Carbon Monoxide Protects against Hyperoxia-inducedEndothelial Cell Apoptosis by Inhibiting ReactiveOxygen Species Formation*

Received for publication, August 9, 2006, and in revised form, November 28, 2006 Published, JBC Papers in Press, November 29, 2006, DOI 10.1074/jbc.M607610200

Xue Wang, Yong Wang, Hong Pyo Kim, Kiichi Nakahira, Stefan W. Ryter, and Augustine M. K. Choi1

From the Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine, University of Pittsburgh MedicalCenter, Pittsburgh, Pennsylvania 15213

Hyperoxia causes cell injury and death associated with reac-tive oxygen species formation and inflammatory responses.Recent studies show that hyperoxia-induced cell death involvesapoptosis, necrosis, or mixed phenotypes depending on celltype, although the underlying mechanisms remain unclear.Using murine lung endothelial cells, we found that hyperoxiacaused cell death by apoptosis involving both extrinsic (Fas-dependent) and intrinsic (mitochondria-dependent) pathways.Hyperoxia-dependent activation of the extrinsic apoptosispathway and formationof thedeath-inducing signaling complexrequired NADPH oxidase-dependent reactive oxygen speciesproduction, because this process was attenuated by chemical inhi-bition, as well as by genetic deletion of the p47phox subunit, of theoxidase. Overexpression of heme oxygenase-1 prevented hyper-oxia-inducedcelldeathandcytochrome c release.Likewise, carbonmonoxide, at low concentrations, markedly inhibited hyperoxia-induced endothelial cell death by inhibiting cytochrome c releaseand caspase-9/3 activation. Carbon monoxide, by attenuatinghyperoxia-induced reactive oxygen species production, inhibitedextrinsic apoptosis signaling initiated by death-inducing signalcomplex trafficking from the Golgi apparatus to the plasmamem-brane anddownstreamactivation of caspase-8.We also found thatcarbon monoxide inhibited the hyperoxia-induced activation ofBcl-2-related proteins involved in both intrinsic and extrinsic apo-ptotic signaling. Carbon monoxide inhibited the activation of Bidand the expression and mitochondrial translocation of Bax,whereas promoted Bcl-XL/Bax interaction and increased Badphosphorylation. We also show that carbon monoxide promotedan interaction of heme oxygenase-1 with Bax. These results definenovel mechanisms underlying the antiapoptotic effects of carbonmonoxide during hyperoxic stress.

The clinical treatment of respiratory failure often requiressupplemental oxygen therapy. Prolonged exposure to an ele-

vated oxygen tension (hyperoxia) in animal models causesacute and chronic lung injury that resembles acute respiratorydistress syndrome. In rodent models, hyperoxia triggers anextensive inflammatory response in the lung that degrades thealveolar-capillary barrier, leading to impaired gas exchange andpulmonary edema (1, 2). Lung tissue damage results from thedirect action of increased intracellular reactive oxygen species(ROS)2 or as a secondary consequence of inflammatoryresponses of the host (3, 4). The source(s) of intracellular ROSduring oxygen exposure remain unclear but may involveincreased mitochondrial generation and/or activation ofNADPHoxidases (5, 6). The pathological changes in hyperoxia-injured lungs coincide with the injury or death of pulmonarycapillary endothelial cells and alveolar epithelial cells (2, 4–7).Epithelial cells maintain the integrity of the alveolar-capillarybarrier and defend against oxidative injury. Compromised epi-thelial cell function may permit fluid and macromolecules toleak into the airspace, resulting in clinical respiratory failureand death (3, 8, 9).The mechanisms underlying hyperoxic lung injury and cell

death vary in a tissue-specific manner and can involve apopto-sis, necrosis, or mixed cell death phenotypes. Apoptosis signal-ing pathways play an important role in hyperoxia-induced lungcell death, regardless of whether the final phenotypic outcomeresembles apoptosis or necrosis (10).Two apoptotic pathways have been identified by which

cells can initiate and execute the cell death process: anintrinsic (mitochondria-dependent) pathway and an extrinsic(receptor-dependent) pathway. Mitochondria play key roles indirecting the fate of cells by maintaining the cellular levels ofATP and by releasing apoptogenic factors upon cellular stimu-lation with pro-apoptotic signals. Members of the Bcl-2 familyof proteins act as critical regulators of the intrinsic apoptoticpathway. Antiapoptoticmembers of this family, such as Bcl-XL,localize to the outer membrane of the mitochondria, whereaspro-apoptotic Bcl-2 familymembers such as Bax and Bid trans-locate to the mitochondria upon cellular stimulation withdiverse agents. Bax forms channels upon assimilating into themitochondria, thus increasing outer mitochondrial membranepermeability and thereby facilitating the release of cytochromec and other proapoptotic molecules from the mitochondrial

* This work was supported by American Heart Association Grants 0335035N(to S. W. R.) and 0525552U (to H. P. K.) and National Institutes of HealthGrants R01-HL60234, R01-HL55330, R01-HL079904, and P01-HL70807 (toA. M. K. C.). The costs of publication of this article were defrayed in part bythe payment of page charges. This article must therefore be herebymarked “advertisement” in accordance with 18 U.S.C. Section 1734 solely toindicate this fact.

1 To whom correspondence should be addressed: Division of Pulmonary,Allergy and Critical Care Medicine, Dept. of Medicine, University of Pitts-burgh Medical Center, 3459 Fifth Ave., MUH 628NW, Pittsburgh, PA 15213.Tel.: 412-692-2210; Fax: 412-692-2260; E-mail: choiam@upmc.edu.

2 The abbreviations used are: ROS, reactive oxygen species; DCFDA, 2�,7�-dichlorodihydrofluorescein diacetate; DISC, Death-inducing signal com-plex; EGFR, epidermal growth factor receptor; HO, heme oxygenase; MLEC,mouse lung endothelial cells; LDH, lactate dehydrogenase.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 282, NO. 3, pp. 1718 –1726, January 19, 2007© 2007 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

1718 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 3 • JANUARY 19, 2007

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

intermembrane space (11–13). Alternatively, hyperoxia maydirectly cause irreversible mitochondrial damage, also leadingto increases in mitochondrial outer membrane permeability(14). Once released to the cytosol, cytochrome c forms an apop-tosome complex with Apaf-1, which activates caspase-9 and, inturn, its downstream caspase-3, resulting in the morphologicalfeatures of apoptosis (15). On the other hand, the extrinsic apo-ptotic pathway initiateswhenadeath ligand, such as theFas ligand(FasL), interacts with its cell surface receptor (i.e. Fas), forming adeath-inducing signal complex (DISC) (16). Activation of Fas trig-gers its oligomerization and the rapid recruitment of FADD (Fas-associated death domain protein) and caspase-8. Activatedcaspase-8 subsequently cleaves Bid into truncated Bid (tBid),which translocates from the cytosol to the mitochondrial mem-brane, where it assists in Bax activation (16).CO can provide protection against hyperoxic injury in mice

(17) and rats (18), but the underlying mechanisms remainincompletely understood. CO occurs in nature as a product ofthe combustion of organic materials. CO also arises endog-enously in cells and tissues as a product of heme oxygenase(HO) activity, which degrades heme to biliverdin-IX� and fer-rous iron (19). HO-1, the inducible form of HO, responds totranscriptional up-regulation by multiple forms of cellularstress. HO-1 confers cytoprotection against oxidative stress invitro and in vivo (20). HO-derivedCOacts as a vasorelaxant andinhibits other vascular functions such as platelet aggregationand smooth muscle proliferation (21). When applied at lowconcentration, CO exerts potent cytoprotective effects mim-icking those of HO-1 induction, which include anti-inflamma-tory effects in macrophages (22, 23), as well as anti-apoptoticeffects in vascular cells (24, 25). HO-1 and/orCOprovide tissueprotection in a number of in vivomodels, including vascularinjury and ischemia/reperfusion injury (26, 27). The mecha-nisms of CO action potentially involve modulation of intra-cellular signaling pathways, including activation of p38mitogen-activated protein kinase and soluble guanylylcyclase (22, 27–29).In the current study, we demonstrate that CO prevented

hyperoxia-induced apoptosis by inhibition of both extrinsicand intrinsic apoptotic pathways by blocking Bax and Bid acti-vation and cytochrome c release from the mitochondria. Addi-tionally, CO inhibited plasma membrane DISC formation byattenuating ROS production and preventing DISC traffickingfrom the Golgi apparatus. We describe for the first time theoccurrence of an interaction between HO-1 with Bax that isenhanced by CO treatment and that suggests a novel antiapo-ptotic mechanism.

EXPERIMENTAL PROCEDURES

Chemicals and Reagents—Antibodies anti-Bax, anti-Bid,anti-caspase-8, anti-caspase-9, anti-cytochrome c, anti-Fas,and protein A-agarose were purchased from Santa Cruz Bio-technology, Inc. (Santa Cruz, CA). Anti-Bax 6A7 antibody waspurchased from BD PharMingen (San Diego, CA). Anti-p110was from Calbiochem (San Diego, CA). Anti-heme oxygen-ase-1 was from Stressgen (Vancouver, Canada). The lactatedehydrogenase (LDH) assay kit was from Roche Applied Sci-

ence. Apocyanin and diphenyleneiodoniun were from Calbio-chem. Digitonin and all other chemicals were from Sigma.Animals—The mice were acclimated for 1 week with rodent

chow and water ad libitum. The animals were housed accord-ing to guidelines from theAmericanAssociation for LaboratoryAnimal Care and Research Protocols and were approved by theAnimal Care and Use Committee (University of PittsburghSchool of Medicine). C57BL/6 wild type mice and p47phox�/�

mice (C57BL/6 background) were purchased from JacksonLaboratory and used for preparation of endothelial cells asdescribed below.Isolation and Culture of Murine Lung Endothelial Cells—

The isolation of mouse lung endothelial cells (MLEC) by animmunobead protocol has been reported elsewhere (30).Briefly, mouse lungs were digested in collagenase and filteredthrough 100-�m cell strainers, centrifuged, and washed twicewith medium. Cell suspensions were incubated with a mono-clonal antibody (rat anti-mouse) against platelet endothelialcell adhesion molecule-1 for 30 min at 4 °C. The cells werewashed twice to remove unbound antibody and resuspended ina binding buffer containingwashedmagnetic beads coatedwithsheep anti-rat immunoglobulin G. Attached cells were washedfour to five times in culture medium and then digested withtrypsin/EDTA to detach the beads. Bead-free cells were centri-fuged and resuspended for culture. After two passages, the cellswere incubated with fluorescent-labeled diacetylated low den-sity lipoprotein, which is only absorbed by endothelial cells andmacrophages, and sorted to homogeneity by fluorescence-acti-vated cell sorting.Cell Culture and Treatments—The MLEC were cultured in

Dulbecco’s modified Eagle’s medium containing 10% fetal calfserum, 6.32 g/liter Hepes, and 3.3 ml of endothelial cell growthsupplements in humidified incubators at 37 °C. For adenoviralinfections, the cells were grown to 30% confluence and changedto serum-freemedium containing 106 plaque-forming units/mlof an adenoviral vector inserted with lacZ, or ho-1. Infectedcells were incubated for 3 h and then restored to Dulbecco’smodified Eagle’smedium containing 10% fetal calf serum for anadditional 2 days of incubation. For hyperoxia treatment, theMLEC were grown to 95% confluence, changed to freshmedium, and transferred to a COY anaerobic chamber (COYLaboratory Products, Inc., Ann Arbor, MI) containing a hyper-oxic atmosphere (95% O2 and 5% CO2) in the absence or pres-ence of CO (250 ppm). Control cells were cultured in standardtissue culture conditions (95% room air, 5% CO2).LDH Release Assay—LDH release was measured using a

commercially available assay (cytotoxicity detection kit; RocheApplied Science). After gentle agitation, 200 �l of medium wasremoved at various times to be used for the assay. The sampleswere incubated (30min) with buffer containing NAD�, lactate,and tetrazolium. LDH converts lactate to pyruvate generatingNADH. TheNADH then reduces tetrazolium (yellow) to form-azan (red), which was detected by absorbance at 490 nm. Addi-tionally, the percentage of cell death was determined by exclu-sion of trypan blue.ROSMeasurement by DCFDA Fluorescence—The formation

of ROS in MLEC was determined by the DCFDA fluorescencemethod (6). MLEC were loaded with 10 �MDCFDA for 30 min

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

JANUARY 19, 2007 • VOLUME 282 • NUMBER 3 JOURNAL OF BIOLOGICAL CHEMISTRY 1719

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

in Dulbecco’smodified Eagle’smedium at 37 °C in a 95% air, 5%CO2 environment. At the end of the incubation, the mediumcontaining DCFDA was aspirated, and the cells were washedonce in media. After the addition of 5 ml of growth media, thecultures were exposed to hyperoxia (95%O2, 5%CO2) for 3 h inthe absence or presence of CO (250 ppm). At the end of expo-sure to hyperoxia, the cultures were scraped on ice, and theresulting cell suspensions were centrifuged at 8,000 � g for 10min at 4 °C. The medium was aspirated, and the cell pellet waswashed twice with ice-cold phosphate-buffered saline. Thecells were then sonicated on ice with a probe sonicator at asetting of 5 for 15 s in 500 �l of ice-cold phosphate-bufferedsaline to prepare cell lysates. The fluorescence of oxidizedDCFDA in cell lysates, an index of formation of ROS, wasmeasured on an Aminco Bowman series-2 spectrofluorom-eter at 490/530-nm excitation/emission, using appropriateblanks. The extent of ROS formation was expressed in arbi-trary fluorescence units.Cell Fractionation—Cellular fractionation was performed

essentially as previously described (31). For mitochondrial iso-lation, at various times after exposure to hypoxia, MLEC wereharvested in 0.05% of digitonin in extraction buffer containing50 mM Hepes, pH 7.5, 50 mM KCl, 5 mM EGTA, and 2 mMMgCl2, 0.1 mg/ml phenylmethylsulfonyl fluoride, 30 �l/mlaprotinin, 1 mM sodium orthovanadate. The cell extracts werespun at 700� g for 10min, and the supernatants weremoved tonew tubes and then centrifuged again at 14,000 � g at 4 °C for20 min. The supernatants (cytosol) and the pellets wereretained for Western blotting.For plasmamembrane isolation, the homogenates of the cells

in MBS (25 mM 2-(N-morpholino)ethanesulfonic acid, pH 6.5,

0.15 M NaCl) containing 1% TritonX-100 were adjusted to 40% sucroseby the addition of 2 ml of 80%sucrose prepared in MBS andplaced at the bottom of an ultracen-trifuge tube. A 5–30% discontinu-ous sucrose gradient was formedabove (4 ml of 5% sucrose, 4 ml of30% sucrose, both in MBS lackingdetergent) and centrifuged at 39,000rpm for 18 h in an SW41 rotor(Beckman Instruments, Palo Alto,CA). A band at the interface of 5 and30% sucrose was collected and usedfor immunoprecipitation andWest-ern blotting.The Golgi complex was isolated

using sucrose density gradientcentrifugation, as described else-where (31). After washing withphosphate-buffered saline, thecells were harvested in G buffer(10 mM Tris-HCl, 0.25 M sucrose, 2mM MgCl2, pH 7.4) containing 10mM CaCl2 and protease inhibitors.The cells were disrupted with 20strokes in a Potter-type homoge-

nizer. The homogenate was centrifuged at 2,500� g for 10min,and the pellet was discarded. The resulting post nuclear super-natant was harvested, and the sucrose concentration wasadjusted to a final concentration of 1.4 M. This suspension wasloaded onto the bottom of an ultracentrifuge tube and overlaidin succession with 1.2, 1.0, and 0.8 M sucrose in G buffer. Thesamples were then centrifuged at 95,000 � g for 2.5 h. Twobands from the top, representing 0.8/1.0 and 1.0/1.2 M inter-faces were carefully removed, diluted with G buffer withoutsucrose, collected by centrifugation at 80,000 � g for 30 min,and used for the experiments. Golgi fraction purity wasassessed by enzymatic activity and immunoblottingmethods aspreviously described (31).Western Blot Analysis and Immunoprecipitation—The pro-

teins were isolated from MLEC cultures in radioimmunopre-cipitation assay buffer (1� phosphate-buffered saline, 1% (v/v)Nonidet P-40, 0.5% (w/v) sodium deoxycholate, 0.1% (w/v)sodium dodecyl sulfate, 0.1 mg/ml phenylmethylsulfonyl fluo-ride, 30 �l/ml aprotinin, and 1mM sodium orthovanadate). Forimmunoprecipitation, 1 �g of antibody (anti-Fas, anti-6A7,anti-Bax, or anti-phosphoserine) was added to 500 �g of totalprotein in 500 �l, rotated for 2 h at 4 °C, and then incubatedwith 20 �l of beads (protein A-sucrose; Santa Cruz Biotechnol-ogy, Inc., Santa Cruz, CA) for 2 h, spun down at 500 � g, andwashed three times with radioimmunoprecipitation buffer.Then 20 �l of loading buffer (100 mM Tris-HCl, 200 mM dithi-othreitol, 4% SDS, 0.2% bromphenol blue, and 20% glycerol)was added. For SDS-PAGE, protein samples at the indicatedand equal amounts were boiled in the loading buffer and sepa-rated on SDS-PAGE, followed by transfer to polyvinylidenedifluoride membranes. The membrane was blocked with 5%

FIGURE 1. HO-1/CO protected against hyperoxia-induced cell death. MLEC at 30% confluence were cul-tured in serum-free medium in the presence of 106 CPU/ml of adeno-HO-1 or adeno-lacZ for 3 h and thenrestored to normal medium. Two days later, the cells were exposed to hyperoxia (95% O2, 5% CO2) for theindicated times. Viability was determined by trypan blue exclusion (A). LDH assay was performed on cell culturesupernatants (B). Cytosolic or mitochondrial fractions were prepared from total lysates as described under“Experimental Procedures.” The fractions were subjected to immunoblotting (IB) to detect cytochrome c, withtotal cytochrome c as the standard (C). Western results are representative of three experiments. �-Actin servedas the loading standard. The data represent the means � S.E. of three independent experiments. *, p � 0.05.

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

1720 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 3 • JANUARY 19, 2007

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

nonfat milk and stained with the primary antibodies for 2 h atthe optimal concentrations. After five washes in phosphate-buffered saline with 0.2% Tween 20, the horseradish peroxi-dase-conjugated secondary antibody was applied, and the blotwas developed with enhanced chemiluminescence reagents(Amersham Biosciences).Caspase Activity Assays—Caspase-3 fluorogenic substrate,

Ac-DEVD-AFC and caspase-8 fluorogenic substrate, and Ac-IETD-AFC were from BD Biosciences (Franklin Lakes, NJ).Caspase activity in cell lysates was determined according to themanufacturer’s instructions, using an Aminco Bowmanseries-2 spectrofluorometer (440/500-nm excitation/emis-sion), and expressed as arbitrary fluorescence units.Statistical Analysis—All of the values are expressed as the

means � S.E. Statistical significance was determined by Stu-dent’s t test, and a value of p � 0.05 was considered significant.

RESULTS

HO-1 and CO Inhibit Hyperoxic MLEC Death—MLEC wereexposed to hyperoxia (95%O2, 5% CO2) in vitro and thenmon-itored for hyperoxia-induced toxicity by assaying LDH in thesupernatants. Hyperoxia caused a time-dependent cell death inMLEC as determined by viability and LDH release assays, evi-

dent at 24 h, which increased until72 h of exposure (Fig. 1, A and B).Hyperoxia caused the time-depend-ent release of cytochrome c from themitochondria and correspondingincrease in cytosolic cytochrome clevels in MLEC (Fig. 1C).Overexpression of HO-1 by infec-

tion with ho-1 containing adeno-virus inhibited hyperoxia-inducedcell death, as assessed by cell viabilityand LDH release assays (Fig. 1,A andB), relative to LacZ-infected cells.Furthermore, ho-1 infection dramati-cally decreased cytochrome c releasefrom the mitochondria and corre-sponding accumulation in thecytosol (Fig. 1C). Because many ofthe effects of HO-1 expression canbe duplicated by the exogenousapplication of CO and vice versa, wealso tested the effect of CO onhyperoxia-induced cell death.The presence of CO at a concen-

tration of 250 ppm in the hyperoxicenvironment significantly inhibitedhyperoxia-induced cell death andLDH release (Fig. 2, A and B). Thepresence of CO (250 ppm) inhibitedhyperoxia-dependent cytochrome crelease from the mitochondria andits corresponding accumulation inthe cytosol (Fig. 2C).We also exam-ined the effect of CO on thehyperoxia-dependent execution of

executioner caspases. CO inhibited hyperoxia-dependentcaspase-9 activation (Fig. 2D) in MLEC. Finally, CO inhib-ited the time-dependent activation of caspase-3 duringhyperoxic treatment of MLEC (Fig. 2E).CO Inhibits the Extrinsic Apoptotic Pathway—We tested the

hypothesis that the antiapoptotic effects of CO during hyper-oxic stress involve the inhibition of the extrinsic apoptoticpathway. MLEC were exposed to hyperoxia for various periodsof time in the absence or presence of CO (250 ppm). Cell lysateswere immunoprecipitated with Fas followed by immunoblot-ting with caspase-8 to detect DISC formation. Exposure ofMLEC to hyperoxia induced time-dependent DISC formationassociated with the recruitment of pro-caspase-8 to Fas, thecleavage of pro-caspase-8, and the appearance of caspase-8activity (Fig. 3A). The presence of CO during the hyperoxiatreatment decreased theDISC formation and delayed the cleav-age of caspase-8 (Fig. 3A). We also examined cellular signalingevents upstream of DISC formation in response to hyperoxia.Previously, several investigations have demonstrated interac-tions between Fas and the epidermal growth factor receptor(EGFR) and demonstrated tyrosine phosphorylation of Fas byEGFR as a prerequisite for Fas plasma membrane traffickingand DISC formation (32, 33).

FIGURE 2. CO protected against hyperoxia-induced cell death. MLEC were cultured under 95% air/5% CO2(normoxia), or 95% O2, 5% CO2 (hyperoxia) in the absence or presence of CO (250 ppm) for the times indicated.Viability was determined by trypan blue exclusion (A). Supernatants were analyzed by LDH release assay (B).Cytosolic or mitochondrial fractions were prepared from total lysates as described under “Experimental Pro-cedures.” The indicated fractions were subjected to immunoblotting (IB) to detect cytochrome c (C) orcaspase-9 (D). The lysates were also assayed for caspase-3 activity (E). Western results are representative ofthree experiments. �-Actin served as the loading standard. The data represent the means � S.E. of threeindependent experiments. *, p � 0.05.

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

JANUARY 19, 2007 • VOLUME 282 • NUMBER 3 JOURNAL OF BIOLOGICAL CHEMISTRY 1721

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

Hyperoxia increased EGFR phosphorylation and promotedits associationwith Fas (Fig. 3B). The presence ofCOduring thehyperoxia treatment significantly inhibited EGFR phosphoryl-ation and decreased the association of Fas with EGFR (Fig. 3B).We hypothesized that hyperoxia-induced ROS generation reg-ulates early events in extrinsic apoptosis. MLEC were treated

with a NADPH oxidase inhibitor,dipheyleneiodonium or apocynin,prior to their exposure to hyperoxia.Interestingly, the inhibition ofNADPH oxidase with such com-pounds dramatically decreased theDISC formation, as represented bythe associations between caspase-8and/or EGFRwith Fas (Fig. 3C). CO,which attenuated DISC formation,also inhibited hyperoxia-inducedROS production (Fig. 3D). To con-firm these observations with agenetic strategy, MLEC were thenisolated from mice deficient in thep47phox subunit of NADPH oxidase.Previous studies show that p47phoxis essential for ROS production inendothelial cells in response tostimuli such as phorbol esters andtumor necrosis factor-� (34).Hyperoxia induced the associationof caspase-8 and Fas (DISC forma-tion) in wild type MLEC. However,hyperoxia did not induce DISC for-mation in p47�/� MLEC (Fig. 3E).These results suggest that NADPHoxidase-derived ROS play a criticalrole in the activation of the extrinsicapoptotic pathway in response tohyperoxia and provide amechanismby which CO inhibits the activationof this pathway.CO Inhibits DISC Trafficking—Re-

cent evidence suggests that DISC formation at the plasmamem-brane is preceded by translocation from the Golgi apparatus.To test the effect of hyperoxia on translocation of theDISC, celllysates from MLEC exposed to hyperoxia were separated intoGolgi complex and plasma membrane fractions. DISC forma-tion was assessed by immunoprecipitation with Fas followed byimmunoblotting with caspase-8. Hyperoxia induced DISC for-mation in the Golgi complex and in the plasmamembrane (Fig.4). Inclusion of CO during the hyperoxic exposure resulted inthe retention of the DISC in the Golgi complex and decreasedDISC formation in the plasma membrane relative to exposureto hyperoxia alone. These results suggested that the DISC ispreformed in the Golgi complex and translocates to the plasmamembrane during hyperoxia exposure, whereas the presence ofCO inhibits the DISC trafficking.CO Inhibits Bax Activation and Translocation—We tested

the hypothesis that CO inhibits hyperoxia-induced cell deathby inhibiting the intrinsic apoptotic pathway. Bax is a majorregulator of intrinsic apoptosis but also amplifies extrinsic apo-ptosis through interaction with Bid, which facilitates its activa-tion. Hyperoxia induced the total expression level of Bax inMLEC in a time-dependent fashion with an apparent maxi-mum at 48 h of exposure, relative to normoxic cultures (Fig.5A). The presence of CO (250 ppm) during the hyperoxia

FIGURE 3. CO inhibits extrinsic apoptotic pathways induced by hyperoxia. MLEC were cultured under 95%room air, 5% CO2 (normoxia) or 95% O2, 5% CO2 (hyperoxia) in the absence or presence of CO (250 ppm) for thetimes indicated (A and B). Their lysates were then subjected to immunoprecipitation with anti-Fas followed byWestern blot analysis to detect caspase-8 (A, upper panel), The total lysates were also subjected to Western blotanalysis to detect caspase-8 (A, middle panel). MLEC were cultured under 95% O2, 5% CO2 (hyperoxia) in theabsence or presence of 250 ppm of CO (B) or NADPH oxidase inhibitors, apocynin (300 �M), or diphenylenei-odoniun (DPI, 10 �M) (C), for the times indicated. Their lysates were then subjected to immunoprecipitationwith anti-Fas followed by Western blot analysis as indicated to detect EGFR (B and C, upper panels), or caspase-8(C, lower panel). The lysates were also subjected to Western blot analysis to detect EGFR phosphorylation (B).Western blots are representative of three independent experiments. Total EGFR served as the standard (B).MLEC were loaded with DCFDA (10 �M) and then exposed to hyperoxia (95% O2, 5% CO2) for 3 h. ROS gener-ation was determined by the fluorimetric measurement of DCFDA oxidation (D) as described under “Experi-mental Procedures.” The measurements represent the means � S.E. of triplicate determinations. **, p � 0.01(D). MLEC isolated from wild type or p47�/� mice were cultured under 95% room air, 5% CO2 (normoxia) or 95%O2, 5% CO2 (hyperoxia) for the indicated times. Their lysates were subjected to immunoprecipitation withanti-Fas followed by Western blot analysis to detect caspase-8 (E). IB, immunoblot; IP, immunoprecipitation;WT, wild type.

FIGURE 4. CO inhibits the DISC trafficking from the Golgi complex to theplasma membrane. MLEC were cultured under 95% O2, 5% CO2 (hyperoxia)in the absence or presence of CO (250 ppm) for the times indicated. Proteinsfrom the Golgi complex or plasma membrane were separated from thecell lysates and subjected to immunoprecipitation with anti-Fas followedby Western blot analysis as indicated to detect caspase-8. Western blotsare representative of three independent experiments. IB, immunoblot; IP,immunoprecipitation.

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

1722 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 3 • JANUARY 19, 2007

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

delayed the time-dependent expression of Bax, resulting indiminished Bax expression at 48 h (Fig. 5A). Furthermore,hyperoxia induced the time-dependent cleavage of Bid,whereas the presence of CO significantly decreased hyperoxia-inducible Bid activation (Fig. 5B).We also tested the hypothesis that CO may inhibit Bax acti-

vation and translocation from cytosol to mitochondria. Celllysates were immunoprecipitated with the anti-Bax mono-clonal antibody (6A7) that specifically recognizes a conforma-tional change in Bax protein associated with its activation (10).The presence of CO inhibited the activation of Bax after 48 h ofexposure to hyperoxia, relative to high levels of activated Baxobserved in cells exposed to hyperoxia alone (Fig. 5C). Next weisolated cytosolic and mitochondrial fractions from MLECafter 48 h of exposure to hyperoxia. At this time point, Bax wasdetected primarily in the mitochondrial fraction relative to thecytosolic fraction (Fig. 5C). In the presence of CO (250 ppm),Bax was detected primarily in the cytosolic fraction relative tothemitochondrial fraction (Fig. 5C). These results indicate thatCO inhibited Bax trafficking from the cytosol to the mitochon-dria during hyperoxia exposure. We have previously shownthat Bid binds to and assists in the conformational change ofBax required for its activation (10). To test the potential asso-ciation of Bid and Bax under these conditions,MLECwere sub-jected to hyperoxia (48 h) in the absence or presence of CO (250ppm), and total cell lysates were immunoprecipitated with 6A7

antibody followed by immunoblot-ting with Bid. In cells exposed tohyperoxia, an association of Bidwith the activated form of Bax wasdetected in the total lysates and inthe mitochondrial fractions. Thepresence of CO inhibited the associ-ation of Bid with activated Bax inthemitochondrial fraction ofMLECexposed to hyperoxia, relative tothat observed inMLEC treated withhyperoxia alone (Fig. 5D). We alsohypothesized that CO may modu-late the interaction of Bax with anti-apoptotic proteins. The presence ofCO during hyperoxia promoted theinteraction of Bcl-XL with Bax andincreased the phosphorylation ofBad relative to hyperoxia treatmentalone (Fig. 6A). Interestingly, COtreatment promoted an associationof HO-1 with Bax, specifically inmitochondrial fractions (Fig. 6B).

DISCUSSION

Previous studies from this labora-tory and others describe apoptosisas a major histological feature ofhyperoxia-induced lung injury invivo (4, 7–9, 35, 36) and murinemacrophage cell lines in vitro (37).In contrast, hyperoxia primarily

causes necrosis in A549 alveolar type II epithelial cells. Simi-larly, necrosis, but not apoptosis, was observed in type II cellsisolated from rats exposed to hyperoxia (10, 38). Recent in vitrostudies indicate that the signaling pathways used to initiate celldeath do not predict the final outcome of cell necrosis or apo-ptosis. For example, A549 cells exposed to hyperoxia, althoughpresenting caspase-8 and caspase-9 activation, did not undergoapoptotic cell death but presented morphological features ofnecrosis (10, 39). However, deletion of Bid or inhibition ofcaspase-8 by c-FLIP protected against cell death in this model(10). Integrin-dependent interactions with the extracellularmatrix may determine whether type II cells repair oxygen-in-duced DNA strand breaks or activate the apoptotic machinery(40). The differential ability to affect repair or apoptosis poten-tially explains why SV40-transformed MLE12 mouse type IIcells died by apoptosis after hyperoxia exposure, whereasMLE15 cells, another SV40-transformed line of mouse type IIcells, died primarily by necrosis (41, 42). In pulmonary endo-thelial cells, we observed both apoptosis and necrosis afterhyperoxia exposure by annexin-V/propidium iodide staining(data not shown).Previous work from this laboratory demonstrated that exog-

enous CO prevented hyperoxia-induced lung injury in rats,even when endogenous HO enzyme activity was inhibited withtin protoporphyrin, a competitive inhibitor of HO activity. Ratsexposed to CO also exhibited a marked attenuation of hyper-

FIGURE 5. CO inhibits Bax activation. MLEC were cultured under 95% room air, 5% CO2 (normoxia) or 95% O2,5% CO2 (hyperoxia) in the absence or presence of CO (250 ppm) for the times indicated (A and B). Their lysateswere then subjected to Western blot analysis to detect Bax (A) or tBid (B). Lysates corresponding to 48 exposurewere also subjected to immunoprecipitation with anti-6A7 followed by immunoblotting with Bax to detectactivated Bax (C) or separated into cytosolic (C, middle panel) and mitochondrial fractions (C, lower panel) andsubjected to immunoblotting to detect Bax. The lysates (48 h) were segregated into total or mitochondrialfractions and subjected to immunoprecipitation with anti-6A7 and Western blot analysis to detect Bid (D). p110mitochondrial protein served as the standard for mitochondrial fractions (C and D). IB, immunoblot; IP,immunoprecipitation.

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

JANUARY 19, 2007 • VOLUME 282 • NUMBER 3 JOURNAL OF BIOLOGICAL CHEMISTRY 1723

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

oxia-induced neutrophil infiltration into the airways and totallung apoptotic index (18). However, the mechanisms of COprotection against hyperoxic lung cell death have not been welldefined. In the current study, we demonstrated that CO inhib-ited cytochrome c and caspase-9 activation in hyperoxicMLECby blocking both extrinsic and intrinsic apoptotic pathways.CO diminished DISC formation and associated activation ofcaspase-8 and Bid. CO exposure inhibited Bax activation andblocked its translocation into mitochondria, thus preventingcytochrome c release and subsequent caspase-9 activation (Fig.5C). We also showed that CO blocked the association of Bidwith the activated form of Bax (Fig. 5D). We previouslyreported that Bid assists in the conformational change associ-ated with Bax activation (10), and the current data suggest thatCO inhibits this processing. tBid assists Bax not only throughprotein-protein interactions but also by protein-lipid interac-tions to form lipid pore-type structures in the outer mitochon-drial membrane through which intermembrane pro-deathmolecules exit the mitochondria during apoptosis (43).Bcl-XL is a pro-survival member of the Bcl-2 family that

inhibits the activities of the pro-apoptotic members of the fam-ily, such as Bax and Bad. Bad occurs in the cytosol of normalcells, where it is phosphorylated at multiple serine residues (11,14). An apoptotic signal triggering Bad dephosphorylation

results in the binding of Bad to Bcl-XL and the inactivation ofthe pro-survival function of Bcl-XL. Uncomplexed Bcl-XL cancounteract the pro-apoptotic activity of Bax by inhibiting Baxmitochondrial translocation and oligomerization, thus pre-venting cytochrome c release (44). In the current study, wefound thatCOpromoted the interaction of Bcl-XLwithBax andenhanced Bad phosphorylation (Fig. 6A). These data suggestthat CO inhibits intrinsic apoptotic signaling pathways throughincreased Bad phosphorylation and by promoting the bindingof Bax with Bcl-XL, which blocks the oligomerization of Baxnecessary for cytochrome c release.In death receptor (Fas)-mediated apoptosis, the current

model suggests that the DISC, consisting of Fas/FADD/caspase-8, forms in the plasma membrane upon ligand (FasL)stimulation. Autoproteolytic generation of caspase-8 from pro-caspase-8 occurs within the DISC. However, the precise mech-anisms underlying DISC formation and the translocations ofFas, FADD, and caspase-8 are not entirely known. In the hy-poxia/reoxygenation model, we previously reported that theDISC forms initially in theGolgi complex and then translocatesto the plasma membrane (31, 45). DISC formation in the Golgicomplex is important for initiating apoptotic signaling (31),whereas its formation in the plasma membrane is critical forcaspase-8 activation.Recent studies in hepatocytes show that FasL-induced apo-

ptosis involved ROS generation and, consequently, the c-JunNH2-terminal kinase-dependent tyrosine phosphorylation andactivation of EGFR and its subsequent associationwith Fas (33).In this model, EGFR-mediated tyrosine phosphorylation of Faswas required for the trafficking of Fas to the membrane, forDISC formation, and for the initiation of the extrinsic apoptoticpathway (33). Consistent with this model, we observed thathyperoxia treatment promoted EGFR phosphorylation and itsassociation with Fas in MLEC, in parallel with DISC activation(Fig. 3B). Fas/caspase-8 association and Fas/EGFR associationdepended on ROS generation, because these processes wereinhibitable by NADPH oxidase inhibitors (Fig. 3C), absent inp47�/�MLEC (Fig. 3E) andwere also inhibited byCO (Fig. 3B),which down-regulated ROS production (Fig. 3D). Recently, wehave also demonstrated that CO inhibited ROS-dependenttrafficking of Toll-like receptor-4 to lipid rafts by down-regu-lating NADPH oxidase activity (47). The dependence of ROS inDISC activation during hyperoxia and the inhibition of theseprocesses by CO provide a distinct mechanism by which COinhibits extrinsic apoptosis.CO has previously been shown to confer cellular or tissue

protection in multiple models of stress or injury. In rodentmodels of hyperoxia-induced lung injury, CO exerts potentanti-inflammatory effects with reduced inflammatory cellinflux into the lungs and marked attenuation in the expressionof pro-inflammatory cytokines (17). CO suppressed arterio-sclerotic lesion formation associated with chronic graft rejec-tion andwith balloon injury (27) and inhibited apoptosis duringischemia-reperfusion injury (26, 48, 49). These effects wereassociated with the CO-dependent activation of the p38 mito-gen-activated protein kinase pathway.HO-1, a stress-inducible protein, responds to transcriptional

up-regulation by hyperoxic conditions. Exogenous administra-

FIGURE 6. CO promotes Bax complex formation with Bcl-XL and HO-1.MLEC were cultured under 95% room air, 5% CO2 (normoxia) or 95% O2, 5%CO2 (hyperoxia) in the absence or presence of CO (250 ppm). Total lysates(0 –24 h of exposure) were subjected to immunoprecipitation with anti-Bcl-XLand Western blot analysis to detect Bax (A, upper panel). The total lysates werealso subjected to Western blot analysis to detect phosphorylated-Bad (A, mid-dle panel). Total Bax served as the standard (A, lower panel). Proteins frommitochondria and cytosol were separated from the cell lysates and wereimmunoprecipitated (IP) with anti-HO-1 followed by Western blot analysis(IB) to detect Bax (B). Western blots are representative of three experiments.�-Actin served as the standard. p110 mitochondrial protein served as thestandard for mitochondrial fractions (B, lower panel). IB, immunoblot; IP,immunoprecipitation.

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

1724 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 3 • JANUARY 19, 2007

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

tion of HO-1 by adenoviral gene transfer increased rat survivaland attenuated lung neutrophil influx and lung cell apoptosisduring hyperoxia (50). Similarly, administration of lowdoseCOmay prevent lung oxidative stress through secondary increasesin HO-1 (21, 46). However, the exact pathway by which HO-1increases cell survival has not yet been elucidated. Consistentwith previously described antiapoptotic roles, we observed thatadenoviral-mediated HO-1 overexpression prevented hyper-oxia-induced cytochrome c release from the mitochondria.HO-1 also inhibited hyperoxia-induced membrane damage, asassessed by LDH release assays, which also suggests generalprotection against necrotic cell death. Here, we described forthe first time that exogenous CO treatment promotes a previ-ously undescribed binding interaction of HO-1 with Bax in themitochondria (Fig. 6B). The functional significance of thisinteraction is currently unknown, although we speculate thatthis association blocks the oligomerization of Bax, thus pre-cluding its ability to destabilizemitochondrial membrane func-tion and initiate cytochrome c release (Fig. 7).

In summary, we have examined cell death pathways in cul-tured MLEC exposed to hyperoxia and found that hyperoxiacaused significant apoptosis, involving both extrinsic (Fas/caspase-8) and intrinsic (Bax/mitochondria)-dependentapoptotic pathways (Fig. 7). CO inhibited hyperoxic celldeath by inhibiting cellular ROS production and antagoniz-ing ROS-dependent DISC formation in the plasma mem-brane, thereby inhibiting caspase-8 activation and Bid cleav-age. CO also blocked Bax mitochondrial translocation byincreasing Bad phosphorylation and promoting a novel

interaction of HO-1 with Bax. Strategies to attenuate hyper-oxia-induced apoptotic pathways may expand the currentlylimited therapeutic options.

REFERENCES1. Horowitz, S., and Davis, J. M. (1997) in Lung Growth and Development

(McDonald, J. A., ed) pp. 577–610, Decker Press, New York, NY2. Crapo, J. D. (1986) Annu. Rev. Physiol. 48, 721–7313. Mantell, L. L., and Lee, P. J. (2000)Mol. Genet. Metab. 71, 359–3704. Freeman, B. A., Topolosky, M. K., and Crapo, J. D. (1982) Arch. Biochem.

Biophys. 216, 477–4845. Li, J., Gao, X., Qian, M., and Eaton, J. W. (2004) Free Radic. Biol. Med. 36,

1460–14706. Parinandi, N. L., Kleinberg, M. A., Usatyuk, P. V., Cummings, R. J., Pen-

nathur, A., Cardounel, A. J., Zweier, J. L., Garcia, J. G., and Natarajan, V.(2003) Am. J. Physiol. 284, L26–L38

7. Crapo, J. D., Barry, B. E., Foscue, H. A., and Shelburne, J. (1980) Am. Rev.Respir. Dis. 122, 123–143

8. Barazzone, C., Horowitz, S., Donati, Y. R., Rodriguez, I., and Piguet, P. F.(1998) Am. J. Respir. Cell Mol. Biol. 19, 573–581

9. Barazzone, C., Donati, Y. R., Rochat, A. F., Vesin, C., Kan, C. D., Pache,J. C., and Piguet, P. F. (1999) Am. J. Pathol. 154, 1479–1487

10. Wang, X., Ryter, S.W., Dai, C., Tang, Z. L.,Watkins, S. C., Yin, X.M., Song,R., and Choi, A. M. (2003) J. Biol. Chem. 278, 29184–29191

11. Martinou, J. C., and Green, D. R. (2001)Nat. Rev. Mol. Cell. Biol. 2, 63–6712. Antonsson, B., Montessuit, S., Lauper, S. Eskes, R., and Martinou, J. C.

(2000) Biochem. J. 345, 271–27813. Huang, D. C., and Strasser, A. (2000) Cell 103, 839–84214. Goping, I. S., Gross, A., Lavoie, J. N., Nguyen,M., Jemmerson, R., Roth, K.,

Korsmeyer, S. J., and Shore, G. C. (1998) J. Cell Biol. 143, 207–21515. Zou, H., Li, Y., Liu, X., and Wang, X. (1999) J. Biol. Chem. 274,

11549–1155616. Walsh C.M., Luhrs, K. A., and Arechiga, A. F. (2003) J. Clin. Immunol. 23,

333–35317. Otterbein, L. E., Otterbein, S. L., Ifedigbo, E., Liu, F., Morse, D. E., Fearns,

C., Ulevitch, R. J., Knickelbein, R., Flavell, R. A., and Choi, A. M. (2003)Am. J. Pathol. 163, 2555–2563

18. Otterbein, L. E., Mantell, L. L., and Choi, A. M. (1999) Am. J. Physiol. 276,L688–L694

19. Tenhunen, R., Marver, H., and Schmid, R. (1969) J. Biol. Chem. 244,6388–6394

20. Ryter, S. W., Alam, J., and Choi, A. M. (2006) Physiol. Rev. 86, 583–65021. Morita, T., Mitsialis, S. A., Koike, H., Liu, Y., and Kourembanas, S. (1997)

J. Biol. Chem. 272, 32804–3280922. Otterbein, L. E., Bach, F. H., Alam, J., Soares, M., Tao Lu, H., Wysk, M.,

Davis, R. J., Flavell, R. A., and Choi, A. M. (2000) Nat. Med. 6, 422–42823. Song, R., Kubo,M., Morse, D., Zhou, Z., Zhang, X., Dauber, J. H., Fabisiak,

J., Alber, S. M., Watkins, S. C., Zuckerbraun, B. S., Otterbein, L. E., Ning,W., Oury, T. D., Lee, P. J., McCurry, K. R., and Choi, A. M. (2003) Am. J.Pathol. 163, 231–242

24. Brouard, S., Otterbein, L. E., Anrather, J., Tobiasch, E., Bach, F. H., Choi,A. M., and Soares, M. P. (2000) J. Exp. Med. 192, 1015–1026

25. Liu, X. M., Chapman, G. B., Peyton, K. J., Schafer, A. I., and Durante, W.(2002) Cardiovasc. Res. 55, 396–405

26. Zhang, X., Shan, P., Alam, J., Davis, R. J., Flavell, R. A., and Lee, P. J. (2003)J. Biol. Chem. 278, 22061–22070

27. Otterbein, L. E., Zuckerbraun, B. S., Haga,M., Liu, F., Song, R., Usheva, A.,Stachulak, C., Bodyak, N., Smith, R. N., Csizmadia, E., Tyagi, S., Akamatsu,Y., Flavell, R. J., Billiar, T. R., Tzeng, E., Bach, F. H., Choi, A.M., and Soares,M. P. (2003) Nat. Med. 9, 183–190

28. Ryter, S., and Otterbein, L. (2004) Bioessays 26, 270–28029. Kim, H. P., Ryter, S. W., and Choi, A. M. (2006)Ann. Rev. Pharm. Tox. 46,

411–44930. Wang, X., Zhou, Y., Kim, H. P., Song, R., Zarnegar, R., Ryter, S. W., and

Choi, A. M. (2004) J. Biol. Chem. 279, 5237–524331. Wang, X., Zhang, J., Kim, H. P., Wang, Y., Choi, A. M., and Ryter, S. W.

(2004) FASEB J. 18, 1826–1833

FIGURE 7. CO inhibits extrinsic and intrinsic apoptotic pathways duringhyperoxia. The diagram depicts the mechanisms by which CO provides pro-tection against cell death. Hyperoxia triggers both mitochondrial (intrinsic)and death receptor-dependent (extrinsic) apoptotic pathways in MLEC. COinhibited hyperoxic cell death by decreasing ROS production, EGFR phospho-rylation, as well as the interaction between EGFR and Fas. CO reduced theformation of the DISC in the plasma membrane, by retaining the DISC in theGolgi apparatus. CO increased Bad phosphorylation and promoted an asso-ciation of Bcl-XL and/or HO-1 with Bax.

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

JANUARY 19, 2007 • VOLUME 282 • NUMBER 3 JOURNAL OF BIOLOGICAL CHEMISTRY 1725

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

32. Eberle, A., Reinehr, R., Becker, S., and Haussinger, D. (2005) Hepatology41, 315–326

33. Reinehr, R., Schliess, F., and Haussinger, D. (2003) FASEB J. 17,731–733

34. Li, J. M.,Mullen, A.M., Yun, S.,Wientjes, F., Brouns, G. Y., Thrasher, A. J.,and Shah, A. M. (2002) Circ. Res. 90, 143–150

35. Mantell, L. L., Horowitz, S., Davis, J.M., andKazzaz, J. A. (1999)Ann. N. Y.Acad. Sci. 887, 171–180

36. Otterbein, L. E., Chin, B. Y.,Mantell, L. L., Stansberry, L., Horowitz, S., andChoi, A. M. (1998) Am. J. Physiol. 275, L14–L20

37. Petrache, I., Choi,M. E.,Otterbein, L. E., Chin, B. Y.,Mantell, L. L., Horow-itz, S., and Choi, A. M. (1999) Am. J. Physiol. 277, L589–L595

38. O’Reilly,M.A., Staversky, R. J.,Watkins, R.H., Reed, C. K., deMesy Jensen,K. L., Finkelstein, J. N., and Keng, P. C. (2001)Am. J. Respir. Cell Mol. Biol.24, 703–710

39. Kazzaz, J. A., Xu, J., Palaia, T. A., Mantell, L., Fein, A. M., and Horowitz, S.(1996) J. Biol. Chem. 271, 15182–15186

40. Roper, J. M., Mazzatti, D. J., Watkins, R. H., Maniscalco, W. M., Keng,P. C., and O’Reilly, M. A. (2004) Am. J. Physiol. 286, L1045–L1054

41. O’Reilly, M. A., Staversky, R. J., Finkelstein, J. N., and Keng, P. C. (2003)Am. J. Physiol. 284, L368–L375

42. Zhang, X., Shan, P., Sasidhar, M., Chupp, G. L., Flavell, R. A., Choi, A. M.,and Lee, P. J. (2003) Am. J. Respir. Cell Mol. Biol. 28, 305–315

43. Terrones, O., Antonsson, B., Yamaguchi, H., Wang, H. G., Liu, J., Lee,R. M., Herrmann, A., and Basanez, G. (2004) J. Biol. Chem. 279,30081–30091

44. Zamzami, N., and Kroemer, G. (2001) Nat. Rev. Mol. Cell. Biol. 2, 67–7145. Wang, X., Wang, Y., Zhang, J., Kim, H. P., Ryter, S. W., and Choi, A. M.

(2005)Mol. Cell. Biol. 25, 4742–475146. Sawle, P., Foresti, R., Mann, B. E., Johnson, T. R., Green, C. J., andMotter-

lini, R. (2005) Br. J. Pharmacol. 145, 800–81047. Nakahira, K., Kim, H. P., Geng, X. H., Nakao, A., Wang, X., Murase, N.,

Drain, P. F., Wang, X., Sasidhar, M., Nabel, E. G., Takahashi, T., Lukacs,N. W., Ryter, S., Morita, K., and Choi, A. M. K. (2006) J. Exp. Med. 203,2377–2389

48. Zhang, X., Shan, P., Otterbein, L. E., Alam, J., Flavell, R. A., Davis, R. J.,Choi, A. M., and Lee, P. J. (2003) J. Biol. Chem. 278, 1248–1258

49. Amersi, F., Shen, X. D., Anselmo, D., Melinek, J., Iyer, S., Southard, D. J.,Katori, M., Volk, H. D., Busuttil, R.W., Buelow, R., and Kupiec-Weglinski,J. W. (2002) Hepatology 35, 815–823

50. Otterbein, L. E., Kolls, J. K., Mantell, L. L., Cook, J. L., Alam, J., and Choi,A. M. (1999) J. Clin. Investig. 103, 1047–1054

Antiapoptotic Effect of Carbon Monoxide in Hyperoxia

1726 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 3 • JANUARY 19, 2007

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from

Augustine M. K. ChoiXue Wang, Yong Wang, Hong Pyo Kim, Kiichi Nakahira, Stefan W. Ryter and

by Inhibiting Reactive Oxygen Species FormationCarbon Monoxide Protects against Hyperoxia-induced Endothelial Cell Apoptosis

doi: 10.1074/jbc.M607610200 originally published online November 29, 20062007, 282:1718-1726.J. Biol. Chem. 

  10.1074/jbc.M607610200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/282/3/1718.full.html#ref-list-1

This article cites 43 references, 14 of which can be accessed free at

by guest on June 7, 2020http://w

ww

.jbc.org/D

ownloaded from