2.5.1 For metabolic profile

In vivo assay: The preparation method of plasma, urine, bile acid and feces sample were according to the previous reports with a minor modification (Lin et al., 2018). The intestine content freeze-dried powder collected from different time points were extracted with appropriate amount of methanol, vortexed and ultrasonicated for 30 min at low temperature. The mixture was centrifuged for 10 min at 14,000 rpm (4 °C) to obtain supernatant. Next, transferred it into a new 1.5 mL Eppendorf tube and evaporated to dryness under the stream of nitrogen. The residue was re-dissolved in 200 μL methanol. Then vortexed and centrifuged (14,000 rpm, 15 min). The supernatants were injected into the UPLC-Q/TOF-MS system for the following analysis.

In vitro phase I metabolism assay: as previously published with a minor adjustment (Xu et al., 2018), the test compounds were respectively mixed into phase I incubation system (200 μL), which contained 50 mM Tris-HCl buffer solution (pH = 7.4), 0.5 mg/ml of rat liver microsomes (RLM) or rat intestinal microsomes (RIM), 5 mM MgCl₂. After 5 min pre-incubation (37 °C), 1 mM of NAPDH solution was added and mixed. After 1 h incubation at 37 °C, the reaction was terminated by 200 μL cold acetonitrile. Then centrifuging at 13000 rpm for 15 min, the supernatant was subjected into UPLC-Q/TOF-MS system.

In vitro glucuronidation assay: as previously described (Xu et al., 2018), the incubation system contained 50 mM Tris-HCl buffer (pH = 7.4), 0.5 mg/ml of RLM or RIM, 0.88 mM MgCl₂, 22 μg/mL alamethicin, 4.4 mM saccharolactone, 3.5 mM UDPGA and test compounds. After 1 h incubation (37 °C), 200 μL cold acetonitrile was added into incubation mixture to terminate the reaction by centrifugation at 13000 rpm for 15 min, and the supernatant were subjected into UPLC-Q/TOF-MS system.

2.5.2 For quantitative study

Plasma: A 100 μL aliquot of plasma sample were mixed with 10 μL of dexamethasone (IS) solution and 300 μL methanol–acetonitrile at a ratio of 1:1 (v/v). The mixture was centrifuged for 10 min at 14,000 rpm (4 °C) to obtain supernatant. Next, transferred it into a new 1.5 mL Eppendorf tube and evaporated to dryness under the soft stream of nitrogen. The residue was reconstituted with 100 μL of methanol. Then, vortexed and centrifuged at 14,000 rpm for 15 min. The supernatants were injected into the UPLC-TQ-XS system for further analysis.

Intestine content: approximately 2 mg intestine content freeze-dried powder were accurately weighted, dissolved with 500 μL methanol and 10 μL of tolbutamide (IS), vortexed and ultrasonicated for 30 min at low temperature. The mixture was centrifuged for 10 min at 14,000 rpm (4 °C) to obtain supernatant. Next, transferred it into a new 1.5 mL Eppendorf tube and evaporated to dryness under the stream of nitrogen. The residue was re-dissolved in 200 μL of methanol containing 0.01 % formic acid (v/v). Then vortexed and centrifuged (14,000 rpm, 15 min). The supernatants were injected into the UPLC-TQ-XS system for further analysis.

2.6.1 Metabolic profile study

Thirty rats were fasted with free access to water in metabolic cages separately over night before experiment. There were divided into three groups: 10 rats were orally administered with GXB at the dose of 20.0 g/kg/day for one week; 10 rats were orally administered with GXB-dePRP at the dose of 13.2 g/kg/day for one week; 3 rats as control group were received an equivalent volume of distilled water for consecutive 7 days. The urine and feces samples for each group were collected separately during 0–12 h, 12–24 h, 24–36 h and 36–48 h on the last two days of intragastric administration. Blank urine and feces samples were collected before drug administration. After the last oral administration, the rats in each group were anesthetized by intraperitoneal injection of 10 % pentobarbital sodium. The blood samples were obtained from hepatic portal vein in heparinized tube at 0.25, 0.5, 1, 2, 4, 6 and 8 h with 7 rats. Blank plasma samples were collected in the same way. The rats (n = 3) in each group were implanted with a PE-0503 cannula into the bile duct after an abdominal incision anaesthetized with 10 % pentobarbital sodium. During the experiment, a heating lamp was used to maintain their body temperature. Bile samples were collected and recorded during 0–4 h, 4–8 h and 8–12 h. Blank bile samples were obtained in the same way.

2.6.2 Pharmacokinetic study

Plasma: twelve rats were fasted with free access to water in metabolic cages separately over night before experiment. Then the rats were divided into two groups (n = 6) and orally administered with GXB and GXB-dePRP extract as prepared at the dose of 20.0 and 13.2 g/kg for a week, respectively. Blood samples were collected from external jugular veins into heparinized eppendorf tubes at 0.25, 0.5, 1, 2, 4, 6, 8, 12, 24, 36, 48, 60, 72, 84, 96 h after last administration, and immediately centrifuged at 14,000 rpm for 15 min. The supernatant plasma samples were transferred and stored at −80 °C until analysis. WinNonlin 6.3 software was employed to calculate the pharmacokinetic parameters of analytes.

Intestine content: After one week acclimation, they were separated into two groups and orally administrated GXB and GXB-dePRP extract as prepared at the dose of 20.0 and 13.2 g/kg for a week, respectively. After last intragastric administration, immediately anaesthetization, the rats were dissected at 0.25, 0.5, 1, 2, 4, 6, 8, 12, 24, 36, 48, 60 h (n = 5 for each time-point) and carefully removed small intestine after tying the ends with tweezers. The small intestine was washed with saline solution, and dried with filter paper. Then the small intestine contents were slowly and evenly flushed by distilled water, and freeze-dried to obtain intestine content freeze-dried powder. All the samples were stored at −80 °C until further analysis. WinNonlin 6.3 software was employed to calculate the pharmacokinetic parameters of analytes.

3.1.1 Metabolic profile of GXB

Since the metabolic profile of GLXB, a GXB-associated formula, had been systematically characterized previously in our laboratory (Lin et al., 2018). Herein, the metabolic profile of GXB was described simply to avoid repetition.

Firstly, on the basis of chemical profile of GXB (Lin et al., 2021), a total of 103 prototypes were observed in dosed rat biological samples by comparing with retention times, accurate molecular ions and fragment ions, and there were respectively 8, 60, 19, 49 and 85 prototypes detected in plasma, urine, bile acid, feces and intestine content by UPLC-Q/TOF-MS. The specific mass spectrometry data of prototypes characterized in rat bio-samples were depicted in Fig. S1 and Table S1.

In addition, the strategy of “representative structure based homologous xenobiotics identification-RSBHXI” proposed in our laboratory previously (Geng et al., 2014) was applied to characterize the metabolites after oral administration of GXB. Considering that the metabolic behaviors of steroidal saponins from AMB and flavonoids from TP had been investigated comprehensively (Qin et al., 2016; Lin et al., 2018), other representative structures were further selected to develop their metabolic fates in vivo and in vitro serving as supplements. Among them, cucurbitacin B is a tetracyclic triterpenoid that was a representative structure original from TP, and a metabolic profile was developed in rat liver microsomes and intestine microsomes as well as rats after intragastric administration. The proposed metabolic pathway of cucurbitacin B was summarized in Fig. 2, and hydroxylation, dehydrogenation, hydrogenation, and glucuronidation were the major transformation reaction. Similarly, liquirtigenin and glycyrrhizic acid were two representative structure components derived from PRP, and their metabolic behaviors were shown in Figs. 3-4.

Close

Fig. 2

The proposed metabolic pathway of cucurbitacin B (P, U, B, F, I, meant metabolites detected in plasma, urine, bile, feces, intestine content, respectively. RLM, RIM meant that cucurbitacin B were incubated in the system of rat liver microsome enzyme and rat intestinal microsome enzyme, respectively; red, blue and black mark meant metabolites detected in different biosamples after dosing GXB extract, single compounds, and both two groups, respectively.) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

Export to PPT

Close

Fig. 3

The proposed metabolic pathway of liquirtigenin (P, U, B, F, I, meant metabolites detected in plasma, urine, bile, feces, intestine content, respectively. RLM, RIM meant that liquirtigenin were incubated in the system of rat liver microsome enzyme and rat intestinal microsome enzyme, respectively; red, blue and black mark meant metabolites detected in different biosamples after dosing GXB extract, single compounds, and both two groups, respectively.) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

Export to PPT

Close

Fig. 4

The proposed metabolic pathway of glycyrrhizic acid (P, U, B, F, I, meant metabolites detected in plasma, urine, bile, feces, intestine content, respectively. RLM, RIM meant that glycyrrhizic acid were incubated in the system of rat liver microsome enzyme and rat intestinal microsome enzyme, respectively; red, blue and black mark meant metabolites detected in different biosamples after dosing GXB extract, single compounds, and both two groups, respectively.) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

Export to PPT

Subsequently, by comparing the retention times, fragment behaviors, accurate molecular weights with that of single components in vivo and in vitro experiments, a total of 118 GXB-related metabolites were tentatively identified. Specifically, there were 12, 25, 25, 63 and 59 observed in dosed plasma, urine, bile acid, feces and intestine content, respectively (Table S2). The structure types of metabolites detected in intestine content were almost similar with those in feces samples, indicating that small intestine also plays an indispensable role in vivo disposal.

3.1.2 Comparison of metabolic profile

After clarified the metabolic profile of GXB, we further characterized the metabolic properties in complex system after oral administration of GXB-dePRP, so as to make a comparison with that of GXB for describing the contribution of PRP on the metabolism of this formula. The prototypes and metabolites identified in GXB-dePRP containing biological samples were also displayed in Fig. S1 and Table S1-S2. Due to the existing of PRP, the prototypes and metabolites related to flavonoids and triterpenoids saponins from PRP were naturally appeared in dosed GXB bio-samples, but not presented in GXB-dePRP dosed bio-samples. However, the xenobiotics derived from TP and AMB did not displayed obvious difference between GXB and GXB-dePRP groups, which inferred that PRP could not significantly influence the metabolite pathway of TP and AMB in GXB.

3.3.1 Plasma

According to the result of metabolic profile of GXB, five major components were selected for quantification and pharmacokinetic study to explore their kinetic features and exposure characteristics in plasma. Among of them, nitrogenous of 2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole-3-carboxylic acid, steroidal saponins of (25R)-timosaponin B II and macrostemonoside B were derived from AMB; other types of (-)-loliolide were originated from TP; and tritepenoid saponins of glycyrrhetic acid from PRP. As shown in Fig. 5 and Table 2, almost all the major prototypes exhibited a lower maximum plasma concentration (C_max) at 14.08 ± 0.52, 2.45 ± 0.13, 9.73 ± 0.50, 8.09 ± 0.46, 53.65 ± 3.40 ng/mL for 2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole-3-carboxylic acid, (25R)-timosaponin B II, macrostemonoside B, (-)-loliolide and glycyrrhetic acid, respectively. The time to reach the maximum concentration (T_max) of 2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole-3-carboxylic acid, (25R)-timosaponin B II, macrostemonoside B were reached quickly than other two absorbed components, whereas almost all these five analytes displayed a long half-time with higher T_1/2 values. The value of area under the concentration–time curve (AUC) could reflect the overall exposure of absorbed constituents in vivo. Glycyrrhetic acid showed the maximum AUC value, followed by macrostemonoside B, 2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole-3-carboxylic acid, (-)-loliolide and (25R)-timosaponin B II, with values of 2232.93 ± 104.29, 280.31 ± 8.29, 198.21 ± 4.20, 188.00 ± 5.53, 88.50 ± 3.55 h • ng/mL, respectively. Considering the ignorable abundance in GXB extract, the high exposure of glycyrrhetic acid in plasma was mainly attributed to de-glucuronization of 18 β-glycyrrhizic acid in vivo, which exhibited a higher content in this formula and that might be highly associated with the microbiota or intestine metabolism.

Close

Fig. 5

Mean plasma concentration–time profile of 5 components and peak area-time profiles of 6 semi-quantitative components after oral administration of GXB and GXB-dePRP (mean ± SD, n = 6).

Export to PPT

Since few prototypes were observed with available sensitivity, several metabolites transformed by prototypes could be detected according to the results of metabolic profile orally administration of GXB (Table S2), and the dynamic changes of metabolites together with those of prototypes were helpful to depict the whole dynamic feature in vivo of GXB (Table 1). Therefore, a total of six metabolites detected in dosed plasma were selected to monitor their in vivo dynamic process, including one metabolite of nitrogenous (pedatisectine D + CH₂), four metabolites of steroidal saponins (macrostemonoside G-2glc-gal, macrostemonoside G-2glc-gal-2, macrostemonoside G-2glc-gal-3, capsicoside A₂) and one metabolite of tritepenoid saponins (glycyrrhetic acid + gluA). Due to the lack of available reference standards, the semi-quantitative method obtained peak area values of these six metabolites at different time points to characterize in vivo dynamic profiles, and was consistent with the following published literature (Qin et al., 2016). The results of mean plasma peak area-time profiles of above metabolites were displayed in Fig. 5 and pharmacokinetic parameters calculated by peak area were illustrated in Table 2. Compared with those of corresponding prototypes, the T_max values were obviously longer for most metabolites in this experiment, and the mean residence time (MRT_0-∞) values were also slightly prolong. As depicted in Fig. 5, especially for four metabolites of steroidal saponins, they were maintained a relatively stable concentration for a long time in rat plasma, which might be attributed to the continuous transformation of these components abundant in GXB formula, suggesting that these metabolites could play an important role in the improvement of CHD syndrome.

3.3.2 Intestine content

For the reason that small intestine possessing relative larger surface area, which is the important place of in vivo disposal, and the majority of prototypes and metabolites after administration of GXB were also observed in rat small intestine content. Thus, the major prototypes and metabolites in intestine content were further analyze to characterize their kinetic feature in vivo. Consequently, a total of 15 prototypes (Fig. 1) were chosen to determine their concentration–time changes, and 5 analytes presented in plasma were also included. As depicted in Fig. 6 and Table 3, nine steroidal saponins from AMB were quickly reached the maximum concentration in intestine content with the values of T_max<2 h. Their value of C_max could reached the level of μg/g, and which were significant higher by comparing with the content levels of two shared steroidal saponins ((25R)-timosaponin B II and macrostemonoside B) in plasma. The values of MRT_0-∞ for these components were<8 h, which indicated that most of steroidal saponins could be quickly metabolized and/or discharged into the next intestinal tract. Whereas these components showed a higher exposure in intestine content, which might be responsible for the high abundance in GXB extract, especially for (25 R)-timosaponin B II and macrostemonoside B. Luteolin-7-O-β-glucoside and rutin were two flavonoids derived from TP, of which reached the maximum concentration within 2 h. Since the higher abundance of rutin in GXB extract, rutin displayed a higher C_max value than that of luteolin-7-O-β-glucoside. Similarly, these two flavonoids also exhibited a short residual time in small intestine tract. 2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole-3-carboxylic acid not only rapidly reach to the C_max within half an hour, but also quickly eliminate in intestine content. (-)-loliolide is originated from TP, compared with that of pharmacokinetic parameters in plasma, it could more quickly reach to the C_max, while other parameters did not exhibit significantly difference in rat plasma and intestine content. Glycyrrhetic acid and 18 β-glycyrrhizic acid are two common triterpenoid saponins derived from PRP, and bimodal phenomena were apparently observed in the dynamic changes of intestine contents, which were attributed to the hepatic and intestinal circulation or biotransformation between aglycone and glycoside. Relative to the abovementioned analytes, the T_max of these two constituents were prolonged more than 7 h. Interestingly, the abundance of glycyrrhetic acid in GXB extract were much lower than 18 β-glycyrrhizic acid (Lin et al., 2021), but the exposure level in intestine content were almost presented at the same level, which might be caused by the quickly biotransformation from 18 β-glycyrrhizic acid or other analogues.

Close

Fig. 6

Concentration-time curves of 15 components in intestine content after oral administration of GXB and GXB-dePRP in rats (mean ± SD, n = 5).

Export to PPT

Generally, the residence time of 15 prototypes were much shorter than corresponding metabolites detected in intestine content from the results of concentration–time curves and pharmacokinetic parameters (Figs. 6 and 7 and Table 3). As we all known, longer residence time in vivo will absolutely enhance the drug action time so as to improve the efficacy of GXB. Therefore, it could speculate that the metabolites detected in intestine content exert a more important therapeutic effects than corresponding prototypes.

Close

Fig. 7

Mean peak area-time profile of 14 metabolites after oral administration of GXB and GXB-dePRP in rats (mean ± SD, n = 5).

Export to PPT

In addition, we longitudinally compared the content distribution of various prototype components in GXB extract (Lin et al., 2021), GXB-contained plasma, and GXB-contained small intestine contents. The results (Fig. 8) can be more intuitive to find that the exposure profile of prototypes in rat plasma were apparently different from that of GXB extract. For example, the steroidal saponins with a very low exposure in rat plasma, but possessing a higher abundance in GXB extract. On the contrary, low percentage of glycyrrhetic acid in GXB extract, while high exposure in rat plasma. The exposure characteristic of prototypes in intestine content were generally similar to those in GXB extract. Especially, steroidal saponins were the major abundant components in GXB extract and intestine content, indicating that the contribution of so many components exposure in the gut to the efficacy of GXB should not be ignored, and the major exposure components of steroidal saponins in small intestine content play an important role in the efficacy of GXB.

Close

Fig. 8

The content of different prototypes in GXB extract, GXB-contained rat plasma and GXB-contained rat intestine content (pink, green, blue, yellow and white columns respectively belongs to steroidal saponins, triterpenoids saponins, nitrogenous, flavonoids, other types components.) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

Export to PPT