with Fuzi to decrease its toxicity and improve pharmacological effects. The authors showed that Ganjiang enhanced the absorption of MDAs and promoted the elimination of DDAs [30]. We could hypothesize that Renshen played the same role in these herb pairs. Our findings showed that nontoxic alkaloids in the M-Group and C-Group absorption increased more than F-Group. The absorption of alkaloids in mixtures or extractions was better than that of pure compounds seen in other experiments [12].
Like many other TCMs, SFD is always taken orally. Ginsenosides showed poor absorption and low bioavailability [31]. Major factors that limited the intestinal absorption of ginsenosides were poor membrane permeability and active biliary excretion. This limited systemic exposure to most ginsenosides and their deglycosylated metabolites [32]. The poor absorption and bioavailability of ginsenoside make them difficult to accurately measure in serum. Thus, we successfully developed a new method to measure ginsenoside previously. The different clearance rate of ginsenosides is considered to be the mechanism of Shenfu injection [10].
In the past decades, researchers have studied traditional Chinese formulas sufficiently, but the material basis and mechanism are still unclear with most data focusing on pharmacokinetics. Most studies are multicomponent herbs and with multiple targets in vivo. In addition to active compounds, there are also inactive and even toxic compounds in herbs. The material basis, pharmacology, component compatibility, and physiological disposition of TCM remain unclear. Understanding the attenuation and synergistic effects on absorption is an important goal of this study. Here, we studied the material basis of attenuation and synergistic effects by measuring the absorbed ingredients.   Figure 9: Extracted ion chromatogram of ginsenoside in negative ion mode.   the blood collection. Zeng-chun Ma and Yue Gao contributed to critical review of the paper.