Multidrug Resistance-Associated Protein (MRP1, 2, 4 and 5

Article pubs.acs.org/molecularpharmaceutics

Multidrug Resistance-Associated Protein (MRP1, 2, 4 and 5) Expression in Human Corneal Cell Culture Models and Animal Corneal Tissue Jessica Verstraelen and Stephan Reichl* Institut für Pharmazeutische Technologie, Technische Universität Braunschweig, Braunschweig, Germany

ABSTRACT: Preclinical studies addressing the transcorneal absorption of ophthalmic drugs are mainly performed using ex vivo animal corneas and in vitro corneal cell culture models, leaving open the question of transferability to humans in an in vivo situation. While passive drug absorption through corneal tissue is well understood, little is known about the expression of transporter proteins and active drug transport in human and animal corneas as well as corneal cell culture models. Therefore, the aim of this study was to conduct an expression analysis of four multidrug resistance-associated proteins (MRP1, 2, 4 and 5) in various in vitro and ex vivo corneal models, leading to a better understanding of the comparability of different corneal models regarding drug absorption and transferability to humans. Two well-established in vitro human corneal models, the HCE-T epithelial model and the more organotypic Hemicornea construct, both of which are based on the SV40 immortalized human corneal epithelial cell line HCE-T, were analyzed, as were excised rabbit and porcine cornea. Specimens of abraded epithelia from human donor corneas were also tested. MRP mRNA expression was determined via reverse transcriptase polymerase chain reaction. Protein expression was examined using Western blot experiments and immunohistochemistry. The functional activity of the MRP efflux transporter was detected in transport assays using specific marker and inhibitor substances. The functional expression of all of the tested MRP transporters was detected in the HCE-T epithelial model. Hemicornea constructs displayed a similar expression pattern for MRP1, 4 and 5, whereas no MRP2 protein expression or activity was detected. However, excised animal corneas exhibited different expression profiles. In porcine cornea, no functional expression of MRP1, 2, or 5 was observed, and we failed to detect MRP4 expression in rabbit cornea. The results suggest that MRP1, 2, 4, and 5 are expressed in the human corneal epithelium and confirm that the transfer of data obtained from animal experiments to an in vivo situation in humans should be performed with caution. KEYWORDS: multidrug resistance-associated proteins, efflux transporters, drug transport, cornea, in vitro model

1. INTRODUCTION Due to the insufficient ocular bioavailability of systemically applied drugs for the treatment of ocular diseases such as glaucoma or inflammatory processes, such drugs are generally applied topically. However, the ocular bioavailability of drugs following instillation in the cul-de-sac is also poor, at below 5%.1 Such low bioavailability results from the short precorneal residence time of drugs due to high nasolacrimal and conjunctival elimination, the small absorption area and the strong epithelial barrier.2 Therefore, ocular drug delivery represents an interesting and challenging field and has been investigated intensively over the last several decades. The cornea is considered to be the main route for and barrier to ophthalmic drugs following topical application, whereas its intrinsic role is to protect the eye against environmental © 2014 American Chemical Society

influences as well as to refract and transmit incoming light. The cornea is a transparent avascular tissue consisting of five main layers: the endothelium, Descemet’s membrane, the stroma, Bowman’s membrane and the epithelium, as the outer layer. The epithelium consists of five to six layers of flat, tightly linked squamous cells forming a hydrophobic barrier limiting the passive diffusion of hydrophilic small molecules and macromolecules due to the presence of tight junctions.3 The stroma Special Issue: Engineered Biomimetic Tissue Platforms for in Vitro Drug Evaluation Received: October 25, 2013 Accepted: January 23, 2014 Published: January 23, 2014 2160

dx.doi.org/10.1021/mp400625z | Mol. Pharmaceutics 2014, 11, 2160−2171

Molecular Pharmaceutics

Article

accounts for approximately 90% of the thickness of the cornea. It forms a hydrophilic barrier consisting of an organized collagen structure, glycosaminoglycans and various proteoglycans, as a matrix with embedded keratocytes. The endothelium is a monolayer of cells providing a hydrophobic barrier and is responsible for maintaining corneal transparency. Endothelial cells play only a marginal role in ocular drug absorption and a greater role in sustaining eye transparence.4 The epithelial characteristics of the cornea limit the penetration rate of ocular drugs following their administration, indicating that the epithelium is the main barrier to drug absorption, leading to interest in different drug absorption routes in this tissue.5−8 The first route is passive and is based on paracellular and transcellular drug diffusion, which has been investigated intensively. The absorption rate is strongly dependent on the function of the tight junctions of the corneal epithelium and the physicochemical properties (e.g., molecular size, charge and lipophilicity) of the administered drug.4,5,9,10 Furthermore, the presence of various esterases, peptidases and proteases influences the absorption and metabolization of ocular prodrugs.11,12 A second pathway includes active transport processes via transporter proteins. This active membrane transporter system comprises the ATP-binding cassette (ABC) proteins, which are a large family of integral membrane efflux transporters. ABC transporters utilize the energy of ATP hydrolysis to translocate specific substrates across membranes. They are able to change drug concentrations in cells by transporting these molecules actively, depending on their location on the membrane. The human genome contains 49 genes for ABC transporters divided into 7 subfamilies.13,14 The ABC transporter family is widely distributed throughout the human body and has been intensively investigated in different tissues, such as the liver, kidney and intestine.14,15 However, the expression, function and physiological role of transporter proteins in corneal tissue have been much less intensively studied than passive transcorneal absorption.1,8,16,17 Preclinical investigations addressing the efficiency of novel drug delivery systems or the pharmacokinetic behavior of new ophthalmic drugs are mainly carried out ex vivo using excised corneas from experimental animals. Due to the limited availability of human donor corneas for such purposes, the ethical concerns related to animal use, the questionable translation of animal results to humans and the poor standardization of animal experiments, a demand for validated in vitro corneal models has arisen in recent years. Hence, several corneal cell culture models, including pure epithelial models and more organotypic, three-dimensional corneal equivalents of animal origin, such as a bovine,18,19 porcine,20,21 canine22 or rabbit origin,23 as well as of human origin,24−28 have been introduced and investigated in drug absorption studies and through cytotoxicity tests.29,30 Recently, we established a new three-dimensional corneal model for drug absorption testing, the Hemicornea construct, consisting of a multilayered epithelium and stromal biomatrix.16 This model was prevalidated in a previous study and found to be a robust tool as an animal replacement for investigating transcorneal drug absorption.31 In another study, an expression analysis of the MDR1, MRP3 and BCRP efflux transporters was performed in Hemicornea constructs and human and animal corneal cells, indicating similarity between the Hemicornea model and the human cornea, but differences in corneal transporter expression were also noted depending on the species.32

In the present study, we were interested in the expression pattern of a particular ABC transporter subfamily, the ABCC or multidrug resistance-associated protein (MRP) family, which includes nine isoforms with a wide range of substrates and expression. The protein expression and functions of four MRP transporters (MRP1, MRP2, MRP4 and MRP5) were investigated in different ex vivo and in vitro corneal models and compared to examine the similarity and transferability between the models. Two human cell culture models, a pure corneal epithelial model and the Hemicornea construct, were included in this study, and excised rabbit and porcine corneas were also investigated as the most frequently used animal tissues in transcorneal drug absorption studies. Furthermore, human corneal epithelial cells obtained from human donor corneas were tested. MRP expression profiles were determined at the mRNA and protein levels using reverse transcriptase polymerase chain reaction (PCR), Western blot and immunohistochemistry experiments. The function of the MRP proteins was examined by performing transport assays with specific substrates and inhibitors.

2. MATERIALS AND METHODS 2.1. Materials. Dimethyl sulfoxide (DMSO), sodium bicarbonate, rat tail collagen, CaCl2·2H2O, triethanolamine hydrochloride (TEA-HCl), ethylenediaminetetraacetic acid (EDTA), dithiothreitol (DTT), glycine, acrylamide/bis-acrylamide solution, sodium hydroxide solution, tetramethylethylenediamine (TEMED), 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI), protease inhibitor cocktail (Cat. No. P8340), Tris base, Tris HCl, 2× sample Laemmli buffer, milk powder (Cat. No. 70166), Tween 20, verapamil, MK571, 5(6)-carboxy2′,7′-dichlorofluorescein diacetate (CDCFDA), indometacin, probenecid and bovine serum albumin (BSA) were obtained from Sigma (Deisenhofen, Germany). Phosphate-buffered saline (PBS), fetal bovine serum (FBS), Dulbecco’s modified Eagle’s medium (DMEM), Ham’s F12, 10× minimum essential medium (MEM), insulin, epidermal growth factor (EGF), nonessential amino acids (NEAA), fibronectin and L-glutamine were purchased from Biochrom (Berlin, Germany). KCl and MgSO4·7H2O were purchased from Acros Organics (Geel, Belgium), and an EDTA disodium salt solution was obtained from MP Biomedicals (Solon, Ohio, US). Trypsin-EDTA, penicillin G sodium salt, streptomycin sulfate, amphotericin B and the MycoTrace kit were purchased from PAA (Linz, Austria). A trypsin inhibitor was acquired from Invitrogen (Karlsruhe, Germany). Tissue culture flasks were purchased from Sarstedt (Nümbrecht, Germany), and Transwell inserts (polycarbonate membrane 1.12 cm2, pore size 3.0 μm) were obtained from Corning Costar (Acton, Massachusetts, US). Keratinocyte basal medium (KBM), referred to as keratinocyte growth medium (KGM) after the addition of the provided SingleQuots, was acquired from Lonza (Rockland, Maine, US). Acetic acid, NaCl, HEPES, sodium dodecyl sulfate (SDS) and D-glucose monohydrate were purchased from Roth (Karlsruhe, Germany). NaH2PO4·H2O was obtained from Merck (Darmstadt, Germany). The radioactive substrates [H3]erythromycin and [H3]adefovir dipivoxil were purchased from Biotrend (Köln, Germany), and scintillation liquid Optiphase Supermix was obtained from Perkin-Elmer (Waltham, Massachusetts, US). Krebs−Ringer buffer (KRB), which was used for transport analysis, contained the following substances in 1000 mL of double-distilled water: 6.8 g of NaCl, 0.4 g of KCl, 0.14 g of NaH2PO4·H2O, 2.1 g of NaHCO3, 3.575 g of HEPES, 1.1 g of 2161



Article

D-glucose monohydrate, 0.2 g of MgSO4·7H2O and 0.26 g of CaCl2·2H2O. The pH of the KRB was adjusted to 7.4. TRIzol was obtained from Invitrogen (Karlsruhe, Germany). The RevertAid First Strand cDNA Synthesis kit, DreamTaq DNA Polymerase and buffer, the GeneRuler Low Range DNA ladder, the ProteoJET membrane protein extraction kit, a Spectra multicolor broad-range protein ladder, Neg50 frozen section medium, polylysine-coated slides and chamber slides were obtained from Thermo Scientific (St. Leon-Rot, Germany). Polyvinylidene fluoride (PVDF) blotting membranes (0.45 μm pore size) were purchased from Peqlab (Erlangen, Germany). The Lumi-LightPlus Western blotting kit was acquired from Roche (Mannheim, Germany). Rabbit polyclonal anti-MRP1, mouse monoclonal anti-MRP2 [M2III-6], rat monoclonal anti-MRP4 [M4I-10] and rat monoclonal anti-MRP5 [M5II-54] antibodies were purchased from Abcam (Cambridge, UK). The secondary antibodies antirabbit IgG-FITC, anti-mouse IgG-FITC and anti-rat IgG-FITC, all produced in goats, were purchased from Sigma (Deisenhofen, Germany). Horseradish peroxidase-conjugated goat secondary antibodies were obtained from Millipore (Schwalbach, Germany). The antibodies were diluted with the antibody diluent DAKO REAL (DAKO, Hamburg, Germany). 2.2. Cell Culture. Cells were routinely cultured in polycarbonate tissue culture flasks at 37 °C in a humidified atmosphere containing 5% CO2. The growth medium was replaced three times per week. Cell counting was performed using a Z2 Coulter Counter (Beckman Coulter, Krefeld, Germany), and the cell cultures were analyzed according to standard procedures for the absence of mycoplasma infection using the MycoTrace PCR detection kit. 2.2.1. HCE-T Epithelial Model. The human corneal epithelial model is a pure epithelial model based on the SV40 immortalized human corneal epithelial cell line HCE-T. This cell line, which was derived from a 49-year-old woman and characterized by Araki-Sakaki et al.,33 was procured from the RIKEN cell bank (Tsukuba, Japan). The cultivation of the HCE-T epithelial model has been described previously in detail.16 Briefly, 100,000 cells/cm2 were seeded onto collagen/ fibronectin-coated Transwell inserts and cultivated for 7 days. After confluence was reached, the monolayer was lifted to the air−liquid−interface (ALI), where it was incubated for an additional 4 days, resulting in the formation of a multilayered epithelium. The cells were cultivated in DMEM/F12 medium supplemented with 15% FBS, 0.5% DMSO, 10 ng/mL EGF, 2 mM glutamine, 5 μg/mL insulin, 100 U/mL penicillin G sodium salt, 100 μg/mL streptomycin sulfate and 0.25 μg/mL amphotericin B. 2.2.2. Hemicornea Construct. The Hemicornea model is a three-dimensional model that is more organotypic than the HCE-T epithelial model due to the addition of a collagenous stromal biomatrix containing human corneal keratocytes (HCK) to the multilayered epithelium. The SV40 immortalized human corneal keratocyte cell line HCK was a kind gift from Dr. Zorn-Kruppa (Hamburg, Germany).26 The Hemicornea was cultivated according to the protocol as described previously.16,31 Briefly, HCE-T and HCK cells were cultivated under serum-free conditions in KGM. Initially, a keratocyte cell suspension in a premixture of 10× MEM, L-glutamine (12.9 mM) and NaHCO3 (16.1 mg/mL) was mixed with an acidic solution of type l rat tail collagen (1.7 mg/mL) and poured into Transwell inserts. After gelling, the HCE-T cells were seeded onto the collagen matrix. The Hemicornea model was

cultivated submerged for another 7 days before being lifted to the ALI and cultivated for 3 additional days. 2.2.3. Caco-2 Cell Line. The Caco-2 cell line, an immortalized cell line consisting of heterogeneous human epithelial colorectal adenocarcinoma cells,34 is a well-characterized cell line that is used as a standardized in vitro model to predict the intestinal absorption rate of therapeutics.35 Caco-2 cells were employed as a positive control in the present study because the expression of the ABC transporters MRP1, MRP2, MRP4 and MRP5 in these cells has already been described.36 The Caco-2 cells were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany, Cat. No. ACC 169) and were maintained in MEM supplemented with 20% FBS, 1% NEAA, 100 U/mL penicillin G sodium salt, 100 μg/mL streptomycin sulfate and 0.25 μg/ mL amphotericin B. To prepare the Caco-2 model, cells were seeded on Transwell inserts at a density of 100,000/cm2 and cultivated submerged for 21 days until complete confluence and acceptable TEER values were reached. 2.2.4. TEER Measurements. The transepithelial electrical resistance (TEER) was measured using an EVOM epithelial voltohmmeter and Endohm chamber (World Precision Instruments, US-Sarasota). The results were corrected for the background value introduced by the blank filter and used as an indicator of the epithelial tightness and functional integrity of the cell culture models. Only HCE-T cell layers showing TEER values >1000 Ω·cm2, Hemicornea constructs displaying TEER values >400 Ω·cm2 and Caco-2 models exhibiting TEER values >250 Ω·cm2 were used in transport assays to confirm the intactness of the epithelial barrier. 2.3. Corneal Tissue. Fresh porcine and rabbit eyes were obtained from a local slaughterhouse and local breeders. The eyes were enucleated within 30 min after sacrifice and transported in KRB, and the corneas were then excised and used immediately. Human corneal epithelial cells (HEC) were obtained from human donor corneas following abrasion and stored in a 0.9% sodium chloride solution at −80 °C until processing. The human cell material was received from the Cornea Bank of Hannover Medical School (Hannover, Germany) in accordance with ethical regulations and was processed in accordance with the guidelines outlined in the Declaration of Helsinki. 2.4. Reverse Transcriptase Polymerase Chain Reaction. Reverse transcriptase polymerase chain reaction (RTPCR) was utilized to profile the mRNA expression of MRP proteins as previously described.32 Briefly, total RNA was isolated using TRIzol and processed according to the manufacturer’s protocol. Due to the tissue strength of the Hemicornea construct and excised corneas, prehomogenization was required prior to RNA isolation. This prehomogenization step was performed with a MM301 Ball Mill (Retsch, Haan, Germany) at 30 Hz for 10 min using glass beads with a diameter of approximately 0.50−0.75 mm. The concentration and purity of the isolated RNA were determined spectrophotometrically using a Spekol 1300 UV spectrometer (Analytik Jena, Jena, Germany). Reverse transcriptase PCR was performed using the RevertAid First Strand cDNA Synthesis kit and a Labcycler thermocycler (SensoQuest, Göttingen, Germany). Primers were custom-made for each transporter using the OligoPerfect Designer from Invitrogen, and their specificity was controlled using BLAST (GenBank accession numbers NM_004996 for MRP1, NM_000392 for MRP2, NM_005845 for MRP4 and NM_005688 for MRP5). Each 2162



Article

were used to study the expression of MRP1. Electrophoresis was performed at 150 V, 80 mA and 25 W for 90 min. Following electrophoresis, the proteins were transferred to a PVDF blotting membrane using a tankblot system (PerfectBlue tank electro blotter, Peqlab). The transfer buffer contained 25 mM Tris base, 192 mM glycine and 0.2% SDS, pH 8.3. The membrane was blocked for 60 min at room temperature in MPBST buffer containing 5% powdered skimmed milk and 0.05% Tween 20 in PBS. Incubation with the primary antibodies diluted in MPBST (1:200 for MRP1, 1:50 for MRP2, 4 and 5, 1:2000 for β-actin) was conducted overnight at 4 °C, followed by incubation with the appropriate horseradish peroxidase-conjugated secondary antibody (diluted 1:5000 in MPBST buffer) for 60 min at room temperature. Peroxidase activity was visualized via chemiluminescence detection with the Lumi-LightPlus Western blotting substrate. 2.6. Immunohistochemistry. The samples were embedded in an optimal cutting temperature compound (Neg50 frozen section medium), and 10 μm frozen sections were cut using a HM550 cryomicrotome (Thermo Scientific, St. LeonRot, Germany). The sections were placed on polylysine-coated slide glasses and air-dried. These cross sections or samples of Caco-2 and HCE-T cells cultivated on chamber slides were fixed with ice-cold methanol for 10 min. The specimens were blocked with 2% BSA in DAKO REAL solution for a minimum of 60 min at room temperature. Following incubation with primary antibodies diluted in DAKO REAL solution (1:50 for MRP1 and MRP2, 1:20 for MRP4, 1:30 for MRP5), the preparations were washed three times with PBS. FITCconjugated secondary antibodies (diluted 1:100) were added for an additional 60 min, followed by washing with PBS. The specimens were finally examined with a CLSM 510 Meta confocal laser scanning microscope (Zeiss, Jena, Germany). The settings were adjusted such that no positive FITC results were observed in the negative control sample. 2.7. Transport Assays. 2.7.1. Bidirectional Permeation Studies. Permeation studies were carried out to determine the functional expression of the MRP2 transporter by studying the bidirectional transport of a specific substrate. The experiments were performed at 37 °C, either directly in Transwell inserts, in the case of the Caco-2 model, HCE-T model and the Hemicornea constructs, or using a vertical Ussing diffusion chamber system (Harvard Apparatus, Holliston, Massachusetts, US), for excised animal corneas. A specific substrate and inhibitors were used as the donor for efflux-mediated transport: [H3]erythromycin (1 μCi/mL) was used as the substrate, and MK571 (50 μM) and verapamil (100 μM) were used as inhibitors for MRP2 and MDR1, respectively. Prior to the transport experiments, the cell culture models and excised tissues were rinsed with KRB and incubated in KRB for 30 min. Subsequently, the donor solution (KRB containing erythromycin with or without an inhibitor) was applied to either the apical or basolateral side. KRB solution acted as the acceptor medium. For the Transwell experiments, the apical volume was 400 μL, and the basolateral compartment contained 1500 μL, whereas in the Ussing chambers, both the apical and basolateral volumes of the donor and acceptor were 2000 μL. The Transwell plates were agitated on an orbital shaker during the experiment. For the Ussing chamber system, the donor and acceptor solutions were mixed constantly with carbogen gas during the entire course of the experiment. Sample aliquots (200 μL) were withdrawn from the acceptor compartment at 30, 90, 150, 210, 270 and 330 min and replaced with the same

PCR run consisted of 30 cycles, with different temperatures and primers being used depending on the transporter (Table 1). Table 1. Summary of the Sequences of the Forward (F) and Reverse (R) PCR Primers, Annealing Temperatures and Resulting RT-PCR Product Sizes for the Studied MRP Transporters and the Positive Control primers MRP1

MRP2

MRP4

MRP5

GAPDH

F 5′-AGG TGG ACC TGT TTC GTG AC-3′ R 5′-ACC CTG TGA TCC ACC AGA AGG-3′ F 5′-TGC TTC CTG GGG ATA ATC AGC-3′ R 5′-CAC GGA TAA CTG GCA AAC CTG-3′ F 5′-CCA TTG AAG ATC TTC CTG G-3′ R 5′-GGT GTT CAA TCT GTG TGC-3′ F 5′-CTT CCC GTG GTT CCT TGT GG-3′ R 5′-GTC AGG GGA GGG AGC CTT GT-3′ F 5′-CAA GGT CAT CCA TGA CAA CTT TG-3′ R 5′-GTC CAC CAC CCT GTT GCT GTA G-3′

annealing temp (°C)

RT-PCR product size (bp)

59.1

183

59.1

211

54.0

238

63.3

529

61.1

496

The program consisted of three steps, starting with an initial phase at 94 °C for 5 min, followed by a phase of 94 °C for 30 s and an annealing phase for 30 s, with final 1 min incubation at 72 °C. Each run included a negative control, a positive control containing Caco-2 cells and a reaction with GAPDH as the internal control. The obtained PCR products were separated via conventional agarose (2%) gel electrophoresis (PerfectBlue Gelsystem Mini S, Peqlab, Erlangen, Germany) and visualized with ethidium bromide staining under UV light (AlphaImager 1220, Alpha Innotech Corporation, San Leandro, CA). A GeneRuler Low Range DNA ladder (25−700 bp) was loaded as a control marker. 2.5. Western Blot Analysis. Western blot experiments were carried out using standard protocols and two different protein extraction methods. First, the ProteoJET membrane protein extraction kit was applied for membrane protein extraction. The procedure resulted in the isolation of the membrane protein fraction, which could be stored at −80 °C until further analysis. In the second extraction procedure, cell pellets were resuspended in 200 μL of lysis buffer containing 50 mM TEA-HCl, 1 mM EDTA, 10 mM DTT and a protease inhibitor cocktail in PBS, pH 7.4, and sonicated with Soniprep (MSE, London, U.K.) at 10 Hz for 18 cycles (10 s on/20 s off) to disrupt cell membranes. The samples were then centrifuged for 30 min at 16000g at 4 °C, and the supernatants containing whole-cell protein extracts were aliquoted and stored at −80 °C until further analysis. The obtained protein content was quantified via the colorimetric method of Bradford.37 Protein samples were separated via sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS−PAGE), according to the method of Laemmli,38 using a Mini-PROTEAN Tetra Cell unit (Biorad). Equal amounts of proteins were loaded into 7.5% gels. Membrane-isolated proteins were separated to determine MRP2, MRP4 and MRP5 expression, while whole-cell fractions 2163



Article

Figure 1. RT-PCR analysis of MRP1 (A), MRP2 (B), MRP4 (C) and MRP5 (D) mRNA in various in vitro and ex vivo corneal models. Specific mRNA primers for each MRP transporter protein and β-actin were used for RT-PCR analysis of cDNA synthesis from total RNA from corneal cell culture models and ex vivo specimens. The expected size of the mRNA for each MRP was detected by each primer. Lane PC represents GAPDH, the internal standard, followed by lane M, showing the DNA ladder marker (low gene ladder, 25 to 700 bp). The Caco-2 cell line (lane 1) was included as a positive control for all four transporters. Negative controls were run and confirmed as negative. Lanes 2 to 6 represent the DNA bands for the HCE-T epithelial model (lane 2), Hemicornea construct (lane 3), porcine cornea (lane 4), rabbit cornea (lane 5) and HEC (lane 6).

Ussing diffusion chamber system, for the excised animal corneas. A CDCFDA solution with or without probenecid was added to the apical compartment, followed by incubation for 60 min. After incubation, the cell culture models and excised tissues were rinsed with KRB to remove any unpenetrated marker. Aliquots (100 μL) were withdrawn from the acceptor compartment at 90 min and replaced with the same volume of tempered KRB. The amount of transported CDCF was analyzed with a fluorescence plate reader (Genios, Tecan, Männedorf, Switzerland) using a 485 nm excitation filter and 535 nm emission filter. To detect the function of MRP4, an additional setup was used. Caco-2 and HCE-T cells were grown as a confluent monolayer in 24-well plates, whereas the other models and excised tissue were transported following cultivation or transfer into 12-well plates. The corneal models were incubated for 120 min with [H3]adefovir dipivoxil (0.5 μCi/mL) as an MRP4 substrate, with or without indometacin (30 μM) as an MRP4 inhibitor.39 Following incubation, the specimens were rinsed with ice-cold KRB to remove the residue of substrate. Epithelial cells were lysed using a 0.1 N NaOH/0.1% SDS mixture for 4 or 24 h depending on the tissue. The amount of adefovir dipivoxil present was determined by scintillation counting, as described above. Significantly higher intracellular concentrations of adefovir in the presence of indometacin indicated MRP4-mediated efflux.

volume of tempered KRB. Following the addition of each sample to 2 mL of scintillation liquid, the quantity of transported substrate was determined using a liquid scintillation counter (LS 6500, Beckman Coulter, Krefeld, Germany). The amount of transported substrate, both in the presence and in the absence of an inhibitor, was analyzed in both the apical to basolateral (ab) and basolateral to apical (ba) directions. Permeation profiles were determined by plotting the permeated amount of erythromycin versus the time. The permeation coefficient, Papp [cm/s], was calculated from the steady-state flux estimated from the linear ascent of the curve. Functional expression of the MRP2 transporter was detected based on significant differences of Papp values for erythromycin in view of the direction of permeation. 2.7.2. Uptake Assays. An uptake assay was implemented using CDCFDA (25 μM in KRB) as a substrate and probenecid (10 mM) as an inhibitor to detect the functional expression of the MRP1 and MRP5 transporters. CDCFDA diffuses passively into intact cells, where the molecule is cleaved by intracellular esterases, producing 5(6)-carboxy-2′,7′-dichlorofluorescein (CDCF). The polar compound CDCF is unable to leave the cells passively. However, cellular efflux can be mediated by active transport regulated by MRP1 and MRP5. The addition of an inhibitor (probenecid) disables the transport of CDCF out of the cells. These assays were performed at 37 °C either in Transwell inserts, in the case of the Caco-2 model, HCE-T model and the Hemicornea constructs, or using a vertical 2164



Article

2.8. Statistical Analysis. The data obtained from the transport assays are presented as the mean ± standard deviation. Statistical analysis of the results was conducted with Mann−Whitney U tests using SPSS software. P values less than 0.05 (* p < 0.05, ** p < 0.01, *** p < 0.001) were considered statistically significant.

3. RESULTS 3.1. mRNA Expression of MRP Transporter Proteins in Corneal Models. The results of the PCR analysis are shown in Figure 1. The Caco-2 cells, used as a positive control, exhibited strong positive signals for all of the tested transporter proteins (MRP1, 2, 4 and 5). Figure 1A clearly demonstrates that MRP1 mRNA expression was detected in all of the corneal models, both in vitro and ex vivo, though a weaker signal was observed in the excised animal corneas. A similar result was found for MRP2 mRNA expression (Figure 1B), which was present in all of the different models, and only deviated in terms of signal strength, particularly compared to Caco-2 cells. Differences in mRNA expression patterns between the in vitro and ex vivo models were obvious for the MRP4 and 5 transporter proteins. The HCE-T epithelial model and Hemicornea construct exhibited bands for MRP4 and 5, whereas no MRP4 signal was detected in porcine cornea, and only very weak signals could be observed in rabbit cornea in the case of MRP4 and 5 (Figures 1C, 1D). In contrast, in HEC, we found weak and strong positive signals of MRP4 and MRP5 mRNA expression, respectively. However, in interpreting the results of the mRNA analysis, it should kept in mind that some results observed in the rabbit and porcine corneas may be due to the nonspecific nucleotide sequences used to detect the animal genes, resulting in an incompatibility between the human primers and their transporter sequences. 3.2. MRP Expression in Corneal Models at the Protein Level. 3.2.1. Western Blot Analysis. Figure 2 presents the results of the Western blot analysis of transporter expression at the protein level. In samples of Caco-2 cells, used as a positive control, all of the transporter proteins were detected. The betaactin band was always clearly detectable in all of the models, except in the HEC specimens. This result was expected because only a small quantity of these cells was available, and the resultant protein concentration was not in the same range as for the other models. Therefore, the results obtained from HEC are difficult to interpret for all of the tested transporters. When MRP1 expression was examined, we observed a protein band of approximately 190 kDa in the HCE-T epithelial model and from the Hemicornea construct and rabbit cornea. The porcine cornea and HEC samples did not show a positive signal. In the case of MRP2 (190 kDa), a protein band was only detected in the HCE-T epithelial model and rabbit cornea. Protein samples from the Hemicornea construct, porcine cornea and HEC did not show a positive signal, suggesting that there was no MRP2 protein expression in these corneal models. In contrast, expression of MRP4, a 160 kDa protein, appeared in both of the HCE-T-based in vitro models (the epithelial model and Hemicornea) and in porcine cornea, whereas in the rabbit cornea and HEC samples, no protein band was detected on the blotting membrane. Unlike the three other MRP proteins, expression of MRP5 (160 kDa) was only observed in the ex vivo rabbit and porcine cornea models. This protein was not detected either in the HCE-T based in vitro models or in the human corneal abrasion epithelium.

Figure 2. Western blot analysis of MRP1, MRP2, MRP4 and MRP5 protein expression. Representative Western blots of cell lysates isolated from Caco-2 cells as a positive control, the HCE-T epithelial model, the Hemicornea construct, porcine cornea, rabbit cornea and HEC. Beta-actin (42 kDa) was used as the control protein and detected in every blot.

3.2.2. Immunohistochemistry. To verify the findings of the Western blot analysis, immunohistochemical analyses were performed. The positive results, indicating MRP protein expression (green fluorescence), are illustrated in Figures 3 to 6; negative results (no fluorescence) are not shown. Caco-2 cells again served as a positive control and are included in every figure. The corneal models that contained stromal tissue (Hemicornea and excised animal cornea) in addition to the epithelial layer showed strong background fluorescence in all

Figure 3. MRP1 immunostaining showing positive expression in Caco2 cells and the HCE-T epithelial model when cultured in chamber slides (A, B), in a cross section of the Hemicornea construct (C) and the rabbit cornea (D). Arrows indicate the membrane localization of MRP1 in the Caco-2 monolayer and the HCE-T epithelial model as well as in the superficial cells of the Hemicornea construct and the rabbit corneal epithelium. 2165



Article

Figure 6. Immunolocalization of the MRP5 protein in Caco-2 cells and the HCE-T epithelial model cultured in chamber slides (A, B), the cross section of the Hemicornea construct (C) and the porcine cornea (D). Arrows indicate the membrane localization of MRP5 in the Caco-2 and HCE-T cells as well as in superficial cells of the Hemicornea construct and porcine corneal epithelium.

Figure 4. Immunostaining of MRP2 with the M2III-6 antibody. Representative immunostaining images of Caco-2 cells and the HCE-T epithelial model cultured in chamber slides (A, B) and a cross section of the rabbit cornea (C). Arrows indicate MRP2 protein expression in Caco-2, HCE-T and rabbit corneal epithelial cells.

strong fluorescence signal for this protein in the epithelial layer, whereas in rabbit cornea no signal could be detected. It should be noted that all of the antibodies used are specific for human antigens, which means that careful consideration is necessary when interpreting the Western blot and immunohistochemistry results for excised animal corneas. 3.3. Transport Assays. 3.3.1. Bidirectional Permeation Studies. Permeation studies were performed using erythromycin as a substrate to analyze the function of the apically located MRP2 transporter. Because erythromycin is also known to act as an MDR1 substrate, verapamil was added to every transport assay setup in both directions as an MDR1 inhibitor to eliminate any MDR1 transporter activity. In Figure 7A, the apparent permeation coefficients for all of the corneal models are given, representing the apical-to-basolateral and basolateralto-apical transport rates of erythromycin. In the case of the Hemicornea construct and porcine cornea no difference in Papp values related to the transport direction of the substrate was detected, suggesting that neither of models express a functional MRP2 transporter. In contrast, in the HCE-T epithelial model and rabbit cornea significantly higher Papp values were detected for the ba compared to the ab direction, indicating efflux activity in these corneal models. In a second experiment using the same experimental setup, MK571 was added to inhibit MRP2 function. The effect was significantly reduced by the addition of MK571, resulting in similar Papp values for both directions (data not shown), indicating that the detected difference was caused by active MRP2 function. 3.3.2. Uptake Assays. The fluorescent compound CDCF acts as both an MRP1 and MRP5 transporter substrate that is actively transported out of cells following the uptake and conversion of CDCFDA. In Figure 7B, the CDCF efflux results detected in the corneal models and Caco-2 cells, as a positive control, both with and without probenecid as an inhibitor, are presented. In the absence of probenecid, we detected a significantly higher CDCF efflux concentration in both human in vitro corneal models and rabbit cornea than after the addition of the inhibitor, suggesting transporter activity. In porcine cornea different behavior was detected. In this case,

Figure 5. Immunostaining using the M4I-10 antibody to examine MRP4 expression (indicated by arrows) in Caco-2 cells and the HCE-T epithelial model cultured in chamber slides (A, B), a cross section of the Hemicornea construct (C) and the porcine cornea (D).

cases, most likely due to nonspecific interactions between the antibodies and the collagen matrix. MRP1 protein expression was detected in the form of strong fluorescence in cell membranes in the HCE-T epithelial model, the Hemicornea construct and rabbit cornea (Figure 3). Porcine cornea exhibited no detectable signal. Figure 4 illustrates the apical immunolocalization of MRP2 in the Caco-2 cell line, the HCE-T epithelial model and rabbit cornea. However, no positive immunostaining was detected in the Hemicornea construct or porcine cornea. The expression of MRP4 protein is demonstrated in Figure 5 for both human in vitro models and porcine cornea, whereas in rabbit cornea no positive staining was observed. Similar results were obtained for the MRP5 transporter, as shown in Figure 6. Both cell culture models, i.e., the HCE-T epithelial model and the Hemicornea construct, and porcine cornea exhibited a 2166



Article

Figure 7. Comparative illustration of the results of the transport assays. Panel A shows the Papp values [cm/s] calculated from bidirectional permeation studies to explore MRP2 function (mean ± SD, n = 3−6). Significant differences in the permeation coefficients obtained for the opposite directions are indicated (* p < 0.05, ** p < 0.01, *** p < 0.001). Panel B represents the CDCF concentrations [μM] in the acceptor compartment following the cellular uptake and conversion of CDCFDA in the absence and presence of probenecid. Significantly higher CDCF concentrations in the absence of probenecid suggest functional expression of MRP1 and MRP5. In panel C, the accumulated cellular concentrations of adefovir [pmol] for all corneal models are depicted, with or without indometacin as an inhibitor. Higher adefovir concentrations in the presence of indometacin indicate MRP4 function.

the rabbit cornea, indicating MRP4 function in the HCE-T-based models and porcine cornea.

quantities of CDCF transported were independent of the addition of probenecid. The functional expression of the MRP4 transporter was tested through adefovir uptake assays in the presence or absence of the specific inhibitor indometacin. The results are presented in Figure 7C. Addition of the inhibitor led to a higher intracellular adefovir concentration in MRP4-expressing cells in comparison to cells lacking MRP4. This effect was detected at significant levels in all of the analyzed corneal models, except

4. DISCUSSION The most common model used in ocular irritation and absorption studies is the rabbit eye, which therefore represents an important tool for the development of new ophthalmological drugs. The best known test for determining a substance’s irritation potential is the Draize rabbit eye irritation 2167



Article

Table 2. Comparison of Expression Patterns of the MRP1, 2, 4 and 5 Transporter Proteins in ex Vivo and in Vitro Corneal Models Developed for Performing Drug Absorption Studies at the mRNA, Protein and Function Levelsa

−: No detection. +: Detection. n.d: Not determined. (−): No detection, but samples contained small amount of protein. 1*: means no detection in Western blot, detection in IHC. 2*: means no detection in IHC, detection in Western blot. a

obtained results are summarized in Table 2 and discussed in detail below. 4.1. MRP1 and MRP5 Transporter Proteins. The results of the mRNA and protein expression analysis for MRP1 showed a clear, consistent pattern for all of the tested corneal models. No difference in mRNA expression was detectable between the in vitro models, human epithelial corneal cells or ex vivo rabbit and porcine corneas. Previous published investigations have reported contradictory findings concerning mRNA expression in human epithelial corneal cells, though the results obtained in the present study are consistent with the majority of data in the literature.41−44 Our results regarding MRP1 protein expression in humans and rabbits are in accord with previous results published by Vellonen et al.41 In contrast, in porcine cornea no MRP1 expression could be detected at the protein level in the current work (through either Western blotting or immunohistochemistry), and no data are reported in the literature for comparison. However, the expression pattern of the MRP5 transporter did not show the same cohesion observed for MRP1. At the mRNA level, similar expression was detected in the HCE-T epithelial model as has previously been reported.41,45 On the other hand, the rabbit cornea specimens did not show a positive signal in the PCR analysis, though this could have been caused by the use of a nonspecific human primer. Conversely, Karla et al. reported mRNA expression in two rabbit corneal cell culture models (primary cells and SIRC cell line), but excised rabbit corneas had not been analyzed previously.45,46 A strong correlation between the PCR results and the immunohistochemistry images was found for the MRP5 transporter; i.e., FITC was detected in all of the tested corneal models except in rabbit cornea, where no signal was observed. This effect has previously been observed in human cornea specimens, and a good correlation between in vitro models and the in vivo situation is therefore indicated.41,45 The Western blotting results for the HCE-T epithelial model were contradictory to previous published results; specifically, we did

test. This test has been employed for years due to the lack of alternative systems but is increasingly criticized due to ethical conflicts and the different characteristics of rabbit and human eyes.40 Thus, in recent decades, several in vitro corneal cell culture models have been established as a replacement for the Draize test as well as for drug absorption studies. However, the extent of equivalence and, hence, the transferability of the results from in vitro corneal models to in vivo human corneas has remained unclear for quite some time. Nevertheless, extensive characterization and validation of such in vitro models is required to validate the use of these models in studies addressing drug regulatory processes. Such characterization must comprise not only histological studies but also an evaluation of barrier properties in comparison with excised tissue. In this regard, both the expression of similar passive diffusion barriers and the equivalent expression of transporter proteins are of interest. Recently, we reported the development of a serum-free cultivated Hemicornea construct for use in drug absorption studies in our laboratory, which exhibits equivalent barrier characteristics for passively transported drugs to those assumed for human corneas.16 Furthermore, in a prevalidation study, we were able to demonstrate, after transfer of the method to three laboratories, that our Hemicornea construct possessed similar permeabilities for seven marker compounds and showed a higher degree of reproducibility, including presenting lower intra- and interlaboratory variability, compared to excised animal corneas.31 In the current study, we compared the expression of the MDR1, MRP3 and BCRP efflux transporters in Hemicornea construct and excised corneas and found strong species-dependent expression patterns as well as a high degree of similarity between the Hemicornea construct and human corneal cells.32 The presented work was performed to further characterize the Hemicornea construct with regard to the expression of the MRP1, 2, 4 and 5 efflux transporters and compare the observed expression pattern with widely used ex vivo models. The 2168



Article

not detect MRP5 protein in the HCE-T-based models.41,45 However, the functional analysis of MRP1 and MRP5 revealed a greater efflux of CDCF in the case of HCE-T epithelial model, Hemicornea construct and rabbit corneas. These findings are in accord with previous results from the literature.41,45 Investigations performed using the Clonetics model (a commercially available cell culture model of the human corneal epithelium, introduced mainly for toxicity testing) showed no functional effect of MRP1, even though MRP1 mRNA expression was reported by these authors.43 These transport analyses were not performed in excised human corneas due to the difficulties of obtaining healthy, intact cornea samples, and only mRNA expression and Western blot data are therefore available. Hence, no final conclusion could be drawn in terms of the comparability of the expression patterns between Hemicornea construct and human cornea. However, taking the results of the present study and results from the literature into consideration, it is probable that MRP1 and, likely, MRP5 are expressed in human corneas as well as in HCE-T-based corneal models. 4.2. MRP2 Transporter Protein. The investigation of apically located MRP2 in the HCE-T epithelial model and the Hemicornea construct revealed some differences in the obtained expression patterns. Both models exhibited bands in the PCR experiments, indicating expression of MRP2 at the mRNA level, which has also been reported in HCE-T cells as well as the Clonetics model and human corneas by three other groups.42,43,47 On the contrary, Vellonen et al. reported detecting no mRNA expression of MRP2 in HCE-T cells.41 Interestingly, at the protein level (based on Western blotting and immunohistochemistry) only the HCE-T epithelial model possessed detectable MRP2 expression, while we failed to detect protein expression in the Hemicornea construct. This finding in the HCE-T epithelial model is confirmed by data from other laboratories obtained by performing Western blotting47and immunofluorescence42 analysis. The transport assays using erythromycin indicated the presence of MRP2 transporter function in the HCE-T epithelial model, but not in the Hemicornea construct. Our results suggest that more organotypic culture conditions, particularly cocultivation with HCK cells in a collagenous stromal biomatrix, led to different MRP2 expression patterns between the pure HCE-T epithelial model and Hemicornea construct. In rabbit cornea a similar pattern to that observed in the HCE-T epithelial model was obtained, as reported previously, indicating that MRP2 is expressed in rabbit corneal tissue.46−48 A few experiments have been performed with human corneas, but no consistent results have been reported previously.41−43 In summary, these findings support the hypothesis that MRP2 is not functionally expressed in human corneas, similar to our results from the Hemicornea construct. Furthermore, in the present study, porcine cornea did not show MRP2 expression, in contrast to rabbit cornea, again demonstrating the existence of species-dependent differences in efflux transporter expression in corneal tissue.32 4.3. MRP4 Transporter Protein. Our results demonstrate the existence of strong mRNA and protein expression as well as MRP4 functionality in both HCE-T-based in vitro models. Unfortunately, the results from our PCR and Western blot experiments do not allow us to make a conclusion regarding MRP4 expression in ex vivo human corneas. In the literature, MRP4 expression analysis of HCE-T cells have been reported, indicating that MRP4 expression occurs at the mRNA and

protein levels, whereas excised human corneas exhibit a positive signal in PCR assays, but show no bands in Western blot analysis.41 The ex vivo animal models examined in the present study showed no PCR bands, but obvious protein expression and MRP4 function were detected in porcine cornea. On the contrary, in the rabbit cornea neither protein expression nor efflux function was observed. Unfortunately, no investigations addressing MRP4 transporter expression in porcine and rabbit corneas have been reported in the literature; the expression of MRP4 has only been confirmed in excised rat corneas.49 However, our results again suggest that there are different expression patterns in porcine and rabbit corneas. The expression of MRP4 in humans remains unclear, although the detection of this transporter in both human in vitro models suggests that it is expressed in human tissue in vivo.

5. CONCLUSION The results of the present study indicate that the human HCE-T epithelial corneal model expresses MRP1, 2, 4 and 5 at the mRNA, protein and function levels. The more organotypic Hemicornea construct exhibits the same MRP expression pattern, although no protein expression of MRP2 is detectable. In abraded human corneal epithelium, mRNA expression of all four examined MRPs was found, as observed for both HCE-T-based models. Unfortunately, more precise predictions regarding the functional expression of MRP transporters in human cornea and direct comparisons with both human in vitro models were not possible due to the lack of donor tissue. In addition, the rabbit and porcine corneas display great deviations in their MRP expression patterns compared to the human in vitro corneal models and in comparison with each other. Our investigation again reveals species-dependent differences in the expression of efflux transporter proteins in corneal tissue and confirms that the transfer of data obtained in animal tissue to humans should be undertaken with caution. Considering our results as well as previous publications, the HCE-T-based models, especially the Hemicornea construct, seem to exhibit better equivalence with the human cornea than the most commonly used rabbit cornea model. Nevertheless, further research is necessary to obtain a better understanding of the transferability between the in vitro and ex vivo corneal models and between cell culture models and human cornea.

■

AUTHOR INFORMATION

Corresponding Author

*Institut für Pharmazeutische Technologie, Technische Universität Braunschweig, Mendelssohnstrasse 1, Braunschweig 38106, Germany. Tel: 0049 531 391 5651. Fax: 0049 531 391 8108. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to the German Federal Institute for Risk Assessment (BfR), which funded this work under grant no. FK 3-1328-306-52054369. J.V. thanks Deutscher Akademischer Austausch Dienst (DAAD) for a scholarship. Furthermore, the authors would like to thank U. Scheider and Dr. M. Meyer (Cornea Bank, MHH Hannover Germany) for supplying the human donor corneal epithelium and Dr. M. Zorn-Kruppa (UKE Hamburg Germany) and Dr. K. Araki-Sasaki (Kagoshi2169



Article

(19) Minami, Y.; Sugihara, H.; Oono, S. Reconstruction of cornea in three-dimensional collagen matrix culture. Invest. Ophthalmol. Visual Sci. 1993, 34, 2316−2324. (20) Reichl, S.; Müller-Goymann, C. C. Development of an organotypic corneal construction as an in vitro model for permeability studies. Ophthalmologe 2001, 98, 853−858. (21) Schneider, A. I.; Maier-Reif, K.; Graeve, T. The use of an in vitro cornea for predicting ocular toxicity. In Vitro Toxicol. 1997, 10, 309− 318. (22) Werner, A.; Braun, M.; Reichl, S.; Kietzmann, M. Establishing and functional testing of a canine corneal construct. Vet. Ophthalmol. 2008, 11, 280−289. (23) Chang, J.-E.; Basu, S.; Lee, V. Air-interface condition promotes the formation of tight corneal epithelial cell layers for drug transport studies. Pharm. Res. 2000, 17, 670−676. (24) Griffith, M.; Osborne, R.; Munger, R.; Xiong, X.; Doillon, C.; Laycock, N.; Hakim, M.; Song, Y.; Watsky, M. Functional human corneal equivalents constructed from cell lines. Science 1999, 286, 2169−2172. (25) Reichl, S.; Bednarz, J.; Müller-Goymann, C. C. Human corneal equivalent as cell culture model for in vitro drug permeation studies. Br. J. Ophthalmol. 2004, 88, 560−565. (26) Zorn-Kruppa, M.; Tykhonova, S.; Belge, G.; Bednarz, J.; Diehl, H. A.; Engelke, M. A human corneal equivalent constructed from SV40-immortalised corneal cell lines. ATLA, Altern. Lab. Anim. 2005, 33, 37−45. (27) Becker, U.; Ehrhardt, C.; Schneider, M.; Muys, L.; Gross, D.; Eschmann, K.; Schaefer, U. F.; Lehr, C. M. A comparative evaluation of corneal epithelial cell cultures for assessing ocular permeability. ATLA, Altern. Lab. Anim. 2008, 36, 33−44. (28) Toropainen, E.; Ranta, V. P.; Suhonen, P.; Urtti, A. Culture model of human corneal epithelium for prediction of ocular drug absorption. Invest. Ophthalmol. Visual Sci. 2001, 42 (12), 2942−2948. (29) Reichl, S. Cell culture models of the human cornea - a comparative evaluation of their usefullness to determine ocular drug absorption in vitro. J. Pharm. Pharmacol. 2008, 60, 299−307. (30) Engelke, M.; Zorn-Kruppa, M.; Gabel, D.; Reisinger, K.; Rusche, B.; Mewes, K. R. A human hemi-cornea model for eye irritation testing: quality control of production, reliability and predictive capacity. Toxicol. In Vitro 2013, 27 (1), 458−468. (31) Hahne, M.; Zorn-Kruppa, M.; Guzman, G.; Brandner, J.; Haltner-Ukomado, E.; Wätzig, H.; Reichl, S. Prevalidation of a human cornea construct as an alternative to animal corneas for in vitro drug absorption studies. J. Pharm. Sci. 2012, 101, 2976−2988. (32) Verstraelen, J.; Reichl, S. Expression analysis of MDR1, BCRP and MRP3 transporter proteins in different in vitro and ex vivo cornea models for drug absorption studies. Int. J. Pharm. 2013, 441, 765−775. (33) Araki-Sasaki, K.; Ohashi, Y.; Sasabe, T.; Hayashi, K.; Watanabe, H.; Tano, Y.; Handa, H. An SV40-immortalized human corneal epithelial cell line and its characterization. Invest. Ophthalmol. Visual Sci. 1995, 36, 614−621. (34) Fogh, J.; Trempe, G. New human tumor cell lines. In Fogh, J., Ed.; Human Tumor Cells In Vitro; Plenum Press: New York, 1975; pp 115−159. (35) Artursson, P.; Palm, K.; Luthman, K. Caco-2 monolayers in experimental and theoretical predictions of drug transport. Adv. Drug Delivery Rev. 2001, 46, 27−43. (36) Maubon, N.; Le Vee, M.; Fossati, L.; Audry, M.; Le Ferrec, E.; Bolze, S.; Fardel, O. Analysis of drug transport expression in human intestinal Caco-2 cells by real-time PCR. Fundam. Clin. Pharmacol. 2007, 21, 659−663. (37) Bradford, M. M. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 1976, 72, 248−254. (38) Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227, 680− 685.

ma Japan) for their generous gifts of the HCK and HCE-T cell lines. We further gratefully acknowledge J. Bartels and M. Busker (Institute of Pharmacology and toxicology, University Braunschweig, Germany) for their support in using the CLSM.

■

REFERENCES

(1) Järvinen, K.; Järvinen, T.; Urtti, A. Ocular absorption following topical delivery. Adv. Drug Delivery Rev. 1995, 16, 3−19. (2) Sasaki, H.; Yamamura, K.; Nishida, K.; Nakamura, J.; Ichikawa, M. Delivery of drugs to the eye by topical application. Prog. Retinal Eye Res. 1996, 15, 583−620. (3) Chen, L.; Ebihara, N.; Fujiki, K.; Murakami, A. Expression and distribution of junctional adhesion molecule-1 in the human cornea. Jpn. J. Ophthalmol. 2007, 51, 405−411. (4) Prausnitz, M. R.; Noonan, J. S. Permeability of cornea, sclera, and conjunctiva: a literature analysis for drug delivery to the eye. J. Pharm. Sci. 1998, 87, 1479−1488. (5) Lawrence, M. S.; Miller, J. W. Ocular tissue permeabilities. Int. Ophthalmol. Clin. 2004, 44 (3), 53−61. (6) Majumdar, S.; Hingorani, T.; Srirangam, R. Evaluation of active and passive transport processes in corneas extracted from preserved rabbit eyes. J. Pharm. Sci. 2010, 99 (4), 1921−1930. (7) Ranta, V. P.; Laavola, M.; Toropainen, E.; Urtti, A. Ocular pharmacokinetic modeling using corneal absorption and desorption rates from in vitro permeation experiments with cultured corneal epithelial cells. Pharm. Res. 2009, 20 (9), 1409−1416. (8) Mannermaa, E.; Vellonen, K. S.; Urtti, A. Drug transport in corneal epithelium and blood-retina barrier: emerging role of transporters in ocular pharmacokinetics. Adv. Drug Delivery Rev. 2006, 58, 1136−1163. (9) Toropainen, E.; Ranta, V. P.; Vellonen, K. S.; Palmgren, J.; Talvitie, A.; Laavola, M.; Suhonen, P.; Hamalainen, K. M.; Auriola, S.; Urtti, A. Paracellular and passive transcellular permeability in immortalized human corneal epithelial cell culture model. Eur. J. Pharm. Sci. 2009, 20, 99−106. (10) Kidron, H.; Vellonen, K.-S.; Del Amo, E.; Tissari, A.; Urtti, A. Prediction of the corneal permeability of drug-like compounds. Pharm. Res. 2010, 27, 1398−1407. (11) Meyer, L.; Bednarz, J.; Müller-Goymann, C. C.; Reichl, S. Esterase activity of human organotypic cornea construct (HCC) as in vitro model for permeation studies. Ophthalmologe 2005, 102, 971− 980. (12) Lee, V.; Carson, L. W.; Stratford, R. E. Metabolic and permeation barriers to the ocular absorption of topically applied enkephalins in albino rabbits. J. Ocul. Pharmacol. 1986, 2 (4), 345− 352. (13) Jones, P. M.; George, A. M. The ABC transporter structure and mechanism: perspectives on recent research. Cell. Mol. Life Sci. 2004, 61, 682−699. (14) Dean, M.; Rzhetsky, A.; Allikmets, R. The human ATP-binding cassette (ABC) transporter superfamily. Genome Res. 2002, 11, 1156− 1166. (15) Borst, P.; Evers, R.; Kool, M.; Wijnholds, J. A family of drug transporters: the multidrug resistance-associated proteins. J. Natl. Cancer Inst. 2000, 92 (16), 1295−1302. (16) Hahne, M.; Reichl, S. Development of a serum-free human cornea construct for in vitro drug absorption studies: the influence of varying cultivation parameters on barrier characteristics. Int. J. Pharm. 2011, 416, 268−279. (17) Dahlin, A.; Geier, E.; Giacomini, K. M. Gene expression profiling of transporters in the solute carrier and ATP-binding cassette superfamilies in human eye substructures. Mol. Pharmaceutics 2013, 10, 650−663. (18) Tegtmeyer, S.; Reichl, S.; Müller-Goymann, C. C. Cultivation and characterization of a bovine in vitro model of the cornea. Pharmazie 2004, 59, 464−471. 2170



Article

(39) Ming, X.; Thakker, D. R. Role of basolateral efflux transporter MRP4 in the intestinal absorption of the antiviral drug adefovir dipivoxil. Biochem. Pharmacol. 2010, 79, 455−462. (40) Wilhelmus, K. R. The Draize eye test. Surv. Ophthalmol. 2001, 45, 493−515. (41) Vellonen, K.; Mannermaa, E.; Turner, H.; Häkli, M.; Wolosin, J.; Tervo, T.; Honkakoski, P.; Urtti, A. Effluxing ABC-transporters in human corneal epithelium. J. Pharm. Sci. 2010, 99, 1087−1098. (42) Becker, U.; Ehrhardt, C.; Daum, N.; Baldes, C.; Schaefer, U. F.; Ruprecht, K. W.; Kim, K. J.; Lehr, C. M. Expression of ABCtransporters in human corneal tissue and the transformed cell line, HCE-T. J. Ocul. Pharmacol. Ther. 2007, 23, 172−181. (43) Xiang, C. D.; Batugo, M.; Gale, D. C.; Zhang, T.; Ye, J.; Li, C.; Zhou, S.; Wu, E. Y.; Zhang, E. Y. Characterization of human corneal epithelial cell model as a surrogate for corneal permeability assessment: metabolism and transport. Drug Metab. Dispos. 2009, 37, 992−998. (44) Zhang, T.; Xiang, C. D.; Gale, D.; Carreiro, S.; Wu, E. Y.; Zhang, E. Y. Drug transporter and cytochrome P450 mRNA expression in human ocular barriers: implications for ocular drug disposition. Drug Metab. Dispos. 2008, 36, 1300−1307. (45) Karla, P. K.; Mitra, A. K. Expression of multidrug resistance assiociated protein 5 (MRP5) on cornea and its role in drug efflux. J. Ocul. Pharmcol. Ther. 2009, 25 (2), 121−132. (46) Karla, P. K.; Mitra, A. K. Molecular evidence and functional expression of multidrug resistance associated protein (MRP) in rabbit corneal epithelial cells. Exp. Eye Res. 2007, 84, 53−60. (47) Karla, P. K.; Mitra, A. K. Molecular evidence and functional expression of a novel drug efflux pump (ABCC2) in human corneal epithelium and rabbit cornea and its role in ocular drug efflux. Int. J. Pharm. 2007, 336, 12−21. (48) Hariharan, S.; Mitra, A. K. Enhanced corneal absorption of erythromycin by modulating p-glycoprotein and MRP mediated efflux with corticosteroids. Pharm. Res. 2009, 26, 1270−1284. (49) Li, B.; Lim, J. C. Characterization of gluthathione uptake, synthesis and efflux pathways in the epithelium and endothelium of the rat cornea. Cornea 2012, 31, 1304−1312.

2171


Multidrug Resistance-Associated Protein (MRP1, 2, 4 and 5

Recommend Documents