Interphotoreceptor Retinoid-Binding Protein Mitigates Cellular Oxidative Stress and Mitochondrial Dysfunction Induced by All-trans-Retinal

Abstract

Purpose

Point and null mutations in interphotoreceptor retinoid-binding protein (IRBP) cause retinal dystrophy in affected patients and IRBP-deficient mice with unknown mechanism. This study investigated whether IRBP protects cells from damages induced by all-trans-retinal (atRAL), which was increased in the Irbp^−/− retina.

Methods

Wild-type and Irbp^−/− mice retinal explants in buffer with or without purified IBRP were exposed to 800 lux light for different times and subjected to retinoid analysis by high-performance liquid chromatography. Purity of IRBP was determined by Coomassie Brilliant Blue staining and immunoblot analysis. Cellular damages induced by atRAL in the presence or absence of IRBP were evaluated in the mouse photoreceptor-derived 661W cells. Cell viability and death were measured by 3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) and TUNEL assays. Expression and modification levels of retinal proteins were determined by immunoblot analysis. Intracellular reactive oxygen species (ROS) and nitric oxide (NO) were detected with fluorogenic dyes and confocal microscopy. Mitochondrial membrane potential was analyzed by using JC-1 fluorescent probe and a flow cytometer.

Results

Content of atRAL in Irbp^−/− retinal explants exposed to light for 40 minutes was significantly higher than that in wild-type retinas under the same light conditions. All-trans-retinal caused increase in cell death, tumor necrosis factor activation, and Adam17 upregulation in 661W cells. NADPH oxidase-1 (NOX1) upregulation, ROS generation, NO-mediated protein S-nitrosylation, and mitochondrial dysfunction were also observed in 661W cells treated with atRAL. These cytotoxic effects were significantly attenuated in the presence of IRBP.

Conclusions

Interphotoreceptor retinoid-binding protein is required for preventing accumulation of retinal atRAL, which causes inflammation, oxidative stress, and mitochondrial dysfunction of the cells.

The light-sensitive chromophore of the visual pigments in cone and rod photoreceptors is 11-cis-retinal (11cRAL). It keeps the visual pigments in an inactive conformation and functions as a molecular switch for activating opsins in response to light stimulation. When light hits the visual pigments, its energy converts 11cRAL to all-trans-retinal (atRAL), inducing structural rearrangement and activation of opsins. The activated opsins then trigger the phototransduction that converts light signal into electrical and neural signals in the photoreceptors. Both 11cRAL and atRAL possess a highly reactive aldehyde group and are the precursors of A2E,^1,2 a cytotoxic byproduct of the retinoid visual cycle necessary for regenerating 11cRAL.³ Several studies have suggested that atRAL is a critical pathogenic factor in light-induced retinal degeneration and retinopathy of mouse model.⁴ In addition, A2E and atRAL dimer have been suggested to contribute to the pathology of Stargardt disease.⁵ These studies indicate that preventing excess accumulation of atRAL and 11cRAL is important for maintaining retinal health.

Interphotoreceptor retinoid-binding protein (IRBP) expressed by photoreceptors is known to bind with 11cRAL, atRAL, 11-cis-retinol (11cROL), and all-trans-retinol (atROL).^6–9 This 140-kDa secretory glycoprotein with four homologous modules is the most abundant soluble protein in the interphotoreceptor matrix (IPM).^10,11 The high abundance implicates its important role in protecting, solubilizing, and detoxifying retinoids in the IPM and photoreceptors. In ex vivo experiments, IRBP has been shown to promote release of 11cRAL from the retinal pigment epithelium (RPE)^12,13 and atROL from the neural retina after photobleaching of the visual pigments.^14–16 Irbp^−/− mice display severe retinal degeneration and vision impairment.^17–21

The significance of IRBP function in human retinal health and vision is reflected by the fact that mutations in IRBP are associated with retinal diseases and vision impairment. Recently, two nonsense mutations (Y510X and E1152X) of IRBP have been identified in children with high myopia and retinal dystrophy.²² In addition, a missense mutation (D1080N) has been found in adults with retinitis pigmentosa (RP).²³ The RP-associated missense mutation has been shown to abolish secretion of IRBP in culture cells, suggesting that the mutation results in loss of IRBP in the patients' IPM.²⁴ Optical coherence tomography and electroretinography have shown that all affected patients with any of the mutations exhibit abnormal retinal structure and severe impairment of both cone and rod visual function,^22,23 suggesting that IRBP is required for normal retinal development or for maintaining retinal structure and function.

We recently have shown that tumor necrosis factor-α (TNF-α), a potent proinflammatory cytokine that can induce cellular apoptosis and necrosis, is increased several-fold in the Irbp^−/− retina and IPM.²⁵ However, the molecular mechanisms leading to upregulation and activation of TNF in Irbp^−/− retina remain unknown. In this study, we first investigated the effect of IRBP deficiency on accumulation of atRAL in the retina. We then analyzed the pathologic roles of atRAL in induction of TNF activation, oxidative stress, mitochondrial dysfunction, and cell death. At the same time, we also investigated protective role of IRBP against atRAL-induced cytotoxicity in the mouse photoreceptor-derived 661W cells.

Methods

Animals and Ex Vivo Experiments

Wild-type (WT) 129S2/Sv (Charles River Laboratories, Wilmington, MA, USA) and Irbp^−/− mice were maintained in 12-hour cyclic light at ∼30 lux. The Irbp^−/− mice were originally generated by Liou et al.¹⁷ In this study, we used Irbp^−/− mice with the 129S2/Sv genetic background.¹⁸ All animal experiments followed the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the protocols were approved by the Institutional Animal Care and Use Committee. For retinal ex vivo experiments, we isolated retinas from 5-week-old dark-adapted WT and Irbp^−/− mice and placed each in 100 μL serum-free Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Carlsbad, CA, USA) containing 20 mM HEPES buffer (pH 7.4). We added 4 μM purified IRBP, which is similar to IRBP concentration in rat IPM,²⁶ into the media of WT retinal explants. After exposure to white light at 800 lux for 10, 20, or 40 minutes, retinoids were analyzed as described below.

High-Performance Liquid Chromatography Analysis of Retinoids

We converted atRAL and 11cRAL to syn- and anti-retinaloxime isomers by using 150 mM hydroxylamine before extraction of retinoids from retinas.¹⁸ Retinoids were extracted with hexane and analyzed by normal-phase HPLC as previously described.²⁷ In brief, retinoids in hexane extractions were evaporated, redissolved in 100 μL hexane, and separated on a silica column (Zorbax-Sil 5 μm, 250 × 4.6 mm; Agilent Technologies, Santa Clara, CA, USA) by gradient (0.2%–10% dioxane in hexane at 2.0 mL/min flow rate) elution of mobile phase in an Agilent 1100 HPLC system.

Expression and Purification of IRBP

The 293T-LC cells²⁸ grown in DMEM containing 10% fetal bovine serum (FBS) were transfected with IRBP-expression plasmid (pIRBP)¹⁸ or pRK5 mock vector, as described previously.²⁹ One day after transfection, we replaced the media with fresh serum-free or 1% FBS-containing DMEM and incubated an additional 16 hours. The media were centrifuged at 20,000g for 30 minutes. Presence of IRBP in the serum-free media of the pIRBP-transfected (IRBP-medium) or pRK5-transfected (mock-medium) cells was analyzed by Coomassie Brilliant Blue (CBB) staining and immunoblot analysis using an antibody against IRBP.²⁴ The IRBP-medium and mock-medium were then concentrated with Amicon Ultra 100-KD molecular-weight cutoff (Millipore, Billerica, MA, USA). To purify IRBP from the 1% FBS-containing media, we precipitated IRBP with 16% saturated ammonium sulfate. The precipitates were resolved in 50 mM Tris-HCl buffer (pH 7.5) containing EDTA-free protease inhibitors, 150 mM NaCl, and 0.1 mM dithiothreitol, and were incubated with concavalin A (ConA) Sepharose 4B (GE Healthcare, Pittsburgh, PA, USA). After washing three times, IRBP was eluted with the buffer containing 10% methyl a-D-mannopyranoside (Sigma-Aldrich Corp., St. Louis, MO, USA). Purity of IRBP was analyzed by CBB staining and immunoblot analysis.

Cell Viability Assay

The 661W cells, a mouse photoreceptor-derived cell line, were kindly provided by Muayyad R. Al-Ubaidi at the University of Oklahoma and were maintained as described previously.³⁰ For cell viability assay, the cells at a density of 3 × 10³ cells/well in a 96-well plate were incubated with the indicated concentration of atRAL in serum-free medium with or without IRBP for 24 hours. Cell viability was determined by using a CellTiter⁹⁶ AQueous assay kit (Promega, Madison, WI, USA) as described previously.³¹ Briefly, the culture media were replaced with 95 μL fresh media containing 5 μL 3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) solution. After 1 hour, the absorbance at 490 nm was measured by using a microplate reader (Molecular Device, Sunnyvale, CA, USA).

TUNEL Assay

Detection and quantification of cell death at single-cell level was determined by TUNEL assay using the in situ cell death detection kit (Roche, Indianapolis, IN, USA), according to the manufacturer's protocol. Briefly, 661W cells on glass coverslips (5 × 10⁵ cells/well in six-well plate) were treated with 1.8 μM atRAL, washed with phosphate-buffered saline (PBS), fixed with 4.0% paraformaldehyde in PBS for 15 minutes, and then permeabilized with 0.5% Triton X-100 in PBS for 10 minutes. After incubation with the mixture of enzyme solution and label solution for 1 hour at 37°C, fluorescent signals were captured with a Zeiss LSM710 Meta confocal microscope (Oberkochen, Germany) with an ×20 objective lens.

Measurement of Intracellular Reactive Oxygen Species (ROS) and Nitric Oxide (NO)

The cellular ROS and NO were measured by using 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) (Sigma-Aldrich Corp.) and 4,5-diaminofluorescein diacetate (DAF-2DA) (Sigma-Aldrich Corp.), respectively. 661W cells in a Corning Costar 96-well black clear-bottom plate at a density of 3 × 10³ cells/well were incubated with 1.8 μM atRAL in the presence or absence of IRBP for 2 hours, and then with 20 μM DCFH-DA or 10 μM DAF-2DA for 20 minutes. After washing with PBS twice, the relative fluorescence unit was measured at an excitation wavelength of 485 nm and an emission wavelength of 528 nm with a fluorescence microplate reader (Molecular Device). The cells incubated in the same condition were used for capturing fluorescent signals with a Zeiss LSM710 Meta confocal microscope with an ×20 objective lens. Each image was collected under the same system settings (e.g., the same laser intensity and scan speed).

Preparation of Cell Lysates and Mitochondrial and Nuclear Fractions

661W cells incubated with atRAL in serum-free media with or without IRBP were washed with ice-cold PBS containing protease and phosphatase inhibitors and incubated with lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, 1% SDS, 1% Nonidet P-40, 1 mM EDTA, 1 mM dithiothreitol, 200 nM aprotinin, 20 μM leupetin, 50 μM phenanthroline, 280 μM benzamidine-HCl) for 30 minutes on ice. After centrifugation at 16,000g for 20 minutes at 4°C, the supernatant was used as the cell lysates. Mitochondrial and nuclear fractions were prepared from 661W cells by using the Cell Fractionation Kit (Abcam, Cambridge, MA, USA), according to the manufacturer's protocol.

Analysis of Mitochondrial Membrane Potentials

661W cells were incubated with 1.8 μM atRAL in fresh serum-free medium, in IRBP-medium, or in mock-medium for 6 hours. After washing with PBS, the cells were incubated with 2.5 μg/mL 5,5′,6,6′-tetrachloro-1,1,3′,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) (Thermo Fisher Scientific, Waltham, MA, USA) for 20 minutes. JC-1 is a green fluorescent cationic carbocyanine dye that enters the negatively charged mitochondrial matrix and forms J-aggregates that exhibit red fluorescence. After washing with PBS, the green fluorescent JC-1 monomers and the red fluorescent J-aggregates in the cells were detected with a Gallios Flow Cytometer (Beckman Coulter, Brea, CA, USA). A total of 30,000 cells were acquired. The data were analyzed by using the Kaluza Analysis Software and displayed in histogram of JC-1 green fluorescence and in dot plot of J-aggregate red fluorescence (y-axis) against JC-1 green fluorescence (x-axis).

Immunoblot Analysis

Proteins were separated in an 8%, 10%, or 12% polyacrylamide gel by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and transferred onto an Immobilon-P membrane (Millipore). The membrane was incubated sequentially in blocking buffer, primary antibody, and horseradish peroxidase–conjugated anti-mouse or rabbit IgG secondary antibody. Antibodies against β-actin (Sigma-Aldrich Corp.), IRBP,²⁴ TNF (Abcam), Adam17 (Chemicon, Billerica, MA, USA), S-nitrosocysteine (Sigma-Aldrich Corp.), apoptosis inducing factor (AIF) (Abcam), NADPH oxidase-1 (NOX1; GeneTex, Irvine, CA, USA), and PARP1 (Cell Signaling Technology, Danvers, MA, USA) were used as the primary antibodies. Immunoblots were visualized and quantified as described.³²

Statistical Analysis

Data were expressed as means ± standard deviation (SD) from at least three separate experiments unless otherwise indicated. Data were analyzed by using 1-way analysis of variance, followed by each pair of Student's t-tests for multiple comparisons. All analyses were done with Microsoft Excel software (Microsoft, Redmond, WA, USA).

Results

Accumulation of atRAL in Irbp^−/− Retina Exposed to Light

Since IRBP has been shown to play an important role in protection and transportation of retinoids, we tested if atRAL accumulates in Irbp^−/− retina after long-time photobleaching of the visual pigments. To do this experiment, we purified IRBP from pIRBP-transfected 293T-LC cell media by a ConA affinity chromatography (Fig. 1A). We then placed dark-adapted WT and Irbp^−/− retinas in IRBP-containing or not containing buffer, which mimics WT and Irbp^−/− IPMs, respectively. After exposure to bright light for different times, we analyzed retinoids in the retinas. Since Irbp^−/− mice had retinal degeneration, we expressed the levels of atRAL as the percentage of total retinoids. As shown in Figure 1B, atRAL in WT retina decreased as light-exposure time increased. The amounts of atRAL in WT retina exposed to light for 40 minutes was significantly lower than those in WT retina exposed to light for 10 minutes (P < 0.01). However, the amounts of atRAL in Irbp^−/− retina were not greatly changed during 10 to 40 minutes of light-exposure time (Fig. 1B). As a result, the contents of atRAL in Irbp^−/− retina exposed to light for longer times were significantly higher than those in WT retina (Figs. 1B, 1C).

Figure 1.

Interphotoreceptor retinoid-binding protein is required for reducing atRAL in retina exposed to light. (A) Media of 293T-LC cells transfected with pRK5 or pIRBP and IRBP purified from the pIRBP-transfected cell medium were analyzed by CBB staining and immunoblot (IB) analysis. (B) Relative contents of atRAL in WT and Irbp^−/− mice retinal explants exposed to 800 lux light for the indicated times. Wild-type and Irbp^−/− retinal explants were incubated in buffer with or without purified IRBP, respectively. Asterisks indicate significant differences between WT and Irbp^−/− mice (*P < 0.05, **P < 0.01). Error bars indicate SD (n = 4). (C) Representative HPLC chromatograms of retinoids in WT and Irbp^−/− retinas exposed to 800 lux light for 40 minutes. The peaks of atROL, syn- and anti-retinaloxime (ROX) isomers of 11cRAL and atRAL are labeled.

IRBP Alleviated Cell Death Induced by atRAL

Since atRAL possesses a highly reactive aldehyde group, we tested the effects of atRAL on cell viability. We incubated 661W cells with a series of increasing concentration (0–5 μM) of atRAL for 24 hours. As shown in Figure 2A, cell viability was significantly decreased by atRAL treatment in a dose-dependent manner. The estimated IC₅₀ was approximately 1.8 μM. To examine the effect of IRBP on the atRAL-mediated inhibition of cell viability, we performed the cell viability assay with 1.8 μM atRAL in the IRBP-medium or mock-medium. Coomassie Brilliant Blue and immunoblot staining revealed that the IRBP-medium, but not the mock-medium, contained relatively pure IRBP (Fig. 2B). Cell viability assay showed that viabilities of 661W cells incubated with atRAL in the mock-medium were similar to those of cells incubated with atRAL in fresh serum-free medium (Fig. 2C). In contrast, the viabilities of the cells incubated with atRAL in the IRBP-medium were increased by at least 25% (Fig. 2C). Consistent with these results, the number of TUNEL-positive nuclei was significantly reduced in the cells incubated with atRAL in the IRBP-medium, as compared to the cells incubated with atRAL in the mock-medium (Fig. 2D).

Figure 2.

Interphotoreceptor retinoid-binding protein alleviated cell death induced by atRAL. (A) Viabilities of 661W cells incubated with the indicated concentration of atRAL in serum-free medium were determined by MTS assay. **P < 0.01 indicates significant differences between control and atRAL-treated cells. (B) Coomassie Brilliant Blue staining and IB analysis of IRBP in serum-free media of 293T-LC cells transfected with pRK5 or pIRBP. (C) Viabilities of 661W cells incubated with 0 or 1.8 μM atRAL in fresh medium or medium of 293T-LC cells transfected with pRK5 (mock) or pIRBP (IRBP). (D) TUNEL assay of 661W cells incubated with 1.8 μM atRAL as in (C). Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). TUNEL-positive nuclei were indicated with arrows. Scale bars: 50 μm.

IRBP Attenuated TNF Activation Triggered by atRAL

In a previous study,²⁵ we have shown that TNF is dramatically increased in soluble fraction of the Irbp^−/− IPM. We therefore tested whether atRAL can promote production of soluble TNF (sTNF), the active form of TNF,³³ in 661W cells. We incubated the cells with a series of increasing concentration of atRAL in serum-free media. Immunoblot analysis showed that the cleaved sTNF was increased in a dose-dependent manner (Fig. 3A). The immunoblot intensity of sTNF in the cells treated with 1.8 μM atRAL was approximately 2-fold of that in control cells (Fig. 3A). To test if IRBP inhibits the atRAL-mediated activation of TNF, we incubated 661W cells with 1.8 μM atRAL either in fresh serum-free medium, IRBP-medium, or mock-medium. As shown in Figure 3B, the atRAL-mediated elevation of sTNF was markedly suppressed by IRBP-medium, but not by mock-medium.

Figure 3.

Interphotoreceptor retinoid-binding protein attenuated TNF activation triggered by atRAL. (A) Immunoblot analysis showing increase of sTNF in 661W cells treated with atRAL for 6 hours. Histogram shows relative immunoblot intensities of sTNF normalized by actin. (B) Immunoblot analysis of sTNF in 661W cells treated with atRAL in serum-free medium, mock-medium, or IRBP-medium. Histogram shows relative immunoblot intensities of sTNF. *P < 0.05 indicates significant differences between cells treated with atRAL in mock- versus IRBP-medium. (C) Immunoblot analysis of Adam17 in 661W cells treated with atRAL for indicated times. Histogram shows normalized intensities of Adam17 immunoblots. (D) Immunoblot analysis of Adam17 in 661W cells treated with atRAL in the indicated media. (E, F) Immunoblot analysis of Adam17 in the retina (E) and RPE (F) of WT and Irbp^−/− mice. **P < 0.01 indicates significant differences between WT and Irbp^−/− tissues. Error bars denote SD (n = 3).

IRBP Inhibited Adam17 Upregulation Induced by atRAL

Since TNF-α converting enzyme (TACE)/Adam17 metalloprotease has been shown to mediate cleavage of membrane-bound pro-TNF in the post-Golgi,³⁴ we tested the effect of atRAL on expression levels of Adam17 in 661W cells. Immunoblot analysis showed that protein levels of Adam17 (approximately 100 kDa) were increased 50% in the cells treated with 1.8 μM atRAL for 4 to 6 hours (Fig. 3C). This upregulation of Adam17 was significantly attenuated by incubation of atRAL with IRBP-medium (Fig. 3D). To test if Adam17 is upregulated in the Irbp^−/− retina and RPE, we compared expression levels of Adam17 in WT and Irbp^−/− tissues. Immunoblot analysis showed that Adam17 was increased at least 2-fold in the Irbp^−/− retina and RPE (Figs. 3E, 3F).

IRBP Suppressed atRAL-Induced Production of ROS in 661W Cells

To analyze the mechanism by which atRAL causes cell damage, we tested if atRAL stimulates ROS generation in 661W cells. DCFH-DA is a cell-permeable fluorogenic probe. Reactive oxygen species in the cells convert DCFH-DA to highly fluorescent DCF. We incubated DCFH-DA with 661W cells treated with 0 to 1.8 μM atRAL in the presence or absence of IRBP. Confocal microscopy showed that the DCF fluorescent signal was dramatically increased in the cells treated with atRAL in fresh medium, as compared to the control cells maintained in fresh medium without atRAL (Figs. 4A, 4B). The DCF fluorescent signal was significantly decreased in the cells treated with atRAL in the IRBP-medium, but not in the mock-medium (Figs. 4A, 4B). We then analyzed expression of NOX1, a catalytic subunit of the superoxide-generating NADPH oxidase, in 661W cells. Immunoblot analysis showed that atRAL stimulated expression of NOX1 in a time-dependent manner (Fig. 4C). To test if IRBP suppresses the atRAL-induced upregulation of NOX1, we incubated 661W cells with atRAL in the IRBP-medium or mock-medium. Immunoblot analysis showed that the IRBP-medium, but not mock-medium, significantly reduced expression levels of NOX1 in the cells (Fig. 4D). To confirm this result in vivo, we analyzed expression levels of NOX1 in the retina and RPE of WT and Irbp^−/− mice. As shown in Figures 4E and 4F, NOX1 was increased more than 2-fold in the Irbp^−/− retina and RPE compared to WT retina and RPE.

Figure 4.

Interphotoreceptor retinoid-binding protein suppressed atRAL-induced production of ROS. (A) Cells treated with atRAL in fresh serum-free, mock-, or IRBP-medium were incubated with the DCFH-DA fluorogenic probe. Reactive oxygen species–induced fluorescence was visualized by confocal microscopy. Scale bars: 50 μm. (B) Relative fluorescence intensities in the cells shown in (A). Similar results were observed in three independent experiments. **P < 0.01 indicates significant differences between cells treated with atRAL in mock- versus IRBP-medium. (C) Immunoblot analysis showing increase of NOX1 in the cells treated with atRAL for indicated times. Histogram shows relative intensities of NOX1 immunoblots normalized by actin. Error bars denote SD (n = 3). (D) Immunoblot analysis of NOX1 in 661W cells treated with atRAL either in serum-free, mock-, or IRBP-medium for 2 hours. Normalized intensities of NOX1 immunoblots are shown in the histogram. (E, F) Immunoblot analysis (E) and relative immunoblot intensities (F) of NOX1 in the retina and RPE of WT and Irbp^−/− mice. **P < 0.01 indicates significant differences between WT and Irbp^−/− tissues.

IRBP Attenuated atRAL-Induced Production of NO

Since NO also plays an important role in induction of oxidative condition,³⁵ we tested if atRAL increases intracellular NO level in 661W cells. To detect NO we used DAF-2DA, a fluorescent detector of NO in living cells. Confocal pictures showed that the fluorescent signal (red) of DAF in 661W cells treated with 1.8 μM atRAL in fresh medium was drastically increased as compared to the control cell maintained in fresh medium without atRAL (Figs. 5A, 5B). The IRBP-containing medium, but not mock-medium, significantly attenuated NO production stimulated by atRAL (Figs. 5A, 5B). To confirm these results, we analyzed the levels of protein nitration in the cells. Immunoblot analysis using an antibody against S-nitrosocysteine showed that atRAL treatment promoted protein nitration in a time-dependent manner (Figs. 5C, 5D). This atRAL-induced protein nitration was significantly attenuated by the IRBP-medium, but not by mock-medium (Figs. 5E, 5F).

Figure 5.

Interphotoreceptor retinoid-binding protein attenuated atRAL-induced production of NO in 661W cells. (A) Cells treated with atRAL in fresh serum-free, mock-, or IRBP-medium were incubated with DAF-2DA fluorescence detector of NO for 20 minutes. Nitric oxide was visualized by confocal microscopy. Scale bars: 20 μm. (B) Relative DAF intensities in the cells shown in (A). Similar results were observed in three independent experiments. **P < 0.01 indicates significant differences between cells treated with atRAL in mock- versus IRBP-medium. (C) S-nitrosylated proteins in the cells treated with atRAL for indicated times were detected by immunoblot analysis with an antibody against S-nitrosocysteine. (D) Normalized relative immunoblot intensities of S-nitrosylated proteins shown in (C). (E) Immunoblot analysis of S-nitrosylated proteins in the cells treated with atRAL in serum-free, mock-, or IRBP-medium. (F) Relative immunoblot intensities of S-nitrosylated proteins shown in (E). **P < 0.01 indicates significant differences between cells treated with atRAL in mock- versus IRBP-medium.

IRBP Mitigated Mitochondrial Dysfunction Induced by atRAL

Oxidative stress has been shown to cause mitochondrial dysfunction.³⁶ We therefore analyzed mitochondrial membrane potentials by using a cell-permeable JC-1 dye, a mitochondrial potential indicator that exists either as a green fluorescent monomer at depolarized membrane potentials or as a red fluorescent J-aggregate at hyperpolarized membrane potentials. As shown in Figures 6A and 6B, the populations of cells with green fluorescence were increased in the cells treated with atRAL in fresh medium (20% ± 1.2%) or mock-medium (23% ± 1.0%), as compared to the untreated control cells (3.9% ± 0.1%) or the cells treated with atRAL in IRBP-medium (4.6% ± 0.8%).

Figure 6.

Interphotoreceptor retinoid-binding protein mitigated atRAL-induced mitochondrial dysfunction in 661W cells. (A, B) Cells maintained either in fresh serum-free, mock-, or IRBP-medium containing 1.8 μM atRAL were incubated with JC-1 for 20 minutes. Fluorescence density of JC-1 monomers (x-axis) was displayed in histogram (A), and fluorescence density of J-aggregates (y-axis) against JC-1 monomers (x-axis) was displayed in dot plot (B). Similar results were observed in three independent experiments. (C) Immunoblot analysis of AIF in nuclear and mitochondrial fractions of 661W cells treated as in (A). PARP1 and COX IV were detected as nuclear or mitochondrial marker, respectively. (D, E) Relative AIF contents in nuclear (D) or mitochondrial (E) fractions of the treated cells were determined by the immunoblot intensities in (C). *P < 0.05 indicates significant differences between cells treated with atRAL in mock-medium versus IRBP-medium.

We then further analyzed the mitochondrial integrity by detecting AIF in mitochondrial and nuclear fractions. Apoptosis inducing factor is normally localized to mitochondria, but when mitochondrion is damaged, it moves to the cytosol and the nucleus.³⁷ Quantitative immunoblot analysis showed that AIF was increased in the nuclear fraction of 661W cells treated with atRAL in fresh medium or mock-medium, as compared to the control cells and the cells treated with atRAL in IRBP-medium (Figs. 6C, 6D). In contrast, AIF was decreased in the mitochondrial fraction of 661W cells treated with atRAL in fresh medium or mock-medium, as compared to the control cells and the cells treated with atRAL in IRBP-medium (Figs. 6C, 6E).

Discussion

Using purified IRBP and retinal explants of WT and Irbp^−/− mice, we demonstrated that IRBP is essential for preventing accumulation of atRAL in the retina under bright light conditions. Furthermore, we found that IRBP plays an important role in protection of the mouse photoreceptor-derived 661W cells from TNF activation, oxidative stress, and mitochondrial dysfunction caused by atRAL. These findings may reflect a mechanism of IRBP function in retinal neuroprotection.

The well-known biochemical function of IRBP is its high-affinity binding with retinoids involved in the visual cycle. Although all retinoid isomers can bind with IRBP, the major composition of retinoids bound to IRBP changes depending on light condition. In dark-adapted eyes, 11cRAL increases while atROL decreases. In light-exposed eyes, however, atROL and atRAL increases, whereas 11cRAL decreases.^7,8 Retinols decompose rapidly into a number of products, including their aldehyde form, retinals.⁹ Interphotoreceptor retinoid-binding protein has been shown to protect retinol from isomerization, oxidation, and photodegradation.^9,38 In this study, we observed that the content of atRAL in Irbp^−/− retinal explants exposed to light for 40 minutes was significantly higher than that in WT retinas under the same light conditions (Fig. 1). Since IRBP promotes the release of atROL from isolated retina after photobleaching of the visual pigments,^14,15 the higher content of atRAL in the Irbp^−/− retina could be due to oxidation of atROL retained in photoreceptors by retinol dehydrogenases (RDHs) and/or by photothermal oxidation.

The neural retina and RPE contain a large amount of retinoids. In dark-adapted eye, at least 80% of total retinoids are 11cRAL and localized to the neural retina,¹⁸ mainly in photoreceptors. Following exposure to bright light, which causes large photobleaching of the visual pigments, at least 60% of total retinoids in the neural retina are atRAL.¹⁸ Mutations in retinol dehydrogenase 12 (RDH12), an atRAL reductase in rods and cones,³⁹ are associated with Leber's congenital amaurosis.⁴⁰ Disruption of atRAL clearance pathways in photoreceptors induces severe photoreceptor dystrophy in Rdh8^−/−Abca4^−/− mice,⁴ suggesting that atRAL contributes at least in part to the pathogenic mechanisms of the diseases. In agreement with these studies, we observed that atRAL inhibited cell viability and induced cell death in cultured 661W cells (Fig. 2). We also found that IRBP significantly alleviated cell death induced by atRAL (Fig. 2). The bovine IRBP binds atRAL with approximately 2 μM of K_D and 3.2 of stoichiometry.⁴¹ Structural studies have identified a restrictive hydrophobic cavity and a shallow hydrophobic cleft as ligand-binding sites in the Xenopus IRBP.^42,43 These findings suggest that IRBP could prevent direct interaction of atRAL with molecules important for cellular functions by holding atRAL in the cavity and cleft.

Lack of IRBP has been shown to result in increase of TNF in the Irbp^−/− retina.²⁵ Tumor necrosis factor is a proinflammatory cytokine that promotes cell death.⁴⁴ Tumor necrosis factor mRNA in the Irbp^−/− retina is 2-fold higher than that in WT retina, whereas the amount of TNF in the Irbp^−/− IPM is more than 10-fold higher than that in WT IPM,²⁵ suggesting that secreted sTNF is increased in the Irbp^−/− IPM. It is known that the secreted sTNF, the more active form of TNF, is derived from the transmembrane-anchored TNF precursor through the TACE/Adam17-mediated proteolytic process in the post-Golgi and plasma membrane.^34,45 In this study, we observed that Adam17 was upregulated in 661W cells by atRAL treatment (Fig. 3C). Consistent with this result, sTNF, but not TNF precursor, was significantly increased in atRAL-treated 661W cells (Fig. 3A). Both sTNF and Adam17 were reduced when the cells were treated with atRAL in the presence of IRBP (Figs. 3B, 3D). These results suggest that atRAL is one of the pathogenic factors that induced upregulation of TNF and Adam17 in the Irbp^−/− retina and RPE (Figs. 3E, 3F) and that Adam17 upregulated by atRAL promoted secretion of sTNF into the Irbp^−/− IPM.

Reactive oxygen species and NO are important intracellular messengers that regulate cellular function, survival, and differentiation.^46,47 Reactive oxygen species induce upregulation of Adam17,⁴⁸ and NO activates Adam17.⁴⁹ We found that atRAL stimulated the generation of ROS and NO in 661W cells (Figs. 4A, 4B, 5A, 5B). These observations suggest that ROS and NO are involved in upregulation and activation of Adam17 in 661W cells treated with atRAL.

Excessive generation of ROS and/or NO induces cellular oxidative and nitrosative stresses by damaging macromolecules important for cellular structure, function, and survival.^35,47 NOX1 plays a critical role in generation of ROS. Reaction between superoxide and NO produces peroxynitrite, a strong oxidant that also causes cellular damage by reacting directly with macromolecules.^35,50 In the present study, we observed that NOX1 expression (Fig. 4C) and protein S-nitrosylation were significantly increased in 661W cells treated with atRAL (Fig. 5C). In addition, both atRAL (Fig. 1) and NOX1 (Figs. 4E, 4F) were increased in the Irbp^−/− ocular tissues. Consistent with these results, IRBP inhibited the atRAL-mediated NOX1 upregulation and protein S-nitrosylation (Figs. 4D, 5E, 5F). Since oxidative and nitrosative stresses are pathologic mechanisms involved in rod and cone photoreceptor degeneration in retinal degenerative diseases,^35,51–53 our results suggest that IRBP plays an important role in protection of photoreceptors from oxidative stress through suppression of atRAL accumulation and its thiol-dependent antioxidant activity.⁵⁴

In pathophysiologic conditions, mitochondria are the main producers of ROS; and excessive ROS induces oxidative stress that results in mitochondrial dysfunction in the cells.^47,55 In addition, mitochondrial dysfunction is often associated with activation of various cell death signals. Translocation of AIF from the mitochondrial intermembranous space to the nucleus is one of the important cell death factors that promotes DNA fragmentation and cell death.³⁷ In the present study, we found that treatment of 661W cells with atRAL in the absence of IRBP caused depolarization of mitochondrial membranes (Figs. 6A, 6B), and this mitochondrial damage was accompanied by translocation of AIF into the nuclei of the cells (Figs. 6C, 6D). These results are consistent with previous studies showing that Vitamin A derivative and β-carotene cleavage products, including retinaldehydes, induce cellular oxidative stress and mitochondrial dysfunction.^56,57 Interestingly, it has been shown that IRBP binds carotenoids with a K_D similar to that of retinoids,⁴¹ suggesting that IRBP may inhibit carotenoid cleavage that produces retinaldehydes. The present study and the previous studies also suggest that atRAL-induced oxidative stress and mitochondrial dysfunction may contribute to retinal degeneration in Irbp^−/− mice.

In a previous study,¹⁸ we have shown that dark rearing could not prevent degeneration of photoreceptors in Irbp^−/− mice. This observation suggests that IRBP has an unidentified function(s) necessary for photoreceptor survival. Expression of IRBP in the developing retina implicates a potential role for IRBP in retinal development and/or maturation.^6,58 Wisard et al.²¹ have found that the size and weight of Irbp^−/− mouse eyes are greater than those of WT mice. The excessive ocular enlargement starts between P7 and P10. Although the mechanism of how IRBP regulates ocular development remains unknown, it is known that the IPM, where IRBP accumulates, contains numerous growth and neurotrophic factors that regulate cell proliferation, differentiation, and survival.^59,60 The significant increase of TNF in the Irbp^−/− IPM suggests that IRBP is required for maintaining IPM homeostasis essential for retinal neurodevelopment and function. Both atRAL-mediated cellular damages and aberrant regulation of extracellular signaling proteins in the IPM may contribute to retinal degeneration in Irbp^−/− mice. Identification of IRBP-interacting protein in the IPM is therefore important for understanding the molecular mechanism of IRBP function and may contribute to the development of therapeutic interventions for retinal dystrophy caused by IRBP mutations.