Abstract
Organic anion-transporting polypeptide (OATP) 1A2 has the potential to be a target for central nervous system drug delivery due to its luminal localization at the human blood-brain barrier and broad substrate specificity. We found OATP1A2 mRNA expression in the human brain to be comparable to breast cancer resistance protein and OATP2B1 and much higher than P-glycoprotein (P-gp), and confirmed greater expression in the brain relative to other tissues. The goal of this study was to establish a model system to explore OATP1A2-mediated transcellular transport of substrate drugs and the interplay with P-gp. In vitro (human embryonic kidney 293 cells stably expressing Oatp1a4, the closest murine isoform) and in vivo (naïve and Oatp1a4 knock-out mice) studies with OATP1A2 substrate triptan drugs demonstrated that these drugs were not Oatp1a4 substrates. This species difference demonstrates that the rodent is not a good model to investigate the active brain uptake of potential OATP1A2 substrates. Thus, we constructed a novel OATP1A2 expressing Madin-Darby canine kidney (MDCK) II wild type and an MDCKII-multidrug resistance protein 1 (MDR1) system using BacMam virus transduction. The spatial expression pattern of OATP1A2 after transduction in MDCKII-MDR1 cells was superimposed to P-gp, confirming apical membrane localization. OATP1A2-mediated uptake of zolmitriptan, rosuvastatin, and fexofenadine across monolayers increased with increasing OATP1A2 protein expression. OATP1A2 counteracted P-gp efflux for cosubstrates zolmitriptan and fexofenadine. A three-compartment model incorporating OATP1A2-mediated influx was used to quantitatively describe the time- and concentration-dependent apical-to-basolateral transcellular transport of rosuvastatin across OATP1A2 expressing the MDCKII monolayer. This novel, simple and versatile experimental system is useful for understanding the contribution of OATP1A2-mediated transcellular transport across barriers, such as the blood-brain barrier.
Introduction
The blood-brain barrier (BBB) presents a major hurdle to the delivery of drugs to the central nervous system (CNS). The capillary endothelium of the BBB is among the most restrictive barriers, with tight junctions that preclude paracellular distribution as well as efflux transporters that effectively limit transcellular passage of more permeable xenobiotics. To enable brain uptake of essential polar nutrients, highly specialized uptake systems are expressed at the BBB, for example, for glucose and amino acids. Although the role of efflux transporters in protecting the brain from exposure to potential toxins is well appreciated, the involvement of BBB uptake transporters in the brain penetration of drugs is poorly understood. Human organic anion-transporting polypeptide (OATP) 1A2 (rodent Oatp1a4) has emerged as a potentially important BBB uptake transporter for drugs (Urquhart and Kim, 2009). An indication of the potential of OATP1A2 as a brain drug delivery target comes from studies with statins that are OATP substrates and demonstrate neuroprotection in hypoxia and inflammatory disorders, including multiple sclerosis (Sierra et al., 2011; Ciurleo et al., 2014).
OATP1A2 (solute carrier family of the organic anion-transporting polypeptides 1A2) is a member of the sodium-independent uptake transporter family that was initially reported to be exclusively expressed at human brain microvessels and was not detected in astrocytes or neurons (Gao et al., 2000; Lee et al., 2005). It was later confirmed to be expressed on the luminal membrane of the BBB in tumorous and healthy adjacent tissue (Bronger et al., 2005), although a more recent paper reports expression in neurons as well (Gao et al., 2014). OATP1A2 transports amphipathic substrates, including bile salts, thyroid hormones, steroid conjugates, organic dyes, and anionic oligopeptides as well as xenobiotics (Franke et al., 2009). The substrate spectrum of OATP1A2 includes neuroactive substrates (i.e., opioid analgesic peptides), drugs with CNS side effects (e.g., the antibacterial levofloxacin, which may cause seizures, toxic psychoses, increased intracranial pressure, and CNS stimulation), and/or toxicities (e.g., methotrexate may cause mild white-matter changes or severe CNS demyelination and encephalopathy) (Gao et al., 2000; Badagnani et al., 2006; Maeda et al., 2007). There is no rodent ortholog reported for OATP1A2. Oatp1a4 and Oatp1c1 are two major isoforms expressed and enriched at the mouse BBB (Cheng et al., 2005; Mayerl et al., 2012). The closest counterpart to OATP1A2 is Oatp1a4 (Slco1a4), which shares a 72% gene homology. Unlike OATP1A2, Oapt1a4 is expressed at both the apical and basolateral membranes of the microvascular endothelium of the BBB (Gao et al., 1999; Ose et al., 2010). Oatp1a4 has a wide substrate spectrum that includes amphipathic organic anions, such as 17β-estradiol-17β-D-glucuronide (E217βG) and statins (Kikuchi et al., 2004), prostaglandin E1 (Taogoshi et al., 2005), and morphine-6-glucuronide (Bourasset et al., 2003). Oatp1a4 functional expression at the rat BBB increases with pain and correlates with taurocholate transport across the BBB (Ronaldson et al., 2011). An earlier study from our laboratory showed that marketed triptans were substrates for OATP1A2 using a BacMam2-OATP1A2 transduced human embryonic kidney (HEK) 293 system, which likely contributed to their ability to traverse the BBB (Cheng et al., 2012). We later confirmed naratriptan to be an OATP1A2 substrate as well (Km = 13.7 µM), using the same OATP1A2 expressing HEK293 system (data not published). Understanding the difference in the expression of these two transporters at the BBB could improve predictions of CNS permeability of drug candidates.
Discovery-stage studies to assess CNS uptake potential are typically conducted in vitro using Madin-Darby canine kidney (MDCK) cells transfected with human P-glycoprotein (P-gp) and in vivo in rodents to quantify brain distribution, as species difference in terms of P-gp substrate specificity and expression are minimal (Feng et al., 2008; Uchida et al., 2011b). Using in vitro human P-gp and rodent in vivo data for evaluating the CNS penetration potential of discovery compounds with > 1% brain free fraction, we found cases where compounds with low P-gp efflux activity were not taken up into the rodent brain, whereas some P-gp and/or breast cancer resistance protein (BCRP) substrates were detected at appreciable concentrations. Although the former could be due to additional efflux transporters (e.g., MRP4 and MRP5), the latter is most likely due to the presence of BBB uptake transporter(s) or low fraction of P-gp and/or BCRP contribution relative to passive permeability in preventing the CNS entry of the compounds (Hsiao and Unadkat, 2014). In this study, we have focused on OATP1A2 to evaluate the potential for this uptake transporter as a target for enhancing transcellular permeability in vitro. We have established novel methodologies using the BacMam virus and quantitatively characterized the role the OATP1A2 transporter plays in the uptake of its substrates in both wild-type (WT) and P-gp transfected cells.
Materials and Methods
Zolmitriptan, almotriptan, sumatriptan, rizatriptan, and naratriptan were purchased from Toronto Research Chemicals (Toronto, Canada). Rosuvastatin calcium salt was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Fexofenadine hydrochloride and atenolol were purchased from Sigma (St. Louis, MO), and naringin was purchased from TCI (Tokyo, Japan). [3H]taurocholate acid (TCA) and [3H]E217βG were purchased from PerkinElmer (Boston, MA). Cell culture reagents were purchased from Invitrogen (Carlsbad, CA). All other reagents used were of bioanalytical grade or higher. For OATP1A2 quantification, a peptide (EGLETNADIIK) standard was purchased from New England Peptides (Boston, MA). The corresponding stable isotope labeled ([13C615N2]lysine) internal standard was obtained from Thermo Fisher Scientific (Rockford, IL). The ProteoExtract native membrane protein extraction kit was procured from Calbiochem (Temecula, CA). The protein quantification bicinchoninic acid assay kit, dithiotrietol, iodoacetamide, and mass spectrometry (MS) grade trypsin were purchased from Pierce Biotechnology (Rockford, IL).
Determination of the Brain-to-Plasma Concentration Ratios of Sumatriptan, Naratriptan, and Zolmitriptan after Intravenous Infusion in Wild-Type and Oatp1a4 (−/−) Mice.
Animal experiments in this study were performed according to the guidelines provided by the Institutional Animal Care Committee (Graduate School of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan). Oatp1a4 (−/−) mice were obtained from Deltagen (San Carlos, CA), and their generation was described previously (Ose, et al., 2010). WT C57BL/6J mice were supplied by Oriental Yeast Co., Ltd. (Tokyo, Japan). All mice were maintained under standard conditions, with a 12-hour reverse dark/light cycle. Food and water were available ad libitum.
Male C57BL/6J mice and Oatp1a4 (−/−) mice (10–18 weeks) weighing approximately 25–35 g were used for the experiments. Under pentobarbital anesthesia (30 mg/kg), the jugular vein was cannulated with a polyethylene-10 catheter for the administration of drugs. The mice then received a constant intravenous infusion of sumatriptan, naratriptan, or zolmitriptan at a rate of 0.335, 0.288, and 0.313 µmol/h per kg, respectively. These compounds were dissolved in saline. Blood samples were collected from the jugular vein at 30 and 60 minutes after treatment, and the brain and liver were excised immediately after blood collection at 60 minutes. Plasma specimens were obtained by centrifugation of the blood samples (10,000g). The plasma, brain, and liver concentrations of sumatriptan, naratriptan, and zolmitriptan were determined using liquid chromatography (LC)-MS/MS analysis.
Determination of Cellular Uptake of Triptans Using a Stable Transfectant of Mouse Oatp1a4 in HEK293 Cells.
Construction of the stable transfectant of Oatp1a4 in HEK293 cells was as described previously (Ose et al., 2010). Uptake was initiated by the addition of the triptans to the incubation buffer after cells had been washed twice and preincubated at 37°C for 10 minutes in Krebs-Henseleit buffer. Uptake was terminated at a specified time by the addition of buffer at 4°C, and cells were washed three times. The amount of triptans associated with the cells and medium was determined by LC-MS/MS analysis. The protein concentrations in the aliquots of the cell lysate were determined using Lowry’s method. Triptan uptake was assessed from the cell-to-medium ligand concentration ratio, which was calculated as the concentration of triptans associated with the cells divided by that in the medium.
Construction of OATP1A2 Expressing MDCKII-WT or MDCKII-MDR1 Cells.
The generation of the recombinant BacMam2-OATP1A2 baculovirus (second generation of the BacMam vector) was described previously (Cheng et al., 2012). Polarized Madin-Darby canine kidney (MDCKII-WT or MDCKII-MDR1) cells were used for the in vitro transport studies and were obtained from the Netherlands Cancer Institute (Amsterdam, the Netherlands). Cell cultures and transport studies were conducted according to the procedure described previously (Rautio et al., 2006). At an appropriate multiplicity of infection (MOI), BacMam2-OATP1A2 was administered to the apical side of 24-well Transwell inserts (Millipore, Bedford, MA) 24 hours post-MDCKII cell seeding. The Transwell insert had a surface area of 0.33 cm2 as specified by manufacturer. The cells were cultured for another 48 hours before the transport experiments.
Transport Studies across BacMam2-OATP1A2 Transduced or Nontransduced MDCKII-WT or MDCKII-MDR1 Cells.
On the day of the transport experiments, donor solutions were prepared by diluting test compounds in transport medium (Dulbecco’s modified Eagle medium supplemented with 4500 mg/l D-glucose, L-glutamine, and 25 mM HEPES, but without sodium pyruvate and phenol red, pH 7.4). The receiver solution was the transport medium. The transport of the test compounds was measured in both directions [apical to basolateral (A→B) and basolateral to apical (B→A)]. Lucifer yellow was used as a paracellular marker to determine the integrity of the MDCKII monolayer, and its concentration was measured using a SpectraMax Gemini cytofluorimeter (Molecular Devices, Sunnyvale, CA) set to an excitation wavelength of 430 nm and an emission wavelength of 540 nm. The permeability at pH 7.4 (Pexact) for the test compounds across the MDCKII monolayer was determined by the method described previously (Tran et al., 2004).
Membrane Localization of OATP1A2 in BacMam2-OATP1A2 Transduced MDCKII-MDR1 Monolayers by Immunocytochemical Staining.
The BacMam2-OATP1A2 transduced (MOI = 50) MDCKII-MDR1 monolayers were fixed in 4% paraformaldehyde on Transwell filter membranes (Millipore, Bedford, MA) for 15 minutes at room temperature. The fixed monolayers were subsequently incubated in 0.2% Triton X-100 for 2 minutes at room temperature, washed three times with phosphate-buffered saline (PBS), and incubated with 10% normal goat serum in PBS to block nonspecific antibody binding. The monolayers were washed three times with PBS and incubated overnight at 4°C in a 1:200 dilution of rabbit anti-OATP1A2 antiserum (Santa Cruz Biotechnology) and a 1:50 dilution of mouse anti-MDR1 antibody (Sigma). After incubation, the monolayers were washed three times with PBS and then incubated for 1 hour at room temperature in a 1:1000 dilution of secondary antibodies (Alexa Fluor 546–conjugated anti-rabbit IgG and Alexa Fluor 488–conjugated anti-mouse IgG; Invitrogen). Subsequently, the monolayers were stained with diamidinophenylindole (Sigma-Aldrich, St. Louis, MO) for 2 minutes at room temperature. Finally, the MDCKII monolayers were transferred and mounted on glass slides and imaged by a Nikon A1R inverted confocal microscope with a 60× oil immersed lens (Plan Apo VC N.A. 1.40; Nikon, Tokyo, Japan). Images were captured and analyzed by NIS Element AR software (Nikon, Tokyo, Japan). Fluorescence intensity along the lines from the apical to basal side were normalized by the averaged intensity and plotted by Microsoft Excel software (Seattle, WA).
Quantification of Test Compounds by LC-MS/MS.
Quantification of zolmitriptan, rosuvastatin, and fexofenadine in MDCKII transport samples was performed by a Waters ACQUITY ultra-performance LC (UPLC) system coupled with a Waters Xevo TQ mass spectrometry system (Waters, Milford, MA). Detection of atenolol in MDCKII transport samples was performed by a Waters ACQUITY UPLC system coupled with an AB Sciex 5000 triple-quadrupole mass spectrometer (AB Sciex, Foster City, CA). The samples were processed by deproteination, with the same volume of acetonitrile containing the appropriate internal standard. The chromatographic separation was achieved on a Waters ACQUITY UPLC ethylene bridged hybrid C18 (50 × 2.1 mm, 1.7 µm) analytical column at 40°C using a gradient of aqueous and organic mobile phase at a flow rate of 600 µl/min. Run time for each compound was 2.5 minutes. Key mass spectrometric settings were optimized to yield the best sensitivity for each test compound. The precursor ion transitions to the strongest intensity product ions were 288.2→58.1 for zolmitriptan, 482.3→258.3 for rosuvastatin, 502.3→466.5 for fexofenadine, and 267.1→190.1 for atenolol.
Bioanalysis of triptans in in vitro HEK293 and in vivo mouse plasma, brain, and liver samples was performed by an AB Sciex Qtrap 5500 mass spectrometer equipped with a Prominence UPLC system (Shimadzu, Kyoto, Japan). The chromatographic separation was achieved on an Atlantis T3 (3 µm, 2.1 mm × 50 mm; WATERS, Tokyo, Japan) at 40°C using a gradient of aqueous (0.1% formic acid) and organic mobile phase (acetonitrile) at a flow rate of 400 µl/min (0 minutes, 3%; 0.3 minutes, 3%; 2.2 minutes, 90%; 3 minutes, 90%; 3.1 minutes, 3%). The precursor ion transitions to the strongest intensity product ions were 296.2→157.1 for sumatriptan, 335.8→58.1 for almotriptan, 335.7→98.1 for naratriptan, 269.7→58.1 for rizatriptan, and 288.1→58.0 for zolmitriptan.
Quantitative Real-Time Reverse-Transcription Polymerase Chain Reaction.
Total RNA was extracted from BacMam2-OATP1A2 transduced or nontransduced MDCKII-WT or MDCKII-MDR1 cells using the Qiagen RNeasy mini kit according to the manufacturer’s instructions (Valencia, CA). Human total RNA samples were purchased from Clontech (Palo Alto, CA). The RNA panel for the 18 human tissues contained different numbers of individuals between 14 and 68 years of age. The lowest number of individuals was one male or female Caucasians for brain (whole), liver, kidney, and colon with mucosa, and the highest number was 64 male/female Caucasians for the thyroid gland. Five hundred nanograms of aforementioned RNA were reversely transcribed to cDNA in a final volume of 20 µl using the Qiagen Omniscript reverse transcription kit.
The SYBR Green polymerase chain reaction (PCR) master mix (Applied Biosystems, Foster City, CA) was used for quantitative gene expression analysis on an ABI 7900HT fast real-time PCR system (Applied Biosystems), with primers as detailed in Table 1. Fifteen microliter reaction mixtures in a MicroAmp fast optical 384-well reaction plate (Applied Biosystems) contained 0.5 µl of cDNA, 7.5 µl of SYBR Green PCR master mix, and a 75 nM forward and reverse primer. Cycling conditions were 2 minutes at 50°C and 10 minutes at 95°C, followed by 40 cycles of alternating 15 seconds at 95°C and 1 minute at 60°C. Cycle time (Ct) values for OATP1A2, OATP2B1, P-gp, and BCRP genes were first normalized with that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in the same sample (ΔCt = CtOATP1A2−CtGAPDH), and then relative gene expression was determined by eq. 2–ΔCt. The relative expression fold difference of OATP1A2 as compared with other genes (BCRP, P-gp, and OATP2B1) was determined by eq. 2–ΔΔCt (ΔΔCt = ΔCtOATP1A2 – ΔCtother gene).
Protein Quantification of OATP1A2 by LC-MS/MS.
Protein quantification of OATP1A2 in OATP1A2-transduced MDCKII membrane extracts was conducted by mass spectrometry–based targeted proteomics using validated LC-MS/MS methods. Briefly, the cells were isolated from the transwell filters using extraction buffer I of the ProteoExtract membrane extraction kit. The total membrane proteins were isolated and digested using the procedure outlined before (Prasad et al., 2014). A triple-quadrupole LC-MS instrument [Xevo TQ-S coupled to ACQUITY UPLC (Waters) was used in the electrospray ionization positive ionization mode]. Approximately 2 µg of the trypsin digest (5 µl) was injected onto the column (Kinetex 1.7 µm, C18 100A; 100 × 2.1 mm; Phenomenex, Torrance, CA) and eluted at 0.3 ml/min. A mobile phase consisting of water containing 0.1% formic acid (A) and acetonitrile containing 0.1% formic acid (B) was employed. A flow rate of 0.3 ml/min was used, with elution starting at 3% B for 3.0 minutes, followed by a linear gradient increasing to 60% B (3–20.0 minutes). This was followed by eluting the column with 90% mobile phase B for 0.9 minutes and re-equilibrating it at 3% B for 4.9 minutes. MS/MS analysis was performed by monitoring the surrogate peptide (mass-to-charge ratio m/z 602.1 to 673.8, 774.7, and 903.8) and the internal standard (606.2 to 681.7, 911.9, and 783.3) using an optimized fragmentor voltage of 33 V and collision energy of 22 eV. The LC-MS/MS data were processed by integrating the peak areas generated from the reconstructed ion chromatograms for the surrogate peptides and the internal standards using MassLynx 4.1 (Waters). The peak response for two transitions from each peptide was averaged for quantification of samples or standards. Calibrators and quality control samples were assays as before (Prasad et al., 2014). Accuracy and precision of the assay was > 80% and < 20%, respectively. OATP1A2 protein expression data (femtomoles per microgram) were expressed relative to the total protein content of the isolated membrane, as determined by the bicinchoninic acid assay. All samples were digested and measured in duplicate.
Data Analysis.
A three-compartment model was set up to describe the concentration-dependent and time course of apical-to-basolateral transcellular transport of rosuvasatin (Fig. 1). The differential equations used to describe the rate of mass change in the apical, cellular, and basolateral compartments are as follows:(1)(2)(3)where dXA/dt, dXB/dt, and dXC/dt represent the substrate flux into and out of the apical, basolateral, and cellular compartments, respectively; CA, CB, and CC represent substrate concentration in each compartment; PSA and PSB represent the passive permeability-surface area product at the apical and basolateral membranes and here assumes PSA = PSB; and PSinflux represents the permeability-surface area product of OATP1A2-mediated uptake activity. When the substrate concentration approached or exceeded the Km, the following equation was used:(4)where PSinflux represents the permeability-surface area product of OATP1A2-mediated influx activity (nL/s); Vmax represents the maximal influx velocity (nmol/min); Km represents the concentration of the substrate when the initial transport rate is at one-half of the maximum (µM); and CA represents apical donor concentration (µM). The models were constructed using SimBiology software, and the data analysis was conducted in MATLAB (version 8.4; MathWorks, Inc., Natick, MA).
Assuming initial unidirectional flux into the basolateral compartment (i.e., CB = 0) and rapid equilibration between the apical and cellular compartments (dXC/dt = 0), the flux in the apical-to-basolateral direction can be described by the following equations:(5)(6)Substitution of eq. 6 into eq. 5 yields the apical-to-basolateral flux equation(7)In the absence of OATP1A2-mediated influx clearance (PSinflux = 0), apical-to-basolateral flux is given by eq. 8(8)Under the sink condition, eq. 7 divided by eq. 8 should equate to the ratio of apical-to-basolateral permeability–surface area product in the presence to absence of OATP1A2-mediated influx(9)According to eq. 8, apical-to-basolateral passive permeability (PSA-to-B, without OATP1A2) is half of PSA under the assumption of equal apical and basolateral passive permeability (PSA = PSB). Thus, OATP1A2-mediated uptake clearance for substrate drugs across the MDCKII monolayer is defined as follows:
(10)Statistical Analysis.
All data are presented as mean ± standard deviation. A two-tailed Student’s t test or one-way or two-way analysis of variance followed by Tukey’s multiple comparison test, where appropriate, was used to determine the statistical significance of differences among two or more groups. Linear regression analysis was performed with Microsoft Excel 2007. In all cases, P < 0.05 was considered to be statistically significant.
Results
Transport of Triptan Drugs by Oatp1a4.
Sumatriptan, naratriptan, and zolmitriptan were administered by continuous infusion of up to 60 minutes to obtain plasma concentrations and brain- and liver-to-plasma ratios between WT and Oatp1a4 knock-out [Oatp1a4 (−/−)] mice. We did not observe any difference in brain- and liver-to-plasma ratios between WT and Oatp1a4 (−/−) mice at 60 minutes. Plasma concentrations at 30 and 60 minutes were similar between the two genotypes of mice (Fig. 2). These results indicated that brain and liver penetration of triptans was not impacted by Oatp1a4 in vivo.
To confirm the role Oatp1a4 played in the in vivo studies, uptake was determined in mock and Oatp1a4-transfected HEK293 cells in the absence and presence of rifampicin, a potent inhibitor of Oatp1a4. E217βG and TCA were used as positive controls. The uptake of all test triptans at 10 minutes was not significantly different between mock and Oatp1a4-transfected HEK293 cells (Fig.3). Rifampicin did not influence the transport activities of triptans in mock and Oatp1a4-transfected HEK293 cells. The Oatp1a4-mediated uptake of E217βG and TCA was evident in Oatp1a4-transfected HEK293 cells, and the uptake could be inhibited by rifampicin, demonstrating Oatp1a4 activity in the in vitro system.
Membrane Localization of OATP1A2 in BacMam2-OATP1A2 Transduced MDCKII-MDR1 Cells.
The membrane localization of OATP1A2 in MDCKII cells was examined by immunocytochemical staining, with P-gp used as a marker of the apical membrane of MDCKII cells. Nuclei were stained with diamidinophenylindole (blue fluorescence). Both P-gp (green fluorescence) and OATP1A2 (red fluorescence) displayed an asymmetric and overlapping distribution pattern (Fig. 4). The results confirm that OATP1A2 is expressed predominantly in the apical membranes of the cultured MDCKII cells.
Tissue Distribution of OATP1A2 mRNAs.
Human brain OATP1A2 mRNA expression was comparable to BCRP and OATP2B1 and much higher than P-gp (Fig. 5A). The expression level was also significantly higher in the brain mRNA pools as compared with the other tissues evaluated, including the liver, kidney, and lung (Fig. 5B). The fetal brain had a much lower OATP1A2 mRNA expression compared with the adult human brain (Fig. 5B).
Expression of OATP1A2 and P-gp in BacMam2-OATP1A2 Transduced MDCKII-WT and MDCKII-MDR1 Cells.
The mRNA and protein expression of OATP1A2 and mRNA expression of P-gp was characterized in OATP1A2 expressing MDCKII-WT and MDCKII-MDR1 cells. Transduction of the BacMam2-OATP1A2 virus in MDCKII-MDR1 cells resulted in a proportional increase in mRNA levels of OATP1A2, with an MOI of up to 1000, whereas the mRNA level of OATP1A2 increased linearly, with an MOI of up to 500, and then leveled off in MDCKII-WT cells (Fig. 6A). mRNA expression of OATP1A2 had a strong correlation with protein levels in the combination of both types of MDCKII cells (R2 = 0.74; Fig. 6B). There was a trend toward the human P-gp mRNA level increasing slightly with increasing MOI of the BacMam2-OATP1A2 virus in MDCKII-MDR1 cells (Fig. 6C). Virus transduction led to a decrease of endogenous canine P-gp in MDCKII-WT cells, whereas it did not affect the expression of canine P-gp in MDCKII-MDR1 cells. MDCKII-WT cells had a higher endogenous canine P-gp expression compared with MDCKII-MDR1 cells (Fig. 6D).
Monolayer Integrity Post–BacMam2-OATP1A2 Transduction.
Introduction of the BacMam2 virus to the Transwell apical compartment did not compromise the integrity of MDCKII cells up to an MOI of 1000, as indicated by the low permeability of the paracellular markers (lucifer yellow and atenolol). The BacMam2 virus caused ∼40% of the monolayers to be leaky in MDCKII-WT cells at an MOI of 2000, whereas it had less effect in MDCKII-MDR1 cells. Resultant data from monolayers transduced with BacMam2-OATP1A2 at an MOI of 2000 were excluded from further analysis due to loss of monolayer integrity. Interestingly, transepithelial electrical resistance (TEER) values of MDCKII monolayers increased with an increasing MOI of BacMam2-OATP1A2 viruses (Fig. 7A; Fig. 7B). Apical-to-basolateral or basolateral-to-apical permeability of the paracellular marker atenolol did not change substantially with the MOI of BacMam viruses in the MDCKII-WT and MDCKII-MDR1 monolayers (Fig. 7C; Fig. 7D). We observed a trend of increasing apical-to-basolateral permeability of atenolol, with corresponding increases in the expression of OATP1A2 in MDCKII-WT cells. However, data variability and a lack of correlation between atenolol and lucifer yellow permeability in these wells precluded us from making a definite conclusion. The impact on the permeability of zolmitriptan and rosuvastatin was expected to be minimal since their permeability was at least 2-fold higher in the OATP1A2-transfected cells.
Transcellular Transport of Zolmitriptan, Rosuvastatin, and Fexofenadine across OATP1A2 Expressing MDCKII-WT and MDCKII-MDR1 Monolayers.
Bidirectional transcellular permeability of zolmitriptan, rosuvastatin, and fexofenadine was examined in MDCKII-WT and MDCKII-MDR1 monolayers expressing different levels of OATP1A2. Apical-to-basolateral transcellular permeability increased with the increasing protein expression level of OATP1A2 in the two types of MDCKII cells (Fig. 8). The efflux ratio of zolmitriptan and fexofenadine was 7.7 and 3.3 in MDCKII-MDR1 cells, respectively, and 1.1 and 3.2 in MDCKII-WT cells, suggesting that these are P-gp substrates, which is in line with previous reports (Cvetkovic et al., 1999; Evans et al., 2003). The efflux ratios of zolmitriptan and fexofenadine at the protein expression of OATP1A2 of 0.17 fmol/µg protein were 0.21 and 0.53 in MDCKII-WT cells and 0.43 and 1.6 in MDCKII-MDR1 cells, respectively, suggesting that OATP1A2 counteracted P-gp efflux effects for cosubstrates. Rosuvastatin is not a P-gp substrate, as indicated by efflux ratios of 1.2 in MDCKII-WT and 1.4 in MDCKII-MDR1 cells. The efflux ratio of rosuvastatin at an OATP1A2 protein expression of 0.17 fmol/µg protein was 0.06 in MDCKII-WT cells and 0.09 in MDCKII-MDR1 cells, respectively, exhibiting no substantial difference in the two cell types. Virus transduction did not change the permeability of midazolam (data not shown here), a non-OATP1A2 substrate that we evaluated previously (Cheng et al., 2012). OATP1A2-mediated uptake clearance was calculated using eq. 10 for zolmitriptan in MDCKII-WT cells and rosuvastatin in both cell types because their transport was not affected by P-gp in these cells. The uptake clearance was directly proportional to the protein levels of OATP1A2 (Fig. 9). The coefficient of correlation (R2) was strong for zolmitriptan in MDCKII-WT cells (0.94; P < 0.01) and rosuvastatin in MDCKII-MDR1 cells (0.96; P < 0.01). The correlation was weaker and not statistically significant for rosuvastatin in MDCKII-WT cells (R2 = 0.54; P = 0.16).
Kinetic Characterization of Transcellular Transport of Rosuvastatin across OATP1A2 Expressing MDCKII-WT Monolayers.
Rosuvastatin is a BCRP and not a P-gp substrate (Huang et al., 2006). Since endogenous canine Bcrp expression was not detectable in our MDCKII cells (data not shown), rosuvastatin was an ideal tool compound to investigate the role of OATP1A2 in transcellular transport in a quantitative manner. To understand the effect of drug concentration on OATP1A2-mediated transcellular transport more mechanistically, apical-to-basolateral permeability of rosuvastatin was examined at concentrations between 0.3 to 300 μM and found to be high and constant at less than 10 µM. Permeability decreased with an increase in concentrations, suggesting saturation of OATP1A2-mediated influx. Permeability decreased in the presence of the OATP1A2 inhibitor naringin at 500 µM (Fig. 10A). Transcellular apical-to-basolateral permeability was inhibited ∼90% in the presence of naringin at low nonsaturable donor concentrations (0.3 and 1 µM). The concentration-dependent apical-to-basolateral permeability of rosuvastatin was fitted using the three-compartment model described in Fig. 1 and correlated well with the observed permeability. The optimized transport kinetic parameters Km, Vmax, and passive permeability were 30.6 µM, 0.02 nmol/min, and 2.2 nm/s, respectively (Fig. 10B). The resultant OATP1A2-mediated intrinsic transport capacity (Vmax/Km) was 10.9 nl/s, which corresponded to 332 nm/s, assuming 0.33 cm2 of the Transwell surface area. The concentration of rosuvastatin over time across the MDCKII-WT monolayer expressing the same level of OATP1A2 could be predicted by the three-compartment model using the above-fitted kinetic parameters (Fig. 10C), further demonstrating the validity of the three-compartment model to quantitatively depict the OATP1A2-mediated transcellular transport of rosuvastatin across the polarized tight barrier.
Discussion
OATP1A2 is highly expressed in the apical membrane of the human BBB, suggesting the potential to be targeted for CNS drug delivery. At this time, we lack the appropriate in vitro and in vivo tools to investigate the role of this transporter in the brain penetration of xenobiotics. We have successfully constructed a versatile in vitro system using the BacMam virus to transiently express OATP1A2 in MDCKII monolayers and used this to examine the effect of OATP1A2 expression and the interplay with P-gp on transcellular transport of substrate drugs.
OATP1A2 was identified to have a high mRNA expression in the human brain (Kullak-Ublick et al., 1995), and this finding was later corroborated (Eechoute et al., 2011). We examined additional transporters in mRNA preparations from whole adult human brains and report comparable OATP1A2 expression relative to BCRP and OATP2B1 and a much higher expression than P-gp.
Protein expression of OATP1A2 has been reported in human brain capillaries by immunostaining (Gao et al., 2000; Bronger et al., 2005; Lee et al., 2005) and quantitated in healthy brain tissue by LC-MS/MS at 0.25 fmol/µg protein (Drozdzik et al., 2014). This is the concentration we achieved in our transduced cell lines, which lead to greatly enhanced transcellular permeability of the substrate drugs. The protein expression of OATP1A2 in purified human brain microvessels may be lower than 0.695 fmol/µg protein, as Uchida et al. (2011b) could not detect the protein in purified human brain microvessels by LC-MS/MS with this sensitivity limit.
During CNS drug discovery, mouse or rat brains are routinely analyzed to check for penetration of discovery compounds, and along with human P-gp in vitro efflux data, were used to make predictions of clinical CNS distribution. P-gp is an important transporter limiting CNS penetration and is similar between the species, unlike the less homologous uptake transporters e.g., OATP1A2 and Oatp1a4 (Bleasby et al., 2006; Feng et al., 2008). Oatp1a4 is highly enriched in brain capillaries (Liu et al., 2014), with protein levels of 2.11 fmol/µg protein in purified mouse brain microvessels (Kamiie et al., 2008). The presence of Oatp1a4 at both the brain and blood side of rodent BBB could imply a role that is divergent from the apically expressed OATP1A2 (Bronger et al., 2005; Ose et al., 2010). As would be expected, OATP1A2 has some overlapping substrate specificity with Oatp1a4, e.g., rocuronium, fexofenadine, estrone-3-sulfate, and thyroid hormone (Hagenbuch et al., 2002). However, OATP1A2 transports compounds, e.g., deltorphin II and microcystin, which are not substrates of Oatp1a4, whereas Leu-enkephalin and digoxin are Oatp1a4-specific substrates (Hagenbuch et al., 2002). We now report additional differences in substrate specificity with triptan drugs. The CNS penetration of systemically administered 14C-labeled sumatriptan, a relatively hydrophilic triptan, was investigated in mice, and only 0.006% of total radioactivity was recovered in the brain (Humphrey et al., 1990), indicating this drug was not a CNS penetrant. Similarly, sumatriptan was not found in the rat brain in whole body autoradiography studies (Dixon et al., 1993). However, in a positron emission tomography study in six migraine patients, subcutaneous sumatriptan (6 mg) normalized the migraine attack–related increases in brain serotonin synthesis (Sakai et al., 2008), demonstrating the central efficacy of sumatriptan and potential brain uptake. We found that sumitriptan was an OATP1A2 substrate (Cheng et al., 2012), and this likely contributed to the species difference. Zolmitriptan is another example of a P-gp substrate that is brain penetrable, as evidenced by a clinical positron emission tomography study with a brain-to-blood ratio at ∼0.25 (Bergstrom et al., 2006). We found this to be an OATP1A2 substrate as well (Cheng et al., 2012). We have now demonstrated in vitro and in vivo that triptans are not substrates for Oatp1a4 and that by not factoring in potential uptake transport, we may not be potentially progressing brain penetrant compounds to the clinic. To further evaluate this, we have developed a unique in vitro MDCK model to investigate OATP1A2-facilitated transcellular transport.
OATP1A2 expression colocalized with P-gp in MDCKII cells, similar to the reported physiologic apical localization at the human BBB. The increase in apical-to-basolateral transcellular transport of zolmitriptan, rosuvastatin, and fexofenadine was consistent with increasing OATP1A2 protein levels. OATP1A2-mediated uptake clearance was proportional to protein levels of OATP1A2. These data could support future endeavors in the extrapolation of the in vitro findings to in vivo by correcting for the difference in OATP1A2 expression levels. Of note, the basolateral-to-apical permeability of zolmitriptan and fexofenadine in MDCKII-MDR1 cells also increased with increasing OATP1A2 protein expression. This could be caused, in part, by increased human P-gp expression and activity after BacMam virus transduction, as immunocytochemical staining confirmed that OATP1A2 is not expressed at the basolateral membrane. At the same OATP1A2 protein expression, efflux ratios of the OATP1A2 and P-gp cosubstrates zolmitriptan and fexofenadine were lower in MDCKII-WT cells as compared with MDCKII-MDR1 cells. This indicates that OATP1A2 counteracts the P-gp efflux effect of cosubstrates. We used rosuvastatin as a model compound to quantitatively illustrate the contribution of OATP1A2-mediated transcellular transport across MDCKII monolayers because it is not a P-gp substrate. OATP1A2-mediated influx contributed to ∼90% of apical-to-basolateral transcellular permeability of rosuvastatin across the MDCKII monolayer at the expression level used in the experiment. The three-compartment model incorporating OATP1A2-mediated influx quantitatively described the concentration-dependent apical-to-basolateral transcellular permeability across OATP1A2 expressing MDCKII-WT monolayers. The fitted intrinsic transport capacity (Vmax/Km) was much larger than passive clearance, indicating the substantial contribution of OATP1A2-mediated influx to the transcellular permeability of rosuvastatin across MDCKII-WT monolayers. Since P-gp and BCRP, along with OATP1A2, are expressed at the human BBB (Kalvass et al., 2013), these likely work in concert to determine the transcellular permeability across the barrier for cosubstrates, such as zolmitriptan and rosuvastatin.
OATP1A2 is an attractive target for enhancing brain penetration because of broad substrate specificity, desirable expression level (at least of mRNA), and BBB localization, and our data indicate that this could be potentially leveraged to counteract P-gp efflux. Accurate and robust quantification of the OATP1A2 protein with membrane extracts of highly purified human brain capillaries will help this further. The physiologic expression data would help us to contextualize the current in vitro findings, as has been previously done for P-gp (Uchida et al., 2011a, 2014). The data could also shed light on the emerging debate on neuronal (Gao et al., 2014) and more widely accepted brain endothelial localization (Gao et al., 2000; Bronger et al., 2005; Lee et al., 2005; Kalvass et al., 2013).
In conclusion, we have successfully constructed OATP1A2 expressing MDCKII monolayers using a BacMam virus system and demonstrated enhanced apical-to-basolateral transcellular permeability in an OATP1A2 expression level–dependent manner. OATP1A2-mediated uptake clearance was proportional to protein levels of OATP1A2. The presence of OATP1A2 in MDCKII-MDR1 cells counteracts the efflux effect of P-gp for their cosubstrates. The three-compartment model incorporating OATP1A2-mediated influx can quantitatively describe the time or concentration-dependent apical-to-basolateral transcellular transport across OATP1A2 expressing MDCKII monolayers. In spite of significant knowledge gaps for this transporter in CNS distribution, our efforts and this model represent a step forward to improve this understanding.
Acknowledgments
The authors thank Dan Pu and Yiwen Wu for bioanalytical support. We also thank Chen-Bing Guan for providing technical help with immunocytochemical staining. We thank Maciej Zamek-Gliszczynski and David Tattersall for their critical review and helpful discussion during the preparation of the manuscript.
Authorship Contributions
Participated in research design: Liu, Unadkat, Kusuhara, Sugiyama, Sahi.
Conducted experiments: Liu, Yu, S. Lu, Ito, Zhang, Prasad, X. Lu, Li, Wang.
Contributed new reagents or analytic tools: An, Xu, Prasad, Unadkat.
Performed data analysis: Liu, Yu, S. Lu, Ito, Zhang, He, X. Lu, Kusuhara.
Wrote or contributed to the writing of the manuscript: Liu, Prasad, Unadkat, Kusuhara, Sahi.
Footnotes
- Received March 8, 2015.
- Accepted April 22, 2015.
↵1 Current affiliation: Pharmaceutical Engineering Center, General Hospital of Ningxia Medical University, Ningxia, China.
This work was presented in part as a poster presentation as follows: Yu N, Lu S, An G, Xu H, Guan C-B, Sahi J, and Liu H (2014) OATP1A2 and P-gp expression level dependent transport of zolmitriptan and fexofenadine across MDCKII monolayers. Clinical & Pharmaceutical Solutions through Analysis; 2014 Apr 16–19; Shanghai, China.
Abbreviations
- BBB
- blood-brain barrier
- BCRP
- breast cancer resistance protein
- CNS
- central nervous system
- E217βG
- 17β-estradiol-17β-D-glucuronide
- HEK
- human embryonic kidney
- LC
- liquid chromatography
- MDCK
- Madin-Darby canine kidney
- MDR
- multidrug resistance protein
- MOI
- multiplicity of infection
- MS
- mass spectrometry
- OATP
- organic anion-transporting polypeptide
- PBS
- phosphate-buffered saline
- PCR
- polymerase chain reaction
- P-gp
- P-glycoprotein
- TCA
- taurocholic acid
- UPLC
- ultra-performance liquid chromatography
- WT
- wild type
- Copyright © 2015 by The American Society for Pharmacology and Experimental Therapeutics