Characterization of Brimonidine Transport in Retinal Pigment Epithelium

Abstract

purpose. To investigate the involvement of carrier-mediated transport mechanisms in brimonidine transport in retinal pigment epithelium (RPE).

methods. The transport of [³H]-brimonidine in bovine RPE-choroid explants was evaluated in a modified Ussing chamber. The uptake of [³H]brimonidine was evaluated in differentiated ARPE-19 cells cultured on permeable transwell filters.

results. The transport of brimonidine into (choroid-to-retina transport [inward]) and out of (retina-to-choroid transport [outward]) the eye in bovine RPE-choroid explants was temperature dependent. Both inward and outward brimonidine transport decreased at 5 μM compared with 10 nM. The melanin pigmentation of RPE did not significantly affect tissue permeability at either brimonidine dose. A saturable component was identified for the inward transport with the apparent Michaelis-Menten constant and a maximum transport rate of 51 μM and 148 pmol/(cm²·h), respectively. Both apical (representing retina-to-choroid transport) and basolateral (representing choroid-to-retina transport) brimonidine uptake in ARPE-19 cells showed temperature dependence. Apical uptake was higher than basolateral uptake at 37°C and was decreased to 70% in the presence of NaN₃ or in the absence of extracellular Na⁺. Besides α2-agonists, apical uptake was inhibited by verapamil, desipramine, and quinidine, but not by MPP⁺ (1-methyl-4-phenylpyridinium), TEA (tetraethylammonium), decynium-22, carnitine, PHA (p-aminohippurate), alanine, or inosine. Basolateral brimonidine uptake increased by 35% at extracellular pH of 6 and decreased by 50% under cell-depolarized conditions of high medium K⁺ and 1 μM valinomycin. Temperature-dependent components of basolateral uptake were not saturated at doses up to 2 mM.

conclusions. A carrier-mediated transport process for brimonidine in RPE was demonstrated in bovine RPE-choroid explants and polarized ARPE-19 cells. This transport system may play a significant role in modulating the movement of brimonidine into and out of the eye.

The retinal pigment epithelium (RPE), constituting the outer blood-retinal barrier, is a monolayer of cells resting on Bruch’s membrane and separating the neural retina from the choroidal blood supply. Like all epithelia, RPE cells exhibit polarity. The apical side of RPE faces the subretinal space and exhibits intercellular tight junctions that retard paracellular diffusion. The ability of RPE to regulate the vectorial transport of endogenous and exogenous compounds, metabolites, ions, and fluid relies on an asymmetric distribution of proteins. Studies in the past several years support RPE as a regulatory barrier by showing the expression of several solute transporters in the RPE, including glucose transporter-1 (GLUT1), ¹ P-glycoprotein (P-gp), ² multidrug resistance-associated protein-1 (MRP1), ³ organic cation transporter-3 (OCT-3), ⁴ monocarboxylate transporters (MCTs), ⁵ and organic anion transporter-2. ⁶  

Most cases of irreversible blindness result from diseases affecting the posterior region of the eye. Such diseases include age-related macular degeneration, diabetic retinopathy, glaucoma-associated retinopathy, and retinitis pigmentosa. ⁷ Development of nonsurgical local administration methods to deliver drugs to the posterior region of the eye to avoid systemic side effects or complications of surgery is thus of great value. Knowledge about routes taken by drugs to get to the retina and vitreous after local delivery and about how to improve drug absorption across ocular barriers is crucial. 

Brimonidine is an α2-adrenergic agonist approved for the treatment of open-angle glaucoma (Fig. 1) . In addition to the effect of intraocular pressure reduction, brimonidine was shown to have a neuroprotective function by promoting retinal ganglion cell survival. ⁸ In a recent study, it was reported that topical brimonidine reduces collateral damage caused by laser photocoagulation for choroidal neovascularization. ⁹ Because of structural and metabolic barriers, it is difficult to achieve therapeutic levels of topically applied drugs at the back of the eye, especially in the retina and the vitreous. However, brimonidine demonstrated good ocular distribution after topical application in pharmacokinetic studies using monkeys and rabbits. Significant concentrations of brimonidine were found in the posterior tissues of the eye. ¹⁰ Interestingly, after topical application of a single eye drop to rats, [³H]-brimonidine was found in optic nerves and tracts and in the corpus callosum of the brain within 5 minutes but in extremely low levels in blood (Abdulrazik M, et al. IOVS 2003;44:ARVO E-Abstract 4271). The mechanism of the rapid absorption remains to be explored. 

Drug absorption can be greatly enhanced by targeting endogenously expressed transport systems. ¹¹ It was shown that brimonidine was able to inhibit the passage of guanidine, a model substrate of the organic cation transporter (OCT), through rabbit conjunctival tissue by 70% at a concentration of 0.1 mM, ¹² indicating a possible role played by the OCT in ocular brimonidine absorption. In the present study, using bovine RPE-choroid explants mounted in a Ussing chamber, we investigated whether a carrier-mediated transport mechanism contributed to the transport of brimonidine in RPE. The brimonidine transport was further characterized in polarized monolayers of ARPE-19 cells, a human RPE cell line, cultured on transwell filters. 

Materials and Methods

Materials

Bovine eyes were obtained from a local slaughterhouse (Manning Beef, Pico-Rivera, CA) shortly after the animals were killed for food processing. 

ARPE-19 cells (passage 20) were obtained from American Type Culture Collection (Manassas, VA) and used between passages 22 and 25. Tissue culture–treated polyester transwells (6.5 mm or 12 mm in outer diameter and 0.4 μm in pore size) were purchased from Costar (Corning, NY). Dulbecco’s modified Eagle’s medium (DMEM) and Ham’s F-12 medium were purchased from Mediatech (Herndon, VA). Mouse laminin was from BD Biosciences (Bedford, MA). Other cell culture reagents and supplies were obtained from Life Technologies (Grand Island, NY). 

Brimonidine tartrate was a generous gift from Allergan, Inc. (Irvine, CA). [³H]Brimonidine (UK-14,304, [imidazolyl-4,5-³H]-) (81.2 Ci/mmol) was purchased from PerkinElmer Life Sciences (Boston, MA). The other chemicals of analytical purity were from Sigma. 

Preparation of Bovine RPE-Choroid Explants

Bovine eyes were kept in an ice bucket during transfer. Typically, 3 to 4 hours elapsed before the eyes were studied in the laboratory. 

After a brief cleaning of the connective tissues and the muscles, the eyecup was opened by a circular cut just posterior to the limbus. The anterior segment and the vitreous were removed. The posterior eyecup with the sclera, choroid, RPE, and retina was cut into approximately 1-inch squares. Unlike the human RPE, the bovine RPE contains nonpigmented areas over the tapetum and tapetum-free melanotic areas. Except in the studies to compare drug transport in pigmented RPE with that in nonpigmented RPE, only the pieces with pigmented RPE, where the choroid layer was thinner, were selected. The sclera was removed by a combination of sharp and blunt dissection under a binocular microscope using microscissors and delicate forceps; only a small amount of choroid was included. The retina-RPE-choroid preparation was then placed on filter paper (Whatman #1; VWR International, West Chester, PA; particle retention, 11 μM), with the retinal side up. The retina was then carefully peeled, and the remaining RPE-choroid was mounted in a clamping chip and sealed between two chamber halves of a custom designed Ussing chamber with an exposed area (A) of 1.0 cm². ¹³  

The potential difference (PD) and the transepithelial electrical resistance (TEER) of the tissues were measured using a voltage-clamp device (558C-5; Bioengineering Department, University of Iowa, Iowa City, IA). In addition to visual examination, the integrity of the RPE was examined by PD and TEER measurements. After 30-minute equilibration at 37°C, the tissues maintained PD of 6.5 ± 0.6 mV (retinal side positive) and TEER of 234 ± 34 Ω·cm² (n = 22), which were comparable to reported values. ^{14 15} PD and TEER were monitored for the entire duration of the transport experiment. 

Transport Study in Bovine RPE-Choroid Explants

The Ussing chamber held 6 mL bicarbonated Ringer solution (BRS; 119 mM NaCl, 3.6 mM KCl, 10 mM glucose, 0.5 mM Na₂Hpo ₄, 2.5 mM MgSO₄, 1.2 mM CaCl₂, and 23 mM NaHco ₃). The solution was gassed with a 5% CO₂–95% air mixture to yield a pH of 7.4 on both sides of the tissue and was water-jacketed to maintain the temperature within the chamber at 4°C or 37°C. 

Choroidal-to-retinal (transport into the eye, inward) or retinal-to-choroidal (transport out of the eye, outward) flux measurements were initiated by adding a specified amount of brimonidine traced with [³H]brimonidine to the choroidal or the retinal fluid. At predetermined time periods up to 3 hours, a 0.6-mL aliquot was collected from the receiver fluid for assay of radioactivity in a liquid scintillation counter (Beckman, Fullerton, CA). Each aliquot removed was immediately replaced with an equal volume of fresh BRS buffer. 

ARPE-19 Cell Monolayers on Permeable Transwell Filters

Transwell filters (0.4 μm pore size, 6.5 mm or 12 mm diameter; Costar) were coated with laminin. ARPE-19 cells were seeded at a seeding density of 1.66 × 10⁵ cells/cm² in DMEM/F-12 culture medium, supplemented with 100 U/mL penicillin-streptomycin, 2 mM l-glutamine, and 1% FBS. A corresponding amount of culture medium was added to the basolateral compartment, leveling the height of the liquid to prevent hydrostatic pressure. The medium was changed twice a week. Cells were cultured for at least a month to form differentiated monolayers, with the apical domain corresponding to the retinal facing side of the RPE monolayer and the basolateral domain corresponding to the choroidal facing side of the RPE monolayer. ^{5 16 17} TEER was measured once a week using a meter (EVOM Epithelial Voltohmmeter; World Precision Instruments, Sarasota, FL). The differentiation of ARPE-19 cells was also precharacterized by staining for Na⁺-K⁺-ATPase and the tight junction protein, zonules occludin-1. 

Uptake Studies by ARPE-19 Cell Monolayers

Unless otherwise specified, all uptake experiments were performed in a humidified atmosphere of 5% CO₂ and 95% air at 37°C in buffered BRS solution at pH 7.4. After the cell layers were preincubated in BRS for 40 minutes, the uptake experiments were initiated by spiking the apical (corresponding to the retina-facing surface and representing transport from the retina to the choroid) or basolateral (corresponding to the choroid-facing surface and representing transport from the choroid to the retina) fluid with a predetermined amount of brimonidine or competitive compounds traced with [³H]brimonidine. After a specified period, uptake was terminated by aspiration of the dosing solution followed by three quick washes of the cell layers in ice-cold BRS buffer. The cell layers were then solubilized in 0.5 mL 1% SDS solution. Five milliliters scintillation cocktail was added for quantification of the radioactivity in a liquid scintillation counter (Beckman, Fullerton, CA). 

Estimation of Kinetic Parameters

Unidirectional flux (J) for brimonidine was estimated from the steady state slope of the cumulative amount appearing in the receiver fluid over time. The apparent permeability coefficient (P_app) was calculated by normalizing the flux against nominal surface area (1.0 cm²) and the initial solute concentration. The P_app for the tissue explants (P_{app, tissue}) was calculated from the P_app for the tissue explants mounted on filter paper (P_app) and the P_app of the filter paper only (P_{app, filter}) using: 1/P_{app, tissue} = 1/P_app – 1/P_{app, filter} (equation 1) . The activation energy (E _a) of transport was estimated from the P_app at 37°C (P_app,37) and 4°C (P_app,4) according to: P_app,37/P_app,₄ = exp {–E _a/R[1/(273 + 37) – 1/(273 + 4)] } (equation 2) , where R is the universal gas constant.  

\[1/\mathrm{P}_{\mathrm{app,\ tissue}}{=}1/\mathrm{P}_{\mathrm{app}}{-}1/\mathrm{P}_{\mathrm{app,\ filter}}\ {\ldots}\]

 

\[\mathrm{P}_{\mathrm{app},37}/\mathrm{P}_{\mathrm{app},4}{=}\mathrm{exp}{\{}{-}E_{a}/R{[}1/(273{+}37){-}1/(273{+}4){]}{\}}\ {\ldots}\]

 

\[V{=}V_{\mathrm{max}}{\times}S/(K_{\mathrm{m}}{+}S)\ {\ldots}\]

 

To estimate the maximum uptake rate (V _max) and the Michaelis-Menten constant (K _m), the uptake rate over various dosing concentrations (S) was fitted based on V = V _max × S/(K _m + S) (equation 3)by means of nonlinear least-squares regression analysis using computer software (KaleidaGraph 3.5; Synergy Software, Reading, PA). All data were plotted as means ± SD, and statistical analysis was performed with analysis of variance (ANOVA) and Student-Newman-Keuls multiple comparison test. The criterion of significant difference was defined as P < 0.05. 

Results

Characterization of Brimonidine Transport in Bovine RPE-Choroid Explants

Directionality and Pigment Effect.

The effect of RPE pigment on brimonidine transport was studied at two substrate concentrations, 10 nM and 5 μM, which represented the approximate concentration of brimonidine in vitreous and choroid, respectively, of cynomolgus monkeys after a single 0.5% topical application. ¹⁰ At 10 nM, the P_app of brimonidine in the retinal-to-choroidal direction (R-to-C, outward) in pigmented and non-pigmented RPE-choroid explants was higher, though statistically insignificant, than those in the choroidal-to-retinal (C-to-R, inward) direction. Both R-to-C and C-to-R brimonidine transport decreased to a similar extent at 5 μM. RPE pigmentation did not significantly change the tissue permeability for brimonidine at either concentration (Fig. 2) . 

Temperature Dependence.

The transport of brimonidine across bovine RPE-choroid explants mounted in Ussing chamber in the C-to-R direction and the R-to-C direction has a profile of pseudo-steady state characteristics with no evident lag time. The transport of 10 nM brimonidine showed temperature dependence (Fig. 3) . When the experimental temperature was lowered to 4°C, the P_app for 10 nM brimonidine decreased to 20.6% (Fig. 3A)and 58.4% (Fig. 3B)of the P_app at 37°C in the R-to-C and the C-to-R directions, respectively. The activation energy (E _a) of brimonidine transport was calculated to be 8.2 kcal/mol and 2.8 kcal/mol in the R-to-C direction and the C-to-R direction, respectively. 

Concentration Dependence.

As shown in Figure 4 , the initial rate of C-to-R brimonidine flux across bovine RPE-choroid explants followed a saturable function of substrate concentration over the range from 1 to 100 μM, with the apparent Michaelis-Menten constant and the maximal flux rate estimated to be 51 ± 17 μM and 148 ± 22 pmol/(cm²·h), respectively. 

Characterization of Brimonidine Transport in ARPE-19 Cells Cultured on Transwell Filters

Directionality and Temperature Dependence.

Apical uptake (representing R-to-C transport) of 20 nM [³H]brimonidine (1.92 μCi/mL) in ARPE-19 cells was 50% higher than basolateral uptake (representing C-to-R transport) in 5 minutes (Fig. 5) . The difference was abolished when uptake temperature was lowered to 4°C. Approximately 30% and 50% brimonidine uptake from the apical and basolateral sides, respectively, was contributed by passive diffusion or surface binding (4°C). 

Time Course.

Apical (representing R-to-C transport) and basolateral uptake (representing C-to-R transport) of 20 nM [³H]brimonidine (1.92 μCi/mL) was determined in ARPE-19 cells cultured on permeable supports. As shown in Figure 6 , brimonidine uptake by ARPE-19 cells was rapid over the initial 5 minutes, with faster uptake from the apical side than from the basolateral side, and reached equilibration within 15 minutes. 

Energy Dependence.

To determine the possible driving force of the carrier-mediated transport mechanism, the apical (representing R-to-C transport) and basolateral (representing C-to-R transport) uptake of 20 nM brimonidine was investigated under conditions of low metabolic energy, lack of extracellular Na⁺, variation in membrane potential, and variation in extracellular proton concentration. As shown in Table 1 , apical and basolateral brimonidine uptake exhibited different profiles in response to these modulations. 

Apical uptake of brimonidine decreased by 30% when the metabolic energy of ARPE-19 cells was lowered by 30-minute pretreatment with 10 mM NaN₃, an inhibitor of oxidative phosphorylation. A decrease of equal magnitude was observed if extracellular Na⁺ was replaced by N-methyl-D-glucamine (NMDG) or by choline on the basis of iso-osmolarity. Apical uptake of brimonidine was sensitive to the changes in extracellular pH, with slower uptake in an acidic environment. However, it was not affected by the elimination of transmembrane proton gradient by 0.5 μM FCCP, a proton ionophore. Changes in membrane potential did not show a significant effect on apical brimonidine uptake in ARPE-19 cell monolayers. 

Basolateral uptake of brimonidine was not significantly decreased by the decreasing metabolic energy state through NaN₃ pretreatment (P = 0.23; Table 1 ). Rather, it decreased moderately (16%) when the extracellular Na⁺ was replaced by choline, but no decrease was observed when Na⁺ was substituted by NMDG. Basolateral brimonidine uptake decreased significantly (by 50%) when ARPE-19 cells were depolarized in a high concentration of extracellular potassium (143 mM) and 1 μM valinomycin. Decreasing the extracellular pH to 6 increased basolateral brimonidine uptake by 35%. Like apical uptake, basolateral uptake was not affected by the elimination of transmembrane proton gradient using 0.5 μM FCCP. 

Substrate Specificity.

The effect of various compounds added to the apical (representing R-to-C transport) dosing solution on brimonidine uptake is shown in Figure 7 . Uptake was carried for 3 minutes to maintain the “sink” condition. Uptake of the carrier-mediated process was estimated by subtracting uptake at 4°C in the presence of inhibitors from that at 37°C to eliminate surface binding. In addition to other α2-receptor agonists, the compounds tested were cationic or anionic compounds that are substrates or inhibitors of various known transport systems: the membrane potential–dependent organic cation transporters (TEA, MPP, verapamil, and decynium-22), the pH-dependent carnitine/organic cation transporters (carnitine), the adenosine triphosphate (ATP)–dependent organic anion transporters (PHA), the Na⁺-dependent amino acid transporters (alanine), the Na⁺-dependent nucleoside transporter (inosine), and the Na⁺-dependent neurotransmitter transporter (MPP⁺ and desipramine). Because approximately 50% of brimonidine exists as cation at physiologic pH, more cation inhibitors were tested in addition to anions and zwitterions. 

As shown in Figure 7 , apical brimonidine uptake was inhibited by α2-receptor agonists, including brimonidine (51% at 0.5 mM), clonidine (34% at 0.5 mM), and oxymetazoline (55% at 1 mM). It was not inhibited by TEA, MPP⁺, or decynium-22, substrates or inhibitors of the OCTs, but it was significantly inhibited by other cationic compounds such as verapamil (64% at 500 μM), quinidine (61% at 250 μM), and desipramine (85% at 1 mM). No significant inhibition of uptake was observed for other compounds tested. Basolateral uptake (representing C-to-R transport) of brimonidine decreased by a similar magnitude at 37°C and 4°C as substrate concentration increased from 0.02 μM to 2 mM (Fig. 8) . 

To further study the potency of various inhibitors, especially the α2-agonists, the inhibition pattern of apical brimonidine uptake over a range of concentrations (20 nM to 5 mM) of verapamil, clonidine, oxymetazoline, and unlabeled brimonidine was investigated, and the concentrations for 50% inhibition (K _i) were estimated to be 976 μM, 1427 μM, 808 μM, and 351 μM, respectively (Fig. 9) . 

Discussion

The blood-retinal barrier (BRB) consists of two major components, the endothelium of retinal blood vessels (inner BRB) and the retinal pigment epithelium (outer BRB). ¹⁸ The retinal blood vessels closely resembling the blood-brain barrier constitute the local blood circulation system for the inner retina, whereas the RPE controls penetration of nutrients, metabolic wastes, and xenobiotics from the choroid blood supply to the photoreceptor layer of retina that is completely avascular. ¹⁹ Although the inner and outer BRB cooperate to create a “privileged” environment in the retina and vitreous, the outer BRB constitutes the major absorption barrier for transscleral ²⁰ or topically ²¹ administrated drugs, especially when low concentrations of drug enter the bloodstream. 

In the present study, we have presented evidence for the existence of a carrier-mediated system facilitating the inward (C-to-R) transport of brimonidine across the outer BRB using bovine RPE-choroid explants. RPE-choroid explants mounted in two-compartment chambers are commonly used as a tissue-level model to study the permeability of the outer BRB. ^{22 23 24 25} In this model, most resistance to solute was from the RPE. The influence from choroid blood flow was excluded, and a contribution from the porous choroid in the explants was minimal. The integrity of the tissue was monitored by measuring the TEER and PD throughout the transport experiments. Brimonidine was found to penetrate the bovine RPE-choroid explants rapidly with no obvious lag time. An approximately twofold difference was observed between C-to-R brimonidine transport at 37°C and at 4°C (Fig. 3 , inset), indicating the existence of a carrier-mediated mechanism that was further confirmed by the identification of a saturable transport component with high affinity (K _m = 51 ± 17 μM) (Fig. 4) . Pharmacokinetic studies in monkeys ¹⁰ showed that after a single eye drop, brimonidine concentration in the sclera (eg, C _max = 6.67 to 8.19 μg-equivalents/g) was several times greater than that in the choroid/retina (eg, C _max = 1.84 μg-equivalents/g) and hundreds of times higher than that in the vitreous (eg, C _max = 0.011 μg-equivalents/g), indicating a net inward transport of brimonidine at the back of the eye across the RPE barrier to the therapeutic site under a concentration gradient. The contribution of active transport in carrier-mediated processes varies according to substrate concentration. If an intermediate concentration was used to roughly represent drug concentration in choroid, it can be estimated from our results (Fig. 4)that, at 5 to 10 μM, more than 50% of the inward brimonidine transport across the RPE barrier might be contributed by the carrier-mediated transport mechanism. A carrier-mediated outward transport process was also found in the RPE-choroid explants. The transport of 10 nM brimonidine in the R-to-C direction decreased by 80% when temperature was lowered from 37°C to 4°C (Fig. 3A) . Compared with the inward transport process accounting for brimonidine absorption on the micromolar concentrations, the outward transport can contribute to the removal of trace amounts of brimonidine from the subretinal fluid. 

Brimonidine is known to bind reversibly to ocular melanin with high affinity. ²⁶ In the present study we studied brimonidine transport in pigmented and nonpigmented bovine RPE-choroid explants at two concentrations, 10 nM (close to drug concentration in vitreous) and 5 μM (close to drug concentration in choroid). No difference in P_app was observed. Although binding with melanin affects the disposition and retention of brimonidine in RPE, our data indicated an insignificant effect of RPE melanin on in vitro brimonidine transport through the RPE. 

Appropriate RPE function relies highly on the maintenance of its polarity. In the past decade, the expression, distribution, and activity of membrane solute transporters or ion channels on the RPE have drawn increased attention. In addition to establishing the physiologic function of the endogenous transport systems in RPE, it is important to study how they affect the absorption, disposition, and elimination of drugs that are recognized as their substrates. In the present study, we also investigated whether the active brimonidine transport is mediated by established nutrient transporters in polarized ARPE-19 cells. ARPE-19, a commercially available human RPE cell line, forms polarized monolayers after prolonged culture (>4 weeks) on permeable support, ¹⁶ and it has been selected as a model of polarized RPE in studies of polarized transport, ²⁷ polarized protein expression/secretion, ^{5 17 28} and barrier breakdown. ²⁹ The apical side of the ARPE-19 cells represents the retina-facing domain of the RPE monolayer, whereas the basolateral side of the ARPE-19 cells represents the choroid-facing side of the monolayer. Consistent with tissue-level observations in bovine RPE-choroid explants, uptake studies in ARPE-19 cells also confirmed the existence of a carrier-mediated process for brimonidine transport in RPE. Furthermore, similar to the finding of a higher outward brimonidine transport compared with the inward transport in RPE-choroid explants, a carrier-mediated uptake process was primarily located on the apical side (representing R-to-C transport) of ARPE-19 cells. Support for this finding came from results on temperature dependence (Fig. 4) , directionality of uptake (Figs. 5 and 6) , energy dependence (Table 1) , substrate specificity (Fig. 7) , and concentration dependence (Fig. 9) . 

Studies of energy dependence indicated that the carrier-mediated brimonidine transport system might be composed of two processes, an active transport process that was driven by Na⁺ gradient and a facilitative transport process that was driven by substrate concentration gradient. Even though the acidic bathing medium lowered apical brimonidine uptake, the decrease was abolished by FCCP, a proton ionophore. Rather than its diminishing the transmembrane proton gradient, this decrease in uptake was more likely caused by an increased amount of ionized brimonidine, with a pKa value of 7.4, because the nonprotonated brimonidine penetrated the lipophilic cell membrane more readily through passive diffusion and the amount of nonprotonated brimonidine had significantly lower acidic pH. 

The existence of Na⁺-dependent and facilitative brimonidine transport processes on the apical side of RPE (representing R-to-C transport) may provide a valid explanation for the observation that outward brimonidine transport was more efficient than inward transport, although brimonidine transport in both directions showed carried-mediated characteristics. One feature of facilitative transporters, bidirectional movement, may be applicable for brimonidine. In the in vitro model of bovine RPE-choroid explants, transport in either direction could be enhanced based on the concentration gradient created after dosing. Under in vivo conditions, this concentration gradient across the RPE depends on various factors, including the modality of administration, the time period after administration, and the efficacy of drug elimination by choroidal blood flow. 

Basolateral brimonidine uptake (representing C-to-R transport) in ARPE-19 cells was approximately 30% higher when extracellular pH was lowered from 7.4 to 6. As a consequence, protonated/nonprotonated brimonidine was estimated to change from 1:1 to 25:1 by Henderson-Hasselbalch equation. In addition, at pH 7.4, brimonidine uptake decreased to approximately 50% when ARPE-19 cells were depolarized with high K⁺ medium containing valinomycin (Table 1) . These results showed that protonated brimonidine might be more readily transported than the noncharged form from the basolateral side of ARPE-19 cells. Basolateral brimonidine uptake showed temperature dependence (Fig. 5) . However, the temperature-dependent uptake (37°C-4°C) of brimonidine was not saturable at concentrations up to 2mM (Fig. 8) . Basolateral brimonidine uptake, taken together, might be carrier mediated but by a low-affinity process, which merits further investigation. 

The existence of α2-adrenergic receptors on the apical side of RPE cells was indicated by receptor stimulation studies. ³⁰ However, apical carrier-mediated transport was more likely accomplished by membrane transporter than by receptor-mediated endocytosis. This can be supported by several lines of evidence: (1) brimonidine was introduced at a concentration 10 times higher than its EC₅₀ (2 nM) to activate the α2-receptors, indicating the saturation of exposed binding site; (2) in addition to α2-adrenergic agonists, brimonidine transport was inhibited by verapamil with even greater potency (Fig. 9) ; (3) brimonidine uptake was sensitive to the existence of extracellular Na⁺; (4) although clonidine was a potent inducer of α2-adrenergic endocytosis, ³¹ it was a less potent inhibitor of brimonidine uptake (Fig. 9) . 

Brimonidine inhibited the inward active transport of guanidine, a model substrate of the organic cation transporters in rabbit conjunctiva. ¹² OCT-3, also recognized as the extraneuronal monoamine transporter, was found in the RPE ⁴ and was further identified primarily on the apical side by immunofluorescent staining and by monitoring MPP⁺ uptake. ³² Hence, we specifically explored the involvement of organic cation transporters in brimonidine uptake. Neither MPP⁺, TEA, decynium-22 (inhibitors of the organic cation transporters), nor carnitine (an endogenous substrate of the pH-dependent organic cation/carnitine transporter) inhibited apical brimonidine uptake at a concentration as high as 2.5 mM. These results indicated that brimonidine uptake in the RPE did not occur through the cloned organic cation transporters. Other tested cationic compounds, including verapamil, desipramine, and quinidine, inhibited brimonidine uptake significantly. Interestingly, unlike the organic cation transporters and the neurotransmitter transporters, ³³ which transport many primary amines and the permanently charged quaternary amines, the brimonidine transport system in ARPE-19 showed specificity to tertiary and heterocyclic amines with sp2-hybridized nitrogen. This inhibition profile resembled that of the novel cationic drug transporter of Han et al. ³⁴ functionally characterized in RPE cells, which recognized verapamil, quinidine, and several β-blockers but not TEA, MPP⁺, or carnitine. Substrates of several other transporters (organic anion transporter, amino acid transporter, and nucleoside transporter) showed no inhibition of brimonidine uptake. 

In conclusion, based on our observations in bovine RPE-choroid explants and polarized ARPE-19 cells, we demonstrated a carrier-mediated transport process for brimonidine in RPE. This transport system may play a significant role in modulating the movement of brimonidine into and out of the eye. 

 Supported by the Arnold and Mabel Beckman Foundation, National Institutes of Health Grants EY01545 and EY03040, and Research to Prevent Blindness (Doheny Eye Institute).

 Submitted for publication February 11, 2005; revised April 25, July 1, and August 15, 2005; accepted November 18, 2005.

 Disclosure: N. Zhang, Allergan, Inc. (F); R. Kannan, None; C.T. Okamoto, None; S.J. Ryan, Allergan, Inc. (C); V.H.L. Lee, None; D.R. Hinton, Allergan, Inc. (F)

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Chemical structure of brimonidine.

Figure 2.

 

Apparent permeability (P_app) of brimonidine in pigmented or nonpigmented bovine RPE-choroid explants in the retinal-to-choroidal (R-to-C) or the choroidal-to-retinal (C-to-R) direction. All experiments were conducted in the presence of tracer alone (0.81 μCi/mL (10 nM) [³H]brimonidine) (A) or of 5 μM unlabeled brimonidine (B). Data represent mean ± SD; n = 3.

Figure 3.

 

R-to-C (A) and C-to-R (B) brimonidine transport showed temperature dependence. [³H]brimonidine at 0.81 μCi/mL (10 nM) was dosed from the retinal or choroidal side of bovine RPE-choroid explants. Insets show the P_app of R-to-C (B) and C-to-R (A) transport of 10 nM brimonidine at 37°C and 4°C. Data represent mean ± SD; n = 3–4.

Figure 4.

 

Total C-to-R brimonidine flux as a function of brimonidine concentration. All experiments were conducted in the presence of 0.81 μCi/mL (10 nM) [³H]brimonidine and 1 to 100 μM unlabeled brimonidine. The linear component of brimonidine flux contributed by passive diffusion was measured at 4°C. Apparent Michaelis-Menten constant and maximal flux of the saturable component contributed by carrier-mediated transport were estimated to be 51 ± 17 μM and 148 ± 22 pmol/(cm²·h), respectively. Data represent mean ± SD; n = 3–4.

Figure 5.

 

Temperature-dependent and directional uptake of 20 nM [³H]brimonidine (1.92 μCi/mL) in ARPE-19 cell monolayers cultured on transwell filters (12 mm in diameter, 0.4 μm in pore size). Apical dosing represents the R-to-C direction, whereas basolateral dosing represents the C-to-R direction of transport. Data represent mean ± SEM; n = 3.

Figure 6.

 

Time-dependent uptake of 20 nM [³H]brimonidine in ARPE-19 cell monolayers cultured on transwell filters (6.5 mm in diameter, 0.4 μm in pore size). Apical dosing represents the R-to-C direction, whereas basolateral dosing represents the C-to-R direction of transport.

Table 1.

 

Energy-Dependent Brimonidine Uptake from the Apical Side or the Basolateral Side of ARPE-19 Cell Monolayers

Parameter	Conditions	Uptake (% of control)
				Apical	Basolateral
Metabolic energy	10 mM NaN3 30-minute pretreatment	70.7 ± 4.9*	86.7 ± 9.8
Extracellular Na+	Replaced by NMDG	64.9 ± 4.5*	102.0 ± 6.2
	Replaced by choline	72.3 ± 6.4*	84.2 ± 5.9*
Membrane potential	+ 1μM valinomycin	111.8 ± 8.0	77.8 ± 9.4
	[K+] = 143 mM	98.2 ± 6.0	76.5 ± 9.4
	+ 1 μM valinomycin ([K+] = 143 mM)	88.2 ± 12.9	49.1 ± 8.4*
Proton gradient	pH 6	67.1 ± 5.0*	134.6 ± 10.5*
	pH 8.5	109.3 ± 3.4*	104.4 ± 9.7
	+ 0.5 μM FCCP (pH 7.4)	97.5 ± 5.4	98.7 ± 4.3

 Experiments were carried out as described in Materials and Methods. Apical (R-to-C) or basolateral (C-to-R) uptake was measured at 37°C for 3 minutes. The amount of [³H]-brimonidine was 20 nM (1.39 μCi/mL). In the control experiments, BRS at pH 7.4 was used as uptake buffer. Carrier-mediated uptake was calculated by subtracting uptake at 4°C from uptake at 37°C. Data represent mean ± SEM; n = 6.

*  Significant differences from control values (P < 0.05).

Figure 7.

 

Inhibition profile of apical uptake (represents the R-to-C direction of transport) of brimonidine in ARPE-19 cell monolayers. The uptake of 1.62 μCi/mL (20 nM) [³H]brimonidine was measured in the presence or absence of various inhibitors. Inhibitors of characterized transport systems were dosed at concentrations commonly used to achieve sufficient inhibition in heterologous systems. ¹Dosed at 500 μM; ²dosed at 1 mM; ³dosed at 2.5 mM; ⁴dosed at 10 μM; ⁵dosed at 250 μM. Bars represent mean ± SD; n = 3. *Statistically significant difference (P < 0.05) from control value, without inhibitors.

Figure 8.

 

Effect of increasing concentrations of unlabeled substrate on basolateral brimonidine uptake (represents C-to-R direction of transport) at 37°C and 4°C. The experiment was carried out as described in Materials and Methods. Bars represent mean ± SD; n = 3.

Figure 9.

 

Dose-dependent inhibition of apical brimonidine (1.62 μCi/mL, 20 nM) uptake (represents R-to-C direction of transport) in ARPE-19 cells by verapamil (▪), unlabeled brimonidine (▴), clonidine (•), and oxymetazoline (▾). Concentrations for 50% inhibition (K_i) were estimated to be 351 μM, 808 μM, 976 μM, and 1427 μM for verapamil, brimonidine, clonidine, and oxymetazoline, respectively. Bars represent mean ± SD; n = 4.

The authors thank Diane Tang-Liu (Allergan, Inc., Irvine, CA) for providing brimonidine tartrate.