Opposite Effects of Interferon Regulatory Factor 1 andOsteopontin on

OppositeEffectsofInterferonRegulatoryFactor1andOsteopontinontheApoptosisofEpithelialCellsInducedbyTNF-ainIn?ammatoryBowelDisease

RuihanTang,MD,PhD,*GuangYang,MD,PhD,*,?ShenghongZhang,MD,PhD,*ChangyouWu,MD,PhD,?andMinhuChen,MD,PhD*

Background:In?ammatoryboweldisease(IBD)ischaracterizedbyadamagedintestinalepitheliumbarrier.Interferonregulatoryfactor1(IRF1)and

osteopontin(OPN)regulatecellsurvivalandgrowthinavarietyofcircumstancesbuttheireffectsontheintestinalepitheliumhavenotbeenelucidated.Inthisstudy,wesoughttodeterminetheeffectsofOPNonintestinalepithelialcellsunderconditionsoftumornecrosisfactor(TNF)-a–inducedin?ammationandwhetherIRF1regulatesOPNexpression,theactivationofdownstreampathways,andin?ammatoryresponses.

Methods:TheexpressionlevelsofOPNandIRF1wereassessedbyimmunohistochemicalanalysesofhumanIBDandexperimentalmousecolitis.The

effectsofIRF1andOPNonin?ammatoryresponseswereinvestigatedinvitroinNCM460andCaco-2cellsstimulatedbyTNF-a.Changesinp-AKT,p-P38,andp-ERKlevelswerequanti?edbywesternblottingassays.TheregulationofOPNexpressionbyIRF1wasdeterminedbyluciferaseactivityandchromatinimmunoprecipitationassays.

Results:IRF1wasupregulatedinhumanIBDandinthecolonepitheliumofmicewithdextransulfatesodium–inducedcolitis.Additionally,IRF1was

correlatedwithhigh-sensitivityC-reactiveprotein,erythrocytesedimentationrate,Crohn’sdiseaseactivityindex,Crohn’sdiseaseendoscopicindexofseverity,andsimpleendoscopicscoreforCrohn’sdiseaseinCrohn’sdiseaseandwithhigh-sensitivityC-reactiveprotein,erythrocytesedimentationrate,Mayoscore,Baronscore,modi?edBaronscore,Rachmilewitzscore,ulcerativecolitisendoscopicindexofseverity,ulcerativecolitiscolonoscopicindexofseverity,anddiseasedurationinulcerativecolitis.TheexpressionofOPNwassigni?cantlydecreasedinpatientswithIBDcomparedwithcontrolsandindextransulfatesodium–inducedexperimentalcolitisandwasalsoinverselycorrelatedwithclinicalandendoscopicactivitiesinbothCrohn’sdiseaseandulcerativecolitis.TNF-atreatmentupregulatedIRF1anddiminishedOPNinbothNCM460andCaco-2cells.TheoverexpressionofOPNandrhOPNamelioratedtheapoptosisinducedbyTNF-a,whereastheoverexpressionofIRF1aggravatedapoptosis,indicatingoppositeeffectsofOPNandIRF1inin?amedepithelialcells.TheluciferaseandchromatinimmunoprecipitationassaysshowedthatIRF1transcriptionallymodulatedtheexpressionofOPN.TNF-ainhibitedtheOPN-inducedupregulationofp-ERK,p-P38,andp-AKT.

Conclusions:Ourdatasuggestthatduringintestinalin?ammation,theTNF-a–mediatedactivationofIRF1isrelatedtothesubsequentsuppressionof

OPNexpression,furtherreducingp-AKT,p-P38,andp-ERKactivitiesandresultinginaggravationoftheinjurytointestinalepithelialcells.(In?ammBowelDis2014;20:1950–1961)

KeyWords:TNF-a,in?ammatoryboweldiseases,osteopontin,IRF1

ReceivedforpublicationJune3,2014;AcceptedJuly13,2014.

Fromthe*DepartmentofGastroenterology,theFirstAf?liatedHospital,SunYat-senUniversity,Guangzhou,China;?DepartmentofMedicalImagingandInter-ventionalRadiology,SunYat-senUniversityCancerCenter,StateKeyLaboratoryofOncologyinSouthChina,Guangzhou,China;and?DepartmentofImmunology,ZhongshanSchoolofMedicine,SunYat-SenUniversity,Guangzhou,China.

M.ChenwassupportedbygrantsfromtheNationalNaturalScienceFoundationofChina(grantno.81270473),theKeyProjectoftheNational12th?ve-yearResearchProgramofChina(grantno.2012BAI06B03),theScienti?candTechnologicalPlanningofGuangzhou(grantno.2011YZ-00004),andtheFundamentalResearchFundsforSunYat-senUniversity(grantno.13YKJC01).

Theauthorshavenocon?ictsofinteresttodisclose.

Reprints:MinhuChen,MD,PhD,DepartmentofGastroenterology,theFirstAf?liatedHospitalofSunYat-SenUniversity,58ZhongshanIIRoad,Guangzhou510080,China(e-mail:chenminhu@http://wendang.chazidian.com).

Copyright©2014Crohn’s&ColitisFoundationofAmerica,Inc.DOI10.1097/MIB.0000000000000192Publishedonline9September2014.

n?ammatoryboweldisease(IBD),whichincludesCrohn’sdis-ease(CD)andulcerativecolitis(UC),ischaracterizedbychronicin?ammationandepithelialbarrierdamage.1,2Accumulatingevi-dencehasshownthattumornecrosisfactor(TNF)-aactsasakeystimulatorinbothcellsurvivalandproapoptoticpathways,exertingitsimpactbyregulatingproin?ammatorycytokinestoinducetissuedegradationorprotection.3–5ElevatedlevelsofTNF-aintheintes-tinaltissueandserumhavebeenwellestablishedtobeassociatedwiththeseverityandextentofbothCDandUC.6,7Additionally,clinicaltrialsonanti-TNFtherapieshaveshowngreatef?cacyininducingIBDmucosalhealing.8,9However,thedetailedpathwaysandmechanismsrequirefurtherelucidation.

IRF1,aninterferonregulatoryfactor,wasoriginallyidenti-?edasakeyregulatorofinterferon-stimulatedgenes.IRF1servesasatranscriptionfactorinmanybiologicalprocesses,suchasin?ammationandtumorproliferation.10,11SeveralstudiesinmurinecolitismodelshavedemonstratedthatTNF-asigni?cantlyinduces

In?ammBowelDis??Volume20,Number11,November2014

I

1950

|http://wendang.chazidian.com

In?ammBowelDis??Volume20,Number11,November2014OppositeEffectsofIRF1andOsteopontin

theexpressionofIRF1.12,13OnlylimitedeffortshavebeendirectedtowardtheinvestigationofIRF1aboutIBD.OnereportsuggestedthatconditionalknockoutoftheIRF1geneprotectedmicefromchronicin?ammation,14whereasanotherreportshowedthatIRF1exhibitedfewanti-in?ammatoryeffectsindextransulfatesodium(DSS)–inducedcolitis.15TheexpressionofIRF1inpatientswithIBDisstillunclear.ArecentstudyconcludedthattherewasnoobviouschangeinIRF1expressioninpatientscomparedwithcontrols.16However,anotherstudyobservedincreasedexpressionofIRF1in72%ofpatientswithCDamongstudycohorts.17Still,littleisknownabouthowIRF1interactswithmultiplecytokinesandactivatesdifferentsignal-ingpathwaysinthecontextofIBD.

Osteopontin(OPN)isasecretedarginine-glycine-asparticacid–containingphosphoproteinthatisubiquitouslyexpressedinavarietyoftissues.18BybindingtointegrinsandCD44variants,itisactiveinmanybiologicalprocesses,includingwoundheal-ing,tumorigenesis,in?ammation,andimmuneresponses.19,20OPNispivotalinmaintainingcellsurvivalbyinhibitingapopto-sis.ThedownregulationofOPNsensitizescellstoapoptosis.21OPN,asolublecytokine,wasalsofoundtobeinvolvedinvar-iousin?ammatorydiseasesandautoimmunedisorders.22–25OPNprimarilyservesasafactorthatfosterssurvivalindifferentsit-uations.However,whetherOPNhaspro-oranti-in?ammatoryeffectsinIBDremainscontroversial.26–28TheregulationofOPNexpressioninthein?amedmucosaofpatientswithIBDhasyettobeinvestigated.

ArecentreportfoundthatIRF1siRNAresultedindecreasedexpressionofOPNincardiac?broblasts.29However,thein?uenceofIRF1ontheexpressionofOPNinothercellsortissueshasnotbeenreportedtodate.WespeculatedthatincreasedTNF-ainthein?amedintestinalmucosamightsup-pressOPNgeneexpressionatthetranscriptionallevelbystim-ulatingIRF1regulation.Inthisstudy,weinvestigatedtheassociationbetweenIRF1/OPNlevelsinepithelialcellsandclinicalindicesofdiseaseactivityinpatientswithIBD.Addi-tionally,werevealedsubtlecrosstalkamongTNF,IRF1,andOPNsignalinganddetectedcytoprotectiveeffectsofOPNagainstTNF-aongrowthandapoptosisinintestinalepithelialcells(IECs),furthercon?rmingthein?uenceofIRF1onOPNexpressionatthetranscriptionallevel.

ParticipantsandDiseaseActivity

BetweenJanuary2010andApril2014,patientsdiagnosedonthebasisofthestandardclinical,endoscopic,andhistologicalcriteriaofIBDattheGastroenterologyDepartmentsoftheFirstAf?liatedHospitalofSun-YatSenUniversitywereincluded.Thestudywasapprovedbythehospitalethicscommittees,andwritteninformedconsentswereobtained.Thebiopsieswereobtainedduringendoscopyorsurgery.Bloodsamplesforthemeasurementofhigh-sensitivityC-reactiveproteinanderythrocytesedimentationrateweretakenwithin1weekbeforeorafterendoscopy.Patientswhohadinfectiouscolitisandcolorectalcancerwereexcluded.Colonbiopsieswereobtainedfromhealthyvolunteersduringendoscopyorincisedfromunaffectedareassurgicallyremovedfrompatientswithcoloncancer.TheclinicaldiseaseactivitywasassessedbythemeasurementoftheCrohn’sdiseaseactivityindexforCDandoftheMayoscoreforUCbyaphysicianwhowasblindedtothecolonoscopicappearanceofthepatient.30,31Endos-copieswereperformedandgradedaccordingtotheCrohn’sdiseaseendoscopicindexofseverityandsimpleendoscopicscoreforCrohn’sdiseasescoringsystemsinpatientswithCDandaccordingtotheBaron,modi?edBaron,Rachmilewitz,ulcerativecolitisendoscopicindexofseverity,andulcerativecolitiscolonoscopicindexofseverityscoresforUC.32–38Toavoidbias,allgastroenter-ologistsperformingtheendoscopieswereunawareoftheresultsoftheCrohn’sdiseaseactivityindex,Mayoscore,high-sensitivityC-reactiveprotein,orerythrocytesedimentationrate.

MiceandExperimentalColitis

MaleC57BL/6mice(n¼24,age6–8wk,weight18–20g)wereobtainedfromGuangdongMedicalLaboratoryAnimalCenter(Guangzhou,Guangdong,China).Themicewererandomlyassignedtotheexperimentalgroup(n¼12)orthecontrolgroup(n¼12).Themiceintheexperimentalgroupweresubjectedtoacutecolitisinducedby2%(wt/vol)DSS(MW36,000–50,000;MPBiochemicals,Solon,OH)intheirdrinkingwaterfor7days.39Thecontrolmiceweregivendistilledwater.Onday8,themicewereeuthanized,andtheircolonswerecollectedforanalysis.40Thecolonlengthwasmeasuredbeforehistologicalevaluation.Thecolonepitheliumwasisolatedbyscrapingthemucosalsurfacewithaglassslide.RNAandproteinwereex-tractedforreverse-transcriptasepolymerasechainreaction(PCR)andwesternblotanalysis.TheremainingcolontissueswereSwiss-rolled,formalin-?xed,andparaf?n-embeddedforhistologicalandimmuno-histochemicalanalyses.Allanimalusewasapprovedandmonitoredbythehospital’sethicscommittees.

MATERIALSANDMETHODS

Materials

RecombinanthumanOPNwaspurchasedfromPeproTech(RockyHill,NJ),preparedinwaterandstoredinaliquotsofappropriatevolumesat2808Cuntiluse.Theantibodiesusedwereanti-OPN(AssayBiotechnology,Sunnyvale,CA)andanti-IRF1(SantaCruz,Dallas,TX),aswellasanti-p-AKT,anti-AKT,anti-p-EKR,anti-P38andanti-actin(Beyotime,Nantong,China).ThesiRNAsagainstOPNandIRF1weresynthesized(Genepharma,Suzhou,China).

CellCultureandTransfection

TheCaco-2celllinewasagiftfromDr.JianjunTangoftheStateKeyLaboratoryofOncologyinSouthernChina,SunYat-SenUniversity,Guangzhou,China.TheCaco-2cellswereculturedinDulbecco’sModi?edEagle’sMediumsupplementedwithheat-inactivatedfetalcalfserum(10%)andpenicillin(100U/mL)orstreptomycin(100mg/mL)in5%CO2at378C.NCM460cellswereobtainedfromJennioBiotechnology(Guangzhou,China)andcul-turedinM3:10TMmedium.Thiscelllineisconsideredtobe

http://wendang.chazidian.com|

1951

TangetalIn?ammBowelDis??Volume20,Number11,November2014

anoncancerouscolonicepithelialcellline.Thetransfectionofplas-midsandsiRNAwasperformedusingLipofectamine3000(Invitro-gen,Carlsbad,CA)accordingtothemanufacturer’sinstructions.

MTTAssay

Thecelllineswereseededin96-wellplatesatadensityof5000cellsperwellinavolumeof150mLofcellculturemediumperwellfor24hoursbeforedrugexposure.OPNandTNF-awereaddedintothewellsatdifferentconcentrationsintriplicate.Theplateswereincubatedat378Cwith5%CO2for48hours.Oncethetransfectionwasperformed,TNF-awasaddedtothewells24hoursaftertransfection.A20-mLsampleofMTTsolution(5g/L,dis-solvedinphosphate-bufferedsaline[PBS])wasaddedtoeachwell,andtheplateswereincubatedat378Cforanadditional4hours.Thesupernatantwasdiscarded,and150mLofdimethylsulfoxidewasaddedtodissolvetheformazanproduct.Theabsorbancevaluesat490nm(A490)weredeterminedusingamultiwellplatereader(Tecan,Männedorf,Switzerland).

withanti-IRF1antibodiesat48Cfor15hours,thecellcultureswerewashed3timeswithPBS.Thecellswerethenincubatedwiththesecondaryantibodyfor1houratroomtemperatureandobservedthroughaninverted?uorescencemicroscopeafterbeingwashed3timeswithPBS.

RNAExtractionand

Reverse-transcriptasePCR

TotalRNAwasextractedfromtheculturedcellsorcolonicbiopsiesfromthepatientswithIBDandcontrolsusingRNAisoPlus(Takara,Otsu,Japan).TotalRNAwasreversetranscribedusingtheReverseTranscriptionSystem(Promega,Madison,WI)accordingtothemanufacturer’sinstructions.PCRwasperformedusingPCRMix(TransGenBiotech,Beijing,China).

WesternBlotAnalysis

Caco-2andNCM460cellswereexposedtovariousexperi-mentalconditionsfortheindicatedtimesbeforebeingharvestedandlysedforproteinextractionusingsodiumdodecylsulfate(Sigma-Aldrich,St.Louis,MO).TheproteinconcentrationsweredeterminedusingtheBio-RadProteinAssayKit(Bio-Rad,Hercules,CA).Theblotswerevisualizedusingtheenhancedchemiluminescencedetec-tionsystem(Amersham,Piscataway,NY),andb-actinwasusedasaloadingcontrol.

FlowCytometryAnalysis

Thesampleswereanalyzedusinga?owcytometer(Beck-manCoulter,Urbana,IL).Todetectapoptoticcells,labelingtestsinvolvingbothpropidiumiodideandannexin-VwereperformedusinganAnnexin-Vstainingkit(Invitrogen)accordingtothemanufacturer’sinstructions.

Immunohistochemistry

AnimmunohistochemicalanalysiswasperformedtoevaluatetheexpressionlevelsofOPNandIRF1inparaf?n-embeddedsections.Thegeneralprocesseswereasfollows:Paraf?nsectionswerebakedfordeparaf?nizationandwashed.Afterassessingtheendogenousperoxidaseactivity,thesectionswereincubatedwith3%H2O2for15minutesatroomtemperatureandwashedwithdistilledwaterandPBS.Then,thesectionsweretreatedwithethylenediami-netetraaceticacidinamicrowaveovenfor10minutesforantigenretrieval.Aftersubsequentblockingwith1%bovineserumalbuminfor10minutes,thesectionswereincubatedwithrabbitanti-humanOPNantibodies(1:100)overnightat48Candthenincubatedwithgoatanti-rabbitEnVisionworkingsolutionfor30minutesatroomtemperature.Horseradishperoxidaseenzymeactivitywasdetectedbyaddingthechromogenicsubstrate3,30-diaminobenzidinetetra-chloride(DAB;Dako,Copenhagen,Denmark)for30seconds,whichresultedinbrownstaining.Allthestepswerefollowedby3washeswithPBSfor5minuteseach.NegativecontrolsweretreatedwithPBSunderthesameconditions,ratherthanwithprimaryanti-bodies.ImageProPlus6.0(MediaCybernetics,Rockville,MD)wasusedincalculatingtheopticaldensityofIRF1andOPN.

LuciferaseReporterAssay

NCM460cellswereseededat5·103perwellin96-wellplates24hoursbeforetransfection.A683-bp(2365to+318)fragmentofthepromoterregionofOPNwasinsertedintothepGL3-Basicluciferasereportervector(Promega).Thecellswerecotransfectedwith0.1mgofthe?re?yluciferasereporterconstruct,0.01mgofthepRL-TKRenillaluciferasereporterplasmid(Prom-ega),andthepcDNA3.1-IRF1vectorusingLipofectamine3000(Invitrogen).Theluciferaseactivitywasexaminedusingadual-luciferasereporterassaysystem(Promega)accordingtothemanu-facturer’sinstructions,andthesignalwasnormalizedtotheinternalRenillacontroltodeterminetheef?ciencyofthetransfection.

ChromatinImmunoprecipitation

NCM460cellswereusedforchromatinimmunoprecipita-tionwiththeEZChIPkit(Millipore,Billerica,MA)accordingtothemanufacturer’sinstructions.Afterelutionandpuri?cation,therecoveredimmunoprecipitatedDNAsampleswereusedforPCR,withtheprimers50-GTGGCCTACTTTACATACC-30(forward)and50-GGACTATATGTGTGTTAAC-30(reverse)toamplifya185-bpsegmentoftheOPNpromotercontainingthepotentialIRF1bindingsites(2284to2272).ThePCRproductswereanalyzedbyagarosegelelectrophoresis.

Immuno?uorescence

AfterbeingtreatedwithTNF-afor6hours,thecellswerewashedwithPBSand?xedfor20minutesatroomtemperaturewith4%paraformaldehyde.TheywerewashedagainwithPBSandper-meabilizedwithTritonX-100for10minutesatroomtemperature.Then,thecellswerewashedwithPBSandincubatedwith3%bovineserumalbumintoblockthenonspeci?cbindingsites.Afterincubation

StatisticalAnalysis

Theresultsfornumericaldataarepresentedasmean6SD.TheMann–WhitneyUtestwasusedtoexploretheassociationsbetweenthenonparametricnumericaldatain2independentgroups,andthettestwasusedforparametricnumericaldata.Theassoci-ationbetween2variableswasassessedbySpearman’srank

1952

|http://wendang.chazidian.com

In?ammBowelDis??Volume20,Number11,November2014OppositeEffectsofIRF1andOsteopontin

TABLE1.DemographicandClinicalCharacteristicsoftheStudyPopulation

CD

N

Male,n(%)Age,yr

Diseaseduration,moAgeatdiagnosis,yr

Diseasephenotype,n(%)B1B1pB2B2pB3B3p

Diseaselocation,n(%)L1L2L3

Diseaseextent,n(%)ProctosigmoiditisLeft-sidedcolitisPancolitis

IBD-relatedsurgeryhistory,n(%)Medication,n(%)5-ASA

CorticosteroidsAzathioprine6-MercaptopurineMethotrexateThalidomide

Anti-TNF-atherapyNomedications

211136

30.1611.1

37.026

169734687

UC

100

62(62.0)

43.2614.3(14–70)

52.5(12–132)—

—————————5219291182331710054(52.0)(19.0)(29.0)(11.0)(82.0)(33.0)(17.0)(1.0)(0.0)(0.0)(5.0)(4.0)

Controls110

61(55.5)

34.566.0(23–49)

——————————————————————

(64.5)(11–65)(1–252)(10–65)(80.1)(3.3)(16.1)(2.8)(3.8)(3.3)

52(24.6)82(38.9)77(36.5)———

37(17.5)50919214918165(23.7)(43.1)(43.6)(6.6)(4.3)(8.5)(7.6)(2.4)

Normallydistributeddataarepresentedasmean6SD,whereasnonparametricdataarepresentedasmedian,minimal,andmaximalvalues.

Locations:L1¼terminalileum;L2¼colon;L3¼ileocolon.PatientswithL4(isolatedupperdisease)wereexcluded.Behaviors:B1¼nonstricturing;B2¼stricturing,B3¼penetrating;P¼perianaldisease.ASA,aminosalicylicacid.

correlationcoef?cient(r)fornonparametriccorrelations.Two-sidedPvalues,0.05wereconsideredstatisticallysigni?cant.

RESULTS

ExpressionLevelsofIRF1andOPNWereInverselyCorrelatedinPatientswithActiveCDorUC

Atotalof211patientswithCD,100patientswithUC,and110controlswereenrolled(Table1).Immunohistochemicalanalyseswasperformedtoexaminetheexpressionandlocaliza-tionofIRF1andOPNinthebiopsiesfromthepatientswithIBD

(CD:n¼211,174endoscopicbiopsysamplesand37surgicalresectionsamples;UC:n¼100,91endoscopicbiopsysamplesand9surgicalresectionsamples),and70normalsurgicalresectionsand40intestinalbiopsieswereusedascontrols(n¼110).IncreasedexpressionofIRF1intheIECswasobservedinboththeCDandtheUCpatientscomparedwiththecontrols(Fig.1B).Thisobservationwascon?rmedbyquantitativeanalysis.Theaverageopticaldensity(AOD)ofIRF1wassigni?cantlyhigherintheCDandUCgroupsthanthatinthecontrolgroup(n¼211,AOD¼0.17160.063;n¼100,AOD¼0.15160.049;andn¼110,AOD¼0.10160.047,respectively;P,0.001)(Fig.1C).TheAODofIRF1wascorrelatedwithseveralindicatorsofdiseaseactivityinbothCDandUC,suggestingthattheIRF1levelisassociatedwithIBDdisease

http://wendang.chazidian.com|

1953

内容需要下载文档才能查看

TangetalIn?ammBowelDis??Volume20,Number11,November2014FIGURE1.TheexpressionlevelsofIRF1andOPNinhumanIBD.A,ThelevelsofOPNandIRF1messengerRNAweredeterminedbyreverse-transcriptasePCR.Therelativelevelswerequanti?edthroughthegrayvalues.ThedataindicatedthatthelevelofIRF1messengerRNAwasincreased,whereasthelevelofOPNmessengerRNAwasdecreasedinbothUCandCDtissues(P,0.001)(Mann–WhitneyUtest).B,RepresentativeimmunostainingofIRF1andOPNinnormalintestinalandIBD-in?amedtissues.OPNexpressionwasrobustinnormalintestinalepitheliumandloworevenundetectedinIBD-damagedepithelium,whereasIRF1showedagreaterexpressioninepitheliumofIBDtissuescomparedwithnormaltissues.C,TheaverageopticaldensitiesofOPNweresigni?cantlyhigherintheCDandUCgroupsthaninthecontrolgroup,P,0.001.IncontrasttoOPNexpression,theaverageopticaldensitiesofIRF1weremuchhigherinIBDtissuesthaninnormaltissues,P,0.001.Furthermore,therewereinversecorrelationsoftheIRF1levelswiththeOPNlevelsinbothCD(r¼20.593,P,0.001)andUC(r¼20.691,P,0.001)epithelium.***P,0.001versuscontrol.

1954|http://wendang.chazidian.com