Polyphenols activate Nrf2 in astrocytes via H2O2, semiquinones, and quinones

Free Radical Biology & Medicine 51 (2011) 2319–2327

Contents lists available at SciVerse ScienceDirect

Free Radical Biology & Medicine journal homepage: www.elsevier.com/locate/freeradbiomed

Original Contribution

Polyphenols activate Nrf2 in astrocytes via H2O2, semiquinones, and quinones Hilla Erlank a, Anat Elmann a, 1, Ron Kohen b, Joseph Kanner a,⁎, 1 a b

Department of Food Science, ARO, Volcani Center, Bet-Dagan 50250, Israel Department of Pharmaceutics, School of Pharmacy, Hebrew University of Jerusalem, Jerusalem, Israel

a r t i c l e

i n f o

Article history: Received 8 May 2011 Revised 27 September 2011 Accepted 28 September 2011 Available online 12 October 2011 Keywords: Astrocytes Polyphenols tBHQ Curcumin Resveratrol Nrf2 EpRE NQO1 H2O2 Quinone Free radicals

a b s t r a c t Polyphenols, which occur both in edible plants and in foodstuff, have been reported to exert a wide range of health effects; however, the mechanism of action of these molecules is not fully understood. One important cellular pathway affected by polyphenols is the activation of the transcription factor Nrf2 via the electrophile response element, which mediates generation of phase 2 detoxifying enzymes. Our study found that Nrf2 nuclear translocation and the activity of NAD(P)H quinone oxidoreductase (NQO1) were increased significantly after treatment of astrocytes with tert-butylhydroquinone (tBHQ), resveratrol, or curcumin, at 20–50 μM. Incubation of tBHQ, resveratrol, and curcumin in the growth medium in the absence of astrocytes caused the accumulation of H2O2. Treatment of cells with either glutathione or metmyoglobin was found to decrease Nrf2 translocation and NQO1 activity induced by polyphenols by up to 40 and 60%, respectively. Addition of both glutathione and metmyoglobin to growth medium decreased Nrf2 translocation and NQO1 activity by up to 100 and 80%, respectively. In conclusion, because metmyoglobin, in the presence of polyphenols and glutathione, is known to interact with H2O2, semiquinones, and quinones, the up-regulation of the antioxidant defense of the cells through activation of the Nrf2 transcription factor, paradoxically, occurs via the generation of H2O2 and polyphenol-oxidized species generated from the exogenous microenvironment of the cells. © 2011 Elsevier Inc. All rights reserved.

Polyphenols, which occur both in edible plants and in foodstuff, form a substantial part of the human diet. Their total dietary intake could be as high as 1000 mg/day, which is much higher than the dietary intake of all other classes of phytochemicals and antioxidant vitamins [1]. It is well accepted that diets rich in polyphenols have health benefits, but the absorption of polyphenols in humans is limited and the mechanism of action of these molecules in the human body is not fully understood [2,3]. Some reported biological effects of polyphenols include antioxidant activity [4,5], amelioration of cardiovascular diseases [2], prevention of several degenerative age-related diseases [6], and prevention of several kinds of cancer [7]. Polyphenols are a large and diverse family of compounds synthesized by plants as secondary metabolites. The benzoic ring, which contains three double bonds, decreases very much the bond strength between hydrogen and oxygen in the linked hydroxyl group, turning it to a very active antioxidant [8,9]. The FDA-approved synthetic food antioxidants are polyphenols such as galates, BHT2 (butylhydroxytoluene),

Abbreviations: BHT, butylhydroxytoluene; EpRE, electrophile responsive element; GSH, glutathione; Keap1, kelch-like ECH-associated protein 1; MtMb, metmyoglobin; Nrf2, nuclear factor erythroid 2p45-related factor 2; NQO1, NAD(P)H:quinine oxidoreductase 1; ROS, reactive oxygen species; tBHQ, tert-butylhydroquinone. ⁎ Corresponding author. E-mail address: [email protected] (J. Kanner). 1 These authors contributed equally to this work. 0891-5849/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.freeradbiomed.2011.09.033

or tBHQ (tert-butylhydroquinone), a metabolite of butylhydroxy anisole. In cell and tissue culture systems, typical dietary plant polyphenolic compounds act as antioxidants, with protective properties [10] but under some circumstances they were found to be pro-oxidants and cytotoxic [3,11,12]. Although polyphenols are strong reducing agents, under in vitro conditions, in the presence of oxygen and metal ions, they could act as pro-oxidants, very much like ascorbic acid [13]. It has been reported that polyphenols undergo autoxidation and oxygen is consumed, generating O2•−, hydrogen peroxide (H2O2), semiquinones, and quinones [11,14,15]. The ability of apple extracts to inhibit proliferation of tumor cells in vitro was attributed to polyphenol antioxidants [16]. Our studies, for the first time, demonstrated that this inhibition was caused indirectly by H2O2 generated through interaction of the polyphenols with the cell culture medium [11,12]. Production of H2O2 by polyphenols in culture media was demonstrated by other researchers [15,17]. H2O2 is now clearly recognized as a part of the normal cell signaling that is involved in responses to specific genes involved in cell replication, regulation of metabolism, apoptosis, and necrosis [18,19]. H2O2 is an activator of the transcription factor nuclear factor erythroid 2p45-related factor-2 (Nrf2), which by translocation to the nucleus induces the activity of the electrophile response element (EpRE) [20,21]. H2O2 is electronically neutral and can freely diffuse through cellular membranes [22]. Compared to more aggressive ROS molecules such as hydroxyl radicals, which react with all molecules they

2320

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

encounter, H2O2 is a rather mild oxidant that primarily targets cysteines in diverse proteins and, therefore, can act as a second messenger inducing various transcription factors [18,19]. Naturally, cells and tissues are empowered with a panel of antioxidants and detoxifying enzymes, which are responsible for inactivating or eliminating ROS and elecrophilic compounds, thereby protecting cellular macromolecule damage caused by these species [5,23,24]. The proximal promoter regions for the antioxidant and detoxification genes contain the EpRE sequence, which is the preferred target of the nuclear transcription factor Nrf2. Nrf2 is sequestered in the cytoplasm as an inactive complex with its cytosolic repressor, Kelch-like ECH-associated protein 1 (Keap1). In response to oxidative or electophilic insults that oxidize two SH groups, Nrf2 is dissociated from the inhibitory protein Keap1 and is translocated to the nucleus to bind to the response element EpRE, leading to transcriptional activation of the antioxidant and cytoprotective genes [25]. Nrf2regulated gene products are phase 2 enzymes such as NAD(P)H:quinone oxidoreductase 1 (NQO1), glutathione peroxidase, glutathione Stransferase (GST), glutathione–cysteine ligase, heme oxygenase 1 (HO-1), and others [26]. A wide variety of phytochemicals have been shown to exert chemopreventive effects by protecting cellular antioxidation or detoxification capacities through activation of Nrf2 signaling. tBHQ [27,28], resveratrol [29], and curcumin [30] have been reported to induce antioxidant and detoxifying enzymes via Nrf2/EpRE signaling. Several studies adopted tBHQ as a positive control for activation of Nrf2 signaling in cells [27,28]. tBHQ was found to induce the synthesis of NQO1 and GST in mouse liver and intestinal mucosa [31]. More recently, it was found that through activation of Nrf2, tBHQ prevents the deposition of amyloid β-protein after oxidative stress, in NT2N neurons, a cell line model for Alzheimer disease [32], and provides effective prophylaxis against cerebral ischemia in vivo [33]. Resveratrol has been reported to elevate the expression and activity of several enzymes connected to the EpRE [34]. Moreover, resveratrol increased the phosphorylation and nuclear translocation of Nrf2 and also induced the activity as well as the expression of NQO1 at both protein and mRNA levels in human leukemia K562 cells [29]. The induction of antioxidant or detoxifying enzymes by curcumin is also mediated via Nrf2/EpRE signaling. Mice injected intraperitoneally with curcumin showed a twofold increase in total brain glutathione levels after treatment with buthionine sulfoximine [35]. Dietary administration of curcumin elevated hepatic GST and NQO1, resulting in increased detoxification of benzo(a)pyrene-treated mice [36]. Oral administration of curcumin also enhanced the nuclear translocation and the EpRE binding of Nrf2, inducing the expression of HO-1 in the liver of male ICR mice, protecting the animals against dimethylnitrosamineinduced hepatotoxicity [37]. Transgenic mice [38] that consumed juices or extracts of polyphenolrich berries had increased γ-glutamylcysteine-synthesized promoter and glutathione in muscles. The data demonstrate that polyphenols from various classes activate the Nrf2/EpRE pathway in ex vivo and in vivo model systems. However, the mechanism by which this action is induced by these molecules is not fully understood [2,3,39]. Astrocytes are multifunctional cells that are important for the maintenance of the normal functions of the central nervous system. One of their responsibilities is to increase the enzymatic redox activity executed by generating a set of phase 2 enzymes [40]. It was found that astrocytes in coculture with neurons at a concentration of 1% of the number of neurons protected the neuronal cells significantly from ROS stress [41]. The objectives of this research were to examine whether the polyphenols tBHQ, resveratrol, and curcumin are able to induce the activation of the Nrf2/EpRE pathway in astrocytes; whether this biological activity can be attributed to either their reduced or their oxidized metabolites; and if such activation could be affected by the exogenous microenvironment of the cells.

Materials and methods Chemicals and reagents Bovine serum albumin, β-nicotinamide adenine dinucleotide reduced (NADPH), BHT, curcumin, dichlorophenolindophenol (DCPIP), flavine adenine dinucleotide reduced (FAD), ferrous ammonium sulfate, glutathione reduced (GSH), luminol, myoglobin from horse skeletal muscle, poly-D-lysine (PDL), protease inhibitors, resveratrol, tBHQ, Triton X-100, and xylenol orange were purchased from Sigma (St. Louis, MO, USA). Dulbecco's modified Eagle's medium (DMEM), Leibovitz-15 medium, glutamine, antibiotics (10,000 IU/ml penicillin and 10,000 g/ml streptomycin), trypsin 0.25% (w/v)–0.53 mM EDTA solution, soybean trypsin inhibitor, and fetal bovine serum (FBS) were purchased from Biological Industries (Beit Haemek, Israel). 4′,6′-Diamidino-2-phenylindol dihydrochloride (DAPI) and dimethyl sulfoxide (DMSO) were obtained from AppliChem GmbH (Darmstadt, Germany). β-Mercaptoethanol, H2O2, Tris, Tris–hydrochloride, and monoclonal mouse anti-actin were obtained from MP Biomedicals (Solon, OH, USA). Acrylamide mix 30%, Laemmli sample buffer, sodium dodecyl sulfate (SDS), skim-milk powder, and Precision Plus dual-color standard protein assay dye reagent were purchased from Bio-Rad Laboratories (Hercules, CA ,USA). Protein fraction-enriched (Pro-FEK) commercial kit was obtained from ITSI Biosciences (Johnstown, PA, USA). Cytotoxicity detection kit (lactate dehydrogenase (LDH)) was obtained from Amresco (Solon, OH, USA). The antibody against Nrf2 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Fluorescein (FITC)-conjugated anti-rabbit IgG, horseradish peroxidase-conjugated anti-rabbit IgG, and antimouse IgG secondary antibodies along with goat serum were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA, USA). All organic solvents were of AR grade, purchased from Bio-Lab Ltd. (Jerusalem, Israel).

Preparation of primary cultures of astrocytes Cultures of primary rat astrocytes were prepared from cerebral cortices of 1- to 2-day-old neonatal Wistar rats. Briefly, while in Leibovitz-15 medium, meninges and blood vessels were carefully removed, brain tissues were dissociated by trypsinization with 0.5% trypsin (10 min, 37 °C, 5% CO2), and cells were washed first with DMEM containing soybean trypsin inhibitor (100 μg/ml) and 10% FBS and then with DMEM containing 10% FBS. Cells were seeded in tissue culture flasks precoated with PDL (20 μg/ml in 0.1 M borate buffer, pH 8.4) and incubated at 37 °C in humidified air with 5% CO2. The medium was changed on the second day and every second day thereafter. At the time of primary cell confluence (day 10), microglial and progenitor cells were discarded by shaking the flasks for 24 h on a horizontal shaking platform. Astrocytes were replated in 24-well PDL-coated plastic plates at a density of 1 × 10 5/well, in DMEM (without phenol red) containing 10% FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin.

Treatment of cells Twenty-four hours after plating, the original medium of the cells was aspirated and fresh medium was added to the cells. Glutathione and metmyoglobin were solubilized in growth medium. Dilutions of the various tested materials were done first in DMSO (resveratrol, tBHQ) or ethanol (curcumin) and then in the growth medium. All dilutions were made freshly from stock solutions just before each experiment and were used immediately. A combined stock solution of glutathione with each of the polyphenols (3:1) was prepared just before the experiment. Each treatment was performed in triplicate.

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

2321

Immunofluorescence Astrocytes were seeded into 24-well plates containing glass coverslips. After treatment, the cells were washed with 0.05% Tween 20 in PBS. Then, these cells were fixed with 4% formaldehyde in PBS containing 0.075 mM D-lysine and permeabilized with cold acetone. The cells were washed again and incubated with 6% goat serum in PBS overnight at room temperature. After the cells were washed, they were incubated for 3 h at room temperature with rabbit anti-human Nrf2 antibody in PBS containing 1% goat serum. The cells were washed again and further incubated for 1 h with FITC-conjugated goat anti-rabbit IgG as a secondary antibody. The cells were counterstained with DAPI to visualize the nuclei. For Nrf2 and DAPI detection, stained cells were viewed using a fluorescence microscope (Leica, Germany, DLMB) with a Fluotar lens. Photographs were taken using a DC 200 camera (40×/0.75 magnification) and processed using IM1000 software (Leica, Germany).

Fig. 1. tBHQ induces Nrf2 nuclear translocation in astrocytes. Astrocytes were treated with DMSO (control) or (A) various concentrations of tBHQ for 2.5 h or (B) 20 μM tBHQ for the indicated time points. Nuclear proteins were extracted, and equal amounts were separated by SDS–PAGE and immunoblotted with specific Nrf2 antibody.

Western blot analysis After treatment, cells were washed twice with ice-cold PBS. Separation of the nuclear proteins from the cytosolic and membranal proteins was conducted using a protein fraction-enriched (Pro-FEK) commercial kit, to which a protease inhibitor cocktail was added. Protein content was determined by Bradford reagent using bovine serum albumin as a standard. Samples were boiled in SDS sample buffer containing 10% β-mercaptoethanol for 5 min. Equal amounts of nuclear proteins (~ 40 μg) were separated by electrophoresis in 8% SDS– polyacrylamide gel and were transferred onto nitrocellulose membranes. After blotting, the membranes were blocked with 5% nonfat dry milk in PBS and incubated with rabbit anti-human Nrf2 antibody overnight at 4 °C followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG as a secondary antibody. To ensure equal loading of protein samples, membranes were stripped off and reprobed with monoclonal mouse anti-actin, followed by incubation with horseradish peroxidase-conjugated donkey anti-mouse IgG secondary antibody. Detection was performed using enhanced chemiluminescence.

reagent (250 μM Fe(NH4)2(SO4)2, 100 μM xylenol orange, 25 mM H2SO4, 4.4 mM BHT, 90% (v/v) methanol). After 30 min incubation at room temperature the samples were centrifuged (2 min, 5200 g) and the absorption was measured at 560 nm. Statistical analysis Statistical analyses were performed with one-way ANOVA followed by multiple comparison tests using GraphPad InStat 3 for Windows (GraphPad Software, San Diego, CA, USA). Results Polyphenols induce Nrf2 nuclear translocation in astrocytes To optimize the experimental conditions, time and dose response of the induction of Nrf2 translocation from the cytosol to the nucleus by tBHQ and resveratrol were determined. Figs. 1 and 2 show that the optimal conditions for induction by tBHQ or resveratrol were 2.5 h of

Determination of NQO1 activity After treatment, cultured cells were washed twice in PBS, and 0.5 ml of ice-cold homogenization buffer (50 mM potassium phosphate, pH 7.4, 1.15% KCl) was added to each well. The cells were scraped off and the entire plate was frozen for 24 h at −80 °C. Thawed cells were sonicated using a probe sonicator (Sonics Vibra Cell) before they were centrifuged at 10,000 g, 4 °C, for 20 min. The supernatants were collected and stored at −80 °C for later use to determine NQO1 enzymatic activities. Protein concentrations were determined by Bradford reagent using bovine serum albumin as a standard. NQO1 activity was determined by a continuous spectrophotometric assay to measure the reduction of its substrate, DCPIP, as described previously [33]. Briefly, 0.1–10 μg of protein from each sample was incubated with 1 ml of the assay buffer (40 μM DCPIP, 0.2 mM NADPH, 5 μM FAD, 25 mM Tris–HCl (pH 7.8), 0.1% (v/v) Tween 20, and 0.023% bovine serum albumin). The rate of DCPIP reduction was monitored over 1 min at 600 nm with an extinction coefficient (ε) of 2.1 mM − 1 cm − 1. The NQO1 activity was calculated as the decrease in absorbance per minute per milligram of total protein of the sample. Determination of H2O2 concentration by the FOX2 method Ten microliters of triphenylphosphine (TPP) solution (10 mM TPP in methanol) was added to 90 μl of fresh culture medium or of conditioned medium from treated astrocytes (18,000 cell equivalents). After 30 min of incubation the samples were diluted with FOX2

Fig. 2. Resveratrol induces Nrf2 nuclear translocation in astrocytes. Astrocytes were treated with DMSO (control) or (A) various concentrations of resveratrol for 2.5 h or (B) 25 μM resveratrol for the indicated time points. Nuclear proteins were extracted, and equal amounts were separated by SDS–PAGE and immunoblotted with specific Nrf2 antibody.

2322

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

incubation with 20 or 25 μM, respectively. Curcumin (30 μM) showed optimal induction after 4 h of incubation (data not shown). Similar results were obtained when astrocytes were treated with the abovementioned polyphenols and were immunostained for Nrf2. Fig. 3A shows representative photographs of the cells, and Fig. 3B shows quantitative analysis of the photographs. The indicated concentrations were not toxic to astrocytes, as examined by LDH cytotoxicity test (data not shown). Polyphenols produce H2O2, which is diminished by astrocytes H2O2 concentration was used as an indicator for the reactive oxygen species formed during polyphenol oxidation. To estimate the autoxidation of various polyphenols, H2O2 production by each polyphenol was measured after incubation in growth medium. In the absence of astrocytes, after 30 min of incubation of tBHQ in the growth medium, H2O2 was generated in concentrations equimolar to those of tBHQ (Fig. 4A). Resveratrol was less active, and H2O2 generated by this polyphenol was 10% of its molar concentrations (Fig. 4B). Results similar to those of resveratrol were obtained for curcumin (data not shown). Once polyphenols were incubated with astrocytes, significantly lower concentrations of H2O2 could be detected in the growth medium, compared to those detected in the absence of astrocytes (Fig. 5).

Scavenging of the oxidized forms of the polyphenols reduces their ability to induce Nrf2 nuclear translocation To distinguish between the biological activities of the reduced forms of the various polyphenols and their respective oxidized forms, the reducing agent glutathione was used. Incubation of astrocytes with tBHQ in the presence of glutathione, which binds to the oxidized form of tBHQ, lowered nuclear levels of Nrf2 (Fig. 6). In the presence of metmyoglobin, which acts as a pseudo-peroxidase and scavenges H2O2, Nrf2 translocation into the nucleus was also inhibited (Fig. 6). Addition of both glutathione and metmyoglobin to tBHQ-treated astrocytes was found to further decrease the nuclear levels of Nrf2 (Fig. 6). Similar results were obtained when astrocytes were treated with resveratrol or curcumin in the presence of either glutathione or metmyoglobin (Fig. 7). Addition of glutathione, metmyoglobin, or both agents inhibited Nrf2 translocation induced by resveratrol or curcumin by 40–60%. Polyphenols increase NQO1 activity in astrocytes, and scavenging of their oxidized forms inhibits their ability to increase the enzymatic activity The activity of NQO1, which is known to be induced by Nrf2, increased by 1.5- to 1.8-fold after treatment of astrocytes with tBHQ, curcumin, or resveratrol (Fig. 8). The effects of glutathione and

Fig. 3. Induction of nuclear translocation of Nrf2 in astrocytes by various polyphenols. Astrocytes were treated for 2.5 h with tBHQ (20 μM) or resveratrol (25 μM), or for 4 h with curcumin (30 μM). Control cells were treated with equal volumes of DMSO or ethanol. Afterward, cells were fixed and immunostained against Nrf2, and DNA was stained with DAPI. (A) Representative microphotographs at × 400 original magnification of immunofluorescent staining of Nrf2 (left) and counterstaining of nuclei with DAPI (right) in astrocytes. In control cells arrows indicate nuclei lacking Nrf2; in resveratrol-treated cells arrows indicate nuclei containing Nrf2. (B) Nrf2 translocation was calculated as the percentage of Nrf2stained nuclei from the overall number of DAPI-stained nuclei. For resveratrol n = 13; for tBHQ n = 4; for curcumin n = 7. *p b 0.05 vs control group.

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

2323

Fig. 7. The effects of glutathione and metmyoglobin on nuclear translocation of Nrf2 induced by resveratrol or curcumin. Astrocytes were treated with resveratrol, glutathione, or metmyoglobin (25, 75, 25 μM, respectively) for 2.5 h or curcumin, glutathione, and metmyoglobin (30, 90, 30 μM, respectively) for 4 h. Control cells were treated with equal volumes of DMSO or ethanol. After treatment, the cells were fixed and immunostained against Nrf2, and DNA was stained with DAPI. Nrf2 translocation was calculated as the percentage of Nrf2-stained nuclei from the overall number of DAPI-stained nuclei. Percentage of stained nuclei of control cells was subtracted from all treatments. *p b 0.05, n = 6.

Fig. 4. H2O2 generation by tBHQ and resveratrol in growth medium in the absence of astrocytes. H2O2 concentration was determined in fresh growth medium containing DMSO (control) or various concentrations of (A) tBHQ or (B) resveratrol at 37 °C, using the FOX2 method.

Fig. 8. Polyphenols induce NQO1 activity in astrocytes. Astrocytes were treated with tBHQ (20 μM, 24 h), curcumin (20 μM, 24 h), or resveratrol (2.5 μM, 18 h). Control cells were treated with an equal volume of DMSO. NQO1 activity (1 μg cellular proteins) was determined as described under Materials and methods. *p b 0.05. Fig. 5. H2O2 generation in growth medium, by various polyphenols, in the absence or presence of astrocytes. The concentrations of H2O2 generated by tBHQ (20 μM), resveratrol (25 μM), or curcumin (30 μM) were determined after 1 h of incubation in the presence or absence of astrocytes, using the FOX2 method. *p b 0.05, n = 3.

metmyoglobin on polyphenol-induced NQO1 activity in astrocytes were determined, and Fig. 9 demonstrates that either glutathione or metmyoglobin significantly decreased NQO1 activity induced by tBHQ, resveratrol, or curcumin. However, maximal inhibition (50–80%) of polyphenol-induced NQO1 activity was achieved in the presence of both glutathione and metmyoglobin.

Discussion Polyphenols are naturally occurring compounds present in fruits, vegetables, and many beverages that have been reported to exert a wide range of health effects; however, the mechanism by which polyphenols neutralize and protect against oxidative stress in humans still remains unclear. The Nrf2 transcription factor regulates major environmental and oxidative stress responses [42]. In the cytoplasm, Nrf2 is negatively regulated by sequestering the protein Keap1. Electrophilic

Fig. 6. The effects of glutathione and metmyoglobin on Nrf2 translocation induced by tBHQ. Astrocytes were treated (2.5 h) with tBHQ (20 μM) in the absence and presence of glutathione (60 μM) and/or metmyoglobin (20 μM). Control cells were treated with an equal volume of DMSO. Nuclear proteins were extracted, and equal amounts were separated by SDS–PAGE and immunoblotted with specific Nrf2 antibody.

2324

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

the electron-releasing power expressed by the HOMO energy, or the ease with which a molecule donates an electron and oxidizes, correlates with the induction of NQO1. Interestingly, the results reported by Zoete et al. [44] are very similar to those published by Lee-Hilz et al. [46], who reported that EHOMO of 21 tested flavonoids correlated with their induction factor for the EpRE-mediated gene transcription. Our results show that tBHQ, curcumin, and resveratrol generated H2O2 in the cell growth medium during their autoxidation. The results are in line with our previous results on the generation of H2O2 by polyphenols in cell growth medium [11,12] and also those of others [15]. tBHQ in cell growth medium generated H2O2 in an amount equimolar to its concentration. The stoichiometric nature of the polyphenol oxidation can be explained by the following equation: Fig. 9. Polyphenol induction of NQO1 activity as affected by glutathione and metmyoglobin. Astrocytes were treated with resveratrol, glutathione, and metmyoglobin (25, 75, 25 μM, respectively) for 2.5 h or curcumin, glutathione, and metmyoglobin (30, 90, 30 μM, respectively) for 4 h. Control cells were treated with equal volumes of vehicle. NQO1 activity was determined, using 1 μg cellular proteins, as described under Materials and methods. *p b 0.05.

compounds can activate Nrf2 primarily by oxidizing or alkylating specific cysteines, causing structural modifications in the two Keap1 molecules clenching the Nrf2 and enabling its release [43] (Fig. 10). One important mechanism by which polyphenols seem to affect human metabolism is by the induction of phase 2 detoxifying enzymes, via an EpRE-mediated response [44-46]. Zoete et al. [44] examined the EHOMO of 30 different polyphenols and their induction factor for NQO1 activation in Hepa 1c1c7 cells. They concluded that

Ph ðOHÞ2 þ O2 →Ph ¼ ðOÞ2 þ H2 O2 :

ð1Þ

Our study demonstrated a significant reduction in exogenous H2O2 by the astrocytic culture. The ability of these cells to detoxify exogenously applied H2O2 was previously determined by Dringer and Hamprecht [47] and corresponds with our results. The reduction of H2O2 seems to occur mostly because of its ability to freely diffuse through cell membranes [22,47] and its intracellular decomposition by glutathione peroxidase or by thioredoxins. A small part of this H2O2 flux seems to affect signaling, translocation of Nrf2, and activation of NQO1. Our previous data [5,48] demonstrated that metmyoglobin strongly interacted with H2O2, forming oxoferryl myoglobin. Although oxoferryl myoglobin may cause oxidative damage under some circumstances, we clearly demonstrated that in our systems,

Fig. 10. The induction of the Nrf2/EpRE pathway in astrocytes by polyphenols.

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

which contained polyphenols [5,48,49], the effect was far less toxic than in the absence of polyphenols. Oxoferryl myoglobin is very efficiently reduced by polyphenolic antioxidants, acting as pseudoperoxidase, as summarized below: Ph OH þ O2 þ culture medium→Ph ¼ O þ H2 O2 ; 3þ

MbFe •

4þ

MbFe

•

4þ

þ H2 O2 → MbFe

–O þ H2 O; 3þ

–O þ PhðOHÞ2 →MbFe

ð2Þ ð3Þ

þ polymerized phenolics:

It may be suggested that the ferryl myoglobin generated in the medium will oxidize all the polyphenols, leaving the system with oxidized ineffective polyphenols for further H2O2 generation. However, from the stoichiometric point of view, for example, resveratrol in the presence of astrocytes generated only 4 μM H2O2/2 h. This amount of H2O2 will activate MtMb, forming ferryl (4 μM). Taking into consideration that a part of ferrylmyoglobin is autoreduced, then about 2 μM ferryl will oxidize 4 μM polyphenols, leaving in the system about 21 μM reduced polyphenols. As we found that the system needs only 2 h for a significant induction of Nrf2 signaling (see Fig. 2), the medium remained with enough reduced polyphenols for further H2O2 generation. In culture medium, metmyoglobin decomposes H2O2 generated by polyphenols much better than catalase [12]. As most of the polyphenols interact with membranal proteins and phospholipids, it seems that generation of H2O2 occurs on the extracellular membrane and its diffusion into cells is much higher than its accumulation in the culture medium. However, it is possible that in the presence of cells, decomposition of H2O2 also occurs by exogenously secreted proteins such as thioredoxins. Previously we showed that catalase decreased H2O2 levels generated by glucose oxidase, but it could not cause the same effect when H2O2 was generated by several polyphenols [12]. This discrepancy is of note because glucose oxidase generates H2O2 in the medium and polyphenols generate H2O2, site specifically, on the membrane. To inhibit such H2O2 flux, myoglobin, a cationized protein, was found to interact with the negatively charged membranes, which makes it a more efficient H2O2 decomposer than catalase, which has a low affinity for membranes, because of its negative charge [12,50]. Our results explain those of Lee et al. (27), who found that catalase could inhibit Nrf2 activation induced by diethyl maleate but not by tBHQ. Diethyl maleate increases production of H2O2 in the cell medium [51]. However, like many other polyphenolic compounds, the polyphenol tBHQ is better associated with the membrane and upon autoxidation produces H2O2 in the proximity of the membrane [52,53]. Our results clearly show that metmyoglobin significantly prevents tBHQ, resveratrol, and curcumin activation of Nrf2; its translocation to the nucleus; and NQO1 activity. As metmyoglobin has no cell permeativity, its activity is carried out exogenously, meaning in the growth medium. Glutathione, like metmyoglobin, has no cell permeativity and therefore also acts exogenously in our system. Glutathione prevents generation of superoxide and H2O2 by scavenging phenoxyl radicals. It also reduces semiquinones by generating GSSG and glutathione–phenoxyl conjugates, as can be seen in Reactions (5)–(7) [54,55]. Glutathione alone was found to decrease Nrf2 translocation induced by tBHQ, resveratrol, and curcumin, most probably by eliminating their oxidation products, H2O2 and quinones, and thus preventing their diffusion into the cells and the activation of Nrf2. Although resveratrol and curcumin generate low concentrations of H2O2, as was found also by Long et al. [56], the activation of Nrf2 was inhibited by glutathione. Glutathione at high concentration (mM) could react with H2O2, albeit slowly; however, at the

2325

concentration we used (μM) it seems to act spontaneously with phenoxyl radicals and quinones: •

•

PhO þ GSH→PhOH þ GS ; •

ð5Þ •

Ph ¼ O þ GSH→PhO þ GS →GS–PhOH; •

ð6Þ

•

GS þ GS →GSSG:

ð7Þ

Nrf2 activation induced by all three polyphenols tested in this study was inhibited more efficiently after the addition of both metmyoglobin and glutathione. This effect was well demonstrated by a low NQO1 enzymatic activity. This outcome was achieved most probably by an additional beneficial effect on the reduction of H2O2 and oxidized polyphenols, as demonstrated by the following reaction: 3þ

3þ

MbFe þ H2 O2 þ 2GSH þ PhðOHÞ2 →MbFe þ GSSG þ 2H2 O:

þ PhðOHÞ2

ð8Þ

The generation of a glutathione–polyphenol conjugated molecule prevented the production of superoxide and H2O2 from the medium. Reaction (8), which was demonstrated by Khalife and Lupidi [57], displays decomposition of H2O2 in the presence of metmyoglobin, polyphenol, and glutathione. Such an interaction could not only affect the reduction of hypervalent myoglobin but also increase the elimination rate of H2O2 from the medium. Upon ingestion, flavonoid glucosides are deglycosylated and the aglycones are metabolized to glucoronide, sulfate, and methyl conjugates [46,58]. Only a small part of the metabolites are transported to the bloodstream and flavonoid glucoronides are the major metabolites present in the circulation [58]. Several metabolites of quercetin were shown to effectively activate EpRE-mediated gene expression, thereby inducing detoxification enzymes such as NQO1. However, the induction by the mixed metabolites constituted only 40–50% of the maximal level induced by quercetin aglycone alone [46]. This seems reasonable because quercetin metabolites generated lower concentrations of H2O2 and quinones [59,60]. Hydrogen peroxide is a well-accepted second messenger [19]. Among many ROS, H2O2 is more stable and plays the role of a survival molecule. The main prosurvival functions are kinase-driven oxidation of cysteines in the active sites of various phosphatases and the regulation of transcription factors such as p53, NF-κB, AP-1, and Nrf2. Hydrogen peroxide concentration and its specific site of generation greatly influence its activity as a second messenger. Thus, for H2O2 to play a direct role in signaling, its target(s) must be localized near its site of production especially because of the high cellular enzymatic activity of peroxidases. The application of an extracellular concentration of H2O2 of 0.1–5.0 μM to cell cultures results in intracellular H2O2 levels of about 0.01–0.5 μM, which directly stimulates cell proliferation and affects cell adaptation (such activation of Nrf2), whereas very high concentrations induce apoptosis and cell death [58]. According to our results, the medium concentration of H2O2 achieved in the presence of cell culture was lower than 5 μM. This most probably resulted in an intracellular concentration in the range of cell adaptation to ROS. Most recently it was found that Keap1 intermolecular disulfide formation via cysteine 151 underlies the activation of Nrf2 by low concentrations of H2O2 [21]. In conclusion, cytosolic Nrf2 levels decreased and nuclear Nrf2 levels as well as NQO1 enzymatic activity increased in astrocytes after treatment with tBHQ, resveratrol, and curcumin. All polyphenols, when present in growth medium, generated H2O2. H2O2 concentrations in the media were affected to a large extent by the presence of astrocytes, which reduced it to 4–5 μM. As most of the polyphenols interact with cell membranes, reduction of H2O2 seems to occur via its partial diffusion into the cells. Addition of metmyoglobin and

2326

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327

glutathione to the medium decreased significantly the translocation of Nrf2 and NQO1 activity, induced by all polyphenols. Because metmyoglobin and glutathione fail to permeate membranes, their mode of action is expressed exogenously. For a long time, ROS generated in cells had been considered harmful mediators because of their highly reactive nature. Although traditionally considered lethal to cells, ROS at low concentrations seem to be involved in redox signaling that may contribute to normal cell function and adaptation as well as disease prevention [61,62]. We hypothesize that low concentrations of polyphenols generate H2O2, at very low concentration, at the level of arterioles and capillaries, in association with the outer surface of the cells, and by diffusion it activates Nrf2 signaling, inducing cell adaptation to oxidative stress. Thus, polyphenols act as nutritional "medicinals," which might have a preventative nature, rather than functioning as therapeutic agents. However, for the same reason, high concentrations of polyphenols could enhance the generation of H2O2 and other metabolites capable of causing cytotoxic events. Acknowledgment The authors thank Mrs. Mordechay Sharon, Mrs. Rindner Miriam, and Mrs. Granit Rina for technical assistance. References [1] Kuhnau, J. The flavonoids: a class of semi-essential food components: their role in human nutrition. World Rev. Nutr. Diet. 24:117–191; 1976. [2] Manach, C.; Williamson, G.; Morand, C.; Scalbert, A.; Remesy, C. Bioavailability and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability studies. Am. J. Clin. Nutr. 81:230S–242S; 2005. [3] Tang, S. Y.; Halliwell, B. Medicinal plants and antioxidants: what do we learn from cell culture and Caenorhabditis elegans studies? Biochem. Biophys. Res. Commun. 394:1–5; 2010. [4] Bors, W.; Saran, M. Radical scavenging by flavonoid antioxidants. Free Radic. Res. Commun. 2:289–294; 1987. [5] Kanner, J.; German, J. B.; Kinsella, J. E. Initiation of lipid peroxidation in biological systems. Crit. Rev. Food Sci. Nutr. 25:317–364; 1987. [6] Stevenson, D. E.; Hurst, R. D. Polyphenolic phytochemicals—just antioxidants or much more? Cell. Mol. Life Sci. 64:2900–2916; 2007. [7] Roginsky, A. On the potential use of flavonoids in the treatment and prevention of pancreatic cancer. In Vivo 19:61–67; 2005. [8] Kanner, J.; Frankel, E.; Granit, R.; German, B.; Kinsella, J. E. Natural antioxidants in grapes and wines. J. Agric. Food Chem. 42:64–69; 1994. [9] Lapidot, T.; Granit, R.; Kanner, J. Lipid peroxidation by "free" iron ions and myoglobin as affected by dietary antioxidants in simulated gastric fluids. J. Agric. Food Chem. 53:3383–3390; 2005. [10] Ramassamy, C. Emerging role of polyphenolic compounds in the treatment of neurodegenerative diseases: a review of their intracellular targets. Eur. J. Pharmacol. 545:51–64; 2006. [11] Lapidot, T.; Walker, M. D.; Kanner, J. Can apple antioxidants inhibit tumor cell proliferation? Generation of H2O2 during interaction of phenolic compounds with cell culture media. J. Agric. Food Chem. 50:3156–3160; 2002. [12] Lapidot, T.; Walker, M. D.; Kanner, J. Antioxidant and prooxidant effects of phenolics on pancreatic beta-cells in vitro. J. Agric. Food Chem. 50:7220–7225; 2002. [13] Kanner, J.; Mendel, H.; Budowski, P. Pro-oxidant and antioxidant effects of ascorbic-acid and metal-salts in a beta-carotene–linoleate model system. J. Food Sci. 42:60–64; 1977. [14] Canada, A. T.; Giannella, E.; Nguyen, T. D.; Mason, R. P. The production of reactive oxygen species by dietary flavonols. Free Radic. Biol. Med. 9:441–449; 1990. [15] Long, L. H.; Clement, M. V.; Halliwell, B. Artifacts in cell culture: rapid generation of hydrogen peroxide on addition of (−)-epigallocatechin, (−)-epigallocatechin gallate, (+)-catechin, and quercetin to commonly used cell culture media. Biochem. Biophys. Res. Commun. 273:50–53; 2000. [16] Eberhardt, M. V.; Lee, C. Y.; Liu, R. H. Antioxidant activity of fresh apples. Nature 405:903–904; 2000. [17] Maeta, K.; Nomura, W.; Takatsume, Y.; Izawa, S.; Inoue, Y. Green tea polyphenols function as prooxidants to activate oxidative-stress-responsive transcription factors in yeasts. Appl. Environ. Microbiol. 73:572–580; 2007. [18] Reth, M. Hydrogen peroxide as second messenger in lymphocyte activation. Nat. Immunol. 3:1129–1134; 2002. [19] Rhee, S. G. Cell signaling: H2O2, a necessary evil for cell signaling. Science 312: 1882–1883; 2006. [20] Shih, A. Y.; Erb, H.; Murphy, T. H. Dopamine activates Nrf2-regulated neuroprotective pathways in astrocytes and meningeal cells. J. Neurochem. 101:109–119; 2007. [21] Fourquet, S.; Guerois, R.; Biard, D.; Toledano, M. B. Activation of NRF2 by nitrosative agents and H2O2involves KEAP1 disulfide formation. J. Biol. Chem. 285: 8463–8471; 2010.

[22] Bienert, G. P.; Schjoerring, J. K.; Jahn, T. P. Membrane transport of hydrogen peroxide. Biochim. Biophys. Acta 1758:994–1003; 2006. [23] Rhee, S. G.; Yang, K. S.; Kang, S. W.; Woo, H. A.; Chang, T. S. Controlled elimination of intracellular H2O2: regulation of peroxiredoxin, catalase, and glutathione peroxidase via post-translational modification. Antioxid. Redox Signal. 7:619–626; 2005. [24] Talalay, P. Chemoprotection against cancer by induction of phase 2 enzymes. Biofactors 12:5–11; 2000. [25] Chen, C.; Kong, A. N. Dietary chemopreventive compounds and ARE/EpRE signaling. Free Radic. Biol. Med. 36:1505–1516; 2004. [26] Wakabayashi, N.; Dinkova-Kostova, A. T.; Holtzclaw, W. D., et al. Protection against electrophile and oxidant stress by induction of the phase 2 response: fate of cysteines of the Keap1 sensor modified by inducers. Proc. Natl. Acad. Sci. U. S. A. 101:2040–2045; 2004. [27] Lee, J. M.; Moehlenkamp, J. D.; Hanson, J. M.; Johnson, J. A. Nrf2-dependent activation of the antioxidant responsive element by tert-butylhydroquinone is independent of oxidative stress in IMR-32 human neuroblastoma cells. Biochem. Biophys. Res. Commun. 280:286–292; 2001. [28] Ben-Dor, A.; Steiner, M.; Gheber, L.; Danilenko, M.; Dubi, N.; Linnewiel, K.; Zick, A.; Sharoni, Y.; Levy, J. Carotenoids activate the antioxidant response element transcription system. Mol. Cancer Ther. 4:177–186; 2005. [29] Hsieh, T. C.; Lu, X.; Wang, Z.; Wu, J. M. Induction of quinone reductase NQO1 by resveratrol in human K562 cells involves the antioxidant response element ARE and is accompanied by nuclear translocation of transcription factor Nrf2. Med. Chem. 2:275–285; 2006. [30] Dickinson, D. A.; Iles, K. E.; Zhang, H. Q.; Blank, V.; Forman, H. J. Curcumin alters EpRE and AP-1 binding complexes and elevates glutamate–cysteine ligase gene expression. FASEB J. 17:473–476; 2003. [31] Prochaska, H. J.; De Long, M. J.; Talalay, P. On the mechanisms of induction of cancer-protective enzymes: a unifying proposal. Proc. Natl. Acad. Sci. U. S. A. 82: 8232–8236; 1985. [32] Eftekharzadeh, B.; Maghsoudi, N.; Khodagholi, F. Stabilization of transcription factor Nrf2 by tBHQ prevents oxidative stress-induced amyloid beta formation in NT2N neurons. Biochimie 92:245–253; 2010. [33] Shih, A. Y.; Li, P.; Murphy, T. H. A small-molecule-inducible Nrf2-mediated antioxidant response provides effective prophylaxis against cerebral ischemia in vivo. J. Neurosci. 25:10321–10335; 2005. [34] Chen, C. Y.; Jang, J. H.; Li, M. H.; Surh, Y. J. Resveratrol upregulates heme oxygenase-1 expression via activation of NF-E2-related factor 2 in PC12 cells. Biochem. Biophys. Res. Commun. 331:993–1000; 2005. [35] Jagatha, B.; Mythri, R. B.; Vali, S.; Bharath, M. M. Curcumin treatment alleviates the effects of glutathione depletion in vitro and in vivo: therapeutic implications for Parkinson's disease explained via in silico studies. Free Radic. Biol. Med. 44:907–917; 2008. [36] Garg, R.; Gupta, S.; Maru, G. B. Dietary curcumin modulates transcriptional regulators of phase I and phase II enzymes in benzo[a]pyrene-treated mice: mechanism of its anti-initiating action. Carcinogenesis 29:1022–1032; 2008. [37] Farombi, E. O.; Shrotriya, S.; Na, H. K.; Kim, S. H.; Surh, Y. J. Curcumin attenuates dimethylnitrosamine-induced liver injury in rats through Nrf2-mediated induction of heme oxygenase-1. Food Chem. Toxicol. 46:1279–1287; 2008. [38] Myhrstad, M. C.; Carlsen, H.; Dahl, L. I.; Ebihara, K.; Glemmestad, L.; Haffner, K.; Moskaug, J. O.; Blomhoff, R. Bilberry extracts induce gene expression through the electrophile response element. Nutr. Cancer 54:94–101; 2006. [39] Gutteridge, J. M.; Halliwell, B. Antioxidants: molecules, medicines, and myths. Biochem. Biophys. Res. Commun. 393:561–564; 2010. [40] Cotrina, M. L.; Nedergaard, M. Astrocytes in the aging brain. J. Neurosci. Res. 67: 1–10; 2002. [41] Shih, A. Y.; Johnson, D. A.; Wong, G.; Kraft, A. D.; Jiang, L.; Erb, H.; Johnson, J. A.; Murphy, T. H. Coordinate regulation of glutathione biosynthesis and release by Nrf2-expressing glia potently protects neurons from oxidative stress. J. Neurosci. 23:3394–3406; 2003. [42] Leiser, S. F.; Miller, R. A. Nrf2 signaling, a mechanism for cellular stress resistance in long-lived mice. Mol. Cell. Biol. 30:871–884; 2009. [43] Dinkova-Kostova, A. T.; Talalay, P. Direct and indirect antioxidant properties of inducers of cytoprotective proteins. Mol. Nutr. Food Res. 52 (Suppl. 1):S128–138; 2008. [44] Zoete, V.; Rougee, M.; Dinkova-Kostova, A. T.; Talalay, P.; Bensasson, R. V. Redox ranking of inducers of a cancer-protective enzyme via the energy of their highest occupied molecular orbital. Free Radic. Biol. Med. 36:1418–1423; 2004. [45] Liu, H.; Dinkova-Kostova, A. T.; Talalay, P. Coordinate regulation of enzyme markers for inflammation and for protection against oxidants and electrophiles. Proc. Natl. Acad. Sci. U. S. A. 105:15926–15931; 2008. [46] Lee-Hilz, Y. Y.; Boerboom, A. M.; Westphal, A. H.; Berkel, W. J.; Aarts, J. M.; Rietjens, I. M. Pro-oxidant activity of flavonoids induces EpRE-mediated gene expression. Chem. Res. Toxicol. 19:1499–1505; 2006. [47] Dringen, R.; Hamprecht, B. Involvement of glutathione peroxidase and catalase in the disposal of exogenous hydrogen peroxide by cultured astroglial cells. Brain Res. 759:67–75; 1997. [48] Kanner, J.; Harel, S. Initiation of membranal lipid peroxidation by activated metmyoglobin and methemoglobin. Arch. Biochem. Biophys. 237:314–321; 1985. [49] Kanner, J.; Lapidot, T. The stomach as a bioreactor: dietary lipid peroxidation in the gastric fluid and the effects of plant-derived antioxidants. Free Radic. Biol. Med. 31:1388–1395; 2001. [50] Kohen, R.; Kakunda, A.; Rubinstein, A. The role of cationized catalase and cationized glucose oxidase in mucosal oxidative damage induced in the rat jejunum. J. Biol. Chem. 267:21349–21354; 1992. [51] Iantomasi, T.; Favilli, F.; Catarzi, S.; Vincenzini, M. T. GSH role on platelet-derived growth factor receptor tyrosine phosphorylation induced by H2O2. Biochem. Biophys. Res. Commun. 280:1279–1285; 2001.

H. Erlank et al. / Free Radical Biology & Medicine 51 (2011) 2319–2327 [52] Koren, E.; Kohen, R.; Ginsburg, I. Polyphenols enhance total oxidant-scavenging capacities of human blood by binding to red blood cells. Exp. Biol. Med. (Maywood) 235:689–699; 2010. [53] Youdim, K. A.; Shukitt-Hale, B.; MacKinnon, S.; Kalt, W.; Joseph, J. A. Polyphenolics enhance red blood cell resistance to oxidative stress: in vitro and in vivo. Biochim. Biophys. Acta 1523:117–122; 2000. [54] Galati, G.; Moridani, M. Y.; Chan, T. S.; O'Brien, P. J. Peroxidative metabolism of apigenin and naringenin versus luteolin and quercetin: glutathione oxidation and conjugation. Free Radic. Biol. Med. 30:370–382; 2001. [55] Cuddihy, S. L.; Parker, A.; Harwood, D. T.; Vissers, M. C.; Winterbourn, C. C. Ascorbate interacts with reduced glutathione to scavenge phenoxyl radicals in HL60 cells. Free Radic. Biol. Med. 44:1637–1644; 2008. [56] Long, L. H.; Hoi, A.; Halliwell, B. Instability of, and generation of hydrogen peroxide by, phenolic compounds in cell culture media. Arch. Biochem. Biophys. 501: 162–169; 2010.

2327

[57] Khalife, K. H.; Lupidi, G. Reduction of hypervalent states of myoglobin and hemoglobin to their ferrous forms by thymoquinone: the role of GSH, NADH and NADPH. Biochim. Biophys. Acta 1780:627–637; 2008. [58] Spencer, J. P.; Abd-el-Mohsen, M. M.; Rice-Evans, C. Cellular uptake and metabolism of flavonoids and their metabolites: implications for their bioactivity. Arch. Biochem. Biophys. 423:148–161; 2004. [59] Stone, J. R.; Yang, S. Hydrogen peroxide: a signaling messenger. Antioxid. Redox Signal. 8:243–270; 2006. [60] Shirai, M.; Yamanishi, R.; Moon, J. H.; Murota, K.; Terao, J. Effect of quercetin and its conjugated metabolite on the hydrogen peroxide-induced intracellular production of reactive oxygen species in mouse fibroblasts. Biosci. Biotechnol. Biochem. 66:1015–1021; 2002. [61] Groeger, G.; Quiney, C.; Cotter, T. G. Hydrogen peroxide as a cell-survival signaling molecule. Antioxid. Redox Signal. 11:2655–2671; 2009. [62] Hultqvist, M.; Olsson, L. M.; Gelderman, K. A.; Holmdahl, R. The protective role of ROS in autoimmune disease. Trends Immunol. 30:201–208; 2009.