The Nitrated Fatty Acid 10-Nitro-oleate Diminishes Severity of LPS-Induced Acute Lung Injury in Mice

13 Pages·2014·3.54 MB·English

HindawiPublishingCorporation
PPARResearch
Volume2012,ArticleID617063,12pages
doi:10.1155/2012/617063
Research Article
The Nitrated Fatty Acid 10-Nitro-oleate Diminishes
Severity of LPS-Induced Acute Lung Injury in Mice
AravindT.Reddy,SowmyaP.Lakshmi,andRajuC.Reddy
DepartmentofMedicine,DivisionofPulmonary,AllergyandCriticalCareMedicine,
EmoryUniversityandAtlantaVAMedicalCenter,Atlanta,GA30033,USA
CorrespondenceshouldbeaddressedtoRajuC.Reddy,[email protected]
Received24February2012;Accepted21April2012
AcademicEditor:JesseRoman
Copyright©2012AravindT.Reddyetal. This is an open access article distributed under the Creative Commons Attribution
License,whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperly
cited.
Acute lung injury (ALI) is an inflammatory condition culminating in respiratory failure. There is currently no effective
pharmacological treatment. Nitrated fatty acids (NFAs) have been shown to exert anti-inflammatory effects. We therefore
hypothesized that delivery of NFAs directly to the site of inflammation would reduce the severity of ALI. Pulmonary delivery
of10-nitro-oleatefollowingendotoxin-inducedALIinmicereducedmarkersoflunginflammationandinjury,includingcapillary
leakage, lung edema, infiltration of neutrophils into the lung, and oxidant stress, as well as plasma levels of proinflammatory
cytokines.Nitro-oleatedeliverylikewisedownregulatedexpressionofproinflammatorygenesbyalveolarmacrophages,keycells
inregulationoflunginflammation.TheseeffectsmaybeaccountedforbytheobservedincreasesintheactivityofPPAR-γand
thePPAR-γ-inducedantioxidanttranscriptionfactorNrf2,togetherwiththedecreasedactivityofNF-κB.Ourresultsdemonstrate
thatpulmonarydeliveryofNFAsreducesseverityofacutelunginjuryandsuggestpotentialutilityofthesemoleculesinother
inflammatorylungdiseases.
1.Introduction Among the many anti-inflammatory effects attributable to
PPAR-γ activationarediminishedincreasesinreactiveoxy-
A variety of pulmonary and extrapulmonary insults can genspecies,cytokines,chemokines,andadhesionmolecules
result in acute lung injury (ALI), which is characterized [10]. A major mechanism underlying these actions is
by capillary leakage and resulting pulmonary edema and decreased activity of proinflammatory transcription factors
hypoxemia[1].ThesemultipleoriginsofALIarereflectedin such as NF-κB, AP-1, and STAT [5, 6]. Synthetic PPAR-
differentanimalmodelsofthedisease,ofwhichpulmonary γ agonists are widely used for treatment of diabetes but
administration of bacterial endotoxin (lipopolysaccharide; have been associated with adverse effects. A wide variety
LPS) is among the most common. Regardless of the pre- of endogenous molecules are known to activate PPAR-
cipitating cause, the earliest phases of ALI feature severe γ, but most either exhibit low potency or are present at
neutrophil-rich alveolar inflammation [2] and associated low concentrations, leading to uncertainty regarding their
oxidant stress [3] that represent the proximate causes of physiologicalrole.
much or all of the subsequent pulmonary injury. ALI Nitrated fatty acids (NFAs) are activating ligands for all
morbidity and mortality remain high and there is no three PPARs, exhibiting their greatest potency as PPAR-γ
effectivepharmacotherapy[4],underliningtheurgentneed agonists [11, 12]. They have also been shown to exhibit a
forimprovedtreatmentmodalities. number of PPAR-γ-dependent effects, including induction
Peroxisome proliferator-activated receptor γ (PPAR-γ) of adipogenesis [11] and of CD36 receptor expression by
plays a central role in many of the feedback loops that macrophages [12]. NFAs are produced endogenously by
normallylimitinflammationandleadtoitsresolution[5–9] nonenzymatic reaction of NO or its inorganic reaction
andisthereforeapromisingtargetforALIpharmacotherapy. products with naturally present unsaturated fatty acids
2 PPARResearch
[13], with positional isomers of nitro-oleic acid (OA- supernatant following BAL fluid centrifugation was deter-
NO ) and nitrolinoleic acid (LNO ) found at the highest mined using the BCA Protein Assay kit (Pierce, Rockford,
2 2
concentrations in human plasma [11, 14]. The suggested IL).
PPAR-γ-mediated anti-inflammatory effects of NFAs have
been supported by in vitro studies [15, 16], but in vivo 2.6. Lung Histopathology. The lungs were inflated and fixed
and potentially translational studies of such effects have with 10% neutral formalin overnight at room tempera-
been limited. These considerations led us to investigate the ture. Lung tissue was dehydrated with increasing ethanol
abilityoftreatmentwithOA-NO2,themostpotentPPAR-γ- (EtOH) concentrations and then embedded in paraffin.
activatingNFA[11],toreducetheseverityofinflammation Five-micrometer-thick paraffin sections were stained with
andlunginjuryinamurinemodelofALIinducedbyLPS. hematoxylinandeosin(H&E).
In order to maximize pulmonary availability, OA-NO was
2
delivereddirectlytothelungviatheintratrachealroute.
2.7. Assessment of Capillary Leakage. To further assess lung
permeability, 50mg/kg of Evans Blue Dye (EBD; Sigma-
2.MaterialsandMethods Aldrich, St. Louis, MO) dissolved in 200μL of PBS was
injectedintothetailveinsofmicefollowingALIinduction.
2.1. Animals. Male C57BL/6 mice were obtained from After 30min, the animals were euthanized and the lungs
JacksonLaboratories(BarHarbor,ME)andwereusedat6–8 perfusedwith5mLPBS,afterwhichthelungswereexcised
weeksofage(20–25g).Allstudieswereperformedaccording enblocandsnap-frozeninliquidnitrogen.Thefrozenlungs
toprotocolsreviewedandapprovedbytheAtlantaVeterans were then homogenized in 2mL PBS. The homogenate
Affairs Medical Center Institutional Animal Care and Use was diluted with 2vol of formamide and incubated at
Committee. 60◦C for 18h, followed by centrifugation at 5,000×g for
30min. The supernatant was collected and absorbance was
measured at 620 and 740nm. The EBD concentration was
2.2. OA-NO and LPS Administration. Mice were anes-
2
determined from standard absorbance curves evaluated in
thetized with 90mg/kg ketamine and 10mg/kg xylazine,
parallel. Correction for contaminating heme pigments was
administered via intraperitoneal injection, and a tra-
calculatedbytheformulaE (EBD)=E −(1.426×E +
cheotomy was performed. ALI was then induced by intra- 620 620 740
0.030).TheEBDconcentrationwasexpressedasμgpergof
tracheal(i.t.)injectionof50μgofendotoxin(LPS)prepared
lung.
from Escherichia coli O111:B6 (Sigma-Aldrich, St. Louis,
MO).Thirtyminlatermicewereinjectedi.t.with50μgOA-
NO (Cayman Chemical, Ann Arbor, MI) in 50μL of 10% 2.8. Measurement of Myeloperoxidase Activity. As an index
2
DMSO or with vehicle. After a further 5.5h the mice were of neutrophil infiltration, BAL fluid and tissue-associated
euthanized, plasma was obtained, and lung and BAL fluid myeloperoxidase (MPO) activity was determined. Frozen
sampleswerecollected.Neutrophilinfiltrationintothelung lung tissues were thawed, weighed, homogenized, and
peaksapproximately6hafteri.t.LPSadministration[17]. sonicated on ice in radioimmunoprecipitation assay buffer
(RIPA).Aftercentrifugationat10,000×g at4◦Cfor20min,
thesupernatantwascollectedandusedfordeterminationof
2.3.LungWet:DryWeightRatio. Asanindexoflungedema,
MPOactivitybyacommerciallyavailablefluorometricassay
theamountofextravascularlungwaterwascalculated.The
kit (700160; Cayman Chemical) according to the manufac-
lowerlobeoftherightlungwasligatedandexcisedandthe
turer’s instructions. Results were expressed as nmol/min-
wet weight was recorded. The lung was then placed in an
ml. Similar measurements were performed on BAL fluid
incubator at 60◦C for 24h to obtain the dry weight. The
supernatant.
wet:dry ratio was calculated by dividing the wet weight by
thedryweight.
2.9. Measurement of Oxidant Stress. Hydrogen peroxide
(H O )productioninlungtissuewasdeterminedusingthe
2 2
2.4. Bronchoalveolar Lavage (BAL) Fluid Collection and Cell AmplexRedHydrogenPeroxideAssaykit(MolecularProbes,
Count. Following removal of the lung’s lower right lobe, Eugene,OR)accordingtothemanufacturer’sdirections.The
BALfluidwascollectedbyflushing3×1mLofphosphate- concentrations of nitrate and malondialdehyde (MDA) in
buffered saline (PBS) containing 0.1mM EDTA into the lunghomogenatesweremeasuredusingcommerciallyavail-
lung via a tracheal cannula. The pooled BAL fluid was ablecolorimetricassaykits(CaymanChemical)accordingto
centrifuged at 500×g at 4◦C for 5min. Pelleted cells were themanufacturer’sinstructions.
then resuspended in 1mL of PBS. Total cell number was
countedbyhemocytometerandadifferentialcellcountwas
2.10. Measurement of Plasma Cytokine Levels. Plasma levels
performed by cytospin staining with Diff-Quik (Siemens,
of tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6),
Newark,DE).
keratinocytechemoattractant(KC),andmacrophageinflam-
matory protein-2 (MIP-2) were measured using enzyme-
2.5.BALFluidProtein. IncreaseinBALfluidproteinconcen- linked immunosorbent assay (ELISA) kits (R&D Systems,
trationwastakenasameasureofincreasedpermeabilityof Minneapolis,MN)accordingtothemanufacturer’sinstruc-
alveolar-capillarybarriers.Totalproteinconcentrationinthe tions.
PPARResearch 3
Table1:Oligonucleotideprimersemployed.
Gene Primersequence Tm◦C Ampliconsize(bp)
F 5(cid:3)-ATCCGCCAACGTCACATT-3(cid:3) 57.6
9srRNA 115
R 5(cid:3)-CCGCCGCCATAAGGAGAAC-3(cid:3) 64.5
F 5(cid:3)-CAGTCGGAGACATGCTTATTGAG-3(cid:3) 60.7
CD36 151
R 5(cid:3)-TTTGCCACGTCATCTGGGTTT-3(cid:3) 62.5
F 5(cid:3)-TGTGACTGTACCCGGACTGG-3(cid:3) 63.0
COX-2 233
R 5(cid:3)-TGCACATTGTAAGTAGGTGGAC-3(cid:3) 60.0
F 5(cid:3)-GGGGCCAGGCTTCTATTCC-3(cid:3) 61.8
FABP4 114
R 5(cid:3)-GGAGCTGGGTTAGGTATGGG-3(cid:3) 61.1
F 5(cid:3)-GGACGCATTGGTCGTCTGG-3(cid:3) 63.0
GAPDH 204
R 5(cid:3)-TTTGCACTGGTACGTGTTGAT-3(cid:3) 60.2
F 5(cid:3)-CAAGGCTGTTCACATTATCCCA-3(cid:3) 60.3
IL-12p40 107
R 5(cid:3)-CCAGTGTGGTCATGGACTTTC-3(cid:3) 60.6
F 5(cid:3)-TTAAAAACCTGGATCGGAACCAA-3(cid:3) 60.1
MCP-1 121
R 5(cid:3)-GCATTAGCTTCAGATTTACGGGT-3(cid:3) 60.7
F 5(cid:3)-ACATCGACCCGTCCACAGTAT-3(cid:3) 62.7
NOS2 177
R 5(cid:3)-CAGAGGGGTAGGCTTGTCTC-3(cid:3) 61.0
F 5(cid:3)-TGCCTGCTCATTTGGCTGT-3(cid:3) 62.2
NOX4 180
R 5(cid:3)-CCGGCACATAGGTAAAAGGATG-3(cid:3) 60.8
F 5(cid:3)-CCATTCTGGCCCACCAAC-3(cid:3) 66.5
PPAR-γ 479
R 5(cid:3)-CTGAAACCGACAGTACTG-3(cid:3) 53.8
F 5(cid:3)-CCTGTAGCCCACGTCGTAG-3(cid:3) 61.5
TNF-α 148
R 5(cid:3)-GGGAGTAGACAAGGTACAACCC-3(cid:3) 61.4
2.11. RNA Isolation and Quantitative Real-Time RT-PCR. 4 TK-luciferase) and a plasmid containing the PPAR-γ
Cells pelleted from BAL fluid were resuspended in DMEM ligand-binding domain fused to the Gal4 DNA-binding
supplemented with 10% fetal bovine serum, allowed to domain or with the luciferase gene under control of the
adhere to tissue culture-treated six-well plates for 1h, and peroxisome proliferator response element (PPRE) isolated
then washed twice to remove nonadherent cells. Adherent fromthefattyacidtransportprotein.Alltransfectionswere
alveolar macrophages were lysed, and RNA was isolated performedusingLipofectamine2000(Invitrogen)according
using RNeasy Mini kit (Qiagen, Valencia, CA), with cDNA tothemanufacturer’sinstructions.Followingtreatmentwith
beinggeneratedfrom100ngoftotalRNAusingMultiScribe test compounds, activation was measured using the Dual-
reverse transcriptase (Applied Biosystems, Foster City, CA) Luciferase Reporter Assay System (Promega, Madison, WI)
employing random and oligo dT primers. Real-time quan- accordingtothemanufacturer’sinstructions.
titative PCR was performed using 100ng cDNA with 2X
SYBR Green Master mix (Applied Biosystems) and specific 2.13. Transcription Factor DNA-Binding Activity Assay.
primersforthegenesofinterest(Table1).Theseexperiments Nuclear proteins were extracted using a nuclear extraction
were performed on an AB 7500 fast thermal cycler using kit(ActiveMotif,Carlsbad,CA) accordingtothemanufac-
athree-step protocol employing the melting curvemethod. turer’sprotocol.Theproteinconcentrationwasdetermined
Theaverageofeachgenecyclethreshold(Ct)wasdetermined using the BCA Protein Assay kit (Pierce). Nuclear extracts
for each experiment. Relative cDNA levels (2−ΔΔCt) for the were used to quantify DNA-binding activity of PPAR-γ,
genesofinterestweredeterminedbyusingthecomparative nuclear factor (erythroid-derived 2)-like 2 (Nrf2), and the
Ct method, which generates the ΔΔCt as the difference p65 subunit of NF-κB using ELISA-based TransAM kits
between the gene of interest and the housekeeping genes (40096, 40696, and 50296; Active Motif) according to the
glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and manufacturer’sinstructions.
9srRNAforeachsample.Eachaveragedexperimentalgene
expression sample was compared to the averaged control
2.14. Molecular Modeling and Computer Simulations of
sample,whichwassetto1.
Binding of OA-NO with PPAR-γ. In silico construction of
2
OA-NO wascarriedoutusingChemOffice(ChemDrawand
2
2.12. Transient Transfection Assay. PPAR-γ activity in A549 Chem3D;CambridgeSoftCorp.,Cambridge,MA).Toavoid
cells (ATCC, Rockville, MD) was determined as previously sterichindranceandclashesthatcanappearinthebuilding
described [18]. Briefly, cells were transiently co-transfected process, the models obtained were subjected to geometry
either with a plasmid containing the luciferase gene under optimization using GaussView with a protocol of 300 steps
regulation by four Gal4 DNA-binding elements (UAS × ofconjugategradients.Eachmodelthenwasoptimizedusing
G
4 PPARResearch
asemi-empiricalmethodsuchasPM3asimplementedinthe compound. As shown in Figures 1(a) and 1(b), OA-NO
2
Gaussian98packageofprograms. is well accommodated within the PPAR-γ ligand-binding
Autodock 4.2 (with Lamarckian Genetic Algorithm) or site,withbestpose energyof−11.50kcal/mol,andexhibits
ArgusLab 4 was used to generate the starting complexes. the appropriate interactions and hydrogen bonds with the
An elitism value of 1 was used, together with probabilities PPAR-γ aminoacidresidues(ARG288,GLN286,HIS449,
of mutation and crossing-over of 0.03 and 0.07, respec- and TYR 473) that are known to be important for PPAR-γ
tively. From the best solutions obtained according to these activation.
parameters,thosedefinedbytheuserasexhibiting thebest
probabilities—in our case, 0.07—were further refined by a 3.2. OA-NO Activates PPAR-γ In Vitro. To confirm the
2
local search method. Autodock defines the conformational ability of NFAs to activate PPAR-γ we utilized the GAL4
space implementing grids over the entire space of possible reporter system in A549 airway epithelial cells. Transfected
solutions.WiththeaimoftestingtheabilityofAutodockto cells were treated with 0.1, 1, or 5μM concentrations of
converge into solutions that are inside the PPAR-γ ligand- OA-NO or, as a positive control, of the synthetic PPAR-γ
2
binding domain, a grid of 70A˚ per side with 0.4A˚ spacing agonist rosiglitazone. Each compound demonstrated dose-
betweenpointswassetupinsuchawaythatitcoveredboth dependentactivationasaresultofbindingtotheconstruct’s
the external surface and the internal cavity of the PPAR- PPAR-γ ligand-binding domain, with similar activation for
γ ligand-binding domain. The following procedure was given molar concentrations (Figure1(c)). To demonstrate
employed for the OA-NO -PPAR-γ ligand-binding domain activationofendogenousPPAR-γ,A549cellsweresimilarly
2
docking simulations: 150 runs were done for each OA- transfectedwiththeluciferasegeneundercontrolofaPPAR
NO -PPAR-γ ligand-binding domain. At the end of each response element. This assay likewise demonstrated dose-
2
run, the complexes were separated into clusters according dependent activation by both OA-NO and rosiglitazone
2
to their lowest root mean square deviation (RMSD), and (Figure1(d)).
thebestscorevaluebasedonafreeempiricenergyfunction
wasdetermined.Clustercomplexeswhoseaveragescorewas 3.3.PulmonaryAdministrationofOA-NO DiminishesSever-
2
−11.50kcal-mol/L with respect to the best energy obtained ity of LPS-Induced Lung Inflammation. Pulmonary inflam-
in that run were selected. The selected final complexes mationisacrucialfeatureofALI.Sincethenuclearreceptor
were optimized using the semi-empirical PM3 method as a PPAR-γ is known to exert a variety of anti-inflammatory
refiningprocedurewithGaussian98. effects and unsaturated long-chain NFAs are activating lig-
Constant-volume, constant-temperature molecular dy- ands for PPAR-γ, we hypothesized that pulmonary delivery
namics(MD)simulationsofthecomplexeswereperformed of an NFA would diminish the severity of ALI. To test this
ontheDiscover,version2.7(BiosymTechnologies,Inc.,San hypothesis, we utilized a well-established murine model of
Diego, CA) and MD simulation programs in Chem Office. ALI induced by i.t. administration of 50μg of LPS. Thirty
Energyminimizationswereconductedfortwosystems,both min after LPS injection, 50μg of OA-NO in 10% DMSO
2
of which assumed 1 molecule of OA-NO , 1 molecule was delivered to the lungs via the i.t. route. Control mice
2
of PPAR-γ, and water molecules. The systems simulated receivedvehiclewithoutNFA.Afterafurther5.5hthemice
were; (i) OA-NO and PPAR-γ separated from each other wereeuthanized,thelowerrightlobeexcisedforassessment
2
by a distance greater than the non-bonded cutoff distance of edema by wet:dry weight ratio, BAL fluid collected,
(>8.5A˚); (ii) OA-NO complexed with the PPAR-γ ligand- and the lungs excised for histopathological examination
2
binding domain. Each molecular system was contained in and measurement of inflammation-association molecular
a box size of 25.0×25.0×37.0A˚ with periodic boundary markers.Plasmawasobtainedatthesametime.
conditions. The step size was 2 femtoseconds (fsec). To A prominent aspect of pulmonary inflammation is
start the simulations, different seed numbers were used infiltration of neutrophils (polymorphonuclear leukocytes;
for initial Maxwellian velocity distribution for each system. PMNs)intothelungsandthusintoBALfluid.Weobserved
Simulationswerecontinuedandthecoordinatesweresaved anLPS-inducedincreaseinthetotalnumberofcellsinBAL
foranalysisevery2fsec. fluid (Figure2(a)), and differential staining indicated that
neutrophilsaccountedformostoftheincrease(Figure2(b)).
2.15.StatisticalAnalysis. Dataarepresentedasmean±SD. Both increases were significantly attenuated by OA-NO
2
Differences between groups were analyzed using ANOVA, treatment. Similar results were seen for measurements of
followed by a Bonferroni multiple comparison test using myeloperoxidase in both lung (Figure2(c)) and BAL fluid
GraphPad Prism 5.03 (GraphPad Software, La Jolla, CA). (Figure2(d)). MPO is found in the pulmonary system in
P<0.05wasconsideredsignificant. association with PMNs and is thus a marker for their
presence.Theseresultswereconfirmedbydirectmicroscopic
3.Results examinationofBALfluid(Figure2(e)).
Inflammation is also characterized by oxidant stress,
3.1. In Silico Binding of OA-NO to PPAR-γ. Although the which directly injures lung tissues. Oxidant stress reflects
2
crystal structure of LNO bound to the PPAR-γ ligand production of reactive oxygen species such as H O and
2 2 2
bindingsitehasbeenreported[19],nosimilarinformation superoxide,predominantlybymacrophagesandneutrophils,
is available for OA-NO . We accordingly used in silico and of the reactive nitrogen species NO by a variety of
2
methods to determine the likely binding mode for this celltypes.Asexpected,measurementofH O demonstrated
2 2
PPARResearch 5
GLN 286
OA-NO2 TYR 473
HIS 449
2.76 A˚ 3.0 A˚
3.0 A˚
2.05 A˚ ARG 288
TYR 473
GLN 286
ARG 288 HIS 449
OA-NO2
(a) (b)
250 6
∗∗∗
∗∗∗
y y
vit 200 vit ∗∗∗ ∗∗∗
cti ∗∗∗ ∗∗∗ cti
e a e a 4 ∗∗∗
as 150 as
er ∗∗∗ ∗∗∗ er
cif cif ∗∗∗
u u
ve l 100 ∗∗∗ ve l 2 ∗∗∗
ati ati
Rel 50 Rel
0 0
DMSO 0.1 μM 1 μM 5 μM DMSO 0.1 μM 1 μM 5 μM
OA-NO2 OA-NO2
Rosiglitazone Rosiglitazone
(c) (d)
Figure1:OA-NO activatesPPAR-γinairwayepithelialcells.BindingofOA-NO tothePPAR-γligand-bindingdomainwasmodeledin
2 2
silico.Resultsarepresentedas(a)space-fillingrepresentation;(b)schematicrepresentationshowinginteractionsofPPAR-γresidueswith
specificgroupsintheOA-NO ligand.Hydrogenbondsareindicatedbydottedlines.A549cellsweretransientlytransfectedwithpRLSV40
2
andwithoneofthefollowingconstructs:(c)PPAR-dependentluciferasereporter,pFATPluc;(d)PPAR-γGAL4reportersystem(UAS ×4
G
TK-luciferase+GAL4-PPAR-γ).Cellswerethenincubatedwithvehicle(DMSO),OA-NO (0.1–5μM),orrosiglitazone(0.1–5μM).After
2
24h,luciferaseactivitywasmeasuredwithadualluciferaseactivitykitandnormalizedtothatofRenillaluciferase.Dataarerepresentative
ofoneoftwoindependentexperiments;n=6;∗∗∗P<0.001versusvehicle.
alarge increase following LPS administration that was 3.4. Pulmonary Administration of OA-NO Diminishes Cap-
2
reducedbyOA-NO treatment(Figure2(f)).Similarresults illary Permeability and Severity of LPS-Induced Lung Injury.
2
were seen for nitrate, the end product of NO metabolism Increased capillary permeability results in lung edema, a
(Figure2(g)). Levels of the lipid oxidation product MDA driving force for the hypoxemia that is observed in ALI.
thatprovidesanindexofoveralloxidativestress,andMDA It also allows escape of plasma proteins into the alveolar
levelsfollowedthesamepatternastheothertwomarkersof space,whichcanthenbedetectedinBALfluid.Wefindthat
oxidantgeneration(Figure2(h)). OA-NO treatment attenuates the LPS-induced increase in
2
Inflammation in ALI is likewise associated with release BALfluidproteinconcentration(Figure3(a)).Theincreased
of proinflammatory cytokines and chemokines by neu- vascularpermeabilityfollowingLPSadministrationallowed
trophils, macrophages, and other cells. We measured the EBD to extravasate from the vasculature into the lung
proinflammatory cytokines TNF-α (Figure2(i)) and IL-6 parenchyma, turning these lungs deep blue (∼0.3μg dye
(Figure2(j)),thechemokineKC,whichactivatesandattracts per g lung). This extravasation of EBD was significantly
neutrophils (Figure2(k)), and the related chemokine MIP- reduced by OA-NO2 treatment, however, resulting in only
2, which has effects similar to KC (Figure2(l)). Plasma a pale blue appearance of the lung (∼0.17μg dye per g
levels of all four markers were greatly increased by LPS lung;Figure3(c)).Thelungedemathatresultsfromcapillary
administration,butthisincreasewassignificantlyreducedby permeabilityisreflectedinthewet:dryweightratio,andLPS-
OA-NO treatment. induced increases in this parameter were also reduced by
2
6 PPARResearch
50 30
∗∗∗
∗∗∗
L)
m40
410/ mL) 20
×otal cells (2300 4×MNs (10/ 10
LF t10 n.s. P
A
B
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(a) (b)
∗∗∗
0.15 0.15
mL) mL) ∗∗∗
PO min/ 0.1 PO min/ 0.1
BALF Mvity (nM/ 0.05 n.s. BALF Mvity (nM/ 0.05
cti cti n.s.
a a
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(c) (d)
LPS + Veh. LPS+OA-NO2
(e)
50000 5
∗∗∗
∗∗∗
U) 40000 4
productionescence, AF 30000 μate (M) 3 n.s.
HO22fluor 20000 Nitr 2
Δ( 10000 n.s. 1
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(f) (g)
Figure2:Continued.
PPARResearch 7
15 15
∗∗∗
∗∗∗
ein) mL)
prot 10 ng/ 10
mg α(
M/ NF-
n n.s. T
A ( 5 ma 5 n.s.
MD as
Pl
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(h) (i)
20 1.5
∗∗∗
∗∗∗
L) 15 L)
m m 1
g/ g/
n n
L-6 ( 10 KC (
ma I n.s. ma 0.5 n.s.
Plas 5 Plas
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(j) (k)
1.5
∗∗∗
L)
m
g/ 1
n
2 (
P-
MI
a 0.5
m
as n.s.
Pl
0
Veh. OA-NO2 Veh. OA-NO2
PBS LPS
(l)
Figure2:PulmonarydeliveryofOA-NO reducesLPS-inducedlunginflammation.InductionofALIbyi.t.injectionofLPS(50μg)was
2
followed30minlaterbyi.t.administration(50μL)ofOA-NO (50μg)orvehicle(10%DMSO).Afterafurther5.5h,BALfluid,plasma,
2
andlungsampleswereobtained.(a)Totalcelland(b)neutrophilnumberinBALfluid.Myeloperoxidaseactivityin(c)lungtissueand
(d) BAL fluid. (e) Microscopic examination following staining of BAL fluid. (f) H O production, (g) nitrate concentration, and (h)
2 2
malonaldehyde/protein ratio in lung. Plasma levels of (i) TNF-α, (j) IL-6, (k) KC, and (l) MIP-2. Data are representative of one of two
independentexperimentswithn=6–8micepergroup;∗∗∗P<0.001.
OA-NO treatment (Figure3(b)). Direct histopathological NFAs exert many of their anti-inflammatory effects by
2
examination of the lung following H&E staining further activating PPAR-γ, which is known to decrease activity of
confirmedsignificantlunginflammationandinjuryinmice the proinflammatory transcription factor NF-κB. PPAR-γ
treated with vehicle but much less severe abnormalities in promotestranscriptionoftheantioxidantfactorNrf2,which
thosethathadreceivedOA-NO2(Figure3(d)). inturnupregulatesPPAR-γexpressioninapositivefeedback
loop [20, 21]. To test our hypothesis, we examined the
3.5. OA-NO Produces Anti-Inflammatory Alterations in effects of OA-NO treatment on activity of PPAR-γ, NF-
2 2
Activity of PPAR-γ, NF-κB, and Nrf2. We propose that κB, and Nrf2. When followed by vehicle treatment, i.t. LPS
8 PPARResearch
600 6
∗∗∗
∗∗∗
L)
m o
μn (g/ 400 dry rati 4 n.s.
F protei 200 ng wet/ 2
AL n.s. Lu
B
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(a) (b)
0.4
LPS + Veh. LPS+OA-NO2 g) ∗∗∗
n
g lu 0.3
e/
y
d
μg 0.2
(
e
u
bl
ns 0.1
a
v
E
0
Veh. OA-NO2
LPS
(c)
LPS + Veh. LPS+OA-NO2
(d)
Figure3:PulmonarydeliveryofOA-NO reducesLPS-inducedlunginjury.InductionofALIbyi.t.injectionofLPS(50μg)wasfollowed
2
30minlaterbyi.t.administration(50μL)ofOA-NO (50μg)orvehicle(10%DMSO).Afterafurther5.5h,BALfluidandlungsampleswere
2
obtained.(a)ProteinconcentrationinBALfluid.(b)Ratiooflungwet:dryweight.(c)ExtravasationofEvansBluedyeintothelungfollowing
intravenousinjectionwasphotographedandquantitatedbyspectrophotometry.(d)ThelungwasexaminedhistologicallyfollowingH&E
staining.Dataarerepresentativeofoneoftwoindependentexperimentswithn=6–8micepergroup;∗∗∗P<0.001.
upregulatedtheDNA-bindingactivityofNF-κB,asexpected inhibitionofNF-κBactivityandupregulationofNrf2,these
(Figure4(c))butdecreasedactivityofPPAR-γ (Figure4(a)) data support the concept that NFAs act in part via PPAR-γ
and Nrf2 (Figure4(b)). OA-NO treatment, however, not activation.
2
only diminished the increase in NF-κB activity but also
increased the DNA-binding activity of PPAR-γ and Nrf2. 3.6. OA-NO Decreases Inflammatory Response of Alveolar
2
Under noninflammatory conditions, OA-NO treatment Macrophages. Alveolar macrophages play a central role in
2
increased PPAR-γ activity but had no significant effect on regulation of the lung’s immune system. When activated
basal levels of NF-κB or Nrf2 activity. Given the role of by stimuli such as LPS, they generate oxidants and secrete
NFAsasPPAR-γagonistsandtheknownanti-inflammatory molecules that attract neutrophils and other immune cells
andantioxidantactivitiesofPPAR-γ,manyexertedthrough to the lung. As PPAR-γ is known to play a major role
PPARResearch 9
0.8 0.4
∗∗∗
γ)activity (AR-450 00..46 ∗∗∗ 2 activity (A)450 00..23 n.s. ∗∗∗
PA 0.2 Nrf 0.1
P
0 0
Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2 Veh. OA-NO2
PBS LPS PBS LPS
(a) (b)
2.5
∗∗∗
) 2
450
A
y ( 1.5
vit
acti 1 n.s.
B
κ
F-
N 0.5
0
Veh. OA-NO2 Veh. OA-NO2
PBS LPS
(c)
Figure 4: OA-NO reverses LPS-induced changes in transcription factor activity. Induction of ALI by i.t. injection of LPS (50μg) was
2
followed30minlaterbyi.t.administration(50μL)ofOA-NO (50μg)orvehicle(10%DMSO).Afteranother5.5hlungswereexcisedand
2
DNA-bindingactivityofthetranscriptionfactors(a)PPAR-γ,(b)Nrf2,and(c)NF-κBwasdetermined.Dataarerepresentativeofoneof
twoindependentexperimentswithn=6–8micepergroup;∗∗∗P<0.001.
in modulating activation of alveolar macrophages [22], we 4.Discussion
investigated the ability of NFAs to suppress LPS-induced
activationofthesecells.Specifically,miceweretreatedwith Our results establish that delivery of OA-NO2 directly to
LPS and either vehicle or OA-NO as previously described. the lung significantly reduces the severity of pulmonary
2
Alveolar macrophages were then isolated and their RNA inflammation and injury. We also find that treatment with
OA-NO suppresses the activated phenotype of alveolar
extracted for determination of relevant gene expression. 2
macrophages, key cells in the regulation of pulmonary
Results (Figure5) demonstrate significant suppression of
inflammation.Inthelung,LPS-inducedincreaseinactivity
the proinflammatory cytokines TNF-α and interleukin-12
oftheproinflammatorytranscriptionfactorNF-κBislargely
(IL-12) and the chemokine MCP-1. Downregulation was
blocked, as are LPS-induced decreases of PPAR-γ and
also observed for the inducible form of nitric oxide syn-
the antioxidant transcription factor Nrf2. Indeed, PPAR-γ
thase(iNOS;NOS2)andthesuperoxide-generatingenzyme
activity is increased over basal levels in both inflammatory
NADPH oxidase 4 (NOX4), both of which contribute to
and noninflammatory conditions. Upregulation of Nrf2
oxidative stress. OA-NO treatment also reduced alveolar
2 expression and activity may account for much of the
macrophageexpressionofcyclooxygenase2(COX-2),which
antioxidantactivityofPPAR-γandconsequentreductionin
synthesizes proinflammatory prostaglandins. Conversely,
inflammation-associatedoxidativeinjury[20,21].
OA-NO2 upregulatedexpressionofPPAR-γ,andthePPAR- This appears to be the first study of NFAs’ ability to
γ target genes fatty acid binding protein 4 (FABP4) and reducetheseverityofALIandoneofveryfewtoinvestigate
CD36,areceptorthatfacilitatesmacrophagephagocytosisof the protective anti-inflammatory actions of NFAs in an
apoptoticandsenescentneutrophilsandtherebycontributes animalmodelofanydisease.Borniquelandcolleagueshave
to the resolution of inflammation. All these measurements recentlyobtainedresultsinamurinemodelofinflammatory
confirm the ability of NFA treatment in vivo to sup- bowel disease that are similar to ours in ALI [23], finding
press the activated, proinflammatory phenotype of alveolar that OA-NO decreased disease severity and the increase in
2
macrophages. NF-κB expression while increasing expression of PPAR-γ.
10 PPARResearch
8 NOS2 15 NOX4 15 COX-2
∗∗ ∗ ∗∗
A 6 A A
N N 10 N 10
R R R
m m m
e 4 e e
v v v
ati ati 5 ati 5
Rel 2 Rel Rel
0 0 0
LPS + veh. LPS+OA-NO2 LPS + veh. LPS+OA-NO2 LPS + veh. LPS+OA-NO2
(a) (b) (c)
IL-12p40
TNF-α MCP-1 ∗∗
6 15 30
∗∗∗ ∗∗∗
A A A
RN 4 RN 10 RN 20
m m m
ve ve ve
ati 2 ati 5 ati 10
Rel Rel Rel
0 0 0
LPS + veh. LPS+OA-NO2 LPS + veh. LPS+OA-NO2 LPS + veh. LPS+OA-NO2
(d) (e) (f)
PPAR-γ FABP4 CD36
5 40 ∗∗∗ 15 ∗∗∗
∗∗
A 4 A 30 A
N N N 10
mR 3 mR mR
ative 2 ative 20 ative 5
Rel 1 Rel 10 Rel
0 0 0
LPS + veh. LPS+OA-NO2 LPS + veh. LPS+OA-NO2 LPS + veh. LPS+OA-NO2
(g) (h) (i)
Figure 5: OA-NO reduces inflammatory phenotype in LPS-activated alveolar macrophages. Induction of ALI by i.t. injection of LPS
2
(50μg) was followed 30min later by i.t. administration (50μL) of OA-NO (50μg) or vehicle (10% DMSO). After another 5.5h BAL
2
fluidwasobtained.AlveolarmacrophageswereisolatedfromtheBALfluidandplatedinDMEM+10%FBS.After1h,RNAwasisolated
andexpressionoftheindicatedgeneswasdeterminedusingreal-timePCR;resultswerenormalizedtovaluesforthehousekeepinggenes
glyceraldehyde-3-phosphatedehydrogenase(GAPDH)and9srRNA.Dataarerepresentativeofoneoftwoindependentexperimentswithn
=6–8micepergroup;∗P<0.05;∗∗P<0.01;∗∗∗P<0.001.
Significantly, the increase in PPAR-γ expression was abol- observe are mediated largely by activation of PPAR-γ. Such
ished by simultaneous administration of a PPAR-γ antago- activation appears adequate to account for most of our
nist.OA-NO wasalsofoundtoreduceinfarctsizeincardiac observations,includingincreasedactivityandexpressionof
2
ischemia-reperfusioninjury[24].Thiswasaccompaniedby both PPAR-γ itself and the antioxidant transcription factor
areductioninNF-κBactivitybutapotentialroleforPPAR-γ Nrf2,whichbothupregulatesandisupregulatedbyPPAR-γ
wasnotinvestigated.BeneficialeffectsofOA-NO ,together [20,21],aswellasdecreasedactivityoftheproinflammatory
2
with reduced inflammation, have also been seen in renal transcriptionfactorNF-κB.
ischemia/reperfusion injury [25]. It appears likely that a Initial studies suggested that OA-NO and LNO were
2 2
numberofotherinflammatoryconditionsmightalsobenefit presentinnormalhumanplasmaatconcentrationsexceed-
fromNFAtreatment. ing0.1μM,whichinconjunctionwiththeirmeasuredbind-
Activation of the nuclear hormone receptor PPAR-γ ingaffinitywouldraisethepossibilitythattheywereendoge-
has been shown to reduce the severity of LPS-induced nous PPAR-γ agonists [11]. Later studies from another
ALI [5, 9], and NFAs are known to activate PPAR-γ [11, group, however, have suggested concentrations below 1nM
12]. That this activation reflects binding to the ligand [26]. Notably, both studies were performed with plasma
binding domain is demonstrated both by ability of NFAs from healthy subjects and do not reflect the increase in
to displace the synthetic ligand rosiglitazone [12] and by NFAconcentrationsanticipatedininflammatoryconditions.
the elucidated crystal structure of LNO bound to this site This controversy nevertheless has only limited relevance
2
[19]. We therefore suggest that the protective effects we to our studies, which address pharmacological effects of