ToxSci Advance Access originally published online on February 14, 2008
Toxicological Sciences 2008 104(1):27-39; doi:10.1093/toxsci/kfn026
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
A Possible Role of Multidrug Resistance–Associated Protein 2 (Mrp2) in Hepatic Excretion of PCB126, an Environmental Contaminant: PBPK/PD Modeling
* Quantitative and Computational Toxicology Group, Department of Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, Colorado 80523-1680
Department of Pharmacy Practice, Faculty of Pharmaceutical Sciences, Naresuan University, Phitsanulok 65000, Thailand
School of Pharmaceutical Sciences, Kitasato University, Tokyo 108-8641, Japan
1 To whom correspondence should be addressed at Quantitative and Computational Toxicology Group, Department of Environmental and Radiological Health Sciences, Colorado State University, 137A Physiology Building, Fort Collins, CO 80523-1680. Fax: (970) 491-7569. E-mail: raymond.yang{at}colostate.edu.
Received November 13, 2007; accepted February 4, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
3,3',4,4',5'-Pentachlorobiphenyl (PCB126) is a carcinogenic environmental pollutant and its toxicity is mediated through binding with aryl hydrocarbon receptor (AhR). Earlier, we found that PCB126 treated F344 rats had 110–400 times higher PCB126 concentration in the liver than in the fat. Protein binding was suspected to be a major factor for the high liver concentration of PCB126 despite its high lipophilicity. In this research, we conducted a combined pharmacokinetic/pharmacodynamic study in male F344 rats. In addition to blood and tissue pharmacokinetics, we use the development of hepatic preneoplastic foci (glutathione-S-transferase placental form [GSTP]) as a pharmacodynamic endpoint. Experimental data were utilized for building a physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) model. PBPK/PD modeling was consistent with the experimental PK and PD data. Salient features of this model include: (1) bindings between PCB126 and hepatic proteins, particularly the multidrug resistance–associated protein (Mrp2), a protein transporter; (2) Mrp2-mediated excretion; and (3) a relationship between area under the curve of PCB126 in the livers and % volume of GSTP foci. Mrp2 involvement in PCB126 pharmacokinetics is supported by computational chemistry calculation using a three-dimensional quantitative structure–activity relationship model of Mrp2 developed by S. Hirono et al. (2005, Pharm. Res. 22, 260–269). This work, for the first time, provided a plausible role of a versatile hepatic transporter for drugs, Mrp2, in the disposition of an important environmental pollutant, PCB126.
Key Words: carcinogenesis; disposition; physiologically based pharmacokinetics; polychlorinated biphenyls; PCB 126; Mrp2.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Polychlorinated biphenyls (PCBs) are halogenated aromatic hydrocarbons that had been widely used in industry. Because of their persistence as environmental pollutants, they had been discontinued from any usage since 1970's. However, significant amount of PCBs is still detectable in foods, human and animal tissues, and in the environment (CDC, 2005
; Safe, 1994; Safe et al., 1985). 3,3',4,4',5'-Pentachlorobiphenyl (PCB126) is the most toxic congener of all PCBs with carcinogenic effects. Structurally, PCB126 is capable of binding with aryl hydrocarbon receptor (AhR) and elicits biological effects, which include the induction of cytochrome P4501A1 and 1A2 (CYP1A1, 1A2), thymic involution, wasting syndrome (Safe, 1994), and carcinogenesis in the liver, lung, and mouth in rats (NTP, 2006).
Despite its high lipophilicity, the levels of PCB126 in the liver are much higher than those observed in adipose tissue (Chu et al., 1994; Dean et al., 2002; Lohitnavy et al., 2004). This is similar to those of chlordecone and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (Abraham et al., 1988; Belfiore et al., 2007), but at a much greater extent. These results suggested that protein binding is responsible for high levels of PCB126 in the liver. From a different perspective, whereas PCB126 is known to be persistent in the environment (Safe, 1994), in our laboratory, earlier results demonstrated that PCB126 could attain its steady state fairly rapidly (Lohitnavy et al., 2004). An initial pharmacokinetic estimation of half-life for this chemical turned out to be around 3 days (unpublished data). These results suggested that there was a relatively high hepatic clearance of PCB126. Therefore, we further considered the possible involvement of transporter protein(s), specifically the multidrug resistance–associated protein 2 (Mrp2).
Excretion of xenobiotics can be mediated through several mechanisms. One of which is biliary excretion involving transporter proteins in the liver. This particular mechanism is responsible for excretion of many drugs and chemicals (Petzinger and Geyer, 2006; Shitara et al., 2006). Mrp2 is an adenosine triphosphate–binding cassette transporter, which is responsible for biliary excretion of many drugs and xenobiotics (Borst et al., 2006; Jedlitschky et al., 2006). Recently, a three-dimensional quantitative structure–activity relationship (3D-QSAR) model of Mrp2 in rats was developed (Hirono et al., 2005). Using a number of molecular indices (i.e., steric field, electrostatic field, and C log P), and computational chemistry, this 3D-QSAR model is capable of assessing the feasibility of Mrp2 binding, as well as estimating binding affinity (Km), of chemicals. Although Mrp2 was reported to having significant roles in biliary excretion in many drugs and xenobiotics (Borst et al., 2006; Jedlitschky et al., 2006), there is thus far no evidence demonstrating that Mrp2 could have a significant role in the disposition of PCB126 or other PCBs. In this paper, for the first time, we demonstrated a possible role of Mrp2 in the disposition of an environmental contaminant, PCB126. By incorporating this suggested role of Mrp2 into physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) model, the computer simulations were consistent with a number of sets of experimental results from different laboratories.
In this study, we also focused on pharmacodynamics of PCB126. To test carcinogenic potential of chemicals, we incorporated pharmacokinetics and pharmacodynamics into the Ito's medium-term liver bioassay (Ito et al., 2003; Shirai, 1997), one of the most extensively studied cancer bioassay protocols. This bioassay involves a single-dose administration of an initiator (diethylnitrosamine, DEN), followed by repeated administration of a promoter (a test chemical; in this case PCB126) to male F344 rats. The promotional effect of this assay is further enhanced by a two-third partial hepatectomy (PH). Based on the development of glutathione-S-transferase placental form (GSTP) foci as a marker for preneoplastic lesions, this experimental model has shown excellent capability in predicting liver carcinogenicity in rats (Ogiso et al., 1990; Shirai, 1997). In our modified protocol, we added multiple sacrificing time points to observe liver GSTP foci development over time, as well as tissue kinetics of PCB126 for the development of PBPK/PD model.
There had been an earlier attempt to a PBPK model of PCB126 with the incorporation of hepatic protein binding to AhR and CYP1A2 (NTP, 2006); however, this model was unable to describe the tissue concentration data accurately (NTP, 2006). In the present study, we utilized the available National Toxicology Program (NTP) experimental data (NTP, 2006) plus our own results in the building of a new PBPK/PD model. In this model, in addition to protein bindings with AhR and CYP1A2, we incorporated a transporter protein, Mrp2, in the disposition of PCB126 based on supporting evidence from 3D-QSAR and computational chemistry (Hirono et al., 2005). The resulting computer simulations were consistent with all the available data.
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
This study consisted of three parts: (1) development of a PBPK/PD model for PCB126 in the rat under normal physiological conditions with the incorporation of binding between PCB126 and two hepatic proteins (i.e., AhR and CYP1A2) and binding/excretion of PCB126 via hepatic Mrp2. The data used in this part were from 3D-QSAR modeling, computational chemistry, and from mining the literature; (2) experimental pharmacokinetic study of PCB126 under the time-course medium-term liver foci bioassay protocol, and simulation of this data set by incorporation of pathophysiological conditions (i.e., two-third PH and recovery); and (3) correlation between internal dosimetry and experimental liver GSTP foci development.
Development of a PBPK/PD Model under Normal Physiological Conditions with Incorporation of Binding Between PCB126 and Hepatic Proteins and Excretion of PCB126 via Hepatic Mrp2
Pharmacokinetic Studies of PCB126 by the NTP
Recently the NTP published a 2-year carcinogenic study of PCB126 (NTP, 2006). In this report, there were single-dose and multiple-dose pharmacokinetic studies. Thus, as described below, we employed these data as working data sets in PBPK model development in rats without pathological conditions.
Single-dose study.
Female Sprague–Dawley rats (SD, 20–22 weeks of age) were orally administered with a single-dose 1000 ng PCB126 in corn oil. PCB126 levels were determined in liver, blood, and fat samples at multiple time points. Group of five rats per time point were bled and the tissues were collected at 0.5, 1, 1.5, 2, 3, 8, 16, or 24 h post-PCB126 administration. Tissue samples were analyzed using a validated gas chromatography–mass spectrometry (GC–MS) method. These data were available in the NTP Technical Report (NTP, 2006).
Multiple-dose 2-year study.
Female SD rats were orally administered with corn oil or 30, 100, 175, 300, 550, and 1000 ng PCB126/kg body weight (five times per week) for 2 years. At week 13, 30, 52, and 104, five animals in each group were sacrificed. Livers, blood, and fat tissue were harvested at the specified time points. Tissue samples were analyzed using a validated GC–MS method. Body weight of the animals was recorded periodically up to 2 years, and liver weights were reported from interim sacrifices up to one year. In the NTP Technical Report on PCB126, there was a PBPK model of PCB126. The NTP model consists of five compartments (liver, kidney, muscle, fat, and lungs) with blood circulating and connecting the compartments; it was based on the original TCDD model constructed by Kohn et al. (1993, 1996). Partition coefficients were estimated from the work of Parham et al. (1997). Even though this PBPK model was unable to describe the PCB126 tissue concentration accurately, this preliminary work by NTP (2006) proved to be very helpful for our further development of a PBPK model as described in this paper.
Data extraction.
The figures illustrating concentration–time courses of the pharmacokinetic studies of PCB126 in the NTP Technical Report (NTP, 2006) were utilized. A digiMatic Program (version 2.1; Richmond, VA) was used to extract numerical co-ordinates from the concentration–time courses of PCB126 presented in the NTP Technical Report.
Computational Chemistry, 3D-QSAR Modeling of Binding of PCB126 to Mrp2
A 3D-QSAR model for rat Mrp2 was recently developed using ligand-based drug design techniques (Hirono et al., 2005). PCB126 is known to be present in human body at extremely low concentrations (CDC, 2005). This congener, being the most toxic of all PCB congeners, is usually studied at very low dose levels in animal experimentation. These realities make in vitro binding studies at realistic in vivo concentrations difficult due to analytical limitations. Therefore, we chose an in silico approach and determined the feasibility of Mrp2 binding by PCB126 based on molecular characteristics such as molecular steric field, molecular electrostatic field and ClogP calculated by SYBYL software package (Tripos, Inc., St Louis, MO,). With the 3D-QSAR modeling and computational chemistry calculation, we found that Mrp2 binding feasibility is at least as good as or better than leukotriene C4, S- or R-grepafloxacin glucuronide, temocaprilate, and L-methotrexate with a binding affinity constant (Km) estimated to be 7760.0nM (log 1/Km = 5.11). We believe that, in the present case, the in silico approach is a reasonable alternative without the necessity of conducting Mrp2 binding experiments because: (1) Hirono et al. (2005), using this same 3D-QSAR modeling and computational chemistry approach, demonstrated that their predicted values of log (1/Km) were within 2% of the experimentally determined values for 16 chemicals in their training set. The largest difference of 13% was seen between predicted and experimental values in one of the two chemicals in their test set. Furthermore, according to our sensitivity analyses, varying the above in silico derived Km values by two times either way resulted in little or no change of simulation results (unpublished data). Thus, we believe that the in silico derived binding affinity constant is adequate for our purpose. In brief, conformational search for PCB126 was performed using the Conformational Analyzer with Molecular Dynamics and Sampling developed by Tsujishita and Hirono (1997). Then, the conformation of the stable conformer of PCB126 was analyzed using the SUPERPOSE program (Iwase and Hirono, 1999). Subsequently, conventional comparative molecular field analysis (CoMFA) was performed using the QSAR option of the SYBYL software. Finally, the CoMFA 3D-QSAR model was generated using a partial least square algorithm.
Strategy in PBPK Model Development
The stepwise development of our PBPK model is given below:
Overall model scheme.
The conceptual PBPK model of PCB126 is illustrated in Figure 1A. The model structure included liver, rapidly perfused, slowly perfused, blood, fat, and gastrointestinal (GI) tract compartments. The model described flow-limited transfer of PCB126 in liver, rapidly perfused, slowly perfused, and fat compartments. All parameters used in the model are summarized in Table 1. Physiological parameters were obtained from Brown et al. (1997). Partition coefficients of PCB126 in tissues were taken from the NTP Technical Report (NTP, 2006).
|
|
Protein binding.
Bindings with AhR and CYP1A2 are responsible for not only toxicological effects of PCB126, but also high levels of PCB126 in the livers. Thus, we incorporated two reversible binding processes between PCB126 and the responsible proteins into our model (Fig. 1B). Furthermore, to describe the Mrp2-mediated excretion process, a Michaelis–Menten equation was added in the liver submodel; the utilization of a Michaelis–Menten equation is consistent with Hirono et al. (2005)
. The binding affinity of Mrp2 (Km) obtained from the 3D-QSAR calculations were incorporated into the PBPK model, whereas maximum binding capacity of PCB126 to Mrp2 was estimated using the Mead–Nelder optimization algorithm as well as other parameters. The equation describing the rate of change in the amount of PCB126 in the liver with Mrp2 excretion was expressed as follows:
 | (1) |
.
Model Validation
Recently, Fisher et al. (2006) conducted a single-dose PCB126 pharmacokinetic study. In this study, male SD rats were orally administered with a single-dose of PCB126 (7.5, 75, and 275 µg/kg), and liver concentration levels of PCB126 were measured from day 1 up to day 22 post-PCB126 administration. Because the data concerning changes in body weight of the animals were available, we also incorporated these body weight changes into our model simulations. The data set in this paper, which is different from those data sets used in constructing the PBPK model, was used for model validation.
Sensitivity Analysis
Sensitivity analysis is a useful approach for identifying important parameters affecting a pharmacokinetic measurement (Clewell et al., 1994). Log-normalized sensitivity parameters (LSPs) were defined as follows:
 | (2) |
Software
The model code was written and the simulations were performed using ACSL Tox (version 11.8.4; Aegis Technologies Group Inc., Marietta, GA). The sensitivity analysis and parameter optimization were carried out using ACSL Math (version 2.5.4; Aegis Technologies Group, Inc.).
Pharmacokinetic/Pharmacodynamic Studies of PCB126 under the Time-Course Medium-Term Liver Foci Bioassay Protocol, and Simulation of these Data Sets by Incorporating Pathophysiological Conditions
Chemicals
PCB126 (> 99% purity) was purchased from AccuStandard (New Haven, CT). 2,2',4,4',5,5'-Hexachlorobiphenyl (PCB74; > 98% purity) was purchased from Ultra Scientific (North Kingstown, RI) and used as an internal standard (I.S.) for GC analyses. DEN was purchased from Sigma Chemical (St Louis, MO). Pentane (high-performance liquid chromatography [HPLC] grade) and sulfuric acid were supplied by VWR Scientific (Denver, CO). Anhydrous sodium sulfate was purchased from Fisher Scientific (Houston, TX). Florisil was supplied by Alltech Associates (Deerfield, IL).
Animals and Experimental Protocol
Male F344 rats, 30 days of age, supplied by Harlan Sprague–Dawley (Indianapolis, IN), were maintained at the Painter Center, Colorado State University. The Center is fully accredited by the American Association for Accreditation of Laboratory Animal Care. The animals were given food (Harlan Teklad NIH-07 diet; Madison, WI) and water ad libitum, and the lighting was set at 12-h light/dark cycle. The study was conducted in accordance with the National Institutes of Health guidelines for the care and use of laboratory animals.
After a 4-week acclimation, the rats were randomly allocated into three groups, and treated according to the time-course medium-term liver foci bioassay (Fig. 2). In brief, on day 0, the animals were administered with a single ip injection of DEN (200 mg/kg body weight) dissolved in normal saline. On day 14, the rats were orally administered with a daily oral gavage (5 ml of dosing solution/kg body weight) of corn oil (control), 3.3 µg PCB126/kg body weight (low dose), or 9.8 µg PCB126/kg body weight (high dose) in corn oil until sacrifices. On day 21, a two-third PH was performed on the rats. On the surgery day and the following 2 days, oral dosing was stopped to decrease stress to the animals while they were recovering from the PH. On day 20, 24, 28, 47, and 56, six rats from each group were sacrificed by aortic exsanguination under anesthesia. The body and liver weights from each rat were recorded at the sacrifice (Table 2). One piece of a liver (approximately 5-mm thickness) from each liver lobe was collected and fixed in formalin for GSTP foci development analysis. The remaining part of the livers was collected for PCB126 tissue concentration analyses; they were stored at –80°C until chemical analysis.
|
|
PCB126 Extraction
Liver samples were weighed (approximately 1–2 g/sample). The samples were chopped and 1.5 ml of water added. Subsequently 250 ng PCB74 was added to the samples as an I.S. Then 3 ml of 60% sulfuric acid was added to the samples and mixed vigorously. Following standing overnight at room temperature for complete tissue digestion, 5 ml of pentane was added to the samples and mixed vigorously. The samples were then centrifuged at 3200 rpm for 15 min at 25°C using a Centrifuge Model 5682 (Forma Scientific, Inc., Marietta, OH) and the organic layer was collected. Two more extractions were carried out and the organic layers combined. To clean up the extracts, the combined organic layers were passed through a clean-up column consisting of 3.0 g of anhydrous sodium sulfate and 500 mg of activated florisil. The cleaned up organic extracts were evaporated under nitrogen streams until dryness. Each sample was reconstituted with 1 ml of pentane (HPLC grade) and analyzed by gas chromatography. The % recovery of PCB126 by this extraction method is about 75%.
GC Analyses
An HP-5890 Series II Plus gas chromatograph (Hewlett Packard, Wilmington, DE) with an electron capture detector was employed to analyze PCB126. A DB-5 (crosslink 5% phenyl methylsilicone, 30 m x 0.53 mm x 0.5 µm film thickness, Supelco, Bellefonte, PA) capillary column was used. The initial temperature was 80°C for 3 min, programmed to 120°C at the rate of 15°C/min and stayed at this temperature level for 5 min and then programmed to 180°C at the rate of 20°C/min. The flow rate of carrier gas, helium, and the make-up gas, nitrogen, were 5 and 80 ml/min, respectively. The temperature of injector and detector were 225°C and 320°C, respectively. The volume of injection was 5–10 µl per sample. The concentration levels of PCB126 were quantified using an I.S. method. A calibration curve was built and fitted using a linear regression equation. The detection limit of the system was 0.1 ng PCB126.
Development of GSTP Foci in the Liver
Quantification of GSTP foci.
Livers were collected at the specified sacrifice time points, fixed in formalin, sliced to 5-µm thickness, and stained for expression of GSTP immunohistochemically as previously described by Dean et al (Dean et al. 2002). In brief, liver sections were deparaffinized in xylene and rehydrated by passage through an alcohol series. Endogenous peroxidase was quenched in 3% hydrogen peroxide for 10 min. Slides were rinsed with deionized water and placed in PBS (pH 7.4; 2.7mM KCl, 0.14M NaCl, 1.5mM KH2PO4). GSTP foci were detected with a primary GSTP antibody using a standard avidin–biotin complex method. Area and number of GSTP foci were measured using an Olympus BX51 light microscope (Olympus Optical Co., LTD, Tokyo, Japan) coupled with an Optronics DEI-750CE microscope mounted digital camera (Optronics, Coleta, CA) and a stage-mounted Microcode II Digital Readout (Boeackeler Instruments, Inc., Tucson, AZ). Image analysis software was the Bioquant Nova for Windows 98 (Version 5.00.8) computerized histomorphometry program (B&M Biometrics Inc., Nashville, TN), installed in an AOpen PIII-700 computer (Aopen, Inc., Taipei, Taiwan). Any GSTP focus with more than two cells or larger (approximately 50 µm diameter) was counted and area of the GSTP focus was recorded. Subsequently the two-dimensional data of GSTP foci development from the tissue slices were used to calculate numbers and volume of GSTP foci in the livers using STEREO (the McArdle Laboratory, Madison, WI) as described earlier by Xu et al. (1998) and Ou et al. (2001).
Statistical analyses.
Statistical comparisons between treatment groups and the concurrent controls were performed using one-way ANOVA. Values were considered to be statistically significant when p < 0.05 (Minitab, Inc., State College, PA).
Correlation between Internal Dosimetry and Experimental Liver GSTP Foci Development
Calculations of internal dose metrics.
We selected AUCLiver and %bound AhR as the two most likely candidates for the internal dose metric. Using our PBPK/PD model under the conditions of time-course medium-term liver bioassay, the values of AUCLiver and %bound AhR over time were determined. WinNonlin Professional (version 4.1; Mountain View, CA) was then employed to determine the correlation between the chosen liver internal metrics and the volume of GSTP foci. A simple maximal effect equation was used to describe the correlation between the internal dose metric and the GSTP foci development. The equation can be described as follows:
 | (3) |
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Model Performance Under Normal Physiological Conditions
PBPK Model Simulations
The single-dose study.
The model simulations of PCB126 concentrations in the liver and fat (Fig. 3A) were consistent with the experimental data reported in the NTP Technical Report (NTP, 2006
). According to our PBPK/PD model, with this dosing scenario (1000 ng PCB126 oral single dose) at 24 h post-PCB126 administration, most of PCB126 (about 55.6% of total administered dose) was found in the liver and some of PCB126 (approximately 2.0%) was excreted out from the body via hepatic Mrp2 (Fig. 3B).
|
The multiple-dose study.
In the multiple-dose 2-year study, there were six dosing levels; 30, 100, 175, 300, 550, and 1000 ng PCB126/kg body weight. The model simulations of PCB126 concentrations in the liver, fat, and blood were consistent with the experimental data (Figs. 4A–F). Further PBPK/PD modeling revealed that % excretion of PCB126 via Mrp2 was increasing, whereas %PCB126 in the livers was decreasing over time in all dosing levels (Figs. 5A and 5B). For instance, at the dosing level of 30 ng PCB126/kg body weight, at 1, 13, 30, 52, and 104 weeks, % excretion of PCB126 via Mrp2 was 9.3, 52.4, 73.3, 82.4, and 88.6%, respectively, whereas %PCB126 in the liver was 49.6, 27.3, 15.8, 10.7, and 7.4%, respectively (Fig. 5A).
|
|
PBPK Model Validation
An entirely different data set from a recent study conducted by Fisher et al. (2006)
was used for model validation. As shown in Figure 6, our PBPK model was utilized to simulate the concentration–time course data taken from this paper. The results, shown in Figure 6, demonstrated fair consistency between the liver concentration–time courses of PCB126 at the dosing levels of 75 and 275 µg/kg. However, at the lowest dosing level of 7.5 µg/kg, the overprediction of the experimental data was more pronounced (Fig. 6). Previously, it was reported that, in younger rats, the expression levels of Mrp2 were significantly different than adult rats (Tomer et al., 2003). In this study, the authors used relatively smaller (151–175 g) and younger rats (Fisher et al., 2006), these physiological factors might contribute to the inconsistency of our model simulation results to the experimental data.
|
Sensitivity Analysis
The sensitivity of hepatic concentration of PCB126 to parameters related to binding with AhR, CYP1A2, and Mrp2 at various time points in the single-dose and multiple-dose (1000 ng dosing level) studies is summarized in Table 3. From the single-dose study, at 24-h post-PCB126 administration, affinity to CYP1A2 (KB2), capacity of CYP1A2 (BM2I), basal level of CYP1A2 (BM2O) and the capacity of AhR (BM1) had the largest effect on the hepatic concentration. Affinity to AhR (KB1), Mrp2 (Km), and maximum binding capacity of Mrp2 (Vmax) had moderate effect, whereas partition coefficient of PCB126 in the liver (PL) had minimal effect on hepatic concentration of PCB126. In the 2-year oral multiple-dose study, at 104 weeks, Km and Vmax of Mrp2 had the most prominent effect on concentration of PCB126 in the livers. Induction rate of CYP1A2 (slope), BM2O, KB2, BM1, and KB1 had moderate effect, whereas PL had the weakest effect on PCB126 concentration in the livers.
|
Model Performance Under the Conditions of Time-Course Medium-Term Liver Foci Bioassay
Body and liver weight and liver/body weight ratio of the rats in the time-course medium-term liver foci bioassay were summarized in Table 4. There was no statistical difference in either body weight or liver weight in all treatment groups as compared with concurrent controls. However, in the low dose group, on day 56, liver/body weight ratio was significantly different compared with concurrent control (p < 0.05). In the high dose group, on days 20 and 56, liver/body weight ratio was significantly different compared with the concurrent controls (p < 0.05).
|
With our own experiment, model simulations of PCB126 concentrations in livers were consistent with experimental results (Figs. 7A and 7B). Oral dosing of PCB126 was started on day 14 until day 21. The dosing was stopped between days 21 and 24. Our PCB126 liver concentration levels on day 24 were lower than the limit of quantification. Using PBPK/PD modeling, % bound AhR in low and high dose group was estimated to be 99.2 and 99.8%, respectively, at 8 weeks (Fig. 8). Area under the curve in the liver (AUCLiver) increased over time in both dosing levels (Fig. 9).
|
|
|
GSTP Foci Development in the Time-Course Medium-Term Liver Foci Bioassay
As shown in Figure 10, when compared with the controls, there was no statistical significance observed in both number and volume of GSTP foci in the rats treated with 3.3 ng PCB126/kg body weight (low dose group). However, in rats treated with 9.8 ng PCB126/kg body weight (high dose group), there were significantly higher numbers of GSTP foci in the livers on day 24, 28, 47, and 56 (p < 0.05). For instance, on day 56, numbers of GSTP foci in the livers in control and high dose groups were 19,949.0 ± 6913.7 and 35,617.8 ± 8806.7 foci (mean ± standard deviation; p < 0.05), respectively, whereas % volume of GSTP foci for control and high dose groups were 0.230 ± 0.062 and 0.520 ± 0.107% (mean ± standard deviation; p < 0.05), respectively.
|
Relationship between AUCLiver and Liver GSTP Foci Development
Using a simple maximal effect equation, AUCLiver correlated well with % volume of GSTP foci in the livers at both dose levels (Figs. 11A and 11B). Parameters of the maximal effect equation, a reflection of relationship between AUCLiver and % volume of GSTP foci, were summarized in Table 4. In the low and high dose group, maximal volumes of GSTP foci (Emax) were 0.402 and 0.717%, respectively, whereas EAUCLiver,50 in the low and high dose group were 288,096 and 1,413,766 nmoles h/l, respectively.
|
Model Simulation of PCB126 Pharmacokinetics Under Pathophysiological Conditions
In model simulations from our initiation–promotion study, the computer modeling results showed a similar trend in increasing fraction of PCB126 in feces via the Mrp2-mediated excretion process (Fig. 12A). Computer simulation also demonstrated a decreasing trend of the ratio between hepatic PCB126 levels and the total administered dose (Fig. 12B).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
For the first time, we demonstrated a possible involvement of the versatile Mrp2 transporter protein in the excretion of a highly persistent environmental contaminant, PCB126. Furthermore, our PBPK/PD modeling work reported herein has the following significance:
First, our PBPK/PD model is capable of simulating blood and tissue kinetics of PCB126 in rats under many different dosing scenarios. These included the NTP single (Fig. 3) and multiple dosing studies up to 2 years (Fig. 4) (NTP, 2006), Fisher et al. studies (Fig. 6) (Fisher et al., 2006), as well as our own studies involving an initiation–promotion experimental protocol (Fig. 7).
Second, the much higher hepatic concentration of PCB126 over that in the fat despite high lipophilicity of PCB126 is most likely the result of protein binding. In the liver subcompartment of our PBPK/PD model, PCB126 binds to AhR and CYP1A2 in a reversible fashion. The level of AhR in the livers was assumed to be constant throughout the testing conditions (Andersen et al., 1993), whereas CYP1A2 was considered to be an inducible protein (Chubb et al., 2004; NTP, 2006). Using the 3D-QSAR model and computational chemistry developed earlier by Hirono et al. (2005), PCB126 was found to be a good substrate for Mrp2 binding with relatively high affinity. Thus, an Mrp2-mediated excretion process of PCB126 was incorporated into the model. The resulting PBPK/PD model simulations were consistent with a number of experimental data sets from different laboratories.
Third, under pathophysiological conditions involving two-third PH, the model can also simulate the time-course liver concentrations. PBPK/PD model simulations under our experimental conditions suggested that AUCLiver was a better internal dose metric than the specific binding between PCB126 and AhR. There was a correlation between AUCLiver and the observed liver GSTP foci development.
In the current PBPK model of PCB126, we incorporated reversible bindings between PCB126 and hepatic proteins (i.e., AhR and CYP1A2), and an excretion of PCB126 from the liver via Mrp2. From the sensitivity analysis (Table 3), in early period (less than 24 h) of the 1000 ng single oral dose study, the parameters related to the bindings between PCB126 and AhR and CYP1A2 (BM1, KB1, BM2I, BM2O, and KB2) had a stronger effect on PCB126 levels in the liver than those for Mrp2 (Km and Vmax). However, when the oral dosing continued, at the same dose level, the parameters related to the Mrp2-mediated excretion became more influential on hepatic concentration of PCB126 than those of AhR and CYP1A2 bindings. It is possible that at the beginning, absorbed PCB126 from the GI tract entered into the liver and bound preferentially to AhR and CYP1A2 because of their higher binding affinities. As AhR and CYP1A2 had low binding capacities, they became saturated with PCB126. The excess PCB126 in the liver then bound to Mrp2, a transporter protein with lower binding affinity but higher capacity. In our time-course medium-term liver bioassay study, a two-third PH was performed on all of the animals on day 21. When the two-third liver is removed, the remaining portion of the liver will regenerate (Taub, 2004). In our PBPK/PD model, we incorporated this liver regeneration process (Lu et al., 2006). To reduce stress to the animals, oral dosing in all of the treatment groups was stopped from day 21 to day 24 (Fig. 2). Interestingly, on day 24 (right before the dosing resumed), the liver concentrations of PCB126 in samples harvested from the rats exposed to PCB126 in both dosing levels were lower than the detection limit (Figs. 7A and 7B). In our modeling, without any changes in Mrp2 affinity or capacity, the model was not able to successfully simulate the liver concentration at this time point. Mrp2 expression was increased by 46% at 12-h postsurgery (Gerloff et al. 1999) and bile flow was increased by 73% at 24 h after PH (Vos et al., 1999). In addition, Villanueva et al. (2005) reported that, after PH, there was an increase in Mrp2-mediated excretion of dinitrophenyl-S-glutathione, a substrate of Mrp2. Thus, in our PBPK/PD model, from day 21 to day 24, the value of the maximal binding capacity of Mrp2 (Vmax) was increased from 64.6 to 2000.0 nmol/h to fit the observed liver concentration data. This result suggested that PH might be a potent stimulator in translocation of Mrp2 from its intracellular storage sites.
It has been recognized that PCB126, a coplanar PCB, exerts its toxicological actions via binding to AhR (Safe, 1994; Safe et al., 1985). However, it has been hypothesized that there may be other factors contributing to its toxicological effects such as free radical production resulting from CYP1A2 induction (Jin et al., 2001; Katynski et al., 2004). It has been suggested that, for chemicals exerting toxicological effects in liver, AUC of the chemical can be used as an internal dose metric (MacGregor et al., 2001). Hence, we chose AUCLiver as our internal dose metric. The resulting PBPK/PD modeling indicated that AUCLiver has correlation with the liver GSTP foci development (Figs. 11A and 11B) and is a better internal dose metric for this pharmacodynamic endpoint.
In the past 13 years or so, a number of clonal growth models were developed to describe liver foci formation in rats treated with carcinogenic chemicals (Conolly and Andersen, 1997; Conolly and Kimbell, 1994; Lu et al., 2007; Ou et al., 2001, 2003; Thomas et al., 2000). These biologically based models were based upon the multistage carcinogenesis theory (Moolgavkar and Knudson, 1981). However, up to the present time, these models have not been linked with any pharmacokinetic models. The present study revealed that AUCLiver correlated well with the formation of liver GSTP foci. It would be of considerable utility in risk assessment if the present PBPK/PD model of PCB126 can be incorporated into the clonal growth model in order to give a more biologically relevant connection between pharmacokinetics to pharmacodynamics.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
National Institute of Occupational Safety and Health/Centers for Disease Control and Prevention grant 1 (RO1 OH07556); NIEHS Training grant 1 (T32 ES 07321); and U.S. Fulbright Foundation and Naresuan University, Thailand, scholarships.
|
ACKNOWLEDGMENTS |
|---|
We thank our colleagues in the Quantitative and Computational Toxicology Group for their excellent technical assistance.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Abraham K, Krowke R, Neubert D. Pharmacokinetics and biological activity of 2,3,7,8-tetrachlorodibenzo-p-dioxin. 1. Dose-dependent tissue distribution and induction of hepatic ethoxyresorufin O-deethylase in rats following a single injection. Arch. Toxicol. (1988) 62:359–368.[CrossRef][Web of Science][Medline]
Andersen ME, Mills JJ, Gargas ML, Kedderis L, Birnbaum LS, Neubert D, Greenlee WF. Modeling receptor-mediated processes with dioxin: Implications for pharmacokinetics and risk assessment. Risk Anal. (1993) 13:25–36.[CrossRef][Web of Science][Medline]
Belfiore CJ, Yang RS, Chubb LS, Lohitnavy M, Lohitnavy OS, Andersen ME. Hepatic sequestration of chlordecone and hexafluoroacetone evaluated by pharmacokinetic modeling. Toxicology (2007) 234:59–72.[CrossRef][Web of Science][Medline]
Borst P, Zelcer N, van de Wetering K. MRP2 and 3 in health and disease. Cancer Lett. (2006) 234:51–61.[CrossRef][Web of Science][Medline]
Brown RP, Delp MD, Lindstedt SL, Rhomberg LR, Beliles RP. Physiological parameter values for physiologically based pharmacokinetic models. Toxicol. Ind. Health. (1997) 13:407–484.
CDC. Third National Report on human exposure to environmental chemicals. National Center for Environmental Health (2005) Atlanta, GA.
Chu I, Villeneuve DC, Yagminas A, LeCavalier P, Poon R, Feeley M, Kennedy SW, Seegal RF, Hakansson H, Ahlborg UG, et al. Subchronic toxicity of 3,3',4,4',5-pentachlorobiphenyl in the rat. I. Clinical, biochemical, hematological, and histopathological changes. Fundam. Appl. Toxicol. (1994) 22:457–468.[CrossRef][Web of Science][Medline]
Chubb LS, Andersen ME, Broccardo CJ, Legare ME, Billings RE, Dean CE, Hanneman WH. Regional induction of CYP1A1 in rat liver following treatment with mixtures of PCB 126 and PCB 153. Toxicol. Pathol. (2004) 32:467–473.
Clewell HJ 3rd, Lee TS, Carpenter RL. Sensitivity of physiologically based pharmacokinetic models to variation in model parameters: Methylene chloride. Risk Anal. (1994) 14:521–531.[CrossRef][Web of Science][Medline]
Conolly RB, Andersen ME. Hepatic foci in rats after diethylnitrosamine initiation and 2,3,7,8-tetrachlorodibenzo-p-dioxin promotion: Evaluation of a quantitative two-cell model and of CYP 1A1/1A2 as a dosimeter. Toxicol. Appl. Pharmacol. (1997) 146:281–293.[CrossRef][Web of Science][Medline]
Conolly RB, Kimbell JS. Computer simulation of cell growth governed by stochastic processes: Application to clonal growth cancer models. Toxicol. Appl. Pharmacol. (1994) 124:284–295.[CrossRef][Web of Science][Medline]
Dean CE Jr, Benjamin SA, Chubb LS, Tessari JD, Keefe TJ. Nonadditive hepatic tumor promoting effects by a mixture of two structurally different polychlorinated biphenyls in female rat livers. Toxicol. Sci. (2002) 66:54–61.
Fisher JW, Campbell J, Muralidhara S, Bruckner JV, Ferguson D, Mumtaz M, Harmon B, Hedge JM, Crofton KM, Kim H, et al. Effect of PCB 126 on hepatic metabolism of thyroxine and perturbations in the hypothalamic-pituitary-thyroid axis in the rat. Toxicol. Sci. (2006) 90:87–95.
Gerloff T, Geier A, Stieger B, Hagenbuch B, Meier PJ, Matern S, Gartung C. Differential expression of basolateral and canalicular organic anion transporters during regeneration of rat liver. Gastroenterology (1999) 117:1408–1415.[CrossRef][Web of Science][Medline]
Hirono S, Nakagome I, Imai R, Maeda K, Kusuhara H, Sugiyama Y. Estimation of the three-dimensional pharmacophore of ligands for rat multidrug-resistance-associated protein 2 using ligand-based drug design techniques. Pharm. Res. (2005) 22:260–269.[CrossRef][Web of Science][Medline]
Ito N, Tamano S, Shirai T. A medium-term rat liver bioassay for rapid in vivo detection of carcinogenic potential of chemicals. Cancer Sci. (2003) 94:3–8.[CrossRef][Medline]
Iwase K, Hirono S. Estimation of active conformations of drugs by a new molecular superposing procedure. J. Comput. Aided Mol. Des. (1999) 13:499–512.[CrossRef][Web of Science][Medline]
Jedlitschky G, Hoffmann U, Kroemer HK. Structure and function of the MRP2 (ABCC2) protein and its role in drug disposition. Expert Opin. Drug Metab. Toxicol. (2006) 2:351–366.[CrossRef][Web of Science][Medline]
Jin X, Kennedy SW, Di Muccio T, Moon TW. Role of oxidative stress and antioxidant defense in 3,3',4,4',5-pentachlorobiphenyl-induced toxicity and species-differential sensitivity in chicken and duck embryos. Toxicol. Appl. Pharmacol. (2001) 172:241–248.[CrossRef][Web of Science][Medline]
Katynski AL, Vijayan MM, Kennedy SW, Moon TW. 3,3',4,4',5-Pentachlorobiphenyl (PCB 126) impacts hepatic lipid peroxidation, membrane fluidity and beta-adrenoceptor kinetics in chick embryos. Comp. Biochem. Physiol. C Toxicol. Pharmacol. (2004) 137:81–93.[CrossRef][Web of Science][Medline]
Kohn MC, Lucier GW, Clark GC, Sewall C, Tritscher AM, Portier CJ. A mechanistic model of effects of dioxin on gene expression in the rat liver. Toxicol. Appl. Pharmacol. (1993) 120:138–154.[CrossRef][Web of Science][Medline]
Kohn MC, Sewall CH, Lucier GW, Portier CJ. A mechanistic model of effects of dioxin on thyroid hormones in the rat. Toxicol. Appl. Pharmacol. (1996) 136:29–48.[CrossRef][Web of Science][Medline]
Lee HB, Blaufox MD. Blood volume in the rat. J. Nucl. Med. (1985) 26:72–76.
Lohitnavy M, Chubb LS, Lohitnavy O, Yang CC, Homberg J, Campain JA, Yang RS. Pharmacokinetic (PK)/Pharmacodynamic (PD) Relationship of PCB126 Under the Conditions of Modified Medium-Term Liver Bioassay. Toxicologist (2004) 78:271.
Lu Y, Lohitnavy M, Reddy MB, Lohitnavy O, Ashley A, Yang RS. An updated physiologically based pharmacokinetic model for hexachlorobenzene: Incorporation of pathophysiological states following partial hepatectomy and hexachlorobenzene treatment. Toxicol. Sci. (2006) 91:29–41.
Lu Y, Lohitnavy M, Reddy MB, Lohitnavy O, Eickman E, Ashley A, Gerjevic L, Xu Y, Conolly RB, Yang RSH. Quantitative analysis of liver GST-P Foci promoted by a chemical mixture of hexachlorobenzene and PCB 126: Implication of size-dependent cellular growth kinetics. Arch. Toxicol. (2007) 82:103–116.[CrossRef][Web of Science][Medline]
MacGregor JT, Collins JM, Sugiyama Y, Tyson CA, Dean J, Smith L, Andersen M, Curren RD, Houston JB, Kadlubar FF, et al. In vitro human tissue models in risk assessment: Report of a consensus-building workshop. Toxicol. Sci. (2001) 59:17–36.
Moolgavkar SH, Knudson AG Jr. Mutation and cancer: A model for human carcinogenesis. J. Natl. Cancer Inst. (1981) 66:1037–1052.[Web of Science][Medline]
NTP. NTP toxicology and carcinogenesis studies of 3,3',4,4',5-pentachlorobiphenyl (PCB 126) (CAS No. 57465-57428-8) in female Harlan Sprague-Dawley rats (Gavage Studies). Natl. Toxicol. Program Tech Rep. Ser. (2006) 4–246.
Ogiso T, Tatematsu M, Tamano S, Hasegawa R, Ito N. Correlation between medium-term liver bioassay system data and results of long-term testing in rats. Carcinogenesis (1990) 11:561–566.
Ou YC, Conolly RB, Thomas RS, Gustafson DL, Long ME, Dobrev ID, Chubb LS, Xu Y, Lapidot SA, Andersen ME, et al. Stochastic simulation of hepatic preneoplastic foci development for four chlorobenzene congeners in a medium-term bioassay. Toxicol. Sci. (2003) 73:301–314.
Ou YC, Conolly RB, Thomas RS, Xu Y, Andersen ME, Chubb LS, Pitot HC, Yang RS. A clonal growth model: Time-course simulations of liver foci growth following penta- or hexachlorobenzene treatment in a medium-term bioassay. Cancer Res. (2001) 61:1879–1889.
Parham FM, Kohn MC, Matthews HB, DeRosa C, Portier CJ. Using structural information to create physiologically based pharmacokinetic models for all polychlorinated biphenyls. Toxicol. Appl. Pharmacol. (1997) 144:340–347.[CrossRef][Web of Science][Medline]
Petzinger E, Geyer J. Drug transporters in pharmacokinetics. Naunyn Schmiedebergs Arch. Pharmacol. (2006) 372:465–475.[CrossRef][Web of Science][Medline]
Safe S, Bandiera S, Sawyer T, Robertson L, Safe L, Parkinson A, Thomas PE, Ryan DE, Reik LM, Levin W, et al. PCBs: Structure-function relationships and mechanism of action. Environ. Health Perspect. (1985) 60:47–56.[Web of Science][Medline]
Safe SH. Polychlorinated biphenyls (PCBs): Environmental impact, biochemical and toxic responses, and implications for risk assessment. Crit. Rev. Toxicol. (1994) 24:87–149.[Web of Science][Medline]
Shirai T. A medium-term rat liver bioassay as a rapid in vivo test for carcinogenic potential: A historical review of model development and summary of results from 291 tests. Toxicol. Pathol. (1997) 25:453–460.
Shitara Y, Horie T, Sugiyama Y. Transporters as a determinant of drug clearance and tissue distribution. Eur. J. Pharm. Sci. (2006) 27:425–446.[CrossRef][Web of Science][Medline]
Taub R. Liver regeneration: From myth to mechanism. Nat. Rev. Mol. Cell. Biol. (2004) 5:836–847.[CrossRef][Web of Science][Medline]
Thomas RS, Conolly RB, Gustafson DL, Long ME, Benjamin SA, Yang RS. A physiologically based pharmacodynamic analysis of hepatic foci within a medium-term liver bioassay using pentachlorobenzene as a promoter and diethylnitrosamine as an initiator. Toxicol. Appl. Pharmacol. (2000) 166:128–137.[CrossRef][Web of Science][Medline]
Tomer G, Ananthanarayanan M, Weymann A, Balasubramanian N, Suchy FJ. Differential developmental regulation of rat liver canalicular membrane transporters Bsep and Mrp2. Pediatr. Res. (2003) 53:288–294.[CrossRef][Web of Science][Medline]
Tsujishita H, Hirono S. CAMDAS: An automated conformational analysis system using molecular dynamics. Conformational Analyzer with Molecular Dynamics And Sampling. J. Comput. Aided Mol. Des. (1997) 11:305–315.[CrossRef][Web of Science][Medline]
Villanueva SS, Ruiz ML, Luquita MG, Sanchez Pozzi EJ, Catania VA, Mottino AD. Involvement of Mrp2 in hepatic and intestinal disposition of dinitrophenyl-S-glutathione in partially hepatectomized rats. Toxicol. Sci. (2005) 84:4–11.
Vos TA, Ros JE, Havinga R, Moshage H, Kuipers F, Jansen PL, Muller M. Regulation of hepatic transport systems involved in bile secretion during liver regeneration in rats. Hepatology (1999) 29:1833–1839.[CrossRef][Web of Science][Medline]
Xu YH, Dragan YP, Campbell HA, Pitot HC. STEREO: A program on a PC-Windows 95 platform for recording and evaluating quantitative stereologic investigations of multistage hepatocarcinogenesis in rodents. Comput. Methods Programs Biomed. (1998) 56:49–63.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||