Flavonoids Protect Human Retinal Pigment Epithelial Cells from Oxidative-Stress–Induced Death

Abstract

purpose. To determine whether specific dietary and synthetic flavonoids can protect human retinal pigment epithelial (RPE) cells from oxidative-stress–induced death.

methods. The efficacy and potency were determined of a variety of dietary and synthetic flavonoids on the survival of human ARPE-19 cells and primary human RPE cells treated with either hydrogen peroxide (H₂O₂) or t-butyl hydroperoxide (t-BOOH). We determined the effective concentrations (EC₅₀s) and the toxicities (LD₅₀s) of the flavonoids after 24 hours, by using the MTT assay. The efficacy of vitamins E and C on RPE cell survival were compared under identical conditions. The ability of specific flavonoids to protect RPE cells from cell death was determined at various time intervals after the cells were exposed to oxidative stress. The ability of flavonoids to block the accumulation of intracellular reactive oxygen species was examined with dichlorofluorescein (DCF) fluorescence. Finally, the ability of flavonoids to induce phase-2 detoxifying enzymes was tested by immunoblot analysis for the transcription factor Nrf2 and the phase-2 gene product heme-oxygenase 1.

results. Specific flavonoids protected human RPE cells from oxidative-stress–induced death with efficacies between 80% and 100% and potencies in the high-nanomolar and low-micromolar range. The toxicities of most of the effective flavonoids were low. The effective flavonoids included the dietary flavonoids fisetin, luteolin, quercetin, eriodictyol, baicalein, galangin and EGCG, and the synthetic flavonoids, 3,6-dihydroxy flavonol and 3,7 dihydroxy flavonol. Several flavonoids can protect RPE cells even when they are added after the cells have been exposed to oxidative stress. The flavonoids acted through an intracellular route to block the accumulation of reactive oxygen species. Many of these flavonoids induced the expression of Nrf2 and the phase-2 gene product heme-oxygenase 1 in human RPE cells.

conclusions. The results identify a select group of flavonoids that protect RPE cells from oxidative-stress–induced death with a high degree of potency and low toxicity. Many of these flavonoids also induce the expression of phase-2 detoxification proteins which could function to provide additional protection against oxidative stress. This select group of flavonoids and the foods that contain high levels of these compounds may have some clinical benefit for patients with retinal diseases associated with oxidative stress.

The excitement associated with the advances in antiangiogenic therapy for patients with age-related macular degeneration (ARMD) has been tempered by the clinical findings that after successful regression of subretinal neovascularization, visual improvement has been limited. In both the TAP (Treatment of AMD with Photodynamic Therapy) and the VIP (Verteporfin in Photodynamic Therapy) studies, only 5% and 8% of patients, respectively, gained vision compared with control subjects after the treatment of subretinal neovascularization. ^{1 2}  

As a result, it is becoming increasingly clear that new approaches for treatment should be focused on both preventing the initial insults that lead to disease progression and rescuing the retinal pigment epithelium (RPE) and photoreceptor cells that have been damaged. 

Oxidative stress is thought to play an important role in the pathogenesis of ARMD. The retina-RPE exists in an environment that is rich in endogenous sources of reactive oxygen species (ROS). Contributing to the production of ROS are the high metabolic rate of the photoreceptors, the phagocytic activity of RPE cells, ^{3 4} the high local oxygen concentration and the chronic exposure to light. ^{5 6} Although multiple physiologic defenses exist to protect the retina-RPE from the toxic effects of light and oxidative damage, mounting evidence suggests that chronic exposure to oxidative stress over the long term may damage the retina-RPE and predispose it to the development of age-related macular degeneration. ^{5 6} Supporting this theory is the observation that large drusen, which are deposited under the RPE in patients with macular degeneration, consist of insoluble aggregates of oxidized lipids and proteins derived from the photochemical reactions of visual transduction. ⁷  

As a preventative measure, various attempts are being made to reinforce the eye’s defenses against oxidative stress. Clinical epidemiologic studies have shown that specific dietary habits are associated with a decreased incidence of advanced ARMD ^{8 9} The AREDS (Age-Related Eye Disease Study) recently confirmed that increasing the body’s defenses against oxidative stress with specific antioxidants and mineral supplements can preserve vision in patients with macular degeneration and reduce the rate of disease progression. ¹⁰ As a result, general interest in dietary management is growing as the number of patients at risk of macular degeneration increases with the aging of the baby-boomer generation. 

Flavonoids are a class of natural biological products that have evolved to protect plants from the oxidative damage induced by chronic exposure to ultraviolet light. They are structurally heterogeneous, polyphenolic compounds that are present in high concentrations in fruits, vegetables and other plant-derived foods, such as teas and other beverages, and are regularly consumed in the human diet. ¹¹ Flavonoids have many physiological health benefits, including protection from cardiovascular disease and cancer, and most of these beneficial effects are thought to stem from their potent antioxidant and free radical scavenging properties, as well as their ability to modulate many cellular enzyme functions. ¹² Notably, many of the foods that have been associated with a reduced risk of macular degeneration in clinical epidemiologic studies contain high concentrations of flavonoids. ¹¹ At least one flavonoid has been identified in the mammalian retina. ¹³ These observations raise the question of whether some of the protection obtained in consuming foods associated with a reduced risk of macular degeneration is related to the flavonoids that they contain. This possibility would in turn indicate that the consumption of flavonoids from specific foods or as nutritional supplements could protect patients at risk of macular degeneration. 

Flavonoids can provide both short and long-term protection against oxidative stress through a variety of mechanisms. Many flavonoids act directly as antioxidants, neutralizing toxic ROS by donating hydrogen ions. ¹⁴ Yet, equally and potentially even more important, flavonoids can modulate cell-signaling pathways. ¹⁵ In particular, they can induce the expression of phase-2 proteins that function to enhance the cell’s natural defenses against oxidative stress. Phase-2 proteins catalyze several different reactions that neutralize reactive oxygen species and increase the intracellular concentrations of compounds that protect against oxidative stress, such as glutathione. ¹⁶ Among these phase-2 proteins are some of the key enzymes involved in glutathione metabolism (glutathione S-transferase [GSH], glutamate cysteine ligase) and other antioxidant enzymes, including heme-oxygenase 1 (HO-1). ¹⁶  

From a biological perspective, there are key advantages for a cell to induce phase-2 proteins to fight oxidative stress. In contrast to direct antioxidants that are consumed immediately after their interaction with a reactive oxygen species, the induction of phase-2 proteins allows the cell to mount a more prolonged and sustained defense that will continue to function after the antioxidants are consumed. Phase-2 proteins are induced by a wide variety of diverse chemical substances, including sulforaphane and oltipraz, as well as compounds that induce oxidative stress. ^{17 18 19 20} These inducers activate the transcription factor Nrf2 (NF-E2-related factor-2) which binds to the antioxidant response element (ARE), an upstream regulatory element that is shared by all the genes encoding phase-2 proteins. ²¹  

Based on these observations, we propose that specific flavonoids can protect eye-derived cells from oxidative-stress–induced death and may be of benefit in the treatment of retinal diseases associated with oxidative stress. In a very recent study, we showed that specific flavonoids could protect retinal ganglion cells from oxidative-stress–induced death. ²² In the work presented herein, we focused on RPE cells and flavonoids found in common fruits and vegetables and asked four questions: (1) Can specific dietary flavonoids protect RPE cells from oxidative-stress–induced cell death? (2) Can these flavonoids prevent cell death after the exposure has occurred? (3) Can flavonoids act through an intracellular route to reduce the accumulation of reactive oxygen species? (4) Do flavonoids activate Nrf2 in human RPE cells and induce the expression of phase 2 genes? The answers to these questions are described in the following pages and indicate that specific flavonoids and the foods that contain high levels of these compounds may be of clinical benefit to patients with retinal diseases associated with oxidative stress. 

Methods

Chemicals

Flavonoids were purchased from Alexis Biochemicals (San Diego, CA) or Sigma-Aldrich, Inc. (St. Louis, MO). Ascorbic acid (vitamin C), vitamin E, 3-(4,5-dimethylthiazol-2-yl-2,5-diphenyl tetrazolium bromide (MTT), and dimethylsulfoxide were also purchased from Sigma-Aldrich, Inc. The structures of the flavonoids used in this study are shown in Fig. 1 . 

Cell Culture

Adult human retinal pigment epithelial (ARPE-19) cells were a gift from Larry Hjelmeland (University of California, Davis, CA). These cells were grown in basal DMEM/F12 medium (Invitrogen, San Diego, CA) containing 10% fetal bovine serum (FBS; Hyclone, Logan, UT), 15 mM HEPES, 2.5 mM l-glutamine, 0.5 mM sodium pyruvate, and 20 mM sodium bicarbonate. ARPE-19 cells grew with undifferentiated characteristics at low passages. Over the course of multiple passages, the cells spontaneously developed a differentiated morphology and expressed cellular markers of differentiation (see later section on RT-PCR). Both the undifferentiated and the differentiated ARPE-19 cells were maintained and passaged in basal media. Primary human RPE159 cells were a gift from Brian McKay (University of Arizona, Tucson, AZ) and were cultured in low-calcium medium containing DMEM high glucose medium (calcium-free; Invitrogen) supplemented with 10% dialyzed FBS (Hyclone), 2 mM l-glutamine, 1 mM sodium pyruvate, and 50 μM CaCl₂. The cells were dissociated from the culture dishes (Versene; 0.53 mM EDTA in Hanks’ balanced salt solution) followed by 0.25% trypsin/EDTA. 

Cytotoxicity Assay

Dose–response assays were performed on both differentiated and undifferentiated RPE cells to determine the concentrations of t-BOOH and H₂O₂ that would reliably kill 80% to 95% of the cells. ARPE-19 cells and human primary 159 RPE cells were seeded onto 96-well plates at various densities (3,200, 5,000, 10,000, 20,000, 40,000, 60,000 or 80,000 cells/well) and grown for 24 hours. On the evening before the start of the cytoprotection assays, the differentiated ARPE-19 cells were switched to a low-calcium medium containing DMEM high glucose (calcium-free; Invitrogen) supplemented with 10% dialyzed FBS (Hyclone), 2 mM l-glutamine, 1 mM sodium pyruvate, and 50 μM CaCl₂. On the following day, differentiated and undifferentiated cells were replenished with fresh culture medium containing 10% dialyzed FBS and preincubated with flavonoids or other natural products for 1 hour before the addition of the chemical oxidants. t-BOOH or H₂O₂ were added at concentrations that had been found to kill more than 80% to 90% of the cultured cells in dose–response assays. After an overnight incubation, cell viability was determined by a modified version of the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay. ²³ The assay was performed by removing the cell culture medium and replacing it with 100 μL fresh culture medium containing 5.0 mg/mL MTT. After 4 hours of incubation at 37°C, cells were solubilized overnight with 100 μL of a solution containing 50% dimethylformamide and 20% SDS (pH 4.7). The absorbance at 560 nm was measured with a microplate reader (Spectromax 190; Molecular Devices Corp., Sunnyvale, CA). To assure that the spectrophotometric readings correlated with cell viability, all cells were examined by microscopy before the addition of the MTT. Each experiment was performed at least three times, and multiple control subjects were included. For each concentration of a specific compound, six wells were analyzed. Of these six wells, the cells in two were treated with the flavonoid alone to determine the toxicity of the compound. The cells in the remaining four wells were treated with the compound of interest and either t-BOOH or H₂O₂. Background absorbance values consisted of blank wells (with no cells) into which medium, MTT dye, and solubilization buffer were added. The background readings were subtracted from the average absorbance readings of the treated wells to obtain an adjusted absorbance reading that represented cell viability. This reading was divided by the adjusted absorbance reading of untreated cells in control wells to obtain the percentage of cell survival. To determine the efficacy, potency and EC₅₀s of the compounds of interest, the dose–response data were analyzed (Prism 4 software; GraphPad, San Diego, CA). 

Microscopy

A (model TE200; Nikon, Melville, NY) microscope equipped with phase contrast optics, a 10 × 0.45 NA objective and a charge-coupled device (CCD) camera (Orca II; Hamamatsu, Bridgewater, NJ) was used to photograph the RPE cells. Fourteen-bit images were collected using the slow-scan mode. 

Isolation of Total RNA and RT-PCR

Total RNA was isolated from undifferentiated and differentiated ARPE-19 cells (RNeasy Mini kit; Qiagen, Valencia, CA) and then treated with RNase-free DNase I to remove any contaminating genomic DNA. The isolated RNA had 28S/18S ratios greater than or equal to 2.0 and optic density (OD) 260/280 ratios greater than or equal to 2.0. One microgram of total RNA was reverse transcribed (RETROscript kit; Ambion, Austin, TX). 

PCR was performed with 3 μL of each cDNA sample with 1 U (Platinum Taq DNA Polymerase High Fidelity; Invitrogen), 0.2 mM dNTP, 2 mM MgSO₄, 1× High Fidelity PCR buffer, and 10 picomoles of each primer, in a total volume of 50 μL for 35 cycles (GeneAmp 9700 thermocycler; Applied Biosystems International [ABI], Foster City, CA). Each cycle consisted of 30 seconds at 94°C. Thirty seconds at 55°C and 45 seconds at 74°C. Specific primers used for RPE65 were forward (F): 5′-AACCTCTTCCATCACATCAACACC-3′ and reverse (R): 5′-GATTCAAGCCAAGTCCATACGC3′; and for FGFR1 were F: 5′-CGGCAGCATCAACCACACATAC-3′ and R: 5′-AGCACCTCCATCTCTTTGTCGG-3′. Fifteen microliters of each amplificate were assessed by agarose gel electrophoresis. 

Delayed-Response Cytoprotection Assays

ARPE-19 cells were seeded at 3200 cells/well in 96-well plates as described earlier. The following day, the medium was replaced with DMEM/F12+10% dialyzed FBS, and the cells were treated with either t-BOOH or H₂O₂ in DMEM/F12+10% dialyzed FBS. At various time points after the addition of the oxidants, flavonoids (50 μM) were added to the wells and the cell survival was analyzed by the MTT assay 24 hours later. To compare the cytoprotective effect of the flavonoids with the cell survival due to the removal of the oxidant from the media, the media containing the oxidant were replaced with basal medium at various time points and the MTT assay was performed 24 hours later. 

ROS Production

Differentiated ARPE-19 cells were seeded at confluence into black, clear-bottomed 96-well plates. The following day, the media were removed, and the cells were washed with PBS+++ (PBS supplemented with 1 mM CaCl₂, 0.5 mM MgCl₂, and 7.5 mM glucose) and incubated for 20 minutes at 37°C in PBS+++ containing 10 μM 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCFDA; Invitrogen, Eugene, OR). The cells were washed with PBS+++, incubated with 500 μM t-BOOH in PBS+++ for 2 hours at 37°C and washed twice with PBS+++ to remove the t-BOOH from the wells. After these washes, the cells were incubated with flavonoids at a concentration of 30 μM in PBS+++ for 1 hour at 37°C and washed with PBS+++ to remove the flavonoids from the wells. Intracellular ROS production was measured on a spectrofluorometer (Gemini EM microplate; Molecular Devices Corp.) with an excitation λ of 485 nm and emission λ of 530 (ex/em = 485/530 nm). 

Western Blot Analysis of HO-1 and Nrf2 Expression

To evaluate the expression of the phase-2 protein HO-1 and the transcription factor Nrf2, we plated the ARPE-19 cells into 60-mm plates at 1.0 × 10⁶ cells/plate. The following day, the cells were replenished with fresh DMEM/F12 medium+10% dialyzed FBS and pretreated with various flavonoids at a concentration of 50 μM for 2 to 24 hours. For HO-1, the cells were treated for 24 hours and then rinsed twice in ice-cold PBS, scraped into lysis buffer (50 mM HEPES [pH 7.5] 50 mM NaCl, 50 mM NaF, 10 mM sodium pyrophosphate, 5 mM EDTA, 1% Triton X-100, 1 mM sodium orthovanadate, complete protease inhibitor cocktail (Roche, Indianapolis, IN), 1 mM phenylmethylsulfonyl fluoride (PMSF), incubated at 4°C for 30 minutes, and centrifuged at 14,000 rpm for 10 minutes at 4°C. Protein levels were determined using the BCA protein assay (Pierce, Rockford, IL). 

To evaluate the expression of Nrf2, ARPE-19 cells were treated with various flavonoids for 2 or 4 hours. Cells were washed twice in ice-cold Tris-buffered saline and scraped into nuclear fractionation buffer (10 mM HEPES [pH 7.9] 10 mM KCl, 0.1 mM EDTA, 1 mM DTT, 1 mM PMSF and 1 mM Na₃VO₄). After a 15-minute incubation at 4°C, 10% NP-40 was added to achieve a final concentration of 0.625%. The cells were vortexed for 10 seconds, centrifuged at 14,000 rpm for 30 seconds to pellet nuclei, resuspended in nuclear fractionation buffer containing NP-40 and sonicated gently to break up nuclei. Protein concentrations were determined using the BCA protein assay. 

Fifteen micrograms of total protein were solubilized in 2× SDS sample buffer containing 10% β-mercaptoethanol, separated on 10% Tris-glycine polyacrylamide gels (Invitrogen), and transferred to nitrocellulose membranes by electroblot. Blots were washed in Tris-buffered saline containing 0.1% Tween 20 and 5% nonfat dairy milk, incubated in antibodies to HO-1 (rabbit polyclonal 1:5000; StressGen Biotechnologies), Nrf2 (rabbit polyclonal 1:500; Santa Cruz Biotechnology, Santa Cruz, CA) or β-actin (mouse monoclonal 1:3000; Sigma) at 4°C overnight. Blots were washed three times, incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:3000; Bio-Rad, Hercules, CA) or HRP-conjugated goat anti-mouse IgG (1:3000; Bio-Rad) and developed using chemiluminescence (SuperSignal West Pico Luminescent; Pierce). 

Results

Oxidative Stress Assay Conditions

The oxidants t-BOOH and H₂O₂ have been used in the design of models of oxidative stress in RPE cells by many investigators because of their rapid membrane permeability and depolarizing effects on the mitochondrial membrane potential. ^{24 25 26 27} For these reasons, we selected t-BOOH and H₂O₂ for our studies and performed a series of dose–response assays to determine the working concentrations that led to a consistent and high degree of cytotoxicity, which we defined as the levels of t-BOOH and H₂O₂ that killed 80% to 95% of the RPE cells after an overnight incubation. We initially used low-density, proliferating cultures of RPE cells because of their high sensitivity to oxidative stress ²⁸ and their ability to provide a simple, high-throughput screen for compounds that protect from oxidative stress. We found that reliable and consistent cytotoxicity was obtained in low-density ARPE-19 cells that were exposed to t-BOOH and H₂O₂ at concentrations of 120 and 250 μM, respectively. 

Protection of RPE Cells

To determine whether flavonoids protect ARPE-19 cells from oxidative-stress–induced cell death, we tested a variety of dietary and nondietary flavonoids from each of the six major classes of flavonoids (Fig. 1) . Included in this list were the dietary flavonoids luteolin and quercetin, which are present in the dark-green, leafy vegetables (spinach and kale) that have been associated with protection from macular degeneration in epidemiologic studies. ^{9 29} We also tested vitamins C and E, both of which were found to be beneficial in reducing the progression of macular degeneration in the AREDS. ¹⁰  

We found that selected flavonoids protected ARPE-19 cells from oxidative-stress–induced death with excellent efficacy. Figure 2shows a representative set of micrographs of ARPE-19 cells that were treated with luteolin. Treatment with either t-BOOH or H₂O₂ led to marked cell death after 24 hours of exposure (Figs. 2C 2E) . In contrast, the cells treated with luteolin alone (Fig. 2B)or luteolin in the presence of either t-BOOH or H₂O₂ (Figs. 2D 2F)appeared similar to the control cells (Fig. 2A) . 

We determined the dose–response curves of various flavonoids, to compare their efficacy and potency. We found that different flavonoids had distinct profiles, suggesting that they may have different pharmacological characteristics. As shown in Figure 3 , luteolin, quercetin, and fisetin protected 100% of the cells at concentrations less than or equal to 50 μM. Eriodictyol, galangin, 3,6-dihydroxyflavone, and 3,7-dihydroxyflavone achieved more than >75% protection in both assays. Baicalein protected >80% of cells treated with t-BOOH but had minimal effect on cells exposed to H₂O₂. 

We calculated the effective half maximal concentrations for protection (EC₅₀s) from the dose–response curves (Table 1) . The most potent flavonoids include the dietary flavonoids fisetin, quercetin, luteolin, galangin, baicalein, and eriodictyol and the synthetic flavonoids 3,6-dihydroxyflavone, and 3,7-dihydroxyflavone. The potencies of these flavonoids are all within the low micromolar range. Some flavonoids had no cytoprotective effect in either the t-BOOH or the H₂O₂ assays. 

We analyzed the toxicities (LD₅₀s) of the flavonoids and found that almost all of the most potent compounds had relatively low toxicity, with the exception of the synthetic flavonoid, 3,7-dihydroxyflavonol. We did not determine the LD₅₀s of the flavonoids that were ineffective (Table 1) . 

We found that vitamins C and E protect ARPE-19 cells from oxidative-stress–induced death, although their efficacies were significantly lower than the efficacies of the flavonoids shown in Figure 3 . The cytoprotective effects of vitamin C and various forms of vitamin E are shown in Figure 4 . The efficacy of quercetin is shown for comparison. 

Effect of ARPE-19 Cell Differentiation and Cell Density

Most of our early studies were performed on low-passage ARPE-19 cells that were quite sensitive to oxidative stress and grew with relatively undifferentiated characteristics (Figs. 2 5A) . However, over the course of many passages, our ARPE-19 cells spontaneously changed to a differentiated phenotype ³⁰ (Fig. 5B)and expressed cellular markers of differentiation, including RPE65 (Fig. 6) . Because RPE cells are differentiated in vivo and are likely to be present at various densities, depending on the health of the retina, we decided to compare the responses of the undifferentiated and differentiated cells at a variety of cell densities to determine whether the state of differentiation and/or the plating cell density influenced the protective effect seen with the flavonoids. 

First, we performed dose–response assays to establish the appropriate conditions for these studies. We found that the concentrations of t-BOOH and H₂O₂ necessary to kill 85% to 90% of the cells were directly proportional to the plating cell density. Subconfluent ARPE-19 cells plated at 3200 cells/well died at relatively low concentrations of oxidants (Figs. 7A 7B) . The concentrations of the oxidants that were necessary to kill the cells increased steadily as the cells became more dense (Figs. 7A 7B) . At confluence, the concentrations of t-BOOH and H₂O₂ that were needed to kill the cells reached 500 to 600 μM and 800 to 1000 μM, respectively. Because we observed some interexperimental variability in the plating density and the cytotoxic concentrations of t-BOOH and H₂O₂ (depicted by the SE bars in Fig. 7 ), we performed dose–response cytotoxicity studies immediately preceding each cytoprotection assay to establish the specific concentrations of t-BOOH and H₂O₂ that would kill 80% to 95% of the cells in a specific experiment. 

Overall, we found that the cytoprotective effect of the flavonoids appeared to be independent of the state of ARPE-19 cell differentiation and independent of the plating cell density. In experiments repeated at least three times for each condition, we observed a cytoprotective efficacy of 90% to 100% with four of the most effective flavonoids—luteolin, fisetin, eriodictyol, and quercetin—in both undifferentiated and differentiated ARPE-19 cells plated at a wide range of cell densities (3,200, 5,000, 10,000, 20,000, 40,000, 60,000, and 80,000 cells/well). The only exception was eriodictyol, which had an efficacy of 75% in low-density cultures (3200 cells/well) but protected at more than 90% efficacy at higher densities (5000+ cells/well). The dose–response curves of confluent and differentiated ARPE-19 cells treated with luteolin, quercetin, fisetin, and eriodictyol are shown in Figure 8 . It should be noted that these dose–response curves are very similar to the curves shown in Figure 3 , which were performed using relatively undifferentiated ARPE-19 cells plated at lower cell densities. The EC₅₀s of luteolin, quercetin, fisetin, and eriodictyol in confluent differentiated cells reflect the similarities in the dose–response curves (Table 2) . 

Protection of ARPE-19 Cells after Exposure to Oxidative Stress

From a clinical perspective, having a compound that could protect against oxidative stress over an extended period in the presence of an oxidant would be highly desirable. To determine whether any of the flavonoids are able to protect ARPE-19 cells that have been exposed to oxidative stress but have not yet died, we performed delayed-response time course studies. These studies were performed by exposing RPE cells to oxidants and waiting for progressively longer periods before adding the flavonoids. Figure 9shows the percentages of surviving ARPE-19 cells that were treated with t-BOOH and H₂O₂ for various times before the addition of three of the most effective flavonoids: luteolin, quercetin, and fisetin. We compared the cell viability in the wells treated with flavonoids and t-BOOH or H₂O₂ with the wells in which the oxidants were removed at various times by replacing the culture medium. These experiments show that an increasing percentage of ARPE-19 cells are fully committed to cell death at various times after exposure to H₂O₂ and t-BOOH. Approximately 85% of the ARPE-19 cells were fully committed to death approximately 2 hours after exposure to H₂O₂ and approximately 4 hours after exposure to t-BOOH. During the interval after the addition of the oxidant but before the cells were committed to death, the addition of luteolin, quercetin, or fisetin protected the cells as effectively as if the oxidants were removed altogether. More than 75% of the ARPE-19 cells were still protected from cell death even 2 hours after exposure to t-BOOH. In the more rapidly progressing H₂O₂ assay, whereas little or no rescue was seen at 4 hours when the oxidant was removed from the cells, some protection was still observed at 4 hours in the presence of the flavonoids. 

Flavonoids and Intracellular ROS Levels

To determine whether the cytoprotection observed with flavonoid treatment is a consequence of an intracellular effect or whether it is simply due to ROS scavenging activity in the extracellular media, we examined whether flavonoid treatment could reduce intracellular ROS levels. In the design of these experiments, the cells were first stressed with t-BOOH, and then the t-BOOH was removed from the wells, and the flavonoids were added for 1 hour before intracellular ROS levels were measured. In this way, any reduction in ROS levels could be attributed only to an intracellular effect, as there is no extracellular contact between the flavonoid and the oxidant. We found that treatment with t-BOOH led to an increase in intracellular ROS levels which was significantly reduced by all the effective flavonoids that we tested, including luteolin, quercetin, fisetin, and 3,6-DHF (Fig. 10) . Luteolin, quercetin, and fisetin reduced intracellular ROS levels by 70% whereas 3,6-DHF reduced intracellular ROS levels by approximately 45%. 

Protection of Primary Human RPE Cells from Oxidative-Stress–Induced Cell Death

To confirm that the protective effects seen with the ARPE-19 cells are also present in primary cultures, we examined the ability of flavonoids to protect primary cultures of human RPE cells from oxidative stress. In these studies, 85% of the primary human RPE cells plated at low density (3200 cells/well) died with 180 to 300 μM t-BOOH and 850 to 1000 μM H₂O₂. We found that more flavonoids were effective at protecting the primary cultures of human RPE cells from t-BOOH- and H₂O₂- induced cell death than the ARPE-19 cell line (Table 3) . Several flavonoids that had no effect in the ARPE-19 cell line, such as myricetin, kaempferol, taxifolin, and EGCG, were protective in the primary cells. In addition, the potencies of most of the effective flavonoids in the primary cells were greater than those in the ARPE-19 cell line. Of note, luteolin and fisetin had the highest potencies, with EC₅₀s of 2 and 3 μM. Indeed, the dose–response curves indicate that some flavonoids were effective in the nM range in the primary cultures of human ARPE cells (Fig. 11) . 

Stimulate of Expression of Nrf2 and HO-1 in ARPE-19 Cells

The induction of phase 2 genes is an important defense mechanism for protecting cells from oxidative stress. ¹⁶ To determine whether some flavonoids could have additional, long-term effects on the ability of RPE cells to resist oxidative stress, we examined the ability of flavonoids to induce the expression of phase-2 proteins in human ARPE-19 cells by blotting for the transcription factor Nrf2 and one of its downstream gene products, HO-1. We found that several of the protective flavonoids induced the expression of Nrf2 and HO-1 (Fig. 12) . Fisetin and quercetin induced Nrf2 and HO-1 but luteolin, which was one of the more potent flavonoids in the oxidative stress assays, did not induce either protein. As expected, the induction of Nrf2 and HO-1 was coordinated, so that all the flavonoids that induced the expression of Nrf2 also induced HO-1 expression (Table 4) . Several flavonoids that were not protective under our conditions also induced Nrf2 and HO-1, suggesting that these compounds may provide some protection from oxidative stress toxicity with long-term treatment. 

Curiously, the expression of HO-1 in adult RPE cells was markedly more robust after exposure to eriodictyol compared with quercetin and fisetin, at the same time that the expression of Nrf2 was similar among all three. The reason for this is not completely clear but may be related to activation of other pathways by eriodictyol, which can induce HO-1 expression. ^{31 32}  

Discussion

Reinforcing the cellular defenses that protect the retina and RPE against oxidative stress has been proposed to be a viable option for reducing the progression of macular degeneration in patients with early signs of the condition. Previous studies have shown that flavonoids can protect nerve cells from oxidative-stress–induced death. ^{22 33} In our study, flavonoids also protected human RPE cells from oxidative-stress–induced death with good efficacy, high potency, and low toxicity. The effective flavonoids include the dietary flavonoids fisetin, luteolin, quercetin, eriodictyol, baicalein, galangin, and EGCG, and the synthetic flavonoids, 3,6-dihydroxy flavonol, and 3,7-dihydroxy flavonol. To a lesser degree, there is also some protection from the dietary flavonoids myricetin, taxifolin, and kaempferol. 

We used t-BOOH and H₂O₂ to induce oxidative stress for several reasons. First, these oxidants have been used by other investigators and their effects on RPE cells have been well characterized. ^{24 25 26 27} Second, assays with these agents can be performed efficiently in 96-well microtiter plates, which permits the screening of a large number, of compounds in a rapid manner. Third, these oxidants induce a high degree of oxidative stress. Compounds that protect under these conditions have the potential to be even more effective under milder, more physiologic conditions of stress. Fourth, these oxidants have different modes of action. t-BOOH induces lipid peroxidation, a self-propagating form of oxidative injury that damages cell membranes and is a particular risk to the RPE and the lipid-rich photoreceptor cells. H₂O₂ is a direct oxidant that is formed by the RPE cell under normal physiologic conditions and must be neutralized on a continual basis. ^{3 34}  

Although several flavonoids were protective in this study, several others were not. For example, catechin, epicatechin, cyanidin, and genistein did not protect against t-BOOH- or H₂O₂-induced oxidative stress. To gain some insight into the structure–activity relationships between the different compounds, we compared the structures of the effective and ineffective flavonoids. We found that only minor structural differences created large differences in efficacy. For example, the only difference between luteolin, which is very effective, and apigenin, which is ineffective, is a single hydroxyl group in the R1 position of the B ring. We analyzed these features further and identified a few useful indicators of efficacy. The most effective classes of flavonoids are the flavones and flavonols. Many of these have structural features that result in good antioxidant activity—that is, the hydroxyl group in the C3 position, the catechol structure, 3′4′ dihydroxy in the B ring and the unsaturation of the C ring. ^{35 36} Quercetin, fisetin, and luteolin all share these features. In addition, we observed that many of the most effective compounds are hydrophobic, indicating that they should readily pass through cell membranes and accumulate intracellularly. Ishige et al. ³³ reported that differences in hydrophobicity was a key factor in explaining the discrepancy between flavonoids with high TEAC (trolox equivalent activity concentration) values that protect nerve cells from oxidative stress and those which fail to protect. (The antioxidant potential is defined by the TEAC value, which is derived by measuring the ability of a given substance in an aqueous solution to neutralize the radical cation of ABTS. The values are calculated in reference to 1 mM Trolox, a water-soluble vitamin E analogue.) 

Some of the flavonoids that had excellent efficacy did not share the structural features that are associated with good antioxidant activity. For example, 3,6-dihydroxyflavonol and 3,7-dihydroxyflavonol do not have as much antioxidant activity as other flavonoids. Nonetheless, they were quite effective in our assays. The flavanones are also not very effective antioxidants, but eriodictyol was very protective in our assays. We believe that these findings can be explained by the fact that flavonoids function through a variety of different mechanisms, some of which can activate specific cellular signaling pathways that protect cells from oxidative stress. ¹⁵ For example, quercetin and fisetin have been shown to increase the intracellular levels of glutathione, the major intracellular antioxidant, ³³ as well as acting themselves as intracellular antioxidants. In contrast, the flavonols 3,6-dihydroxyflavonol and 3,7-dihydroxyflavonol can block Ca⁺² influx in nerve cells, which is one of the last steps in the cell death cascade. ³³ Thus, the differences between the potency and efficacy of specific flavonoids may be due to a variety of factors that are dependent on very small differences in their chemical structures. 

We performed our initial studies with low-density cultures of ARPE-19 cells due to their high sensitivity to oxidative stress and their ability to serve as a high-throughput screening cell line for compounds that protect from oxidative stress. In establishing the conditions for these assays, we found that the concentrations of oxidants necessary to kill the cells were directly proportional to the plating cell density. Our findings are consistent with those of other investigators who reported a similar density-dependent cytotoxicity of H₂O₂ in primary RPE and ARPE-19 cells. ^{28 37 38} These researchers suggested that it is actually the amount of oxidant per cell rather than the concentration that is important in oxidative toxicity. ³⁸ Thus, consistent with our results, denser cultures required higher concentrations of oxidants to achieve the same amount of oxidant per cell. ³⁸ In our study as well as theirs, there was a narrow window between the concentrations of t-BOOH and H₂O₂ that killed 50% of the cells and the concentrations that killed 100% of the cells. Therefore, we found it necessary to perform a dose–response curve on each set of 96-well plates before each cytoprotection experiment. This allowed us to standardize our cell death assays at 80% and 95% death so as to be able to compare the results from experiment to experiment. 

We were not able to detect any differences in the cellular responses to flavonoid treatment after oxidative stress between either low-density or high-density cultures or undifferentiated and well-differentiated cells. After numerous experiments performed at least three times each with both early- and late-passage ARPE-19 cells in a broad range of densities, we found that the cytoprotective efficacy and potency of four of the most effective flavonoids—quercetin, luteolin, fisetin, and eriodictyol—was independent of cellular differentiation and cell density. Although changes in cell density over a 16- to 24-hour period did affect the concentration of oxidants necessary to kill the cells, our controls demonstrated that changes in cell density did not affect the outcome over the 2- to 4-hour time window during which the cells become committed to cell death in these assays. 

Selected flavonoids protect RPE cells from death after the cells are exposed to oxidative stress but before the cells are committed to die. This protection is as effective as removing the oxidants from the cells. From a clinical perspective, this finding is interesting, because patients frequently present after the initial signs of damage have appeared but before there has been complete loss of RPE cells. Conceptually, a treatment that is as effective as removing the oxidant insult altogether may be quite useful for reducing the progression of disease in patients, since, unlike in cell cultures, it is not possible in vivo to remove the insult directly. 

Although we did not attempt to examine mechanistic questions in this article, we were able to rule out the possibility that the flavonoids were acting solely by neutralizing the oxidants in the extracellular media. We designed our ROS experiments to avoid any extracellular contact between the flavonoid and the oxidant. We added the oxidant first, removed the oxidant and then added the flavonoids for a short incubation. Using this approach, we were able to avoid any possibility that the flavonoids were sticking to the well or the cell surface and neutralizing the oxidant extracellularly. These studies demonstrated that quercetin, fisetin, luteolin, and eriodictyol can reduce intracellular ROS, compared with controls, even 2 hours after the addition of the oxidant. The ability of these flavonoids to reduce accumulated intracellular ROS demonstrates that the protective effects of flavonoids are not simply limited to their ability to scavenge ROS in the extracellular media and are likely to play a role in the protection seen in the delayed-response studies (Fig. 9) . Future studies will be focused on examining the mechanisms of action of the effective flavonoids in more detail. 

ARPE-19 cells were used in our screening assay to look for compounds that protect from oxidative stress. This cell line allowed us to screen a large number of compounds in an efficient and rapid manner and to validate our most interesting findings in human primary RPE cells as a second step. We observed that many of the flavonoids were more effective at protecting primary human RPE cells compared with the transformed ARPE-19 cells. Indeed, the EC₅₀s of luteolin and fisetin in the human primary RPE cell assays were in the 2- to 5-μM range and the dose–response profiles indicated that these two flavonoids had activity in the high nanomolar range. Taken together, these findings validate the use of the transformed ARPE-19 cells for the rapid screening of potential protective compounds. 

Another important finding in this study is that selected flavonoids induce the expression of phase-2 detoxification enzymes in human RPE cells. This is a useful property, because phase-2 proteins enhance the cell’s natural defenses against oxidative stress. ¹⁶ Recently, other investigators have demonstrated that the activation of the phase 2 response genes can protect human RPE cells from oxidative-stress–induced death. ^{17 18 19 39} It is unlikely that the protection seen in our study can be attributed to activation of the phase-2 enzymes due to the time frame necessary to accumulate significant levels of these proteins in cells. Nevertheless, the ability of specific flavonoids to induce these proteins is likely to make them even more beneficial for protecting RPE cells in vivo from oxidative stress. We will be examining this possibility further in future studies. 

In this study, we observed that specific flavonoids were significantly more effective than either vitamins C or E in protecting human RPE cells from oxidative stress in vitro. High doses of vitamin E (100 μM) protected 50% or less of the cells in the t-BOOH and the H₂O₂ assays, whereas lower doses of quercetin protected >90% of cells. Vitamin C protected most of the RPE cells from t-BOOH-induced cell death, but it was completely ineffective at protecting cells from H₂O₂-induced toxicity. Some insight into these findings can be gained by examining the stoichiometry of the reactions and the thermodynamic properties of vitamins C and E. Overall, vitamin E is known to be a better antioxidant than vitamin C. ¹⁴ It has a higher standard reduction potential, ¹⁴ consistent with its ability to provide partial protection for cells in both the t-BOOH and H₂O₂ assays. Vitamin C has less favorable thermodynamic properties for neutralizing oxidants. ¹⁴ Although it can “theoretically” reduce H₂O₂, the thermodynamics of this reaction are not favorable. ¹⁴ In contrast, the reaction between vitamin C and t-BOOH is thermodynamically more favorable, which is consistent with our findings. Although vitamins E and C did not protect RPE cells as well as the flavonoids in our assays, it is important not to conclude that quercetin or the other flavonoids are necessarily more beneficial antioxidants in vivo than either vitamins E or C. Nor is it fair to translate these findings into clinical recommendations. The health benefits of vitamins E and C are well known. They are both well absorbed and widely distributed throughout the body. They have minimal to low toxicity, even at high doses. In comparison, there are many complex issues relating to the in vivo bioabsorption, metabolism, and distribution of quercetin and other flavonoids that still should be carefully evaluated. 

Nevertheless, the results raise the question of whether flavonoids may be among the group of natural compounds that contribute to the ocular benefits associated with the dietary products identified in the epidemiologic studies of Seddon et al. ⁹ and others. ²⁹ These investigators showed a strong inverse relationship between the incidence of macular degeneration and the intake of certain foods and beverages, such as spinach, dark leafy greens, and wine. ^{9 29} Spinach and other dark leafy greens are known to contain multiple, important biological antioxidants, including vitamins C and E, glutathione, lycopene, lutein, and zeaxanthin. In addition, they contain quercetin and luteolin, which were two of the most potent and efficacious flavonoids in our study. ¹¹ Quercetin is also found in red wine. ¹¹ Although it is unknown at this time whether any of these flavonoids contribute to the clinical benefits seen in the epidemiologic studies, we believe that these associations are worth noting. As indicated, although there are many questions regarding flavonoids that remain to be investigated, it is reasonable to suggest that dietary products that contain the most effective flavonoids should be considered for their nutritional value in the diets of patients with ocular diseases. 

The U.S. Department of Agriculture (USDA) has compiled a database that provides a partial list of the flavonoid contents of selected foods (http://www.nal.usda.gov/fnic/foodcomp/data/flav/flav.html). ¹¹ Quercetin is listed at high concentrations in yellow onions, capers, ancho peppers, cranberries, fennel, cocoa, currants, buckwheat, black tea, spinach, and wild greens. It is also present in bilberries, blueberries, broccoli, tomato puree, apples, and yellow and green beans. Luteolin is present in spinach and wild greens, hot peppers, peppermint, parsley, rosemary, and thyme. Eriodictyol is found in peppermint and a variety of juices, including lemon juice, sour orange juice, and lime juice. Fisetin is not included in the USDA database, but it has been reported at high concentrations in strawberries and at lower concentrations in persimmons, tomatoes, onion, kiwi, oranges, apples, peaches, and grapes. ⁴⁰  

Our results are consistent with other recent reports that have shown that specific flavonoids can protect primary cultures of cortical neurons from oxidative-stress–induced cell death resulting from glutamate toxicity, hypoglycemia, buthionine sulfoximine (BSO) treatment and H₂O₂ exposure. ³³ We have also observed that retinal ganglion cells can be protected from cell death when exposed to flavonoids either before or after an oxidative insult. ²² These findings are noteworthy and suggest that specific flavonoids may be capable of protecting both RPE cells and retinal neurons from oxidative injuries. It also opens up the possibility that these or related compounds could be beneficial for the treatment of multiple retinal disorders associated with oxidative-stress–induced injuries. 

²  Present affiliation: The Salk Institute, La Jolla, California.

 Supported by joint Mericos/Scripps Research Institute (TSRI) Neurobiology and Vision Science Research Program, and National Eye Institute Core Grant for Vision Research P30EY012598, and National Center for Neurological Disorders and Stroke Grant NS28121 (PM).

 Submitted for publication November 23, 2004; revised March 15 and December 5, 2005; accepted May 1, 2006.

 Disclosure: A. Hanneken, None; F.-F. Lin, None; J. Johnson, None; P. Maher, None

 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.

 

Structures of the six major classes of flavonoids.

Figure 2.

 

Protection of ARPE-19 cells from t-BOOH- or H₂O₂-induced toxicity by luteolin. ARPE-19 cells were plated at 3200 cells/well in 35-mm plates and maintained in culture for 18 hours. The media were replaced, and the cells were untreated (A), treated with luteolin alone (B), treated with 120 μM t-BOOH (C) or 250 μM H₂O₂ (E) alone, or pretreated with luteolin followed by 120 μM t-BOOH (D) or 250 μM H₂O₂ (F). The cells were photographed 24 hours later.

Figure 3.

 

Dose–response curves of selected flavonoids on ARPE-19 cell survival in the presence of t-BOOH or H₂O₂. Cells were preincubated with luteolin (A), quercetin (B), fisetin (C), eriodictyol (D), 3,6-dihydroxyflavone (E), 3,7-dihydroxyflavone (F), baicalein (G) and galangin (H) for 1 hour and exposed to 120 μM t-BOOH or 250 μM H₂O₂ treatment. Data represent the percentage of ARPE-19 cells protected by flavonoids compared with control cells that were treated identically without t-BOOH or H₂O₂ (mean ± SEM, n = 3 to 5 experiments).

Table 1.

 

The Potency of Various Flavonoids in Protecting ARPE-19 Cells from Oxidative Stress-Induced Cell Death

Flavonoid	Free Hydroxyl Positions	Common Name	EC50 (μM)			LD50 (μM)
							t-BOOH	H2O2
Flavone	4′,5,7	Apigenin	No	No	<50
	5,6,7	Baicalein	14 ± 1	No	84
	3′,4′,5,7	Luteolin	14 ± 1	9 ± 1	104
Flavonol	3,6		27 ± 1	18 ± 1	78
	3,7		21 ± 1	22 ± 1	40
	3,5,7	Galangin	32 ± 1	31 ± 1	112
	3,3′,4′,7	Fisetin	15 ± 1	11 ± 1	101
	3,4′,5,7	Kaempferol	No	No	∼50
	3,3′,4′,5,7	Quercetin	18 ± 1	19 ± 1	230
	3,3′methoxy,4′,5,7	Isorhamnetin	No	No	>50
	3,3′,4′,5,5′,7	Myricetin	No	No	>50
Isoflavone	4′,5,7	Genistein	No	No	>>50
Flavanone	4′,5,7	Naringenin	No	No	>>50
	3′,4′,5,7	Eriodictyol	6 ± 1	17 ± 1	153
	3,3′,4′,5,7	Taxifolin	No	No	>100
Flavanol	3,3′,4′,5,7	Catechin	No	No	>50
	3,3′,4′,5,7	Epicatechin	No	No	>50
		Epigallocatechin-3-gallate	No	No	>50
Anthocyanidin	3,3′,4′,5,7	Cyanidin	No	No	>100
	3,3′methoxy,4′,5,7	Peonidin	No	No	>>50
	3,3′5′dimethoxy,4′,5,7	Malvidin	No	No	>>50

 Half maximal effective concentrations (EC₅₀) were determined by exposing ARPE-19 cells to t-BOOH or H₂O₂ in the presence of flavonoids. The doses of the compounds that cause 50% of cell death (LD₅₀) were measured after exposure to flavonoids alone. Data represent the average results of two to five independent experiments. Each study was conducted with two to four replicates. “No” indicates that there was little or no efficacy at doses up to 50 μM.

Figure 4.

 

Comparison of the efficacy of quercetin, vitamin C and vitamin E. ARPE-19 cells were incubated with 50 μM quercetin or 100 μM vitamin E and vitamin C, then were subjected to treatment with 120 μM t-BOOH or 250 μM H₂O₂. Data represent the percentage of ARPE-19 cells protected by flavonoids compared with control cells that were treated identically without t-BOOH or H₂O₂ (mean ± SEM, n = 3 to 4 experiments). * and **P < 0.001 compared with quercetin in t-BOOH or H₂O₂ oxidative stress, respectively, by ANOVA followed by the Newman-Keuls test.

Figure 5.

 

Morphology of undifferentiated and differentiated human adult RPE19 cells. (A) Undifferentiated and (B) differentiated human adult RPE-19 cells.

Figure 6.

 

RT-PCR detection of RPE65 and FGFR1 mRNA in undifferentiated and differentiated ARPE-19 cells. Total RNA from undifferentiated or 14-day and 76-day differentiated ARPE-19 cells was reverse transcribed and amplified using primers for RPE65 and FGFR1. A reaction omitting the cDNA was used as a negative control.

Figure 7.

 

Cytotoxic concentrations of t-BOOH and H₂O₂ are proportional to ARPE-19 cell density. Various densities of differentiated ARPE-19 cells were plated in 96-well plates and grown for 24 hours. The media was replaced and the cells were treated with a range of different concentrations of (A) t-BOOH or (B) H₂O₂ to determine the concentration that induced 80% to 95% cell death after an overnight incubation. The cytotoxicity was determined using the MTT assay.

Figure 8.

 

Dose–response curves of selected flavonoids on ARPE-19 cell survival in the presence of t-BOOH or H₂O₂. Confluent, differentiated ARPE-19 cells were preincubated with luteolin, quercetin, fisetin, and eriodictyol for 1 hour and then exposed to 500 μM t-BOOH. Data represent the percentage of ARPE-19 cells protected by flavonoids compared with control cells that were treated identically without t-BOOH (mean ± SEM, n = 3 experiments).

Table 2.

 

Comparison of the Potency of Flavonoids in Differentiated and Undifferentiated Cultures of ARPE-19 Cells

Flavonoid	EC50
		Differentiated	Undifferentiated
Eriodictyol	8 ± 1	6 ± 1
Luteolin	12 ± 1	14 ± 1
Quercetin	17 ± 1	19 ± 1
Fisetin	26 ± 1	15 ± 1

Figure 9.

 

Time course of the delayed rescue effect of flavonoids on ARPE-19 cells. ARPE-19 cells were plated into 96-well microtiter plates at 3200 cells/well and exposed to (A) t-BOOH or (B) H₂O₂ the next day. Twenty-four hour exposure to either t-BOOH or H₂O₂ resulted in less than 10% survival. Luteolin, quercetin, or fisetin at concentrations of 50 μM were added either before (−1 hour), at the same time (0), or at various times after the addition of t-BOOH or H₂O₂, and cell survival was assessed by the MTT assay 24 hours later. In one set of wells, the t-BOOH or H₂O₂ oxidants were removed at various times and replaced with normal media. Arrows: indicate the time that t-BOOH or H₂O₂ were added to the cells. The percentage of surviving ARPE-19 cells was determined 24 hours later (mean ± SEM, n =3 to 5 experiments).

Figure 10.

 

Fluorescence detection of intracellular ROS in ARPE-19 cells. Cells were preincubated with H₂DCFDA, exposed to t-BOOH for 2 hours, and treated with quercetin, 3,6- dihydroxyflavone, fisetin, or luteolin for 1 hour. Intracellular ROS production was measured on a spectrofluorometer. Data represent the relative ROS levels in ARPE-19 cells exposed to flavonoids (30 μM) and t-BOOH (500 μM) compared with the control cells exposed to t-BOOH alone (mean ± SEM, n =3 experiments).

Table 3.

 

The Potency of Various Flavonoids in Protecting Primary Human RPE159 Cells from Oxidative Stress-Induced Cell Death

Flavonoid	Free Hydroxyl Positions	Common Name	EC50 (μM)			LD50 (μM)
							t-BOOH	H2O2
Flavone	5,6,7	Baicalein	8 ± 1	21 ± 1	>>100
	3′,4′,5,7	Luteolin	2 ± 1	3 ± 1	>50
Flavonol	3,6		7 ± 1	11 ± 1	>>50
	3,7		9 ± 1	8 ± 1	27
	3,5,7	Galangin	26 ± 1	61 ± 1	70
	3,3′,4′,7	Fisetin	3 ± 1	5 ± 1	>50
	3,4′,5,7	Kaempferol	∼50	No	∼50
	3,3′,4′,5,7	Quercetin	6 ± 1	11 ± 2	>50
	3,3′methoxy,4′,5,7	Isorhamnetin	>>50	No	>50
	3,3′,4′,5,5′,7	Myricetin	>50	No	>>50
Isoflavone	4′,5,7	Genistein	No	No	>50
Flavanone	4′,5,7	Naringenin	No	No	>>50
	3′,4′,5,7	Eriodictyol	7 ± 1	11 ± 1	>100
	3,3′,4′,5,7	Taxifolin	>50	No	>>50
Flavanol	3,3′,4′,5,7	Catechin	No	No	>>50
	3,3′,4′,5,7	Epicatechin	No	No	>>50
		Epigallocatechin-3-gallate	22 ± 1	30 ± 2	>100
Anthocyanidin	3,3′,4,4′,5,7	Cyanidin	No	No	>>50

 Half maximal effective concentrations (EC₅₀) were determined by exposing ARPE159 cells to t-BOOH or H₂O₂ in the presence of flavonoids. The doses of the compounds that cause 50% of cell death (LD₅₀) were measured after exposure to flavonoids alone. Data represent the average results of two to three independent experiments. Each study was conducted with two to four replicates. “No” indicates that there was little or no efficacy at doses up to 50 μM.

Figure 11.

 

Dose–response curves of flavonoids on human primary RPE cell survival in the presence of t-BOOH or H₂O₂. Cells were plated at 3200 cells/well, preincubated with luteolin (A), quercetin (B), fisetin (C), eriodictyol (D), 3,6-dihydroxyflavone (E), 3,7-dihydroxyflavone (F), galangin (G), and baicalein (H) for 1 hour and exposed to 180 to 300 μM t-BOOH or 850 to 1000 μM H₂O₂. Data represent the percentage of primary human RPE159 cells protected by flavonoids compared with control cells that were treated identically without t-BOOH or H₂O₂ (mean ± SEM, n =3 experiments).

Figure 12.

 

Some flavonoids induce Nrf2 and HO-1 protein expression in ARPE-19 cells. ARPE cells were untreated or treated for 2 to 4 hours to induce Nrf2 expression or 24 hours to induce HO-1 expression with 50 μM luteolin, quercetin, fisetin, eriodictyol, 3,6-dihydroxyflavone, or 3,7-dihydroxyflavone. Immunoblots were performed with anti-Nrf2 antibody; anti-HO-1 antibody, and anti-β-actin antibody as a loading control for the anti-HO-1 blot.

Table 4.

 

The Expression of Nrf2 and HO-1 in ARPE-19 Cells Induced by Various Flavonoids

Flavonoid	Nrf2	HO-1
3,6-dihydroxyflavone	−	−
3,7-dihydroxyflavone	−	−
Luteolin	−	−
Fisetin	+	+
Myricetin	+	+
Apigenin	−	−
Quercetin	+	+
Eriodictyol	+	+
Taxifolin	+	+
Epicatechin	+	+
Epigallocatechin-3-gallate	+	+

 APRE19 cells were treated with 50 μM of each of the flavonoids for 2 or 4 hours and then nuclear proteins were isolated and analyzed for Nrf2 expression. For HO-1 expression, the cells were stimulated with 50 μM each of the flavonoids for 24 hours. Cells were lysed and cytoplasmic proteins were assayed for HO-1 expression.

The authors thank Brian McCay, PhD, for his gift of primary human RPE159 cells and the Mericos Eye Institute at Scripps Memorial Hospital for ongoing support.