Advertisement
Research ArticleArticle

SGLT-1 Transport and Deglycosylation inside Intestinal Cells Are Key Steps in the Absorption and Disposition of Calycosin-7-O-β-d-Glucoside in Rats

Jian Shi, Haihui Zheng, Jia Yu, Lijun Zhu, Tongmeng Yan, Peng Wu, Linlin Lu, Ying Wang, Ming Hu and Zhongqiu Liu
Drug Metabolism and Disposition March 2016, 44 (3) 283-296; DOI: https://doi.org/10.1124/dmd.115.067009

Abstract

Hydrolysis by lactase-phloridzin hydrolase (LPH) is the first and critical step in the absorption of isoflavonoid glucosides. However, the absorption characteristics of calycosin-7-O-β-d-glucoside (CG) slightly differ from other isoflavonoid glucosides. In this study, we used the rat intestinal perfusion model and performed pharmacokinetic studies and in vitro experiments to determine the factors influencing CG absorption and disposition. After oral administration of isoflavonoid glucosides, LPH was found to play minimal or no role on the hydrolysis of CG, in contrast to that of daidzin. CG was mainly transported into the small intestinal cells by sodium-dependent glucose transporter 1 (SGLT-1) as intact. This pathway could be the main mechanism underlying the high permeability of CG in the small intestine. CG was likely to be hydrolyzed in enterocytes to its aglycone calycosin by broad-specific β-glucuronides (BSβG) and glucocerebrosidase or rapidly metabolized. Calycosin was also rapidly and extensively metabolized to 3′-glucuronide in the enterocytes and liver, and the glucuronidation rates of calycosin and CG were much higher in the former. The metabolites were also transported into lumen by breast cancer resistance protein and multidrug resistance-associated protein 2. In conclusion, the enterocytes could be an important site for CG absorption, deglycosylation, and metabolism in rats. This study could contribute to the theoretical foundation and mechanism of absorption and disposition of flavonoid compounds.

Introduction

Isoflavonoids, a class of flavonoid polyphenolic compounds, are mainly found in soy products as conjugated glucoside and confer beneficial health effects (Murphy et al., 2002; Nurmi et al., 2002). Absorption and metabolism of soy isoflavonoid glucosides have been extensively studied in recent years to elucidate their physiologic functions (Walsh et al., 2007; Islam et al., 2014). Deglycosylation by lactase-phloridzin hydrolase (LPH), an enterocyte-specific enzyme localized to the brush-border membrane in small intestine, is a critical step in the absorption and metabolism of isoflavonoid glucosides. This process generates aglycones that can enter enterocytes through passive diffusion (Day et al., 2000; Walsh et al., 2007). After cellular accumulation, aglycones undergo extensive metabolism in enterocytes (Chen et al., 2005). The metabolites are either excreted into the digestive tract or passed into the portal vein by specific transporters and then metabolized by phase II enzymes in liver (Scalbert and Williamson, 2000; Del Rio et al., 2013). This pathway mainly causes low bioavailability of isoflavonoid glucosides. However, in our previous study, we found that the absorption characteristic of the main isoflavonoid glucoside in Astragali Radix differs from that of other soy isoflavonoid glucosides (Shi et al., 2015).

Astragali Radix is a medicinal herb with immunogenic properties. It is used to enhance immunity and treat cardiovascular diseases, such as myocardial infarction and heart failure (Zhao et al., 2008). More than 400 isoflavonoids have been identified in Astragali Radix (Ruan et al., 2015), including calycosin-7-O-β-d-glucoside (CG), a marker used routinely to monitor the quality of the herb. CG exhibits several pharmacological activities, such as anti-inflammatory effect in rabbit osteoarthritis model and antioxidative and neuroprotective effects in vitro (Choi et al., 2007; Jian et al., 2015). A recent study showed that the CG metabolite selectively modulates estrogen receptors, which promote angiogenesis in zebrafish embryos (Ruan et al., 2015). The aglycone calycosin of CG also exerts antidiabetic, antioxidative, and neuroprotective effects (Fan et al., 2003; Yu et al., 2005; Shen et al., 2006). CG and calycosin are used as the main components of a pharmaceutical drug that has been patented for prevention and treatment of estrogen receptor-mediated diseases and diabetic nephropathy (Ruan et al., 2015). Nevertheless, studies on the characteristics of CG disposition remain limited.

Recently, our group studied the pharmacokinetic profiles of the active components of Astragali Radix and found that CG and its glucuronide were rapidly absorbed, with a peak concentration of about 100 ng/ml, after oral administration of the water extract of this herb (Shi et al., 2015). This suggested that CG could traverse the enterocytes intact. This finding contradicts that of a previous study, which reported that isoflavonoid glucosides cannot cross the Caco-2 monolayer (Islam et al., 2014) and hydrolysis by LPH is not essential for CG absorption. Glucosides exhibit high polarity, which allows them to diffuse across the cellular membrane. As a result, some of these compounds can be transported into enterocytes via glucose transporters (GLTs) such as sodium-dependent glucose transporter 1 (SGLT-1) and glucose transporter 2 (GLUT-2) (Wolffram et al., 2002; Day et al., 2003). Thus far, studies on the effect of GLTs on isoflavonoid glucosides are limited and the underlying mechanism remain unclear.

Although LPH may not play a key role in CG hydrolysis, it was recently reported that calycosin glucuronide was the major component in plasma after oral administration of CG (Ruan et al., 2015). This finding suggested that most of the absorbed CG was hydrolyzed in enterocytes. Generally, after being transported into enterocytes, glucosides are deglycosylated by broad-specific β-glucuronides (BSβG) and glucocerebrosidase, which are two major intracellular β-glucosides hydrolases (Day et al., 2003). However, studies on the characteristics of CG hydrolysis in enterocytes are limited. Although a recent study examined calycosin glucuronidation (Ruan and Yan, 2014), it only focused on liver microsomes, and its metabolic characteristics in intestinal microsomes remain unexplored. UDP-glucuronosyltranferases (UGT) enzymes are also abundantly expressed in small intestines, which are the sites of first isoflavonoids metabolism after oral administration. Therefore, it is necessary to study the corresponding metabolic characteristics of calycosin and CG in intestinal microsomes.

In this work, we investigated CG absorption, hydrolysis, and metabolism after oral administration of pure CG to determine its physiologic functions. We used the rat intestinal perfusion model and performed in vivo pharmacokinetic studies with in vitro hydrolysis experiments and Caco-2 cell line to investigate CG absorption, hydrolysis, metabolism, and transportation in rats. The enzyme kinetics of calycosin and CG in rat liver/intestinal microsomes (RLMs and RIMs) was also assessed to characterize enzyme catalysis in liver and intestines.

Materials and Methods

Materials.

CG, calycosin, and daidzin (≥98%, HPLC grade) were purchased from Chengdu Must Biotechnology Co., Ltd. (Chengdu, China) and used to prepare standard solutions of phase II metabolites in our laboratory (Shi et al., 2015). Glucurolactone, CaCl2, MgSO4, KCl, KH2PO4, NaCl, Na2HPO4, NaHCO3, DMSO, uridine diphosphate glucuronic acid, alamethicin, MgCl2, MK571, ko143, d-saccharic-1,4-lactone monohydrate, HEPES, and Hanks' balanced salt solution (HBSS) were purchased from Sigma-Aldrich. Gluconolactone was obtained from Tokyo Chemical Industry Co., Ltd., and conduritol B epoxide was purchased from Cayman Chemical Company. Erlotinib, used as internal standard (IS), was purchased from Cayman Chemical Company. RLMs, RIMs, and rat intestinal S9 were prepared in our laboratory by using a previously published method (Jia et al., 2004). All other chemicals (typically analytical grade) were used as received.

Animals.

Male Sprague-Dawley rats (80–110 days old), weighing 250–300 g, were obtained from the Laboratory Animal Center of Guangzhou University of Chinese Medicine. Four rats were housed per cage in a room with unidirectional air flow under 40–70% relative humidity, 20–24°C, and 12-hour light/dark cycle. Isoflavonoids were not detected in the blank samples of plasma, bile, and perfusate, which were obtained through perfusion of a segment of the rat upper small intestine; this finding indicates the absence of isoflavonoids in the samples. The rats were fasted overnight and given access to water before the experiments. The animal protocol used in the study was approved by the Southern Medical University’s Ethics Committee.

Cell Culture.

Culture conditions for Caco-2 cells have been described previously (Hu et al., 1994). Briefly, a cell monolayer was prepared by seeding 4×105 cells per insert (Nunc, Thermo Fisher Scientific, Waltham, MA; surface area, 4.6 cm2; pore size, 3 μm). Cells were maintained at 37°C with 90% humidity and 50% CO2, fed every day, and determined ready for transport experiments (19–21) days after seeding. The integrity of each monolayer was determined by measuring the transepithelial electrical resistance (Millicell ERS, Merck Millipore, Darmstadt, Germany) before the experiments were conducted. Cell monolayers with transepithelial electrical resistance values less than 460 Ω/cm2 were discarded (Yang et al., 2011).

In Vivo Pharmacokinetic Study in Rats.

CG was dissolved in 20% (w/v) hydroxypropyl β-cyclodextrin aqueous solution, sterilized, and filtered through a 0.22 μm filter. Single dose of CG (0.5 mg/kg) was administered via the tail vein (intravenous). Blood samples (0.2 ml) were collected from the vein of the eye sockets and placed into dried heparinized tubes at 2, 5, 10, 20, 30, 45, 60, 90, 120, 180, 240, 360, 480, and 720 minutes. For the oral administration and intraperitoneal injection groups, blood samples (0.2 ml) were collected at 5, 10, 20, 30, 45, 60, 90, 120, 180, 240, 360, 480, and 720 minutes after administration of a single dose of CG (10 mg/kg). The blood samples were immediately centrifuged at 5510 g for 8 minutes, and plasma was stored at −80°C until analysis.

Plasma samples (50 μl) were mixed with 10 μl of IS solution (5.95 ng/ml erlotinib in methanol) and 200 μl of methanol. After vortexing to homogenize, the mixture was centrifuged at 19,357 g for 30 minutes. The supernatant was transferred into a new tube and evaporated to dryness under a nitrogen stream. The residue was reconstituted in 100 μl of 50% methanol solution and centrifuged at 19,357 g for 30 minutes. 1 μl of the supernatant was injected into the ultraperformance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) column for analysis.

Pharmacokinetic parameters were calculated using the WinNonlin 3.3 software (Pharsight Corporation, Mountain View, CA). Systemic bioavailability of CG after oral (Foral) administration was calculated as follows:Embedded Image(1)where AUCoral (0-∞) and AUCi.v (0-∞) are areas under the plasma concentration curves and calculated after oral and intravenous administration, respectively.

In Situ Absorption and Hydrolysis Experiments Using the Perfused Rat Intestinal Model.

Single-pass intestinal perfusion was performed similar to previously described methods with minor modifications (Xia et al., 2012; Cai et al., 2013). The perfusate buffer, a pH 6.5 phosphate buffer, contained 1.26 mM CaCl2, 0.81 mM MgSO4, 5.37 mM KCl, 0.44 mM KH2PO4, 170.67 mM NaCl, 0.34 mM Na2HPO4, 4.43 mM NaHCO3, 25.02 mM HEPES, and 2.5 or 10 μM CG. The perfusate buffer were then added with glucose or different concentrations of the inhibitors of LPH (glucurolactone), GLTs (phloridzin for SGLT-1, and phloretin for GLUT-2), or efflux transporters [ko143 for breast cancer resistance protein (BCRP) and MK571 for multidrug resistance-associated protein 2 (MRP2)] to investigate the effects of these compounds on the intestinal absorption and metabolism of CG. In addition, daidzin, which can be hydrolyzed by LPH (Wilkinson et al., 2003), was selected as the positive control to further confirm the effect of LPH on CG. After the rats were anesthetized by intraperitoneal injection of 1.5 g/kg urethane (50%, w/v), four segments of the rat intestine were simultaneously perfused. The duodenum, jejunum, ileum, and colon were cannulated using 1.5- to 2.0-cm-long cannulas, which can be easily disconnected or reconnected to the main perfusion tube. The main tube was attached to a syringe driven by an infusion pump (Model PHD2000 Harvard Apparatus, Cambridge, MA), and the flow rate was maintained at 10 ml/h. The inlet cannula was insulated, stored in a circulating water bath at 37°C, and perfused with two CG concentrations (10 and 2.5 μM). After washing for 30 minutes, which is sufficient to achieve steady-state absorption, the perfusate, bile, and plasma samples were collected every 30 minutes during the 2.5-hours period. After perfusion, the length of every intestinal segment was measured, and each tube containing the sample was weighed as described previously (Xia et al., 2012).

Briefly, 50% methanol with 5.95 ng/ml IS was used to dilute the perfusate and bile samples (1:10) for detection of CG and its metabolites. Plasma samples were also prepared using the same method. The prepared samples were centrifuged at 19,357 g for 30 minutes. One microliter of the supernatant was injected into UPLC-MS/MS columns for analysis.

CG permeability was represented by P*eff, which can be calculated using a previously described method (Wang et al., 2006).Embedded Image(2)P*eff represents the apparent permeability of the compound, which indicates the amount loss in the perfusate samples. Co and Cm are the inlet and outlet concentrations, which were corrected for water flux by using the sample weight, respectively; Gz (Gz=πDL/2Q) is a scaling factor incorporating flow rate (Q) and intestinal length (L) and diffusion coefficients (D) to obtain a dimensionless permeability value.

The amounts of absorbed CG (Mab) and the corresponding calycosin glucuronide excreted into the intestinal lumen (Mmet) were calculated as described previously (Wang et al., 2006). Generally, Mab is expressed asEmbedded Image(3)where τ is the sampling interval (30 minutes) and other parameters were identical with those defined in eq. 1. The amounts of metabolites (Mmet) were determined asEmbedded Image(4)where Cmet is the outlet concentration (nmol/ml) of the metabolites corrected for water flux.

Transport Experiments of Analytes in Cultured Caco-2 Cells.

Triplicate experiments were performed in HBSS; the pH of HBSS in both apical (AP) side and the basolateral (BL) side was 7.4. Protocol for transport experiments were similar to those described previously (Ye et al., 2013). In brief, the testing solution (2 ml, containing 5 μM calycosin) was loaded onto one side of the cell monolayer and blank HBSS (2 ml) was loaded onto the other side. Five sequential samples (0.5 ml) were taken at different times (0, 0.5, 1, 1.5, and 2 hours) from both sides of the cell monolayer and 250 μl of IS solution (5.95 ng/ml erlotinib in methanol) was immediately added to each sample to stabilize them. The same volume of testing solution or blank HBSS was immediately added to replace the samples obtained. Transport experiments were also performed in the presence of ko143 (10 μM) and MK571 (10 μM).

Intracellular concentrations of analytes were determined when the transport experiments were conducted. The procedure used to determine the analytes was the same as that described previously (Jeong et al., 2005). Briefly, the cell membranes were carefully washed three times with ice-cold blank HBSS at the end of the transport study. Cells attached to the polycarbonate membranes were cut off from the inserts, immersed in 0.3 ml blank HBSS, and sonicated for 30 minutes in an ice bath (4°C), and 50 μl of IS solution (5.95 ng/ml erlotinib in methanol) was immediately added to 100 μl each sample to stabilize them.

Briefly, 50% methanol was used to dilute the samples (1:5) for detection of analytes. The prepared samples were centrifuged at 19,357 g for 30 minutes. One microliter of the supernatant was injected into UPLC-MS/MS columns for analysis.

The rates of transport (Bt) were obtained from the rate of change in the concentration of the transported calycosin glucuronide, which can be expressed as a function of time and the volume of the sampling chamber (V):Embedded Image(5)The apparent intrinsic clearance (CL) was calculated here because the extracellular concentration of calycosin glucuronide could be very different from its intracellular concentration. CL was determined from the rate of glucuronide excretion (Bex) divided by the intracellular concentration of glucuronide (Ci):Embedded Image(6)where Bmax is the excretion rate of glucuronide and Km is the Michaelis constant. We assumed that the average volume of the engineered Caco-2 cells is 4 μl/mg protein (Yamaguchi et al., 2000). The intracellular concentration of metabolite was calculated from the total amount of intracellular glucuronide divided by the total volume of intracellular protein.

The fraction of the dose metabolized (fmet) was calculated here as a proper parameter to describe the extent of calycosin metabolism in the Caco-2 monolayer (Pang et al., 2009), where Mettotal is the total metabolite concentration and Caltotal is the total calycosin combined in the receiver side, the donor side, and within the cell at 2 hours:

Embedded Image(7)

Hydrolysis Experiment Using Rat Intestinal S9.

The hydrolysis characteristics of CG and the functions of different glucosidases were determined using rat intestinal S9. Incubation mixtures (total volume 400 μl) containing 50 mM potassium phosphate buffer (pH 7.4), CG (25, 100, and 400 μM), and 0.1 mg protein/ml rat intestinal S9 (duodenum, jejunum, and ileum) were incubated in the absence or presence of conduritol B epoxide (potent glucocerebrosidase inhibitor) or gluconolactone (specific BSβG inhibitor). Several studies reported the specificity of different enzyme inhibitors for intestinal glucosidases (Daniels et al., 1981; Day et al., 1998). The low, middle, and high concentrations of conduritol B epoxide were 1.25, 2.5, and 5 mM, whereas those of gluconolactone were 2.5, 5, and 10 mM, respectively. The reactions were performed at 37°C for 15 minutes and terminated by adding 200 μl of methanol with 5.95 ng/ml IS. The reaction mixtures were centrifuged at 19,357 g for 30 minutes, and 1 μk of the supernatant was injected into the UPLC-MS/MS for analysis. All samples were performed in triplicate.

In Vitro Enzymes Kinetic Studies.

Calycosin and CG were incubated with MgCl2 (0.88 mM), saccharolactone (4.4 mM), alamethicin (0.022 mg/ml), and RLMs and RIMs (final concentration = 0.01325 mg protein per ml) in 50 mM potassium phosphate buffer (pH 7.4). This reaction mixture, with a total volume of 400 μl with 0.5–200 μM calycosin and 10–800 μM CG, was used to measure glucuronidation kinetics. The reaction was initiated by adding 3.5 mM uridine diphosphate glucuronic acid, stored for 15 minutes at 37°C, and terminated by adding 200 μl of methanol with 5.95 ng/ml IS. The reaction mixtures were vortexed immediately, followed by centrifugation at 19,357 g for 30 minutes. Subsequently, 1 μl of the supernatant was injected into the UPLC-MS/MS for analysis. All samples were performed in triplicate.

Kinetic parameters were obtained based on the profile of Eadie-Hofstee plots as described previously (Wang et al., 2006). If Eadie-Hofstee plot was linear, formation rates (V) of glucuronide of calycosin and CG at different substrate concentrations (C) were fit to the standard Michaelis-Menten equation:Embedded Image(8)where Km is the Michaelis-Menten constant and Vmax is the maximum rate of glucuronidation.

If Eadie-Hofstee plots showed characteristic profiles of atypical kinetic (Hutzler and Tracy, 2002), the data from these atypical profiles were fit to eq. 6 using the ADAPT II program. To confirm the best-fit model, the model candidates were discriminated using the minimum Akaike’s information criterion value, and the rule of parsimony was employed.

The following eq. 6 describes enzyme reactions with biphasic kinetic:Embedded Image(9)where Vmax1 is the maximum enzyme velocity of the high-affinity phase, Vmax2 is the maximum velocity of the low-affinity phase, Km1 is concentration of substrate to achieve half of Vmax1 for high-affinity phase, and Km2 is concentration of substrate to achieve half of Vmax2 for low-affinity phase.

Analysis and Identification of CG and its Metabolites.

All samples were analyzed by using Agilent 6460 Triple Quadrupole mass system and 6540 accurate-mass Q-TOF MS system (Santa Clara, CA). The conditions were as follows: column, Zorbax SB-C18, 1.8 μm, 3.0 × 100 mm; mobile phase A, 100% aqueous buffer (0.1%, v/v formic acid, pH 2.5), mobile phase B, 100% acetonitrile; flow rate, 0.3 ml/min; gradient, 0–0.5 minutes, 15–15% B, 0.5–4 minutes, 15–60% B, 4–4.5 minutes, 60–90% B, 4.5–6 minutes, 90–90% B, 6–6.5 minutes, 90–15% B; and injection volume, 1 μl. Electrospray ionization was performed in positive mode.

The main mass working parameters for the mass spectrometers were set as follows: capillary voltage 3.5 kV, nozzle voltage 500 V, gas temperature 280°C, and sheath temperature 380°C with sheath gas flow of 11 l/min and gas flow of 5 l/min,. Data were recorded and the system was controlled using the MassHunter software (version B.06.00, Agilent Technologies).

This LC-MS/MS method has been validated. In plasma, the intraday precision relative standard deviation (RSD) of analysts ranged between 1.41 and 11.2%, and accuracy relative error (RE) ranged between Embedded Image12.86 and 7.92%. The interday precision (RSD) of analysts ranged between 4.89 and 9.21%, and accuracy (RE) ranged between Embedded Image11.06 and 5.43%. The recovery ranged from 91.4 to 112.39% and the matrix effects were between 52.9 and 78.4%. In bile, the intraday precision (RSD) of analysts ranged between 5.46 and 10.9%, and accuracy (RE) ranged between Embedded Image8.44 and 10.89%. The interday precision (RSD) of analysts ranged between 3.32 and 8.97%, and accuracy (RE) ranged between Embedded Image9.61 and 3.28%. The recovery ranged from 89.7 to 99.01%, and the matrix effects were between 74.3 and 89.2%. In perfusate, the intraday precision (RSD) of analysts ranged between 1.96 and 12.33%, and accuracy (RE) ranged between Embedded Image6.77 and 13.89%. The interday precision (RSD) of analysts ranged between 5.28 and 10.78%, and accuracy (RE) ranged between Embedded Image10.11 and 5.87%. The recovery ranged from 91.4 to 102.41% and the matrix effects were between 87.3 and 92.4%.

The position of glucuronic acid substitution of calycosin was determined through 1HNMR spectroscopy. All readings were recorded on Bruker AV-500 spectrophotometer (Bruker, Faellanden, Switzerland). The purified major metabolite of calycosin (M1, 2.5 mg) was stored at −20°C before dissolution in deuterated DMSO for 1HNMR analysis. Chemical shifts were given on the δ scale and referenced to tetramethylsilane at 0 ppm for 1HNMR (500 MHz).

Statistical Analysis.

One-way analysis of variance with or without Tukey-Kramer multiple comparison (post hoc) tests were used to evaluate statistical difference. Student’s t test was also performed. Differences were considered significant at P values <0.05.

Results

Analysis and Identification of CG and its Metabolites.

The LC-MS/MS method for CG and its metabolites exhibited a run time of 6.5 minutes (Fig. 1). The tested linear response range was 0.04–10 μM (for nine concentrations), with a lower limit of quantification (LLOQ) of 0.02 μM.

Fig. 1.
Fig. 1.

Representative multiple reaction monitoring (MRM) chromatograms and structures of CG, calycosin, and their respective glucuronides. LC-MS/MS was used to separate and quantify CG and its metabolites in the experimental samples. (A–D) represent the MRM chromatograms showing the retention time and structure of CG, calycosin, and their glucuronides in perfusate, bile, plasma, and incubation system samples.

The retention time of CG, CG glucuronide, and calycosin were 4.13, 3.12, and 5.98 minutes in the samples (Fig. 1, A–C) respectively. CG concentration in the bile sample was lower than the LLOQ of the LC-MS/MS analysis method. Calycosin was not detected in the bile and plasma samples. There were two monoglucuronides of calycosin were detected in the incubation system (Fig. 1D), which was consistent with previous reports (Ruan and Yan, 2014; Ruan et al., 2015). The obtained monoglucuronides were eluted at 4.61 (M1) and 4.19 minutes (M2). However, in the perfusate, bile, and plasma samples, only one monoglucuronide of calycosin was detected. High-resolution mass spectrometry was performed to determine the MS spectrum of the metabolite. The CG glucuronide showed a pseudomolecule ion [M+H]+ of m/z 623.1600 in full-scan mass spectra and fragment ions at m/z 447.1 and 285.2; this finding indicates that this glucuronide exhibited the molecular formula of C28H30O16 (or CG-7-glucuronide). M1 and M2 showed a pseudomolecule ion [M+H]+ of m/z 461.1074 in full-scan mass spectra and fragment ions at m/z 285.0; hence, these monoglucuronides demonstrated a molecular formula of C22H20O11 (or calycosin-3′-glucuronide or calycosin-7-glucuronide). The substitution position of the proton of 3′-OH was confirmed through nuclear overhauser effect spectroscopy [correlation between Glu-H-1′′ (δH 5.08, 1H, d, J=7.0 Hz) and H-2′ (δH 7.28, 1H, d, J=2.0 Hz)] (Supplemental Figs. S1–S3). Therefore, the structure of M1, which was found to be the major metabolite of calycosin, was unambiguously identified as calycosin-3′-glucuronide. As such, M2 was tentatively assigned as calycosin-7-glucuronide.

Stability of CG and Its Metabolites.

The stability of CG and its metabolites in the perfusate buffer, blank bile, and blank plasma was tested to ensure that CG metabolism is not due to chemical instability. The results illustrated that the stabilities of the analytes in the blank perfusate and bile samples showed 90–110% recoveries after incubation at 37°C for 1, 2, 3, and 4 hours (Supplemental Tables S1 and S2).The stability of CG in the rat blank plasma also showed no less than 85% recovery after the following storage conditions: 25°C for 6 hours, 37°C for 6 hours, −80°C for 15 days, three freeze-thaw cycles (−20°C and 25°C), and 5°C for 12 hours (Supplemental Table S3). These results demonstrate that CG and its metabolites were stable in the perfusate buffer, blank bile, and blank plasma.

In Vivo Pharmacokinetic Studies of CG and Its Metabolites in Rats.

The mean plasma concentration-time profiles of CG and its metabolites after intravenous (0.5 mg/kg), intraperitoneal (10 mg/kg), and oral (10 mg/kg) administrations of CG are shown in Fig. 2. Pharmacokinetic parameters were calculated using the noncompartmental method with WinNonlin 3.3 (Pharsight, Mountain View, CA), and the results are presented in Table 1.

Fig. 2.
Fig. 2.

The pharmacokinetic profile of CG and its metabolites after oral (10 mg/kg) (A), intraperitoneal (10 mg/kg) (B), and intravenous (0.5 mg/kg) (C) administration. The plasma samples were analyzed by UPLC-MS/MS. The plasma concentration-time curves of CG and its metabolites were obtained by plotting the mean plasma concentration versus time. Each data point is the average of six determinations, and the error bar represents the standard deviation of the mean (n = 6).

View this table:
TABLE 1

Pharmacokinetic parameters of CG and its metabolites in rat plasma after intravenous (0.5 mg/kg), intraperitoneal (10 mg/kg), and oral (0 mg/kg) administration of CG (n = 6)

After oral administration of CG, the Cmax and AUC0-∞ values of calycosin-3′-glucuronide were 1189.66 ± 346.95 ng/ml and 111.16±34.18 min·μg/ml, respectively, which were much higher than those of CG glucuronide and CG. After intraperitoneal administration, CG exhibited the highest Cmax (965.24 ± 133.53 ng/ml), followed by CG glucuronide and calycosin-3′-glucuronide. However, the AUC0–∞ of CG (72.64±22.08 min·μg/ml) was not the highest because the elimination of CG was very fast (t1/2 = 77.5 ± 17.91 minutes), and its plasma concentration was lower than 10 ng/ml after 4 hours. The AUC0-∞ of calycosin-3′-glucuronide was 99.11±45.90 min·μg/ml. After intravenous administration of CG, the plasma concentrations of CG glucuronide and calycosin-3′-glucuronide were considerably low, and the AUC0-∞ of CG was 5.39±1.02 min·μg/ml. Thus, the absolute oral bioavailability of CG was 0.304%. Moreover, calycosin was not detected in the plasma after CG administrations.

Regional Transport and Metabolism of CG in a Perfused Rat Intestinal Model.

After perfusion with 10 and 2.5 μM CG, it was rapidly absorbed in the duodenum, jejunum, and ileum, with rates higher than those in the colon. Moreover, P*eff values of 2.5 μM in all intestinal segments were higher than those of 10 μM, but the difference was not significant (Fig. 3A). CG was converted into calycosin-3′-glucuronide and calycosin in the rat intestine and large amounts of calycosin-3′-glucuronide were detected in the perfusate sample, with the highest excretion in the jejunum and the lowest in the colon (Fig. 3C). In addition, low amounts of CG glucuronide were detected in the perfusate sample, but the concentration was lower than the LLOQ of the analytical method (0.02 μM). In the bile sample, the glucuronides of calycosin and CG were detected, and the concentration of calycosin-3′-glucuronide was higher than that of CG glucuronide. Moreover, calycosin and CG were not detected in the bile sample. Calycosin-3′-glucuronide, CG, and CG glucuronide were detected in the plasma sample, with calycosin-3′-glucuronide having the highest concentration. By contrast, calycosin was not detected in the plasma. Furthermore, CG sulfate was not detected in the perfusate, bile, and plasma samples.

Fig. 3.
Fig. 3.

Effect of different inhibitors on the absorption and metabolism of CG and daidzin in perfusate samples. Effective permeabilities of CG or daidzin; amounts absorbed of CG or daidzin; and amounts of glucuronide, of calycosin, or daidzein excreted in the perfusate samples after CG was perfused alone at the concentration of 10 and 2.5 μM (A–C), CG was perfused alone or coperfused with GCL and glucose (D–F), daidzin was perfused alone or coperfused with GCL and glucose (G–I), CG was perfused alone or coperfused with phloridzin (J–L), CG was perfused alone or coperfused with phloretin (M–O), and CG was perfused alone or coperfused with MK571 and ko143 (P–R) were determined and normalized over a 10 cm intestinal length. Each column represents the average of four determinations, and the error bar is the S.D. (n = 4). *Statistically significant difference at P < 0.05; **P < 0.01.

Effect of LPH and Glucose in Absorption of CG and Daidzin.

The LPH inhibitor glucurolactone (20 and 40 mM) was used to inhibit the hydrolysis of CG and daidzin to their corresponding aglycones and to determine the effect of this mechanism on absorption and metabolism. Surprisingly, glucurolactone had no influence on CG absorption and calycosin-3′-glucuronide excretion in the perfusate sample at two concentrations (Fig. 3, D–F). In addition, the concentrations of CG and its metabolites did not change in the bile (Fig. 4, C and D) and plasma (Fig. 5, D–F) samples. By contrast, glucurolactone concentration dependently reduced the absorption of daidzin, which was selected as the positive control, and the excretion of daidzin glucuronide in the perfusate sample (Fig. 3, G–I). The effect of glucurolactone on the excretion of daidzein glucuronide was more remarkable in the upper small intestine than that in the colon, as evidenced by the higher LPH quantity in the small intestine. In addition, daidzein glucuronide concentration was significantly reduced in the bile (Fig. 4E) and plasma (Fig. 5G) samples. By contrast, daidzin concentration in the plasma (Fig. 5H) was significantly enhanced in the presence of glucurolactone.

Fig. 4.
Fig. 4.

Effect of different inhibitors on the concentration of metabolites of CG and daidzin in bile samples. The concentrations of glucuronide, of calycosin, or daidzein; glucuronide of CG in the bile samples after CG was perfused alone at the concentration of 10 and 2.5 μM (A and B), CG was perfused alone or coperfused with GCL and glucose (C and D), daidzin was perfused alone or coperfused with GCL and glucose (E), CG was perfused alone or coperfused with phloridzin (F and G), CG was perfused alone or coperfused with phloretin (H and I), and CG was perfused alone or coperfused with MK571 and ko143 (J and K) were determined. Each curve represents the average of four determinations, and the error bar is the S.D. (n = 4). *Statistically significant difference at P < 0.05; **P < 0.01.

Fig. 5.
Fig. 5.

Effect of different inhibitors on the concentration of metabolites of CG and daidzin in plasma samples. The concentrations of glucuronide, of calycosin, or daidzin; CG or daidzin; and glucuronide of CG in the plasma samples after CG was perfused alone at the concentration of 10 and 2.5 μM (A–C), CG was perfused alone or coperfused with GCL and glucose (D–F), daidzin was perfused alone or coperfused with GCL and glucose (G and H), CG was perfused alone or coperfused with phloridzin (I–K), CG was perfused alone or coperfused with phloretin (L–N), and CG was perfused alone or coperfused with MK571 and ko143 (O–Q) were determined. Each curve represents the average of four determinations, and the error bar is the S.D. (n = 4). *Statistically significant difference at P < 0.05; **P < 0.01.

Glucose, which can saturate GLTs, reduced CG absorption and calycosin-3′-glucuronide excretion in the perfusate samples of the upper small intestine (duodenum, jejunum, and ileum) (Fig. 3. D–F). Glucose did not affect CG absorption in the colon but significantly reduced the concentration of CG and its metabolites in the bile and plasma samples; this finding was consistent with reduced absorption in the perfusate sample (Fig. 4, C and D, and Fig. 5, D–F). Glucurolactone (40 mM) was also added to the perfusate with glucose to further determine the effect of LPH on CG. Similarly, glucurolactone played minimal or no role on CG absorption and metabolism in the rat intestine (Fig. 3, D–F; Fig. 4, C and D; and Fig. 5, D–F). As the positive control, the absorption of daidzin and the excretion of its metabolite significantly decreased in the perfusate sample added with glucose (Fig. 3, G–I). The concentrations of daidzein glucuronide and daidzin in the bile and plasma samples also decreased in the presence of glucose in the perfusate sample (Fig. 4E and Fig. 5, G–H). Moreover, glucurolactone reduced daidzin absorption and daidzein glucuronide excretion in the perfusate sample containing glucose (Fig. 3, G–I), which is consistent with previous reports (Liu and Hu, 2002). Daidzein glucuronide concentration significantly decreased in the bile (Fig. 4E) and plasma samples (Fig. 5G), and daidzin concentration significantly increased in the plasma sample (Fig. 5H). These results demonstrated that LPH played minimal to no role on CG hydrolysis.

Importance of GLTs in CG Absorption.

SGLT-1 and GLUT-2 are main intestinal GLTs (Zou et al., 2014). The SGLT-1 inhibitor phloridzin and the GLUT-2 inhibitor phloretin were used to determine the effects of GLTs on CG absorption and metabolism. After perfusion with phloridzin (100, 200, and 400 μM), CG absorption and calycosin-3′-glucuronide excretion were significantly and concentration dependently reduced in the perfusate sample in duodenum, jejunum, and ileum (Fig. 3, J–L). Similar to glucose, phloridzin also played no role on CG absorption in the colon, as evidenced by the absence of SGLT-1 expression in the rat colon (Balen et al., 2008). Meanwhile, the concentrations of CG and its metabolites in the bile and plasma samples significantly decreased in the presence of phloridzin in the perfusate sample, which was consistent with reduced absorption in the perfusate sample (Fig. 4, F and G, and Fig. 5, I and J). Moreover, CG glucuronide concentration was moderately reduced in the plasma (Fig. 5K), but the effect was not statistically significant. By contrast, phloretin (50, 100, and 200 μM) had no evident effect on CG absorption and its metabolite excretion in the perfusate sample of each intestinal segment at different concentrations (Fig. 3, M–O). In the bile (Fig. 4, H and I) and plasma (Fig. 5, L–N) samples, the concentrations of analytes remained constant in the presence of phloretin in the perfusate sample at different concentrations.

Role of Efflux Transporters on Absorption and Excretion of CG and Its metabolites.

After perfusion with ko143 (50 μM), the inhibitor of BCRP, and MK571 (50 μM), the inhibitor of MRP2, CG absorption was enhanced in the perfusate sample of duodenum and jejunum (Fig. 3, P–R), but the increase was not statistically significant. The excretion of calycosin-3′-glucuronide significantly reduced in all four segments of the intestine (Fig. 3R). Moreover, the concentrations of calycosin-3′-glucuronide and CG glucuronide in the bile significantly increased in the presence of ko143 and MK571 in the perfusate (Fig. 4, J and K). By contrast, the concentrations of calycosin-3′-glucuronide, CG, and CG glucuronide only slightly (but not significantly) increased in the plasma samples (Fig. 5, O–Q).

Effects of Inhibitors of BCRP and MRP2 on the Efflux of Calycosin Glucuronide in Caco-2 Cells.

Bidirectional calycosin 5 μM (positive control) transport across Caco-2 cells was studied in the presence of ko143 (10 μM) and MK571 (10 μM). In the presence of 10 μM ko143 or 10 μM MK571, the efflux rates of calycosin-3′-glucuronide from the cells to the AP side and BL was decreased in the bidirectional transport studies (Fig. 6, A and B). Similarly, ko143 and MK571 showed inhibitory effects on the CL of calycosin glucuronide from cells to the AP side (71% for A–B, 64% for B–A by MK571; 54% for A–B, 57% for B–A by ko143) or the BL side (52% for A–B, 54% for B–A by MK571; 53% for A–B, 33% for B–A by ko143, P < 0.05) in the bidirectional transport studies (Fig. 6, C and D). On the other hand, the fmet values were significantly lower (P < 0.05) and the intracellular concentrations of calycosin glucuornide were significantly higher in the presence of ko143 or MK571 in both A–B and B–A directional transports (Fig. 6, E and F).

Fig. 6.
Fig. 6.

Effects of inhibitors on the rates of calycosin glucuronide efflux from cellular membrane to the AP side (A) or BL side (B), on the clearance of calycosin glucuronide at AP side (C) or BL side (D) of the Caco-2 cell monolayer, on fmet from the cellular membrane to the AP side or BL side (E), and on the intracellular concentration of calycosin glucuronide (F). Calycosin (5 μM) without inhibitors (open bar), with the MRP2 inhibitor MK571 at 10 μM (gray bar), and with the BCRP inhibitor ko143 (black bar) was added to the AP side (A–B) or BL side (B–A). The inhibitors were added to BL side. *Statistically significant difference at P < 0.05.

Effects of Chemical Inhibitors on the Hydrolysis of CG in S9 of Duodenum, Jejunum, and Ileum.

The effect of LPH on the brush border enterocytes on CG hydrolysis was negligible. The hydrolytic site of CG is likely to be inside the enterocytes. Intestinal S9 was considered as the model for intestinal deconjugation study (Islam et al., 2014). Moreover, intracellular glucosidases were found in intestinal S9. Therefore, the inhibitory effects of gluconolactone and conduritol B epoxide on CG hydrolysis in intestinal S9 of duodenum (Fig. 7, A and B), jejunum (Fig. 7, C and D), and ileum (Fig. 7, E and F) were investigated to confirm the effect of glucocerebrosidase and BSβG on CG hydrolysis in vitro (Day et al., 1998; Islam et al., 2014). In the S9 of duodenum, jejunum, and ileum, gluconolactone significantly inhibited calycosin formation at three CG concentrations. Three concentrations of gluconolactone also inhibited more than 75% of CG hydrolysis at 25 μM. Similar results were also observed in the inhibition of conduritol B epoxide on CG hydrolysis. However, the inhibitory effect of conduritol B epoxide was weaker than that of gluconolactone.

Fig. 7.
Fig. 7.

Effects of chemical inhibitors of BSβG and glucocerebrosidase on the hydrolyzation of CG in S9 of duodenum (A and B), jejunum (C and D), and ileum (E and F). Hydrolyzation of CG by S9 were measured at three concentrations (25, 100, and 400 μM) and incubated in the presence of inhibitors at 2.5, 5, and 10 mM for gluconolactone and 1.25, 2.5, and 5 mM for conduritol B epoxide. All protein concentrations were 0.1 mg/ml, incubation times for CG in S9 ranged from 15 to 45 minutes. Control experiments were incubated without chemical inhibitors. All incubations were performed in triplicate. Each column represents mean percentage control of formation rates of these three metabolites, and error bars are standard deviations of the mean. *Statistically significant difference at P < 0.05.

Kinetics of Calycosin and CG Glucuronidation by RLMs and RIMs.

Glucuronidation rates of calycosin and CG in RLMs and RIMs were determined using different substrate concentrations. Within the tested concentration ranges, 3′-glucuronide (M1) was found to be the major metabolite of calycosin in RLMs and RIMs (Fig. 8, A and B); hence, RLMs and RIMs are preferential sites for catalysis of calycosin glucuronidation at 3′-OH rather than 7-OH (M2). Both RLMs and RIMs, which mediated M1 glucuronidation, exhibited classic Michaelis-Menten kinetic characteristics. Furthermore, M2 exhibited biphasic kinetic profile (Fig. 8, A and B), as evidenced by the Eadie-Hofstee plot (Fig. 8, a–d). In addition, only CG glucuronide was formed in RLMs and RIMs. According to the Eadie-Hofstee plot (Fig. 8, e and f ), CG glucuronide formations by RLMs and RIMs followed classic Michaelis-Menten kinetic characteristics and substrate inhibition profile, respectively (Fig. 8, C and D). All kinetic parameters of the glucuronidation of calycosin and CG are shown in Table 2.

Fig. 8.
Fig. 8.

Kinetics of calycosin glucuronidation by RLMs (A), RIMs (B), and CG glucuronidation by RLMs (C) and RIMs (D) (n = 3). The glucuronidation rates were measured at substrate concentration ranging from 1.25 to 200 μM (calycosin) and 10 to 800 μM (CG). The reaction time and protein concentration were controlled so that substrate concentration did not decrease substantially (usually less than 30%) at the end of experiments, which lasted for of up to 15 minute. Each data point is the average of three determinations, and the error bar represents the standard deviation of the mean. The data points in figures are observed glucuronidation rates, and the curves are plotted using the fitted kinetic parameters (Table 2).

View this table:
TABLE 2

Kinetic parameters of calycosin and CG glucuronidation obtained from RLMs and RIMs

The intrinsic clearance (CL, Vmax/Km) of M1 was much higher than that of M2 in RLMs and RIMs (M1, CL 2570.4 ml/min/mg for RLMs, CL 2955.74 ml/min/mg for RIMs; M2, CL 304.35 ml/min/mg for RLMs, CL 57.14 ml/min/mg for RIMs). RIMs exhibited lower Km and higher Vmax values for the formation of CG glucuronide than RLMs. As such, the CL of CG, through glucuronidation in RIMs, was about 30 times higher than that in RLMs (1846.39 ml/min/mg for RIMs and 67.54 ml/min/mg for RLMs).

Discussion

In this study, we found that CG and its metabolites rapidly appeared in the plasma sample after oral administration, which indicated that CG was not hydrolyzed by LPH and crossed the enterocytes intact (Fig. 2). This result differed from previous reports, which showed that hydrolysis by LPH was a critical first step in the disposition of isoflavonoid glucosides. We used the rat intestinal perfusion model, which has been demonstrated to be suitable for studying the role of LPH (Dai et al., 2015), and found that LPH inhibitor played minimal role in CG absorption and metabolism in the intestine (Fig. 3,D–F; Fig. 4, C and D; and Fig. 5, D–F). Daidzin, the substrate of LPH and GLTs, was chosen as positive control to ensure that the results were reliable. Consistent with a previous report (Liu and Hu, 2002), absorption and metabolism of daidzin were affected by LPH inhibitor (Fig. 3, G–I; Fig. 4E; and Fig. 5, G and H). In addition, to determine whether GLTs affected the role of LPH on the absorption of CG, we added glucose into perfusate buffer to saturate GLTs. We found that CG could not be hydrolyzed by LPH after the GLTs saturated (Fig. 3, D–F; Fig. 4, C and D; and Fig. 5, D–F). However, the absorption and metabolism of daidzin were affected by LPH inhibitor after the GLTs saturated (Fig. 3, G–I; Fig. 4E; Fig. 5, G and H). Thus, the effect of LPH was not influenced after the GLTs saturated. These results suggest that CG may not be hydrolyzed by LPH.

After oral administration of CG, calycosin-3′-glucuronide was found to be the main form circulating in plasma (Fig. 2). Hence, deglycosylation of CG to calycosin was an important step before entry into systemic circulation. Because small intestine and liver can hydrolyze isoflavonoid glucosides (Day et al., 1998), we studied the pharmacokinetics of CG after intraperitoneal administration to avoid intestinal disposition. The results showed that CG rapidly entered systemic circulation and became the main circulating component in the plasma, thereby limiting hydrolysis in the liver (Fig. 2). The gut microbiota has also been thought to hydrolyze glucosides (Bokkenheuser et al., 1987). It was reported that rutin was not absorbed until it was hydrolyzed by gut microbiota, and the absorption peak of its metabolites occurred at 7 hours (He et al., 2013). However, we found that, after oral administration, CG and its metabolites were rapidly absorbed, with an absorption peak at 45 minutes (Fig. 2). This indicated that the role of gut microbiota in CG hydrolysis could be significantly small. Therefore, the site of CG hydrolysis is very likely to be the upper small intestine.

CG cannot easily cross cellular membranes through diffusion because of its large molecular weight. However, we found that the Peff values of CG in the small intestine were ∼1.5 (Fig. 3, A–C; Fig. 4, A and B; and Fig. 5, A–C). Generally, Peff values exceeding 1.0 are considered to indicate good absorption (Johnson and Amidon, 1988). Furthermore, glucose, which exhibits high affinity with GLTs, significantly reduced the absorption and metabolism rates of CG in the intestine, and the concentrations of CG and its metabolites in bile and plasma samples (Fig. 3, D–F; Fig. 4, C and D; and Fig. 5, D–F). Therefore, the most likely carrier molecule to facilitate intestinal absorption of CG could be GLTs, which could transport some flavonoid glucosides (Walgren et al., 2000). By using specific inhibitors of GLTs, our results showed that CG transport from the lumen to enterocytes was mediated by SGLT-1 (Fig. 3, J–L; Fig. 4, F and G; and Fig. 5, I–K), and the role of GLUT-2 was negligible (Fig. 3, M–O; Fig. 4, H and I; and Fig. 5, L–N). Additionally, Peff values of CG were the highest in the jejunum and the lowest in the colon (Fig. 3A–C). This finding could be related to the distribution of SGLT-1 in the intestine. Previous studies reported that SGLT-1 expression levels in the small intestine of rats exhibited the following order: jejunum > duodenum = ileum but was not observed in the colon (Balen et al., 2008). Therefore, the main absorption site of CG could be the upper small intestine, which should be considered in the design of CG oral preparations.

Given that CG could be hydrolyzed in enterocytes, we used intestinal (duodenum, jejunum, and ileum) S9, which is an appropriate model to study intestinal deconjugation because it contains all cytosolic and membrane-bound enzymes, and inhibitors of BSβG and glucocerebrosidase (Islam et al., 2014), to identify the enzyme that hydrolyzes CG. Conduritol B epoxide, the specific inhibitor of glucocerebrosidase, showed complete inhibition at 5 mM (Day et al., 1998). Our results showed that conduritol B epoxide inhibited CG hydrolysis in the intestinal S9 in a dose-dependent manner (Fig. 7, B, D, and F). In contrast to conduritol B epoxide, gluconolactone was not considered a specific inhibitor; this compound completely inhibited BSβG at 10 mM and exhibited weak inhibitory effect on glucocerebrosidase (Islam et al., 2014). Gluconolactone also showed concentration-dependent effects on CG. This result indicated that CG may be hydrolyzed by BSβG and glucocerebrosidase. However, the kinetics and inhibition mechanisms of GLT inhibitors remain unclear because we did not obtain pure enzymes of BSβG and glucocerebrosidase. Thus, identifying the mechanisms by which conduritol B epoxide and gluconolactone exert inhibitory effects requires further studies, and in vitro data on their inhibition should be interpreted with caution.

MRP2 and BCRP are important transporters for the efflux of glucuronides (Jiang et al., 2012). We found that the amount of calycosin-3′-glucuronide in the perfusate samples significantly decreased in the presence of MK571 and ko143 (Fig. 3R). Furthermore, experiments conducted on Caco-2 cells showed that MK571 and ko143 could decrease the efflux rates, CL and fmet, and increase the intracellular concentrations of calycosin-3′-glucuronide (Fig. 6). Taken together, these results showed that MRP2 and BCRP were involved in the excretion of calycosin-3′-glucuronide. However, unlike MK571 and ko143 that increased the concentration of metabolites of CG in the bile and plasma samples (Fig. 4, J and K, and Fig. 5, O–Q), these inhibitors reduced the efflux rates and CL of calycosin-3′-glucuronide from Caco-2 cells to the BL side (Fig. 6). A possible reason for this could be that there was no mucus covering the Caco-2 cell; therefore the inhibitors could reach the BL side to inhibit the efflux (Liu et al., 2012). As a result, after BCRP and MRP2 were inhibited, most of calycosin-3′-glucuronide was transported to the bile and plasma in the intestinal perfusion model.

The results of the kinetic assays showed that the glucuronidation rate of calycosin at 3′-OH was higher than that at 7-OH in RLMs and RIMs. Moreover, the glucuronidation rate of calycosin at 3′-OH in RIMs was twice as high as that in RLMs (Fig. 8). Rapid glucuronidation at 3′-OH could be the reason why only calycosin-3′-glucoronide was detected in the perfusate, bile, and plasma samples, and aglycone calycosin was not detected in the plasma sample. Similar to the case of calycosin, the glucuronidation rate of CG in RIMs was higher than that in RLMs (Fig. 8) and the CG glucuronidation by RIMs showed substrate inhibition kinetics. Such an atypical kinetics profile has been described as a two-site model in which one binding site is productive, whereas the other one is inhibitory at high substrate concentration, resulting in decreased velocity with increased substrate concentration (Hutzler and Tracy, 2002). It has been reported that isoflavonoids, including calycosin, can be metabolized by multiple UGT enzymes (Tang et al., 2009; Ruan and Yan, 2014), and the rates and profiles of metabolism by individual UGT enzymes are different. Distribution of these enzymes in RLMs and RIMs are diverse and the rates and profiles of isoflavonoids glucuronidation in microsomes are determined by UGT enzymes (Zhou et al., 2010), which could explain why the profile of CG glucuronidation in RIMs was different from that of calycosin glucuronidation in RLMs and RIMs and of CG glucuronidation in RLMs. Although the liver is the main metabolic organ and several studies have focused on calycosin metabolism at this site (Ruan et al., 2015), our results indicate that, as the first site of metabolism after oral administration, more attention should be paid to the importance and characteristics of metabolism of calycosin and CG in the intestine.

In conclusion, the present study describes the characteristics of CG absorption and disposition in the rat intestine. We found that the intestine could be the main organ responsible for the hydrolysis and first-pass metabolism of CG. In contrast to the case of other structurally analogous isoflavonoid glucosides such as daidzin, hydrolysis by intestinal LPH, which was thought to be capable of hydrolyzing various flavonoids and isoflavonoid glucosides, was not necessary for CG absorption. This study is the first to determine that isoflavonoid glucosides can be transported by SGLT-1 into enterocytes using the rat intestinal perfusion model. We also demonstrated that CG was likely to be hydrolyzed by BSβG and glucocerebrosidase in enterocytes. Thus, this study reveals a novel mechanism for the absorption and disposition of isoflavonoid glucosides in the small intestine and elucidates the underlying mechanism. Our findings contribute to improved understanding of the absorption and disposition of flavonoid compounds, which is essential for determining their potential as chemopreventive agents for LPH-deficient individuals.

Authorship Contributions

Participated in research design: Shi, Hu, and Liu.

Conducted experiments: Shi, Zheng, and Yu.

Contributed new reagents or analytic tools: Zhu and Lu.

Performed data analysis: Shi, Yan, Wu, and Wang.

Wrote or contributed to the writing of the manuscript: Shi, Hu, and Liu.

Footnotes

    • Received August 31, 2015.
    • Accepted December 10, 2015.

  • This work was financially supported by the grants of National Natural Science Foundation of China [Grant No. 81120108025], Science and Technology Project of Guangzhou City [Grant No. 201509010004] and Guangdong Natural Science Foundation [Grant No. 2015A030312012].

  • dx.doi.org/10.1124/dmd.115.067009

  • ↵Embedded ImageThis article has supplemental material available at (dmd.aspetjournals.org).

Abbreviations

AP
apical
BCRP
breast cancer resistance protein
BL
basolateral
BSβG
broad-specific β-glucuronides
CG
calycosin-7-O-β-d-glucoside
GLTs
glucose transporters
GLUT-2
glucose transporter 2
HBSS
Hanks' balanced salt solution
IS
internal standard
LLOQ
lower limit of quantification
LPH
lactase-phloridzin hydrolase
MRM
multiple reaction monitoring
MRP2
multidrug resistance-associated protein 2
RE
relative error
RIMs
rat intestinal microsomes
RLMs
rat liver microsomes
RSD
relative standard deviation
SGLT-1
sodium-dependent glucose transporter 1
UGT
UDP-glucuronosyltranferases
UPLC-MS/MS
ultraperformance liquid chromatography tandem mass spectrometry

References

  1. ↵
    1. Balen D,
    2. Ljubojevic M,
    3. Breljak D,
    4. Brzica H,
    5. Zlender V,
    6. Koepsell H, and
    7. Sabolic I
    (2008) Revised immunolocalization of the Na+-D-glucose cotransporter SGLT1 in rat organs with an improved antibody. Am J Physiol Cell Physiol 295:C475–C489.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Bokkenheuser VD,
    2. Shackleton CH, and
    3. Winter J
    (1987) Hydrolysis of dietary flavonoid glycosides by strains of intestinal Bacteroides from humans. Biochem J 248:953–956.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Cai Z,
    2. Huang J,
    3. Luo H,
    4. Lei X,
    5. Yang Z,
    6. Mai Y, and
    7. Liu Z
    (2013) Role of glucose transporters in the intestinal absorption of gastrodin, a highly water-soluble drug with good oral bioavailability. J Drug Target 21:574–580.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Chen J,
    2. Lin H, and
    3. Hu M
    (2005) Absorption and metabolism of genistein and its five isoflavone analogs in the human intestinal Caco-2 model. Cancer Chemother Pharmacol 55:159–169.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Choi SI,
    2. Heo TR,
    3. Min BH,
    4. Cui JH,
    5. Choi BH, and
    6. Park SR
    (2007) Alleviation of osteoarthritis by calycosin-7-O-beta-D-glucopyranoside (CG) isolated from Astragali radix (AR) in rabbit osteoarthritis (OA) model. Osteoarthritis Research Society 15:1086–1092.
    OpenUrlCrossRef
  6. ↵
    1. Dai P,
    2. Zhu L,
    3. Luo F,
    4. Lu L,
    5. Li Q,
    6. Wang L,
    7. Wang Y,
    8. Wang X,
    9. Hu M, and
    10. Liu Z
    (2015) Triple recycling processes impact systemic and local bioavailability of orally administered flavonoids. AAPS J 17:723–736.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Daniels LB,
    2. Coyle PJ,
    3. Chiao YB,
    4. Glew RH, and
    5. Labow RS
    (1981) Purification and characterization of a cytosolic broad specificity beta-glucosidase from human liver. J Biol Chem 256:13004–13013.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Day AJ,
    2. Cañada FJ,
    3. Díaz JC,
    4. Kroon PA,
    5. Mclauchlan R,
    6. Faulds CB,
    7. Plumb GW,
    8. Morgan MR, and
    9. Williamson G
    (2000) Dietary flavonoid and isoflavone glycosides are hydrolysed by the lactase site of lactase phlorizin hydrolase. FEBS Lett 468:166–170.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Day AJ,
    2. DuPont MS,
    3. Ridley S,
    4. Rhodes M,
    5. Rhodes MJ,
    6. Morgan MR, and
    7. Williamson G
    (1998) Deglycosylation of flavonoid and isoflavonoid glycosides by human small intestine and liver beta-glucosidase activity. FEBS Lett 436:71–75.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Day AJ,
    2. Gee JM,
    3. DuPont MS,
    4. Johnson IT, and
    5. Williamson G
    (2003) Absorption of quercetin-3-glucoside and quercetin-4′-glucoside in the rat small intestine: the role of lactase phlorizin hydrolase and the sodium-dependent glucose transporter. Biochem Pharmacol 65:1199–1206.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Del Rio D,
    2. Rodriguez-Mateos A,
    3. Spencer JP,
    4. Tognolini M,
    5. Borges G, and
    6. Crozier A
    (2013) Dietary (poly)phenolics in human health: structures, bioavailability, and evidence of protective effects against chronic diseases. Antioxid Redox Signal 18:1818–1892.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Fan Y,
    2. Wu DZ,
    3. Gong YQ,
    4. Zhou JY, and
    5. Hu ZB
    (2003) Effects of calycosin on the impairment of barrier function induced by hypoxia in human umbilical vein endothelial cells. Eur J Pharmacol 481:33–40.
    OpenUrlCrossRefPubMed
  13. ↵
    1. He J,
    2. Feng Y,
    3. Ouyang HZ,
    4. Yu B,
    5. Chang YX,
    6. Pan GX,
    7. Dong GY,
    8. Wang T, and
    9. Gao XM
    (2013) A sensitive LC-MS/MS method for simultaneous determination of six flavonoids in rat plasma: application to a pharmacokinetic study of total flavonoids from mulberry leaves. J Pharm Biomed Anal 84:189–195.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Hu M,
    2. Chen J,
    3. Tran D,
    4. Zhu Y, and
    5. Leonardo G
    (1994) The Caco-2 cell monolayers as an intestinal metabolism model: metabolism of dipeptide Phe-Pro. J Drug Target 2:79–89.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Hutzler JM and
    2. Tracy TS
    (2002) Atypical kinetic profiles in drug metabolism reactions. Drug Metab Dispos 30:355–362.
    OpenUrlFREE Full Text
  16. ↵
    1. Islam MA,
    2. Punt A,
    3. Spenkelink B,
    4. Murk AJ,
    5. Rolaf van Leeuwen FX, and
    6. Rietjens IM
    (2014) Conversion of major soy isoflavone glucosides and aglycones in in vitro intestinal models. Mol Nutr Food Res 58:503–515.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Jia X,
    2. Chen J,
    3. Lin H, and
    4. Hu M
    (2004) Disposition of flavonoids via enteric recycling: enzyme-transporter coupling affects metabolism of biochanin A and formononetin and excretion of their phase II conjugates. J Pharmacol Exp Ther 310:1103–1113.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Jian L,
    2. Xin L,
    3. Yufang M, and
    4. Yifan H
    (2015) Protective effect of calycosin-7-O-β-D-glucopyranoside against oxidative stress of BRL-3A cells induced by thioacetamide. Pharmacogn Mag 11:524–532.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Jiang W,
    2. Xu B,
    3. Wu B,
    4. Yu R, and
    5. Hu M
    (2012) UDP-glucuronosyltransferase (UGT) 1A9-overexpressing HeLa cells is an appropriate tool to delineate the kinetic interplay between breast cancer resistance protein (BRCP) and UGT and to rapidly identify the glucuronide substrates of BCRP. Drug Metab Dispos 40:336–345.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Jeong EJ,
    2. Jia X, and
    3. Hu M
    (2005) Disposition of formononetin via enteric recycling: metabolism and excretion in mouse intestinal perfusion and Caco-2 cell models. Mol Pharm 2:319–328.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Johnson DA and
    2. Amidon GL
    (1988) Determination of intrinsic membrane transport parameters from perfused intestine experiments: a boundary layer approach to estimating the aqueous and unbiased membrane permeabilities. J Theor Biol 131:93–106.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Liu W,
    2. Feng Q,
    3. Li Y,
    4. Ye L,
    5. Hu M, and
    6. Liu Z
    (2012) Coupling of UDP-glucuronosyltransferases and multidrug resistance-associated proteins is responsible for the intestinal disposition and poor bioavailability of emodin. Toxicol Appl Pharmacol 265:316–324.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Liu Y and
    2. Hu M
    (2002) Absorption and metabolism of flavonoids in the caco-2 cell culture model and a perused rat intestinal model. Drug Metab Dispos 30:370–377.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Murphy PA,
    2. Barua K, and
    3. Hauck CC
    (2002) Solvent extraction selection in the determination of isoflavones in soy foods. J Chromatogr B Analyt Technol Biomed Life Sci 777:129–138.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Nurmi T,
    2. Mazur W,
    3. Heinonen S,
    4. Kokkonen J, and
    5. Adlercreutz H
    (2002) Isoflavone content of the soy based supplements. J Pharm Biomed Anal 28:1–11.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Pang KS,
    2. Maeng HJ, and
    3. Fan J
    (2009) Interplay of transporters and enzymes in drug and metabolite processing. Mol Pharm 6:1734–1755.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Ruan JQ,
    2. Li S,
    3. Li YP,
    4. Wu WJ,
    5. Lee SM, and
    6. Yan R
    (2015) The presystemic interplay between gut microbiota and orally administered calycosin-7-O-β-D-glucoside. Drug Metab Dispos 43:1601–1611.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Ruan JQ and
    2. Yan R
    (2014) Regioselective glucuronidation of the isoflavone calycosin by human liver microsomes and recombinant human UDP-glucuronosyltransferases. Chem Biol Interact 220:231–240.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Scalbert A and
    2. Williamson G
    (2000) Dietary intake and bioavailability of polyphenols. J Nutr 130(8S, Suppl)2073S–2085S.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Shen P,
    2. Liu MH,
    3. Ng TY,
    4. Chan YH, and
    5. Yong EL
    (2006) Differential effects of isoflavones, from Astragalus membranaceus and Pueraria thomsonii, on the activation of PPARalpha, PPARgamma, and adipocyte differentiation in vitro. J Nutr 136:899–905.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Shi J,
    2. Zheng L,
    3. Lin Z,
    4. Hou C,
    5. Liu W,
    6. Yan T,
    7. Zhu L,
    8. Wang Y,
    9. Lu L, and
    10. Liu Z
    (2015) Study of pharmacokinetic profiles and characteristics of active components and their metabolites in rat plasma following oral administration of the water extract of Astragali radix using UPLC-MS/MS. J Ethnopharmacol 169:183–194.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Tang L,
    2. Singh R,
    3. Liu Z, and
    4. Hu M
    (2009) Structure and concentration changes affect characterization of UGT isoform-specific metabolism of isoflavones. Mol Pharm 6:1466–1482.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Walgren RA,
    2. Lin JT,
    3. Kinne RK, and
    4. Walle T
    (2000) Cellular uptake of dietary flavonoid quercetin 4′-beta-glucoside by sodium-dependent glucose transporter SGLT1. J Pharmacol Exp Ther 294:837–843.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Walsh KR,
    2. Haak SJ,
    3. Bohn T,
    4. Tian Q,
    5. Schwartz SJ, and
    6. Failla ML
    (2007) Isoflavonoid glucosides are deconjugated and absorbed in the small intestine of human subjects with ileostomies. Am J Clin Nutr 85:1050–1056.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Wang SW,
    2. Chen J,
    3. Jia X,
    4. Tam VH, and
    5. Hu M
    (2006) Disposition of flavonoids via enteric recycling: structural effects and lack of correlations between in vitro and in situ metabolic properties. Drug Metab Dispos 34:1837–1848.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Wilkinson AP,
    2. Gee JM,
    3. Dupont MS,
    4. Needs PW,
    5. Mellon FA,
    6. Williamson G, and
    7. Johnson IT
    (2003) Hydrolysis by lactase phlorizin hydrolase is the first step in the uptake of daidzein glucosides by rat small intestine in vitro. Xenobiotica 33:255–264.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Wolffram S,
    2. Blöck M, and
    3. Ader P
    (2002) Quercetin-3-glucoside is transported by the glucose carrier SGLT1 across the brush border membrane of rat small intestine. J Nutr 132:630–635.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Xia B,
    2. Zhou Q,
    3. Zheng Z,
    4. Ye L,
    5. Hu M, and
    6. Liu Z
    (2012) A novel local recycling mechanism that enhances enteric bioavailability of flavonoids and prolongs their residence time in the gut. Mol Pharm 9:3246–3258.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Yamaguchi H,
    2. Yano I,
    3. Hashimoto Y, and
    4. Inui K
    (2000) Secretory mechanisms of grepafloxacin and levofloxacin in the human intestinal cell line caco-2. J Pharmacol Exp Ther 295:360–366.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Yang Z,
    2. Gao S,
    3. Wang J,
    4. Yin T,
    5. Teng Y,
    6. Wu B,
    7. You M,
    8. Jiang Z, and
    9. Hu M
    (2011) Enhancement of oral bioavailability of 20(S)-ginsenoside Rh2 through improved understanding of its absorption and efflux mechanisms. Drug Metab Dispos 39:1866–1872.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Ye L,
    2. Yang X,
    3. Yang Z,
    4. Gao S,
    5. Yin T,
    6. Liu W,
    7. Wang F,
    8. Hu M, and
    9. Liu Z
    (2013) The role of efflux transporters on the transport of highly toxic aconitine, mesaconitine, hypaconitine, and their hydrolysates, as determined in cultured Caco-2 and transfected MDCKII cells. Toxicol Lett 216:86–99.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Yu D,
    2. Duan Y,
    3. Bao Y,
    4. Wei C, and
    5. An L
    (2005) Isoflavonoids from Astragalus mongholicus protect PC12 cells from toxicity induced by L-glutamate. J Ethnopharmacol 98:89–94.
    OpenUrlCrossRefPubMed
  43. ↵
    1. Zhao P,
    2. Su G,
    3. Xiao X,
    4. Hao E,
    5. Zhu X, and
    6. Ren J
    (2008) Chinese medicinal herb Radix Astragali suppresses cardiac contractile dysfunction and inflammation in a rat model of autoimmune myocarditis. Toxicol Lett 182:29–35.
    OpenUrlCrossRefPubMed
  44. ↵
    1. Zhou Q,
    2. Zheng Z,
    3. Xia B,
    4. Tang L,
    5. Lv C,
    6. Liu W,
    7. Liu Z, and
    8. Hu M
    (2010) Use of isoform-specific UGT metabolism to determine and describe rates and profiles of glucuronidation of wogonin and oroxylin A by human liver and intestinal microsomes. Pharm Res 27:1568–1583.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Zou TB,
    2. Feng D,
    3. Song G,
    4. Li HW,
    5. Tang HW, and
    6. Ling WH
    (2014) The role of sodium-dependent glucose transporter 1 and glucose transporter 2 in the absorption of cyanidin-3-o-β-glucoside in Caco-2 cells. Nutrients 6:4165–4177.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

Transport of Calycosin-7-O-β-d-Glucoside by SGLT-1 in Rats

Jian Shi, Haihui Zheng, Jia Yu, Lijun Zhu, Tongmeng Yan, Peng Wu, Linlin Lu, Ying Wang, Ming Hu and Zhongqiu Liu
Drug Metabolism and Disposition March 1, 2016, 44 (3) 283-296; DOI: https://doi.org/10.1124/dmd.115.067009
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement