Elevated signal transducers and activators of transcription 1 correlates with increased C-C motif chemokine ligand 2 and...

MISSING IMAGE

Material Information

Title:
Elevated signal transducers and activators of transcription 1 correlates with increased C-C motif chemokine ligand 2 and C-X-C motif chemokine 10 levels in peripheral blood of patients with systemic lupus erythematosus
Series Title:
Arthritis Research and Therapy
Physical Description:
Mixed Material
Creator:
Paul R Dominguez-Gutierrez
Angela Ceribelli
Minoru Satoh
Eric S Sobel
Westley H Reeves
Edward KL Chan
Publisher:
Arthritis Research and Therapy
Publication Date:

Notes

Abstract:
Introduction: The present study examines the levels of recently reported biomarkers, adenosine deaminase acting on RNA (ADAR), C-C motif chemokine ligand 2 (CCL2), C-X-C motif chemokine 10 (CXCL10), signal transducers and activators of transcription 1 (STAT1), and miR-146a in systemic lupus erythematosus (SLE) patients over multiple visits. Methods: Peripheral blood leukocytes were collected from 65 healthy donors and 103 SLE patients, 60 of whom had samples from 2 or more visits. Total RNA was isolated and analyzed for the expression of mRNA and microRNA using Taqman real time PCR assays. Relative expression of I-IFN signature genes, chemokines, and miR-146a were determined by the ΔΔCT method. Results were correlated with clinical data and analyzed by Wilcoxon/Kruskal-Wallis test and Fisher’s exact test. Results: Levels of ADAR, CCL2, CXCL10, and STAT1 in SLE were significantly elevated compared with the healthy controls (P <0.0001). ADAR, CCL2, and CXCL10 showed significant correlation with IFN score in both healthy donors (P <0.0033) and SLE patients (P <0.0001). In SLE patients, miR-146a level was not significantly different from healthy controls nor correlated to the IFN score. Two STAT1 populations were identified: a low STAT1 and a high STAT1 group. High STAT1 patient visits displayed higher (P ≤0.0020) levels of CCL2 and CXCL10 than the low STAT1 patient visits. STAT1 levels correlated with IFN score in low STAT1 group but not in high STAT1 group. More importantly, high STAT1 levels appeared as an enhancer of CCL2 and CXCL10 as indicated by the significantly stronger correlation of CCL2 and CXCL10 with IFN score in high STAT1 patient visits relative to low STAT1 patient visits. Conclusion: Our data indicate a novel role for STAT1 in the pathogenesis of SLE as an expression enhancer of CCL2 and CXCL10 in SLE patients with high levels of STAT1. Future study is needed to examine the exact role of STAT1 in the etiology of SLE.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution.
System ID:
AA00020063:00001

Full Text

PAGE 1

Elevated signal transducers and activators of transcription 1 correlates with increased C C motif chemokine ligand 2 and C X C motif chemokine 10 levels in peripheral blood of patients with systemic lupus erythematosusPaul R. Dominguez Gutierrez1, Angela Ceribelli1, Minoru Satoh2,3, Eric S. Sobel2, Westley H. Reeves2and Edward K.L. Chan1* 1Department of Oral Biology, University of Florida, P.O. Box 100424, 1395 Center Drive, Gainesville, FL 32610 0424, USA, 22Division of Rheumatology and Clinical Immunology, Department of Medicine, 3Department of Pathology, Immunology, and Laboratory Medicine, University of Florida, P.O. Box 100221, 1600 SW Archer Rd, Gainesville, FL 32610 0221, USAAdditional file 1 Figure S1 4

PAGE 2

Figure S1 Anti dsDNAlevel, IFN score, STAT1, CCL2, and CXCL10 in individuals with different ethnic background. A D. Anti dsDNAtiter was significantly higher in SLE patient visits of AA, EA, and LA race compared to HD of the respective race. E L,Q T. IFN score, STAT1, and CXCL10 were higher ( P 0.040) in AA, EA, and AsApatients than HD, while no significant difference was observed for LA M P CCL2 was only significantly higher in AA but not in EA, LA, and AsA.

PAGE 3

Figure S2 Comparison of SLEDAI, anti dsDNAtiter, IFN score, STAT1, CCL2, and CXCL10 in patients with different ethnic background. To examine whether difference of race could affect clinical and biomarkers, AA, EA, LA, and AsApatient visits were compared to each other. Overall, AA patient visits displayed generally higher SLEDAI ( A ), anti dsDNAtiter ( B ), IFN score ( C ), STAT1 ( D ), CCL2 ( E ), and CXCL10 ( F ) than SLE patient visits of any other races.

PAGE 4

Figure S3 IFN score, CCL2, and CXCL10 in individuals with different ethnic background and STAT1 levels. IFN score ( A ), CCL2 ( B ), and CXCL10 ( C ) levels were compared segregated based on race and high and low STAT1. A IFN scores were not significantly different between high and low STAT1 groups of the same race. In high STAT1 patient visits did not display significant differences among the different ethnic groups. There were too few patients in AsA high STAT1 group to be included. However, in the low STAT1 patient visits, AA were higher than EA and LA, EA were higher than LA, and LA had the lowest levels. B C High STAT1 patient visits displayed higher CCL2 and CXCL10 compared to their corresponding low STAT1 groups of the same race. Ethnic differences were not observed for AA, EA, and LA in the high STAT1 patients; however, in low STAT1 patient visits, similar difference appeared to have the same effect on CCL2 and CXCL10 expression as described above for IFN scores.

PAGE 5

Figure S4 CCL2, CXCL10, and IFN score in individuals with different ethnic background and high vslow STAT1 groups. An analysis similar to that in Figure 7 was performed, but in addition, patient visits were segregated by race and by high and low STAT1. The effects of high and low STAT1 on the association of CCL2 and CXCL10 with IFN score did not appear to be influenced by AA ( A,B ), EA ( C,D ) and LA ( E,F ) ethnicity.



PAGE 1

RESEARCHARTICLEOpenAccessElevatedsignaltransducersandactivatorsof transcription1correlateswithincreasedC-Cmotif chemokineligand2andC-X-Cmotifchemokine10 levelsinperipheralbloodofpatientswithsystemic lupuserythematosusPaulRDominguez-Gutierrez1,6,AngelaCeribelli1,4,5,MinoruSatoh2,3,EricSSobel2,WestleyHReeves2andEdwardKLChan1*AbstractIntroduction: Thepresentstudyexaminesthelevelsofrecentlyre portedbiomarkers,adenosinedeaminaseactingon RNA(ADAR),C-Cmotifchemokineligand2(CCL2),C-X-Cmotifchemokine10(CXCL10),signaltransducersandactivators oftranscription1(STAT1),andmiR-146ainsystemiclupu serythematosus(SLE)patientsovermultiplevisits. Methods: Peripheralbloodleukocyteswerecollectedfrom65healthydonorsand103SLEpatients,60ofwhomhad samplesfrom2ormorevisits.TotalRNAw asisolatedandanalyzedfortheexpressionofmRNAandmicroRNAusingTaqman realtimePCRassays.Relativeexpressi onofI-IFNsignaturegenes,chemokines,andmiR-146aweredeterminedbytheCT method.ResultswerecorrelatedwithclinicaldataandanalyzedbyWilcoxon/Kruskal-WallistestandFisher ’ sexacttest. Results: LevelsofADAR,CCL2,CXCL10,andSTAT1inSLEweresigni ficantlyelevatedcomparedwiththehealthycontrols ( P <0.0001).ADAR,CCL2,andCXCL10showedsignificantcorrelationwithIFNscoreinbothhealthydonors( P <0.0033) andSLEpatients( P <0.0001).InSLEpatients,miR-146alevelwasnotsi gnificantlydifferentfromhealthycontrolsnor correlatedtotheIFNscore.TwoSTAT1populationswereid entified:alowSTAT1andahighSTAT1group.HighSTAT1 patientvisitsdisplayedhigher( P 0.0020)levelsofCCL2andCXCL10thanthelowSTAT1patientvisits.STAT1levels correlatedwithIFNscoreinlowSTAT1groupbutnotinhighST AT1group.Moreimportantly,highSTAT1levelsappeared asanenhancerofCCL2andCXCL10asindicatedbythesignificantlystrongercorrelationofCCL2andCXCL10withIFN scoreinhighSTAT1patientvisitsrelativetolowSTAT1patientvisits. Conclusion: OurdataindicateanovelroleforSTAT1inthepathog enesisofSLEasanexpressionenhancerofCCL2and CXCL10inSLEpatientswithhighlevelsofSTAT1.FuturestudyisneededtoexaminetheexactroleofSTAT1inthe etiologyofSLE.IntroductionSystemiclupuserythematosus (SLE)isachronicsystemic autoimmunediseasecharacterizedbyperiodsofincreased diseaseactivity,referredtoasflare-ups,andperiodsofremission.Severalgeneticandenvironmentalfactorshave beenimplicatedinSLEetiopathogenesis,butinrecent yearsincreasedtypeIinterferon(IFN-I,IFN andIFN ) expressionhasbeendiscoveredtoplayakeyroleinthemajorityofSLEpatients,despitebeingknownforover30years thatitiselevatedinSLEpatients[1-4].Becauseofthetechnicalchallengesinmeasuringthenumerousisoformsof IFN ,onecommonwaytoevaluateIFN-Iexpressionisto examinethelevelsofcommonIFN-induciblegenes,such as2 ,5 -oligoadenylatesynthetase(OAS1),myxovirusresistance1(MX1),andlymphocyteantigen6complexlocus E(LY6E);themRNAlevelsoftheseIFN-I-induciblegenes *Correspondence: echan@ufl.edu1DepartmentofOralBiology,UniversityofFlorida,P.O.Box100424,1395 CenterDrive,Gainesville,FL32610-0424,USA Fulllistofauthorinformationisavailableattheendofthearticle 2014Dominguez-Gutierrezetal.;licenseeBioMedCentralLtd.Thisisanopenaccessarticledistributedunderthetermsof theCreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse, distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20 http://arthritis-research.com/content/16/1/R20

PAGE 2

arethenusedtocalculateth eIFNscore[1,5-7].Another interferoninduciblegenethatplaysanimportantantiviral andimmunomodulatoryfunctionistheadenosinedeaminaseactingonRNA(ADAR).ADARisanenzymethatcatalyzestheconversionfromade nosine(A)toinosine(I)in double-strandedRNA(dsRNA)substrate[8,9],withanimpactonRNAatdifferentlevels,suchasmRNAsplicing anddegradation[10,11].Furthermore,ADAR1hasbeen observedtosuppressinterferonregulatoryfactor(IRF)3 andproteinkinaseRNA-activated(PKR)andtherefore blockingIFNinduction[12-14].TheabilityofADAR1to respondandregulateIFN-Ip roductionmakesitanintriguingIFN-I-induciblegenetoexamineinSLE.Uptonow, ADAR1expressionhasonlybeenobservedinT-cellsof SLEpatients,asshowninalimitednumberofstudies [15-17].Infact,Laxminarayana etal .showedthatADAR1 isupregulatedapproximately 3-foldinSLEpatients[15]. Thesamegrouplaterobservedtheincreasedediting ofADAR2byADAR1inT-cellsofSLEpatients[16].Additionally,duetoincreasedADAR1inSLEpatients, Orlowski etal .observedanincreaseofphosphodiesterase 8A1,whichparticipatesintheterminationofcyclicnucleotidesignalingbyhydrolyzingcAMPandcGMPandisactivatedbyIFNandenhancesT-celladhesion[17]. OtherIFN-I-induciblegenesincludesignaltransducers andactivatorsoftranscription(STAT)1and2.STAT1is involvedintypeI,II,andIIIIFNsignalingandhasbeen observedtobeelevatedinSLE[18].Inresponsetotype IIFN,STAT1causesIFNreceptor(IFNAR)1and2 dimerization,activationandphosphorylationofIFNAR byTyk2andJak1,andthusdockingandphosphorylation ofSTAT1andSTAT2[19].TheheterodimerSTAT1STAT2isthentranslocatedintothenucleuswhereit canbindspecificpromotersplayingakeyroleinIFN signalingandproduction[20]. BesidesSTAT1andADAR,IFN-regulatedchemokines havebecomeanotherimportanttopicofresearchinrecent years[21].Twoofthesechemokineshavebeenshownto beSLEbiomarkers,andtheyarecalledC-Cmotifchemokineligand2(CCL2)andC-X-Cmotifchemokine10 (CXCL10)[22].CCL2,formerlyreferredtoasmonocyte chemotacticprotein-1(MCP-1),isapotentrecruiterof monocytes,T-cells,basophils,anddendriticcellstothesite ofinfectionortissuedamage,butithasnoeffectonneutrophilsoreosinophilsunlesstheN-terminusofCCL2is cleaved[18,23].Somecelltypessuchasmonocytes,macrophages,anddendriticcellscanprimarilysecreteCCL2, whichsignalsviathecellsurfacereceptorsCCR2and CCR4andisupregulatedbyIFN andIFN [24,25].The roleofCCL2isbeneficialinclearingpathogens,butithas alsobeeninvolvedinsomepathologicalprocesses.Ina glomerulonephritismousemodel,CCL2playsarolein crescentformationandintersti tialfibrosissupportedbythe observationthatanti-CCL2antibodiescanreducecrescent formation,renalimpairment,andscarring,aswellasTcell andmacrophageinfiltration[26].CCL2hasbeenobserved intherecruitmentofTcellsandmonocytes/macrophages inlupusnephritisandblockadeofCCL2ameliorateslupus nephritisinMRL-(Fas)lprmice[23,27].Inaserologic proteomestudybyantibodymicroarrayinSLE,CCL2was identifiedasoneofthetwelve upregulatedproteins;furthermoreCCL2wasoneofthreechemokinesthatwould precedelupusflare,indicatingthattheyaregoodpredictors ofincreasedSLEactivity[21]. CXCL10,alsoknownasIFNgamma-inducedprotein 10(IP-10),isachemokineoftheC-X-Cmotiffamily. SimilartoCCL2,CXCL10isapotentattractorofmonocytes,macrophages,T-cells,naturalkiller(NK)cells,and dendriticcellstositesoftissuedamageandinfection [28,29].CXCL10isanIFN-responsecytokinethatbinds itsCCL3receptorandactsviaJak/STATpathwayactivation[30-32].EventhoughCXCL10isapotentimmune responderforbacterialandviralinfectionsandacritical biomarkerfororgantransplantrejection,itsroleinthe pathogenesisofautoimmunediseasesisnotclear[33,34]. Furthermore,thecombinationofCXCL10andCCL2 proteinlevelscouldbeusefulaspredictionfactorforupcomingflares[22]. ThereasonbehindupregulationandcontrolofIFNin SLEisnotknown,butsomestudieshaverecently focusedonthepossibleroleplayedbyselectedmicroRNAs(miRNAs).MiRNAsaresmallnon-encoding20to23-nucleotide-longRNAs,thatregulatetheirtarget mRNAbybindingtothe3 UTR,causingtranslational repressionand/ordegradationoftargets.miR-146ais oneofthemostsignificantmiRNAsinregulatinginnate immuneresponseandtolerance[35]anditwasfirst showntobeinvolvedintoll-likereceptor(TLR)regulation throughthenuclearfactor(NF)Bpathway[36].miR146awouldfunctiontoattenuatetheimmuneresponse andregulateinflammationinnormalimmuneresponse andautoimmunedisorders,anditisalsoacriticalregulatorofendotoxin-inducedtoleranceandcross-tolerance [37-39].Todate,miR-146ahasbeenfoundinassociation withautoimmunediseasessuchasSjgren  ssyndrome [40],psoriasis[41,42],andrheumatoidarthritis[43-45]. Tang etal .reportedthatmiR-146awasunder-expressed inperipheralbloodmononuclearcells(PBMCs)ofChinese SLEpatients[46].miR-146awassignificantlylowerinpatientswithactiveSLEwithproteinuriacomparedtothose withinactiveSLE[46].Additionally,SLEpatientsdisplayed aninversecorrelationbetweenmiR-146aexpressionand IFNscore[46].Tang etal.alsodemonstratedthatreduction ofmiR-146amayenhancethesignalingduetoelevated levelsofSTAT1andIRF5whichleadstoincreasedproductionofIFN[46].ThereducedlevelsofmiR-146aobserved inChineseSLEpatientscouldp otentiallyexplainelevation ofIFNbylossofregulationofSTAT1expression.Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page2of14 http://arthritis-research.com/content/16/1/R20

PAGE 3

Ourpresentstudyevaluatestheinteractionamong STAT1,ADAR,CCL2,CXCL10,andmiR-146ainSLE patientsandhealthycontrols,demonstratingthatallexceptformiR-146acorrelatewithIFNscoreinbothSLE patientsandhealthydonors.MethodsHealthydonors ’ andSLEpatients ’ demographicdataWholebloodwascollectedfromatotalof103SLEpatientsand65healthycontrolsenrolledintheUniversity ofFloridaCenterforAutoimmuneDiseasesregistry from2008to2011.Healthydonors(HD)wereselected basedonnohistoryofautoimmunedisease,andallSLE patientssatisfiedtheAmericanCollegeofRheumatology(ACR)criteria[47].Healthydonorsonlyvisitedthe cliniconce,therefore,theyrepresentasinglevisit. Therewereatotalof180SLEvisitswithsequential samplescollectedin60SLEpatients(Table1).SLEpatientsandhealthycontrolsweresegregatedbyethnic profile(Table1).Allhumanbloodsampleswereobtainedfromenrolledindividualswiththeapprovalof theinstitutionalreviewboard(IRB)attheUniversityof Florida.Thisstudymeetsandisincompliancewithall ethicalstandardsinmedicineandinformedconsentwas obtainedfromallpatientsaccordingtotheDeclaration ofHelsinki.LeukocytesandRNApurificationPeripheralbloodleukocyteswerecollectedfromwhole bloodusingAmbionLeukoLOCKkit(Ambion,Austin,TX, USA).LeukoLOCKfilterswerewashedtwicewith3mlof PBSandstabilizedwith3mlofRNAlatersolution.Stabilizedfilterswerestoredforaminimumof24hat 80CbeforecollectingtotalRNA.TotalRNA,includingsmall RNAs,wascollectedusingthe  AlternativeProtocol Ž (version0602,Ambion)fortheextractionofRNAfromcells capturedonLeukoLOCKfiltersusingTRIreagent.mRNAandmicroRNAquantitativeRT-PCROAS1(Hs00973637_m1),MX1(Hs00895608_m1),LY6E (Hs00158942_m1),STAT1(Hs01013996_m1),CCL2 (Hs00234140_m1),CXCL10(Hs00171042_m1),ADAR (Hs00241666_m1),TNF (Hs00174128_m1),andprimiR-146a(Hs033303259_pri)levelswereanalyzedby TaqManmRNAassayprimers(AppliedBiosystems, FosterCity,CA,USA).mRNAqRT-PCRwasperformed asaduplexwith18SrRNAassayedasthenormalizer. mRNAwastranscribedtocDNAusingtheTaqManHighCapacitycDNAReverseTranscriptionKitfollowedby quantitative(q)PCRusingTaqManFastAdvancePCR MasterMix(AppliedBiosystems).miR-146a(000468; Catalogue#4427975)wasanalyzedbymiRNAqRT-PCR usingtheTaqManMicroRNAReverseTranscriptionKit, TaqManFastAdvancePCRMasterMix,andTaqMan MicroRNAprimers(AppliedBiosystems).Allreactions wereanalyzedusingStepOneReal-TimePCRSystem (AppliedBiosystems).Anti-dsDNAELISAAfterthecollectionofleukocyteswiththeLeukoLOCKfilters,theleukocytefreebloodwastransferredto10ml VacutainerSSTplusblood-collectiontubes(BD,Franklin Lakes,NJ,USA).Bloodwascentrifugedat1,000gfor 20minutes.Theplasmawastransferredtoa15-mlconical tubeandstoredat 20C.Anti-dsDNAELISAwasperformedaspreviouslydescribed[48].Inbrief,anti-human IgGsecondaryantibodywasusedandsampleswereconsideredpositivewhentheabsorbancewasgreaterthanthe meanplusthreeSDfromthehealthycontrols.ComplementlevelsC3andC4complementlevelswereobtainedfromclinicaldata.C3levelslowerthan90mg/dlandC4levels lessthan15mg/dlwereconsideredaslowcomplement levelsintheanalysis. Table1DemographicdataofSLEpatientsand healthydonorsSLEHD Numberofcases10365 Numberofvisits(2ormore)180(60)65(0) Meanage,years(range)44(25 – 68)33(19 – 59) Sex,female/male/unknown,n90/13/031/9/24 Race,AA/EA/LA/AsA/IrA/unknown1,n35/50/12/3/3/010/18/6/3/2/25 Racebyvisit (AA/EA/LA/AsA/IrA/unknown)1,n 64/86/20/7/3/010/18/6/3/2/25 Active/inactivebySLEDAI, numberofpatients 49/131N/A Mainclinicalfeatures, numberofpatients Malarrash9/94N/A Discoidlesions2/101N/A Photosensitivity8/95N/A Oralulcers7/96N/A Arthritis17/86N/A Serositis4/99N/A Pleuritis5/98N/A Nephritis49/54N/A Seizures2/101N/A Psychosis1/102N/A1AA,AfricanAmericans;AsA,AsianAmericans;EA,EuropeanAmericans;HD, healthydonorswithnohistoryofautoimmunedisease;IrA,interracial Americans;LA,LatinoAmericans;unknown,undisclosedrace;N/A, notapplicable.Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page3of14 http://arthritis-research.com/content/16/1/R20

PAGE 4

IFNscoreandSLEactivityTheexpressionofthreeknowntype-IIFNsignaturegenes, MX1,OAS1,andLY6E,wasz-transformedintoIFNscore aspreviouslyshown[1,49].TheSLEdiseaseactivityindex (SLEDAI)wasusedtoclassifythepatientsintoactive (SLEDAI>4)orinactive(SLEDAI 4)atthetimeofthe visit(Table1)[50,51].CellcultureandinnateimmuneligandstimulationHumanTHP-1cellswereobtainedfromtheAmerican TypeCultureCollection(ATCC,Manassas,VA,USA). THP-1cellsweremaintainedinRPMIcontaining10%(v/v) FBS(Mediatech,Manassas,VA,USA)and100U/mlpenicillin-streptomycin(Mediatech).ForanalysisofTHP-1 monocyteresponsetoligand invitro ,log-phasecellswere seededat5105cells/mlina24-wellplate.Cellswere stimulatedwiththefollowing agonists:1,000ng/mloflipopolysaccharidefrom S.enterica serotypeMinnesotaRe595 (LPSSe,TLR4ligand,Ultrap uregrade,Sigma-Aldrich Corp.St.Louis,MOUSA),0.10and1.0ng/mlIFN 2(PBL InterferonSource,Piscat away,NJ,USA),and0.10or 1.0ng/mlIFN (PBLInterferonSource).TLR4ligands werereconstitutedinendotoxin-freewaterandusedat concentrationsasreportedbefore[38].IFN 2andIFN werereconstitutedinendotoxin-freePBSwith1mg/ml BSAtomake5g/mlstocksstoredat 80C.DataanalysisThecopynumberofmiR-146awasnormalizedtototal loadedRNA,whereasmRNAlevelswerenormalizedto 18SRNA.ThecopynumberofmiR-146awasdetermined usingastandardcurvewithsyntheticmiR-146a(Integrated DNATechnologiesInc.,Coralville,IA,USA)[52].Relative expressionofmRNAcompared tocontrolswasdetermined bytheCT(cyclethreshold)method[53].Analyseswere performedusingSASversion9.2andJMPGenomics version5(SAS,Cary,NC,USA) .TheWilcoxon/KruskalWallistestwasusedtoevaluatesignificancebetween groups,whereastheWilcoxonsignedranktestformatched pairswasusedtoevaluateSLEpatientswithtwovisits. P -values<0.05wereconsidered significant.Beforeapplying ordinarylinearregressionan alyses,thedistributionsof datasetswereconfirmedfornormality.Thecoefficientof determination( r2)wasusedtodeterminelinearcorrelation. Significantdifferencesbetweenslopeswasevaluatedby analysisofcovariance(ANCOVA).TheGeneralized EstimatingEquation(GEE)modelforRepeatedMeasures wasusedtoaccountforpossiblewith-insubjecteffects frompatientswithmultiplevisits[54].ResultsExpressionofcandidatebiomarkersintheSLEcohortTodeterminewhetherpreviouslyreportedbiomarkers wereelevatedinourSLEpatientcohort,wemeasured thebiomarkerexpressionlevelsinHD,activeSLE,and inactiveSLEpatientvisits(Figure1).TheSLEcohort wassegregatedbySLEDAIintoactiveSLE(49visits, SLEDAI>4)andinactiveSLE(131visits,SLEDAI 4). ThelevelofIFN-IwasestimatedbyquantifyingtheexpressionofIFN-induciblegenes.TheIFNscore,STAT1, ADAR,CCL2,andCXCL10levelsweresignificantlyelevatedatbothactiveandinactiveSLEpatientvisitscomparedtoHD(Figure1A-E),establishingandconfirming thatthesebiomarkerswereabe rrantlyoverexpressedinour SLEpatients.Toexploreifthesebiomarkerswerecapable ofdistinguishingdiseaseactiv itystatus,activeandinactive patientvisitswerecomparedtooneanother.Nosignificant differencewasobservedbetweenactiveandinactiveSLE patientvisitsforIFNscore (Figure1A,meanSD,62.7 6.1unitsversus57.84.9units),ADAR(Figure1C,5.27 0.31foldversus5.270.23fold),andCXCL10(Figure1E, 158.1-fold26.6versus120.0-fold10.5),butSTAT1 (Figure1B,44.810.7vs34.46.6fold, P =0.033)and CCL2(Figure1D,18.2-fold3.1versus9.96-fold1.42, P =0.0061)weresignificantlyelevatedinactiveSLEcomparedtoinactiveSLEpatientvisits.TNF ,whichisnot generallyinvolvedinthepathogenesisofSLE,wasusedas anegativecontrol.Asexpected,TNF wasnotsignificantly differentamongthethreegroups(Figure1F).Similarly miR-146adidnotdisplayanysignificantdifferenceamong activeSLE,inactiveSLE,andHD(Figure1G).Tovalidate this,wedeterminedthelevelsoftheprimarytranscriptof miR-146a(pri-mir-146a)whichalsodidnotdemonstrate anysignificantdifferenceamongactiveSLE,inactiveSLE, andHD.WiththeexceptionofmiR-146a,theseresultsare consistentwithreportsonSLEpatientswithelevatedIFN scorecomparedtoHD[1,49]aswellasupregulatedlevels ofIFNsignaturegenes(STAT1andADAR)[15-17]and chemokines(CCL2andCXCL10)[21]. Theclinicalandexpressiondatawerecorrelatedwith anti-dsDNAautoantibodylevel,whichisanindicatorfor patients  diseaseactivityincertainpatients[55-58].DecreasesinC3andC4levelscorrelatedwithSLEactivityand renaldamageaswellasincreasedlevelsofanti-dsDNA antibodies[59].Anti-dsDNAautoantibodylevelshavealso beenusedforsub-classificationofSLEpatients[60,61].SLE patientvisitsandHDweresegregatedintoanti-dsDNA(+) andanti-dsDNA( ).Patientvisitsthatwereanti-dsDNA(+) displayedhigherSLEDAIanddecreasedC3andC4levels (Figure2A-C).Theresultsfortheremainingbiomarkers (Figure2D-K)closelyresembledthosefromactiveversus inactiveSLEDAIresults(Figure1). Theinfluenceofraceinanti-dsDNA,IFNscore,STAT1, CCL2,andCXCL10werealsoexamined.AfricanAmericans (AA)andEuropeanAmericans(EA)contributedto83.3% ofthevisits,followedbyLatinAmericans(LA)andAsian Americans(AsA)for15%,andinterracialAmericans(IrA) forlessthan2%ofpatientvisits(Table1).DuetothesmallDominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page4of14 http://arthritis-research.com/content/16/1/R20

PAGE 5

samplesize,IrAwereexcludedinallsubsequentanalyses. Ingeneral,resultsshowthathigherlevelsofanti-dsDNA, IFNscore,STAT1,CCL2,andCXCL10wereobservedinall racegroupsanalyzed(Additio nalfile1:FigureS1).Thelack ofstatisticallysignificantd ifferencesbetweenSLEandHD incertaingroups,suchasLA ,mightbeduetothesmaller samplesizes. Bycomparingpatientsofdifferentrace(Additionalfile1: FigureS2),thelevelsoftheparametersexaminedwereall higherinAAthanEA,LA,andAsA.Inparticular,AAhad significantlyhigherSLEDAI( P =0.024),anti-dsDNAlevel ( P =0.044),IFNscore( P =0.0005),STAT1( P =0.0011), CCL2( P =0.0004),andCXCL10( P =0.0004)thanEA.Furthermore,AAhadsignificantly( P 0.014)higherIFNscore, Figure1 CorrelationofIFNscore,STAT1,ADAR,CCL2,CXCL10,andmiR-146alevelstosystemiclupuserythematosus(SLE)disease activity.(A) IFNscore, (B) STAT1, (C) ADAR, (D) CCL2,and (E) CXCL10weresignificantlyelevatedinSLEpatientvisits(activeandinactivedisease activityareindicatedas+and-respectively)comparedtohealthydonors(HD).Nostatisticaldifferencewasdetectedbetweenactiveandinactive SLEforIFNscore,ADAR,orCXCL10. (F-H) TNF ,pri-miR-146a,andmiR-146adidnotshowanysignificantdifferenceamongthegroups.STAT, signaltransducersandactivatorsoftranscription;ADAR,adenosinedeaminaseactingonRNA;CCL2,C-Cmotifchemokineligand2;CXCL10, C-X-Cmotifchemokine10. Figure2 CorrelationofIFNscore,STAT1,ADAR,CCL2,CXCL10,andmiR-146alevelstoanti-double-strandardDNA(dsDNA) autoantibodies.(A) Systemiclupuserythematosusdiseaseactivity(SLEDAI)scoresweresignificantlyhigherinanti-dsDNA(+)thananti-dsDNA( ) patientvisits. (B C) C3andC4weresignificantlylowerinanti-dsDNA(+)thaninanti-dsDNA( )patients. (D) IFNscore, (E) STAT1, (F) ADAR, (G) CCL2,and (H) CXCL10weresignificantlyelevatedinSLEpatientvisitscomparedtohealthydonors(HD),butnostatisticaldifferenceswere detectedbetweenanti-dsDNA(+)andanti-dsDNA( )patientsforIFNscore,ADAR,orCXCL10. (I K) TNF ,pri-miR-146a,andmiR-146adidnot showanysignificantdifferenceamongthegroups.STAT,signaltransducersandactivatorsoftranscription;ADAR,adenosinedeaminaseactingon RNA;CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotifchemokine10. Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page5of14 http://arthritis-research.com/content/16/1/R20

PAGE 6

STAT1,CCL2,andCXCL10thanLA(Additionalfile1: FigureS2B-F).Alsointhiscase,thelackofadditionalstatisticallysignificantresultsforLAandAsAmightbedue tothesmallsamplesizes.However,AAclearlydisplayed increasedbiomarkerlevelsmorethananyotherrace.BiomarkerinterrelationshipinSLEpatientswith returnvisitsToexpandupontheinterrelationshipofthesebiomarkers, datafromSLEpatientswithtwoconsecutivevisitswere segregatedforanalysesbyincreasingordecreasingIFN scorebyatleast50%fromthefirsttothesecondvisit.PatientswithincreasingIFNscorefromonevisittothenext (n=13; P =0.0001,Figure3A)displayedasignificantincreaseinSTAT1( P =0.0017),CCL2( P =0.0086),CXCL10 ( P =0.038),andmiR-146a( P =0.0034).Similarly,forSLE patientswithincreasingSTAT1byatleast50%between thefirstandsecondvisit(n=25; P <0.0001,Figure3B), significantincreaseswereobservedforIFNscore( P = 0.027),CCL2( P <0.0001),CXCL10( P =0.0003),andmiR146a( P =0.0078).ThestrongcorrelationbetweenSTAT1, CCL2,andCXCL10wereexpected;however,correlation betweenIFNscoreandincreasingSTAT1wasweaker thanexpected.ThismaybeindicatingthathighSTAT1 levelsdonotnecessarilytranslateintohighlevelsofIFN-I. ThehighlysignificantcorrelationbetweenmiR-146alevels andIFNscoreinthereturnvisitswasunexpected,asthe levelofmiR-146ainSLEwasnotsignificantlydifferent fromHD(Figures1Hand2H)andalsoitwaspreviously reportedtobedecreasedinSLEandinverselycorrelated withIFNscoreinaChineseSLEcohort[46]. SLEpatientswhohaddecreasingIFNscorebyatleast 50%betweenfirstandsecondvisit(N=32; P <0.0001, Figure3C)displayedasignificantdecreaseinSTAT1( P = 0.0002)andCXCL10( P =0.0002),butnotinCCL2and miR-146a.Similarly,SLEpatientswithdecreasingSTAT1 (n=13; P =0.0001,Figure3D)hadsignificantdecreasein IFNscore( P =0.0001)andCXCL10( P =0.0004),whereas nosignificantchangesinCCL2andmiR-146awereobserved.ByrankingpatientsaccordingtodecreasingIFN scoreorSTAT1,thereversaloftheresultsfromranking byincreasingIFNscoreorSTAT1shouldideallyhave beenobserved.Interestingly,theexceptionwasobserved onlyforCCL2andmiR-146a(Figure3C,D).RelationshipofIFNscoretootherbiomarkersTobetterunderstandwhethertheassociationofIFN scorewiththeotherbiomarkersinpairedpatient-visits couldbeexpanded,levelsofADAR,CCL2,andCXCL10 fromtheentirecohortofSLEpatientvisitsandHDwere correlatedtoIFNscore(Figure4).ADAR,CCL2,and CXCL10displayedsignificantcoefficientofdetermination( r2)inbothSLEandHD(Figure4).Theconsistent significantcorrelationsofthesegenestoIFNfromthe Figure3 Systemiclupuserythematosus(SLE)patientswithtwovisitsrankedbyincreasingordecreasingIFNscoreandSTAT1. Data fromthefirstandsecondvisitsforeachpatientisdenotedbyanindividualcolorline. (A) SLEpatientsrankedbyincreasingIFNscorefromthe firsttothesecondvisitshowedsignificantincreaseinSTAT1,CCL2,CXCL10,andmiR-146a. (B) PatientsrankedbyincreasingSTAT1alsoshowed significantincreaseinIFNscore,CCL2,CXCL10,andmiR-146a. (C) SLEpatientsrankedbydecreasingIFNscorefromthefirsttothesecondvisit showedsignificantdecreaseonlyinSTAT1andCXCL10. (D) PatientsrankedbydecreasingSTAT1showedsignificantdecreaseinIFNscoreand CXCL10.STAT,signaltransducersandactivatorsoftranscription;CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotifchemokine10. Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page6of14 http://arthritis-research.com/content/16/1/R20

PAGE 7

lowlevelsobservedinHD(Figure4,rightpanels)toaberrantlyhighpathogeniclevelsofIFNinSLEpatientvisits (Figure4,leftpanels)wasindicativeofanormalintrinsic responseofADAR,CCL2,andCXCL10toIFNproduction.Contrarytoanearlierreportshowingthatthelevel ofmiR-146awasnegativelycorrelatedwithIFNscore[46], miR-146aaswellaspri-miR-146adidnotdisplayanysignificantcorrelationwithIFNscoreineitherHDorSLE patients(datanotshown).Surprisingly,inthesametype ofanalysis,STAT1didnotdisplayasignificantcorrelation toIFNscoreeither(datanotshown).Furtheranalysisof STAT1expressionrevealedtwopopulationsafterapplying alog10transformation(Log[STAT1])inbothHDandSLE patients(Figure5A).Usinganarbitrarycutoffof1.50Log [STAT1]tosegregateSTAT1results,valuesbelow1.50 werereferredasthelow-STAT1groupandabove1.50 werethehigh-STAT1group(Figure5B,C).Inthelow STAT1group,SLEpatientvisitsdisplayedsignificantly higherexpressionofSTAT1comparedtoHD(2.44-fold, P <0.0001,Figure5B),butinthehigh-STAT1group,no significantdifferencewasobserved(Figure5C).Furthermore,thelow-STAT1groupdisplayedsignificantpositive associationbetweenSTAT1andIFNscoreinbothHD (Figure5D)andSLEpatients(Figure5E).Incontrast,in thehigh-STAT1grouptherewasnocorrelationbetween STAT1andIFNscore(datanotshown).STAT1levelscorrelatewithSLEactivityTheeffectsofhighandlowSTAT1onIFNscoreand ADARappearedtoberelatedtotheactiveversusinactive statusofSLE(SLEDAI,Figure1A,C)andanti-dsDNA(+) versus( )patients(Figure2A,C)whereIFNscoreand ADARweresignificantlyhigherthaninHD,butnotsignificantlydifferentbetweenSLEpatientvisitswithhigh andlowSTAT1(Figure6A,B).CCL2wassignificantlydifferentbetweenactiveandinactiveSLE,andbetweenHD andactiveandinactiveSLEaswell(Figure1D),whichresemblestheresultsofanti-dsDNA(+versus-)(Figure2D) andhigh-versuslow-STAT1comparisons(Figure6D). SimilarobservationsarevalidforCCL2,withtheaddition thatthereisadifferenceinCCL2expressionbetween high-andlow-STAT1SLE(Figure6C).AsbothSLEDAI activeandanti-dsDNA(+)areindicatorsofincreaseddiseaseactivity,theseresultsindicatethatpatientswithhigh STAT1arealsoinamoreactivediseasestatethanthose withlowSTAT1. Todeterminewhetherethn icitycouldbeaconfoundingfactorfortheeffectsofhighandlowSTAT1,IFN score,CCL2,andCXCL10levelsweresegregatedbased onethnicityandhighandlowSTAT1(Additionalfile1: FigureS3A-C).Overall,highSTAT1patientvisitsdid notshowasignificantdifferenceamongAA,EA,andLA. However,low-STAT1AApatientsshowedsignificantly Figure4 ADAR,CCL2,andCXCL10levelscorrelatewithIFNscoreinbothsystemiclupuserythematosus(SLE)patientsandhealthy donors(HD).(A) ADAR, (B) CCL2,and (C) CXCL10displaydirectandsignificant r2withIFNscoreforbothSLEvisitsandHD.ADAR,adenosine deaminaseactingonRNA;CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotifchemokine10. Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page7of14 http://arthritis-research.com/content/16/1/R20

PAGE 8

higherIFNscore,CCL2andCXCL10comparedtoother groups(Additionalfile1:FigureS3).Theseresultsindicatedthathigh-andlow-STAT1groupswereidentified essentiallyinallethnicities,anddifferencesinIFN score,CCL2,andCXCL2levelswereobservedamong low-STAT1groupsbutnotamongthehigh-STAT1 groups.STAT1influencesthecovarianceofIFNscorewithADAR, CCL2,andCXCL10TodeterminewhetherhighversuslowSTAT1levelsaffectedthecorrelationbetweenIFNscoreandtheother biomarkers,weanalyzedtheseparametersinpatientswith highversuslowSTAT1expression(Figure7).Eventhough ADARexpressionwasreportedtobeSTAT1-independent Figure5 BimodaldistributionofSTAT1intohighandlowgroups.(A) Thelog10transformationofSTAT1showsabimodaldistributionof STAT1withtwopopulations(high-andlow-STAT1groups)withacutoffat1.5log[relativefold-change]ofSTAT1(Log[STAT1])forbothhealthy donor(HD)andSLEpatientvisits. (B) Thelow-STAT1groupsdisplayedsignificantdifferenceofSTAT1inSLEpatientvisitscomparedtoHD. (C) Ontheotherhand,thehigh-STAT1groupsshowednosignificantdifferencebetweenSLEandHD. (D E) Inthelow-STAT1group,STAT1levels displayadirectcorrelationtotheIFNscoreinSLEpatientvisitsandHD.STAT,signaltransducersandactivatorsoftranscription. Figure6 HighlevelsofCCL2,CXCL10,andmiR-146acomparedtolowSTAT1inhighSTAT1systemiclupuserythematosus(SLE) patients.(A) IFNscore, (B) ADAR, (C) CCL2, (D) CXCL10, (E) pri-miR-146a,and (F) miR-146awerecomparedinSLEpatientvisitsandhealthy donors(HD),whichweresegregatedbyhighandlowSTAT1levelsdemonstratingthatpatientswithhighSTAT1SLEexpressedhigherlevelsof CCL2,CXCL10,andmiR-146athanthosewithlowSTAT1.STAT,signaltransducersandactivatorsoftranscription;ADAR,adenosinedeaminase actingonRNA;CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotifchemokine10. Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page8of14 http://arthritis-research.com/content/16/1/R20

PAGE 9

[62,63],patientvisitswithlow-STAT1SLE(red, r2=0.29, P <0.0001),high-STAT1SLE(blue, r2=0.35, P =0.0002) patients,andlow-STAT1HD(black, r2=0.24, P <0.0001) displayedsignificantassociationbetweenADARandIFN score(Figure7A).Similarly,CCL2wassignificantlyassociatedwithIFNscoreinpatientvisitswithlowSTAT1SLE ( r2=0.07, P <0.0010),patientvisitswithhigh-STAT1SLE ( r2=0.76, P <0.0001),andHDwithlowSTAT1( r2=0.08, P =0.0002);alsoCXCL10displayedsignificantassociation withIFNscoreforSLEpatientswithlowSTAT1( r2=0.09, P =0.0003),patientvisitswithhighSTAT1SLE( r2=0.30, P =0.0008),andHDwithlowSTAT1( r2=0.08, P =0.027, Figure7B,C). Theslopeofthelinearregressionrepresentstherate ofchangeofADAR,CCL2,andCXCL10perunitof changeinIFNscore.Thisledtotheintriguingpossibility thatpatientvisitswithhighSTAT1haveahigherslope thanthosewithlowSTAT1.ANCOVAwasusedtotest iftheslopesweresignificantlydifferent(Figure7). ADAR/IFNscoreswerenotsignificantlydifferentbetweenhigh-andlow-STAT1patients(Figure7A,blue versusredline, P -valuenotshown),butCCL2/IFN scoreandCXCL10/IFNscoreslopesweresignificantly higherinthehigh-STAT1(HS)patientscomparedto thelowSTAT1(LS)patients(Figure7B,C,blueversus redline).ThissuggeststhathighSTAT1levelsmayenhanceCCL2andCXCL10expressionpotentiallyinducedbyIFN. Next,westudiedwhethereth nicbackgroundcouldinfluencetheassociationofIFNscorewithCCL2andCXCL10 andalteredtheeffectsofhighandlowSTAT1(See Additionalfile1:FigureS4).Influenceofethnicbackground appearedtobeminimalonCCL2inhigh-STAT1patient visits.CCL2inhigh-STAT1AA,EA,andLAdisplayedvery goodlinearcorrelation( r2>0.59, P 0.0018)withIFNscore (SeeAdditionalfile1:FigureS4A,C,E).Low-STAT1EA andLAalsoshowedgoodlinearcorrelation( r2 0.16, P 0.039,red,Additionalfile1:FigureS4C,E);however, low-STAT1AAdidnotdisplayalinearcorrelationbetween CCL2andIFNscore(SeeAdditionalfile1:FigureS4A). CXCL10hadasignificantcorrelation( r2>0.24, P 0.032) withIFNscoreforhigh-STAT1AAandEA(SeeAdditional file1:FigureS4B,D);however,CXCL10hadsignificantcorrelation( r2>0.25, P 0.0002)withIFNscoreforlowSTAT1EAandLA(SeeAdditionalfile1:FigureS4A,D, F).AsAcouldnotascertainsignificantcorrelationsfor CCL2/IFNscoreandCXCL10,probablyduetothesmall samplesize(datanotshown). Figure7 EffectofhighversuslowSTAT1expressioninADAR,CCL2,andCXCL10correlationwithIFNscore.(A) ADAR, (B) CCL2,and (C) CXCL10displayedasignificantlinearcoefficientofdetermination( r2)withIFNscoreforhighSTAT1(HS,blue,leftpanels).FortheSLEpatient visitswithlowSTAT1(LS,red,leftpanels) r2resembledthatoftheLShealthydonors(HD)(black,rightpanels). (B C) TheslopesofhighSTAT1for CCL2andCXCL10weresignificantlyhigherthanthoseforpatientvisitswithlowSTAT1(leftpanels).ADAR,adenosinedeaminaseactingonRNA; CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotifchemokine10. Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page9of14 http://arthritis-research.com/content/16/1/R20

PAGE 10

InductionofSTAT1,CCL2,andCXCL10inTHP-1cellswith typeIIFNTLRshavebeenimplicatedto playaroleinSLEpathogenesis.TomodeltheresponseofSTAT1,CCL2,andCXCL10 aswellasIFN-I,TLR4wasstimulatedinhumanmonocytic THP-1for24hwith1,000ng/mlofLPS.IFNscoreincreasedataround4handpeakedaround8h(Figure8A). In1.0ng/mlofIFN 2-treatedand0.1ng/mlofIFN treatedTHP-1cells,IFNscoredisplayedasimilartrendas inLPStreatment(Figure8F,K);howeverfor1.0ng/ml IFN -treatedcells,IFNscoreincreaseduptill12h (Figure8K),whereas0.1ng/ml -treatedcellsdisplayedlittle change(Figure8F).TheseresultsdemonstratedTHP-1responsivenesstoIFN-IaswellasthefactthattheywerecapableofIFN-Iproduction. Interestingly,whereasLPSdisplayedagradual,longtermincreaseofCCL2andCXCL10,IFN 2andIFN treatmentsdisplayedrapidincreasesfollowedbydecreases ofCCL2andCXCL10.AfterLPSstimulation,STAT1did notincreasetill4handreacheditspeakexpressionat8h Figure8 THP-1responsetoIFN ,IFN ,andLPSoveraperiodof24h. THP-1cellsweretreatedwithdifferentdosesofIFN ,IFN ,andLPS andlysateswereharvestedatvarioustimesfrom2to24hforRNAisolationandanalyses.IFNscore (A F K) ,andtheexpressionofSTAT1 (B G L) CCL2 (C H M) ,CXCL10 (D I N) ,andmiR-146a (E J O) wereevaluatedat0.1and1.0ng/mlofIFN 2andIFN aswellas1,000ng/mlofLPS.LPS, lipopolysaccharide;STAT,signaltransducersandactivatorsoftranscription;CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotifchemokine1 0; UTX,untreated. Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page10of14 http://arthritis-research.com/content/16/1/R20

PAGE 11

(Figure8B);howeverinTHP-1cellsstimulatedwith IFN 2orIFN ,STAT1increasedat2h,peakingat8h (Figure8G,L).CCL2increasedat2hinLPS-treated THP-1cellsandcontinuedtoincreaseduringthe24-h period(Figure8C);howeverthiswasnotuntilaftermaximumexpressionofSTAT1wasreached(Figure8B),and CCL2begantorapidlyincrease(Figure8C).CCL2increasedat2hin0.1and1.0ng/mlIFN 2-treatedaswell as0.1ng/mlIFN -treatedTHP-1cells,butitpeakedat 4handbegantodecreaserapidly(Figure8H).For1.0ng/ mlIFN treatmentofTHP-1cells,thepeakwasshiftedby 2hsothatCCL2peakedat2handbegantorapidlydecrease(Figure8M).CXCL10displayedatrendsimilarto CCL2for1.0ng/mlIFN 2-treatedand0.1ng/mlIFN treatedTHP-1cells(Figure8I,N).In1.0ng/mlIFN treatmentofTHP-1cells,CXCL10continuedtill8h (Figure8N).TheseresultsindicatedthatCCL2and CXCL10rapidlyrespondedtoIFN 2andIFN stimulationwhereasTLR4stimulationappearedtoinduceaslow gradualincrease,butthenarapidincreaseafterSTAT1 reacheditsmaximumexpression. miR-146aappearedtodifferinitsresponsefromthe otherbiomarkers.LPSupregulatedmiR-1463-foldand itrapidlyreachedapeakofan11-foldincreaseat12h (Figure8E).miR-146ainIFN 2-orIFN -treatedcells showedamodestof3-to4-foldpeakat8h,potentially indicatingthatIFN-IdidnotinducesignificantproductionofmiR-146a(Figure8J,O).DiscussionInthisstudy,expressionofpreviouslyidentifiedSLEbiomarkerswasexaminedandcorrelationtestedwithdemographicandclinicalparameters,focusingontheanalysisof apossiblecorrelationamongthem.Theprimaryanalyses usedordinarylinearregressi on,evenfordatafrommultiple visits,asreportedinFigures4,5,and7.Alternatively,the GEEmodelforrepeatedmeasureswasalsousedtoaccountforpossiblewithin-subje cteffectsfrompatientswith multiplevisits[54].Whenwecomparedtheparameters (slopeandsignificance)fromtheGEEandordinarylinear regression,theresultswerepracticallyidentical(datanot shown).Itisknownthatunlessthevastmajorityofthe sampleshaverepeatedmeasures(patientswithmultiple visits),theordinarylinearregressionisexpectedtoclosely approximatetheGEEmodel[64].Furthermore,evenif therewasstrongcorrelationbetweenvisitsofpatients,ordinarylinearregressionwoul dunderestimatethecorrelationbecauseitassumesthatt hevisitsareindependent; therefore,thecorrelationsofo rdinarylinearregressionsare morestringentthanthoseofGEE[64].Inaddition,wealso assessedthenormalityofeachdatasetbeforeapplyinglinearregression.WiththeexceptionofSTAT1,theIFNscore, ADAR,CCL2,andCXCL10resemblednormaldistributions(datanotshown).Inmostcaseswhendealingwith suchlargedatasets,evenmoderatedeviationsfromnormalcyarenotcriticalduetothecentrallimittheorem[65]. Forthesereasons,wedecidedtoreportordinarylinearregressionratherthanthemorecomplexGEEmodelforrepeatedsamples.BiomarkerassessmentOurresultsshowthatADAR,STAT1,CCL2,andCXCL10 levelsweresignificantlyelevatedintheSLEcohortas expected.Thisisinpartvalid atedbypreviouslypublished resultsshowingincreasedlevelsofthesebiomarkersand theircorrelationtoIFN-IproductioninSLEpatients [1,2,6,21,22,66].Furthermore ,ourstudyshowsthatTHP-1 cellstreatedwithIFN 2orIFN displayupto18-foldincreaseofSTAT1,25-foldincreaseofCCL2,and700-fold increaseofCXCL10,confirmingthatthesegenesrespond toIFN-Istimulation. Tang etal .reportedmiR-146aunder-expressioninSLE PBMCs[46],whereaswedidnotobserveadecreaseora differencebetweenpatientswithactiveorinactiveSLEfor miR-146aexpressioninperiphe ralbloodleukocytesofSLE patientsinourcohort.Luo etal .[67]hypothesizethata functionalvariantinthemiR-146apromotermayberesponsiblefordecreasedlevelsofmiR-146ainSLE,sothe pri-miR-146alevelsshouldbedecreasedinourpopulation; however,nosignificantdifferencesinpri-miR-146aexpressionwereobservedinourpopulation.Furthermore,Tang etal .reportedinversecorrel ationbetweenmiR-146aand IFNscoreintheirSLEcohort,whilewedidnotobservea significantcorrelationinourcohort.Asignificantincrease inmiR-146awasobservedonlyinSLEpatientswithincreasingIFNscorebetweentheinitialandthesecondvisit [36].Otherpossibleexplanationsforthediscrepancybetweenthetwodatasetscouldbethedifferenceincellpopulationsandracialcompositioninourcohortversustheone examinedbyTang etal .andLuo etal .[46,67].Asforthe THP-1monocytecellmodel,IFN-IweaklystimulatedmiR146aexpressioncomparedtoLPS.Alltheseresultssuggest thattheroleofmiR-146ainregulatingIFN-Iinourcohort ofSLEpatientsmayhavebeenlimited.BiomarkerconnectionsPreviousreportshavedemonstratedtheinvolvementof ADARmRNAandCCL2andCXCL10proteininSLE [15-17,22].Inpublishedliterature,ADARmRNAand CCL2,CXCL10proteinlevelsdisplayedapositiveassociationwithIFNscore[11,21,22,68].Similarlyinourcohort, directlycorrelationbetweenIFNscoreandmRNAlevelsof ADAR,CCL2,andCXCL10wasobserved.ThiswasobservednotonlyinSLEbutinHDaswell,potentiallyindicatingthatthesegenesarerespondingnormallytoIFN evenwhenatlevelsaberrantlyelevated.Unlikereportsfrom previousstudies,STAT1didnotcorrelatewellwiththe IFNscoreintheSLEpatientpopulation[69,70].Instead,Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page11of14 http://arthritis-research.com/content/16/1/R20

PAGE 12

patientswithlow-STAT1SLEandHDwithlow-STAT1, theexpressionwasassociatedwithIFNscore.Patients pairedbytwovisitsthatwererankedbyincreasingIFN scoredemonstratestrongco variancewithSTAT1,butthe covariancebetweenIFNscoreandincreasingSTAT1appearedtobeweaker.InpairedSLEpatientvisits,decreasing IFNscoresorSTAT1levelisaccompaniedbyadecreaseof theotherbiomarkerssuggestingthatSTAT1andIFN-I maybedrivingfactors. WhenSLEpatientvisitsaregroupedintohighandlow STAT1,high-STAT1SLEpatientvisitsshowedsignificantly higherlevelsofCCL2andCXCL10.Aftergroupingbyhigh andlowSTAT1,thehigh-STAT1patientvisitsshowedasignificantlyincreasedslopeforCCL2/IFNandCXCL10/IFN scorescomparedtolow-STAT1SLEpatientvisits.ThisenhancedresponsebyCCL2andCXCL10toIFN-IinhighSTAT1patientsmaybedueinparttotheroleofSTAT1in activationofCCL2andCXCL10[71-73].Hence,STAT1 levelsappeartobeenhancingchemokineresponsetoIFN-I. Furthermore,THP-1cellstreatedwithIFN 2,IFN ,or evenLPS,demonstratedthatIFNscore,STAT1,CCL2, CXCL10andmiR-146awereupregulatedinatimedependentmanner.IFN 2orIFN treatmentofTHP-1 cellsshowsthatcellsexpresseddecreasedlevelsofCCL2 andCXCL10shortlyafterreachingtheirpeakexpression, whereasLPStreatmentdisplayedasteadyincreaseof CCL2andCXCL10withalessrapidinductioncompared totheirexpressionafterIFN 2orIFN stimulation.After STAT1peakexpressioninLPS-treatedTHP-1cells,CCL2 andCXCL10expressionrapidlyaccelerated.Onthecontrary,IFN 2andIFN treatmentofTHP-1cellsshows thatCCL2andCXCL10bothstarteddecreasingafter reachingtheirpeakexpression,whereasSTAT1continued toincreasewithIFN 2orIFN stimulation.Theseresults indicatethatCCL2andCXCL10responddifferentlyto TLR4stimulationcomparedtoIFNsignaling.ItalsoindicatesthatCCL2andCXCL10responsetoIFN-Iisrapid butshortcomparedtoTLRsignaling,asIFNscorecorrelateswithgreaterincreaseofCCL2andCXCL10inthe high-STAT1patientsthaninlow-STAT1patients.TheresultsofTLR4stimulationsuggestthatatleastinthehighSTAT1patientpopulationCCL2andCXCL10arebeing drivenbyTLRsignalingratherthanIFN-Idirectlysince IFN-Istimulationcausedarapidincreasefollowedbyan equallyrapiddecreaseofCCL2andCXCL10independent ofSTAT1expression. ItisunclearwhySTAT1waselevatedtosuchhighlevels insomeoftheSLEpatientsandHD.Onepossibilityis fromTLRactivationasseenintheLPSstimulations.AnotherpossibilityisimpairmentintheexpressionofmiR146a,whichisknowntotargetSTAT1[46].Inthepaired SLE-patientvisits,miR-146amightbeincreasedasaresponsetoSTAT1increases,butitisunabletodownregulate STAT1.OnepotentialreasonthatmiR-146aisunableto downregulateSTAT1isduetoalternativesplicing.STAT1 existsasalongform(STAT1a)andshortform(STAT1b). AccordingtothemiRNAtargetpredictionsite,TargetScan. com,STAT1bhasashorter3 UTRcomparedtoSTAT1a 3 UTR.Theshorter3 UTRinSTAT1blacksmiR-146a bindingsites,whichwouldpreventmiR-146adownregulationofSTAT1b.SeveralHDalsodisplayedveryhigh STAT1levels,howeverCCL2andCXCL10,eventhoughelevatedcomparedtolow-STAT1HD,weresignificantly lowerthaninSLEpatients.ApotentialreasonisthatIFN-I drivesCCL2andCXCL10expression,andhighSTAT1 primestheimmunesystemtoamplifyCCL2andCXCL10 expressionwhenIFN-Iisprese nt.WithoutIFN-I,thehigh STAT1levelsmaystillprimetheimmunesystembutthey lacktheignitiontodrivetheprocessforward.ConclusionsTheresultsofthisstudyshowthatSTAT1mRNAexpressioninPBMCsfromlupuspatientsandhealthycontrolsis segregatedintohigh-orlow-STAT1groups.STAT1may beanimportantdriveroflupuspathogenesiswithSTAT1 servingasanexpressionenhancerofCCL2andCXCL10 inpatientswithhighlevelsofSTAT1.AdditionalfileAdditionalfile1:FigureS1. Anti-dsDNAlevel,IFNscore,STAT1,CCL2, andCXCL10inindividualswithdifferentethnicbackgrounds. FigureS2 ComparisonofSLEDAI,anti-dsDNAtiter,IFNscore,STAT1,CCL2,and CXCL10inpatientswithdifferentethnicbackground. FigureS3 .IFN score,CCL2,andCXCL10inindividualswithdifferentethnicbackground andSTAT1levels. FigureS4 .CCL2,CXCL10,andIFNscoreinindividuals withdifferentethnicbackgroundandhighvslowSTAT1groups.dsDNA, double-strandedDNA;STAT,sign altransducersandactivatorsof transcription;CCL2,C-Cmotifchemokineligand2;CXCL10,C-X-Cmotif chemokine10;SLEDAI,systemiclupuserythematosusdiseaseactivityindex. Abbreviations AA: AfricanAmericans;ACR:AmericanCollegeofRheumatology; ADAR:adenosinedeaminaseactingonRNA;ANA:antinuclearantibody; ANCOVA:analysisofcovariance;AsA:AsianAmericans;BSA:bovineserum albumin;CCL2:C-Cmotifchemokineligand2;CXCL10:C-X-Cmotifchemokine 10;dsDNA:double-strandedDNA;dsRNA:double-strandedRNA;EA:European Americans;ELISA:enzyme-linkedimmunosorbentassay;FBS:fetalbovineserum; HD:healthydonors;IFNAR:interferonreceptor;IFN-I:typeIinterferon; IgG:immunoglobulinG;IrA:interracialAmericans;IRF:interferonregulatory factor;LA:LatinAmericans;LPS:lipopolysaccharide;LY6E:lymphocyteantigen6 complexlocusE;miRNA:microRNA;MX1:myxovirusresistance1;OAS1: 2 ,5 -oligoadenylatesynthetase;OD:opticaldensity;PBMC:peripheralblood mononuclearcells;PBS:phosphate-bufferedsaline;PKR:proteinkinaseRNAactivated;SLE:systemiclupuserythematosus;SLEDAI:systemiclupus erythematosusdiseaseactivityindex;STAT:signaltransducersandactivatorsof transcription;TLR:toll-likereceptor;TNF:tumornecrosisfactor;UTR:untranslated region. Competinginterests Theauthorsdeclarethattheyhavenocompetinginterests. Authors ’ contributions PRDGcarriedouttheexperiments.PRDG,MSandEKLCdesignedthestudy. PRDG,AC,andMSperformedthestatisticalanalysis.ESS,AC,andWHR enrolledpatientsforthestudy,collectedinformationandmaintainedtheDominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page12of14 http://arthritis-research.com/content/16/1/R20

PAGE 13

database.PRDG,AC,andEKLCdraftedthemanuscript.Allauthorsreadand approvedthefinalmanuscript. Acknowledgements SupportedinpartbyagrantfromtheLupusResearchInstituteandthe NationalInstitutesofHealthgrantAl47859.PRDGwassupportedbyNIH traininggrantT90/R90DE007200.WethankallthestaffattheDivisionof Rheumatologyforcollectionofbloodandclinicalinformation. Authordetails1DepartmentofOralBiology,UniversityofFlorida,P.O.Box100424,1395 CenterDrive,Gainesville,FL32610-0424,USA.2DivisionofRheumatologyand ClinicalImmunology,DepartmentofMedicine,UniversityofFlorida,P.O.Box 100221,1600SWArcherRd,Gainesville,FL32610-0221,USA.3Schoolof HealthSciences,UniversityofOccupationalandEnvironmentalHealth,Japan, 1-1Isei-ga-oka,Yahata-nishi-ku,Kitakyushu,Fukuoka807-8555,Japan.4Currentaddress:RheumatologyandClinicalImmunology,Humanitas ClinicalandResearchCenter,ViaA.Manzoni56,20089Rozzano,Italy.5Currentaddress:BIOMETRADepartment,UniversityofMilan,Milan,Italy.6Currentaddress:DepartmentofUrology,UniversityofFlorida,1600SW ArcherRoad,Gainesville,FL32610-0247,USA. Received:23April2013Accepted:14January2014 Published:23January2014 References1.BaechlerEC,BatliwallaFM,KarypisG,GaffneyPM,OrtmannWA,EspeKJ, SharkKB,GrandeWJ,HughesKM,KapurV,GregersenPK,BehrensTW: Interferon-induciblegeneexpressionsignatureinperipheralbloodcells ofpatientswithseverelupus. ProcNatlAcadSciUSA 2003, 100: 2610 – 2615. 2.BennettL,PaluckaAK,ArceE,CantrellV,BorvakJ,BanchereauJ,PascualV: Interferonandgranulopoiesissignaturesinsystemiclupus erythematosusblood. JExpMed 2003, 197: 711 – 723. 3.CrowMK: Interferonpathwayactivationinsystemiclupus erythematosus. CurrRheumatolRep 2005, 7: 463 – 468. 4.PrebleOT,BlackRJ,FriedmanRM,KlippelJH,VilcekJ: Systemiclupus erythematosus:presenceinhumanserumofanunusualacid-labile leukocyteinterferon. Science 1982, 216: 429 – 431. 5.NikpourM,DempseyAA,UrowitzMB,GladmanDD,BarnesDA: Association ofageneexpressionprofilefromwholebloodwithdiseaseactivityin systemiclupuserythaematosus. AnnRheumDis 2008, 67: 1069 – 1075. 6.CrowMK,KirouKA,WohlgemuthJ: MicroarrayanalysisofinterferonregulatedgenesinSLE. Autoimmunity 2003, 36: 481 – 490. 7.KirouKA,LeeC,GeorgeS,LoucaK,PapagiannisIG,PetersonMG,LyN, WoodwardRN,FryKE,LauAY,PrenticeJG,WohlgemuthJG,CrowMK: Coordinateoverexpressionofinterferon-alpha-inducedgenesin systemiclupuserythematosus. ArthritisRheum 2004, 50: 3958 – 3967. 8.BassBL,WeintraubH: Anunwindingactivitythatcovalentlymodifiesits double-strandedRNAsubstrate. Cell 1988, 55: 1089 – 1098. 9.WagnerRW,SmithJE,CoopermanBS,NishikuraK: Adouble-strandedRNA unwindingactivityintroducesstructuralalterationsbymeansof adenosinetoinosineconversionsinmammaliancellsandXenopuseggs. ProcNatlAcadSciUSA 1989, 86: 2647 – 2651. 10.NishikuraK: FunctionsandregulationofRNAeditingbyADAR deaminases. AnnuRevBiochem 2010, 79: 321 – 349. 11.GeorgeCX,GanZ,LiuY,SamuelCE: Adenosinedeaminasesactingon RNA,RNAediting,andinterferonaction. JInterferonCytokineRes 2011, 31: 99 –117. 12.NieY,HammondGL,YangJH: Double-strandedRNAdeaminaseADAR1 increaseshostsusceptibilitytovirusinfection. JVirol 2007, 81: 917 – 923. 13.TothAM,LiZ,CattaneoR,SamuelCE: RNA-specificadenosinedeaminase ADAR1suppressesmeaslesvirus-inducedapoptosisandactivationof proteinkinasePKR. JBiolChem 2009, 284: 29350 – 29356. 14.LiZ,WolffKC,SamuelCE: RNAadenosinedeaminaseADAR1deficiency leadstoincreasedactivationofproteinkinasePKRandreduced vesicularstomatitisvirusgrowthfollowinginterferontreatment. Virology 2010, 396: 316 – 322. 15.LaxminarayanaD,KhanIU,KammerG: Transcriptmutationsofthealpha regulatorysubunitofproteinkinaseAandup-regulationoftheRNAeditinggenetranscriptinlupusTlymphocytes. Lancet 2002, 360: 842 – 849. 16.LaxminarayanaD,O ’ RourkeKS,MaasS,OlorenshawI: Alterededitingin RNAeditingadenosinedeaminaseADAR2genetranscriptsofsystemic lupuserythematosusTlymphocytes. Immunology 2007, 121: 359 – 369. 17.OrlowskiRJ,O ’ RourkeKS,OlorenshawI,HawkinsGA,MaasS,LaxminarayanaD: Alterededitingincyclicnucleotidephosphodiesterase8A1genetranscripts ofsystemiclupuserythematosusTlymphocytes. Immunology 2008, 125: 408 – 419. 18.KaronitschT,FeierlE,SteinerCW,DalwigkK,KorbA,BinderN,RappA, SteinerG,ScheineckerC,SmolenJ,AringerM: Activationoftheinterferongammasignalingpathwayinsystemiclupuserythematosusperipheral bloodmononuclearcells. ArthritisRheum 2009, 60: 1463 – 1471. 19.UsachevaA,SmithR,MinshallR,BaidaG,SengS,CrozeE,ColamoniciO: TheWDmotif-containingproteinreceptorforactivatedproteinkinaseC (RACK1)isrequiredforrecruitmentandactivationofsignaltransducer andactivatoroftranscription1throughthetypeIinterferonreceptor. JBiolChem 2001, 276: 22948 – 22953. 20.IvashkivLB: TypeIinterferonmodulationofcellularresponsesto cytokinesandinfectiouspathogens:potentialroleinSLEpathogenesis. Autoimmunity 2003, 36: 473 – 479. 21.BauerJW,BaechlerEC,PetriM,BatliwallaFM,CrawfordD,OrtmannWA, EspeKJ,LiW,PatelDD,GregersenPK,BehrensTW: Elevatedserumlevels ofinterferon-regulatedchemokinesarebiomarkersforactivehuman systemiclupuserythematosus. PLoSMed 2006, 3: e491. 22.BauerJW,PetriM,BatliwallaFM,KoeuthT,WilsonJ,SlatteryC,Panoskaltsis-MortariA, GregersenPK,BehrensTW,BaechlerEC: Interferon-regulatedchemokinesas biomarkersofsystemiclupuserythematosusdiseaseactivity:avalidation study. ArthritisRheum 2009,60: 3098 – 3107. 23.YadavA,SainiV,AroraS: MCP-1:chemoattractantwitharolebeyond immunity:areview. ClinChimActa 2010, 411: 1570 – 1579. 24.LaGrutaNL,KedzierskaK,StambasJ,DohertyPC: Aquestionofselfpreservation:immunopathologyininfluenzavirusinfection. ImmunolCell Biol 2007, 85: 85 – 92. 25.MarscheG,SemlitschM,HammerA,FrankS,WeigleB,DemlingN,SchmidtK, WindischhoferW,WaegG,SattlerW,MalleE: Hypochlorite-modified albumincolocalizeswithRAGEinthearterywallandpromotesMCP-1 expressionviatheRAGE-Erk1/2MAP-kinasepathway. FASEBJ 2007, 21: 1145 – 1152. 26.LloydCM,MintoAW,DorfME,ProudfootA,WellsTN,SalantDJ,GutierrezRamosJC: RANTESandmonocytechemoattractantprotein-1(MCP-1) playanimportantroleintheinflammatoryphaseofcrescenticnephritis, butonlyMCP-1isinvolvedincrescentformationandinterstitialfibrosis. JExpMed 1997, 185: 1371 – 1380. 27.KulkarniO,PawarRD,PurschkeW,EulbergD,SelveN,BuchnerK,NinichukV, SegererS,VielhauerV,KlussmannS,AndersHJ: Spiegelmerinhibitionof CCL2/MCP-1ameliorateslupusnephritisinMRL-(Fas)lprmice. JAmSoc Nephrol 2007, 18: 2350 – 2358. 28.DufourJH,DziejmanM,LiuMT,LeungJH,LaneTE,LusterAD: IFN-gammainducibleprotein10(IP-10;CXCL10)-deficientmicerevealaroleforIP-10in effectorTcellgenerationandtrafficking. JImmunol 2002, 168: 3195 – 3204. 29.LusterAD,JhanwarSC,ChagantiRS,KerseyJH,RavetchJV: Interferoninduciblegenemapstoachromosomalbandassociatedwitha(4;11) translocationinacuteleukemiacells. ProcNatlAcadSciUSA 1987, 84: 2868 – 2871. 30.LoetscherM,LoetscherP,BrassN,MeeseE,MoserB: Lymphocyte-specific chemokinereceptorCXCR3:regulation,chemokinebindingandgene localization. EurJImmunol 1998, 28: 3696 – 3705. 31.WengY,SicilianoSJ,WaldburgerKE,Sirotina-MeisherA,StaruchMJ, DaughertyBL,GouldSL,SpringerMS,DeMartinoJA: Bindingand functionalpropertiesofrecombinantandendogenousCXCR3 chemokinereceptors. JBiolChem 1998, 273: 18288 – 18291. 32.HanC,FuJ,LiuZ,HuangH,LuoL,YinZ: DipyrithioneinhibitsIFNgamma-inducedJAK/STAT1signalingpathwayactivationandIP-10/ CXCL10expressioninRAW264.7cells. InflammRes 2010, 59:809 – 816. 33.RomagnaniP,CrescioliC: CXCL10:acandidatebiomarkerin transplantation. ClinChimActa 2012, 413: 1364 – 1373. 34.LiuM,GuoS,HibbertJM,JainV,SinghN,WilsonNO,StilesJK: CXCL10/IP10ininfectiousdiseasespathogenesisandpotentialtherapeutic implications. CytokineGrowthFactorRev 2011, 22: 121 – 130. 35.NahidMA,SatohM,ChanEKL: MicroRNAinTLRsignalingandendotoxin tolerance. CellMolImmunol 2011, 8: 388 – 403.Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page13of14 http://arthritis-research.com/content/16/1/R20

PAGE 14

36.TaganovKD,BoldinMP,ChangKJ,BaltimoreD: NF-kappaB-dependent inductionofmicroRNAmiR-146,aninhibitortargetedtosignaling proteinsofinnateimmuneresponses. ProcNatlAcadSciUSA 2006, 103: 12481 – 12486. 37.NahidMA,RiveraM,LucasA,ChanEK,KesavaluL: Polymicrobialinfection withperiodontalpathogensspecificallyenhancesmicroRNAmiR-146ain ApoE / miceduringexperimentalperiodontaldisease. InfectImmun 2011, 79: 1597 – 1605. 38.NahidMA,SatohM,ChanEKL: MechanisticroleofmicroRNA-146ain endotoxin-induceddifferentialcross-regulationofTLRsignaling. JImmunol 2011, 186: 1723 – 1734. 39.NahidMA,PauleyKM,SatohM,ChanEKL: miR-146aiscriticalfor endotoxin-inducedtolerance:Implicationininnateimmunity. JBiolChem 2009, 284: 34590 – 34599. 40.PauleyKM,StewartCM,GaunaAE,DupreLC,KuklaniR,ChanAL,PauleyBA, ReevesWH,ChanEK,ChaS: AlteredmiR-146aexpressioninSjogren ’ s syndromeanditsfunctionalroleininnateimmunity. EurJImmunol 2011, 41: 2029 – 2039. 41.SonkolyE,StahleM,PivarcsiA: MicroRNAs:novelregulatorsinskin inflammation. ClinExpDermatol 2008, 33: 312 – 315. 42.SonkolyE,WeiT,JansonPC,SaafA,LundebergL,Tengvall-LinderM,NorstedtG, AleniusH,HomeyB,ScheyniusA,StahleM,PivarcsiA: MicroRNAs:novel regulatorsinvolvedinthep athogenesisofpsoriasis? PLoSOne 2007, 2: e610. 43.NakasaT,MiyakiS,OkuboA,HashimotoM,NishidaK,OchiM,AsaharaH: ExpressionofmicroRNA-146inrheumatoidarthritissynovialtissue. ArthritisRheum 2008, 58: 1284 – 1292. 44.PauleyKM,SatohM,ChanAL,BubbMR,ReevesWH,ChanEKL: UpregulatedmiR-146aexpressioninperipheralbloodmononuclearcells fromrheumatoidarthritispatients. ArthritisResTher 2008, 10: R101. 45.StanczykJ,PedrioliDM,BrentanoF,Sanchez-PernauteO,KollingC,GayRE, DetmarM,GayS,KyburzD: AlteredexpressionofMicroRNAinsynovial fibroblastsandsynovialtissueinrheumatoidarthritis. ArthritisRheum 2008, 58: 1001 – 1009. 46.TangY,LuoX,CuiH,NiX,YuanM,GuoY,HuangX,ZhouH,deVriesN, TakPP,ChenS,ShenN: MicroRNA-146Acontributestoabnormal activationofthetypeIinterferonpathwayinhumanlupusbytargeting thekeysignalingproteins. ArthritisRheum2009, 60: 1065 – 1075. 47.TanEM,CohenAS,FriesJF,MasiAT,McShaneDJ,RothfieldNF,SchallerJG, TalalN,WinchesterRJ: The1982revisedcriteriafortheclassificationof systemiclupuserythematosus. ArthritisRheum 1982, 25: 1271 – 1277. 48.YamasakiY,NarainS,HernandezL,BarkerT,IkedaK,SegalMS,RichardsHB, ChanEK,ReevesWH,SatohM: Autoantibodiesagainstthereplication proteinAcomplexinsystemiclupuserythematosusandother autoimmunediseases. ArthritisResTher 2006, 8: R111. 49.FengX,WuH,GrossmanJM,HanvivadhanakulP,FitzGeraldJD,ParkGS, DongX,ChenW,KimMH,WengHH,FurstDE,GornA,McMahonM,TaylorM, BrahnE,HahnBH,TsaoBP: Associationofincreasedinterferon-inducible geneexpressionwithdiseaseactivityandlupusnephritisinpatients withsystemiclupuserythematosus. ArthritisRheum 2006, 54: 2951 – 2962. 50.BombardierC,GladmanDD,UrowitzMB,CaronD,ChangCH: Derivationof theSLEDAI.Adiseaseactivityindexforlupuspatients.TheCommittee onPrognosisStudiesinSLE. ArthritisRheum 1992, 35: 630 – 640. 51.ToumaZ,GladmanDD,UrowitzMB: CinicalMeausres,Metrics,and indices. In Dubois ’ LupusErythematosusandRelatedSyndromes. 8thedition. EditedbyWallaceD,HahnBH.Philadelphia,PA:ElsevierHealthSciences; 2012:563 – 581. 52.NahidMA,YaoB,Dominguez-GutierrezPR,KesavaluL,SatohM,ChanEKL: RegulationofTLR2-MediatedToleranceandCross-Tolerancethrough IRAK4ModulationbymiR-132andmiR-212. JImmunol 2013, 190: 1250 – 1263. 53.LivakKJ,SchmittgenTD: Analysisofrelativegeneexpressiondatausing real-timequantitativePCRandthe2( DeltaDeltaC(T))Method. Methods 2001, 25: 402 – 408. 54.HanleyJA,NegassaA,EdwardesMD,ForresterJE: Statisticalanalysisof correlateddatausinggeneralizedestimatingequations:anorientation. AmJEpidemiol 2003, 157: 364 – 375. 55.HoA,MagderLS,BarrSG,PetriM: Decreasesinanti-double-strandedDNA levelsareassociatedwithconcurrentflaresinpatientswithsystemic lupuserythematosus. ArthritisRheum 2001, 44: 2342 – 2349. 56.BootsmaH,SpronkP,DerksenR,deBoerG,Wolters-DickeH,HermansJ, LimburgP,Gmelig-MeylingF,KaterL,KallenbergC: Preventionofrelapses insystemiclupuserythematosus. Lancet 1995, 345: 1595– 1599. 57.terBorgEJ,HorstG,HummelEJ,LimburgPC,KallenbergCG: Measurement ofincreasesinanti-double-strandedDNAantibodylevelsasapredictor ofdiseaseexacerbationinsystemiclupuserythematosus.Along-term, prospectivestudy. ArthritisRheum 1990, 33: 634 – 643. 58.IsenbergDA,GartonM,ReichlinMW,ReichlinM: Long-termfollow-upofautoantibodyprofilesinblackfemalelupuspatientsandclinicalcomparison withCaucasianandAsianpatients. BrJRheumatol 1997, 36: 229 – 233. 59.BengtssonAA,SturfeltG,TruedssonL,BlombergJ,AlmG,VallinH, RonnblomL: ActivationoftypeIinterferonsysteminsystemiclupus erythematosuscorrelateswithdiseaseactivitybutnotwithantiretroviral antibodies. Lupus 2000, 9: 664 – 671. 60.ChingKH,BurbeloPD,TiptonC,WeiC,PetriM,SanzI,IadarolaMJ: Two majorautoantibodyclustersinsystemiclupuserythematosus. PLoSOne 2012, 7: e32001. 61.ToCH,PetriM: Isantibodyclusteringpredictiveofclinicalsubsetsand damageinsystemiclupuserythematosus? ArthritisRheum 2005, 52: 4003 – 4010. 62.UmareddyI,TangKF,VasudevanSG,DeviS,HibberdML,GuF: Dengue virusregulatestypeIinterferonsignallinginastrain-dependentmanner inhumancelllines. JGenVirol 2008, 89: 3052 – 3062. 63.PerryST,BuckMD,LadaSM,SchindlerC,ShrestaS: STAT2mediatesinnate immunitytoDenguevirusintheabsenceofSTAT1viathetypeI interferonreceptor. PLoSPathog 2011, 7: e1001297. 64.WalkerGA,ShostakJ: Commonstatisticalmethodsforclinicalresearchwith SASexamples. 3rdedition.Cary,NC:SASInstituteInc.;2010. 65.SainaniKL: Dealingwithnon-normaldata. PMR 2012, 4: 1001 – 1005. 66.QingX,PuttermanC: Geneexpressionprofilinginthestudyofthe pathogenesisofsystemiclupuserythematosus. AutoimmunRev 2004, 3: 505 – 509. 67.LuoX,YangW,YeDQ,CuiH,ZhangY,HirankarnN,QianX,TangY,LauYL, deVriesN,TakPP,TsaoBP,ShenN: AfunctionalvariantinMicroRNA-146a promotermodulatesitsexpressionandconfersdiseaseriskforsystemic lupuserythematosus. PLoSGenet 2011, 7: e1002128. 68.GeorgeCX,DasS,SamuelCE: OrganizationofthemouseRNA-specific adenosinedeaminaseAdar1gene5 -regionanddemonstrationof STAT1-independent,STAT2-dependenttranscriptionalactivationby interferon. Virology 2008, 380: 338 – 343. 69.TassiulasI,HuX,HoH,KashyapY,PaikP,HuY,LowellCA,IvashkivLB: AmplificationofIFN-alpha-inducedSTAT1activationandinflammatory functionbySykandITAM-containingadaptors. NatImmunol 2004, 5: 1181 – 1189. 70.HuX,HerreroC,LiWP,AntonivTT,Falck-PedersenE,KochAE,WoodsJM, HainesGK,IvashkivLB: SensitizationofIFN-gammaJak-STATsignaling duringmacrophageactivation. NatImmunol 2002, 3: 859 – 866. 71.FulkersonPC,ZimmermannN,HassmanLM,FinkelmanFD,RothenbergME: PulmonarychemokineexpressioniscoordinatelyregulatedbySTAT1, STAT6,andIFN-gamma. JImmunol 2004, 173: 7565 – 7574. 72.KokSH,HongCY,KuoMY,WangCC,HouKL,LinYT,GalsonDL,LinSK: OncostatinM-inducedCCL2transcriptioninosteoblasticcellsismediatedbymultiplelevelsofSTAT-1andSTAT-3signaling:animplication forthepathogenesisofarthritis. ArthritisRheum 2009, 60: 1451 – 1462. 73.ValenteAJ,XieJF,AbramovaMA,WenzelUO,AbboudHE,GravesDT: A complexelementregulatesIFN-gamma-stimulatedmonocytechemoattractantprotein-1genetranscription. JImmunol 1998, 161: 3719 – 3728.doi:10.1186/ar4448 Citethisarticleas: Dominguez-Gutierrez etal. : Elevatedsignaltransducers andactivatorsoftranscription1correlateswithincreasedC-Cmotif chemokineligand2andC-X-Cmotifchemokine10levelsinperipheral bloodofpatientswithsystemiclupuserythematosus. ArthritisResearch& Therapy 2014 16 :R20.Dominguez-Gutierrez etal.ArthritisResearch&Therapy 2014, 16 :R20Page14of14 http://arthritis-research.com/content/16/1/R20