Abstract
This study was designed to characterize breast cancer resistance protein (Bcrp) knockout Abcg2(−/−) rats and assess the effect of ATP-binding cassette subfamily G member 2 (Abcg2) deletion on the excretion and pharmacokinetic properties of probe substrates. Deletion of the target gene in the Abcg2(−/−) rats was confirmed, whereas gene expression was unaffected for most of the other transporters and metabolizing enzymes. Biliary excretion of nitrofurantoin, sulfasalazine, and compound A [2-(5-methoxy-2-((2-methyl-1,3-benzothiazol-6-yl)amino)-4-pyridinyl)-1,5,6,7-tetrahydro-4H-pyrrolo[3,2-c]pyridin-4-one] accounted for 1.5, 48, and 48% of the dose in the Abcg2(+/+) rats, respectively, whereas it was decreased by 70 to 90% in the Abcg2(−/−) rats. Urinary excretion of nitrofurantoin, a significant elimination pathway, was unaffected in the Abcg2(−/−) rats, whereas renal clearance of sulfasalazine, a minor elimination pathway, was reduced by >90%. Urinary excretion of compound A was minimal. Systemic clearance in the Abcg2(−/−) rats decreased 22, 43 (p < 0.05), and 57%, respectively, for nitrofurantoin, sulfasalazine, and compound A administered at 1 mg/kg and 27% for compound A administered at 5 mg/kg. Oral absorption of nitrofurantoin, a compound with high aqueous solubility and good permeability, was not limited by Bcrp. In contrast, the absence of Bcrp led to a 33- and 11-fold increase in oral exposure of sulfasalazine and compound A, respectively. These data show that Bcrp plays a crucial role in biliary excretion of these probe substrates and has differential effects on systemic clearance and oral absorption in rats depending on clearance mechanisms and compound properties. The Abcg2(−/−) rat is a useful model for understanding the role of Bcrp in elimination and oral absorption.
Introduction
Breast cancer resistance protein (BCRP), encoded by ATP-binding cassette subfamily G member 2 (ABCG2), is a member of the ATP-dependent transporter family that can actively extrude compounds out of cells. BCRP is expressed in many tissues including canalicular membranes of hepatocytes, apical membranes of enterocytes, apical membranes of kidney proximal tubule cells, blood-brain barrier endothelium, placenta, and lactating mammary gland (Maliepaard et al., 2001; Fetsch et al., 2006). Substrate specificity of BCRP is broad and includes chemotherapy agents, statins, antibiotics, carcinogens, and sulfate conjugates (Robey et al., 2007).
BCRP polymorphisms may contribute to interindividual variability of pharmacokinetics (PK), and BCRP inhibition may lead to drug-drug interactions in the clinic. Single-nucleotide polymorphisms (SNPs) in the ABCG2 gene are common and vary among ethnic groups, in particular, a nonsynonymous 421C>A SNP. This SNP was not detected in African Americans, but 25.9% of European Americans and 50 to 60% of Asian Americans had at least one variant allele (Zamber et al., 2003). ABCG2 421C>A SNP results in a lysine to glutamine acid change at codon 141 (Q141K), which leads to decreased protein expression level or reduced drug resistance to anticancer agents in transfected cells (Imai et al., 2002; Mizuarai et al., 2004; Morisaki et al., 2005; Tamura et al., 2006, 2007; Furukawa et al., 2009). The Q141K variant was associated with increased risk for gefitinib-induced diarrhea (Cusatis et al., 2006). Oral exposure of sulfasalazine and rosuvastatin in the subjects with the ABCG2 421AA genotype was approximately 2- to 3.5-fold higher than in the control subjects (ABCG2 421CC), although the effect on sulfasalazine exposure seems to be influenced by the dose and formulation (Zhang et al., 2006; Yamasaki et al., 2008; Keskitalo et al., 2009; Adkison et al., 2010). On the other hand, oral exposure of other known BCRP substrates (nitrofurantoin and pitavastatin) is not altered in the subjects with the ABCG2 421AA genotype (Ieiri et al., 2007; Adkison et al., 2008). Several additional SNPs, including V12M, Q126Stop, and P269S, affect expression or activities of the transporter, which may also have clinical implications (Kondo et al., 2004; Mizuarai et al., 2004; Tamura et al., 2006; Lee et al., 2007). Although few clinical drug-drug interactions attributed to BCRP inhibition have been reported, coadministration of curcumin, a BCRP inhibitor, increased sulfasalazine exposure by 3-fold in humans (Kusuhara et al., 2012). Therefore, it is important to assess BCRP's contribution to PK in drug discovery and development.
Abcg2(−/−) mice have been used to delineate the role of Bcrp in drug disposition. Biliary excretion of nitrofurantoin was abolished in Abcg2(−/−) mice (Merino et al., 2005). Likewise, biliary clearance (CL) of rosuvastatin and pitavastatin was decreased approximately 90% in the Abcg2(−/−) mice (Hirano et al., 2005; Kitamura et al., 2008). Plasma exposure of nitrofurantoin was 2- and 4-fold higher in the Abcg2(−/−) mice than in the Abcg2(+/+) mice after intravenous and oral administration, respectively (Merino et al., 2005). The change in sulfasalazine PK was more dramatic; exposure in the Abcg2(−/−) mice increased 13- and 111-fold after intravenous and oral administration, respectively (Zaher et al., 2006). However, there are potential species differences in BCRP. Although the ABCG2 gene is highly conserved across species (Robey et al., 2009) and no species differences in substrate specificity have been reported yet, BCRP expression is species-dependent. The intestinal mRNA level of ABCG2 in humans is lower than in other species (Bleasby et al., 2006). Although hepatic mRNA expression of Abcg2 is higher in dogs and humans than in mice, rats, and monkeys (Bleasby et al., 2006), the absolute amount of Bcrp in mouse liver as measured by mass spectrometry is 8-, 6-, 2-, and 4- fold higher than that in human, monkey, dog, and rat liver, respectively (Lai, 2009).
Rats are routinely used for PK and excretion studies. However, the lack of specific inhibitors presents challenges to the assessment of Bcrp contribution to PK in rats (Kalgutkar et al., 2009). Abcg2 knockout homozygous rats were developed recently by using targeted, CompoZr ZFN technology in Sprague-Dawley embryos (www.sageresearchmodels.com). This new model permits the examination of Bcrp's contribution to elimination and PK in the species most commonly used in drug discovery and development. The aims of this study were: 1) to determine whether the Abcg2(−/−) rat is an appropriate model for elucidating roles of Bcrp in drug disposition, and 2) to probe the effect of Abcg2 deletion on the excretion and PK properties of the BCRP substrates nitrofurantoin, sulfasalazine, and compound A [2-(5-methoxy-2-((2-methyl-1,3-benzothiazol-6-yl)amino)-4-pyridinyl)-1,5,6,7-tetrahydro-4H-pyrrolo[3,2-c]pyridin-4-one].
Materials and Methods
Materials.
Dulbecco's modified Eagle's medium, Hanks' balanced salt solution, phosphate-buffered saline, heat-inactivated fetal bovine serum, real-time quantitative polymerase chain reaction (RT-PCR) enzymes, and reagents were purchased from Invitrogen (Carlsbad, CA). RNA extraction and purification kits were from QIAGEN (Valencia, CA). Noncollagen-coated 24-well transwells (0.4-μm pore size) were purchased from Millipore Corporation (Billerica, MA). Compound A was synthesized at the Department of Medicinal Chemistry at Amgen (Cambridge, MA). All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO).
Madin-Darby canine kidney (MDCK) cells were purchased from the American Type Culture Collection (Manassas, VA). Transfection of MDCK cells with the human MDR1 gene (MDR1-MDCK), human ABCG2 gene (ABCG2-MDCK), rat Abcg2 gene (Abcg2-MDCK), or control vector (VC-MDCK) was conducted at Amgen (Thousand Oaks, CA).
Animals.
All animal procedures were conducted under protocols approved by the Amgen Institutional Animal Care and Use Committee (Cambridge, MA). Male Sprague-Dawley Abcg2(+/+) and Abcg2(−/−) rats, 6 to 7 weeks old, were obtained from Sigma Sage Laboratories (St. Louis, MO). The rats were housed in a temperature- and humidity-controlled environment subject to a 12-h light/dark cycle and had access to water and a standard laboratory rodent diet ad libitum. Animals were housed for 3 to 5 weeks before use.
Gene Expression of Transporters and Metabolizing Enzymes in Liver, Brain, Kidney, and Intestine.
Rats were euthanized by exposure to 100% CO2 gas. Brain, kidney, and liver were harvested and sectioned into 5-mm segments. Intestines were removed, cut lengthwise, and rinsed twice in phosphate-buffered saline. The intestine was then sectioned into segments of duodenum, jejunum, ileum, and large intestine. All tissue segments were weighed and then immersed in chilled RNAlater (Ambion, Austin, TX) at a volume of 1 ml per 100 mg of tissue. Samples were stored frozen at −80°C until analysis.
Brain tissues were homogenized gently in nine volumes of Hanks' balanced salt solution for isolation of the capillary vessel-enriched fraction. The brain homogenates were then mixed with dextran 70 at a final concentration of 15% and centrifuged at 5800g. The resulting pellet, containing the capillary vessel-enriched fraction, was used for RNA extraction.
Total RNA isolation from tissues and subsequent RT-PCR analysis were performed as described previously (Tchaparian et al., 2011). In brief, RNA was prepared by using QIAGEN RNA mini-kits and quantified by spectrophotometry (Thermo Fisher Scientific, Waltham, MA). Total RNA (1 μg) was reverse-transcribed according to the manufacturer's protocol by using random hexamer primers with the SuperScript III First-Strand Synthesis System for RT-PCR kit (Invitrogen). The resulting cDNA was treated with RNase H to remove residual RNA. Customized TaqMan Arrays using 384-Well Micro Fluidic Cards (Applied Biosystems, Foster City, CA) were used to determine the relative expression and profiling of the selected genes. The selected PCR primers were gene-specific and designed to span an exon-exon junction.
Quantitation of the transcription levels and statistical analysis using the parametric t test were performed as described previously (Tchaparian et al., 2011). Gene expression values were normalized to the level of 18S rRNA, glyceraldehyde-3-phosphate dehydrogenase, or β-actin. For each gene, the false discovery rate was estimated and used to determine the corresponding adjusted p value. Only genes exhibiting expression changes at a significant level (adjusted p < 0.05; fold change of 2) were considered to be differentially expressed.
Cyp3a activity in the pooled liver microsomes was measured by testosterone 6β-hydroxylation (Anderson et al., 1998). P-glycoprotein (Pgp) expression in the pooled liver homogenate was detected by Western blot analysis using a human MDR1 (D-11) antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) that cross-reacts with rodent Mdr1a/b (Tchaparian et al., 2011).
Bi-Directional Transport across MDCK cells.
Cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and hygromycin B (500 μg/ml), a selection agent for the vector. Transport studies were conducted in 24-well transwell plates (duplicate or triplicate) as described previously (Huang et al., 2010) with some modification. In brief, compounds were tested at 5 μM in the incubation buffer (Hanks' balanced salt solution containing 0.1% bovine serum albumin) 5 days postseeding. Transport studies were conducted at 37°C in the air with shaking (60 rpm) for 120 min.
For quantification of test compounds in the samples, calibration curve standards (ranging from 0.001 to 1 μM) were prepared from the 5-μM incubation solution. Aliquots from the apical receiver and donor chambers were diluted 2- and 5-fold with the incubation buffer, respectively. A mixture containing 0.1% formic acid, 66.6% acetonitrile, and 33.3% water was added to samples at a 4:1 ratio to precipitate protein. Samples were analyzed as described under LC-MS/MS Analysis. Detection limit ranged from 0.001 to 0.037 μM.
The apparent permeability (Papp), and efflux ratios were calculated according to the following equations (Huang et al., 2010): where A is the membrane surface area, C0 is the donor drug concentration at t = 0, and dQ/dt is the amount of drug transported within a given time period. PappB→A and PappA→B are the Papp values in the basolateral to apical and apical to basolateral direction, respectively. For Papp of <1 × 10−6 cm/s, specific values were not reported in consideration of LC-MS/MS detection sensitivity and variability.
Biliary Excretion and Urinary Excretion Study after Intravenous Administration.
Rats were anesthetized with ketamine (30 mg/kg) and dexmedetomidine (0.2 mg/kg) administered by intramuscular injection. The left femoral vein and artery were catheterized for compound administration and blood collection, respectively. Silastic catheters were implanted in the bile duct for bile collection and in the proximal duodenum for bile circulation and infusion of replacement fluids. The catheters were protected by a Covance infusion harness (Covance Research Products, Princeton, NJ) and connected to permit bile recirculation. The rats were housed in metabolic cages throughout the experiment for the collection of excreta.
After a recovery period of 2 to 3 days, rats were administered a single dose of test material by bolus injection into the femoral vein catheter, and the catheter was flushed with saline to ensure full delivery of the dose. Bile samples were collected from 0 to 0.5, 0.5 to 1, 1 to 2, 2 to 4, 4 to 6, 6 to 8, and 8 to 24 h (digoxin at 0.25 mg/kg; sulfasalazine and compound A at 1 mg/kg) or 0 to 0.5, 0.5 to 1, 1 to 2, 2 to 4, 4 to 8, and 8 to 24 h (nitrofurantoin and compound A at 5 mg/kg). Urine samples were collected from 0 to 8 h and 8 to 24 h postdose. A solution containing 25 mM taurocholic acid, 0.9% NaCl, and 0.05% KCl was infused through the duodenal catheter at a rate of 1 ml/h to replace bile salts, electrolytes, and fluids lost because of bile collection. Blood samples were collected and processed as described under PK Studies in Abcg2(+/+) and Abcg2(−/−) Rats.
For quantification of test compounds in the bile, samples were first diluted 200-fold with blank rat bile from Abcg2(+/+) rats. Standard curve and quality-control samples were prepared in blank rat bile. Diluted samples (20 μl) were extracted by the addition of 200 μl of solvent containing 90% methanol, 0.1% formic acid, and an internal standard followed by centrifugation. The supernatant was mixed with 0.1% trifluoroacetic acid in water at a ratio of 1:1 before LC-MS/MS analysis. Similar standard curve and quality-control samples were prepared in blank urine. Samples were extracted by addition of equal volume of a solvent containing 90% methanol, 0.1% formic acid, and an internal standard. Bile and urine samples were analyzed as described under LC-MS/MS Analysis.
PK Studies in Abcg2(+/+) and Abcg2(−/−) Rats.
Rats were administered a single dose of test material (formulated in dimethyl sulfoxide) by bolus injection into the femoral vein catheter. The catheter was flushed with saline to ensure full delivery of the dose. For oral administration, rats were given a single dose of test compound by orogastric gavage after an overnight fast. The oral doses were formulated as follows: digoxin (1 mg/kg; 0.1 mg/ml) in 0.1% Tween 80 and 2% hydroxypropylmethylcellulose (HPMC); nitrofurantoin (10 mg/kg;1 mg/ml) in 0.5% HPMC; sulfasalazine (10 mg/kg; 1 mg/ml) in 2% Tween 80 and 2% HPMC; and compound A (10 mg/kg;1 mg/ml) in 1% Tween 80 and 2% HPMC. Blood samples were collected via the femoral artery catheter at 0.05 (intravenously only), 0.25, 0.5, 1, 2, 4, 6, 8, and 24 h postdose.
Plasma was separated from blood cells by centrifugation and stored in a freezer that was maintained at approximately −70°C. Plasma samples (50 μl) were mixed with 300 μl of solvent containing 0.1% formic acid, 90% methanol (digoxin, nitrofurantoin, and compound A), or 150 μl of acetonitrile containing 0.1% formic acid (sulfasalazine), and an internal standard. After the mixture was centrifuged, the supernatant was transferred to a 96-well plate, which contained an equal volume of 0.1% trifluoroacetic acid in water, for LC-MS/MS analysis.
PK parameters, including the area under the concentration-time curve from time 0 to infinity (AUC0-inf), and clearance (CL), were calculated by using Small Molecules Discovery Assay Watson software (version 7.0.01; InnaPhase Corp., Philadelphia, PA). The maximum concentration observed (Cmax) and time of Cmax (Tmax) after oral administration were taken directly from the concentration-time data. Oral bioavailability (F) was calculated according to the following equation:
LC-MS/MS Analysis.
The LC-MS/MS system consisted of two LC-10AD HPLC pumps and a DGU-14A degasser (Shimadzu, Kyoto, Japan), a CTC PAL autoinjector (LEAP Technologies, Carrboro, NC), and an API4000 mass spectrometer (Applied Biosystems). Samples from in vitro experiments were injected onto a Sprite Armor C18 analytical column (20 × 2.1 mm, 10-μm pore size; Analytical Sales and Products, Pompton Plains, NJ) with a 0.5-μm polyether ether ketone guard filter. Analytes were separated by using a gradient solvent system consisting of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile) at a flow rate of 0.4 ml/min. The percentage of solvent B was increased in a linear fashion from 2 to 95% over 2 min.
Drug concentrations in extracted samples from in vivo studies were determined by using reversed-phase liquid chromatography with a solvent system consisting of solvent A (0.1% formic acid in water) and solvent B (acetonitrile containing 0.1% formic acid) at a flow rate of 0.55 ml/min. Samples were injected onto a Varian Pursuit C18 column (30 × 2 mm, 5 μm pore size) (Varian, Inc., Palo Alto, CA) and separated by using a gradient with a linear increase of solvent B from 10 to 90% with a total run time of 2.5 min for each sample. LC-MS/MS analysis was performed by using atmospheric pressure chemical ionization and multiple reactions monitoring in the positive ion mode.
Statistics.
For the in vivo studies that had n = 3, statistical differences were determined by unpaired t test, one-tailed, using Prism 5.0 software (GraphPad Software Inc., San Diego, CA). The significance level was p < 0.05.
Results
Gene Expression of Drug Transporters and Metabolizing Enzymes in Abcg2(−/−) Rats.
There was no obvious difference between Abcg2(+/+) and Abcg2(−/−) rats in terms of appearance and body weight. Gene expression analysis was conducted to assess mRNA expression of various transporters and metabolizing enzymes in the liver, brain, kidney, and intestine (Table 1). The absence of the corresponding mRNA confirmed deletion of the target gene in the Abcg2(−/−) rats. The Abcg2(+/+) and the Abcg2(−/−) rats showed comparable mRNA expression for most of the genes examined with a few exceptions. Expression of Mrp3 (brain, duodenum), Mrp6 (jejunum), OctN1 (jejunum), Asbt (ileum), Oatp2b1 (duodenum), and OctN2 (brain) increased, whereas the expression of Mdr1b (liver), Mrp2 (ileum), Oatp1a4 (brain), and Pept2 (jejunum) decreased in the Abcg2(−/−) rats. The magnitude of these changes was approximately 2- to 3-fold relative to the Abcg2(+/+) rats.
Cyp3a activity in the pooled liver microsomes was 0.99 and 1.28 nmol/min/mg protein for the Abcg2(+/+) and Abcg2(−/−) rats, respectively. Although hepatic mRNA expression of Mdr1b decreased modestly in the Abcg2(−/−) rats, Pgp expression as measured by densitometry of Western blots was unaffected (data not shown).
In Vitro Transport of Digoxin, Sulfasalazine, Nitrofurantoin, and Compound A.
Table 2 summarizes Papp values and efflux ratios of digoxin, nitrofurantoin, sulfasalazine, and proprietary compound A in the control cells and the cells expressing human Pgp, human BCRP, or rat Bcrp. No significant differences in transport were observed between human and rat BCRP for the compounds examined. Digoxin showed relatively high efflux in VC-MDCK cells, which was probably caused by the high level of endogenous Pgp (Kuteykin-Teplyakov et al., 2010). Efflux ratio of digoxin in the MDR1-MDCK cells was much higher than that in the control cells. In contrast, digoxin efflux in ABCG2-MDCK or Abcg2-MDCK was similar to VC-MDCK cells.
Nitrofurantoin and compound A (structure shown in Fig. 1) showed good permeability (average Papp values of both directions = 7.3 and 23.9 × 10−6 cm/s in the VC-MDCK cells, respectively). Efflux ratio of nitrofurantoin in the BCRP-expressing cells, but not Pgp-expressing cells, was much higher than that in the control cells, indicating that this compound was a BCRP-specific substrate. Compound A showed an efflux ratio of >35 in Pgp- or BCRP-expressing cells, indicating that it was an excellent substrate for both BCRP and Pgp. Sulfasalazine showed minimal efflux in BCRP- or Pgp-expressing cells with low permeability (<1 × 10−6 cm/s) (Table 2).
Excretion and PK Parameters of Digoxin in Abcg2(−/−) and Abcg2(+/+) Rats.
Digoxin was used as a probe substrate to assess Pgp activity in vivo. Approximately 6 and 20% of the dose was excreted into bile and urine as parent in either the Abcg2(+/+) or Abcg2(−/−) rats, respectively. Consistent with the excretion data, systemic plasma CL of digoxin was unaffected in the Abcg2(−/−) rats. Exposure and bioavailability of digoxin were comparable between the Abcg2(+/+) and Abcg2(−/−) rats after single oral administration at 1 mg/kg (p > 0.05) (Table 3).
Biliary and Urinary Excretion of BCRP Probe Substrates in Bile Duct-Catheterized Rats after Intravenous Administration.
In the Abcg2(+/+) rats, biliary excretion accounted for 48% of the dose for sulfasalazine and compound A, but was a minor elimination pathway for nitrofurantoin (1.6% of the dose) (Fig. 2). Biliary excretion or CL of the three substrates decreased approximately 70 to 90% in the Abcg2(−/−) rats (Table 4). Residual biliary excretion of sulfasalazine and compound A accounted for 10 to 15% of the dose in the Abcg2(−/−) rats (Fig. 2). Compound A was administered at 1 and 5 mg/kg; the percentage of the dose excreted into bile was similar between the two doses in the Abcg2(+/+) or Abcg2(−/−) rats (Fig. 2). Biliary CL of compound A in the Abcg2(+/+) rats decreased from 0.89 to 0.23 l/h/kg with the dose increased from 1 to 5 mg/kg, although the difference was not statistically significant because of relatively large intersubject variability (p > 0.05). In contrast, biliary CL of compound A was dose-independent in the Abcg2(−/−) rats (Table 4).
Urinary excretion was quantitatively an important elimination pathway for nitrofurantoin, whereas renal CL of sulfasalazine and compound A was a minor pathway or minimal. Renal CL in the Abcg2(−/−) rats was unaffected for nitrofurantoin and compound A, but was decreased by more than 90% for sulfasalazine (Table 4).
Plasma Pharmacokinetics in Bile Duct-Intact Rats after Intravenous and Oral Administration.
To assess the effect of Bcrp on CL, intravenous PK studies were conducted for the three probe substrates in bile duct-intact rats. Systemic plasma CL was unaffected for nitrofurantoin, but decreased significantly (40%; p < 0.05) for sulfasalazine in the Abcg2(−/−) rats compared with the Abcg2(+/+) rats (Table 5). Intravenous PK of compound A was assessed at two doses (1 and 5 mg/kg). The absence of Bcrp led to a 2.6-fold increase in the systemic exposure of compound A at 1 mg/kg, but had no effect on its exposure at 5 mg/kg (Fig. 3). In the Abcg2(+/+) rats, CL of compound A decreased from 2.7 to 1.1 L/h/kg (p < 0.01; unpaired t test between the 1 and 5 mg/kg dose groups) when the dose was increased from 1 to 5 mg/kg. In contrast, CL in the Abcg2(−/−) rats was dose-independent (Table 5).
After oral administration, nitrofurantoin was well absorbed, exhibiting comparable oral bioavailability between the Abcg2(+/+) and Abcg2(−/−) rats (Table 6; Fig. 4). In contrast, AUC0-inf, Cmax, and F in the Abcg2(−/−) rats increased 33-, 21-, and 17-fold, respectively, for sulfasalazine and 10-, 5-, and 4-fold, respectively, for compound A. However, oral bioavailability of sulfasalazine and compound A remained low in the Abcg2(−/−) rats. The Tmax of sulfasalazine increased in the Abcg2(−/−) rats (Table 6).
Discussion
Potential changes in the expression of other transporters and metabolizing enzymes, which may confound data interpretation, are concerns for the use of knockout animals in PK studies. The present study confirmed the deletion of the target gene in the Abcg2(−/−) rats, while leaving gene expression unaffected for most of the other transporters and metabolizing enzymes. mRNA expression of some transporters changed modestly in the Abcg2(−/−) rats including decreases in Mdr1b in the liver and increases in Oatp2b1 in the duodenum. Although mRNA expression is informative, it does not always reflect protein level and functional activity (Gry et al., 2009). In vivo Pgp activity assessed by biliary and urinary excretion and PK parameters of digoxin, a Pgp probe substrate, was similar between the Abcg2(+/+) and Abcg2(−/−) rats. Consistent with hepatic Cyp3a and Mrp2 mRNA levels, Cyp3a activity in liver microsomes measured by testosterone metabolism and Mrp2 activity evaluated by biliary CL of carboxydichlorofluorescein (Zamek-Gliszczynski et al., 2012), an Mrp2 probe substrate, were unaffected in the Abcg2(−/−) rats. Furthermore, basal bile flow was similar between the Abcg2(+/+) (25.7 ± 5.8 ml/24 h; n = 17) and Abcg2(−/−) rats (24.5 ± 6.5 ml/24 h; n = 17), as opposed to decreased bile flow in multiple drug resistance protein 2-deficient rats (Huang et al., 2000). These data suggest that Abcg2(−/−) rats have an appropriate degree of selectivity for assessing the contribution of Bcrp transport to excretion and CL.
Nitrofurantoin, sulfasalazine, and compound A were selected as BCRP probe substrates for in vivo evaluation because of their different characteristics in terms of permeability, Pgp, and BCRP substrate status. Nitrofurantoin exhibits good permeability and is a substrate for BCRP, but not for Pgp (Merino et al., 2005). Compound A was an excellent substrate for both Pgp and BCRP with high passive permeability. Sulfasalazine, a poorly permeable compound (Mahar Doan et al., 2002), has been identified as a BCRP substrate in Caco-2 cells by using the BCRP inhibitor 3-(6-isobutyl-9- methoxy-1,4-dioxo-1,2,3,4,6,7,12,12a-octahydropyrazino(1′,2′-1,6)pyrido(3,4-b)indol-3-yl)propionic acid tert-butyl ester (KO143) and membrane vesicles expressing BCRP (Jani et al., 2009; Lin et al., 2011). Papp values of sulfasalazine across BCRP-MDCK or Bcrp-MDCK cells were low in both directions, similar to those of atenolol, a marker of paracellular diffusion, suggesting that sulfasalazine may not be able to penetrate into cells in the absence of uptake transporters. The lack of sulfasalazine efflux in these BCRP-expressing cells probably was caused by its inability to access BCRP.
The deletion of Abcg2 had different effects on biliary and urinary excretion of the probe substrates. Recoveries of parent compounds in the bile from the Abcg2(−/−) rats decreased substantially, indicating that Bcrp played a key role in biliary excretion of the three compounds in rats. Although both Bcrp and Pgp transported compound A efficiently in vitro, Bcrp played a predominant role in the biliary excretion of compound A. In contrast, Bcrp did not seem to play a primary role in urinary excretion of compound A and nitrofuranoin in rats. Urinary excretion of compound A was minimal in both the Abcg2(+/+) and Abcg2(−/−) rats. Nitrofurantoin urinary excretion did not decrease in the Abcg2(−/−) rats. Its unbound renal CL (17.4 ml/min/kg; calculated based on an unbound free fraction of 0.43 in rat plasma) in the Abcg2(−/−) rats exceeded the glomerular filtration rate (5.2 ml/min/kg) (Davies and Morris, 1993; Kari et al., 1997), suggesting that transporters other than Bcrp are involved in active secretion of nitrofurantoin in rats. Renal CL of sulfasalazine, although a minor elimination pathway, was abolished in the Abcg2(−/−) rats, suggesting a possible role of Bcrp in urinary excretion in rats.
The impact of Abcg2 deletion on systemic CL depended on whether Bcrp-mediated efflux was a major CL mechanism. Consistent with the biliary excretion data, the absence of Bcrp led to a 40% reduction in systemic CL of sulfasalazine. As expected, the substantial decrease in biliary CL of nitrofurantoin did not result in a significant change in systemic CL because biliary excretion was a minor elimination pathway. It is noteworthy that plasma and biliary CL of compound A were dose-dependent in the Abcg2(+/+) rats, but not in the Abcg2(−/−) rats. Saturation of Bcrp-mediated biliary CL could be a mechanism for the nonlinear PK observed in the Abcg2(+/+) rats. An additional mechanism could be the saturation of Bcrp-mediated intestinal secretion, which was not measured in this study. Biliary excretion of compound A was a major elimination pathway and was decreased substantially in the Abcg2(−/−) rats after intravenous administration at both 1 and 5 mg/kg. However, the effect of Bcrp on systemic CL was observed only at the dose of 1 mg/kg. Further investigation is needed to understand the mechanisms for the differential effect of Bcrp on biliary excretion and systemic CL of compound A at the dose of 5 mg/kg.
Abcg2 deletion had differential effects on oral absorption of the probe substrates, depending on their permeability and solubility. As reported previously (www.sageresearchmodels.com; Zamek-Gliszczynski et al., 2012), the absence of Bcrp led to a profound increase in the oral exposure of sulfasalazine, a poorly permeable and poorly soluble compound (Benet et al., 2011). Oral absorption of compound A, a highly permeable compound with low solubility (solubility in phosphate-buffered saline = 5 μg/ml; unpublished Amgen data), was also increased significantly in the Abcg2(−/−) rats. In contrast, Bcrp did not limit oral absorption of nitrofurantoin, which showed good permeability (average Papp of the two directions = 7.3 × 10−6 cm/s in VC-MDCK cells) and good solubility (solubility in water = 190 μg/ml; Kari et al., 1997). The present results extend the analysis of Wu and Benet (2005) to drugs subject to Bcrp-mediated efflux in rats; transporters may limit oral absorption of compounds with poor solubility or low permeability, but have minimal effect on oral absorption of compounds with good solubility and permeability. Oral bioavailability of sulfasalazine and compound A was still very low in the Abcg2(−/−) rats. In addition to Bcrp and poor solubility, Pgp and poor permeability could be limiting factors in the oral absorption of compound A and sulfasalazine.
BCRP contribution to nitrofurantoin PK seemed comparable between humans and rats. Similar to rats, nitrofurantoin is well absorbed in humans (Hoener and Patterson, 1981); oral exposure and urinary excretion of nitrofurantoin is unaffected in the subjects with the ABCG2 421 AA genotype (Adkison et al., 2008). On the other hand, Abcg2 had a more pronounced effect on oral exposure of sulfasalazine in rats relative to humans. ABCG2 421C>A SNP led to a 2- to 3.5-fold increase in oral exposure of sulfasalazine in humans (Urquhart et al., 2008; Yamasaki et al., 2008; Adkison et al., 2010), as opposed to a 33-fold increase observed in the Abcg2(−/−) rats. The reason for the species difference in apparent BCRP contribution to the oral exposure of sulfasalazine remains to be delineated. One explanation could be species difference in BCRP expression. Other potential causes include 1) species differences in metabolizing enzymes and other transporters that may be involved in its oral absorption and elimination, and 2) ABCG2 421C>A SNP may retain partial transport activity for some substrates (Morisaki et al., 2005; Tamura et al., 2006).
In summary, Bcrp plays a crucial role in the biliary excretion of the probe substrates in rats and exhibits differential effects on systemic CL and oral absorption, depending on CL mechanisms and compound properties. Abcg2(−/−) rats are better suited to studies that require serial bile and blood collection than Abcg2(−/−) mice. An unexplored area is the utility of this model for probing the role of Bcrp in brain penetration. The Abcg2(−/−) rat is a useful model for understanding the role of BCRP in elimination and oral absorption.
Authorship Contributions
Participated in research design: Huang, Colletti, Wong, and Jin.
Conducted experiments: Huang, Be, Tchaparian, Colletti, Roberts, Langley, and Ling.
Performed data analysis: Huang, Be, Tchaparian, and Jin.
Wrote or contributed to the writing of the manuscript: Huang, Colletti, Wong, and Jin.
Acknowledgments
We thank Brett Janosky for contributions to in vitro transport data; the staff at the Department of Medicinal Chemistry, Amgen (Cambridge, MA) for compound A synthesis; Drs. Jasmine Lin and Zhiyang Zhao for scientific discussions and critical review of the manuscript; and Dr. Jiunn Lin for critical comments.
Footnotes
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS:
- BCRP/Bcrp
- human/rat breast cancer resistance protein
- ABCG2/Abcg2
- human/rat ATP-binding cassette subfamily G member 2
- AUC0-inf
- area under the curve from time 0 to infinity
- CL
- clearance
- Cmax
- maximum concentration observed after oral administration
- Tmax
- time of Cmax
- F
- oral bioavailability
- HPMC
- hydroxypropylmethylcellulose
- LC-MS/MS
- liquid chromatography-tandem mass spectrometry
- MDCK
- Madin-Darby canine kidney
- MDR
- multidrug resistance
- Papp
- apparent permeability
- Pgp
- P-glycoprotein
- PK
- pharmacokinetics
- RT-PCR
- real-time quantitative polymerase chain reaction
- SNP
- single-nucleotide polymorphism
- VC
- vector control
- B→A
- basolateral to apical
- A→B
- apical to basolateral
- KO143
- 3-(6-isobutyl-9-methoxy-1,4-dioxo-1,2,3,4,6,7,12,12a-octahydropyrazino(1′,2′-1,6)pyrido(3,4-b)indol-3-yl) propionic acid tert-butyl ester.
- Received May 31, 2012.
- Accepted July 24, 2012.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics