GEN

ERA

LT

HO

RA

CIC

Protection by 20-5,14-HEDGE Against SurgicallyInduced Ischemia Reperfusion Lung Injury in RatsIrshad Ali, PhD, Stephanie Gruenloh, BS, Ying Gao, MS, Anne Clough, PhD,John R. Falck, PhD, Meetha Medhora, PhD, and Elizabeth R. Jacobs, MD, MBADepartments of Medicine and Radiation Oncology, Medical College of Wisconsin, and Departments of Mathematics, Statistics,

and Computer Sciences, Marquette University, Milwaukee, Wisconsin; and Department of Biochemistry, University of TexasSouthwestern Medical Center, Dallas, Texas

Background. We previously reported that the cyto-chrome P450 product 20-hydroxyeicosatetraenoic acid hasprosurvival effects in pulmonary artery endothelial cellsand ex vivo pulmonary arteries. We tested the potentialof a 20-hydroxyeicosatetraenoic acid analog N-[20-hy-droxyeicosa-5(Z),14(Z)-dienoyl]glycine (20-5,14-HEDGE)to protect against lung ischemic reperfusion injury inrats. Furthermore, we examined activation of innate im-mune system components, high mobility group box 1(HMGB1) and toll-like receptor 4 (TLR4), in this model aswell as the effect of 20-5,14-HEDGE on this signalingpathway.

Methods. Sprague-Dawley rats treated with 20-5,14-HEDGE or vehicle were subjected to surgically induced,unilateral lung ischemia for 60 minutes followed byreperfusion for 2 hours in vivo. Injury was assessedhistologically by hematoxylin and eosin, and with iden-tification of myeloperoxidase immunohistochemically.The HMGB1 and TLR4 proteins were identified by

Western blot. Caspase 3 activity or 3-(4,5-Dimethylthi-

Development, Zablocki VA, Medical College of Wisconsin, 9200 WWisconsin Ave, Milwaukee, WI 53226; e-mail:ejacobs@mcw.edu.

© 2012 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

azol-2-yl)-2,5-diphenyltetrazolium bromide, a yellowtetrazole, incorporation were used to measure apoptosisand cell survival.

Results. The ischemia reperfusion injury evoked atel-ectasis and hemorrhage, an influx of polymorphonuclearcells, and increased TLR4 and HMGB1 expression.Caspase 3 activity was increased, and 3-(4,5-Dimethylthi-azol-2-yl)-2,5-diphenyltetrazolium bromide incorporationwas decreased. The 20-5,14-HEDGE protected against eachof these endpoints, including infiltration of polymorpho-nuclear cells, with no changes in caspase 3 activity in otherorgans.

Conclusions. Lung ischemia reperfusion produces ap-optosis and activation of the innate immune systemincluding HMGB1 and TLR4 within 2 hours of reperfu-sion. Treatment with 20-5,14-HEDGE decreases activa-tion of this response system, and salvages lung tissue.

(Ann Thorac Surg 2012;93:282–9)

© 2012 by The Society of Thoracic Surgeons

Inflammatory responses by the innate immune systemcharacterize ischemia reperfusion (IR) injury in the

heart, liver, and brain and are triggered in part bybinding of endogenous danger molecules such as highmobility group box 1 (HMGB1) to toll-like receptors(TLRs), in particular, TLR4 [1, 2]. Although they havedistinct pathophysiologic mechanisms, conditions suchas crush injury to the chest, lung transplantation, cardio-pulmonary bypass, and others are believed to share somemanifestations of IR injury. The role of TLR4 in warmischemic (in vivo) lung injury is supported by bluntedmicrovascular permeability injury and decreased activa-tion of nuclear factor kappa-beta (NF-�B) and mitogen-activated protein kinases (MAPKs) in TLR4-deficientmice, and well as limited gap formation evoked byhypoxia-reoxygenation in human microvascular endo-thelial monolayers treated with TLR4 blockers [1, 2].Moreover, TLR4 may mediate inflammation of lung en-dothelial cells in hemorrhagic shock [3]. Sustained IR

Accepted for publication Aug 24, 2011.

Address correspondence to Dr Jacobs, MCW and ACOS Research and

injury results in loss of viability in tissue associated withincreased caspase-3 expression (marker of apoptosis).

Our previous studies have shown prosurvival effects ofthe cytochrome P450 metabolite of arachidonic acid,20-HETE, with decreased apoptosis in pulmonary arter-ies treated with 20-HETE then exposed to hypoxia reoxy-genation [4]. The present investigations were undertakento determine whether an analog of 20-HETE, N-[20-hydroxyeicosa-5(Z),14(Z)-dienoyl]glycine (20-5,14-HEDGE) would also impart protection to rat lung tissueexposed in vivo to warm ischemia by reducing caspase 3activation, or modify innate immune system mediatorsHMGB1 and TLR4. Because 20-HETE has been reportedto be proapoptotic to brain [5], kidneys [6], and heart [7]in some injury models, we examined the effects oftreatment by 20-5,14-HEDGE on apoptosis on thesevital organs.

Material and Methods

MaterialsThe Caspase 3 Fluorimetric Assay Kit (Sigma-Aldrich, St.Louis, MO), the Vybrant MTT Cell Proliferation Assay Kit

(Invitrogen, Carlsbad, CA), and the Protein Determina-

0003-4975/$36.00doi:10.1016/j.athoracsur.2011.08.074

blcis

p

Ntpwbrt7asm(

abcAroswfac

ipD[lsn

iwaw((mpnTsl

283Ann Thorac Surg ALI ET AL2012;93:282–9 20-5,14-HEDGE IN IR LUNG INJURY

GEN

ERA

LT

HO

RA

CIC

tion Kit (BioRad, Hercules, CA) were used. The 20-5,14-HEDGE was synthesized in the laboratories of Dr JohnFalck. The structures of 20-HETE and 20-5,14-HEDGEappear in Figure 1. The 20-HETE analog 20-5,14-HEDGE

ehaves as 20-HETE [8]. Whereas 20-HETE is metabo-ized rapidly in vivo, is a substrate for lipoxygenases andyclooxygenases, and also autooxidizes, 20-5,14-HEDGEs more stable in vivo and, unlike 20-HETE, is not aubstrate for lipoxygenases or cyclooxygenases.

THE IR MODEL. All studies were performed under approvalof the Medical College of Wisconsin Institutional AnimalCare and Use Committee and in compliance with theNational Research Council’s “Guide for the Care and Useof Laboratory Animals.” Male Sprague-Dawley rats (200g to 400 g) were used for these experiments. Sixtyminutes before surgery, vehicle (100 mM NaPO4 buffer,

Fig 1. (A) Chemical structure of 20-HETE and 20-5,14-HEDGE. (B,C) Ischemia-reperfusion (IR) lungs from vehicle-treated and 20-5,14-HEDGE–treated rats. The IR resulted in gross hemorrhage in the left(L) lung, and to a lesser extent, right (R) lung. Treatment with 20-5,14-HEDGE decreased visible hemorrhage. (D, E) Hematoxylin andeosin sections revealed hemorrhage (arrows point to red blood cellsin intraalveolar spaces) in ischemic left lungs, which exceeded thatin right lungs. (20-5, 14-HEDGE � N-[20-hydroxyeicosa-5(Z),14(Z)-dienoyl]glycine.)

H 9.0 � 0.1% ethanol) or 20-5,14 HEDGE (30 mg/kg in

aPO4 buffer) was injected intraperitoneally (IP). Anes-hesia was achieved with isoflurane throughout the ex-eriment. Body temperature was maintained with aarming table. A femoral arterial line was placed forlood collection as well as for blood pressure and heartate measurements. A tracheotomy was performed, andhe rat ventilated with a FiO2 of 0.6, with tidal volumes ofmL/kg at a rate of approximately 85 breaths per minutedjusted to maintain eucapnea. Rats were given atropineulfate (0.4 mg/kg, IP), heparin (500 units, IP [APP Phar-aceuticals, Schaumburg, IL]), and 0.9% sodium chloride

1.5 mL, IP every hour).A midline incision was made, and the chest opened to

ccess the left hilum. The left pulmonary artery, vein, andronchus were stripped of connective tissue, then oc-luded with a microvascular clamp to induce ischemia.fter 60 minutes, the clamp was removed to permit

eperfusion and ventilation. Arterial blood gases werebtained before clamping and 120 minutes after reperfu-ion. Rats were exsanguinated, and the heart and lungsere removed en bloc as well as brain, kidneys, and heart

or studies detailed in the following text. Sham-operatednimals underwent the same procedures except thelamp was not applied to the hilum.

CASPASE 3 ACTIVITY AND MTT ASSAY. Caspase 3 activity is anndicator of apoptosis through the intrinsic or extrinsicathways. Caspase-3 activity was determined using Ac-EVD-AMC as a substrate, as previously reported by us

9]. The 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazo-ium bromide (MTT) assay was applied as an index of cellurvival [4]. The results are expressed as optical density (540m)/g wet weight lung tissue (mean � SEM).

WESTERN BLOTS OF TLR4 AND HMGB1. Whole lung homogenatesn a buffer supplemented with protease inhibitor cocktailere centrifuged for 10 minutes at 20,000g. Western blot

nalysis was performed as described previously [9]. Blotsere developed with a primary antibody to TLR4

AF1478; R&D Systems, Minneapolis, MN) �-actinA2228; Sigma), or HMGB1 (MAB1690; R&D Systems),

atched secondary antibodies conjugated to horseradisheroxidase, then visualized using enhanced chemilumi-escent detection reagent (32106; Pierce, Rockford, IL).he relative densities of protein bands were compared incanned images with �-actin used to correct for proteinoading.

HISTOLOGY. Sections of paraffin-embedded lungs werestained for hematoxylin and eosin (H&E), or myeloper-oxidase by immunohistochemistry. Primary antibodiesfor myeloperoxidase (ab45977; Abcam, Cambridge, MA)with matched isotype HRP-conjugated secondary anti-bodies were used in immunohistochemical studies.Hemorrhage and myeloperoxidase were quantified usingMetaMorph to threshold based on hue, saturation, andintensity in images by a person masked to the treatmentgroups. Threshold settings were kept constant for allimages in all groups, and the total threshold area was

recorded.

t

ble

1.H

emod

ynam

icV

aria

bles

and

Art

eria

lBl

ood

Gas

esin

Rat

sin

Isch

emia

-Rep

erfu

sed

Left

Lung

MA

P(m

mH

g)H

eart

Rat

e(b

eats

/min

)H

emog

lobi

n(g

/dL

)p

H(m

mH

g)p

CO

2(m

mH

g)p

O2

(mm

Hg)

HC

O3

Star

tE

nd

Star

tE

nd

Star

tE

nd

Star

tE

nd

Star

tE

nd

Star

tE

nd

Star

tE

nd

am(n

�5)

98�

1010

8�

2110

7�

720

0�

1513

.8�

112

.7�

1.2

7.52

�.0

27.

46�

.04

29.5

�2.

731

.0�

4.4

232.

2�

14.5

225.

8�

14.5

24.6

�1.

221

.5�

1.3

hic

le(n

�8)

92�

1197

�14

170

�21

200

�12

14.3

�1.

112

.2�

2.4

7.46

�0.

087.

36�

0.07

34.9

�5.

838

.1�

8.6

308.

1�

89.2

248.

3�

94.5

24.0

�1.

420

.8�

3.0

5,14

-HE

DG

En

�5)

96�

583

�31

162

�10

194

�17

14.6

�1.

614

.2�

1.6

7.46

�0.

097.

46�

0.1

31.2

�7.

736

.4�

10.7

327.

0�

7035

2�

52a

21.2

�1.

919

.9�

1.8

�0.

05�

20-5

,14-

HE

DG

Eve

rsu

sve

hic

le(a

nal

ysis

ofva

rian

ce,p

�0.

025)

.

O3

�bi

carb

onat

e;M

AP

�m

ean

arte

rial

pre

ssu

re;

pC

O2

�p

arti

alp

ress

ure

(ten

sion

)of

carb

ond

ioxi

de;

pH

�h

ydro

gen

ion

con

cen

trat

ion

;p

O2

�p

arti

alp

ress

ure

(ten

sion

)of

gen

;20

-5,1

4-H

ED

GE

�N

-[20

-hyd

roxy

eico

sa-5

(Z),1

4(Z

)-d

ien

oyl]

glyc

ine.

284 ALI ET AL Ann Thorac Surg20-5,14-HEDGE IN IR LUNG INJURY 2012;93:282–9

GEN

ERA

LT

HO

RA

CIC

Statistical AnalysisFor all endpoints, three groups of animals were studied:sham, vehicle, or 20-5,14-HEDGE pretreatment groups.All analyses were not performed on all rats, as some (eg,histologic analysis and caspase 3 activity) were mutuallyexclusive. For each assay, typically 5 to 8 separate ratswere studied. The number (n) of rats for each groupappears in the figure legends, figures, or tables. Com-parisons between groups for all experiments wereperformed using one-way analysis of variance. Whenbetween-group differences were found, a Holm-Sidakpost hoc test was employed for pairwise multiplecomparisons. Unadjusted p values are provided, withasterisks to indicate those of less than critical levels,and thus significant. All grouped data are presented asmean � SEM.

Results

Hemodynamic and Arterial Blood Gas Variables in IRModelTable 1 shows values for blood pressure, heart rate,hemoglobin, and arterial blood gases for sham-operatedanimals, vehicle-treated, and 20-5,14-HEDGE–treatedrats. All animals maintained good oxygenation, bloodpressures, and acid-base balance throughout the exper-iments. Mean arterial pressures were not increased inrats treated with 20-5,14-HEDGE, important observationsbecause increased synthesis of this lipid is associatedwith systemic hypertension [10]. Rats treated with 20-5,14-HEDGE exhibited higher pO2 values at the end ofhe experiment than rats treated with vehicle.

MorphologyThe gross appearance of sham lungs and most of theright lung of IR animals was light pink at harvesting.Small areas of hemorrhage were observed in the rightlung of some of the vehicle IR animals. The entireischemic left lung looked hemorrhagic in vehicle-treatedIR animals (Fig 1B). In rats preconditioned with 20-5,14-HEDGE, less gross hemorrhage was visible (Fig 1C).

Microscopic Studies of LungsLeft lungs from vehicle-treated rats exhibited areas ofatelectasis and microscopic hemorrhage (Fig 1D). Toquantify the hemorrhage, the hue and intensity of redcolor in H&E slides was determined. Red color wasincreased in left IR lungs of vehicle-treated rats (Table 2)over that of contralateral right lungs, lending supportand quantification to observations of gross hemorrhage.Treatment with 20-5,14-HEDGE prevented hemorrhage.Vehicle IR lungs also exhibited infiltrating cells in H&Esections. Immunohistochemical analysis with myeloper-oxidase antibodies confirmed the presence of neutrophils(Fig 2A to D). Quantification of the brown signal formyeloperoxidase showed that the degree of polymorpho-nuclear cell infiltration was attenuated in the rats pre-

conditioned with 20-5,14-HEDGE (Table 3). Ta Sh V

e20

- (

ap

HC

oxy

cadCaltmsdp

osa-5(

285Ann Thorac Surg ALI ET AL2012;93:282–9 20-5,14-HEDGE IN IR LUNG INJURY

GEN

ERA

LT

HO

RA

CIC

Effect 20-5,14-HEDGE on Caspase 3 ActivityIschemia-reperfusion resulted in an increase in caspase 3activity in the left lung compared with the right lung inrats preconditioned by vehicle (p � 0.05; Table 4). Pre-onditioning with 20-5,14-HEDGE decreased caspase 3ctivity in the ischemic lung to levels that were notifferent from those of the nonischemic right lung.aspase 3 activity in the lungs of the sham-operatednimals was lower than that observed in the ischemic leftungs, but not different from that of either lung in ratsreated with 20-5,14-HEDGE. Caspase 3 activity was also

easured in kidney, brain, and heart tissue (data nothown). Caspase 3 activation in these organs was notifferent in any of the three groups (analysis of variance� 0.2).

Cell Viability Determined by MTT ReductionAs a second marker of cell injury, we examined MTTincorporation into lung tissue. The MTT reduction (opticaldensity normalized to wet weight lung tissue) was reducedin ischemic relative to nonischemic lungs only in the vehiclepreconditioned animals (97 � 3 to 88 � 2; p � 0.04). Ratiosof left to right MTT values for sham or 20-5,14-HEDGEgroups or comparisons of left or right MTT values betweengroups were not different.

Table 2. Percent Area of Slide Positive for Hemorrhage in Lu

ANOVA p � 0.02 Left Lung R

Vehicle (n � 10) 5.99 � 1.32

20-5,14-HEDGE (n � 10) 2.97 � 0.75

a Unadjusted p values. b p � critical level.

ANOVA � analysis of variance; 20-5,14-HEDGE � N-[20-hydroxyeic

Effect of 20-5,14-HEDGE on IR-Stimulated HMGB1and TLR4 ExpressionThe TLR4 protein was increased in homogenates ofischemic left lungs compared with that of right lungs invehicle controls (Fig 3A and B). Preconditioning with20-5,14-HEDGE attenuated the IR-stimulated expressionof TLR4 in the ischemic left lung so that it was notdifferent from the right lungs, and was lower than that ofleft lungs from vehicle-treated animals.

Like TLR4, HMGB1 expression was increased in left IRlungs over that in right lungs from IR rats (Fig 4A and B).Pretreatment with 20-5,14-HEDGE blunted IR-inducedincreases in HMGB1 expression so that it was not differ-ent from the right lungs as well as the levels determinedin the sham-operated animal’s left and right lungs.

Comment

Our work describes several novel findings. HMGB1 andTLR4 expression are increased within 2 hours after isch-emic injury to the lung. Ischemia-reperfusion injury isassociated with gross and microscopic evidence of hem-orrhage and atelectasis, as well as an influx of polymor-phonuclear cells. The 20-5,14-HEDGE protects against (1)IR-associated hemorrhage, (2) decreases in oxygenation,

ctions

Lung Holm-Sidaka

0.34 p � 0.01 left vehicle versus right vehicleb

p � 0.007 left vehicle versus left HEDGEb

0.46 p � 0.84 left HEDGE versus right HEDGEp � 0.82 right vehicle versus right HEDGE

Z),14(Z)-dienoyl]glycine.

Fig 2. Representative immunohistochemicalanalyses demonstrate the presence of myelo-peroxidase-positive cells (see arrow in A) inboth left and right lungs of vehicle-treatedand 20-5,14-HEDGE–treated animals. Themyeloperoxidase expression was quantitatedas brown color in photomicrographic imagesanalyzed by an investigator masked to treat-ment groups. (A) Vehicle ischemia-reperfusion(IR) left lung; (B) vehicle IR right lung; (C)20-5,14 HEDGE IR left lung; (D) 20-5,14HEDGE IR right lung. (IR � ischemiareperfusion; 20-5, 14-HEDGE � N-[20-hydroxyeicosa-5(Z),14(Z)-dienoyl]glycine.)

ng Se

ight

2.96 �

2.73 �

colatlg[ctetw

os

osa-5(

286 ALI ET AL Ann Thorac Surg20-5,14-HEDGE IN IR LUNG INJURY 2012;93:282–9

GEN

ERA

LT

HO

RA

CIC

and (3) increases in HMGB1, TLR4, and caspase 3 in thismodel. These data suggest that 20-5,14-HEDGE decreasesthe inflammatory response and protects against apoptosisthrough pathways including diminished activation of theinnate immune system. Finally, a single dose of 20-5,14-HEDGE does not increase apoptosis in vital organs otherthan the lung including brain, heart, and kidneys.

Apoptosis is a well-recognized feature of transplantedlungs subjected to IR [11, 12], with as many as 35% of theells in human transplanted lung exhibiting terminal de-xynucleotidyl transferase (TdT)-mediated dUTP nick end

abeling positivity 120 minutes after reperfusion. Feng andssociates [13] identified an increase in caspase 3 mRNAhat peaks 4 hours after reperfusion in a rat model of warmung IR. Within 24 hours of reperfusion, caspase 3 messen-er RNA had returned to baseline. Zhang and colleagues14] described decreased injury in rat lung IR after silencingaspase 3 activity with small interfering RNA deliveredhrough inhalation 48 hours before injury. We now reportlevated caspase 3 activity and decreased MTT incorpora-ion in this rodent model of warm ischemia of the lungithin 2 hours of reperfusion.Both gross examination and H&E studies confirm hem-

rrhage as a consistent feature of left IR lungs in ourtudies, as well as atelectasis. Better maintenance of pO2

values in 20-HETE–treated animals supports less atelec-tasis or hemorrhage, or decreased microvascular perme-ability injury. Our data also identify an influx of inflamma-

Table 3. Percent Area of Slide Myeloperoxidase-Positive in Lu

ANOVA p � 0.001 Left Lung R

Sham (n � 8) 1.14 � 0.34

Vehicle (n � 12) 4.51 � 0.56

20-5,14-HEDGE (n � 12) 2.69 � 0.49

a Unadjusted p values. b p � critical level.

ANOVA � analysis of variance; 20-5,14-HEDGE � N-[20-hydroxyeic

Table 4. Caspase 3 Activity

ANOVA p � 0.001 Left Lung

Sham (n � 4) 108.4 � 22.8

Vehicle (n � 12) 335.9 � 51.0b

20-5,14 HEDGE (n � 8) 180.5 � 28.7

a Unadjusted p values. b p � critical level.

ANOVA � analysis of variance; 20-5,14-HEDGE � N-[20-hydroxyeicosa-5(

tory cells in IR lungs 2 hours after reperfusion in theischemic lung. Myeloperoxidase immunohistochemicalstudies confirm neutrophil influx. Feng and colleagues [13]also observed a recruitment of polymorphonuclear cellsaccompanied by an increase in myeloperoxidase activity intheir studies of rat lung IR. These values peaked 4 hoursafter injury, then fell to baseline values after 3 days.

Lung reperfusion after ischemia invokes an inflamma-tory response that has been reported to activate compo-nents of the innate immune system including comple-ment and TLRs, particularly TLR4 [1]. Engagement ofTLR4 receptors activates NF-�B, phosphorylation of p38,ERK, and JNK, and signaling pathways that result innoncardiogenic pulmonary edema and inflammation inpart due to cytokine release [1, 2, 15]. The TLR4-deficientmice enjoy protection from lung IR based on decreasedvascular permeability, diminished bronchoalveolar la-vage accumulation of neutrophils, and decreased activa-tion of NF-�B and JNK [2]. Using chimeric mice, Zanottiand coworkers [1] reported that TLR4 on lung parenchy-mal cells, as opposed to myeloid cells, is critical forincreased vascular permeability in lung IR. The IR re-mote from the lung also appears to increase lung TLR4expression; intestinal ischemia in mice results in en-hanced lung TLR4 expression and inflammatory cellinflux, and pulmonary neutrophil recruitment is substan-tially reduced in TLR4 deficient hosts [16]. Our studiesdemonstrate an increase in TLR4 protein in ischemic left

ections

Lung Holm-Sidaka

0.48 p � 0.22 left sham versus right shamp � 0.001 left sham versus left vehicleb

p � 0.8 R sham versus right vehicle0.14 p � 0.004 left vehicle versus right vehicleb

p � 0.006 left vehicle versus left HEDGE0.54 p � 0.82 left HEDGE versus right HEDGE

p � 0.55 right HEDGE versus right vehiclep � 0.43 left HEDGE versus left sham

Z),14(Z)-dienoyl]glycine.

t Lung Holm-Sidaka

� 11.9 p � 0.9 left sham versus right shamp � 0.0001 left sham versus left vehicleb

p � 0.34 R sham versus right vehicle� 25.9 p � 0.004 left vehicle versus right vehicleb

p � 0.001 left vehicle versus left HEDGEb

� 25.1 p � 0.37 left HEDGE versus right HEDGEp � 0.91 right HEDGE versus right shamp � 0.48 left HEDGE versus left sham

ng S

ight

2.14 �

1.97 �

2.56 �

Righ

100

177.8

136.8

Z),14(Z)-dienoyl]glycine.

A

#(b

287Ann Thorac Surg ALI ET AL2012;93:282–9 20-5,14-HEDGE IN IR LUNG INJURY

GEN

ERA

LT

HO

RA

CIC

lungs over that of matched nonischemic right lungswithin 2 hours of reperfusion (Fig 3). Because (1) TLR4expression is increased coincident with neutrophil recruit-ment in our model, (2) neutrophils are known to expressTLR4, which contributes to lung injury after donor braindeath [17], and (3) IR lungs from hosts treated with 20-5,14-HEDGE exhibit decreased TLR4 expression as well asdecreased myeloperoxidase by immunohistochemicalstudy, we postulate that neutrophils contribute substan-tially to and are essential for the full increase in TLR4expression characteristic of this model. The beneficial effectof 20-5,14-HEDGE on both TLR4 and myeloperoxidase inIR lungs could be attributable to effects on either neutro-phils (Fig 2) or lung parenchymal cells [1].

A key endogenous ligand for TLR4 is HMGB1, which is anonhistone DNA binding protein normally present in thenucleus of cells. With injury, this protein can be releasedfrom damaged cells, where it amplifies inflammationthrough binding to receptors including TLR2, TLR4, andreceptors for advanced glycation products [18]. Binding ofHMGB1 to TLR4 activates NF-�B and mitogen-activatedprotein kinase, and increases expression of intercellularadhesion molecule-1 and E-selectin in rat aortic endothelialcells [18]. This pathway is thought to be key to recruitmentof leukocytes to injured tissues [19]. Increases in HMGB1have not been previously reported in lung IR, but thesedata are consistent with lung tissue injury.

Cytochrome P450 is the primary isoform that catalyzes

Fig 3. Western blots for toll-like receptor 4 (TLR4). (A) Representa-tive images of Western blots probed with a primary antibody toTLR4 show increased band density in samples from vehicle-treatedleft lungs compared with right lungs. Treatment with 20-5,14-HEDGE partially blunted IR-induced increases in TLR4 protein. (B)Average densities and standard error bars of TLR4 normalized for�-actin. *Indicates p � critical level relative to right lung vehicle.Indicates p � critical level relative to left lung vehicle. Numbersn) for paired studies appear in the bar for right lung values. (Blackars � left lung; white bars � right lung.) (IR � ischemia reperfu-

sion; 20-5, 14-HEDGE � N-[20-hydroxyeicosa-5(Z),14(Z)-dienoyl]glycine.)

the conversion of arachidonic acid into 20-HETE, and is

expressed in several cell types in the lung includingepithelium, alveolar macrophages, vascular smooth mus-cle cells, and uniquely, pulmonary artery endothelialcells from small arteries [20]. We have previously re-ported that 20-HETE has prosurvival effects in (1) pulmo-nary artery endothelial cells injured by starvation orlipopolysaccharide, and (2) ex vivo pulmonary arteriesstressed by hypoxia reperfusion [4]. This lipid enhancesnitric oxide release from pulmonary artery endothelialcells [21], which contributes to the vasorelaxant andprosurvival effects in the lung [22, 23]. The 20-HETE–induced nitric oxide production in pulmonary arteryendothelial cells is mediated by nicotinamide adeninedinucleotide phosphate-oxidase, H2O2, and PI3-kinase/

kt [9]. For these reasons, we tested the effect of the20-HETE analog 20-5,14-HEDGE in a rat lung IR model,where we note impressive protection against elevations incaspase 3 activity, TLR4, and HMGB1 expression. Oneattractive hypothesis is that 20-5,14-HEDGE protectsagainst apoptosis and cell injury by decreasing release ofHMGB1, with less binding to TLR4 and ultimately lessinflammatory response in hosts treated with this lipid.

Reduction of MTT is an index of cell viability; less MTTincorporation indicates decreased cell viability. In our IRmodel, caspase 3 activity was roughly doubled in the leftlung of vehicle-treated rats relative to the right lungwhereas the MTT incorporation was decreased by lessthan 10%. In the short time frame after reperfusion whenlungs were harvested, the extent of apoptosis would be

Fig 4. Western blots for high mobility group box 1 (HMGB1). (A)Representative images of Western blots probed for HMGB1. Immu-nospecific band density was increased in samples from ischemia-reperfusion (IR) lungs of vehicle-treated hosts. Treatment with 20-5,14-HEDGE blunted the increase in HMGB1. Lungs from sham-operated animals showed no difference in HMGB1 levels. (B) Graphdepicts average densities and SEM of HMGB1 normalized for �-ac-tin. *Indicates p � critical level relative to right lung vehicle. Num-bers (n) for paired studies appear in the bar for right lung values.(Black bars � left lung; white bars � right lung.) (20-5, 14-

HEDGE � N-[20-hydroxyeicosa-5(Z),14(Z)-dienoyl]glycine.)

288 ALI ET AL Ann Thorac Surg20-5,14-HEDGE IN IR LUNG INJURY 2012;93:282–9

GEN

ERA

LT

HO

RA

CIC

expected to be greater than overt cell death. Protection by20-5,14-HEDGE based on both MTT incorporation andcaspase-3 activation was significant.

Despite protection by 20-5,14-HEDGE in this model,several issues need additional attention. Details of sig-naling mechanisms underlying 20-5,14-HEDGE protec-tion remain to be established. For example, 20-5,14-HEDGE increases nitric oxide production in pulmonaryartery endothelial cells [21], but the role of an increase innitric oxide production to the prosurvival effects in thismodel is unknown. Based on elevations in HMGB-1 andTLR4, one may speculate that treatments that interferewith this signaling pathway could afford protection, butsuch interventions have not been tested in this model.Finally, the potential of posttreatment (after reperfusion)with 20-5,14-HEDGE to protect lungs is not tested. Re-gardless, our data suggest the potential to salvage pul-monary tissue from ischemic injury in perioperativeperiods associated with transplant, cardiopulmonary by-pass, crush injury, or others by a single administration ofa naturally occurring lipid without apparent injury toother vital organs. Therefore, additional studies to ex-plore the mechanisms underlying and the requirementsto enjoy this protection are warranted.

The authors acknowledge contributions to this work byJayashree Narayanan, Dr Said Audi, and Dr Kirkwood Pritchard.Histologic experiments were facilitated by Dr Paula North andthe Children’s Research Institute Imaging Core. Grant supportwas provided by the National Heart, Lung, and Blood Institute(HL068627, HL-49294 to Dr Jacobs) and GM31278 (to Dr Falck).

References

1. Zanotti G, Casiraghi M, Abano JB, et al. Novel critical role oftoll-like receptor 4 in lung ischemia-reperfusion injury andedema. Am J Physiol Lung Cell Mol Physiol 2009;297:L52–63.

2. Shimamoto A, Pohlman TH, Shomura S, Tarukawa T, TakaoM, Shimpo H. Toll-like receptor 4 mediates lung ischemia-reperfusion injury. Ann Thorac Surg 2006;82:2017–23.

3. Li Y, Xiang M, Yuan Y, et al. Hemorrhagic shock augmentslung endothelial cell activation: role of temporal alterationsof TLR4 and TLR2. Am J Physiol Regul Integr Comp Physiol2009;297:R1670–80.

4. Dhanasekaran A, Bodiga S, Gruenloh S, et al. 20-HETEincreases survival and decreases apoptosis in pulmonaryarteries and pulmonary artery endothelial cells. Am JPhysiol Heart Circ Physiol 2009;296:H777–86.

5. Renic M, Klaus JA, Omura T, et al. Effect of 20-HETEinhibition on infarct volume and cerebral blood flow aftertransient middle cerebral artery occlusion. J Cereb BloodFlow Metab 2009;29:629–39.

6. Nilakantan V, Maenpaa C, Jia G, Roman RJ, Park F. 20-HETE-mediated cytotoxicity and apoptosis in ischemic kid-

ney epithelial cells. Am J Physiol Renal Physiol 2008;294:F562–70.

4 (TLR4) and high-mobility group box 1 (HMGB1). This is

© 2012 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

7. Nithipatikom K, Gross ER, Endsley MP, et al. Inhibition ofcytochrome P450 omega-hydroxylase: a novel endogenouscardioprotective pathway. Circ Res 2004;95:e65–71.

8. Cuez T, Korkmaz B, Buharalioglu CK, et al. A syntheticanalogue of 20-HETE, 5,14-HEDGE, reverses endotoxin-induced hypotension through increased 20-HETE levels as-sociated with decreased iNOS protein expression and vaso-dilator prostanoid production in rats. Basic Clin PharmacolToxicol 2009;106:378–88.

9. Bodiga S, Gruenloh SK, Gao Y, et al. 20-HETE-induced nitricoxide production in pulmonary artery endothelial cells ismediated by NADPH oxidase, H2O2, and PI3-kinase/Akt.Am J Physiol Lung Cell Mol Physiol 2010;298:L564–74.

10. Williams JM, Murphy S, Burke M, Roman RJ. 20-hydroxyeicosatetraeonic acid: a new target for the treatmentof hypertension. J Cardiovasc Pharmacol 2010;56:336–44.

11. Fischer S, Cassivi SD, Xavier AM, et al. Cell death in humanlung transplantation: apoptosis induction in human lungsduring ischemia and after transplantation. Ann Surg 2000;231:424–31.

12. Stammberger U, Gaspert A, Hillinger S, et al. Apoptosisinduced by ischemia and reperfusion in experimental lungtransplantation. Ann Thorac Surg 2000;69:1532–6.

13. Feng D, Zhang S, Hu Z, et al. Dynamic investigation ofalveolar type II cell function in a long-term survival model ofrat lung ischemia-reperfusion injury. Scand J Clin Lab Invest2010;70:364–73.

14. Zhang Q, Chatterjee S, Wei Z, Liu WD, Fisher AB. Rac andPI3 kinase mediate endothelial cell-reactive oxygen speciesgeneration during normoxic lung ischemia. Antioxid RedoxSignal 2008;10:679–89.

15. Andersson U, Tracey KJ. HMGB1 is a therapeutic target forsterile inflammation and infection. Annu Rev Immunol 2011;29:139–62.

16. Ben DF, Yu XY, Ji GY, et al. TLR4 mediates lung injury andinflammation in intestinal ischemia-reperfusion. J Surg Res2011 Jan 5 [E-pub ahead of print].

17. Rostron AJ, Cork DM, Avlonitis VS, Fisher AJ, Dark JH,Kirby JA. Contribution of toll-like receptor activation to lungdamage after donor brain death. Transplantation 2010;90:732–9.

18. Hayakawa K, Qiu J, Lo EH. Biphasic actions of HMGB1signaling in inflammation and recovery after stroke. Ann NYAcad Sci 2010;1207:50–7.

19. Yang J, Huang C, Yang J, Jiang H, Ding J. Statins attenuatehigh mobility group box-1 protein induced vascular endo-thelial activation: a key role for TLR4/NF-kappaB signalingpathway. Mol Cell Biochem 2010;345:189–95.

20. Zhu D, Zhang C, Medhora M, Jacobs ER. CYP4A mRNA,protein, and product in rat lungs: novel localization invascular endothelium. J Appl Physiol 2002;93:330–7.

21. Chen Y, Medhora M, Falck JR, Pritchard KA, Jacobs ER.Mechanisms of activation of eNOS by 20-HETE and VEGF inbovine pulmonary artery endothelial cells. Am J PhysiolLung Cell Mol Physiol 2006;291:L378–85.

22. Birks EK, Bousamra M, Presberg K, Marsh JA, Effros RM,Jacobs ER. Human pulmonary arteries dilate to 20-HETE, anendogenous eicosanoid of lung tissue. Am J Physiol 1997;272:L823–9.

23. Jacobs ER, Zhu D, Gruenloh S, Lopez B, Medhora M.VEGF-induced relaxation of pulmonary arteries is mediated

by endothelial cytochrome P-450 hydroxylase. Am J PhysiolLung Cell Mol Physiol 2006;291:L369–77.

INVITED COMMENTARY

In this article, Dr Ali and colleagues [1] provide a role forapoptosis and the innate immune system in lung isch-emia reperfusion injury (LIRI) through toll-like receptor

the first report of an increment of HMGB1 in LIRI to myknowledge. They also show that a 20-hydroxyeicosatet-raenoic acid (20-HETE) analog N-[20-hydroxyeicosa-

5(Z),14(Z)-dienoyl]glycine (20-5,14-HEDGE) suppresses

0003-4975/$36.00doi:10.1016/j.athoracsur.2011.09.060