Identifying the active components of Baihe–Zhimu decoction that ameliorate depressive disease by an effective integrated strategy: a systemic pharmacokinetics study combined with classical depression model tests

Baihe (Lilium brownii var. viridulum) and Zhimu (Rhizoma anemarrhenae) were purchased from Shanghai Kangqiao Chinese Medicine Tablet Co., Ltd. (Shanghai China) and authenticated by Dr. Zhixiong Li (Shanghai Research Center for Modernization of TCM, Shanghai Institute of Materia Medica, Shanghai, China). Mangiferin, neomangiferin, timosaponin BII, timosaponin BIII and timosaponin AIII (purity > 98%) were supplied by Chengdu Biopurify Phytochemicals Co., Ltd. (Chengdu, China), and fluoxetine hydrochloride was provided by Eli Lilly & Co., Ltd. (Indianapolis, USA). HPLC-grade agents, including methanol, acetonitrile and formic acid, were obtained from Merck & Co., Inc. (Darmstadt, Germany). All other chemicals were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China) and were of analytical grade. The deionized water was purified by a Milli-Q system (Millipore, Billerica, MA, USA).

Baihe (300 g) and Zhimu (150 g) were cut into slices and mixed together at a weight ratio of 2:1. Then, the mixture (450 g) was extracted twice with boiling water (4500 mL) for 2 h each time. After filtration, the supernatant was condensed to 300 mL under reduced pressure to obtain Baihe–Zhimu decoction (BZD). The chemical characteristics of BZD were further investigated by HPLC-QQQ MS to ensure the chemical consistency of the tested sample. Five major components, including timosaponin BII, timosaponin BIII, mangiferin, neomangiferin and timosaponin AIII, were identified by comparing the retention times with those of the chemical standards, and their concentrations were determined to be 8979.80, 4191.10, 2649.02, 1624.27 and 442.91 μg/g BZD, respectively. In addition, each herb in BZD was individually extracted with boiling water using the same method as motioned above. The prepared Zhimu decoction (ZD) was fractionated by microporous resin to obtain three fractions, including polysaccharides of Zhimu (PZ, elution with water), xanthones of Zhimu (XZ, elution with 20% EtOH) and saponins of Zhimu (SZ, elution with 60% EtOH).

The HPLC-QQQ MS method was developed according to a previous study in our group [16]. Briefly, five compounds were simultaneously determined in plasma, liver and brain samples using a 1260 Series liquid chromatography system (Agilent Technologies, Palo Alto, CA, USA) coupled to a 6460 triple quadrupole mass spectrometer with an electrospray ionization source (Agilent Technologies, Palo Alto, CA, USA). Chromatographic separation was performed on an ACE Super C18 column (100 mm × 2.1 mm, 3.0 µm, Advanced Chromatography Technologies Ltd. Aberdeen, Scotland) at a temperature of 40 °C. The mobile phase consisted of water containing 0.1% formic acid (A) and acetonitrile containing 0.1% formic acid (B). The gradient program was performed as follows: 0–3 min, 92% A; 3.5–4.0 min, 88–60% A; 5.5–6 min, 60–55% A; 6.6–7 min, 55–5% A; 11–11.01 min, 5–92% A; 11.01–13.5 min, 92% A. The flow rate was set to 0.35 mL/min in the time range of 0–6.6 min and kept at 0.45 mL/min between 7 and 13.5 min. Five components were monitored in negative multiple reaction monitoring (MRM) mode. The parameters of the mass spectrometer were as follows: capillary voltage, 3500 V; nebulizer, 45 psi; gas temperature, 350 °C; gas flow rate, 12 L/min; sheath gas temperature, 400 °C; sheath gas flow rate, 8 L/min. MassHunter Workstation software (Agilent Technologies, Palo Alto, CA, USA) was used for system operation and data analysis.

All Sprague–Dawley (SD) rats and ICR mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. and kept in the breeding room at 22 ± 2 °C and 50 ± 10% humidity in Shanghai Institute of Materia Medica (SIMM). The experimental protocols were approved by the Institutional Animal Care and Use Committee in SIMM. After 1 week of acclimation, all the SD rats were orally administered BZD (15 g/kg). The biological samples were collected from the rats after fasting for 12 h. After anaesthetization, the rats were dissected at 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 7 h, 10 h, 20 h and 40 h (n = 5 for each time-point). The hepatic portal vein plasma (2 mL) and the systemic plasma (6–8 mL) were subsequently harvested, and finally the liver and brain tissue samples were collected. The plasma sample were centrifuged at 9500×g for 5 min, and the tissue samples were washed with saline solution. All samples were stored at − 80 °C until analysis. The samples were prepared according to the method described in the previous study by our group [16].

The antidepressant activity of the drugs was evaluated by classic behavioral measurements, such as the forced swimming test (FST) and tail suspension test (TST). All the ICR mice were randomly divided into several groups (n = 8) as follows: a control group with stimulation and different test groups subjected to FST and TST after the oral administration of different drugs for 8 days, including fluoxetine (10 mg/kg), BZD (3 g/kg), ZD (3 g/kg), Baihe decoction (BD, 3 g/kg), PZ (21.2 mg/kg), XZ (21.2 mg/kg), SZ (21.2 mg/kg), timosaponin BII (10 mg/kg) and timosaponin BIII (10 mg/kg).

The FST was performed according to a previously described method with minor modifications [17]. Briefly, each mouse was individually placed in an open water-filled cylinder (H: 50 cm; Ø: 20 cm; water depth: 35 cm; temperature: 23–25 °C) and allowed to swim for a period of 6 min. The total time of immobility was recorded during the last 4 min of the testing duration. Immobility is defined as the mice floating in the water without movement. The TST was conducted as described in the literature [18]. Briefly, the mice were suspended 20 cm above the floor for 6 min. The time of immobility was recorded after the first 2 min. Mice in a completely motionless state were considered immobile.

The contents of dopamine (DA) and serotonin (5-HT) in the plasma were determined using ELISA Kits (Shanghai Jianglai Biotech Co., Ltd., Shanghai, China) according to the manufacturer’s instructions.

A noncompartmental analysis was carried out using WinNonlin software (Pharsight 6.2, NC, USA) to calculate the PK parameters. The significance of the results in behavior assessment was analyzed by unpaired Student’s t-tests. A P value less than 0.05 was considered significant. The liver extraction ratio (ER) indicated the fraction of hepatic clearance and first-pass effect, and the calculation formula was as reported in a previous study by our group [16].

The HPLC-QQQ MS method has already been developed according to the previous study in our group [16]. In brief, full validation of the selectivity, linearity, accuracy, precision, matrix effect, extraction recovery, and stability, was performed for the simultaneous determination of two xanthones (neomangiferin and mangiferin) and three saponins (timosaponin BII, timosaponin BIII and timosaponin AIII) in the biological matrix by HPLC-QQQ MS. These results confirmed that two xanthones and three saponins in the biological matrix could be simultaneously determined by developed method [16]. Based on the developed method, five major components, including timosaponin BII, timosaponin BIII, mangiferin, neomangiferin and timosaponin AIII, were selected for simultaneous determination and pharmacokinetic analysis in the portal vein plasma, liver tissue, systemic plasma and brain tissue. The related pharmacokinetic parameters are summarized in Table 3.

Portal vein plasma is the site after gut absorption but before hepatic disposition. As displayed in Fig. 2, five compounds were accurately quantified in the portal vein plasma after oral administration of BZD. Figure 2 shows that timosaponin BII, timosaponin BIII, mangiferin and timosaponin AIII exhibited the double-peak phenomenon in the concentration–time curves, which may be caused by enterohepatic recirculation. The plasma concentrations of timosaponin BII, timosaponin BIII, mangiferin, neomangiferin and timosaponin AIII reached a maximum plasma concentration (C_max) at 880.38 ± 159.95 ng/mL, 226.39 ± 43.92 ng/mL, 1757.12 ± 219.18 ng/mL, 64.64 ± 12.00 ng/mL and 61.79 ± 25.59 ng/mL, respectively. The T_max values of timosaponin BII, timosaponin BIII and neomangiferin were greater than those of mangiferin and timosaponin AIII, and the rank order of t_1/2 was timosaponin BII > timosaponin BIII > timosaponin AIII > mangiferin > neomangiferin. This result revealed that timosaponin BII was eliminated more slowly. The area under the concentration–time curve (AUC) is usually regarded as the objective marker for exposure to chemical components and as predictive of pharmacological responses [19]. The AUC values of timosaponin BII, timosaponin BIII, mangiferin, neomangiferin and timosaponin AIII were 1094.92 ± 183.89 ng/mL, 369.20 ± 74.63 ng/mL, 4020.89 ± 397.02 ng/mL, 34.36 ± 2.60 ng/mL and 582.66 ± 104.67 ng/mL, respectively. The large AUC values of the four compounds other than neomangiferin indicated good absorption and utility in the portal vein plasma.

As displayed in Fig. 3, five compounds were determined accurately in the liver after oral administration of BZD. In the liver, timosaponin AIII had the maximum exposure, and the AUCs of timosaponin BII, timosaponin BIII, mangiferin, neomangiferin and timosaponin AIII were 6518.05 ± 411.22 ng/g, 859.42 ± 120.72 ng/g, 2520.63 ± 118.68 ng/g, 751.52 ± 87.96 ng/g and 199,598.41 ± 7787.31 ng/g, respectively. These values were different from those in the portal vein plasma, especially the highest value of timosaponin AIII. These results suggested that timosaponin AIII largely accumulated in the liver.

After hepatic disposition, timosaponin BII, timosaponin BIII, mangiferin, neomangiferin and timosaponin AIII were transported to the systemic plasma. Figure 4 shows the similar pharmacokinetic properties of the five compounds detected in the systemic plasma and the portal vein plasma. Mangiferin showed the maximum AUC, followed by timosaponin BII, timosaponin BIII, timosaponin AIII and neomangiferin, with values of 4217.27 ± 177.38 ng/mL, 975.80 ± 253.95 ng/mL, 326.47 ± 54.29 ng/mL, 156.12 ± 21.63 ng/mL and 78.15 ± 5.66 ng/mL, respectively. Except for mangiferin and neomangiferin, the AUCs of the other compounds were much higher in the portal vein plasma than in the systemic plasma, corresponding to the effective recovery (ER) of timosaponin BII, timosaponin BIII and timosaponin AIII at 10.88%, 11.57% and 73.21%, respectively. In contrast, the ER values of mangiferin and neomangiferin were − 4.88% and − 127.44%, respectively. These results indicated that some other components may be converted to mangiferin and neomangiferin after hepatic metabolism in vivo.

As shown in Fig. 5, only two components were accurately quantified in the cerebellum and hippocampus after oral administration of BZD. Similar to those in the systemic blood, timosaponin BII and timosaponin BIII exhibited an obvious double-peak phenomenon in the time-concentration curves in the cerebellum. However, a multiple-peak phenomenon occurred in the hippocampus. This observation may be attributed to the multiple sites of intestinal absorption. The C_max and AUC of timosaponin BII were 371.90 ± 153.80 ng/g and 9093.43 ± 1034.21 ng/g, respectively, with a larger T_max at 10 h than that of timosaponin BIII (1 h). The C_max and AUC of timosaponin BIII were 198.23 ± 122.31 ng/g and 2559.67 ± 928.52 ng/g, respectively. These results indicated that timosaponin BII and timosaponin BIII, which achieved exposure in the brain tissue, may be the main effective components of BZD.

To verify the screening results of the method, the FST and the TST were employed to assess the antidepressant effects of the chemical compounds detected in the brain. As shown in Table 4, both timosaponin BII and BIII markedly attenuated the time of immobility compared with that of control mice in both the FST (P < 0.05) and TST (P < 0.01), suggesting that they could alleviate depressive disorder. Furthermore, the timosaponins BII and BIII were used to screen for the active mechanism by using 5-HT and DA assays. As shown in Table 5, the levels of 5-HT in the timosaponin BII-treated group and the timosaponin BIII-treated group were both significantly decreased compared with that of the control group. These results further confirmed that timosaponins BII and BIII should be considered effective components of BZD.