lable at ScienceDirect

Biochimie 95 (2013) 2404e2414

Contents lists avai

Biochimie

journal homepage: www.elsevier .com/locate/b iochi

Research paper

Ethyl gallate attenuates acute lung injury through Nrf2 signaling

Kamiya Mehla, Sakshi Balwani, Anurag Agrawal, Balaram Ghosh*

Molecular Immunogenetics Laboratory, CSIR-Institute of Genomics and Integrative Biology, Mall Road, Delhi 110007, India

a r t i c l e i n f o

Article history:Received 31 December 2012Accepted 30 August 2013Available online 6 September 2013

Keywords:LPSInflammationMonocytesAntioxidantsNrf2

* Corresponding author. Tel.: þ91 11 27662580; fa27416489.

E-mail address: bghosh@igib.res.in (B. Ghosh).

0300-9084/$ e see front matter � 2013 Elsevier Mashttp://dx.doi.org/10.1016/j.biochi.2013.08.030

a b s t r a c t

Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) is the clinical syndrome ofpersistent lung inflammation caused by various direct and indirect stimuli. Despite advances in theunderstanding of disease pathogenesis, few therapeutic have emerged for ALI/ARDS. Thus, in the presentstudy we evaluated the therapeutic potential of ethyl gallate (EG), a plant flavanoid in the context of ALIusing in vivo (BALB/c) and in vitromodels (human monocytes). Our in vivo data supports the view that EGalleviates inflammatory condition in ALI as significant reduction in BALF neutrophils, ROS, proin-flammatory cytokines and albumin levels were observed with the single i.p of EG post LPS exposure. Also,histochemical analysis of mice lung tissue demonstrated that EG restored LPS stimulated cellular influxinside the lung airspaces. Unraveling the mechanism of action, our RT-PCR and western blot analysissuggest that enhanced expression of HO-1 underlies the protective effect of EG on ROS level in mice lungtissue. Induction of HO-1 in turn appears to be mediated by Nrf2 nuclear translocation and consequentactivation and ablation of Nrf2 activity through siRNA notably abrogated the EG induced protective effectin LPS induced human monocytes. Furthermore, our results indicate that EG generated moderateamounts of H2O2 could induce Nrf2 translocation in the in vitro systems. However, given the insignificantamount of H2O2 recorded in the injected material in the in vivo system, additional mechanism for EGaction could not be excluded. Nevertheless our results highlight the protective role of EG in ALI andprovide the novel insight into its usefulness as a therapeutic tool for the treatment of ALI.

� 2013 Elsevier Masson SAS. All rights reserved.

1. Introduction

Acute lung injury (ALI) represents a state of diffuse and het-erogeneous lung injury, which often leads to impaired gas ex-change and respiratory failure in critically ill patients [1,2]. Itsetiology is diverse and despite advances in therapeutic modalities,non-ventilatory interventions remain elusive. It is convincible thatoxidative stress and inflammatory pathways are intimatelyinvolved in the progression and outcome of ALI. Interestingly,drastic imbalance of oxidant/antioxidant pathways has even shownto confer mortality in some cases. To counterbalance the delete-rious effect of reactive oxidants, a number of cytoprotectivemechanism come into their action as an adaptive cellular response.Nrf2 (Nuclear factor E2 P45-related factor 2) is one such cellularsensor, which maintains redox homeostasis under oxidative insultsthrough coordinated induction of a large battery of antioxidantenzymes. This is supported by the observation that Nrf2 knockoutcardiomyocytes are significantly more susceptible to H2O2 induced

x: þ91 11 27667471, þ91 11

son SAS. All rights reserved.

cell injury compared to wild type cells [3]. In addition, Nrf2-deficient mice display increased sensitivity to various oxidative aswell as inflammatory insults [4,5]. Thus from a therapeutic stand-point, activation of Nrf2 regulated cytoprotective pathways throughexogenous molecules may provide potential benefits in the treat-ment of ALI.

A number of plant derived molecules such as curcumin, caffeicacid phenyl esters, catechins and resveratrol [6] have been shownto induce Nrf2 regulated antioxdiant enzyme, HO-1 expressionin vitro. However, in vivo activity of these metabolites against ALIhas not been investigated. Hence the present study was designed toevaluate the protective effect of a plant polyphenol, ethyl gallate(EG) for treatment of ALI. Our results indicate the amelioration ofLPS induced inflammation in EG treated mice. In addition, our dataalso provides an insight into the possible role of HO-1/Nrf2signaling in mediating the protective effect of EG.

2. Material and methods

2.1. Materials

Lipopolysaccharide (LPS, from Escherichia coli Serotype: 027:B8)and H2DCF-DA, M-199, L-glutamine, antibiotic and antimycotic

K. Mehla et al. / Biochimie 95 (2013) 2404e2414 2405

solution, endothelial cell growth factor (ECGF), trypsin, HEPES,glutathione and myoglobin from horse skeletal muscle were pur-chased from Sigma Chemical Co., USA. Anti-ICAM-1, anti-VCAM-1,anti-GR1þ and TNF-a ELISA kit were purchased from BD Phar-mingen, USA. Anti-Nrf2, antieHOe1, anti-Lamin-B1 and anti-b-actin were purchased from Santa Cruz biotechnology, USA. Anti-Keap-1 was purchased from Abcam, USA. Fetal calf serum wasobtained from Biological Industries, Israel. IL-1b and MIP-2a ELISAkits were purchased from R&D systems, USA. Albumin ELISA kit waspurchased from Geneway, USA. Oxiselect-hydrogen peroxide kit(STA-343) was purchased from cell biolabs.

2.2. Procurement of plant material and isolation of ethyl gallate

Dried galls of Pistacia integerrimawere purchased from the localmarket, was identified and validated by Dr. H.B. Singh, Head, RawMaterials Herbarium and Museum, National Institute of ScienceCommunication and Information Resources (NISCAIR), Delhi.Specimen sample has been deposited in the herbarium, voucher no.NISCAIR/RHMD/Consult/2010-11/1451/49. Ethyl gallate was iso-lated from the Pistacia integerrima following the previous protocols[7]. The purity of isolated compound was determined on HPLC bycomparing with commercially available ethyl gallate.

2.3. Animals and development of mouse model of ALI

Male BALB/c mice (age 8� 10 weeks, weighing 18e20 gm) wereobtained from National Institute of Nutrition, Hyderabad, India andacclimatized for 1 week under standard laboratory conditionsbefore starting the experiments. The Institutional Animals EthicsCommittee approved the experimental protocol. ALI model wasdeveloped by exposing BALB/c mice to aerosolized LPS dissolved insterile 0.9% NaCl for 30 min in a plexiglass chamber as discussedpreviously [8]. The aerosol was generated by a nebulizer (OmronCX3, Japan) with an airflow rate of 9 l/min. Control mice however,were exposed to sterile 0.9% NaCl aerosol only.

2.4. Drug treatment

Post aerosol challenge, mice were divided into groups (n ¼ 6)and named according to challenge/treatment: Saline/VEH, [salinechallenged and vehicle treated (0.5% DMSO)], LPS/VEH, [LPS chal-lenged and vehicle treated (0.5% DMSO)], LPS/Rolipram, [LPSchallenged and Rolipram treated (10 mg/kg), LPS/EG, [LPS chal-lenged and EG treated (at different doses i.e 1, 5, 10 mg/kg)]. i.pinjection of VEH/Rolipram/EG was chosen as a route of adminis-tration post 30 min of LPS challenge.

2.5. Pulmonary mechanics and bronchoalveolar lavage fluid (BALF)collection

Four hour after i.p injection, mice were sedated and pulmonarymechanics were determined by invasive method using the flex-iVent system (Scireq), as described previously [9]. Immediatelyafter pulmonary mechanics measurement, mice were sacrificed,bronchoalveolar lavage was performed and analyzed for ROS pro-ducing neutrophils and total cell count as described previously [9].

2.6. Lung histological examination and lung injury score

The lung tissues were formalin-fixed and stained withhematoxylin-eosin (H&E) to assess airway inflammation asdescribed earlier [9]. Also, lung injury score was assessed, usingsemiquantitative histopathologic scoring system from previousstudies [10].

2.7. Assays for cytokines, chemokines and MPO activity

Levels of TNF-a, IL-1b and MIP-2awere quantified in BALF usingspecific ELISA kits, according to the manufacturer’s instructions.MPO activity was measured in BALF cell lysate according to man-ufacturer’s kit protocol (Cayman chemicals, USA). MPO activity wasexpressed as nmol/min/ml of the cell lysate.

2.8. Hydrogen peroxide detection

Hydrogen peroxide levels were determined in the cell culturesupernatant as per the manufacturer’s kit protocol (H2O2 colori-metric kit).

2.9. Real time-polymerase chain reaction (RT-PCR)

The transcript levels of antioxidant enzymes were determinedby real-time RT-PCR in a LightCycler (Roche Diagnostics) by usingSYBR Green I (Sigma) method. The reaction mixture consisted ofFastStart DNA SYBR Green I master mix, forward and reverseprimers and cDNA. The mRNA levels were quantified by the stan-dard curve method by using a serially diluted standard templateand were normalized to the mRNA of 18S in each sample.

SOD1: Sense 50-GCGGTGAACCAGTTGTGTTGTC-30

SOD1: Antisense 50-CAGTCACATTGCCCAGGTCTCC-30

SOD2: Sense 50-TGGCTTGGCTTCAATAAGGA-30

SOD2: Antisense 50-AAGGTAGTAAGCGTGCTCCCACAC-30

SOD3: Sense 50-AGGTGGATGCTGCCGAGAT-30

SOD3: Antisense 50-TCCAGACTGAAATAGGCCTCAAG -30

HO-1: Sense 50-AACAAGCAGAACCCAGTCT-30

HO-1: Antisense 50-CCTTCTGTGCAATCTTCTTC-30

18S: Sense 50-GTAACCCGTTGAACCCCATT-30

18S: Antisense 50-CCATCCAATCGGTAGTAGCG-30

2.10. Preparation of cytoplasmic (CE) and nuclear extracts (NE) andwestern blot analysis

Cytoplasmic and nuclear extracts were prepared from themouse lung tissue and human monocytes as described earlier [11].The protein concentration of the extract was determined using BCA(Bicinchoninic acid) method. For western blot analysis, protein wasresolved on SDS-PAGE followed by transfer on PVDF membrane.The membranes were probed with specific antibodies to HO-1,Nrf2, Lamin-B1 and Keap-1 followed by detection with HRPecon-jugated secondary antibody. Blot was finally developed using 3,30-Diaminobenzidine (DAB) substrate. Signals were detected by spotdensitometry using Alpha EaseFC software from Alpha Innotech.

2.11. Electrophoretic mobility shift assay (EMSA)

To determine Nrf2 activation, electrophoretic mobility shiftassay (EMSA) was performed as described previously [11].

2.12. Immunocytochemistry

THP-1 cells were seeded on the poly lysine coated coverslip,which were grown in 75-mm culture flasks. Cells were treated withor without 30 mM of EG for 2 h. Following induction, cells werewashed with phosphate-buffered saline, fixed with para-formaldehyde (4% in PBS, 15 min) and blocked with blocking buffer(1% goat sera in PBS, containing 0.5% triton A �100 and 0.01%Tween 20) for additional 30 min. Following incubation, cells werewashed with phosphate-buffered saline and permeabilized with

Fig. 1. Ethyl gallate attenuates LPS induced lung injury in BALB/c mice: (A) H&E staining of lung tissue section. Post sacrifice, mice lung tissues were fixed in 10% formalin andparaffin embedded sections were prepared and stained with H&E. a, Alveolus; b, bronchus; v, vessel. Broken arrow indicates cellular infiltration. (B) Lung mean injury score. (C)Airway hyperresponsiveness in LPS challenged mice. Airway hyperresponsiveness was determined by invasive method post 4 h of LPS challenge. Histological data is representativeof three independent experiments where n ¼ 6 mice/group (H&E staining, X20). Pulmonary mechanics and Lung injury data was expressed as mean � S.E.M, *P < 0.05 vs. salinecontrol group, #P < 0.05 vs. LPS group and ##P < 0.01 vs. LPS group.

K. Mehla et al. / Biochimie 95 (2013) 2404e24142406

0.5% triton�100 for 10min. Cells werewashed again and incubatedwith anti-Nrf2 (1:500, in 1% goat sera) for 1 h at 37 �C followed bywashing and incubation with FITC-conjugated secondary antibody(1:250) for 1 h in dark. After incubation, cover slips were washedand mounted on glass slides using Prolong Gold antifade reagentwith DAPI (Invitrogen, USA). Fluorescent images were obtained byusing Nikon Eclipse TI, Nikon USA. Control groups were treatedwith vehicle control and stained with rabbit IgG isotypic controls.

2.13. siRNA transfections

Predesigned siRNA against human Nrf2 (Cat no: L-003755-00-0005) and control-scrambled siRNA (Cat no: CN-002000-01-05)were purchased from Dhramacon, USA. THP-1 cells were platedat a density of 3 � 104 cells per 60 mm dish. Cells were transfectedwith 50 nM Nrf2 or scrambled siRNA using HiPerFect transfectionreagent (Qiagen, USA). After 6 h incubation, fresh medium wasadded and the cells were cultured for another 36 h. The cells werethen treated with or without 1 mg/ml of LPS for 4 h followed bytreatment with or without EG (30 mM) for additional 6 h. Afterincubation, cells and culture supernatants were collected sepa-rately. Culture supernatants were assayed for TNF-a releasewhereas total cells lysates were prepared from the harvested cellsand analyzed for Nrf2 and HO-1 expression.

2.14. Statistical analysis

Results are expressed as mean � SE with n indicating thenumber of experiments. Statistical analysis was done using GraphPad Prism software version 4.03 (Graph Pad Software Inc., La Jolla,CA). Differences between the two groups were assessed by un-paired student’s t-test. However for multi-groups, statistical anal-ysis was performed using one-way ANOVA followed by Bonferronipost test. p values <0.05 were considered significant (*).

3. Results

3.1. Ethyl gallate reduced LPS induced lung injury in BALB/c mice

LPS aerosol has previously been shown to mediate structuralchanges in the mice lung tissue [12]. These alterations includeintra-alveolar edema, hemorrhage and neutrophil accumulation inand around perivascular and alveolar regions. To determine theextent of LPS induced injury in our model, lower lobe of right lungwas taken for histology. Histopathological evaluation was per-formed using semiquantitative scoring system based on the pre-vious studies [10]. Representative lung sections for saline/VEH andLPS/VEH mice are shown in Fig. 1A. No histopathological abnor-malities were present in the saline challenged mice. LPS challengewas associated with marked neutrophil infiltration along withalveolar hemorrhage, which resulted in significantly increased lunginjury scores. Treatment of LPS challenged mice with EG, wasassociatedwith significant reduction of neutrophilic infiltration in adose dependent manner. LPS mice treated with EG (10 mg/kg)showed maximal effect, as revealed by notable lower lung injuryscore than the mice treated with VEH alone, implying the resolu-tion of inflammatory response with EG treatment (Fig. 1B).

3.2. EG improves airway hyperresponsiveness in endotoxinchallenged mice

Previously, inhalation and exposure studies have shown theeffect of bacterial endotoxin on airway hyperresponsiveness inhuman and mouse lungs. LPS dose as low as 0.3 mg/ml is sufficientto trigger methacholine induced bronchoconstriction post 3e4 hexposure [13]. Hence to determine the protective effect of EG onLPS induced bronchoconstriction, we measured airway resistance(R), an indicator of airway constriction, during mechanical venti-lation. LPS challenged mice had increased R compared to saline

Fig. 2. EG attenuates inflammatory indices in BALF of LPS challenged BALB/c mice: (A) Dot plot for ROS producing neutrophils in mice BALF. A proportion of BALF, collected post 4 hof LPS challenge, was analyzed on FACS for granulocytes accumulation and ROS generation by double staining method. (B) Histogram plot for total number of inflammatory cells.Total cell count was obtained with hemocytometer based cell counting method in mice BALF. (C) Histogram plot for proinflammatory cytokines/chemokine levels in BALF. Levels ofproinflammatory cytokines (TNF-a, IL-1b) and chemokine (MIP-2a) were determined by ELISA in mice BALF samples. (D) Histogram plot for MPO activity. (E) Histogram plot foralbumin level in mice BALF. Albumin level, marker of vascular leakage was determined by ELISA. Dot plot is a representative of three independent experiments where n ¼ 6 mice/group. Data was expressed as mean � S.E.M., *P < 0.05 vs. saline control group, #P < 0.05 vs. LPS group. ##P < 0.01 vs. LPS group.

K. Mehla et al. / Biochimie 95 (2013) 2404e2414 2407

challenged group. Conversely, EG (10 mg/kg) treatment relievedheightened airway resistance in LPS challenged mice, demon-strating the protective effect of EG on the lung tissue integrity(Fig. 1C).

3.3. EG attenuates inflammatory indices in bronchoalveolar lavagefluid of LPS challenged BALB/c mice

Transendothelial and transepithelial migration of neutrophil isanother important event in endotoxic induced ALI. Kinetic studiesof neutrophil migration in LPS induced injury have shown thatgranulocytes trafficking in lung parenchyma begin as early as by2 h, with significant accumulation in BALF by 4 h [14,15]. In oursystem also, we analyzed BALF for the levels of total cell population,neutrophil accumulation, ROS and proinflammatory markersfollowing LPS challenge. The total number of neutrophils gener-ating reactive oxygen species in BALF was analyzed by flow cyto-metric double staining using APC-labeled GR-1 monoclonalantibody and ROS specific fluorescent dye, H2DCFDA. Flow cyto-metric analysis of BALF from LPS treated group demonstrated thatapproximately 87% of cells were double positive cells, indicating

intense neutrophilia. Interestingly, dose response studies with EG(1e10 mg/kg) administered 30 min post LPS showed notablereduction of neutrophils in BALF with the maximum effect at thehighest dose (Fig. 2A). Next, we determined the total number ofinflammatory cells. In our study, BALF from LPS challenged groupshowed significant cell influx as compared to saline treated group(p < 0.001) (Fig. 2B). Among the various cells, neutrophilsaccounted for the major proportion (data not shown). In addition toROS, neutrophils are also the vital source of tissue damagingenzyme, myeloperoxidase (MPO), which is often used as a markerof neutrophil accumulation in the tissue sample [13]. The MPOactivity in BALF from LPS challenged group was significantly high,which correlatedwith the high level of neutrophils. However, in theEG treated group a significant decrease in MPO activity wasobserved (Fig. 2D).

Neutrophil generated ROS and proteases orchestrate the tissuedamaging effect by oxidizing membrane phospholipids, whichimpairs the membrane integrity and leads to increased vascularpermeability for the serum proteins [16]. In ALI, vascular perme-ability is assessed by albumin level measurement in the BALFsamples. Encouraged by our previous observations, protective

K. Mehla et al. / Biochimie 95 (2013) 2404e24142408

effect of EG on vascular leakage was explored by determining thealbumin levels in mice BALF. In our study, high albumin level,indicating vascular injury, was observed in LPS challenged group ascompared to saline group. However, EG treatment restored theelevated levels of albumin to normal in a dose dependent manner(Fig. 2E).

In addition to albumin, level of proinflammatory cytokines andchemokines were also measured in the BALF. Elevated level of TNF-a, IL-1b and MIP-2a were recorded in LPS challenged mice. Incontrast, post treatment with EG significantly reduced the level ofTNF-a and IL-1b by 64 � 11% and 59 � 3.4% respectively {p < 0.05,LPS vs. EG (10 mg/kg)} (Fig. 2C) Similarly, MIP-2a, was found to benotably down regulated by EG (Fig. 2C) in a dose dependentmanner.

3.4. EG upregulates HO-1 expression in LPS challenged mice

Because EG administration reduced the neutrophil counts andassociated mediator in the mice BALF, the status of various antioxi-dant enzymes were assessed in mice lung homogenate. Superoxidedismutase (SOD) family of antioxidants represents the first line ofdefense against ROS [17]. Transcript levels of SOD family isoformswere assessed by real time quantitative PCR in the mice lung tissue.Transcript levels were normalized with the 18S. Levels of SOD1mRNAwere found to beminimally affected in all the cases however;a marginal increase in SOD2 was observed in LPS challenged groupas compared to EG treatment group. In contrast with SOD2, posttreatmentwith EGmoderately up regulated the level of SOD3,majorisoform present in the lung, in a dose dependent manner (Fig. 3A).Besides SOD, HO-1 is another protective antioxidant enzyme, whichhavebeen shown to reduce the levels of TNF-a and IL-1b in ratmodelof endotoxemia [18]. As marked reduction of these cytokines werealso observed with the EG treatment, status of HO-1 enzyme wasinvestigated in the same experimental sets. Our real time data,Fig. 3A, indicated marked up regulation of HO-1 transcripts in EG

Fig. 3. EG upregulates HO-1 expression in LPS challenged mice: (A) Transcript levels of antidetermined by real-time quantitative PCR. The mRNA levels of genes were normalized with bdetermined in lung cytoplasmic extract by western blot. b-actin was used as a loading corepresentative of three independent experiments where n ¼ 6 mice/group. Data is express

treated group as compared to LPS and saline challenged groups.Worth to notice that significant difference in HO-1 level betweencontrol and LPS challenged group could not be observed. In previousreports, LPS was shown to markedly induced HO-1 in mice, thoughat the later time points (12 h post LPS exposure) [19]. This suggeststhat in our acutemodel (4e5 h), pro-inflammatory signals dominateand counteractive anti-oxidant signals are ineffective in driving theprotection in LPS challengedmice. Moreover, western blot data alsorevealed similar profile for HO-1 level in the control and LPS chal-lengedmice. Interestingly, EG treatedmice (5 and10mg/kg) showedhigh levels of HO-1 (Fig. 3B and C).

3.5. EG enhances nuclear accumulation of Nrf2

Nrf2, a redox sensitive transcription factor plays a central role intranscriptional regulation of HO-1 [20]. Under homeostasis, nuclearlevel of Nrf2 is tightly regulated by its constant degradation insidethe cytoplasm. To evaluate the possible involvement of Nrf2 in EGinduced HO-1 induction, nuclear level of Nrf2 was quantified infreshly harvested mice lung tissue. As with the HO-1, no significantdifference in the Nrf2 levels were observed between LPS challengedmice and sham mice (data not shown) and hence results werecompared only in relation with control group. In our study EGtreatment significantly increased Nrf2 level in a dose dependentmanner (Fig. 4A, B). A similar observation was made regarding theNrf2 DNA binding activity in lung nuclear extracts. Electrophoreticmobility shift assay (EMSA) demonstrated significant Nrf2 bindingactivity in EG treated mice (5 and 10 mg/kg) as compared to salinecontrol groups (Fig. 4C; lane 5 vs. 1). Specificity of the bindingcomplex was determined by competitive experiments with anexcess amount of unlabeled Nrf2 oligonucleotides. A 100-foldexcess of unlabeled Nrf2 nucleotides completely abolished theshifted bands (lane 7). However, no such disappearance wasobserved when excessive amount of unlabeled NF-kB oligonucle-otides were used (lane 8).

oxidant enzymes. Transcript levels of SOD1, SOD2, SOD3, and HO-1 in lung tissue were-actin mRNA. (B) Protein levels of HO-1 in mice lung tissue. Protein levels of HO-1 werentrol in protein measurement. (C) Densitometry quantitation of HO-1 levels. Data ised as mean � S.E.M., *P < 0.05 vs. saline control group, **P < 0.01 vs. control group.

Fig. 4. EG augments nuclear accumulation of Nrf2 in LPS challenged mice: (A) Western blot represent nuclear level of Nrf2. Nrf2 level was determined in the nuclear extractprepared from freshly harvested mice lung tissue. (B) Densitometry quantitation of Nrf2. (C) EMSA represents Nrf2 activity. Nrf2 activity was analyzed in the mice lung nuclearextracts. Specificity of the band was determined using an excess amount of unlabeled oligonucleotides in the respective gels. Lane 1, free probe; lane 2, sham mice; lane 3, LPSchallenged and VEH treated mice; lane 4, LPS challenged and EG (1 mg/kg) treated mice; lane 5, LPS challenged and EG (10 mg/kg) treated mice; lane 6, LPS challenged and EG(5 mg/kg) treated mice; lane 7, cold chase with Nrf2 oligonucleotides; lane 8, cold chase with Nf-kB oligonucleotides. (D) Western blot represents Keap-1 levels in mice lungcytoplasmic extract. (E) Densitometry quantitation of Keap-1. Data is representative of three independent experiments where n ¼ 6 mice/group. Data was expressed asmean � S.E.M. *P � 0.05 vs. saline control group.

K. Mehla et al. / Biochimie 95 (2013) 2404e2414 2409

Under normal condition, Nrf2 is sequestered in cytoplasm,bounded by cytoskeleton associated Keap-1 protein. Nrf2 can betranslocated into the nucleus to bind to the antioxidant responseelement (ARE) when Nrf2-Keap-1 complex is destabilized undersome cellular stimuli. This suggests that EG induced Nrf2 activationmay be linked to the relative level of Nrf2-Keap-1 complex, whichultimately depends on cytosolic Keap-1. Thus the effect of EG onsteady-state level of Keap-1 in lung cytoplasmic extract was stud-ied. As shown in Fig. 4D, E, EG treatment significantly decreasedKeap-1 protein in the lung samples. This indicates that the pro-found influence of EG on Nrf2 translocation is mediated by thereduction of Keap-1 protein level.

3.6. EG upregulates HO-1 expression via Nrf2 in human monocytes

In light of the capacity of EG in inducing HO-1 level in mice lungtissue, we explored the efficacy of EG in vitro. Monocytes play a keyrole in orchestrating the immune response during LPS inducedacute lung injury as it initiates the neutrophil influx events by

releasing an arrays of cytokines and chemokines [21]. Hence, hu-man monocytic cell line, THP-1 was chosen as in vitro model tomimic the inflammatory condition of LPS induced acute lung injury.To test the influence of EG on HO-1 protein, we determined HO-1protein expression using western blot analyses. EG (30 mM)significantly increased HO-1 protein level in a time and concen-tration dependent manner (Fig. 5AeC). We next examined nuclearNrf2 in the similar experimental sets. In THP-1, as shown in Fig. 5EeG, EG treatment increased nuclear Nrf2 level in a time and con-centration dependent manner. The highest induction of Nrf2 levelwas 3.7 fold with EG (Fig. 5H). To validate our western blot findings,we also used immunofluorescence staining for Nrf2 translocationin human monocytes. In untreated monocytes, Nrf2 displayeddiffused cytoplasmic staining and very low, almost undetectablestaining inside the nucleus. However, 2 h after the addition of EG(30 mm), Nrf2 displayed intense cytoplasmic as well as nuclearstaining compared to the control cells (Fig. 6). This suggests that EGtreatment induced rapid translocation of Nrf2 in the humanmonocytes. Finally, protein level of Keap-1 was also investigated in

Fig. 5. EG upregulates HO-1 expression via Nrf2 in human monocytes, THP-1: (A&C) Western blot display time and concentration dependent increase in HO-1 level by EG. THP-1cells were incubated with EG for the indicated concentration and time points and analyzed for HO-1 levels in total cell lysates. Protein level of HO-1 was normalized with b-actinlevels. (B&D) Densitometry quantitation of HO-1 levels in the respective blots. (E&G) Western blot display time and concentration dependent nuclear accumulation of Nrf2. THP-1cells were incubated with EG for the indicated concentration and time period and nuclear lysates were prepared and analyzed for Nrf2 levels. Under the similar condition, Lamin-B1was used as a loading control. (F&H) Densitometry for Nrf2. (I) Western blot represents time dependent effect on Keap-1 level by EG. THP-1 cells were incubated with EG (30 mM) atdifferent time points and analyzed for Keap-1 levels in cytoplasmic extract. (J) Densitometry for Keap-1 level. Data was expressed as mean � S.E.M., *P < 0.05 vs. control group,**P < 0.01 vs. control group.

K. Mehla et al. / Biochimie 95 (2013) 2404e24142410

the EG treated THP-1 cells. Our results demonstrated time depen-dent reduction in Keap-1 level with restoration at the later timepoints (Fig. 5I, J). Observed effect of EG on the steady state level ofKeap-1 could be either due to its modification or degradation.

As discussed earlier, monocyte primarily controls the inflam-matory response elicited by LPS in ALI by producing TNF-a and IL-1b. To further confirm the potential role of HO-1-Nrf2 signaling inEG mediated protective effect in ALI, we assessed the therapeuticeffect of EG in LPS stimulated THP-1 cells, mimicking the in vivomodel of ALI. In this system, TNF-a level was measured in LPSstimulated human monocyte in the presence and absence of EG. Asshown in Fig. 7A, LPS treatment resulted inmarked release of TNF-aas compared to the untreated cells. However, this effect was notably

abrogated in the presence of EG in a dose dependent manner(p< 0.05). The effect of EG on TNF-awas appeared to be dependenton the HO-1 as significant increase in HO-1 proteinwas observed inthe EG treated monocytes which correlated with decreased TNF-alevel (Fig. 7B and C). Thus, mirroring the in vivo model of ALI, ourresult further verify the role of HO-1-Nrf2 signaling in EG mediatedprotective effect in inflammatory response.

3.7. Silencing Nrf2 augments LPS-induced inflammatory responsesin human monocytes

To further corroborate the role of Nrf2 in EG induced protectiveeffects, we used Nrf2 siRNA to silence the Nrf2 protein and

Fig. 6. EG promotes nuclear accumulation of Nrf2 in human monocytes: Immunocytochemistry depicts nuclear translocation of Nrf2 in the presence of EG. THP-1 cells were grownon sterile cover slips and exposed to 30 mM of EG for 2 h. After treatment, cells were examined for Nrf2 levels and visualized under a fluorescent microscope. White arrows representNrf2 staining inside and outside nuclei (blue color). The results are shown as a representative of at least three independent experiments. Data is representative of three independentexperiments.

K. Mehla et al. / Biochimie 95 (2013) 2404e2414 2411

determined the effect of EG on LPS induced THP-1 cells. THP-1 cellswere either untransfected or transfectedwith control siRNA or withNrf2 siRNA for 36 h.Western blot analysis of treated samples showsthat Nrf2 siRNA caused significant knockdown of Nrf2 protein by36 h compared to the scrambled siRNA. We also determined theexpression of HO-1, where samples treated with Nrf2 siRNAshowed partial but significant inhibition of HO-1 (Fig. 7D, E). Toconfirm that Nrf2-HO-1 signaling mediated by EG is responsible forcountering the LPS induced inflammation; THP-1 cells weretransfected with Nrf2 siRNA followed by treatment with LPS andEG. As shown in Fig. 7F, treatment with Nrf2 siRNA enhanced theTNF-a secretion form human monocytes post LPS exposure ascompared to scrambled siRNA treated group. However, this in-crease in TNF-a level could not be attenuated by the addition of EG;indicating that EG mediated HO-1/Nrf2 signaling is required forinhibition of LPS induced inflammatory responses in the humanmonocyte.

3.8. EG mediated H2O2 generation underlies Nrf2 translocation inTHP-1 monocytes

As per the accumulating evidences, activation of Nrf2 bymost ofpolyphenols in part is mediated through their autoxidation andconsequent H2O2 production in the cell culture medium [33e35].To assess the same, we measured the H2O2 levels in the EG treatedcells and medium alone using the H2O2 detection kit. EG treatedmedium-alone increased the amounts of H2O2 as compared to theuntreated medium (Fig. 8A). However, the sharp increase in H2O2was found retrograde in the presence of EG treated THP-1 mono-cytes. This suggests that H2O2 generated by EG was quenched bythe detoxification system activated by Nrf2 in EG treated mono-cytes. Furthermore, we measured the Nrf2 activation potential ofEG in the presence of peroxide radical scavengers such as gluta-thione and myoglobin. As shown in Fig. 8B, EG treatmentaugmented Nrf2 localization inside the nucleus, however this in-crease was mitigated in the presence of glutathione and myoglobinalone (Fig. 8B). Moreover, the observed effect on Nrf2 was morepronounced in cells treated with glutathione and myoglobintogether. Our finding thus suggests that EG generated H2O2 pro-motes Nrf2 translocation in THP-1 monocytes (Fig. 8B). H2O2generating potential of EG was further explored in the injectablematerial which comprise of water with 0.5% DMSO. As injectablematerial is always freshly prepared, 10e20 min time points were

taken in consideration. Interestingly, under this medium and timepoints range negligible amounts of H2O2 could be detected (Fig. 8C).Moreover, it is not known whether the small amounts of H2O2generated in the injectable medium can last long inside the animal.Although, this needs further investigation, our studies demon-strated the role of antioxidant potential of EG in triggering Nrf2activation under in vivo scenario.

4. Discussion

Over the decades acute lung injury (ALI)/ARDS has been a majorcause of respiratory failure in critically ill patients. Since it wasdiscovered, range of pharmacological intervention has been fol-lowed to improve the outcome in ALI/ARDS patients. However,efficacy and safety of many drugs due to unwanted side effects stillposes a challenge for effective treatment of ALI. As the use of nat-ural products in drug development is becoming increasingly pop-ular, in the current study we explored the therapeutic potential of anatural molecule, Ethyl gallate (EG) from Pistacia integerrima, in thein vivo model of LPS induced ALI. Previously, EG has been shown tolimit LDL oxidation and cell adhesion molecule expression in thein vitromodel of atherosclerosis and as these events are common inacute lung injury; beneficial effect of EG in ALI model was expected.In accordance with our assumption, our study has demonstratedthat EG provided significant protection against LPS inducedinflammation in mice.

The current study employed the well established LPS model ofALI in BALB/c mice where, 30 min LPS aerosol (0.3 mg/ml) evokedinflammatory response within 4 h of challenge and marked levelsof proinflammatory cytokines (TNF-a, IL-1b), neutrophilic infiltrateand ROS production were observed in the mouse BALF samples.Evaluating the protective effect of EG, our study showed that asingle i.p (intraperitoneal) of EG given 30 min after LPS challengeremarkably suppressed neutrophilic counts and associated MPOactivity in BALF samples. In addition, a concomitant decrease inspontaneous oxidative burst (ROS production) and airway hyper-reactivity was also observed in the EG treated mice. The plausibleexplanation for this observation is that EG reduces airway sensi-tivity towards methacholine by regulating neutrophil influx and itsspontaneous oxidative burst, as a close relationship between earlyAHR and neutrophil sequestration in the lung parenchyma wasreported earlier [13]. Besides, ROS is itself known to evoke airwayhyperreactivity [22]. We suggest that quenching effect of EG on ROS

Fig. 7. Silencing Nrf2 augments LPS-induced inflammatory responses in human monocytes: (A) Histogram represents TNF-a level in THP-1 cells. THP-1 cells were stimulated withLPS (1 mg/ml) in the presence and absence of EG (30 mM) for 6 h. Post incubation, cells were harvested and supernatants were collected and analyzed for TNF-a levels by ELISA. (B)Western blot represent HO-1 levels in THP-1 cells. THP-1 cells were stimulated with LPS (1 mg/ml) in the presence and absence of EG (30 mM) for 6 h. Post incubation, cells wereharvested and analyzed for HO-1 level. (C) Densitometry for HO-1. (D) Western blots represent Nrf2 and HO-1 level in Nrf2 siRNA transfected THP-1 cells. THP-1 cells weretransfected with or without 50 nM of either control or Nrf2 siRNA for 36 h. Following transfection, cells were treated with or without LPS for 2 h and EG (30 mM) for additional 6 h.Post incubation, total cell lysates were prepared and Nrf2 and HO-1 levels were determined by western blot. (E) Levels of Nrf2 and HO-1 were normalized with b-actin protein. (F)Histogram represents TNF-a level in the untransfected and transfected cell supernatants. Levels of TNF-a were measured post 36 h by ELISA. Data was expressed as mean � S.E.M.,*P < 0.05 vs. LPS group, **P < 0.01 vs. LPS group.

K. Mehla et al. / Biochimie 95 (2013) 2404e24142412

level could be underlying its protective effect on early AHR.Consistent with our findings, a similar report has shown the pro-tective effect of EG on histamine contracted guinea pig’s trachea[23]. Interestingly, protective effect of EG on BHR and airway neu-trophilia is particularly striking when compared with dexametha-sone, a gold standard inmany pharmacological studies. Only partialinhibition of BHR and MPO titers has been reported with dexa-methasone at 1 mg/kg, higher dose being ineffective [13]. Similarly,budesonide, and NAC were ineffective in preventing LPS inducedbronchoconstriction in endotoxemic rats [24]. EG inhibited neu-trophils counts were further associated with reduced proin-flammatory cytokines/chemokines (TNF-a, IL-1b, MIP-2a) levelsand permeability marker, albumin in mice BALF samples. Thesefindings strongly points to the protective role of EG as high plasmalevels of IL-8 (Human analog of mouse MIP-2a), TNF-a and IL-1bhave shown to be correlated with increased mortality in septicpatients [25]. These finding therefore suggest that EG amelioratesALI by preventing neutrophil sequestration and associated tissueinjury.

Among the possible mechanism of action for EG, antioxidantproperty of this molecule was presumed to underlie its protectiveeffect, since ROS have widely been shown to activate NF-kB, whichcontributes to the expression of inflammatory mediators in thelungs of endotoxemic or hemorrhagic mice. Though a plethora ofreports has shown the radical scavenging property of EG, our in vivoand in vitro data clearly demonstrates the involvement of HO-1, animportant antioxidant enzyme for the first time. HO-1 expressionhas proven to be associated with the regulation of inflammationand an avalanche of studies has shown HO-1 mediated protectionin ALI/ARDS. Furthermore, numerous phytochemicals has alreadybeen shown to exert anti-inflammatory activity through HO-1 in-duction [6]. Our results thus suggest the essential role of HO-1 inmediating anti-inflammatory effect of EG. Very recently, Park et al.has shown the anti-inflammatory activity of ethyl gallate throughHO-1 in LPS induced mouse macrophages cell line [26], whichfurther lends a support to our study.

Central to the induction of HO-1 is Nrf2/Keap1 complex, whichtightly regulates nuclear translocation of Nrf2 and consequent

Fig. 8. EG mediated H2O2 generation underlies Nrf2 translocation in THP-1 monocytes (A) Histogram represents H2O2 level in culture medium with or without THP-1 monocytes.Culture medium (RPMI 1640) with or without THP-1 cells were incubated with varying concentration of EG for 1 h and measured for H2O2 level using H2O2 calorimetric assay kit.(B) Western blot represents nuclear Nrf2 in EG (30 mM) treated THP-1 monocytes. THP-1 cells were treated with glutathione (60 mM) and myoglobin (20 mM) in the presence orabsence of EG (30 mM) for 2 h. Post incubation, cells were harvested and analyzed for nuclear Nrf2 levels. Levels of Nrf2 were normalized with respective control in the nuclearextract. (C) Histogram represent H2O2 levels in the injectable medium (water with 0.5% DMSO). EG (30 mM) was added in the injectable material and incubated for two differenttime points (10 and 20 min).

K. Mehla et al. / Biochimie 95 (2013) 2404e2414 2413

activation of its targeted genes. Nrf2 targeted genes including HO-1,NQO1 and GPX-1 contributes to the detoxification and excretion ofxenobiotics and any abrogation in its activity skews host’s innateimmune response against oxidative damage. Thus, nrf2 deficientmice exhibit increased oxidative damage against cigarette smoke[27], diesel exhaust particle [28], LPS and allergen [5] exposure.Moreover, a direct association of Nrf2 polymorphism withincreased susceptibility to ALI in trauma and septic patients hasbeen revealed by a genetic study [29]. Against this background, weinvestigated the status of Nrf2 in EG treated mice. Our studydemonstrated increased nuclear level and activity of Nrf2 in EGprotected ALI mice. In addition, notable reduction of Keap-1 level,correlating with Nrf2 nuclear activity, was also observed in thesame experimental sets. This suggests that Nrf2 mediated HO-1induction through Keap-1 reduction underlies the protective ef-fect of EG in LPS challenged BALB/c mice. Our finding is in agree-ment with numerous reports, which show the involvement of Nrf2in meditating cytoprotective and chemoprotective effect of variousplant and synthetic molecules [6].

Mimicking the in vivo results, human monocytes also demon-strated similar profile for Nrf2 and Keap-1 with EG treatment,further lending a support to our previous observations. To dissectout the anti-inflammatory activity of EG through Nrf2 induction,TNF-a levels were also measured in the human monocytes. Previ-ously, Stuart et al. [30] has reported an inverse correlation betweenHO-1 and TNF-a level in LPS challenged human monocytes. HO-1deficient human monocytes mount an excessive inflammatoryresponse against LPS and over expression of the same rescues thecells. In addition, studies from Otterbein et al. suggest that HO-1limits detrimental effect of LPS by preventing release of early aswell as late inflammatory cytokines such as TNF-a, IL-1b andHMGB-1 in ALI/ARDS model [31]. As anticipated, notable reductionof TNF-awith concomitant increase in HO-1 was observed with EGtreatment in a dose dependent manner. In addition, Nrf2 silencingthrough siRNA abrogated the EG induced protective effect in LPSinduced human monocytes. This finding is of significant impor-tance in the context of critical role played by TNF-a in neutrophilinflux and tissue damage in ALI. Encouraged by these results, statusof NF-kB was also evaluated in mice lung tissue and human

monocytes (data not shown). Interestingly, no effect of EG on NF-kBcould be observed suggesting that EG exerts its protective effectessentially through Nrf2-HO-1 signaling. Over the past few years,many studies have surfaced which demonstrate the instability ofpolyphenol in cell culture medium [33e35]. These studies indicatethat polyphenols such as curcumin, rosameric acid and resveratrolgenerate hydrogen peroxide in the cell culture which upon gainingan access into the cell via cellular membrane activates Nrf2 andpromotes its subsequent translocation inside the nucleus. Our re-sults also demonstrated that EG indeed generates moderateamount of H2O2, which in turn promotes Nrf2 activation in thein vitro system, thus implying the role of H2O2 in EG mediated anti-inflammatory response in human monocytes. In light of thisfinding, an insignificant amount of H2O2 was also observed in theinjectable material. However the significance of such a low amountof H2O2 in the in vivo system is unclear as no study has measuredthe efficacy of injected H2O2 in alleviating ALI. Thus, it could beinteresting to investigate the possibility of H2O2 in alleviating ALIphenotypes alone or in combination of EG in the future studies. Insummary, our study indeed highlights the therapeutic potential ofEG which together with other key anti-inflammatory molecule canbe exploited for the treatment of ALI.

5. Conclusion

Collectively, we provide the evidence for the therapeutic po-tential of EG in LPS induced ALI for the first time. It is important tomention that Mink et al. have already demonstrated the protectiveeffect of EG on cardiovascular system in the canine model of septicshock [32]. However, this is the first report, which suggests bene-ficial effect of EG on the airway inflammation in the in vivomodel ofALI. Our study highlights that protective effect of EG in the in vitrosystem is attributed to the H2O2 mediated increased HO-1/Nrf2signaling. While further studies are required to gain more insightinto the mechanisms underlying the EG regulated HO-1/Nrf2 axisin the in vivo system, present study provides a novel mechanisticinsight towards the effects of EG and shows that this moleculealong with its analogs may be useful in the treatment of ALI.

K. Mehla et al. / Biochimie 95 (2013) 2404e24142414

Acknowledgments

KM acknowledges CSIR for her fellowship. The studywas fundedby the Task Force Project BSC0116 of the Council of Scientific andIndustrial Research (CSIR), Govt. of India.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.biochi.2013.08.030.

References

[1] R. Rylander, M.C. Snella, Endotoxin and the lung: cellular reactions and risk fordisease, Prog. Allergy 33 (1982) 332e344.

[2] O.R. Luhr, K. Antonsen, M. Karlsson, S. Aardal, A. Thorsteinsson, C.G. Frostell,J. Bonde, Incidence and mortality after acute respiratory failure and acuterespiratory distress syndrome in Sweden, Denmark, and Iceland: the ARFstudy group, Am. J. Respir. Crit. Care Med. 159 (1999) 1849e1861.

[3] H. Zhu, Z. Jia, B.R. Misra, L. Zhang, Z. Cao, M. Yamamoto, M.A. Trush, H.P. Misra,Y. Li, Nuclear factor E2-related factor 2-dependent myocardiac cytoprotectionagainst oxidative and electrophilic stress, Cardiovasc. Toxicol. 8 (2008) 71e85.

[4] R.K. Thimmulappa, H. Lee, T. Rangasamy, S.P. Reddy, M. Yamamoto,T.W. Kensler, S. Biswal, Nrf2 is a critical regulator of the innate immuneresponse and survival during experimental sepsis, J. Clin. Invest. 116 (2006)984e995.

[5] R.K. Thimmulappa, C. Scollick, K. Traore, M. Yates, M.A. Trush, K.T. Liby,M.B. Sporn, M. Yamamoto, T.W. Kensler, S. Biswal, Nrf2-dependent protectionfrom LPS induced inflammatory response and mortality by CDDO-imidazolide,Biochem. Biophys. Res. Commun. 351 (2006) 883e889.

[6] E. Balogun, M. Hoque, P. Gong, E. Killeen, C.J. Green, R. Foresti, J. Alam,R.Motterlini, Curcumin activates the heme oxygenase-1 gene via regulation ofNrf2 and the antioxidant-responsive element, Biochem. J. 371 (2003) 887e895.

[7] K. Mehla, S. Balwani, A. Kulshreshtha, D. Nandi, P. Jaisankar, B. Ghosh, Ethylgallate isolated from Pistacia integerrima Linn. inhibits cell adhesion moleculesby blocking AP-1 transcription factor, J. Ethnopharmacol. 137 (2011) 1345e1352.

[8] A. Ram, U. Mabalirajan, S.K. Singh, V.P. Singh, B. Ghosh, Mepacrine alleviatesairway hyperresponsiveness and airway inflammation in a mouse model ofasthma, Int. Immunopharmacol. 8 (2008) 893e899.

[9] U. Mabalirajan, T. Ahmad, G.D. Leishangthem, D.A. Joseph, A.K. Dinda,A. Agrawal, B. Ghosh, Beneficial effects of high dose of l-arginine on airwayhyperresponsiveness and airway inflammation in a murine model of asthma,J. Allergy Clin. Immunol. 125 (2010) 626e635.

[10] E. Kaczmarek, A. Górna, P. Majewski, Techniques of image analysis forquantitative immunohistochemistry, Rocz. Akad. Med. Bialymstoku 49 (2004)155e158.

[11] S. Kumar, V. Singhal, R. Roshan, A. Sharma, G.W. Rembhotkar, B. Ghosh,Piperine inhibits TNF-a induced adhesion of neutrophils to endothelialmonolayer through suppression of NF-kB and IkB kinase activation, Eur. J.Pharmacol. 575 (2007) 177e186.

[12] M. Ochs, Stereological analysis of acute lung injury, Eur. Respir. Rev. 15 (2006)115e121.

[13] J. Lefort, L. Motreff, B.B. Vargaftig, Airway administration of Escherichia coliendotoxin to mice induces glucocorticosteroid-resistant bronchoconstrictionand vasopermeation, Am. J. Respir. Cell. Mol. Biol. 24 (2001) 345e351.

[14] J. Reutershan, A. Basit, E.V. Galkina, K. Ley, Sequential recruitment of neu-trophils into lung and bronchoalveolar lavage, Am. J. Physiol. Lung Cell. Mol.Physiol. 289 (2005) L807eL815.

[15] G.S. Worthen, C. Haslett, A.J. Rees, R.S. Gumbay, J.E. Henson, P.M. Henson,Neutrophil-mediated pulmonary vascular injury: synergistic effect of traceamounts of lipopolysaccharide and neutrophil stimuli on vascular perme-ability and neutrophil sequestration in the lung, Am. Rev. Respir. Dis. 136(1987) 19e28.

[16] K.J. Johnson, J.C. Fantone 3rd, J. Kaplan, P.A. Ward, In vivo damage of rat lungsby oxygen metabolites, J. Clin. Invest. 67 (1981) 983e993.

[17] I.N. Zelko, T.J. Mariani, R.J. Folz, Superoxide dismutase multigene family: acomparison of the cu, zn-sod (sod1), mn-sod (sod2), and ec-sod (sod3) genestructures, evolution, and expression, Free Radical Biol. Med. 33 (2002) 337e349.

[18] S.W. Ryter, L.E. Otterbein, D. Morse, A.M. Choi, Heme oxygenase/carbonmonoxide signaling pathways: regulation and functional significance, Mol.Cell. Biochem. 234e235 (2002) 249e263.

[19] L.E. Otterbein, B.Y. Chin, S.L. Otterbein, V.C. Lowe, H.E. Fessler, A.M. Choi,Mechanism of hemoglobin-induced protection against endotoxemia in rats: aferritin-independent pathway, Am. J. Physiol. 272 (1997) L268eL275.

[20] M. Kobayashi, M. Yamamoto, Molecular mechanisms activating the Nrf2-Keap1 pathway of antioxidant gene regulation, Antioxid. Redox Signaling 7(2005) 385e394.

[21] M. Guha, N. Mackman, LPS induction of gene expression in human monocytes,Cell. Signalling 13 (2001) 85e94.

[22] G. Sadeghi-Hashjin, G. Folkerts, P.A. Henricks, A.K. Verheyen, H.J. Van derLinde, I. Van Ark, A. Coene, F.P. Nijkamp, Peroxynitrite induces airwayhyperresponsiveness inguinea pigs in vitro and in vivo, Am. J. Respir. Crit. CareMed. 153 (1996) 1697e1701.

[23] N. Paulino, M.G. Pizollatti, R.A. Yunes, V.C. Filho, T.B. Creczynski-Pasa,J.B. Calixto, The mechanisms underlying the relaxant effect of methyl andethyl gallates in the guinea pig trachea in vitro: contribution of potassiumchannels, Naunyn Schmiedeberg’s Arch. Pharmacol. 360 (1999) 331e336.

[24] A.H. Jansson, C. Eriksson, X. Wang, Effects of budesonide and N-acetylcysteineon acute lung hyperinflation, inflammation and injury in rats, Vasc. Phar-macol. 43 (2005) 101e111.

[25] G.U. Meduri, S. Headley, G. Kohler, F. Stentz, E. Tolley, R. Umberger, K. Leeper,Persistent elevation of inflammatory cytokines predicts a poor outcome inARDS. Plasma IL-1 and IL-6 are consistent and efficient predictors of outcomeover time, Chest 107 (1995) 1062e1073.

[26] P.H. Park, J. Hur, Y.C. Kim, R.B. An, D.H. Sohm, Involvement of hemeoxygenase-1 induction in inhibitory effect of ethyl gallate isolated from GallaRhois on nitric oxide production in RAW 264.7 macrophages, Arch. Pharm.Res. 34 (2011) 1545e1552.

[27] T. Iizuka, Y. Ishii, K. Itoh, T. Kiwamoto, T. Kimura, Y. Matsuno, Y. Morishima,A.E. Hegab, S. Homma, A. Nomura, T. Sakamoto, M. Shimura, A. Yoshida,M. Yamamoto, K. Sekizawa, Nrf2-deficient mice are highly susceptible tocigarette smoke-induced emphysema, Genes Cells 10 (2005) 1113e1125.

[28] Y.J. Li, H. Taikzawa, A. Azuma, T. Kohyama, Y. Yamauchi, S. Takahashi,M. Yamamoto, T. Kawada, S. Kudoh, I. Sugawara, Nrf2 is closely related toallergic airway inflammatory responses induced by low-dose diesel exhaustparticles in mice, Clin. Immunol. 137 (2010) 234e241.

[29] J.M. Marzec, J.D. Christie, S.P. Reddy, A.E. Jedlicka, H. Vuong, P.N. Lanken,R. Aplenc, T. Yamamoto, M. Yamamoto, H.Y. Cho, S.R. Kleeberger, Functionalpolymorphisms in the transcription factor NRF2 in humans increase the riskof acute lung injury, FASEB J. 21 (2007) 2237e2246.

[30] S.A. Rushworth, D.J. MacEwan, M.A. O’Connell, Lipopolysaccharide-inducedexpression of NAD(P)H: quinoneoxidoreductase 1 and heme Oxygenase-1protects against excessive inflammatory responses in human monocytes,J. Immunol. 181 (2008) 6730e6737.

[31] L.E. Otterbein, F.H. Bach, J. Alam, M. Soares, M. Lu, H. Tao, M. Wysk, R.J. Davis,R.A. Flavell, A.M. Choi, Carbon monoxide has anti-inflammatory effectsinvolving the mitogen-activated protein kinase pathway, Nat. Med. 6 (2000)422e428.

[32] S.N. Mink, H. Jacobs, J. Gotes, K. Kasian, Z.Q. Cheng, Ethyl gallate, a scavengerof hydrogen peroxide that inhibits lysozyme-induced hydrogen peroxidesignaling in vitro, reverses hypotension in canine septic shock, J. Appl. Physiol.110 (2011) 359e374.

[33] M. Akagawa, T. Shigemitsu, K. Suvama, Production of hydrogen peroxide bypolyphenols and polyphenol-rich beverages under quasi-physiological con-ditions, Biosci. Biotechnol. Biochem. 67 (2003) 2632e2640.

[34] L.H. Long, A. Hoi, B. Halliwell, Instability of, and generation of hydrogenperoxide by, phenolic compounds in cell culture media, Arch. Biochem. Bio-phys. 501 (2010) 162e169.

[35] H. Erlank, A. Elmann, R. Kohen, J. Kanner, Polyphenol activate Nrf2 in astro-cytes via H2O2, semiquinones, and quinones, Free Radical Biol. Med. 51 (2011)2319e2327.