Influence of wine-processing on the pharmacokinetics of anthraquinone aglycones and glycosides from rhubarb in hyperlipidemic hamsters

Abstract

Crude and wine-processed rhubarb products are routinely used as drastic purgative and hypolipidemic agents respectively. To better understand this discrepancy in therapeutic applications, their pharmacokinetics should be investigated. In this study, a reliable high performance liquid chromatography-tandem mass spectrometry method was first developed for the simultaneous determination of anthraquinone aglycones and glycosides from rhubarb in plasma. With satisfactory performance, the developed method was then applied to investigate the pharmacokinetic differences of these anthraquinones after oral administration of crude and wine-processed rhubarb to hyperlipidemic hamsters. The results showed that wine-processing could improve the exposures of emodin, chrysophanol, aloe-emodin, chryphanol-8-O-β-D-glucoside and physcion-8-O-β-D-glucoside, which are considered as hypolipidemic constituents in rhubarb. This is the first attempt to evaluate the effect of wine-processing on the pharmacokinetic behaviors of anthraquinones in rhubarb, and the results could be useful to guide the clinical applications of different rhubarb products and reveal the processing mechanism.

---

1. Introduction

Hyperlipidemia is recognized as a major risk factor of coronary heart disease, which is the leading cause of death worldwide.^1,2 Many chemical drugs, such as statins, fibrates, ezetimibe, niacin, and bile acid sequestrants are widely used for the treatment for hyperlipidemia. However, these drugs might cause many adverse effects, such as muscle myopathy,³ pancreatitis,⁴ vomiting,⁵ hepatotoxicity⁶ and so on. Thus, the management of hyperlipidemia is still a major challenge.

Traditional Chinese Medicines (TCMs) have been increasingly accepted worldwide for minimal adverse effects.⁶ In China, TCMs play an important role in the treatment of hyperlipidemia.^7,8 Rhubarb, stipulated as the dried root of Rheum palmatum L., Rheum tanguticum Maxim. ex Balf. or Rheum officinale Baill. in Chinese Pharmacopoeia, is one of the most frequently employed TCMs in treating hyperlipidemia.^9–11 However, the clinical use of crude rhubarb (DH) might lead to adverse effects, such as diarrhea and abdominal pain.¹² To moderate its drastic purgative action and alleviate the abdominal pain, DH is directly stewed or steamed with Chinese rice wine for long time to prepare wine-processed rhubarb (SDH). With such discrepancy in therapeutic effects, an understanding of their pharmacokinetic differences is crucial to guide their rational applications.

According to previous literature, anthraquinone derivatives including rhein, emodin, aloe-emodin, chrysophanol, physcion and their glycosides are considered as the major hypolipidemic constituents of rhubarb.^11,13 Except for hypolipidemic efficacy, a large body of literature has demonstrated that anthraquinone aglycones process various pharmacological actions such as cathartic,¹⁴ anticancer,¹⁵ hepatoprotective,¹⁶ anti-inflammatory,¹⁷ and antibacterial¹⁸ effects. Meanwhile, anthraquinone glycosides could accelerate the purgative activity of sennosides in rhubarb.¹⁹ Therefore, both aglycones and glycosides of anthraquinones should be used as marker constituents of rhubarb in pharmacokinetic studies. During the last decades, different analytical techniques have been described for the quantitative determination of anthraquinone derivatives in various biological specimens, including liquid chromatography coupled with different detectors, such as UV-Vis,^20,21 fluorimetric^22,23 and tandem MS detector.^24–27 However, none of these reported methods were to simultaneously quantify both aglycones and glycosides of anthraquinone derivatives in biological specimens.

Therefore, the objective of this paper is to develop a reliable liquid chromatography-electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS) method to simultaneously determine eleven naturally occurring anthraquinone derivatives, namely emodin (EM), rhein (RH), aloe-emodin (AL), chrysophanol (CH), physcion (PH), emodin-1-O-β-D-glucoside (EM-1-G), emodin-8-O-β-D-glucoside (EM-8-G), rhein-8-O-β-D-glucoside (RH-8-G), aloe-emodin-8-O-β-D-glucoside (AL-8-G), chryphanol-8-O-β-D-glucoside (CH-8-G) and physcion-8-O-β-D-glucoside (PH-8-G) (Fig. 1) in hamster plasma. The established method was then applied to clarify the influence of processing on pharmacokinetics of these bioactive constituents in rhubarb after oral administration to hyperlipidemic hamsters. The results of this study would provide evidence for the interpretation of processing mechanism and clinical application of rhubarb products.

Fig. 1 Chemical structures of the eleven anthraquinone derivatives and IS. Glc: glycoside.

2. Experimental

2.1. Chemicals and materials

Reference standards of EM, RH, AL, CH, PH and loxoprofen sodium (IS) were purchased from the National Institutes for Food and Drug Control (Beijing, China). EM-1-G, EM-8-G, AL-8-G, and RH-8-G were purchased from Shanghai Yilin Biotechnology Co., Ltd (Shanghai, China). CH-8-G was purchased from Chengdu MUST Biotechnology Co., Ltd (Sichuan, China). PH-8-G was purchased from Chengdu Chroma-Biotechnology Co., Ltd (Sichuan, China). The purity of reference standards was higher than 98% determined by HPLC-DAD. Sulfatase (type H-1, 15 275 units per g, from Helix pomatia, containing 300 000 units per g of β-glucuronidase) was purchased from Sigma Chemical Co. (St. Louis, MO). Methanol (HPLC grade) and formic acid (analytical grade) were purchased from Merck (Darmstadt, Germany) and the First Chemical Company of Nanjing (Jiangsu, China) respectively. Deionized water was prepared by a Milli-Q system (Millipore, MA, USA).

Rhubarb (the dried root of Rheum palmatum L., DH) was collected from Li County, Gansu Province and authenticated according to the current standard of Chinese Pharmacopoeia. DH was thoroughly mixed with Chinese rice wine, and steamed in a suitable non-ferrous container. After both the inner and outer parts turned into black, DH was dried in the shade to prepare SDH.⁹ The voucher specimens were deposited at the Herbarium of China Pharmaceutical University, Nanjing, China.

2.2. Apparatus and operation conditions

The analysis was performed on a Finnigan™ TSQ Quantum Discovery MAX™ LC-MS/MS system equipped with a Finnigan Surveyor LC pump, a Finnigan Surveyor autosampler and a triple quadrupole TSQ Quantum mass spectrometer (Thermo Fisher, Palo Alto, CA). Chromatographic separation was achieved on a Phenomenex Kinetex C18 column (100 mm × 2.1 mm, 2.6 μm) maintained at 40 °C. The chromatographic separation was achieved at a flow rate of 0.2 mL min⁻¹ with a mobile phase composed of acetonitrile (mobile phase A) and 0.1% formic acid (mobile phase B). The gradient program was as follows: 25–80% A at 0–2 min, 80–93% A at 2–14 min.

The mass spectrometer was operated in electrospray negative ionization using multiple reaction monitoring (MRM) mode. The MS/MS parameters were set as follows: spray voltage, 4000 V; vaporizer temperature: 300 °C, sheath gas pressure, 40 au (arbitrary units); auxiliary gas pressure, 10 au; capillary temperature: 300 °C; collision pressure: 1.5 mTorr. Previously optimized MRM ion pair transitions and collision energy of each compound were used in this study²⁸ and shown in Table 1.

Table 1 MS/MS detection parameters, regression data and quantification (LOQ) of the 11 anthraquinone derivatives

Compounds	[M − H]^− (m/z)	MRM transitions	Collison energy (V)	Regression equations^a	r^2	Linear range (ng mL^−1)	LOQ (ng mL^−1)
a A is the peak area ratio of mass detection (peak area of analyte/peak area of IS), C is the compound concentration injected and r^2 is the correlation coefficient of the equation.
RH	283.0	283.0 → 238.9	15	C = 82.37A − 2.674	0.9922	20–5000	20
EM	269.0	269.0 → 224.9	27	C = 2.155A − 0.3604	0.9981	2–500	2
AL	269.0	269.0 → 239.9	23	C = 88.85A − 5.476	0.9962	4–500	4
CH	253.0	253.0 → 224.9	30	C = 629.6A + 6.717	0.9968	50–6250	50
PH	283.0	283.0 → 239.9	27	C = 716.0A + 4.782	0.9974	50–6250	50
RH-8-G	445.0	445.0 → 238.9	34	C = 49.68A + 2.482	0.9930	4–500	4
AL-8-G	431.0	431.0 → 269.0	13	C = 47.24A + 2.438	0.9942	10–1250	10
EM-1-G	431.0	431.0 → 269.0	30	C = 7.692A + 0.6768	0.9965	2–250	2
EM-8-G	431.0	431.0 → 269.0	30	C = 4.122A + 0.3963	0.9954	2–250	2
CH-8-G	415.1	415.1 → 252.9	28	C = 29.32A + 0.8379	0.9929	2–250	2
PH-8-G	445.0	445.0 → 283.0	30	C = 22.97A + 0.3348	0.9950	1–125	1

---

2.3. Plasma sample preparation

The concentrations of both naturally occurring and total forms of anthraquinone derivatives were determined following a previously published procedure with minor modifications.^12,13 Plasma was divided into two aliquots. To quantify total forms of anthraquinone derivatives, 25 μL plasma was mixed with 25 μL sulfatase (1500 units per mL in pH 5.0 sodium acetate buffer, containing 30 000 units per mL of β-glucuronidase) and 5.0 μL ascorbic acid (250 mg mL⁻¹), then incubated at 37 °C for 4 h, which has been determined by a preliminary experiment. After hydrolysis, 5.0 μL blank methanol and 5.0 μL internal standard working solution was added into the plasma. The mixture was acidified with 5.0 μL of 0.1 N HCl and extracted twice with ethyl acetate (500 μL per time). After centrifugation at 14 000 rpm for 10 min, the combined supernatants were evaporated to dryness under a stream of nitrogen at 37 °C. The residue was reconstituted in 80 μL mobile phase consisting of acetonitrile and 0.1% formic acid (80 : 20, v/v) and centrifuged at 14 000 rpm for 10 min. 10 μL of supernatant were injected into the HPLC-MS/MS system. For quantification of anthraquinone aglycones and their glycosides, 25 μL plasma was subjected to the method described above expect for the addition of enzyme-free buffer and without incubation.

2.4. Preparation of rhubarb decoction

Pieces of rhubarb were soaked in water for 1 h (25 : 500, wt/v), and extracted by gentle heating until the solution was condensed to 250 mL. Finally the decoction was filtered while hot, condensed to 25 mL and stored at −80 °C for later use. The contents of anthraquinone aglycones and glycosides were determined before and after acid hydrolysis according to our previous method.²⁸

2.5. Animal treatment

Male LVG Syrian hamsters (eight weeks old; from Vital River Laboratory Animal Technology Co., Ltd, Peking, China) were housed and fed with a standard chow diet for one week before experiments. All animal experiments were performed in accordance with Guide for the Care and Use of Laboratory Animals and approved by the Animal Ethics Committee of China Pharmaceutical University. Following acclimatization, the food was switched to high fat diet (10% coconut oil (w/w) and 0.5% cholesterol (w/w)). After induction for two weeks, serum total cholesterol (TC), triglyceride (TG), and low-density lipoprotein-cholesterol (LDL-C) levels were measured with commercial kits to confirm the establishment of hyperlipidemic model.

Hyperlipidemic hamsters were randomly divided into two groups with seven in each group, and intragastrically administered with either DH decoction or SDH decoction at a dosage of 10 g kg⁻¹. Retro-orbital blood samples of 120 μL were collected with heparinized tubes before and 0.083, 0.167, 0.333, 0.75, 1.5, 3, 5, 8, 12, 18, 24, 30, 36 h after administration. Blood samples were immediately centrifuged at 5000 rpm for 10 min to obtain plasma. The plasma samples were stored at −80 °C until LC-MS/MS analysis.

2.6. Method validation

The method was validated in terms of specificity, calibration curve, sensitivity, precision, accuracy, matrix effect, recovery, stability, and dilution integrity in accordance with the Food and Drug Administration (FDA) bioanalytical method validation guidance.

2.6.1. Preparation of calibration standard, quality control (QC) samples. Stock solutions at 1.0 mg mL⁻¹ for each compound were prepared in methanol. An appropriate amount of each compound stock solution were mixed together to yield final concentrations of 125 μg mL⁻¹ for CH and PH, 100 μg mL⁻¹ for RH, 25 μg mL⁻¹ for AL-8-G, 10 μg mL⁻¹ for EM, AL and RH-8-G, 5 μg mL⁻¹ for EM-1-G, EM-8-G and CH-8-G, 2.5 μg mL⁻¹ for PH-8-G. Stock solution was serially diluted to obtain working solutions (dilution factors = 4, 10, 20, 40, 100, 200, 500, 1000). The calibration standards were prepared by spiking 5.0 μL of the corresponding working solutions into 25 μL blank plasma. Quality control (QC) samples at high QC (HQC), middle QC (MQC), and low QC (LQC) concentration levels were prepared in the same way as calibration standards based on linear ranges of the analytes.

2.6.2. Specificity. Chromatograms of blank plasma, blank plasma spiked with standards and hamster plasma samples obtained 0.167 h after oral administration of rhubarb decoction were compared to evaluate the specificity of the method.

2.6.3. Calibration curve and sensitivity. The calibration samples were prepared by spiking the corresponding standard solutions into blank plasma. The calibration curves were established by fitting the peak area ratio of analytes to IS versus the nominal concentration of analytes with weighted (1/c²) least square linear regression. Method sensitivity was evaluated by lower limit of quantification (LOQ), which was defined as the lowest concentration on the standard curve with the precision below 20% and accuracy within ±20%.

2.6.4. Precision, accuracy and dilution integrity. The intra- and inter-day precisions of the method were evaluated by assessing QC samples at LQC, MQC and HQC three concentration levels on three consecutive days with five repetitions each. Precision and accuracy, expressed as relative standard deviation (RSD%) and relative error (RE%) respectively, were calculated. Over-curve dilution was evaluated by analyzing three replicates (diluted 20 000 ng mL⁻¹ RH, 2000 ng mL⁻¹ EM and AL four times with blank plasma) daily over three days. The diluted samples were processed and analyzed as described in this method.

2.6.5. Matrix effect and recovery. The extraction recoveries of analytes at LQC, MQC and HQC levels were determined by comparing the peak area ratios obtained from the extracted QC samples with those from post-extracted blank plasma spiked with standards at the same concentration. The matrix effects were evaluated by comparing the peak areas of post-extraction blank plasma spiked with analytes with those of standard solutions at the same concentration.

2.6.6. Stability. Stability was evaluated at three QC concentrations under different conditions that occurred during sample analysis. The short-term stability and long-term stability were evaluated by exposing QC samples at room temperature for 12 h and low temperature (−80 °C) for 30 days respectively. The autosampler stability was measured by placing prepared QC samples under the autosampler condition (4 °C) for 24 h. The freeze–thaw stability was tested by analyzing QC samples undergoing three freeze (−80 °C)–thaw (room temperature) cycles on consecutive days. All these stability tests were run in triplicate.

2.7. Data analysis

Plasma concentrations of the anthraquinone derivatives were calculated by their calibration curves, respectively. The maximum plasma concentration (C_max) and the time to reach the maximum plasma concentration (t_max) were obtained directly from the measured data. Other pharmacokinetic parameters, including the area under the plasma concentration–time curve (AUC), mean residence time (MRT) and terminal elimination half-life (t_1/2), were estimated by non-compartment modeling using DAS 2.0 (Mathematical Pharmacology Professional Committee of China, Shanghai, China). All data are expressed as mean ± standard deviation. An unpaired Student's t-test was utilized to compare the differences in pharmacokinetic behaviors between the two groups. Values of p < 0.05 were considered to be statistically significant.

The pharmacokinetic parameters of anthraquinone derivatives after oral administration of SDH were calculated by the concentration of the wine-processed group and the normalized concentration of the wine-processed group, respectively. The normalized concentrations of the wine-processed group (C_n) were calculated as follows: C_n = C^*_wA_c/A_w, where C_w is the plasma concentration of wine-processed group; A_w is the content of analyte in SDH decoction; A_c is the content of analyte in DH decoction.

3. Results and discussion

3.1. Method validation

As shown in the representative chromatograms of blank plasma, blank plasma spiked with standards and plasma samples obtained 0.167 h after oral administration of rhubarb decoction (Fig. 2), there was no endogenous interference in the plasma under the established chromatographic condition. Although baseline separations of some analytes with different masses were not achieved, MRM transitions permitted unambiguous peak integrations for quantitative analysis. The calibration curves covering wide dynamic ranges showed good linear relationships with correlation coefficients (r²) higher than 0.992 (Table 1). The intra- and inter-day precisions (RSD%) were in the ranges 5.46–14.87% and 6.59–14.75%, respectively (Table 2). The values of accuracy were within 15% of the nominal values. The precision and accuracy at the upper limit of quantification and for over-curve dilution samples were within the acceptance criteria of bioanalysis.

Fig. 2 Representative MRM chromatograms of IS (1), RH (2), EM (3) AL (4), CH (5), PH (6), RH-8-G (7), AL-8-G (8), EM-1-G (9), EM-8-G (10), CH-8-G (11) and PH-8-G (12). (A) Blank plasma; (B) blank plasma spiked with eleven anthraquinone derivatives at LLOQ and IS; (C) 0.167 h sample plasma after administration of DH decoction.

Table 2 The precision, accuracy, matrix effect and recovery of 11 marker compounds in hyperlipidemic hamster plasma

Compounds	Concentration	Intraday precision		Interday precision		Matrix effect		Recovery
		Accuracy	RSD (%)	Accuracy	RSD (%)	Mean	RSD (%)	Mean	RSD (%)
RH	LQC	109.26	10.36	99.76	8.72	112.45	9.82	92.48	12.09
	MQC	97.65	9.72	89.63	9.75	103.94	9.02	89.75	7.62
	HQC	88.23	9.90	97.06	10.86	106.92	8.09	88.85	10.00
EM	LQC	108.21	5.46	108.36	14.20	91.93	12.68	87.89	14.50
	MQC	105.23	10.69	94.26	12.34	99.93	8.47	90.65	9.63
	HQC	98.67	8.76	106.75	13.72	91.54	12.40	87.92	11.31
AL	LQC	104.66	11.35	107.65	9.76	85.82	11.37	89.82	14.76
	MQC	89.56	8.67	90.75	11.67	86.01	10.40	86.19	12.34
	HQC	110.89	13.98	106.96	14.75	90.42	14.24	91.79	14.25
CH	LQC	114.38	13.87	104.54	14.62	110.97	10.32	89.87	11.18
	MQC	105.49	11.65	86.64	13.24	108.82	10.38	90.75	13.24
	HQC	86.98	14.87	86.65	13.75	103.92	9.10	89.06	12.48
PH	LQC	92.73	14.49	85.34	14.45	103.86	10.51	86.34	13.76
	MQC	106.34	10.89	105.35	11.67	89.85	11.65	87.75	9.71
	HQC	90.01	12.76	97.75	9.86	101.43	9.04	87.04	13.06
RH-8-G	LQC	109.11	14.61	112.67	10.76	114.65	12.93	92.83	10.34
	MQC	99.67	8.96	89.63	11.78	110.82	10.01	86.86	12.39
	HQC	104.34	8.65	102.38	6.59	114.21	9.92	88.76	10.83
AL-8-G	LQC	87.92	13.42	88.78	11.67	104.87	13.02	88.86	14.89
	MQC	101.54	13.23	103.29	8.90	113.02	12.43	89.78	13.08
	HQC	114.73	10.32	103.76	11.46	102.43	8.69	91.87	14.97
EM-1-G	LQC	89.91	14.11	111.76	12.79	87.02	10.93	89.87	6.34
	MQC	106.34	13.24	86.49	9.66	90.94	9.82	87.43	9.65
	HQC	99.78	9.67	87.37	11.38	93.04	8.94	90.75	10.24
EM-8-G	LQC	89.10	11.64	86.34	10.82	97.96	9.90	86.38	13.44
	MQC	110.34	9.99	103.92	8.82	103.73	11.24	92.90	8.89
	HQC	104.70	10.90	106.75	9.82	93.24	8.30	87.81	8.92
CH-8-G	LQC	111.89	8.63	89.93	10.22	113.07	10.18	90.56	12.49
	MQC	98.99	10.05	112.36	14.24	108.88	13.49	85.09	10.67
	HQC	108.34	11.11	110.65	12.43	100.43	9.43	88.80	8.01
PH-8-G	LQC	114.91	6.39	113.83	14.53	112.97	9.96	87.76	9.98
	MQC	110.23	7.36	105.90	11.56	102.39	9.89	87.62	10.76
	HQC	101.87	6.89	90.38	9.86	114.32	10.18	92.72	9.76

---

The matrix effect and recovery of the method were systematically studied and the results are shown in Table 2. No significant matrix effects for analytes were observed, indicating that no co-eluting substance influenced the ionization of them. It can also be seen from Table 2 that more than 85% recovery was achieved for all analytes.

According to the results of the stability experiments (Table 3), all analytes were confirmed to be stable during the sample storing and processing procedures, and there were no stability related problems during the routine pharmacokinetic analysis.

Table 3 The stability of 11 marker compounds in hyperlipidemic hamster plasma

Compounds	Concentration	Long-term stability		Short-term stability		Freeze and thaw stability		Post-preparative stability
		Mean	RSD (%)	Mean	RSD (%)	Mean	RSD (%)	Mean	RSD (%)
RH	LQC	89.13	9.72	101.38	9.82	94.22	8.62	89.83	9.24
	MQC	92.02	8.02	92.47	7.26	99.47	9.03	95.37	8.85
	HQC	99.24	14.25	89.26	9.94	103.28	8.29	109.39	14.29
EM	LQC	103.91	11.82	88.29	12.48	85.28	10.87	92.84	12.85
	MQC	91.04	12.89	102.38	9.28	104.29	13.42	104.28	11.32
	HQC	87.93	11.33	88.20	12.22	102.92	10.34	87.29	10.98
AL	LQC	110.41	14.29	86.01	10.24	94.20	13.95	102.74	14.61
	MQC	102.63	9.99	107.83	13.44	102.58	8.29	88.02	13.49
	HQC	101.87	14.29	87.20	12.49	103.45	9.27	106.78	12.49
CH	LQC	99.32	10.09	113.02	13.38	103.45	10.42	104.71	14.29
	MQC	89.34	14.27	89.26	10.26	102.48	14.29	92.45	12.43
	HQC	92.39	12.39	95.72	12.32	93.47	10.34	89.74	10.74
PH	LQC	109.05	12.92	109.28	12.57	89.21	12.73	108.88	13.24
	MQC	88.28	10.38	103.27	9.27	88.70	9.92	88.19	11.91
	HQC	103.20	12.26	103.26	10.76	111.38	12.33	111.24	9.26
RH-8-G	LQC	99.34	14.29	104.21	11.17	103.99	14.82	89.99	12.48
	MQC	88.21	9.65	89.37	13.77	103.92	9.71	102.48	12.67
	HQC	89.02	13.39	92.37	11.38	98.21	12.89	102.47	11.48
AL-8-G	LQC	102.43	13.44	103.48	12.38	88.67	12.67	102.47	13.77
	MQC	98.02	10.87	102.99	11.82	92.47	13.29	96.30	9.88
	HQC	99.02	13.02	110.37	10.47	102.32	13.22	88.39	14.21
EM-1-G	LQC	104.20	13.28	113.82	11.99	105.02	11.39	110.92	14.59
	MQC	89.88	10.05	109.27	11.23	110.38	13.23	86.98	11.11
	HQC	104.27	11.37	99.10	11.29	98.27	12.42	110.37	10.05
EM-8-G	LQC	86.99	12.72	109.85	10.27	103.76	10.08	102.47	12.48
	MQC	104.71	11.92	92.46	10.24	87.31	11.30	92.47	10.08
	HQC	102.34	12.30	109.25	12.08	98.29	12.37	94.28	8.78
CH-8-G	LQC	102.36	13.31	113.84	9.92	87.92	12.42	99.26	12.02
	MQC	108.39	12.28	92.31	13.29	92.44	11.98	102.49	7.29
	HQC	111.34	13.49	88.31	13.34	103.28	10.36	87.31	13.24
PH-8-G	LQC	89.32	14.30	85.92	10.31	87.47	13.21	102.38	10.32
	MQC	102.29	8.69	87.20	14.72	87.21	13.88	94.21	11.32
	HQC	103.48	10.25	89.19	13.2	102.48	10.08	109.83	9.89

---

3.2. Method application

Since previous researches^27,29 and the TCM theory of syndrome differentiation³⁰ have indicated that a drug would produce harmful effects for the healthy persons, it would be more reasonable to characterize the pharmacokinetics in pathological states. Therefore, high-fat diet induced hyperlipidemic hamsters were selected to compare the pharmacokinetic characteristics of DH and SDH because of its similarity to humans in terms of lipid profiles and high susceptibility to dietary cholesterol.^31,32

3.2.1. Confirmation of the success of hyperlipidemic model. The high-fat diet induced hyperlipidemic hamster model has been well characterized and widely used. The feeding of high-fat diet successfully induced hyperlipidemia in hamsters, as evidenced by the significantly increased serum levels of TC, TG and LDL-C compared with the normal hamsters (ESI Fig. S1†). Meanwhile, hepatic histopathological examination with H&E staining was performed (ESI Fig. S2†). The liver of hyperlipidemic hamsters showed microvesicular steatosis and mild infiltration of inflammation cells. Based on the above results, the hyperlipidemic model was considered to be successfully established.

3.2.2. Pharmacokinetics studies. The validated method was employed in a pharmacokinetic study of the eleven naturally occurring anthraquinone derivatives in hyperlipidemic hamsters treated with DH or SDH. The mean plasma concentration–time profiles of anthraquinones are illustrated in Fig. 3, and the corresponding pharmacokinetic parameters are summarized in Table 4.

Fig. 3 Mean plasma concentration–time profiles of the naturally occurring and total forms of anthraquinone derivatives in hyperlipidemic hamster after oral administration of DH or SDH. Results are mean + standard deviation. (A) RH, (B) EM, (C) AL, (D) CH-8-G, (E) PH-8-G, (F) RH-8-G, (G) total RH, (H) total EM, (I) total AL, (J) total CH.

Table 4 Pharmacokinetic parameters for anthraquinone derivatives in hyperlipidemic hamster following oral administration of DH or SDH (mean ± SD, n = 7)

Analytes	Group	AUC0–∞^a (ng h mL^−1)	MRT0–∞ (h)	t1/2 (h)	Cmax (ng mL^−1)	tmax (h)
a p < 0.05, versus oral administration of DH decoction.b Total form of corresponding anthraquinone derivatives.
RH	DH	43 932.80 ± 19 608.56	9.35 ± 0.73	7.48 ± 1.60	12 486.39 ± 2396.93	0.25 ± 0.24
	SDH	19 641.48 ± 7199.81^a	8.22 ± 1.42^a	7.82 ± 2.07	8838.11 ± 8048.06^a	0.26 ± 0.24
	Normalized SDH	38 339.23 ± 14 053.69	8.22 ± 1.42^a	7.82 ± 2.07	17 251.57 ± 15 709.44	0.26 ± 0.24
EM	DH	475.90 ± 122.64	10.68 ± 1.98	7.30 ± 2.38	105.85 ± 72.39	0.20 ± 0.09
	SDH	390.41 ± 115.12	10.89 ± 1.83	7.93 ± 1.40	89.93 ± 19.61	0.17 ± 0.00
	Normalized SDH	710.45 ± 209.49^a	10.89 ± 1.83	7.93 ± 1.40	163.65 ± 35.69^a	0.17 ± 0.00
AL	DH	379.91 ± 38.91	7.90 ± 3.31	5.94 ± 2.94	97.98 ± 14.42	0.19 ± 0.06
	SDH	452.17 ± 95.62	7.35 ± 2.10	5.21 ± 1.76	101.15 ± 27.22	0.17 ± 0.00
	Normalized SDH	720.09 ± 152.27^a	7.35 ± 2.10	5.21 ± 1.76	161.08 ± 43.35^a	0.17 ± 0.00
CH-8-G	DH	63.03 ± 21.95	5.75 ± 4.58	5.06 ± 4.34	39.96 ± 24.98	0.30 ± 0.21
	SDH	47.56 ± 12.27	4.80 ± 1.55	4.46 ± 0.96	32.84 ± 8.57	0.21 ± 0.12
	Normalized SDH	51.75 ± 13.35	4.80 ± 1.55	4.46 ± 0.96	35.74 ± 9.33	0.21 ± 0.12
PH-8-G	DH	53.32 ± 8.69	8.80 ± 2.39	6.02 ± 1.76	9.29 ± 2.96	0.24 ± 0.08
	SDH	51.04 ± 31.04	9.83 ± 2.12	8.22 ± 3.87	7.43 ± 0.77	0.19 ± 0.06
	Normalized SDH	67.92 ± 23.02	9.83 ± 2.12	8.22 ± 3.87	9.88 ± 1.02	0.19 ± 0.06
RH-8-G	DH	837.66 ± 405.88	5.83 ± 1.02	4.96 ± 1.18	237.34 ± 77.30	0.32 ± 0.21
	SDH	692.72 ± 236.41	6.35 ± 1.91	5.85 ± 2.36	254.49 ± 150.92	0.52 ± 0.29
	Normalized SDH	700.32 ± 239.00	6.35 ± 1.91	5.85 ± 2.36	257.29 ± 152.57	0.52 ± 0.29
RH^b	DH	89 366.40 ± 19 612.87	10.43 ± 1.67	8.58 ± 2.30	13 939.75 ± 3000.92	0.35 ± 0.29
	SDH	39 399.57 ± 7321.93^a	7.42 ± 0.90^a	6.16 ± 1.52^a	8416.99 ± 1600.80^a	0.36 ± 0.29
	Normalized SDH	70 114.12 ± 13 029.86	7.42 ± 0.90^a	6.16 ± 1.52^a	14 978.58 ± 2848.72	0.36 ± 0.29
EM^b	DH	22 918.42 ± 7336.11	12.61 ± 1.77	5.53 ± 2.23	1293.18 ± 463.49	5.67 ± 4.07
	SDH	14 104.30 ± 5815.87^a	10.02 ± 1.08^a	4.47 ± 0.60	988.57 ± 440.89	5.74 ± 3.60
	Normalized SDH	26 465.43 ± 10 912.95	10.02 ± 1.08^a	4.47 ± 0.60	1854.97 ± 827.29	5.74 ± 3.60
AL^b	DH	4066.53 ± 376.57	7.00 ± 0.68	4.75 ± 0.60	837.76 ± 97.24	0.17 ± 0.00
	SDH	3303.71 ± 518.39^a	7.06 ± 1.20	5.06 ± 0.92	750.84 ± 100.11	0.17 ± 0.00
	Normalized SDH	4504.42 ± 706.79	7.06 ± 1.20	5.06 ± 0.92	1023.73 ± 136.49^a	0.17 ± 0.00
CH^b	DH	8040.42 ± 767.54	13.757 ± 2.603	9.43 ± 2.34	732.773 ± 141.771	0.64 ± 0.46
	SDH	6688.68 ± 2205.33	13.72 ± 5.53	10.17 ± 4.16	694.90 ± 151.09	0.57 ± 0.22
	Normalized SDH	9898.48 ± 3263.63	13.89 ± 5.37	10.09 ± 4.24	1028.37 ± 223.59^a	0.51 ± 0.22

---

In order to quantify both the naturally occurring (aglycones and glycosides) and total (aglycones, glycosides, glucuronides, and sulfates) forms of anthraquinones, plasma samples were assayed before and after hydrolysis with sulfatase (containing β-glucuronidase). Before hydrolysis, RH and EM were present in all specimens; AL, RH-8-G, CH-8-G and PH-8-G could be detected up to 18 h or 12 h. However, the other anthraquinone derivatives either existed only transiently in the early phase after dosing, or could hardly be detected. As shown in Fig. 3, both anthraquinone aglycones and glycosides reached a maximum blood concentration at about 0.2 h post-dose, demonstrating a rapid absorption of anthraquinone derivatives from gastrointestinal tract. Anthraquinone aglycones and their total forms exhibited double-peak phenomena, which echoes the findings of reported literatures and might result from enterohepatic circulation, transformation of different compounds, double-site absorption and intestinal efflux.^{21,22,24–26} In addition, the AUC values of the total forms of RH, EM and AL after administered with DH were about 2.03, 48.16 and 10.70-fold of each corresponding free from, and the ratios for SDH were about 2.00, 36.13 and 7.31, implying glucuronide and sulfate were the predominant existing forms of anthraquinones in plasma of hyperlipidemic hamsters administered with either DH or SDH.

The pharmacokinetic differences of anthraquinone aglycones and their glycosides between DH and SDH were first estimated. Although the overall pharmacokinetic profiles of the anthraquinone derivatives in SDH and DH were similar, the pharmacokinetic parameters and normalized ones, calibrated based on the in vitro contents of anthraquinone derivatives in DH and SDH (ESI Fig. S3†), showed remarkable differences.

DH is one of the most frequently employed TCMs in treating hyperlipidemia, however, it would cause diarrhea and abdominal pain because of its drastic purgative action.^9–13 To alleviate these side effects, it should be used in the form of SDH. The extremely high plasma concentration and AUC of RH in hyperlipidemic hamsters administered with DH would provide some evidence for its side effects. Sennosides are regarded as the most potent purgative constituents in rhubarb.^33,34 In vivo, sennosides are metabolized in the gastrointestinal tract by intestinal flora followed by being absorbed into blood as the form of RH. The extremely high plasma concentration of RH due to the high content of sennosides would reflect the diarrhea and abdominal pain of DH. The in vitro contents of EM, AL, CH-8-G, PH-8-G before acid hydrolysis and that of CH after acid hydrolysis in SDH were significantly lower than those in DH (ESI Fig. S3†), however, the AUC of these constituents in hyperlipidemic hamsters administered with SDH were low but without significance when compared with DH. Furthermore, the normalized AUC values of free EM and AL increase significantly (p < 0.05) after wine-processing. Previous researches¹⁰ have proven that EM, CH and CH-8-G could significantly reduce plasma lipid levels. EM regulates lipid metabolism through the activation of AMP activated protein kinase³⁵ and PPARγ,³⁶ and CH improves lipid homeostasis by inhibiting protein tyrosine phosphatase 1B.³⁷ Briefly speaking, the elevated exposures of EM, AL, CH-8-G, PH-8-G and CH would enhance hypolipidemic efficacy of SDH. On the other hand, the side effects of DH were diminished after wine-processing, which was reflected by the reduced exposure of RH.

Other than anthraquinones, there still co-exist hundreds of compounds in rhubarb, such as tannins, naphthalins, phenylbutanones and stilbenes.^38,39 Meanwhile, the non-volatile compounds from Chinese rice wine, such as amine acid, biogenic amines, organic acid, phenols and esters,^40,41 were residual in wine-processing products. It remains to be further studied whether and how these co-existing compounds influence pharmacokinetics of anthraquinones and hypolipidemic effects of rhubarb.

4. Conclusion

In this study, a reliable LC-MS/MS method was first established to simultaneously determine both anthraquinone aglycones and glycosides in plasma. The method was fully validated and then applied to clarify the influence of processing on pharmacokinetics of these bioactive constituents in rhubarb after oral administration to hyperlipidemic hamsters. The decreased exposure of RH and the elevated exposures of EM, CH, AL, CH-8-G and PH-8-G after wine-processing would diminish the diarrhea side effects and enhance the hypolipidemic efficacy of DH. The information obtained in this study would be helpful to guide their rational applications and reveal the wine-processing mechanism.

Acknowledgements

This project was financially supported by the National Natural Science Foundation of China (No. 81274108 and 81403314), the Natural Research Foundation of Jiangsu Province (No. BK2012349), Fundamental Research Funds for the Central Universities (No. 2015PT032), Technology Innovation Program for post-graduate of Jiangsu province (No. KYLX_0617), and the Open Project Program of MOE Key Laboratory of Drug Quality Control and Pharmacovigilance (No. DQCP2015MS03).

References

1. A. G. Crawford, C. Cote, J. Couto, M. Daskiran, C. Gunnarsson, K. Haas, S. Haas, S. C. Nigam and R. Schuette, Popul. Health Manag., 2010, 13, 151–161.
2. J. H. Best, B. J. Hoogwerf, W. H. Herman, E. M. Pelletier, D. B. Smith, M. Wenten and M. A. Hussein, Diabetes Care, 2011, 34, 90–95.
3. S. D. Maggo, M. A. Kennedy and D. W. Clark, Drug Saf., 2011, 34, 1–19.
4. M. Lee, J. L. Saver, A. Towfighi, J. Chow and B. Ovbiagele, Atherosclerosis, 2011, 217, 492–498.
5. M. A. Al-Mohaissen, S. C. Pun and J. J. Frohlich, Mini-Rev. Med. Chem., 2010, 10, 204–217.
6. (a) N. E. Husain, A. T. Hassan, W. M. Elmadhoun and M. H. Ahmed, Expert Opin. Drug Saf., 2015, 14, 1445–1455; (b) M. Jiang, J. Yang, C. Zhang, B. Liu, K. Chan, H. Cao and A. Lu, Planta Med., 2010, 76, 2048–2064.
7. J. X. Chen, H. H. Zhao, Y. Yang, B. Liu, J. Ni and W. Wang, J. Ethnopharmacol., 2011, 134, 919–930.
8. X. F. Song, J. S. Wang, P. R. Wang, N. Tian, M. H. Yang and L. Y. Kong, J. Pharm. Biomed. Anal., 2013, 78, 202–210.
9. The State Pharmacopoeia Commission of PR China, Pharmacopoeia of P.R. China, Chemical Industry Press, Beijing, 2010, pp. 22–23.
10. S. K. Mishra, S. Tiwari, A. Shrivastava, S. Srivastava, G. K. Boudh, S. K. Chourasia, U. Chaturvedi, S. S. Mir, A. K. Saxena, G. Bhatia and V. Lakshmi, J. Nat. Med., 2014, 68, 363–371.
11. S. H. Tseng, T. Y. Chien, J. R. Chen, I. H. Lin and C. C. Wang, J. Evidence-Based Complementary Altern. Med., 2011, 2011, 249–254.
12. W. D. Xie, Y. N. Zhao and L. J. Du, J. Ethnopharmacol., 2012, 140, 345–367.
13. S. K. Mishra, S. Tiwari, A. Shrivastava, S. Srivastava, G. K. Boudh, S. K. Chourasia, U. Chaturvedi, S. S. Mir, A. K. Saxena, G. Bhatia and V. Lakshmi, J. Nat. Med., 2014, 68, 363–371.
14. Y. Qin, J. B. Wang, W. J. Kong, Y. L. Zhao, H. Y. Yang, C. M. Dai, F. Fang, L. Zhang, B. C. Li, C. Jin and X. H. Xiao, J. Ethnopharmacol., 2011, 133, 1096–1102.
15. B. Jelassi, M. Anchelin, J. Chamouton, M. L. Cayuela, L. Clarysse, J. Y. Li, J. Gore, L. H. Jiang and S. Roger, Carcinogenesis, 2013, 34, 1487–1496.
16. X. Y. Xing, Y. L. Zhao, W. J. Kong, J. B. Wang, L. Jia, P. Zhang, D. Yan, Y. W. Zhong, R. S. Li and X. H. Xiao, J. Ethnopharmacol., 2011, 135, 575–581.
17. B. Y. Hu, H. Zhang, X. L. Meng, F. Wang and P. Wang, J. Ethnopharmacol., 2014, 153, 846–853.
18. J. E. Kim, H. J. Kim, S. Pandit, K. W. Chang and J. G. Jeon, Fitoterapia, 2011, 82, 352–356.
19. K. Takayama, H. Tsutsumi, T. Ishizu and N. Okamura, Biol. Pharm. Bull., 2012, 35, 2204–2208.
20. H. L. Gong, W. F. Tang, H. Wang, Q. Xia and X. Huang, Phytother. Res., 2011, 25, 74–80.
21. A. H. Zhang, H. Sun, Y. Yuan, W. J. Sun, G. Z. Jiao and X. J. Wang, Fitoterapia, 2011, 82, 1160–1168.
22. C. S. Shia, S. Y. Tsai, J. C. Lin, M. L. Li, M. H. Ko, P. D. L. Chao, Y. C. Huang and Y. C. Hou, J. Ethnopharmacol., 2011, 137, 1388–1394.
23. C. S. Shia, S. H. Juang, S. Y. Tsai, P. H. Chang, S. C. Kuo, Y. C. Hou and P. D. L. Chao, Planta Med., 2009, 75, 1386–1392.
24. W. J. Wu, R. Yan, M. C. Yao, Y. Zhan and Y. T. Wang, Biomed. Chromatogr., 2014, 28, 564–572.
25. X. Liu, H. Li, L. Wu, J. W. Xing, Y. H. Poh, H. Cai and B. C. Cai, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2015, 980, 88–93.
26. C. J. Yang, S. H. Wang, X. W. Guo, J. H. Sun, L. Liu and L. J. Wu, J. Ethnopharmacol., 2015, 169, 305–313.
27. F. Fang, J. B. Wang, Y. L. Zhao, C. Jin, W. J. Kong, H. P. Zhao, H. J. Wang and X. H. Xiao, J. Ethnopharmacol., 2011, 137, 1492–1497.
28. M. Wang, Y. Tian, M. Y. Lv, F. G. Xu, Z. J. Zhang and R. Song, Anal. Methods, 2015, 7, 5375–5380.
29. J. B. Wang, H. P. Zhao, Y. L. Zhao, C. Jin, D. J. Liu, W. J. Kong, F. Fang, L. Zhang, H. J. Wang and X. H. Xiao, PLoS One, 2011, 6, e24498.
30. M. Jiang, C. Lu, C. Zhang, J. Yang, Y. Tan, A. P. Lu and K. Chan, J. Ethnopharmacol., 2012, 140, 634–642.
31. D. Y. Shao, G. E. Bartley, W. Yokoyama, Z. L. Pan, H. J. Zhang and A. Zhang, Food Chem., 2013, 139, 589–596.
32. F. C. Guo, C. H. Huang, X. L. Liao, Y. M. Wang, Y. He, R. N. Feng, Y. Li and C. H. Sun, Mol. Nutr. Food Res., 2011, 55, 1809–1818.
33. R. Kon, N. Ikarashi, C. Nagoya, T. Takayama, Y. Kusunoki, M. Ishii, H. Ueda, W. Ochiai, Y. Machida, K. Sugita and K. Sugiyama, J. Ethnopharmacol., 2014, 152, 190–200.
34. M. Matsumoto, A. Ishige, Y. Yazawa, M. Kondo, K. Muramatsu and K. Watanabe, PLoS One, 2012, 7, e31700.
35. T. F. Tzeng, H. J. Lu, S. S. Liou, C. J. Chang and I. M. Liu, Planta Med., 2012, 78, 943–950.
36. J. F. Xue, W. J. Ding and Y. Liu, Fitoterapia, 2010, 81, 173–177.
37. W. J. Lee, G. Yoon, Y. R. Hwang, Y. K. Kim and S. N. Kim, BMB Rep., 2012, 45, 141–146.
38. Q. X. Zheng, H. F. Wu, J. Guo, H. J. Nan, S. L. Chen, J. S. Yang and X. D. Xu, Chin. Herb. Med., 2013, 5, 9–32.
39. M. Ye, J. Han, H. B. Chen, J. H. Zheng and D. Guo, J. Am. Soc. Mass Spectrom., 2007, 18, 82–91.
40. M. Arbulu, M. C. Sampedro, A. Gómez-Caballero, M. A. Goicolea and R. J. Barrio, Anal. Chim. Acta, 2015, 858, 32–41.
41. Y. Huang, W. W. Lu, B. Chen, M. Wu and S. G. Li, Food Anal. Methods, 2015, 8, 825–832.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra27273d

---