The Flavonoid, Eriodictyol, Induces Long-term Protection in ARPE-19 Cells through Its Effects on Nrf2 Activation and Phase 2 Gene Expression

Abstract

purpose. Eriodictyol, a flavonoid found in citrus fruits, is among the most potent compounds reported to protect human RPE cells from oxidative stress–induced cell death. The present study sought to determine whether eriodictyol-induced phase 2 protein expression further enhances the resistance of human ARPE-19 cells to oxidative stress.

methods. The ability of eriodictyol to activate Nrf2 and to induce the phase 2 proteins heme-oxygenase (HO)-1 and NAD(P)H:quinone oxidoreductase (NQO)-1, and the cellular antioxidant glutathione (GSH) were analyzed. Cytoprotection assays in ARPE-19 cells that were overexpressing HO-1 or NQO-1 were performed, cell survival after short-term and long-term eriodictyol treatment was compared, and the mechanism of protection using a dominant negative Nrf2 and shRNA specific for HO-1 was tested.

results. Eriodictyol induced the nuclear translocation of Nrf2, enhanced the expression of HO-1 and NQO-1, and increased the levels of intracellular glutathione. ARPE-19 cells that overexpress HO-1 or NQO-1 were more resistant to oxidative stress–induced cell death than control cells. Eriodictyol induced long-term protection significantly greater than its short-term protection. This effect was correlated temporally with the activation of Nrf2 and the induction of phase 2 enzymes and could be blocked with the use of a dominant negative Nrf2 and shRNA specific to HO-1.

conclusions. These findings indicate that the greatest benefit from eriodictyol may be its ability to regulate gene expression and enhance multiple cellular defenses to oxidative injury.

Vitamins, antioxidants, carotenoids, and zinc are considered valuable dietary supplements for preventing vision loss in patients with age-related macular degeneration (ARMD). The Age-Related Eye Disease Study (AREDS) showed that antioxidants and mineral supplements can preserve vision in patients with macular degeneration. ¹ Clinical practice guidelines now emphasize the importance of maintaining a high dietary intake of fruits, vegetables, antioxidants, and vitamins to block the progression of macular degeneration. 

Flavonoids are structurally heterogeneous, polyphenolic compounds regularly consumed in the human diet and are present at high concentrations in fruits, vegetables, and other plant-derived foods, such as teas and other beverages. ² Many of the health benefits associated with the Mediterranean diet have been attributed to flavonoids, including protection from cardiovascular disease and cancer. ³ Evolutionary studies suggest that flavonoids have evolved to protect plant tissue from chronic exposure to ultraviolet light, and epidemiologic studies have identified flavonoids in many foods associated with reduced risk for advanced macular degeneration. ^{4 5 6} Flavonoids accumulate in the mammalian eye, ^{7 8 9} and anecdotal cases report improved visual function after the consumption of bilberry fruits that contain high concentrations of anthocyanins (for a review, see Ref. ¹⁰ ). As a result, there is considerable interest in understanding the potential role and benefit of flavonoids in ocular health and disease prevention. 

Flavonoids can provide short- and long-term protection against oxidative stress through a variety of mechanisms. They are potent antioxidants that neutralize toxic reactive oxygen species by donating hydrogen ions. ¹¹ Yet, potentially even more important, flavonoids can modulate multiple cell signaling pathways and induce the expression of phase 2 proteins, ¹² which protect against oxidative stress by catalyzing a wide variety of reactions that neutralize reactive oxygen species, toxic electrophiles, and carcinogens. Phase 2 proteins are regulated by the transcription factor NF-E2–related factor 2 (Nrf2), which binds to the antioxidant-response element (ARE), a cis-acting enhancer sequence in the promoter of phase 2 proteins, resulting in the activation of gene transcription. ¹³ Numerous in scope, phase 2 proteins are regulated as a group in a coordinated manner through the Nrf2/ARE pathway and include enzymes such as γ-glutamyl cysteine ligase and glutathione synthetase, which regulate the major components in glutathione biosynthesis; heme-oxygenase (HO)-1, which catalyzes the breakdown of heme proteins into iron, the vasodilator carbon monoxide, and biliverdin (which is further reduced to the antioxidant bilirubin); NAD(P)H:quinone oxidoreductase (NQO)-1, a reducing agent that plays a role in antioxidant defenses with the cofactor NADH or NADPH; thioredoxin, a key cellular antioxidant; and others. ^{14 15 16} The phase 2 proteins HO-1 and NQO-1 are expressed in the retina and the retinal pigment epithelium and are upregulated in response to ultraviolet light and inflammation. ^{17 18} Because of the coordinated expression of phase 2 enzymes, HO-1 and NQO-1 are useful markers of retinal and retinal pigment epithelial (RPE) phase 2 protein expression. ¹⁴  

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, which 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 to oxidative injury that will function long after the direct antioxidants are consumed. Phase 2 proteins are induced by a variety of compounds that use phase 2 enzymes for detoxification and degradation, including pro-oxidant xenobiotics, chemically reactive carcinogens, chemopreventive agents (such as sulforaphane and oltipraz), antioxidants, and specific bioflavonoids. ^{14 19 20}  

In a recent study, we showed that specific dietary flavonoids could protect human primary and ARPE-19 cells from oxidative stress–induced cell death when added simultaneously with, or within 1 hour before, oxidant exposure. ²¹ These experiments analyzed the effects of short-term treatment with flavonoids on the resistance of ARPE-19 cells to oxidative stress. After this study, the question arose as to whether flavonoid-induced phase 2 protein expression could provide a significant additional benefit to RPE cells by allowing them to resist higher levels of oxidative stress than they could resist with short-term treatment. Thus, as a follow-up to our initial study, we pretreated RPE cells with flavonoids for several hours to induce phase 2 protein expression, and we compared the outcome of this treatment with that seen after the addition of the flavonoid at the time of oxidant exposure. We focused on the flavonoid eriodictyol because we had previously shown that it had the highest potency of all flavonoids tested in our oxidative stress assay, and it strongly induced the activation of Nrf2 and heme-oxygenase. ²¹ The goal of the present study was to address several questions: Does eriodictyol lead to time-dependent and dose-dependent increases in the expression of Nrf2 and the phase 2 proteins HO-1 and NQO-1? Does the expression of these enzymes protect ARPE-19 cells from oxidative stress, and does the knockdown of one of these proteins abate this protection? Is there a correlation between eriodictyol-induced phase 2 protein expression and eriodictyol-induced long-term protection of ARPE-19 cells? Is the long-term protection greater than the short-term protection induced as a result of the direct antioxidant properties of eriodictyol? Is the long-term protection mediated through the Nrf2/ARE pathway, and can it be blocked by a dominant negative to Nrf2? The answers to these questions show that the induction of phase 2 protein expression by eriodictyol significantly enhances the resistance of RPE cells to oxidative injury and provides a mechanism whereby flavonoids can augment their activity as direct antioxidants and induce more sustained resistance to oxidative damage over the long term. 

Methods

Reagents

Eriodictyol, 3-(4,5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide (MTT), dimethyl sulfoxide, and anti–β-actin monoclonal antibody (AC-15) were purchased from Sigma-Aldrich (St. Louis, MO). Anti–Nrf2 (H-300) rabbit polyclonal antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), and anti–NQO-1 mouse monoclonal and anti–HO-1 rabbit polyclonal antibodies were purchased from Assay Designs (Ann Arbor, MI). Horseradish peroxidase–linked goat anti–rabbit and anti–mouse IgG (H+L) antibodies were purchased from Bio-Rad (Hercules, CA). All antibodies were used at the dilutions recommended by the manufacturer. 

Plasmid Constructs

The dominant negative mutant of Nrf2, Nrf2M, was generated by Jawed Alam (Alton Ochsner Medical Foundation, New Orleans, LA) ²² and was supplied to us by Andy Y. Shih (University of California, San Diego, CA). Nrf2M, lacking the N-terminal transcriptional activation domain, was generated by deleting amino acid residues 1–392. It sequesters Nrf2 dimerization protein(s), competes for Nrf2 DNA binding domains, and blocks the induction of HO-1. ²² The human wild-type NQO-1 expression plasmid (pcDNA-hNQO-1) was a kind gift from David Ross (School of Pharmacy, Department of Pharmaceutical Sciences, University of Colorado Health Sciences Center, Denver, CO). The empty control plasmid pcDNA3.1(−) (CMV3) was purchased from Invitrogen Life Technologies (San Diego, CA). Human heme oxygenase (HMOX)-1 expression plasmid, shRNA HO-1 plasmid, and their control plasmids were purchased from OriGene Technologies (Rockville, MD). 

Cell Culture

Human adult retinal pigment epithelial (ARPE)-19 cells were a gift from Larry Hjelmeland (University of California, Davis, CA) and were grown in DMEM/F12 medium (Invitrogen) containing 10% fetal bovine serum (Hyclone, Logan, UT), 15 mM N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid (HEPES), 2.5 mM l-glutamine, 0.5 mM sodium pyruvate, and 20 mM sodium bicarbonate. As previously reported, ²¹ the ARPE-19 cells used in these experiments grow with differentiated morphology, express cellular markers of differentiation, including RPE65, and behave similarly to primary human RPE cultures. Cells were dissociated from the culture dishes with the use of versene (0.53 mM EDTA [EDTA] in Hanks balanced salt solution) followed by 0.25% trypsin/EDTA. 

Cytotoxicity Assay

ARPE-19 cells were seeded onto 96-well plates at 20,000 cells/well, grown for 24 hours, replenished with fresh culture media containing 10% dialyzed FBS (Hyclone, Logan, UT), and preincubated with eriodictyol for 1 to 24 hours before the addition of the chemical oxidants. t-BOOH was added at concentrations that had been found to kill more than 80% to 90% of the cultured cells in dose-response assays. In a previous study in which we showed that the concentrations of t-BOOH, which are necessary to kill ARPE-19 cells, are density dependent but not differentiation dependent, further details of this assay can be found. ²¹ After overnight incubation, cell viability was determined by a modified version of the MTT assay. ²³ The MTT assay was performed by removing the cell culture medium and replacing it with 100 μL fresh culture medium containing 0.5 μg/μL MTT. After 4 hours of incubation at 37°C, cells were solubilized overnight with 100 μL solution containing 50% dimethyl formamide and 20% SDS (pH 4.7). Absorbance at 560 nm was measured with a microplate reader (Spectromax 190; Molecular Devices, Sunnyvale, CA). To ensure 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 controls were included. For each concentration, six wells were analyzed. Of these six wells, the cells in two of these wells were treated with eriodictyol alone to screen for any potential toxicity. Cells in the remaining four wells were treated with eriodictyol and t-BOOH. Background absorbance values consisted of blank wells (with no cells) into which media, MTT dye, and solubilization buffer were added. 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 a percentage survival value. Cell survival data were analyzed with commercial software (GraphPad Prism 4; GraphPad Software, Inc., San Diego, CA). 

Cell Transfections

ARPE-19 cells were transfected in 96-well plates the day before transfection. Plasmid DNA was diluted with serum-free medium (without antibiotics; Opti-MEM; Invitrogen), and the transfection complex was formed at room temperature for 20 minutes before incubation with the cells. Transfections were performed (Lipofectamine 2000 or Lipofectamine LTX [Invitrogen]; or Fugene HD [Roche Applied Science, Indianapolis, IN]) according to the manufacturer’s instructions. ARPE-19 cells were exposed to the transfection complex of HMOX-1 or pcDNA-hNQO-1 for 16 to 24 hours. Cells transfected with one reagent (Lipofectamine 2000; Invitrogen) and HMOX-1 were allowed to recover for 24 to 36 hours before treatment with t-BOOH. Cells transfected with one transfection reagent (Fugene HD; Roche) and pcDNA-hNQO-1 were treated with t-BOOH after 8 hours, which was adequate time for recovery. The other transfection reagent (Lipofectamine LTX; Invitrogen) was used for the transfection of cells with dominant negative Nrf2 and with shRNA HO-1. For these experiments, the cells were exposed to the transfection complex for 24 hours before treatment with 50 μM eriodictyol for 16 hours and then were exposed to t-BOOH for 24 hours before cell viability was determined, as described earlier. 

Eriodictyol Treatment and Phase 2 Protein Expression

To evaluate the effect of eriodictyol on the activation of the transcription factor Nrf2 and the expression of the phase 2 proteins HO-1 and NQO-1, ARPE-19 cells were plated onto 60-mm² plates in DMEM/F12 medium supplemented with 10% dialyzed fetal bovine serum and 20 mM sodium bicarbonate and were treated the following day with eriodictyol at a concentration of 0 to 100 μM for 2 to 24 hours. To ensure that all cells for the various time points were in fact in culture for the same amount of time, eriodictyol was added as follows: at time 0, all the cells were transferred to 10% DFCS in DMEM/F12, and eriodictyol was added to the wells of the 24-hour time points. Eight hours before the end of the 24-hour period, eriodictyol was added to the 8-hour plate. This was repeated for the 6- and 4-hour plates such that all plates reached end point at the same time and, thus, were harvested at the end of the 24-hour period. For Nrf2, cells were washed twice in ice-cold Tris-buffered saline (TBS) and scraped into nuclear fractionation buffer (10 mM HEPES [pH 7.9], 10 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride (PMSF). After 15-minute incubation at 4°C, 10% Nonidet P-40 (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 wit the use of the BCA protein assay, and nuclear extracts were analyzed by Western blotting. For HO-1 and NQO-1, cells were rinsed twice in ice-cold TBS, scraped into cell lysis buffer (50 mM HEPES [pH 7.5], 50 mM NaCl, 5 mM EDTA, 1% Triton X-100, complete protease inhibitor cocktail [Roche], 1 mM PMSF), incubated at 4°C for 30 minutes, and centrifuged at 14,000 rpm for 10 minutes at 4°C. Protein levels were determined with the BCA protein assay (Pierce, Rockford, IL). Western blot analysis was performed with 15 μg total protein, as previously described. ²¹  

GSH Analysis

ARPE-19 cells were plated into six-well dishes at 7.5 × 10⁵ cells/well, replenished the following day with fresh DMEM/F12 supplemented with 10% dialyzed fetal bovine serum and 20 mM sodium bicarbonate, and treated with eriodictyol in a dose- and time-dependent manner. As described, the time-course experiments were performed such that all cells were incubated for the same amount of time, thus avoiding any differences in GSH levels as a consequence of culture time. The following day cells were trypsinized, pelleted, washed twice with ice-cold PBS, and resuspended in 1 mL PBS. Cell suspension was sonicated and centrifuged at 4°C. The supernatant was transferred to two microfuge tubes, one for protein determination using the Bio-Rad protein assay (Bio-Rad Laboratories) and the other for the GSH assay. The sample was deproteinated immediately after supernatant collection, and GSH assay was performed (NWLSS Glutathione Assay; Northwest Life Science Specialties, Vancouver, WA) according to the modified Tietze method, as described in the instruction booklet. 

Results

Eriodictyol Activation of Nrf2 and the Expression of the Phase 2 Proteins HO-1 and NQO-1

To examine the effect of eriodictyol on phase 2 protein expression, we performed a dose-response study in ARPE19 cells and found that eriodictyol induces the nuclear translocation of Nrf2 in a dose-dependent manner (Fig. 1A) . The activation of Nrf2 correlates with an increase in the expression of HO-1 (Fig. 1A)and a modest, though significant, increase in the expression of NQO-1 (Figs. 1A 1B) . In the absence of eriodictyol, ARPE-19 cells did not express HO-1, whereas there was some constitutive expression of NQO-1. Given the results of this study, we used 50 μM eriodictyol for our further experiments. 

To determine the temporal sequence of phase 2 protein expression, we analyzed the activation of Nrf2 during the first 8 hours after treatment with 50 μM eriodictyol. Nuclear translocation of Nrf2 occurred within 2 hours, reached a maximum by 4 hours, and was still elevated 8 hours after treatment (Fig. 2A , NE). Although the nuclear level of Nrf2 was significantly increased in response to eriodictyol, the cytosolic level remained relatively unaffected (Fig. 2A , Cyto). 

If the expression of phase 2 enzymes were induced by Nrf2 activation, we reasoned that we should observe a temporal correlation between the upregulation of HO-1 and NQO-1 and the activation of Nrf2. As shown in Figure 2A , HO-1 expression began within 2 hours of the activation of Nrf2 and reached a maximum by 6 to 8 hours. A modest, but significant, time-dependent increase in NQO-1 expression was also observed (Figs. 2A , NQO-1, and 2B). As expected, the upregulation of HO-1 and NQO-1 correlated with the induction of Nrf2. 

Overexpression of HO-1 and NQO-1: Protection of Cells from Oxidative Stress

If phase 2 proteins increase cellular defenses to oxidative stress, we reasoned that RPE cells that overexpress phase 2 proteins would have greater resistance to oxidative injury. To examine this, we transfected ARPE-19 cells with HO-1 and NQO-1 expression constructs and found that ARPE-19 cells that overexpress phase 2 protein HO-1 or NQO-1 are more resistant to oxidative stress–induced cell death than cells transfected with the control plasmid. Overexpression of HO-1 results in greater than 90% cell survival at concentrations of t-BOOH that kill up to 60% of the control cells (Fig. 3A) . Similarly, overexpression of NQO-1 results in approximately 75% cell survival at concentrations of t-BOOH that kill more than 75% of control cells (Fig. 4A) . Western blot analyses showing the overexpression of HO-1 and NQO-1 from these constructs are shown in Figures 3B and 4B , respectively. 

Eriodictyol Enhancement of Long-term Resistance to Oxidative Stress–Induced Cell Death

Given that eriodictyol induces the expression of HO-1 and NQO-1 and that these proteins enhance cellular resistance to oxidative stress, we reasoned that we should be able to demonstrate a correlation between HO-1 and NQO-1 expression and enhanced cell survival in oxidative stress–induced cell death assays. 

We treated ARPE-19 cells with eriodictyol for times sufficient to induce phase 2 protein expression and looked at the effect on cell survival. We used progressively higher levels of t-BOOH in dose-response studies until we found a concentration that surpassed the ability of eriodictyol to protect the cells in short-term protection studies. With this concentration of t-BOOH as our starting point, we exposed ARPE-19 cells to eriodictyol for times that were sufficient to induce phase 2 proteins and determined what effect this treatment had on cell survival. We found that pretreating ARPE19 cells with eriodictyol led to a significant increase in cell survival (Fig. 5) , a response that was consistent with the time frame of phase 2 protein expression (Fig. 2) . Cell survival approached 100% with pretreatment times of 8 hours or more. 

To further validate the idea that long-term protection by eriodictyol is not caused by the direct antioxidant activity of eriodictyol, we performed mass spectrometry with LC-MS to determine the stability of eriodictyol in tissue culture media after overnight incubation. We found that more than 90% of eriodictyol is lost from the culture media after overnight incubation (results not shown), supporting the idea that the long-term protective effect of eriodictyol is a secondary phenomenon. 

Eriodictyol Induces Glutathione Synthesis

The binding of Nrf2 to the ARE leads to a coordinated induction of multiple phase 2 proteins, including both subunits of glutamine cysteine ligase, the rate-limiting enzyme for glutathione biosynthesis. We reasoned that treatment with eriodictyol should lead to an increase in glutathione synthesis in ARPE19 cells. To test this hypothesis, we measured glutathione levels in ARPE-19 cells after extended incubation with eriodictyol. The amount of total glutathione increased by almost 100% compared with levels in untreated cells (Fig. 6A) . The increase in total glutathione levels (Fig. 6B)correlated temporally with the time course of long-term protection from eriodictyol (Fig. 5) . However, the increased protection demonstrated in the dose-response study with eriodictyol cannot be attributed to glutathione because the responses of total glutathione to 25, 50, and 100 μM eriodictyol were almost identical (Fig. 6A) . Thus, our data support possible roles for glutathione-dependent and -independent pathways in the mechanisms underlying eriodictyol-induced protection. 

Dominant Negative Nrf2 Reduction of Long-term Cytoprotective Effect of Eriodictyol in ARPE-19 Cells

To provide additional support for the hypothesis that the mechanism of cytoprotection observed with long-term eriodictyol treatment is a consequence of the activation of Nrf2, we transfected cells with a dominant negative to Nrf2 to determine whether it could reduce the cytoprotective effect observed after long-term eriodictyol treatment. The RPE cells were transfected with Nrf2M or the empty vector pEF for 24 hours and then were treated with eriodictyol for another 24 hours. We found that transfection with Nrf2M significantly reduced the survival of eriodictyol-treated cells compared with the survival of control cells transfected with pEF and with eriodictyol (Fig. 7A) . Western blot analyses showed that HO-1 expression was strongly downregulated in response to the Nrf2 dominant negative (Fig. 7B) . However, the expression of NQO-1 was less sensitive to Nrf2M expression and was unaffected by the expression of the dominant negative Nrf2 expression vector. These results, together with those showing that Nrf2 upregulation preceded the increase in HO-1 expression but correlated to a lesser extent with NQO-1 expression, suggested that the Nrf2-initiated protective effect was mediated more by upregulation of HO-1 expression than NQO-1 expression. On the other hand, we cannot rule out a possible role of Nrf2 on NQO-1 posttranslational modifications that favor the protective mechanism. 

To further analyze the role of HO-1 in the mechanism of eriodictyol-induced cytoprotection, we took an shRNA-targeting approach. Figure 8Ashows that the downregulation of HO-1 expression significantly reduced the protection effect of eriodictyol on oxidative-mediated damage. A Western blot demonstrating the downregulation of HO-1 in response to eriodictyol is shown in Figure 8B . Overall, these results strongly suggest that HO-1 plays a central role in the mechanism underlying eriodictyol-mediated protection against oxidant-mediated damage in ARPE-19 cells. 

Discussion

Enhancing the cellular defenses that protect the retina and retinal pigment epithelium against oxidative stress has been a valuable approach to reducing the progression of macular degeneration in patients with moderate to late stages of the condition. ¹ In previous studies, we showed that short-term exposure to flavonoids protects RPE cells, retinal ganglion cells, and other central nervous system neurons from oxidative stress–induced death. ^{21 24 25} In the present study, we focused on the ability of the flavonoid, eriodictyol, to modulate the activation of Nrf2 and the expression of phase 2 proteins, which are recognized as among the most important properties of flavonoids, and which likely surpass their benefit as direct antioxidants. ¹² To examine the impact of Nrf2 activation and phase 2 protein expression on RPE cell survival in the setting of oxidative stress, we exposed RPE cells to eriodictyol for periods of time sufficient to induce phase 2 protein expression, and we examined the long-term survival of ARPE-19 cells during these periods. We used two prototypical phase 2 proteins, HO-1 and NQO-1, as general markers for phase 2 protein expression because they can be easily detected with specific antibodies and are induced in a coordinated manner with other phase 2 proteins. We also analyzed GSH levels because the two subunits of glutamate cysteine ligase, the rate-limiting enzyme in GSH synthesis, are also phase 2 proteins. We observed that eriodictyol induced the expression of these phase 2 proteins in a manner that correlated temporally with increased cell survival in the setting of oxidative stress. One of the major findings in this study was that long-term treatment with eriodictyol increased the resistance of RPE cells to levels of oxidative stress much higher than cells can resist with short-term (1- to 4-hour) exposure, which is a result of the antioxidant activity of eriodictyol. These results emphasize the benefit of inducing phase 2 enzymes to combat oxidative stress compared with using direct antioxidants that are rapidly consumed or have a limited lifespan. In support of this conclusion, we found that simply overexpressing HO-1 and NQO-1 reductase increases the resistance of RPE cells to oxidative stress–induced cell death. 

In examining the mechanism of action of eriodictyol, we found that Nrf2 plays a central role in the protection of RPE cells from oxidative injury. There is a temporal correlation between the long-term survival of ARPE-19 cells and the time course of eriodictyol treatment, Nrf2 activation, HO-1 and NQO1 expression, and glutathione induction. It is reasonable to consider that the effect of phase 2 enzymes is additive and that the effect of a single enzyme cannot account for the full degree of protection achieved. For example, maximal total GSH levels were observed at eriodictyol concentrations lower than necessary for maximal protection against oxidative damage, suggesting that glutathione alone is not sufficient to exert complete protection and that other enzymes are involved. 

Here we showed that expression of a dominant negative Nrf2 that competed with eriodictyol-induced Nrf2 for binding to the ARE reduced the long-term protection of eriodictyol-treated RPE cells, as would be expected if the mechanism was dependent on Nrf2 activation. With the use of a similar dominant negative Nrf2 construct, other investigators have demonstrated the importance of Nrf2 expression in the protection of primary cultures of CNS neurons and glia from oxidative stress–induced cell death caused by glutamate toxicity. ²⁶ A central role for Nrf2 in the regulation of phase 2 protein expression and protection from oxidative toxicity has been demonstrated in Nrf2 transcription factor-deficient mice in which a deficiency of phase 2 proteins leads to increased rates of toxicity and neoplasia. ²⁷  

The evidence linking Nrf2 activation and phase 2 protein expression with the protection of RPE cells raises the possibility that these proteins might enhance the ability of the retinal pigment epithelium and retina to withstand long-term exposure to oxidative stress in situ. Other studies have shown that RPE cells are protected from a variety of oxidative insults in vitro by increasing intracellular GSH levels or by inducing the phase 2 proteins glutamate cysteine ligase and glutathione synthetase. ^{28 29 30 31 32} HO-1, which is present in the macula and peripheral RPE cells of human eyes, is enhanced in the setting of oxidative stress after intense light exposure and is increased in the retinas of patients with neovascular membranes from ARMD. ^{33 34} HO-1 mRNA levels show a decline in aged eyes, but correlations do not support causality, and it is not known whether these changes in expression have any relevance to the development of age-related diseases. ¹⁷ The expression of NQO-1, which is present throughout the retina and retinal pigment epithelium of healthy donor eyes, ¹⁸ correlates with the survival of RPE cells after oxidative injury in vitro, ³⁰ yet the significance of these findings are unknown. 

Limited clinical insight into the importance of phase 2 protein expression can be obtained from epidemiologic studies of diabetic patients with Gilbert syndrome, who have elevated levels of circulating bilirubin, a powerful and versatile antioxidant and an end product of HO-1 bioactivity. These patients have a significantly lower adjusted odds ratio for diabetic retinopathy (0.22; 95% confidence interval [CI], 0.2–0.45; P < 0.001) and other vascular diseases and lower levels of oxidative stress and inflammatory markers. ³⁵ They also have a significantly lower prevalence of ischemic cardiovascular disease, hypertension, and cancer (for a review, see Ref. ³⁶ ). 

Our results are consistent with reports that the activation of Nrf2 and downstream phase 2 proteins is associated with neuroprotection in models of retinal degeneration. ^{37 38} In the tubby (tub/tub) mutant mouse, photoreceptor degeneration is blocked after the activation of Nrf2 and the expression of the phase 2 protein thioredoxin by the oral agent sulforaphane. ³⁷ In a rodent model of light toxicity, neuroprotection is induced by light adaptation through the activation of the Nrf2/ARE pathway and the expression of thioredoxin. ³⁸  

The results presented here raise the question of whether the consumption of eriodictyol and other bioflavonoids that induce Nrf2 activation and phase 2 protein expression may, in part, be responsible for some of the ocular benefits associated with specific dietary products identified in epidemiologic studies of macular degeneration. ^{4 5 6} According to the United States Department of Agriculture database (http://www.nal.usda.gov/fnic/foodcomp/Data/Flav/flav.html), eriodictyol is present in a variety of citrus juices, including lemon juice, orange juice, and lime juice. ² In a large-scale prospective study of diet and macular degeneration, Cho et al. ⁵ identified fruit consumption as inversely related to the development of advanced neovascular ARMD. Surprisingly, the fruit with the strongest inverse association with neovascular ARMD was the orange. It is interesting to speculate whether eriodictyol was responsible for this effect, especially because eriodictyol is present in oranges and the authors could not attribute this finding to vitamin C or any other common antioxidants in oranges. ⁵  

Recently, Milbury et al. ³⁹ demonstrated that anthocyanidins found in bilberry fruit modulate HO-1 expression and block increased intracellular ROS levels in response to H₂O₂ in ARPE19 cells. Combined with our results, these findings support the possibility that eriodictyol or other flavonoids that induce the activation of Nrf2 and the expression of phase 2 enzymes may be capable of protecting RPE cells from oxidative injury and may be beneficial for the treatment of retinal disorders associated with oxidative stress. 

 Supported by the joint Mericos/Scripps Research Institute Neurobiology and Vision Science Research Program and by National Eye Institute Core Grant P30EY012598 for Vision Research.

 Submitted for publication March 27, 2008; revised November 2 and December 10, 2008; accepted March 19, 2009.

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

 

Eriodictyol activates the expression of Nrf2, HO-1, and NQO-1 in a dose-response manner. ARPE-19 cells were pretreated with eriodictyol (0–100 μM) for 2 hours to induce Nrf2 expression or overnight to induce HO-1 and NQO-1 expression. (A) Immunoblot analysis of Nrf2, HO-1, NQO-1, and β-actin expression. 15 μg protein/lane was loaded onto gels. The blots were incubated for 16 hours with Nrf2 (1:500), HO-1 (1:500), NQO-1 (1:5000), and β-actin (1:4000) in blotto (TBS containing 2% NFM). Nuclear and cytoplasmic extracts were prepared to analyze the expression of Nrf2, HO-1, NQO-1, and β-actin, respectively. (B) Densitometry and statistical analysis of NQO-1 expression in the eriodictyol dose response. Signals for NQO-1 expression in APRE-19 cells after pretreatment with eriodictyol (0–100 μM) were quantified, and statistical analysis was performed. Results are mean ± SE of three different experiments. *P < 0.05.

Figure 2.

 

Time course of eriodictyol-induced expression of Nrf2, HO-1, and NQO-1. (A) Immunoblot analysis of Nrf2, HO-1, NQO-1, and β-actin expression. ARPE-19 cells were pretreated with 50 μM eriodictyol for 0, 2, 4, 6, and 8 hours, after which nuclear and cytoplasmic extracts were prepared for analysis of the subcellular distribution of Nrf2, and cytoplasmic extracts were to prepared for analysis of the time course of HO-1, NQO-1, and β-actin expression. (B) Densitometry and statistical analysis of NQO-1 expression in the eriodictyol time course. Signals for NQO-1 expression in APRE-19 cells after pretreatment with eriodictyol (0–8 hours) were quantified, and statistical analysis was performed. Results are mean ± SE of three different experiments. *P < 0.05.

Figure 3.

 

Overexpression of HO-1 protects cells from oxidative stress–induced cell death. ARPE-19 cells were transiently transfected with the human wild-type HO-1 expression plasmid HMOX-1 (pHO-1) or with control plasmid pCMV6 (pCMV). (A) Cell survival of HO-1–overexpressing ARPE-19 cells. On the day after transfection, the cells were exposed to t-BOOH for 24 hours, and cell viability was determined by a modified version of the MTT assay. Results are mean ± SE of triplicates (*P < 0.05) and are a representative experiment from a set of five independent studies. (B) Immunoblot detection of HO-1 expression plasmid. Twenty-four hours after transfection, ARPE-19 cells were harvested, and cell lysates were prepared for the determination of HO-1 and β-actin expression by immunoblotting.

Figure 4.

 

Overexpression of NQO-1 protects cells from oxidative stress–induced cell death. ARPE-19 cells were transiently transfected with the human wild-type NQO-1 expression plasmid pcDNA-hNQO-1 (NQO-1) or with empty control plasmid pcDNA3 (CMV3). (A) Cell survival of NQO-1–overexpressing ARPE-19 cells. On the day after transfection, the cells were exposed to t-BOOH for 24 hours, and cell viability was determined by a modified version of the MTT assay. Results are mean ± SE of triplicates (*P < 0.05) and are a representative experiment from a set of five independent studies. (B) Immunoblot detection of NQO-1 expression plasmid. Twenty-four hours after transfection, ARPE-19 cells were harvested, and cell lysates were prepared for determination of NQO-1 and β-actin expression by immunoblotting.

Figure 5.

 

Long-term protection from eriodictyol in oxidative stress–induced cell death assays. ARPE-19 cells were pretreated with and without 50 μM eriodictyol for various times (0, 2, 4, 6, 8, and 24 hours) to initiate phase 2 protein expression before t-BOOH was added to the cells. All cells were maintained in culture for the same time period. On the following day, cell viability was determined by a modified version of the MTT assay. ²¹ Results are representative of a set of four independent studies of the time course of cell survival with eriodictyol pretreatment. Results are mean ± SD (*P < 0.05). Controls, consisting of cells exposed to no eriodictyol, had a cell survival rate of 0%.

Figure 6.

 

Eriodictyol increases total GSH levels. (A) Total GSH levels after eriodictyol treatment; dose response. ARPE-19 cells in six-well dishes were treated with 25, 50, and 100 μM eriodictyol for 24 hours. On the following day, the cells were harvested, and deproteinated cell lysates were assayed for total GSH. Results are an average of seven independent experiments (± SE). (B) Total GSH levels after eriodictyol treatment; time course study. ARPE-19 cells in six-well dishes were treated with 50 μM eriodictyol for 0 to 24 hours. At the end of the last time point, all cells were harvested, and the deproteinated cell lysates were assayed for total GSH. Results are mean ± SE of three independent experiments. Untreated cells were designated as 100% control.

Figure 7.

 

Dominant negative Nrf2 blocks the cytoprotection induced by eriodictyol. ARPE-19 cells were transiently transfected with pEF-Nrf2M expressing dominant-negative Nrf2 or the empty vector pEF. On the following day, the cells were treated with 50 μM eriodictyol for 24 hours. (A) Cell survival of Nrf2M-overexpressing ARPE-19 cells. Twenty-four hours after eriodictyol treatment, the cells were exposed to t-BOOH overnight. Cell viability was determined the following day by a modified version of the MTT assay. Similar findings were obtained in five independent experiments. Results are mean ± SE of triplicates (*P < 0.05) and are representative of three independent experiments. At the t-BOOH concentrations used for these studies, all cells untreated with eriodictyol had a cell survival rate of 0%. (B) Immunoblot analysis of HO-1, NQO-1, and Nrf-2M expression in ARPE-19 cells. Twenty-four hours after eriodictyol treatment, the cells were harvested. Both nuclear extracts were prepared to examine the expression of dominant negative Nrf2, and cell lysates were prepared to examine the expression of HO-1, NQO-1, and β-actin.

Figure 8.

 

Knockdown of HO-1 using an shRNA targeting approach blocks the cytoprotection induced by eriodictyol. ARPE-19 cells were transiently transfected with a shRNA HO-1 expression plasmid or the scrambled control vector, shRNA control. (A) Cell survival of HO-1 and control shRNA-expressing ARPE-19 cells. On the day after transfection, the cells were treated with 50 μM eriodictyol for 24 hours and were exposed to t-BOOH overnight. Cell viability was determined the following day by a modified version of the MTT assay. Results are mean ± SE and are representative of three independent experiments. (B) Immunoblot analysis of HO-1 and actin expression in ARPE-19 cells transfected with HO-1 shRNA and control shRNA. On the day after transfection, the cells were treated with 50 μM eriodictyol for 24 hours, and lysates were prepared for examination of the expression of HO-1 and β-actin.

The authors thank the staff at the Scripps Mericos Fonseca Research Fund at Scripps Memorial Hospital for their ongoing support and David Ross (University of Colorado Health Sciences Center) for his gift of the human wild-type NQO-1 expression plasmid (pcDNA-hNQO-1).