|Molecular Vision 2002;
Received 5 September 2002 | Accepted 6 November 2002 | Published 11 November 2002
P-glycoprotein expression in human retinal pigment epithelium
Brian G. Kennedy,
Nancy J. Mangini
Northwest Center for Medical Education, Indiana University School of Medicine, Gary, IN
Correspondence to: Brian G. Kennedy, PhD, Indiana University School of Medicine, 3400 Broadway, Gary, IN, 46408; Phone: (219) 980-6520; FAX: (219) 980-6566; email email@example.com
Purpose: The retinal pigment epithelium (RPE) is a transporting epithelial monolayer that controls hydration and composition of the subretinal space. P-glycoprotein is an ATP-binding cassette transport protein known to transport a wide range of hydrophobic compounds. The expression of P-glycoprotein in barrier epithelial cells suggests that it could serve a normal protective function, possibly clearing potentially harmful substances from sensitive compartments, like the subretinal space. The present study is designed to determine the expression and activity of P-glycoprotein in normal human RPE.
Methods: RT-PCR and direct sequencing were employed to examine the presence of mdr1 mRNA in cultured human RPE. P-glycoprotein-specific antibodies were employed in Western blotting to identify P-glycoprotein in cultured human RPE and in an established RPE cell line (D407). Anti-P-glycoprotein antibodies were also used to localize the protein in frozen, formaldehyde-fixed sections of native human RPE/choroid by immunohistochemistry. Finally, rhodamine uptake was performed in cultured human RPE monolayers to assess P-glycoprotein activity. The inhibitory antibody 4E3 and reversins 121 and 205 were used to block transport activity.
Results: P-glycoprotein is expressed, and is active, in human RPE tissue not exposed to any known inducers of P-glycoprotein. RT-PCR yielded a 546 bp product that was 100% identical in sequence to published data for the mdr1 isoform of human P-glycoprotein. Western blotting demonstrated expression at the protein level, with specific bands observed at about 220 and 165 kD. In native tissue, P-glycoprotein immunoreactivity was predominantly membrane associated, with localization to both apical and basolateral cell membranes. Finally, P-glycoprotein expressed in human RPE is active. Steady-state rhodamine accumulation was increased in the presence of compounds reported to block P-glycoprotein mediated rhodamine efflux.
Conclusions: Human RPE, not exposed to inducer treatment, expresses P-glycoprotein with localization to both apical and basal cell surfaces. Basolateral P-glycoprotein could serve a protective function for the neural retina helping to clear unwanted substances from subretinal space. The finding that P-glycoprotein is also on the apical surface suggests possible additional roles for P-glycoprotein in the RPE.
The retinal pigment epithelium (RPE), situated between the sensory retina and its choroidal blood supply, forms a diffusion barrier controlling access to subretinal space, with RPE membrane transport proteins regulating hydration and ionic composition of the subretinal space. Photoreceptor sensitivity and viability thus depends on RPE catalyzed transport activity. Proteins functioning in ionic, sugar, peptide and water transport have been characterized in the RPE . Members of the ATP-binding cassette (ABC) family of transport proteins have been shown to perform multiple critical functions in a range of epithelial cells . However, relatively little has been done to characterize these transport proteins in the RPE. The present study was designed to determine if P-glycoprotein, an ATP-dependent drug efflux pump, is expressed and functional in normal human RPE.
Human P-glycoprotein, encoded by the mdr1 gene, is a member of the ABC family of ATPases . P-glycoprotein, notable for its broad substrate specificity, was first described in tumor cells where it confers (in conjunction with other transport systems) multidrug resistance [4,5]. P-glycoprotein expression has also been documented in normal tissues, most notably in epithelial cells, where it is postulated to serve a protective function, limiting the accumulation of cytotoxic compounds [6-11]. In this regard, P-glycoprotein has been detected in both pigmented and non-pigmented bovine ciliary epithelium . Additional functions in which P-glycoprotein has been implicated include ATP release [13-15], Cl- transport [12,16], steroid secretion  and lipid translocation [18-20]
The antimitotic agent daunomycin, a potent inducer of P-glycoprotein expression , has been used clinically to prevent reproliferation after vitrectomy . P-glycoprotein expression has been documented in native RPE cells from patients exposed to intravitreal daunomycin, as well as in cultured RPE cells grown in the presence of the drug . This earlier work shows that P-glycoprotein expression can be induced in RPE cells and that this expression can be important in a clinical setting. It remains an open question whether P-glycoprotein is constitutively expressed in normal (i.e., uninduced) RPE and, if expressed, what its ongoing function may be.
The present work is intended to assess activity and subcellular localization of P-glycoprotein in RPE. Expression was examined in adult human RPE, employing both cultured cells and native RPE/choroid preparations.
Cell cultures were established from adult human donor eyes obtained from the National Disease Research Interchange (NDRI). Institutional Review Board (Indiana University, Purdue University, Indianapolis) approval for use of human donor tissue was obtained. No ocular disease was reported for any donor. As described previously [23-25], to initiate cultures small patches of RPE/choroid (about 0.2 cm2) were dissected and placed, RPE side down, in 60 mm culture dishes (Falcon, Primaria, BD Biosciences, Bedford MA). RPE cells, which grow out from the explant, were harvested by trypsinization and then maintained in 25 cm2 flasks (Falcon, Primaria). Cells were cultured in Eagle's Minimum Essential Medium with Earle's salts (Mediatech Cellgro, Fisher Scientific) with 10% fetal calf serum, 5% newborn calf serum, essential and non-essential amino acids, 4 mM l-glutamine, 0.5 μg/ml amphotericin B and 10 μg/ml gentamicin at 37 °C with 5% CO2 in air. Cells were subcultured at 1 to 2 week intervals by trypsinization. These cells are reactive for cytokeratin-18 and can attain a hexagonal packing array with pigmentation . These cells have been used to examine the Na+: Ca2+ exchange protein , plasma membrane calcium ATPase , arrestin , transthyretin , volume regulation , Na-K-Cl cotransport , and creatine kinase  in RPE.
D407 cells, an established RPE cell line, were obtained as a generous gift from Dr. Richard Hunt (University of South Carolina Medical School, Columbia SC). The RPE characteristics of these cells have been described . D407 cells were maintained and passed as described by Hunt and coworkers .
The membrane preparation, comprising a crude membrane fraction, was prepared from confluent human RPE cultures grown on 175 cm2 flasks. Flasks (typically four flasks per preparation) were rinsed free of culture medium with Ca2+/Mg2+-free buffer (145 mM NaCl, 5 mM KCl, 5 mM NaHCO3, 15 mM Hepes, 8 mM Tris base, 0.5 mM EDTA; pH 7.4, osmolarity 290 mOs). Cells were harvested by scraping into standard buffer (140 mM NaCl, 5 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 5 mM glucose, 15 mM Hepes, 8 mM Tris base; pH 7.4, osmolarity 290 mOs). The suspension was pelleted by centrifugation at 1500x g for eight minutes and the cells washed once with standard buffer. A final pellet was obtained by centrifugation at 2000 g for 12 min. Cells were suspended in mannitol buffer (200 mM mannitol, 50 mM Hepes, 40 mM Tris; pH 7.4) supplemented with a protease inhibitor cocktail (Complete Protease Inhibitor, Roche Applied Science) and disrupted by nitrogen cavitation (900 to 1200 psi for 30 min at 4 °C) in a cell disruption bomb (Parr Instruments). Unbroken cells were removed by centrifugation (1500x g for 8 min at 4 °C). The resulting supernatant was centrifuged at 30,000x g for 30 min at 4 °C to obtain the final membrane pellet. The pellet was suspended in mannitol buffer, sampled for protein determination and then stored at -80 °C.
Western blot analysis
Western blot analysis was performed on membrane samples (prepared as described above) and on whole cell lysates (from cultured adult human RPE or from D407 cells). Whole cell lysates were obtained from confluent monolayers growing on 25 cm2 culture flasks. Monolayers were washed free of culture medium with either Ca2+/Mg2+-free buffer or standard buffer and then harvested by scraping into the washing buffer. Cells were pelleted by centrifugation at 1500x g for 8 min. The pellet was suspended in either Ca2+/Mg2+-free buffer or standard buffer, sampled for protein determination and then dissolved in electrophoresis buffer (2% SDS, 10% glycerol, 200 mM Hepes at pH 6.8, 1 mM EDTA, 0.1% bromophenol blue and protease inhibitor cocktail). Samples in electrophoresis buffer were stored at -20 °C.
For electrophoresis, samples were heated for 10 min at 37 °C and then resolved on 6% Laemmli gels . Protein was transferred to PVDF membranes by semi-dry blotting. To detect P-glycoprotein immunoreactive bands, membranes were first blocked (1 h at room temperature) with Super Block (Pierce). Incubation with primary antibody was for 2 h at room temperature. Primary antibodies were diluted in Tris-buffered saline (500 mM NaCl, 25 mM Tris base, pH 7.5) containing 0.4% blotting grade non-fat dry milk (Bio-Rad Laboratories) and 0.2% tween-20. Two different anti-P-glycoprotein primary antibodies were used for Western blot studies. Monoclonal antibody (mAb) JSB-1 (Chemicon, Temecula CA), which recognizes a conserved intracellular epitope of plasma membrane associated P-glycoprotein , was used at a final concentration of 0.5 μg/ml. Polyclonal antibody mdr (Ab-1; Oncogene Research Products, Boston MA), raised against a 21 amino acid C-terminal peptide, was used at 1 μg/ml. Secondary antibodies (Vector Labs, Inc., Burlingame, CA) were alkaline phosphatase conjugated and were diluted (1:5000) in the same solution used for the primary antibodies. Blots were reacted with secondary antibodies for 1 h at room temperature. The final incubation (30 min at room temperature) employed CDP-Star chemiluminescent substrate (PerkinElmer Life Sciences). The chemiluminescent signal was digitally captured with an Image Station 440 (Kodak); band densities and molecular sizes were computed with 1D Image Analysis software (Kodak).
Total RNA was isolated from primary cultures of human adult RPE or the established RPE cell line, D407, using RNAqueous (Ambion). For single-strand cDNA synthesis, 5 μg of total RNA from each sample was reverse transcribed using Superscript II reverse transcriptase (Gibco, Life Technologies) and random hexamer priming. Amplification was performed with 2-μl aliquots of cDNA, using High Fidelity PCR Master (Roche Diagnostics, Indianapolis, IN) and a human mdr1-specific primer set selected based on published work  and that spanned more than two intron-exon junctions . The forward primer (5'-tttcattttggtgcctggcagc-3') and the reverse primer (5'-agaaggccagagcataagatgc-3') correspond to the nucleotide sequences 517-538 and 1062-1040 for human mdr1 (Genbank accession number AF016535). The predicted size of the PCR product was 546 base pairs. PCR was carried out in a GeneAmp 9600 thermocycler (Applied Biosystems) using the following cycles: initial 2 min step at 94 °C then 30 cycles of 94 °C for 30 s, 55 °C for 30 s, 68 °C for 3 min; ending with 72 °C for 7 min. PCR products were separated by electrophoresis through 1% agarose-0.5x TBE gels containing 0.5 μg/ml ethidium bromide. As a negative control, reverse transcription of RPE RNA was performed without reverse transcriptase. To confirm the identity of the expected PCR products (546 bp), PCR DNAs were purified (Hi-Pure PCR Purification Kit, Roche, Indianapolis IN) and directly sequenced using the mdr1 primers.
Post mortem eyes were obtained from the National Disease Research Interchange under local IRB approval for the use of human tissue. For the present immunocytochemistry study, eyes from six donors (age range 21 to 79 years) were fixed in 4% formaldehyde in phosphate buffered saline (time into fix approximately 2 to 12 h post enucleation). The data presented in Figure 1A was from a 66 year old male donor (time from enucleation to fixation, 7.5 h) and the data presented in Figure 1B,C was from a 79 year old female donor (time from enucleation to fixation, 2 h). RPE/choroid patches were dissected and 6-8 μm-thick sections were obtained by cryostat.
Immunohistochemistry was performed with three anti-P-glycoprotein antibodies, each recognizing a different epitope. MAb 4E3 (DAKO Corporation, [36,37]) and mAb UIC2 (Immunotech [38,39]) bind to different, conformational extracellular peptide epitopes and can inhibit P-glycoprotein function; mAb JSB-1 (Chemicon) reacts with a highly conserved cytoplasmic epitope . MAb 4E3 reactivity was examined in tissue from three donors; JSB-1 and UIC2 were tested on tissue from 5 donors. Control incubations with non-immune mouse IgG replacing primary antibody were non-reactive.
Prior to antibody incubation, sections reacted with mAb JSB-1 were permeabilized with 1% SDS in PBS (130 mM NaCl, 2.5 mM KCl, 8 mM Na2HPO4, 1.5 mM KH2PO4) for 5 min at room temperature, while sections reacted with mAbs binding to extracellular epitopes (4E3 and UIC2) were not permeabilized. All sections were treated with 0.25% KMNO4 in PBS for 15 min at room temperature to minimize tissue autofluorescence [30,40,41]. Sections were blocked with 20% horse serum for 1.5 h at 37 °C in a humidified chamber prior to primary and secondary antibody incubations. Primary and secondary antibodies were diluted in PBS containing 1% goat serum. MAb JSB-1 was used at a final concentration of 10 μg/ml; while mAbs UIC2 and 4E3 were used at 20 μg/ml. FITC-conjugated anti-mouse secondary antibody (Vector) was used at a dilution of 1:100. Antibody incubations were carried out at 37 °C in a humidified chamber for 1.5 to 2 h. Sections were mounted in Vectashield (Vector Labs, Inc.) containing DAPI (Figure 1A) or propidium iodide (Figure 1B,C) to stain cell nuclei. Anti-P-glycoprotein localization was visualized either using an inverted microscope equipped for epifluorescence (Zeiss Axiovert 100M) and analyzed using deconvolution software (Slidebook, Intelligent Imaging Innovations, Denver, CO; see Figure 1A), or using a Leitz Laborlux S epifluorescence microscope with images captured using a Magnifier CCD camera (Figure 1B,C). Digital images were further processed using Photoshop (Adobe Systems).
Rhodamine uptake, performed in cultured HRPE monolayers grown to confluence on 35 mm culture dishes, followed an established protocol . Monolayers were washed free of growth media with standard buffer (composition noted above) and then preincubated for 20 min at 37 °C in a humidified chamber with standard buffer supplemented with inhibitor or control vehicle (as specified in the figure legend). Uptake was initiated by a change of medium to one containing rhodamine at a concentration of 20 μm. Monolayers were incubated for 45 min at 37 °C in a humidified chamber. Uptake was terminated by aspiration of the rhodamine-containing medium followed by three rapid washes with ice-cold rhodamine-free standard buffer. Cells were lysed by exposure to 2 ml SDS buffer (20 mM Tris, 0.2% SDS; pH 7.7), with the entire lysate transferred to a fluorometer cuvette. Total rhodamine fluorescence was measured with a spectrofluorometer (Farand Instruments). Excitation wavelength was 500 nm and emission wavelength was 530 nm.
Rhodamine 123 (Molecular Probes) was diluted from a 20 mM stock solution in ethanol. Reversin 121 and reversin 205 (Sigma) were diluted from 20 mM stock solutions in DMSO.
Results from RT-PCR analysis are presented in Figure 2. Lane 1 is from cultured human RPE while lane 2 is from the RPE cell line, D407. Both lanes contain a prominent single band at about 546 bp, the size expected based on the set of mdr1 primers used. The PCR DNA's were purified and directly sequenced using the same primers. The resulting sequences were identical to the published data for human mdr1.
Western blot analysis
An example of a western blot of P-glycoprotein obtained with the polyclonal antibody mdr (Ab-1) is shown in Figure 3. Figure 3A,B are identical except in Figure 3B the primary antibody was preabsorbed with immunizing peptide. Lanes 1 and 4 contain a membrane preparation from cultured human RPE. The other lanes are whole cell human RPE lysates (lanes 2 and 5 were from cells at passage 7 and lanes 3 and 6 were from cells at passage 1). In Figure 3A, two bands (identified by the arrows) are detected at about 224 and 165 kD in all the samples. (A band at 224 kD is present in lane 1, though the staining intensity was less than in the other samples). Figure 3B shows that staining of both these bands was completely eliminated by preabsorption with the specific immunizing peptide.
Western blot analysis using antibody mdr (Ab-1) was performed on five different gels, examining whole cell lysates from five different donor cultures. In all cases, both of the bands shown in Figure 3 were detected. The mean molecular weights evaluated over all the blots and including all extracts were 221±0.7 kD for the upper band and 165±0.9 kD for the lower band. In addition, membrane preparations were run on four blots. Average molecular weights were as follows: 222±1.2 kD for the upper band and 161±1.8 kD for the lower band (the four values quoted above are all mean±SEM).
To confirm P-glycoprotein expression in human RPE, another antibody reactive to a distinct protein epitope, was also examined in Western blotting experiments. JSB-1 is a mAb reactive to a conserved intracellular epitope. As shown in Figure 4, JSB-1 reacts with a single band in all samples. The molecular weight of the band, in this run, is about 225 kD. JSB-1 was tested on blots examining whole cell lysates from human RPE cultures established from seven different donors. The banding pattern was equivalent in all runs, with an average molecular weight of 226±1.2 kD (mean±SEM). For comparison, D407 cells were also examined (see Figure 4, lanes 3 and 4). As for the primary RPE cultures, a single band at about 225 kD was detected. Finally, membrane preparations were also examined with JSB-1 (data not shown). In four separate experiments, a single band at about 222±5.4 kD (mean±SEM) was detected.
To summarize, a high molecular weight band running at 220 to 225 kD was detected with both antibodies. The mdr (Ab-1) antibody was used with whole cell human RPE lysates and with the membrane preparation. JSB-1 was used with both of those samples as well as with cell lysates from D407 cells. The high molecular weight band was present in all five conditions. Though not proving that the same band is being detected in all conditions, a single factor ANOVA indicated that the average molecular weight for this band did not differ in any of the five conditions (p-value > 0.2).
The above results indicate that P-glycoprotein, the product of the mdr1 gene, is expressed in cultured human RPE cells. The following experiments were designed to determine whether the protein, as expressed, is functional. The activity assay used follows published work . As described by Sharom et al. , rhodamine 123 is a P-glycoprotein substrate that the protein exports from the cell. In cells expressing functional P-glycoprotein, steady-state rhodamine content is determined by passive drug influx balanced by all efflux mechanisms. Application of an agent that inhibits P-glycoprotein will cause an increase in the steady-state level of intracellular rhodamine. The absolute magnitude of P-glycoprotein-mediated transport cannot be determined since steady-state rhodamine levels depend on all ongoing rhodamine transport modalities. However, increased rhodamine levels, after P-glycoprotein inhibition, clearly indicates the presence of functional P-glycoprotein in the cell. In this analysis, the main caveat concerns the P-glycoprotein inhibitor that is employed. The chosen inhibitor should be specific for P-glycoprotein and, ideally, two unrelated inhibitors should be employed. In this regard, specific, high affinity P-glycoprotein inhibitors, the reversins, have recently been described . To complement functional studies employing the reversins, an extracellularly reactive, inhibitory mAb 4E3 is also available.
Figure 5 summarizes rhodamine uptake data in human RPE monolayers grown on 35 mm dishes. Reversin 205 data was obtained from five separate experiments, reversin 121 from three experiments. Triplicate dishes were run for each condition, in each experiment. For each experiment, rhodamine content was normalized to the control condition (DMSO vehicle). Specific P-glycoprotein inhibitors, the reversins, approximately double steady-state rhodamine accumulation, indicating that P-glycoprotein is active in cultured human RPE.
To further document P-glycoprotein functional activity in cultured human RPE, another highly specific P-glycoprotein inhibitor, chemically and mechanistically unrelated to the reversins, was also employed in the rhodamine assay. MAb 4E3, directed against an extracellular epitope of the protein, has been reported to block activity . Figure 6 shows that mAb 4E3, in a concentration-dependent manner, increases rhodamine content in cultured human RPE cells. The effect is specific for the anti-P-glycoprotein antibody since non-immune mouse IgG, as control, slightly decreased rhodamine content (data not shown). The effect of the anti-P-glycoprotein antibody is consistent with the data presented in Figure 5, indicating that P-glycoprotein is expressed, and active in cultured human RPE.
To further characterize P-glycoprotein activity in human RPE, the concentration dependence of reversin 121 interaction with the transporter were determined. Figure 7 presents intracellular rhodamine content, measured as total rhodamine fluorescence, as a function of reversin 121 concentration. Reversin 121 inhibits P-glycoprotein-mediated rhodamine transport (with a resulting increase in rhodamine content) in a concentration dependent manner. The K1/2 for the effect of reversin 121 is 1.6 μm. This is noteworthy, since the K1/2 for reversin inhibition of P-glycoprotein activity in a P-glycoprotein over-expressing cell line was 2.5 μM .
The final series of experiments were intended to document the presence of P-glycoprotein in native human RPE and to examine its subcellular localization in this epithelium. Immunostaining was performed on frozen sections of adult human RPE-choroid tissue obtained from six different postmortem donor eyes. Localization was assessed with three different anti-P-glycoprotein mAbs (UIC2, JSB-1 and 4E3) each of which reacts with a distinct epitope on the protein. P-glycoprotein reactivity was detected in sections of RPE/choroid from all six donors, though not every antibody was run on each donor. Figure 1 illustrates staining obtained with each antibody. P-glycoprotein immunoreactivity was observed in both the RPE layer (tissue orientation in the figure is RPE up) and in the choroid. P-glycoprotein reactivity in the RPE was robust. Notably, staining in native human RPE was observed on both apical and basal cell membranes. Sections reacted with non-immune mouse IgG, as a negative control, were unstained.
The purpose of this study was to document the expression, activity and membrane localization of P-glycoprotein in human RPE. One thorough study recently examined P-glycoprotein expression in RPE, predominantly in a clinical context . In that study, P-glycoprotein was detected in all specimens of epiretinal membranes obtained from patients exposed to the known P-glycoprotein inducer, daunomycin. By contrast, weak staining was seen in only 15% of membranes from control, unexposed subjects. In this same study, though mdr1 mRNA was detected in cultured human RPE cells, protein expression was only detected after induction with daunomycin. Recent work has documented P-glycoprotein expression in a wide range of normal, uninduced epithelial tissues [6-10]. Constitutive expression in these cells has been postulated to serve ongoing, essential functions. Specifically, P-glycoprotein-mediated transport activity has been implicated in drug/metabolite efflux [5-10], which is postulated to serve a protective function. In addition, the protein has also been implicated in ATP release [13-15], volume sensitive Cl- transport [12,16] and steroid secretion . Since the RPE contributes to the blood retinal barrier, with apically mediated transport activity regulating solute and ionic composition in subretinal space, ongoing transport activity in the RPE is critical to sustain photoreceptor function and viability. Given that, as noted above, P-glycoprotein can be expressed in the RPE after daunomycin induction, it is important to determine if the RPE, like many other barrier epithelia, also constitutively expresses the protein. If constitutively expressed, P-glycoprotein activity would contribute to the normal transport function of the RPE.
To summarize, the present study demonstrated P-glycoprotein expression in both cultured and native human RPE. Confirming published work , mdr1 mRNA was detected in cultured human RPE. Critically, the present work, in addition, showed mdr1 expression at the protein level in non-induced human RPE. Protein expression was documented in both cultured and native preparations. The protein was functional, as assessed by the rhodamine 123 assays.
Based on amino acid composition, unmodified P-glycoprotein should exhibit a molecular weight of about 140 kD. The reported molecular mass of the mature protein, estimated by SDS-PAGE, varies significantly from tissue to tissue, possibly due to differences in the extent of protein glycosylation. Reported molecular weights range from 200 kD [43-45] to approximately 150 kD [46,47]. The molecular size estimates of P-glycoprotein in human RPE is in line with these measurements.
Figure 3 shows a Western blot probed with the polyclonal antibody mdr (Ab-1). Two bands (about 224 and 165 kD) are clearly resolved. Reactivity is specific for P-glycoprotein since both bands are eliminated by preincubation with immunizing peptide (Figure 3B). Mi et al  also describe two bands in Western blot analyses using this polyclonal antibody to assess P-glycoprotein expression in several multidrug resistant cell lines. Additionally, Figure 3 shows that the relative intensity of the high and low molecular weight bands differs in the membrane preparation, as compared to whole cell lysates (lanes 2 and 3). This finding could indicate differential distribution between plasma membrane and a cytosolic membrane compartment. Meschini et al.  describe significant levels of intracellular P-glycoprotein in colon adenocarcinoma cells. Furthermore, inhibition of P-glycoprotein export from the cytosolic to the plasma membrane compartment has been proposed as one potential mechanism to decrease multidrug resistance 
P-glycoprotein was first described in multidrug resistant neoplastic cells, where its overexpression, induced by drug exposure, contributed to the resistance phenomenon [3,4]. Subsequently, constitutive expression of P-glycoprotein has been documented in a wide range of uninduced non-neoplastic cells [3,4]. In particular, this transporter is expressed in barrier cells, both epithelial and endothelial, where it is postulated to serve a protective function, eliminating toxic drugs or metabolic by-products into bile, urine or the intestinal lumen [3,11]. For example, P-glycoprotein has been detected in lens , ciliary epithelium , cornea , brain , intestine [8,9] and kidney [6,7]. The RPE contributes to the blood retinal permeability barrier, regulating ionic and solute composition in the subretinal space . P-glycoprotein activity in the RPE could serve to clear unwanted metabolites from the subretinal space, thus serving a protective function for the neural retina. This model would require basolateral expression of the transporter. Amphipathic compounds (the natural substrates for P-glycoprotein transport, ) would passively diffuse from subretinal space into the RPE cell, across its apical membrane. P-glycoprotein-mediated transport would export the compounds across (or out of) the basolateral membrane, for washout into the choroidal vasculature. Consistent with this model, P-glycoprotein is present on the RPE basolateral membrane (Figure 1).
Notably, in the RPE, P-glycoprotein is localized to the apical, as well as the basolateral, membrane surface (Figure 1). As there is no evidence that P-glycoprotein binds and transports its substrates from the extracellular medium into the cell [3,4], it is unlikely that apical P-glycoprotein would assist in drug clearance from subretinal space. It thus seems possible that apically localized P-glycoprotein serves an additional function in the RPE. There is evidence that P-glycoprotein can mediate ATP efflux [13-15], function as a lipid translocase [18-20], modulate volume sensitive Cl- efflux [12,16], catalyze steroid secretion , and/or transport retinoids .
If apically-localized P-glycoprotein does mediate ATP efflux, it could be involved in autocrine/paracrine signaling since purinergic receptors are present on the RPE surface [53-55] as well as in Mueller cells . In this regard, apical release of ATP has been demonstrated in the RPE  though any contribution of P-glycoprotein to that release has yet to be documented. In addition, recent studies show that P-glycoprotein can translocate short-chain lipid analogs across the plasma membrane [18-20]. This finding raises the possibility that apically-localized P-glycoprotein could serve a paracrine function by delivering bioactive lipids into subretinal space.
P-glycoprotein in bovine ciliary epithelium modulates a volume-sensitive chloride current . Human RPE cells catalyze regulatory volume decrease with increased Cl- transport observed after hypotonic challenge . It is possible, as described in the ciliary epithelium by Wu et al. , that P-glycoprotein in human RPE could serve to modulate volume-sensitive chloride current and hence be involved in cell volume regulation.
Finally, apically localized P-glycoprotein could serve to deliver lipophilic substrates to subretinal space for uptake by photoreceptors. For example, P-glycoprotein has been shown to transport steroids [17,58] and has also been implicated in transmembrane retinoid transport . Additional work is required to establish the ongoing function(s) of P-glycoprotein in the RPE.
This work was supported by NIH grant R01 EY11308. The deconvolution microscope facility used for this study was supported by NIH core grant P01 EY01492 and unrestricted funds from Research to Prevent Blindness, Inc. Parts of this work were presented at the 2001 Annual Meeting of the Biophysical Society, Boston, MA and the 2002 Annual Meeting of the Association for Research in Vision and Ophthalmology, Ft. Lauderdale, FL.
1. Hughes BA, Gallemore RP, Miller SS. Transport mechanisms in the retinal pigment epithelium. In: The retinal pigment epithelium: function and disease. Marmor MF, Wolfensberger TJ, editors. New York: Oxford University Press; 1998. p. 103-34.
2. Dean M, Allikmets R. Complete characterization of the human ABC gene family. J Bioenerg Biomembr 2001; 33:475-9.
3. Blackmore CG, McNaughton PA, van Veen HW. Multidrug transporters in prokaryotic and eukaryotic cells: physiological functions and transport mechanisms. Mol Membr Biol 2001; 18:97-103.
4. Fardel O, Lecureur V, Guillouzo A. The P-glycoprotein multidrug transporter. Gen Pharmacol 1996; 27:1283-91.
5. Gottesman MM, Pastan I, Ambudkar SV. P-glycoprotein and multidrug resistance. Curr Opin Genet Dev 1996; 6:610-7.
6. Ernest S, Bello-Reuss E. Expression and function of P-glycoprotein in a mouse kidney cell line. Am J Physiol 1995; 269:C323-33.
7. Dudley AJ, Brown CD. Mediation of cimetidine secretion by P-glycoprotein and a novel H(+)-coupled mechanism in cultured renal epithelial monolayers of LLC-PK1 cells. Br J Pharmacol 1996; 117:1139-44.
8. Terao T, Hisanaga E, Sai Y, Tamai I, Tsuji A. Active secretion of drugs from the small intestinal epithelium in rats by P-glycoprotein functioning as an absorption barrier. J Pharm Pharmacol 1996; 48:1083-9.
9. Sparreboom A, van Asperen J, Mayer U, Schinkel AH, Smit JW, Meijer DK, Borst P, Nooijen WJ, Beijnen JH, van Tellingen O. Limited oral bioavailability and active epithelial excretion of paclitaxel (Taxol) caused by P-glycoprotein in the intestine. Proc Natl Acad Sci U S A 1997; 94:2031-5.
10. Rao VV, Dahlheimer JL, Bardgett ME, Snyder AZ, Finch RA, Sartorelli AC, Piwnica-Worms D. Choroid plexus epithelial expression of MDR1 P glycoprotein and multidrug resistance-associated protein contribute to the blood-cerebrospinal-fluid drug-permeability barrier. Proc Natl Acad Sci U S A 1999; 96:3900-5.
11. Kawazu K, Yamada K, Nakamura M, Ota A. Characterization of cyclosporin A transport in cultured rabbit corneal epithelial cells: P-glycoprotein transport activity and binding to cyclophilin. Invest Ophthalmol Vis Sci 1999; 40:1738-44.
12. Wu J, Zhang JJ, Koppel H, Jacob TJ. P-glycoprotein regulates a volume-activated chloride current in bovine non-pigmented ciliary epithelial cells. J Physiol 1996; 491:743-55.
13. Abraham EH, Prat AG, Gerweck L, Seneveratne T, Arceci RJ, Kramer R, Guidotti G, Cantiello HF. The multidrug resistance (mdr1) gene product functions as an ATP channel. Proc Natl Acad Sci U S A 1993; 90:312-6.
14. Roman RM, Lomri N, Braunstein G, Feranchak AP, Simeoni LA, Davison AK, Mechetner E, Schwiebert EM, Fitz JG. Evidence for multidrug resistance-1 P-glycoprotein-dependent regulation of cellular ATP permeability. J Membr Biol 2001; 183:165-73.
15. Roman RM, Wang Y, Lidofsky SD, Feranchak AP, Lomri N, Scharschmidt BF, Fitz JG. Hepatocellular ATP-binding cassette protein expression enhances ATP release and autocrine regulation of cell volume. J Biol Chem 1997; 272:21970-6.
16. Idriss HT, Hannun YA, Boulpaep E, Basavappa S. Regulation of volume-activated chloride channels by P-glycoprotein: phosphorylation has the final say! J Physiol 2000; 524:629-36.
17. Bello-Reuss E, Ernest S, Holland OB, Hellmich MR. Role of multidrug resistance P-glycoprotein in the secretion of aldosterone by human adrenal NCI-H295 cells. Am J Physiol Cell Physiol 2000; 278:C1256-65.
18. Borst P, Zelcer N, van Helvoort A. ABC transporters in lipid transport. Biochim Biophys Acta 2000; 1486:128-44.
19. van Helvoort A, Smith AJ, Sprong H, Fritzsche I, Schinkel AH, Borst P, van Meer G. MDR1 P-glycoprotein is a lipid translocase of broad specificity, while MDR3 P-glycoprotein specifically translocates phosphatidylcholine. Cell 1996; 87:507-17.
20. van Helvoort A, Giudici ML, Thielemans M, van Meer G. Transport of sphingomyelin to the cell surface is inhibited by brefeldin A and in mitosis, where C6-NBD-sphingomyelin is translocated across the plasma membrane by a multidrug transporter activity. J Cell Sci 1997; 110:75-83.
21. Wiedemann P, Leinung C, Hilgers RD, Heimann K. Daunomycin and silicone oil for the treatment of proliferative vitreoretinopathy. Graefes Arch Clin Exp Ophthalmol 1991; 229:150-2.
22. Esser P, Tervooren D, Heimann K, Kociok N, Bartz-Schmidt KU, Walter P, Weller M. Intravitreal daunomycin induces multidrug resistance in proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci 1998; 39:164-70.
23. Kennedy BG, Mangini NJ. Plasma membrane calcium-ATPase in cultured human retinal pigment epithelium. Exp Eye Res 1996; 63:547-56.
24. Albert DM, Tso MO, Rabson AS. In vitro growth of pure cultures of retinal pigment epithelium. Arch Ophthalmol 1972; 88:63-9.
25. Reising CA, Kennedy BG, Getz RK, Mangini NJ. Immunodetection of an arrestin-like protein in human retinal pigment epithelium. Curr Eye Res 1996; 15:9-15.
26. Mangini NJ, Haugh-Scheidt L, Valle JE, Cragoe EJ Jr, Ripps H, Kennedy BG. Sodium-calcium exchanger in cultured human retinal pigment epithelium. Exp Eye Res 1997; 65:821-34.
27. Getz RK, Kennedy BG, Mangini NJ. Transthyretin localization in cultured and native human retinal pigment epithelium. Exp Eye Res 1999; 68:629-36.
28. Kennedy BG. Volume regulation in cultured cells derived from human retinal pigment epithelium. Am J Physiol 1994; 266:C676-83.
29. Kennedy BG. Na(+)-K(4)-Cl- cotransport in cultured cells derived from human retinal pigment epithelium. Am J Physiol 1990; 259:C29-34.
30. Kennedy BG, Haley BE, Mangini NJ. Creatine kinase in human retinal pigment epithelium. Exp Eye Res 2000; 70:183-90.
31. Davis AA, Bernstein PS, Bok D, Turner J, Nachtigal M, Hunt RC. A human retinal pigment epithelial cell line that retains epithelial characteristics after prolonged culture. Invest Ophthalmol Vis Sci 1995; 36:955-64.
32. Laemmli UK, Quittner SF. Maturation of the head of bacteriophage T4. IV. The proteins of the core of the tubular polyheads and in vitro cleavage of the head proteins. Virology 1974; 62:483-99.
33. Scheper RJ, Bulte JW, Brakkee JG, Quak JJ, van der Schoot E, Balm AJ, Meijer CJ, Broxterman HJ, Kuiper CM, Lankelma J, et al. Monoclonal antibody JSB-1 detects a highly conserved epitope on the P-glycoprotein associated with multi-drug-resistance. Int J Cancer 1988; 42:389-94.
34. Kobayashi H, Takemura Y, Kawai Y, Miyachi H, Kawabata M, Matsumura T, Yamashita T, Mori S, Furihata K, Shimodaira S, Motoyoshi K, Hotta T, Sekiguchi S, Ando Y, Watanabe K. Competitive reverse transcription-polymerase chain reaction assay for quantification of human multidrug resistance 1 (MDR1) gene expression in fresh leukemic cells. J Lab Clin Med 2000; 135:199-209.
35. Chen CJ, Clark D, Ueda K, Pastan I, Gottesman MM, Roninson IB. Genomic organization of the human multidrug resistance (MDR1) gene and origin of P-glycoproteins. J Biol Chem 1990; 265:506-14.
36. Arceci RJ, Stieglitz K, Bras J, Schinkel A, Baas F, Croop J. Monoclonal antibody to an external epitope of the human mdr1 P-glycoprotein. Cancer Res 1993; 53:310-7.
37. Saha P, Yang JJ, Lee VH. Existence of a p-glycoprotein drug efflux pump in cultured rabbit conjunctival epithelial cells. Invest Ophthalmol Vis Sci 1998; 39:1221-6.
38. Schinkel AH, Arceci RJ, Smit JJ, Wagenaar E, Baas F, Dolle M, Tsuruo T, Mechetner EB, Roninson IB, Borst P. Binding properties of monoclonal antibodies recognizing external epitopes of the human MDR1 P-glycoprotein. Int J Cancer 1993; 55:478-84.
39. Mechetner EB, Roninson IB. Efficient inhibition of P-glycoprotein-mediated multidrug resistance with a monoclonal antibody. Proc Natl Acad Sci U S A 1992; 89:5824-8.
40. Guntern R, Vallet PG, Bouras C, Constantinidis J. An improved immunohistostaining procedure for peptides in human brain. Experientia 1989; 45:159-61.
41. Weng TX, Godley BF, Jin GF, Mangini NJ, Kennedy BG, Yu AS, Wills NK. Oxidant and antioxidant modulation of chloride channels expressed in human retinal pigment epithelium. Am J Physiol Cell Physiol 2002; 283:C839-49.
42. Sharom FJ, Interaction of the P-glycoprotein multidrug transporter (MDR1) with high affinity peptide chemosensitizers in isolated membranes, reconstituted systems, and intact cells. Biochem Pharmacol 1999; 58:571-86.
43. Greiner B, Eichelbaum M, Fritz P, Kreichgauer HP, von Richter O, Zundler J, Kroemer HK. The role of intestinal P-glycoprotein in the interaction of digoxin and rifampin. J Clin Invest 1999; 104:147-53.
44. Demeule M, Shedid D, Beaulieu E, Del Maestro RF, Moghrabi A, Ghosn PB, Moumdjian R, Berthelet F, Beliveau R. Expression of multidrug-resistance P-glycoprotein (MDR1) in human brain tumors. Int J Cancer 2001; 93:62-6.
45. Hamilton KO, Backstrom G, Yazdanian MA, Audus KL. P-glycoprotein efflux pump expression and activity in Calu-3 cells. J Pharm Sci 2001; 90:647-58.
46. Axiotis CA, Bani D, Bianchi S, Pioli P, Tanini A, Brandi ML. P-glycoprotein is expressed in parathyroid epithelium and is regulated by calcium. Calcif Tissue Int 1995; 56:170-4.
47. Demeule M, Jodoin J, Gingras D, Beliveau R. P-glycoprotein is localized in caveolae in resistant cells and in brain capillaries. FEBS Lett 2000; 466:219-24.
48. Mi Q, Cui B, Silva GL, Lantvit D, Lim E, Chai H, You M, Hollingshead MG, Mayo JG, Kinghorn AD, Pezzuto JM. Pervilleine A, a novel tropane alkaloid that reverses the multidrug-resistance phenotype. Cancer Res 2001; 61:4030-7.
49. Meschini S, Calcabrini A, Monti E, Del Bufalo D, Stringaro A, Dolfini E, Arancia G. Intracellular P-glycoprotein expression is associated with the intrinsic multidrug resistance phenotype in human colon adenocarcinoma cells. Int J Cancer 2000; 87:615-28.
50. Loo TW, Clarke DM. Blockage of drug resistance in vitro by disulfiram, a drug used to treat alcoholism. J Natl Cancer Inst 2000; 92:898-902.
51. Merriman-Smith BR, Young MA, Jacobs MD, Kistler J, Donaldson PJ. Molecular identification of P-glycoprotein: a role in lens circulation? Invest Ophthalmol Vis Sci 2002; 43:3008-15.
52. Kizaki M, Ueno H, Yamazoe Y, Shimada M, Takayama N, Muto A, Matsushita H, Nakajima H, Morikawa M, Koeffler HP, Ikeda Y. Mechanisms of retinoid resistance in leukemic cells: possible role of cytochrome P450 and P-glycoprotein. Blood 1996; 87:725-33.
53. Sullivan DM, Erb L, Anglade E, Weisman GA, Turner JT, Csaky KG. Identification and characterization of P2Y2 nucleotide receptors in human retinal pigment epithelial cells. J Neurosci Res 1997; 49:43-52.
54. Peterson WM, Meggyesy C, Yu K, Miller SS. Extracellular ATP activates calcium signaling, ion, and fluid transport in retinal pigment epithelium. J Neurosci 1997; 17:2324-37.
55. Ryan JS, Baldridge WH, Kelly ME. Purinergic regulation of cation conductances and intracellular Ca2+ in cultured rat retinal pigment epithelial cells. J Physiol 1999; 520:745-59.
56. Liu Y, Wakakura M. P1-/P2-purinergic receptors on cultured rabbit retinal Muller cells. Jpn J Ophthalmol 1998; 42:33-40.
57. Mitchell CH. Release of ATP by a human retinal pigment epithelial cell line: potential for autocrine stimulation through subretinal space. J Physiol 2001; 534:193-202.
58. Ueda K, Okamura N, Hirai M, Tanigawara Y, Saeki T, Kioka N, Komano T, Hori R. Human P-glycoprotein transports cortisol, aldosterone, and dexamethasone, but not progesterone. J Biol Chem 1992; 267:24248-52.