Induction of the Cystine/Glutamate Exchanger SLC7A11 in Retinal Pigment Epithelial Cells by the Antipsoriatic Drug Monomethylfumarate

Abstract

Purpose.: Oxidative stress is a common pathological factor in degenerative retinal diseases; therefore, identifying novel strategies for its limitation is critically important and highly relevant clinically. Along these lines, our present goal was to evaluate the effect(s) of the fumarate ester and antipsoriatic agent monomethylfumarate (MMF) on the expression and functional activity of the cystine/glutamate exchanger SLC7A11 (system x_c ⁻), a transport system critical to potentiation of antioxidant signaling in retina.

Methods.: ARPE-19 and primary mouse RPE cells were cultured in the presence or absence of varying concentrations of MMF (0–5000 μM) for 0 to 24 hours. MMF (10 mM) was also delivered intravitreally to mouse eyes. RT-PCR, radiolabeled uptake, Western blotting, and glutathione (GSH) assays were then used to evaluate the effects of MMF on endogenous antioxidant machinery.

Results.: MMF induced system x_c ⁻, Nrf2, and hypoxia-inducible factor 1α (Hif-1α) in cultured RPE cells. Additionally, the compound was recognized as a transportable substrate by the Na⁺-coupled monocarboxylate transporter SLC5A8 (SMCT1). In vivo these factors were evidenced by a significant increase in retinal levels of GSH.

Conclusions.: MMF stimulates multiple pathways in retinal cells that potentiate cellular events leading to the upregulation of genes/mechanisms that function to protect retina against various forms of insult; upregulation of system x_c ⁻ is one such consequence. To our knowledge, this is the first report that fumarate esters, compounds already employed clinically for other indications, are effective in retina via x_c ⁻ induction. This novel, hitherto unknown mechanism helps to explain the antioxidant feature of these compounds and highlights their therapeutic potential in retina.

Introduction

Fumaric acid esters (FAE) have been used for many years in the treatment of psoriasis, ¹ and more recently multiple sclerosis (MS). ² Though pathophysiologically two very different diseases, the critical involvement of oxidative stress in the pathogenesis/progression of each is a major commonality between the two and may explain, at least in part, the high efficacy of FAE in the treatment of these two diseases, given that the beneficial effects of FAE stem purportedly from their potent ability to enhance endogenous cellular antioxidant responses. ^3,4 Interestingly, however, little is known regarding the exact cellular/molecular mechanisms underlying these actions. 

Given the successes of FAE in the treatment of psoriasis and MS, most studies on their mode of action have focused on keratinocytes and immune cells, cell types critically involved in the respective diseases. However, recent studies report potential benefits in other cell types, ^5–7 leading to speculation that FAE therapy can be extrapolated to treatment of additional diseases, particularly those in which oxidant-induced cellular damage is majorly involved. This is directly relevant to retina, a tissue in which oxidative stress wreaks havoc. Indeed, oxidant-induced damage to retinal cells is a major causative factor in the pathogenesis of a number of degenerative retinal diseases including age-related macular degeneration and diabetic retinopathy, ^8,9 two of the leading causes of blindness worldwide. Hence, novel therapies for protecting retina from the detrimental effects of reactive oxygen species (ROS) are sorely needed. Therapeutic use of antioxidant supplements for this very purpose is being explored in many laboratories. Recent studies suggest, however, that the greatest promise may lie in those therapies capable of boosting the endogenous antioxidant machinery of retinal cells, either as an alternative to or in addition to exogenous antioxidant supplementation. ^10–13 Whether FAE may be of benefit in this regard has received little attention to date. 

In the present study, we address both of the aforementioned issues: the potential efficacy of FAE in retina and the underlying mechanisms responsible. Emphasis here is placed on retinal pigment epithelial (RPE) cells, a retinal cell type critically involved in glutathione production/redox regulation in retina, ^14,15 and on monomethylfumarate (MMF), the major bioactive component/metabolite in the principal FAE-based drugs that have been employed clinically (i.e., Fumaderm, BG-12). ^16,17 Specifically, we evaluate the effects of MMF on expression and activity of the cystine/glutamate exchanger, also known as system x_c ⁻, in human RPE (ARPE-19) and primary mouse RPE cells. System x_c ⁻ is a critical regulator of endogenous glutathione synthesis and hence a key determinant of the overall antioxidant capacity of the cell. ^18,19 The critical importance of this transport system to the maintenance of cellular redox status is emphasized further by the fact that it is upregulated readily in response to oxidative stress, providing protection against oxidant-induced cell death/damage. ^18,20 Compounds capable of increasing system x_c ⁻ activity therefore have extremely high potential for therapeutic use in retina. Others have reported previously on the differential effects of FAE on glutathione/glutathione-related enzymes and offered some speculation as to the mechanisms responsible. ^21,22 Interestingly, however, none have directly evaluated the involvement of system x_c ⁻, a transport system upon which endogenous synthesis of glutathione critically hinges. Given the critical role of system x_c ⁻ as a regulator of endogenous glutathione synthesis/cellular redox status, studies such as the present, aimed at understanding the effects of FAE on regulation of this transport system, are broadly relevant and have high clinical significance. 

Materials and Methods

Materials

[³H]-Glutamate was purchased from American Radiolabeled Chemicals (St. Louis, MO). Fumarate, MMF, dimethylfumarate (DMF), monoethylfumarate (MEF), diethylfumarate (DEF), pyruvate, and ibuprofen were all obtained from Sigma-Aldrich (St. Louis, MO). Glutathione (GSH) Glo Assay kit was from Promega (Madison, WI), and histone deacetylase (HDAC) activity assay kit was from BioVision (Mountain View, CA). SLC5A8 cDNA was originally cloned from human intestine. ²³ Rabbit polyclonal Nrf2 antibody was from AbCam (Cambridge, MA), and monoclonal Hif-1α antibody from BD Biosciences (Franklin Lakes, NJ). Preparation of anti-xCT antibody has been described previously. ²⁴  

Animals and Cell Culture

Slc5a8^−/− and Gpr109a^−/− mice have been described previously. ^25,26 Frogs were obtained from Xenopus-I (Ann Arbor, MI). The care and use of animals in these studies adhered to the Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication No. 85-23) and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and was approved by the Institutional Committee for Animal Use in Research and Education of the Georgia Health Sciences University. Human RPE cells (ARPE-19) were obtained from American Type Culture Collection (Manassas, VA) and maintained as described previously. ¹¹ Primary RPE (mRPE) and Müller cells were isolated from wild-type, Slc5a8^−/− , and Gpr109a^−/− mouse eyes and cultured in accordance with our previously described protocols. ^27,28  

Assay of System x_c ⁻ Transport Activity

Activity of the cystine/glutamate exchanger, system x_c ⁻, was assayed in ARPE-19, primary mouse RPE, and Müller cells by monitoring the cellular uptake of [³H]-glutamate under Na⁺-free conditions according to our previously published method. ²⁹ In brief, uptake was initiated by the addition of 250 μL uptake buffer (25 mM HEPES/Tris, 140 mM N-methyl-D-glucamine chloride, 5.4 mM KCl, 1.8 mM CaCl₂, 0.8 mM MgSO₄, and 5 mM glucose, pH 7.5) spiked with 2 μCi/mL [³H]-glutamate, in the presence or absence of various dosages of fumarate, DMF, MMF, or MEF. Cells were incubated for 15 minutes at 37°C before the buffer was removed, and the cells were washed twice with ice-cold uptake buffer. The cells were then solubilized with 0.5 mL 1% (w/v) SDS/0.2M NaOH, and radioactivity was determined. Protein was measured using the Bio-Rad (Hercules, CA) protein assay reagent. Noncarrier-mediated uptake (i.e., diffusional component) was determined by measuring the uptake of [³H]-glutamate in the presence of excess unlabeled glutamate (5 mM). x_c ⁻ transport activity was then calculated by subtracting the diffusional component from total uptake. Kinetic analysis of system x_c ⁻ was performed by measuring the transport activity over a wide range of glutamate concentrations (2.5–1000 μM) and by analyzing the data according to the Michaelis-Menten equation describing a single saturable transport system. K_t and V_max were calculated using the nonlinear regression method and then confirmed using the linear regression method. 

Transport of MMF via Human SLC5A8 in the Xenopus laevis Oocyte Heterologous Expression System

The preparation of capped cRNA from human SLC5A8 cDNA has been described. ²³ Mature oocytes from X. laevis were injected with 50 ng cRNA. Water-injected oocytes served as controls. The oocytes were used for electrophysiological studies 3 to 6 days after cRNA injection. Electrophysiological studies were performed by the two-microelectrode/voltage-clamp method. ^23,30 Oocytes were perifused with NaCl-containing buffer (100 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 1 mM CaCl₂, and 10 mM HEPES/Tris, pH 7.5) followed by the same buffer containing fumarate, DEF, DMF, MEF, or MMF. Pyruvate served as a positive control for SLC5A8 transport activity. ²³ The membrane potential was clamped at −50 mV, and the differences between the steady-state currents measured in the presence or absence of substrates were considered the substrate-induced currents. In the analysis of the saturation kinetics of MEF- or MMF-induced currents, the K_t was calculated by fitting the values of the MEF- or MMF-induced currents at different concentrations of these compounds to the Michaelis-Menten equation. The Na⁺-activation kinetics of MMF-induced currents were analyzed by measuring the MMF-specific currents in the presence of increasing concentrations of Na⁺, and the data were analyzed according to the Hill equation to determine the Hill coefficient (the number of Na⁺ ions involved in the activation process) and K_t for Na⁺. Because the expression levels varied significantly from oocyte to oocyte, data were normalized, by taking the maximally induced SLC5A8-specific current in each oocyte as 1. The kinetic parameters were determined using SigmaPlot software (version 10; Systat Software, Chicago, IL). The effect of ibuprofen, an inhibitor of SLC5A8, ³¹ on SLC5A8-mediated MMF transport was also studied. The dose–response relationship for ibuprofen-induced inhibition of MMF-induced currents in these oocytes was investigated by fitting the data to the Michaelis-Menten equation, which enabled us to determine the value for K_0.5 (i.e., the concentration of ibuprofen necessary to cause 50% maximal inhibition of MMF-induced currents). Each uptake experiment was repeated at least three times with separate oocytes, and the data are presented as mean ± SEM. 

Intravitreal Injection of MMF

Male Balb/c mice (4 weeks; n = 6) were used for intravitreal injection of MMF following our published protocol. ³² Animals were weighed and anesthetized using 17 μL (1 μL/g body weight) of a solution of ketamine (80 mg/mL) and xylazine (12 mg/mL). Then 5 μL of proparacaine solution (5% w/v) was administered topically to the eyes. MMF (1 μL; 10 mM solution prepared in PBS) was then injected into the vitreous body of the right eye of each animal at the limbus; the left eye served as a contralateral control and received an equal volume of PBS. At 24 hours postinjection, mice were sacrificed via CO₂ inhalation, and eyes were harvested. 

Glutathione Measurement in Cells and Tissues

Cellular glutathione levels were measured in ARPE-19 cells cultured in the presence or absence of 100 μM MMF for various lengths of time using Promega's GSH Glo Assay kit, following both the manufacturer's instructions and our previously published protocols. ^11,29 Levels of glutathione were also measured in ocular tissues obtained from the eyes of mice injected intravitreally with MMF as described above. Following removal of the lens, eyes were weighed, homogenized in PBS/2 mM EDTA (10 mg tissue/mL), and centrifuged at 14,000 rpm for 5 minutes at 4°C. Fifty microliters of the supernatant from each sample was then taken for measurement of glutathione using the GSH Glo Assay kit (Promega) per the manufacturer's instructions. Results were expressed as nmol/mg tissue. 

Western Blot Analysis of xCT, Nrf2, and Hif-1α Proteins in RPE Cells

ARPE-19 or primary RPE (wild type, Slc5a8^−/− , or Gpr109a^−/− ) cells were cultured in the presence or absence of MMF (100 μM) as described above. Cells were collected at various times postincubation; protein lysates were prepared and subjected to SDS-PAGE. Samples were then transferred to nitrocellulose membranes and incubated with antibody against hypoxia-inducible factor 1 α (Hif-1α), xCT, or Nrf2 overnight at 4°C, followed by incubation with horseradish peroxidase–conjugated goat antimouse or antirabbit IgG antibody. After washing, antibody-specific bands were detected using chemiluminescence (Enhanced ChemiLuminescence kit; PerkinElmer Life Sciences, Inc., Waltham, MA). β-Actin served as the loading control. 

Reverse Transcriptase (RT)-PCR

RNA was isolated from ARPE-19 or primary RPE cells cultured in the presence or absence of MMF and used for RT-PCR using a GeneAmp RT-PCR kit (Applied Biosystems, Grand Island, NY). The sequences of the forward and reverse primers used to analyze human and mouse xCT and 4F2hc mRNA expression have been reported. ²⁹ Details of the primer sequences and expected product sizes of all other primers utilized in this study are provided in the Table. 18S ribosomal RNA was used as an internal control. PCR was performed using Taq polymerase kit (TaKaRa Bio, Inc./Clontech Laboratories, Mountain View, CA) for 30 cycles with a denaturing phase of 1 minute at 94°C, an annealing phase of 1 minute at 60°C, and an extension phase of 1 minute at 72°C for all the primer sets. 

Table

 

. PCR Primers

Gene	Primer Sequences	Expected Product Size, bp
Human Hif1-α	Fwd: 5′-ACCTATGACCTGCTTGGTGCTGAT-3′	617
	Rev: 5′-CAGTTTCTGTGTCGTTGCTGCCAA-3′
Mouse Hif1-α	Fwd: 5′-AAGCCCTCCAAGTATGAGCACAGT-3′	294
	Rev: 5′-AGGCTCCTTGGATGAGCTTTGTCT-3′
Human Nrf2	Fwd: 5′-GGTTTCTTCGGCTACGTTT-3′	314
	Rev: 5′-ACTTCTTTTTCCATTGAGGGTATA-3′
Mouse Nrf2	Fwd: 5′-TAAAGCTTTCAACCCGAAGCACGC-3′	572
	Rev: 5′-TACAGTTCTGGGCGGCGACTTTAT-3′
Human HO-1	Fwd: 5′-ATTGCCAGTGCCACCAAGTTCAAG-3′	106
	Rev: 5′-ACGCAGTCTTGGCCTCTTCTATCA-3′

Data Analysis

Electrophysiologic measurements were repeated at least three times with separate oocytes. For studies with primary RPE or Müller cells, two independent cell preparations were made from Slc5a8^−/− , Gpr109a^−/− , and/or wild-type (Gpr109a^+/+ , Slc5a8^+/+ ) mouse eyes. All cell treatments were performed in duplicate and measurements were performed in triplicate. For all experiments, data are presented as means ± SEM. An unpaired Student's t-test was used to determine statistical significance. P < 0.05 was considered to be statistically significant. 

Results

MMF-Induced Upregulation of System x_c ⁻ mRNA, Protein, and Functional Activity

System x_c ⁻ is a heterodimer composed of two subunits: a regulatory subunit, known as 4F2hc, common to many amino acid transporters and important for membrane localization, and the catalytic subunit, known as xCT, responsible for transport activity. ^19,33 To assess the effects of MMF on system x_c ⁻ expression at the mRNA level, RT-PCR was carried out for xCT and 4F2hc using RNA obtained from ARPE-19 cells cultured in the presence or absence of 100 μM MMF for 16 hours (Fig. 1A). xCT mRNA expression was higher in cells cultured in the presence of MMF than in control untreated cells. 4F2hc levels remained unchanged. To ascertain whether there was a parallel increase in xCT protein, Western blotting was performed using protein obtained from additional cells that were cultured in a manner identical to that described above. The resulting data demonstrated that the MMF-induced upregulation of xCT expression is evident also at the protein level (Fig. 1B). We then compared the transport activity of system x_c ⁻ in ARPE-19 cells cultured in the presence or absence of MMF (100 μM, 16-hour incubation). Na⁺-independent [³H]-glutamate uptake, indicative of system x_c ⁻ activity, was 3- to 4-fold higher in ARPE-19 cells cultured in the presence of MMF (Fig. 1C). The substrate specificity of the induced transport system, as evident from the selective interaction with glutamate and cystine, confirmed that system x_c ⁻ was indeed responsible for the observed transport activity (Fig. 1D). 

Figure 1

 

Monomethylfumarate (MMF)-induced upregulation of system x_c ⁻ in ARPE-19 cells. (A) RT-PCR analysis of mRNA transcripts specific for xCT and 4F2hc in control (Con) and ARPE-19 cells cultured in the presence of MMF (100 μM) for 16 hours. (B) Western blot analysis of xCT protein in control and MMF-treated ARPE-19 cells. (C) Activity of system x_c ⁻ in control and MMF-treated ARPE-19 cells. Uptake of [³H]-glutamate (2.5 μM) was measured for 15 minutes in control and MMF (100 μM)-treated ARPE-19 cells at 37°C in the absence of Na⁺. The values represent transport activity specific for system x_c ⁻. *P < 0.001. (D) Substrate selectivity of glutamate uptake in control and MMF-treated ARPE-19 cells. Uptake of [³H]-glutamate (2.5 μM) was measured in control and MMF-treated cells in the absence of Na⁺ for 15 minutes at 37°C in the absence or presence of unlabeled amino acids glutamate, aspartate, cystine, cysteine, and leucine, each at a 1 mM concentration. *P < 0.0001. (E) Saturation kinetics of glutamate uptake in control and MMF (100 μM)-treated ARPE-19 cells were evaluated by monitoring the uptake of [³H]-glutamate in Na⁺-free medium with various concentrations of unlabeled glutamate. (F) Eadie-Hofstee transformation of the data. The experiment was repeated three times.

Kinetic Analysis of System x_c ⁻ in APRE-19 Cells Cultured in the Presence or Absence of MMF

The influence of MMF on the kinetics of system x_c ⁻ transport function was analyzed by monitoring the Na⁺-independent uptake of glutamate over a concentration range of 2.5 to 1000 μM. The uptake process was saturable in cells cultured in both the presence and absence of MMF (100 μM) (Figs. 1E, 1F). The increase in the transport activity of system x_c ⁻ observed in MMF-treated ARPE-19 cells compared to control cells was associated with an increase in the maximal velocity of the transport system, with no significant change in substrate affinity. The maximal velocity was ∼5-fold greater in MMF-treated ARPE-19 cells than in control cells (11.7 ± 0.4 compared to 2.2 ± 0.2 nmol/mg protein/15 min; P < 0.001). The values for Michaelis constant were comparable (139 ± 9 compared to 121 ± 28 μM; P > 0.05). 

Influence of Various Fumaric Acid Esters on System x_c ⁻ Activity

To determine whether stimulation of system x_c ⁻ activity is specific to MMF or observable also in the presence of fumarate or other fumaric acid esters, ARPE-19 cells were incubated with 100 μM MMF, DMF, MEF, or fumarate for 16 hours, followed by analysis of Na⁺-independent [³H]-glutamate uptake. Each of the FAE tested evoked marked stimulation of system x_c ⁻ activity (>4-fold stimulation). However, only ∼1.8-fold stimulation was observed in the presence of the parent compound fumarate (Fig. 2). 

Figure 2

 

Specificity of fumaric acid esters for stimulation of system x_c ⁻ activity. ARPE-19 cells were treated for 16 hours with fumarate and its esters: monomethylfumarate (MMF), dimethylfumarate (DMF), and monoethylfumarate (MEF). Cells cultured in the absence of any of these compounds served as controls (con). Uptake of [³H]-glutamate (2.5 μM) was measured in the absence of Na⁺ for 15 minutes at 37°C. Results are means ± SEM (n = 3). (*P < 0.001; **P < 0.05.)

MMF Is a Substrate for the Na⁺-Coupled Monocarboxylate Transporter SLC5A8 (SMCT1)

The induction of system x_c ⁻ expression and functional activity in RPE cells by FAE is a novel finding, and one that may help to explain the antioxidant/glutathione-stimulating actions that have been reported in association with the use of these compounds in various cell types. However, experimental evidence suggests strongly a pleiotropic mode of action for FAE. ² As such, it is plausible to speculate that the beneficial effects associated with FAE involve multiple mechanisms. We reported recently on the potential involvement of SLC5A8, a Na⁺-dependent transporter of monocarboxylates (i.e., lactate, pyruvate, butyrate), in the regulation of glutathione homeostasis in retina/RPE. ¹¹ We did not know whether MMF interacts also with this transport system and/or if the transporter plays any obligatory role in the observed effects of this compound on system x_c ⁻ activity. To address this issue, we sought first to determine whether MMF is a substrate for SLC5A8. SLC5A8 is an electrogenic transporter. ^23,30 Transport of monocarboxylate substrates via this transport system occurs with a Na⁺:substrate stoichiometry of 2:1. Hence, there is a net positive charge entering the cells during the cotransport of Na⁺ and the monocarboxylate substrate via the transporter. This causes depolarization of the cell membrane, a phenomenon that is detected as inward currents under voltage-clamp conditions. With this rationale, we expressed human SLC5A8 in X. laevis oocytes and monitored its transport function by electrophysiological means. Water-injected oocytes served as negative controls. Pyruvate, a known transportable substrate for the transport system, was used as a positive control. ²³ Exposure of human SLC5A8-expressing oocytes to 1 mM pyruvate induced marked inward currents in the presence of Na⁺ (Fig. 3A). Such currents were not detectable in SLC5A8-expressing oocytes in the absence of Na⁺, nor in water-injected oocytes (data not shown). We then examined whether MMF was recognized as a substrate by SLC5A8 by monitoring the inward currents in human SLC5A8-expressing oocytes upon exposure to this compound in the presence of Na⁺. At a concentration of 1 mM, MMF induced inward currents in SLC5A8-expressing oocytes. As with pyruvate, currents induced by MMF were obligatorily dependent upon the presence of Na⁺. Similar results were obtained with three different oocytes. These data show that MMF is indeed a transportable substrate for SLC5A8; and like transport of all other SLC5A8 substrates characterized to date, SLC5A8-mediated transport of this compound is Na⁺-coupled and electrogenic. This was further confirmed via kinetic analyses. Na⁺-activation kinetics indicated that the relationship between SLC5A8-mediated MMF transport and Na⁺ concentration was sigmoidal (Fig. 3B), suggesting involvement of more than one Na⁺ in the activation process. The Hill coefficient was 2.0 ± 0.1, indicating that for every molecule of MMF transported, two Na⁺ ions are also transported. The concentration of Na⁺ necessary for half-maximal activation of MMF-induced currents was 38 ± 1 μM. As an additional means of confirming that MMF is in fact a transportable substrate for SLC5A8, studies were conducted also in the presence of ibuprofen, a blocker of human SLC5A8. ³¹ If MMF is truly a transportable substrate for SLC5A8, then the addition of ibuprofen to the perifusion medium should interfere with MMF-induced inward currents in SLC5A8-expressing oocytes. To test this, we monitored inward currents in SLC5A8-expressing oocytes perifused with MMF (1 mM) in the presence of Na⁺ and increasing concentrations of ibuprofen. Indeed, ibuprofen decreased the magnitude of MMF-induced currents in a dose-dependent manner (Fig. 3C). The concentration of ibuprofen needed for half-maximal blockade of these currents was 12.9 ± 0.4 μM. 

Figure 3

 

Demonstration of human SLC5A8-mediated MMF transport in the Xenopus laevis oocyte expression system. (A) SLC5A8 cRNA-injected oocytes were perifused with 1 mM MMF in the presence of NaCl (Na⁺) or NMDG-Cl (-Na⁺). Pyruvate (1 mM) served as a positive control. Currents were monitored by the two-microelectrode voltage-clamp technique. (B) MMF (1 mM)-induced inward currents were monitored in SLC5A8 cRNA-injected oocytes in the presence of increasing concentrations of Na⁺ (2.5–100 mM). The concentration of Cl⁻ was maintained at 100 mM by appropriately substituting NaCl with NMDG chloride. The experiment was performed with three different oocytes. To adjust for variations in the expression levels in the different oocytes, the currents were normalized by taking the maximal current induced by the highest concentration of Na⁺ (100 mM) as 1 and then calculating the currents induced by MMF at other concentrations of Na⁺ as a fraction of this maximal current. Inset: Hill plot. (C) Dose-dependent blockade of human SLC5A8-mediated MMF transport by ibuprofen. MMF (100 μM)-induced currents were monitored in SLC5A8 cRNA-injected oocytes in the presence of NaCl and increasing concentrations of ibuprofen. The percent inhibition of MMF-induced currents by each concentration of ibuprofen was then calculated and used to determine the concentration of ibuprofen needed to cause 50% maximal inhibition by fitting the Michaelis-Menten equation to the data. Inset: Eadie-Hofstee plot. I, percent inhibition; S, ibuprofen concentration in μM.

We showed previously that FAE other than MMF are also capable of inducing system x_c ⁻ activity in RPE cells. Therefore, we examined the interaction of various FAE with SLC5A8. All FAE examined in the study induced inward currents in SLC5A8-expressing oocytes. However, fumarate did not (Fig. 4A). We also monitored the voltage dependence of the inward currents induced by FAE, with pyruvate as a positive control. The concentration of all compounds tested was 1 mM. The general characteristics of the currents induced in the presence of each compound were similar in terms of Na⁺ dependence (data not shown); however, the magnitude of the inward currents induced was greatest for MMF (Fig. 4A). Hyperpolarization of the oocyte membrane enhanced the magnitude of induced currents, as expected given the known electrogenic nature of SLC5A8. With the exception of fumarate, all of the compounds tested were transported by SLC5A8. The magnitude of the inward currents induced by these compounds was in the following order: pyruvate > MMF > MEF > DMF > DEF > fumarate. We compared the kinetic features of SLC5A8-mediated transport of two of these compounds, MMF and MEF. SLC5A8-mediated transport of both compounds was saturable. The Michaelis constant (K _m) was 860 ± 70 μM for MMF (Fig. 4B) and 1130 ± 155 μM for MEF (Fig. 4C). 

Figure 4

 

Voltage dependence of currents induced by fumarate esters in human SLC5A8-expressing oocytes and saturation kinetics of human SLC5A8-mediated transport. (A) Oocytes expressing human SLC5A8 were exposed to fumarate or its esters: diethylfumarate, dimethylfumarate, monoethylfumarate, and monomethylfumarate (1 mM each), and the substrate-induced currents were monitored at different testing membrane potentials using the two-microelectrode voltage-clamp technique. Pyruvate (1 mM) served as a positive control for SLC5A8-mediated transport. (B) Inward currents were monitored in SLC5A8 cRNA-injected oocytes in the presence of increasing concentrations of MMF in perfusion buffer. The experiment was performed in four different oocytes. Because the expression levels of SLC5A8 varied to some extent in different oocytes, the data were normalized by taking the maximal current induced by the highest concentration of MMF (5.0 mM) as 1 in each oocyte and then calculating the currents induced by MMF at other concentrations as a fraction of this maximal current. Inset: Eadie-Hofstee plot. (C) Identical experiments were performed for monoethylfumarate. Inset: Eadie-Hofstee plot.

Here we demonstrate, for the first time, MMF to be a transportable substrate of SLC5A8. Interestingly, many substrates of this transport system (i.e., butyrate, pyruvate, and 3-bromopyruvate) function quite effectively as histone deacetylase (HDAC) inhibitors and also as agonists of the G-protein–coupled receptor GPR109A. In fact, these properties explain directly many of the cellular actions elicited by such compounds. ^34–41 We therefore evaluated MMF with respect to each of these properties in order to determine whether either underlies the stimulatory effect of MMF on system x_c ⁻ expression and function. Along these lines, we found that while MMF is transported by SLC5A8, it does not significantly impact HDAC activity (see Supplementary Material and Supplemental Fig. S1); however, the compound does, as reported by others, activate GPR109A (see Supplementary Material and Supplemental Fig. S2). ³⁷ To determine whether transport by SLC5A8 and/or interaction with GPR109A is required for the observed induction of system x_c ⁻ expression by MMF, we evaluated xCT expression in Gpr109a^−/− and Slc5a8^−/− (knockout; KO) mouse RPE cells and corresponding wild-type (Gpr109a^+/+ , Slc5a8^+/+ ) cells cultured in the presence or absence of MMF (100 μM) for 16 hours by RT-PCR. xCT expression increased substantially in all cells cultured in the presence of MMF irrespective of Slc5a8 or Gpr109a expression (Fig. 5A). To confirm this at the functional level, we monitored the Na⁺-independent uptake of glutamate in wild-type, Gpr109a^−/− , and Slc5a8^−/− mouse RPE (Figs. 5B, 5C). In agreement with our findings at the mRNA level, the stimulation of Na⁺-independent [³H]glutamate uptake/system x_c ⁻ activity in the presence of MMF was present in wild-type as well as Gpr109a^−/− and Slc5a8^−/− mRPE. These data indicate that although MMF is both a transportable substrate for SLC5A8 and an agonist for GPR109A, neither of these factors is involved directly in the stimulatory effects of system x_c ⁻ activity induced in RPE by this compound. 

Figure 5

 

MMF-induced increase in xCT mRNA is independent of Gpr109a and Slc5a8. Wild-type (WT), Gpr109a^−/− , and Slc5a8^−/− mouse RPE were cultured in the presence or absence of MMF (100 μM) for 16 hours. (A) xCT mRNA expression was analyzed by RT-PCR. Uptake of [³H]glutamate (2.5 μM) was monitored in the presence (NaCl) and absence of Na⁺ (NMDG-Cl) for 15 minutes at 37°C using (B) wild-type and Gpr109a^−/− RPE and (C) wild-type and Slc5a8^−/− RPE cells.

Upregulation of Hif-1α and Nrf2 by MMF

Hif-1α and Nrf2 are critically involved in the regulation of the cellular response to oxidative stress and have been implicated previously in the regulation of the cellular actions associated with FAE. ^36,42–48 Thus, here we evaluated the effects of MMF on the expression of these factors at the RNA and protein level. ARPE-19 cells were cultured in the presence or absence of MMF (100 μM) for various periods of time followed by preparation of total RNA and protein. RT-PCR and Western blotting were then carried out using primers or antibodies specific for Hif-1α or Nrf2. Robust induction of Hif-1α expression was detected in RNA isolated from cells cultured in the presence of MMF (Fig. 6A). Western blotting analysis revealed an associated increase in Hif-1α protein that persisted up to 24 hours postexposure to this compound (Fig. 6B). A similar phenomenon was observable with respect to Nrf2 protein expression (Fig. 6C). To determine whether the early activation of Nrf2 and Hif-1α observed in our experimental system is associated with modulation of downstream factors in a manner consistent with reduced oxidative stress/elevation of cellular antioxidant responses, we evaluated also the expression of heme oxygenase 1 (HO-1) and monitored cellular levels of glutathione in control versus MMF-treated cells. Indeed, HO-1 expression and glutathione levels increased significantly in cells following MMF treatment (Figs. 6D, 6E, respectively). Though the induction of system x_c ⁻ expression appears to occur independently of SLC5A8 and GPR109A, we did not know whether interaction of MMF with either of these factors plays any role in the observed induction of Hif-1α and Nrf2. Thus, we evaluated also the influence of MMF on the expression of these transcription factors in mRNA samples obtained from primary RPE cells isolated from Gpr109a^−/− and Slc5a8^−/− retinas (Fig. 6F). MMF induced the expression of Hif-1α and Nrf2 mRNA in RPE cells even in the absence of Gpr109a or Slc5a8 as observed for the induction of system x_c ⁻. 

Figure 6

 

Evaluation of MMF-induced Hif-1α and Nrf2 RNA and protein expression in ARPE-19 and primary mouse RPE. ARPE-19 and primary mouse RPE cells were incubated with or without MMF (100 μM) and RNA and protein samples collected at various intervals in time postincubation. (A) RT-PCR analysis of Hif1-α expression in ARPE-19 cells 6 hours postincubation with or without MMF. (B) Time-dependent analysis of Hif1-α protein in ARPE-19 cells as analyzed by Western blot. (C) Time-dependent analysis of Nrf2 protein in ARPE-19 cells as analyzed by Western blot. (D) RT-PCR analysis of HO-1 expression in ARPE-19 cells 16 hours postexposure to MMF. (E) Measurement of glutathione levels in ARPE-19 cells 16 hours postexposure to MMF; *P < 0.05. (F) RT-PCR analysis of Hif1-α and Nrf2 mRNA in wild-type and Gpr109a^−/− and Slc5a8^−/− mouse RPE cells cultured in the presence or absence of MMF (100 μM) for 6 hours.

The MMF-induced stimulation of system x_c ⁻ activity in RPE has tremendous implications in terms of protecting these cells and thereby outer retina from oxidant- and/or inflammation-induced damage. System x_c ⁻ expression in retina, however, is not limited to RPE. In fact, expression of the transport system has been shown previously in Müller glia and ganglion cells. ^24,49 Given the differential effects elicited by FAE in various cell types, determining the effect of these compounds and their specific mode of action in different retinal cell types is highly warranted and absolutely necessary if this therapy is to be extrapolated successfully to treatment of patients with retinal disease. We therefore conducted additional studies using primary mouse Müller cells and found that in that cell type, MMF also induces system x_c ⁻ (see Supplementary Material and Supplemental Fig. S3). Hence, the MMF-induced stimulation of system x_c ⁻ expression/activity is not a phenomenon exclusive to RPE, but is instead one that occurs also in other retinal cell types. These data provide added support for the possible benefit of this compound as a therapeutic agent in retina. 

Glutathione Levels in Mouse Eyes following Intravitreal Injection of MMF

Glutathione is the most abundant endogenous antioxidant in retina, and is hence essential for protection of retinal cells against oxidative stress. ^50–53 A number of studies document the differential effects of various FAE on levels of intracellular glutathione in cell types including RPE. ²² Interestingly, however, the in vivo efficacy of these compounds in the eye has not been studied. Our present findings show that MMF stimulates system x_c ⁻ activity in RPE and other retinal cells. The extremely high dependence of endogenous glutathione synthesis on the functional activity of this transport system as the provider of cysteine for glutathione synthesis, its distribution in retina, and our present evidence of its stimulation in the presence of MMF provide strong support for possible benefit in terms of elevation of overall ocular levels of glutathione in vivo. To ascertain the validity of this prediction, MMF was delivered to mouse eyes via intravitreal injection. Eyes injected with PBS served as controls. Animals were sacrificed at 24 hours postinjection, and ocular levels of glutathione were measured. There was a significant increase in glutathione concentration in MMF-injected eyes compared to PBS-injected controls (Fig. 7, P < 0.05). These data suggest that the MMF-induced increase in system x_c ⁻ expression in retinal cells contributes to an increase in the level of this antioxidant in ocular tissues. It is important, however, to emphasize here that entire eye cups rather than isolated retinas were used to calculate the concentration of glutathione, a factor that may not only explain why the values obtained are several times lower than that reported in the literature on retinal glutathione content in mice, ^54–56 but, more importantly, suggests that the positive impact of MMF on levels of glutathione in retina specifically may be somewhat underestimated in our assay. 

Figure 7

 

Effects of intravitreal injection of MMF on ocular levels of glutathione. MMF (1 μL; 10 mM solution prepared in PBS) was injected into the vitreous body of the right eye of each animal; the left eye served as a contralateral control and received an equal volume of PBS (n = 6). At 24 hours postinjection, animals were euthanized, and eyes were harvested. Following removal of the lens, eyes were homogenized in PBS, and glutathione concentration was measured (*P < 0.001).

Discussion

Oxidative stress and resultant inflammation are crucial factors in the development/progression of many pathologic conditions, including those affecting retina. ^8,9 Therefore, understanding mechanisms and developing novel strategies for combating these factors are important. Indeed, there is a burgeoning literature suggesting strongly that therapies targeted at limiting oxidative stress and inflammation may block early biochemical/cellular alterations in retina even before they are clinically evident. ^57,58 Such therapies, therefore, have tremendous potential for slowing/preventing progression of degenerative retinal diseases. Whether FAE may fit this bill, at present, we do not know. However, based upon the demonstrated immunomodulatory, antioxidant, and protective properties of these agents in other cell types/disease states, this is highly plausible and thus strongly warrants further investigation. As such, we began to explore this novel concept by evaluating the effects of MMF, the primary active ingredient/metabolite in the antipsoriatic drug Fumaderm, on the expression and functional activity of system x_c ⁻, a critical regulator of endogenous glutathione synthesis/cellular redox status. In these experiments, we have focused on RPE because of the major role of this cell type in protecting the retina against oxidative and/or inflammatory insult and in the maintenance/preservation of overall retinal health and visual function. 

Here we have demonstrated, for the first time, the MMF-induced upregulation of system x_c ⁻ expression and functional activity in ARPE-19 and primary mouse RPE cells. Stimulation of system x_c ⁻ expression/functional activity significantly increases the entry of cystine in mammalian cells. This cystine is in turn promptly reduced to cysteine, the rate-limiting amino acid in glutathione synthesis. Glutathione is the most abundant endogenous antioxidant in retina and in fact the most abundant low molecular weight thiol in the entire cellular redox system. As such, this compound is essential not only for quenching ROS, thereby protecting cells from potentially deleterious reactions, but also for the transmission of redox signals/overall regulation of redox balance. Though we have focused primarily on RPE here, given the widespread nature of system x_c ⁻ expression, the phenomena described herein are likely not unique to this one retinal cell type. Indeed, supplemental studies conducted using primary mouse Müller cells provide tangible evidence supportive of this claim. Hence, the stimulation of system x_c ⁻ by MMF in multiple retinal cell types may have important implications in terms of antioxidant regulation in retina/protection of this tissue as a whole. This is further supported by our in vivo finding of increased retinal levels of glutathione following intravitreal injection of MMF. 

The actions of FAE within a single cell/tissue type are known to involve multiple mechanisms. Thus we evaluated also the involvement of certain specific mechanisms in the observed effects of MMF in RPE. We confirmed MMF's ability to activate the anti-inflammatory G-protein–coupled receptor GPR109A and showed, for the first time, its recognition as a transportable substrate by the Na⁺-coupled monocarboxylate transporter SLC5A8. Though we found that neither of these properties influences directly the stimulatory actions of MMF on system x_c ⁻, this does not nullify the importance of the receptor or the transporter in the mediation of potentially other beneficial actions of MMF, but rather lends support to the very diverse nature of the mechanisms/cellular actions associated with MMF. Further studies are needed to ascertain the physiologic significance of MMF-induced activation of GPR109A- and SLC5A8-mediated transport of this compound in retina and other tissues. Additionally, little is known regarding the expression of SLC5A8 or GPR109A in retina under pathologic conditions, a factor that may influence the response of cells and animals to FAE treatment. This topic is also worthy of further investigation. In addition to evaluating the interaction of MMF with proteins expressed on the cell membrane (i.e., GPR109A and SLC5A8), and the influence thereof on the observed stimulation of system x_c ⁻ in our experimental system, we looked also at the effects of the MMF on molecules critically involved in regulation of the cellular response to oxidative stress at the transcriptional level, Hif-1α and Nrf2. Indeed, there are a number of recent reports citing a mechanistic link between FAE/MMF administration and induction/stabilization of Nrf2^2–4; our present finding of Nrf2 induction in the presence of MMF in cultured RPE cells is highly congruent with these reports. Nrf2 regulates the expression of many genes involved in the propagation of cellular antioxidant responses including xCT, the catalytic subunit of system x_c ⁻. Hence, the observed stimulation of xCT expression/system x_c ⁻ functional activity and the increase in glutathione concentrations following MMF treatment may be due in part to the induction/stabilization of Nrf2. Interestingly, there is a burgeoning literature supportive of the targeting of this pathway (Nrf2/Keap1) for therapeutic management of neurovascular diseases, a category in which degenerative diseases of the retina like age-related macular degeneration and diabetic retinopathy would also fall. ⁵⁹ The fact that Hif-1α was also induced in MMF-treated RPE suggests that this transcription factor may also play a role. Like Nrf2 and other transcription factors, Hif-1α regulates the expression of a plethora of genes. Others have reported the induction/stabilization of Hif-1α by the parent compound fumarate and esters such as MEF in association with the anti-inflammatory and antiangiogenic effects elicited by FAE. ^42–45 Furthermore, the very recent report by Sims et al. ⁴⁶ demonstrating the regulation of system x_c ⁻ by Hif-1α supports the feasibility of the prediction that in addition to regulating anti-inflammatory/antiangiogenic effects, the induction of Hif-1α by FAE such as MMF is important also for regulation of antioxidant signaling, specifically upregulation of system x_c ⁻/increased glutathione synthesis. 

Taken together, data from the present investigation show that MMF stimulates multiple pathways in RPE cells, resulting ultimately in the potentiation of cellular events leading to the upregulation of genes/mechanisms that function to protect cells against various forms of insult. Increased antioxidant signaling/upregulation of system x_c ⁻ likely represents just one such consequence. Indeed, development of novel combinatorial therapies (coadministration of compounds targeting one or more pathologic processes) for treatment/prevention of degenerative retinal diseases is an area of much ongoing investigation. ^57,60 In this regard, FAE/MMF therapy is an interesting and highly attractive option. MMF represents a single compound that by itself can affect multiple processes (oxidative stress, inflammation, and angiogenesis) in a beneficial manner, a phenomenon that has broad clinical implications in retina. An added bonus is the fact that much clinical information is already available regarding the pharmacokinetics and toxicology of FAE/MMF in humans. However, efforts to move forward with implementation of FAE therapy for use in retina and other disease states (e.g., multiple sclerosis) hinge critically on elucidation of the mechanisms by which these compounds work at the cellular level in various cell/tissue types, as we have begun to do here in RPE cells. Future studies will be devoted to determining the bioavailability of FAE in retina following various modes of administration (e.g., oral, intravitreal), as well as the in vivo efficacy of these compounds in limiting oxidative stress/inflammation in retina and thereby preserving or improving overall retinal function. 

Supplementary Materials

pdf S1 

Acknowledgments

The Gpr109a^−/− animals employed in this study were a generous gift from Stefan Offermanns, Max Planck Institute, Bad Nauheim, Germany. 

Footnotes

 Supported by National Eye Institute (NEI) Grant EY018053 (principal investigator, PMM).

Footnotes

 Disclosure: S. Ananth, None; E. Babu, None; R. Veeranan-Karmegam, None; B.R. Bozard Baldowski, None; T. Boettger, None; P.M. Martin, None