Effect of dietary pristane and other saturated mineral oils (MOSH) on autoimmune arthritis in rats

(1)

ContentslistsavailableatScienceDirect

Toxicology Reports

jo u rn al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / t o x r e p

Effect of dietary pristane and other saturated mineral oils (MOSH) on autoimmune arthritis in rats

Monica Andreassen

^a,∗,1

, Hege Hjertholm

^a,1

, Jean-Pierre Cravedi

^b

, Koni Grob

^c

, Jan Alexander

^a

, Unni C. Nygaard

^a

aNorwegianInstituteofPublicHealth,POBox4404,NO-0403Oslo,Norway

bToxalim,INRA,ENVT,INP-EIPurpan,UniversitédeToulouse,F-31027Toulouse,France

cOfﬁcialFoodControlAuthorityoftheCantonofZurich,P.O.Box1471,CH-8032Zurich,Switzerland

a r t i c l e i n f o

Articlehistory:

Received21October2016

Receivedinrevisedform12January2017 Accepted13February2017

Availableonline16February2017

Keywords:

Mineraloilsaturatedhydrocarbons(MOSH) Autoimmunearthritis

DarkAgoutirat

a b s t r a c t

Pristaneandotheradjuvantsbasedonmineraloilsaturatedhydrocarbons(MOSH)mayinduceautoim- munityinrodentsafterintradermalinjection;howeverthereisalackofinformationonimmuneeffects afteroralMOSHexposure.Theaimofourstudywastodeterminetheimpactofdietaryexposureto pristaneandotherMOSHonthedevelopmentofautoimmunearthritis.

DarkAgouti(DA)ratsweregivenfeedcontaining4000mg/kgpristaneorabroadMOSHmixturein variousconcentrations(0–4000mg/kg)for90days,orasingleintradermalinjectionof200␮lpristane (positivecontrol).Arthritisscores,andserumandsplenocytemarkerspreviouslyassociatedwitharthritis development,weredetermined.

Allratsinjectedwithpristanedisplayedarthritissymptomsandhigherlevelsofcertainserummarkers.

NoneoftheratsfedpristaneorMOSHdevelopedarthritissymptomsordemonstratedclearchangesin anymeasuredarthritis-associatedbiologicalmarkersinserumorsplenocytes.

Theabsenceofclinicalarthritissymptomsoranyincreaseincommonarthritis-associatedbiological markersinseraandspleenfollowingdietaryexposuretopristaneorabroadMOSHmixtureinasub- chronicratmodelofarthritissuggestthatdietaryMOSHhavelowcapacitytopromotedevelopmentof autoimmunity.

1. Introduction

Mineral oils are commonly used as adjuvants to boost the immunesystemresponsetoanantigeninhumanandveterinary vaccines[1–3].Apartfromtherapeuticuse,humansareexposedto mineraloilsmainlyviathefood(seebelow),Whileoraladmin- istration of MOSH generally haslow acute toxicity, MOSH has beenreportedtoaccumulateintissuesinbothhumans[4,5]and animals[6].InFischer-344rats,theformationoflivermicrogranu- lomasassociatedwithinﬂammatoryresponseswasobservedafter feedingwithwhitemineraloils[7,6,8].Whilethere isgenerally

Abbreviations: MOH,mineraloilhydrocarbons;MOSH,mineraloilsaturated hydrocarbons;DA,DarkAgouti;i.d.,intradermal;RF,rheumatoidfactor;TLR,toll likereceptor.

∗ Correspondingauthorat:DomainforInfectionControlandEnvironmental Health,DepartmentofToxicologyandRiskAssessment,NorwegianInstituteof PublicHealth,Oslo,Norway.

E-mailaddress:[email protected](M.Andreassen).

1 Theseauthorscontributedequallytothiswork.

littleinformationaboutthepotentialoflongtermdietaryMOSH toaffectimmunefunctions,singleintradermalandintraperitoneal injectionsofcertainmineraloilsinduceautoimmuneresponsesin rodentmodelssharingbothclinicalandpathologicalfeatureswith humanrheumatoidarthritis (RA)[9–11].Afew epidemiological studiessuggestanassociationbetweenexposuretohighdosesof MOHandanincreasedriskofdevelopingautoimmunediseases.Ina casecontrolstudyinSweden,self-reportedoccupationalexposures toMOHprimarilyviaskinandinhalationwereassociatedwithan increasedrelativeriskofdevelopingrheumatoidarthritis(RA)in men[12].Furthermore,asignificantlyhigherprevalenceofRAand Systemiclupuserythematosuswasobservedinapopulationliving closetoanoilfieldwastesitecomparedtoanothercommunitywith noknownexposuresofthistype[13].Inthelatestscientificopinion ofmineraloilsinfood,TheEuropeanFoodSafetyAuthority(EFSA) identifiedaknowledgegaponpotentialeffectsonsystemicautoim- munediseasesoralteredimmunefunctionafterdietaryexposure [14].

Mineraloilhydrocarbons(MOH)mayunintentionallycontami- natethefoodchainatvariousstagesoffoodproductionormigrate

http://dx.doi.org/10.1016/j.toxrep.2017.02.002

(2)

fromfoodpackagingmaterials[15],whilesomeMOHareintention- allyusedasfoodadditivesandinpesticides[14].Oralanddermal exposuretoMOHmayalsooccurthroughuseofcosmetics and pharmaceuticalsandmedicinaluse[16,17].Somesaturatedhydro- carbonsoccurnaturallyinmarineandterrestrialbiota[14].MOH consistofcomplexmixturesofmineraloilaromatichydrocarbons (MOAH)andmineraloilsaturatedhydrocarbons(MOSH).Dueto carcinogenicpropertiesofMOAH,foodgradeMOH-productsare treatedinsuchawaythattheMOAHcontentisminimized,while technicalgradesofMOHtypicallycontain15–35%MOAH.MOSH arepresentatdifferentlevelsinnearlyallfoods,withthehighest concentrationsdetectedinbread,rolls,grains(mainlyrice)con- fectionary(non-chocolate),vegetableoil,cannedﬁshandoilseeds [14].TheestimateddietaryMOSHexposureinEuroperangedfrom 0.03to0.3mg/kgbodyweight(b.w.)perday,andhigherinyounger consumersandchildren,probablyduetoahigherintakeoffoodper kgb.w.aswellasage-relateddifferencesindietaryhabits[14].

One model used for studies on arthritis is the arthritis- susceptibleDarkagouti(DA)rats.Inthismodelasingleintradermal injectionofmedicinalwhiteoil(commonlyusedforfood,pharma- ceuticalandcosmeticuse)aswellascommoncommercialcosmetic productscontainingupto80%mineraloils(likebodylotionand babyoil),inducedarthritis symptoms(joint synovitis).Percuta- neousapplicationoftheseproductsonabradedskinresultedin similar,butmilderandtransient,clinicalarthritissymptomsin5 outof10animalsbeingexposedtoacertainbabyoil[10].Although shortterm(ﬁvesubsequent)peroraldosesofmedicinalwhiteoil didnothaveanyapparenteffectinthisarthritismodel,theauthors speculatethatoraladministrationofadjuvantsinconjunctionwith inﬂammationofthegutbyotheragentscouldparallelthearthritic effectsofpercutaneousexposureonabradedskin.

Asdescribed above,severalMOSHhavetheabilitytoinduce autoimmuneresponsesinrodentsafterdermal/intradermalexpo- sures,butsofar,nostudieshaveinvestigatedwhetherlongterm (sub-chronicorchronic)dietaryexposuretoMOSHcanpromote autoimmunity[14,18].Asingleintradermalpre-administrationof IncompleteFreund’sadjuvantorhexadecanewasabletoprevent developmentofdiseaseinductionbytheintradermalinjectionwith completeFreund’sadjuvant(includingbacterialcomponentsand mineraloils)[19].WhileincompleteFreund’sadjuvanthasbeen reportednot todemonstrateany signiﬁcantimmuneresponses afteroral administrationSilin etal., 2007,neitheroftheabove studiesinvestigatedthepotencytoinduceorpreventautoimmu- nityafteroralexposure.Anotheraspectofconcernisthatprevious safetyevaluationsregardingMOSHhavebeenbasedonchemical andphysicalproperties(suchasviscosity)insteadofsub-classesof mineraloils.Thepresentstudyprovidesnovelexperimentaldata onthepotentialeffectsonautoimmunityofthewholeMOSHrange towhichhumansareexposedtoviathediet.Moreprecisely,we conductedasubchronicstudytodeterminewhether90dayswith dietaryexposurestoabroadMOSH mixturecaninduceclinical signsand/orbiological markersforautoimmunearthritis inthe arthritisproneDAratmodel.

2. Materialsandmethods

2.1. Animals

InbredDAmaleandfemalerats(DA/OlaHsd),7–8weeksofage, wasobtainedfrom HarlandLaboratories (Indianapolis,USA). In thepilotexperiment,thefemaleratswererandomlyassignedto twogroups(pristaneinjectionandcontrol,n=6)andinthemain experiment5femaleand5maleratswererandomlyassignedto oneofsixexperimentalunits(seeexperimentaloverviewinSec- tion3.2.1;Table2).Theratswerehousedtwoorthreeanimalsin

eachcage,andacclimatizedforaminimumof9days.Themakrolon cages,containingcageenrichments,wererandomlyplacedinven- tilatedScantainerﬁltercabinets,12hlight/darkcycle,temperature 21^◦C±2,relativehumidity35–75%±10.Thecagepositioninthe rackwaschangedatleasttwiceaweek.Feedandtapwaterwere givenadlibitum.Theratsweregivenastandard dietuntilstart of the experiment (Teklad2018S, Envigo, Cambridgeshire,UK.) andstandardrodentbedding(NESTPAKS,DatesandLtd).Allrats weremarkedby earpuncturebeforeentering theexperiments.

TheexperimentswereperformedinconformitywithNorwegian lawsandregulationsforliveanimals,andapprovalwasgivenby theNorwegianAnimalResearchAuthorityundertheMinistryof Agriculture(FOTS7084).

2.2. Chemicalsandreagents

Pristane(tetramethylpentadecane,purity≥97%),Concanavalin A(ConA),Lipopolysaccharide(LPS),and10%bufferedformalinwere purchasedfromSigma-AldrichCorp.(St.Louis,MO,USA).AllELISA kitswerepurchasedfromeBioscience(SanDiego,CA,USA),with theexceptionoftheRFkit(fromMyBioSource,SanDiego,CAUSA).

AllCBA kitswere purchased fromBD Bioscience (SanJose, CA, USA).AntibodiesfordetectionofToll-likereceptor(TLR)2andTLR3 werepurchasedfromSantaCruzBiotechnologyInc.(Dallas,Texas 75220USA);anti-TLR2 (H-175,rabbitpolyclonalIgG),anti-TLR3 (N-14,goatpolyclonalIgG),goatanti-rabbitIgG-FITCanddonkey anti-goatIgG-APC.Hank’sBalancedSaltSolution(HBSS),fetalcalf serum(FCS)andthecellgrowthmediumRPMI1640werepur- chasedfromGibco(ThermoFisherScientiﬁc,Waltham,MAUSA), andpenicillin/streptomycinmixwasacquiredfromPAATheCell CultureCompany(GEHealthcareLittleChalfont,Buckinghamshire, UK).

2.3. PreparationoftheMOSH−mixture

AMOSHmixturerangingfromaboutC₁₄toC₅₀waspreparedby combiningthefollowingproducts:295g/kgparaffinwaxPhEur, low viscosity,Fluka76233 (Buchs,Switzerland),295g/kgparaf- finwaxPhEur,highviscosity,Fluka76234,147g/kgCatenexPh 941FU(Shell)and263g/kgdistillatefromparaffinhighlyliquid PhEur,BPNF,MerckJPK432100742091.17174.1000(Darmstadt, Germany).Thisdistillatewasobtainedbydiscardingthefirst25ml from700mlandconsecutivefractionsof25ml.Fractions2and3 weremixedataratioof1:2.

2.4. PreparationofMOSH-andpristanecontainingfeed

Astandardpelleteddietforrats(AIN–93M),asdescribedby Reeves et al. Reeves et al. (1993), was selected. Prior to the preparationofthefeed,themajoringredientswereanalysedby on-lineHPLC-GC-FIDforMOSHorpolyoleﬁnoligomericsaturated hydrocarbons(POSH)toruleoutdisturbinginterferences.Forall ingredients,thecontaminationwasbelow15mg/kgandwascon- sidered acceptable.Beforebeingincorporated intothediet, the MOSHmixtureandpristanewasdissolvedinsoybeanoilandstirred for4hat40^◦C.MOSHwereincorporatedintothedietat40,400 and4000mg/kgandpristaneat4000mg/kg,andineachcasean equivalentmassofsoybeanoilwasreplacedbytheMOSHsolu- tion.DietconcentrationswereveriﬁedinMarch2014andagainin September2014.Themeasureddoseswerewithin88–95%ofthe nominalconcentrations.

2.5. Anaesthetics

Animalsreceivedanaestheticspriortoandduringintradermal injectionsandattermination(3%Isoﬂurangasanaesthetics(Isoba

(3)

vet; Intervet/Schering-Plough Animal Health, Lysaker, Norway), administeredinsurgicalO2).

2.6. Arthritisscoreandethicalconsiderations

Allfourlimbs,toes,footandankle,wereassessedforarthritis symptoms(redness,swellingetc.)andgivenascorefrom0to4for eachlimb,were0isnosymptomsand4areseveresymptoms.Max- imumtotalscorewas16.Iftheratsshowedsignsofseveredistress, theanimalsreceivedpainrelief,0.3mg/mlTemgesic(Buprenorﬁn, RBPharmaceuticals,Berkshire,UK),asubcutaneouslyinjectionof 0.1mleveryeighttotwelvehours.Iftheanimalsgotascoreoverten theywereeuthanizedduetoanimalwelfarereasons.Iftheweight reductionwas>10%perweek,theanimalswerealsoeuthanized.

Theinvestigatorwasblindedwithrespecttodietarytreatments.

Thepristaneinjectedratshadvisibleshavedareasatthebaseof thetail,andconsequentlytheinjectiontreatmentwasnotblinded totheinvestigator.

2.7. Establishingthearthritismodelandthecorresponding immunemarkers(pilot)

Twogroupsof6femalerats,onegroupservingasanegative control,weregivenasingleintradermalinjection,of200␮lphys- iologicalbufferor200␮lpristane.Theratswereobservedtwice aweek,forupto40days,forsymptomsofarthritis.Attermina- tion,serumwascollectedaswellasmesentericlymphnodesand spleens.

2.8. MainexperimentwithMOSHinfeed,pristaneinfeedand pristaneinjected

Theratswererandomlyassigned tooneof6groups, 5male and5femaleineachtreatmentgroup.Thepositivecontrolgroup for arthritis was injected intradermally once with 200␮l pris- taneandgiventhecontrolfeedcontainingnoMOSH.Theother groupsweregiventhedifferentdietsaccordingtoTable2.Therats wereobservedtwiceaweek,for40(thepositivecontrolgroup)or 90days,forsymptomsofarthritis.Allanimalswereweighedonce aweekforthewholeexperimentalperiod,i.e.atotalof40(the positivecontrolgroup)or90days.Ifthescorewereabove10or thebodyweightwasreducedby10%ormoreinaweek,theani- malswereeuthanizedandbloodwascollectedfromtheheartinto tubeswithoutanticoagulant.Abloodsample(100␮l),alsowith- outanticoagulant,wascollectedfromfemoralveinofallanimals onday10,30and60.Attheendoftheexperimenttheanimals wereanesthetizedandbloodwascollectedfromtheheart(with- outanticoagulant),and thespleenand mesentericlymphnodes wereharvested.Fromallbloodsamples,serumwaspreparedand storedinaliquotsat−80^◦Cforlateranalysesofserumrheumatoid factorandcytokines.

2.9. Preparationofasinglecellsuspensionofsplenocytes

Thesplenocyteswereretrievedusinga70␮mcellstrainerand rinsedwithicecoldHBSScontaining2%fetalcalfserumand1%

penicillin/streptomycin toproduce single cell suspension.After centrifugationfor5minat10–15^◦Cand420gthesupernatantwas discardedandthecellsresuspendedin1mlRPMI1640with10%

fetalcalfserumand1%penicillin/streptomycin.Twenty␮lofthe cellsuspensionwasmixedwith10mlIsotonIIsolution.Thenum- berofcellswascountedinaCellcounter(boththeCoulterIsoton IIdiluentandthecellcounter,CoulterZseries,camefromBeck- manCoulter,Inc.LifeScienceDivisionHeadquartersIndianapolis IN,USA).

2.10. Splenocyteanalyses

2.10.1. Tolllikereceptor2and3expression

Additionalbiologicalmarkersofarthritis,i.e.expressionoftoll likereceptor(TLR)2andTLR3onsplenocytes,weredeterminedin themainexperiment.Spleensinglecellsuspensionswereprepared andcountedaspreviouslydescribedandresuspendedinawash buffercontainingPBS,0.1%sodiumazide and1% FCS,for acell concentrationof2×10⁷cells/ml.Onemillioncellsfromeachcell suspensionwereﬁrstincubatedwithBDFcblocker(0.5␮l/well;

BDBiosciences,CA)at4^◦Cforﬁveminutes.Thereafter,cellsfrom eachanimalwasstainedwithamixtureofantibodiesagainstTLR2 (H-175)andTLR3(N-14)for30minat4^◦C,followedbyawashstep andsecondarystainingwithgoatanti-rabbitIgG-FITCanddonkey anti-goatIgG-APC(allfromSantaCruzBiotechnologyInc.,Dallas, USA).Anegativecontrol,onesampleperspleenwasstainedeither onlywithsecondaryantibodiesorwithnoantibodies.Cellswere washedandresuspendedinwashbufferbeforeanalysesonaLSRII ﬂowcytometer(BDBioscience).

2.10.2. Splenocytecytokinesecretion

ThesplenocyteswereresuspendedinRPMI1640with10%fetal calfserumand1%penicillin/streptomycin,10×10⁶cells/ml.Ina 24wellplate,1mlofthesuspensionwastransferredtoeachwell induplicates,and100␮lConA,LPS(5g/ml)ornostimulantwere added.Thecellswerecultivatedfor48handthesupernatantwere collectedafter centrifugation for 10min at870g, and stored at

−80^◦Cuntildetectionofcytokinesbycytometricbeadarray(CBA) orELISA(seebelow).

2.11. ELISA

Serumwerepreparedaccordingtoprotocol,asrecommended bythemanufacturers,forELISAanalysisofserumconcentrations ofrheumatoidfactor(RF;IgGandIgM)andthecytokinesIL-17, TNF␣andIL-1b.Supernatantfromsplenocytesincubatedinculture mediumandstimulatedwithLPSorConAornotstimulatedwere analyzedforthesecretionofIL-17byELISAassuggestedbythe manufacturer.TheELISAplateswerereadonaEL808ELISAplate reader(BIOTEK,HighlandPark,Winoonski,Vermont,USA).

2.12. Cytometricbeadarray(CBA)

Splenocytes stimulated with LPS were analyzed for the cytokinesIL-2 (E5),IFN␥(A6)andTNF␣(C8).Splenocytes stim- ulatedwithConAwereanalyzedforthecytokinesIL-2(E5),IFN␥ (A6),IL-4(B9)andIL-10(A8),usingBDCytometricbeadarray(CBA) Ratsolubleproteinﬂexset,acquiredonaLSRIIﬂowcytometer,and analyzedbytheFCAPArraysoftware(allfromBDBiosciences).

2.13. Statisticalanalysisanddatapresentation

Thestatisticalanalysiswasperformed(SigmaPlot13.0)using MannWhitneytest(pilotstudy)ortwo-wayANOVA,with‘treatment’ and ‘sex’ as thetwo factors. If signiﬁcant for the factor

‘treatment’,subsequentpairwisecomparisonsbetweenthetreat- ment groups wereconducted using theTukeypost hoctest. It shouldbe emphasized that theserum and splenocyte analyses werecarriedoutinmaterialcollectedattermination.Thus,since thepristaneinjectedratswereterminatedearly(≤40days)dueto theseverityoftheirsymptomsandthegroupsexposedviafeed wereterminatedatday90,thelevelscannotbedirectlycompared.

Therefore,thegroupintradermallytreatedwithpristanewasnot includedinthestatisticalanalyses.

(4)

3. Results

3.1. Pilotexperiment

3.1.1. Arthritisincidence,severityanddayofonset

A single intradermal injection of 200␮l pristane on day 0 inducedarthriticsymptomsinallsixfemaleDA-rats(Table1).Four ofthepristane-injectedratswereeuthanizedduringtheexperi- ment(ondays19and22)duetoseveresymptoms,whilethetwo remainingpristane-injectedratsdisplayedclearbutmildersymp- toms(maximumscore7and9)andshowedremissionofsymptoms fromday19and22,respectively.Theseratsweremonitoredforthe full40dayperiod.Noneofthecontrolratsintradermallyinjected

Table1

Arthritisincidence,maximumscoreanddayofonsetinfemaleDarkAgoutirats intradermally(i.d)injectedday0with200␮lpristane(n=6)orphysiologicalsaline buffer(n=6)andmonitoredtwiceaweekfor40days.

Treatment Arthritisincidence Maximumscore^a Dayofonset Pristanei.d. 6/6 7,11,11,10,10,9 12,12,8,8,12,8

Salinei.d. 0/6 0 0

aRatswitharthriticscore>10and/orweightreduction>10%perweekweresac- riﬁcedforanimalwelfarereasons.

withphysiologicalsalinedisplayedarthriticsymptoms(Fig.1a).

Allsaline-injectedratsshowedasteadybodyweightincreasedur- ingtheexperiment, while thebody weightofpristane-injected

Fig.1.Pilotstudy.(a)Arthriticscores,(b)bodyweight(gram;g),andserumlevelsatterminationof(c)IgG-RF,(d)IgM-RF,(e)IL-17(f),IL-1␤and(g)TNF␣infemaleDA-rats intradermallyinjectedonday0with200␮lpristane(n=6)orphysiologicalbuffer(controls,n=6)andmonitoredtwiceaweekforupto40days.Thedotsrepresentthe valuefortheindividualanimalswhilethelinesrepresentthegroupmedianvalue.ThedottedlineindicatesthelowerdetectionlimitfortheELISA-assay.Asterisk(*)denote statisticallysigniﬁcanthigherlevelscomparedtothecontrolgroup(p<0.05,MannWhitneytest).

(5)

Table2

ExperimentaloverviewanddietarymineraloilexposuresinmaleandfemaleDarkAgoutirats,calculatedfromtheMOSH-concentrationsinthefeedandfeedconsumption atweek2and10.

Exposure Route n

(females+males)

Dietarymineraloilexposure mg/kg/b.w./day(median) (females+males)

Duration(days)

Week2 Week10

200␮lPristaneday0 i.d. 5+5 0+0 0+0 40^a

4000mgPristane/kgfeed p.o. 5+5 283.3+355.1 260.0+244.5 90

0mgMOSH/kgfeed p.o. 5+5 0+0 0+0 90

40mgMOSH/kgfeed p.o. 5+5 3.8+3.3 2.9+2.3 90

400mgMOSH/kgfeed p.o. 5+5 30.3+30.8 29.4+25.7 90

4000mgMOSH/kgfeed p.o. 5+5 317.6+309.3 280.1+238.8 90

i.d.,intradermal;p.o.,peroral.

aPristane-injectedratswereincludedintheexperimentforamaximumof40days.Ratswitharthriticscore>10and/orweightreduction>10%perweekweresacriﬁced foranimalwelfarereasons.

ratsdeclinedand/orﬂattenedoutduringthesymptomaticphase (Fig.1b).

3.1.2. Serumanalyses

Attermination,theserumlevelsof IgG-RF (Fig.1c)wassig- niﬁcantlyhigherinratsinjectedwithpristanecomparedtorats injectedwithsaline(MannWhitneytest,p=0.002),whilethepres- enceof IgM-RF in serum (Fig. 1d) was low and did not differ betweenthegroups.Thepristane-injectedratshadsigniﬁcantly higherserumlevelsofIL-17(Fig.1e)thanthecontrolrats(Mann Whitneytest,p<0.05),whileserumlevelsofIL-1␤(Fig.1f)and TNF␣(Fig.1g)didnotdifferbetweenthegroups.IL-10andIL-6 werebelowthelowerassaylimitsofdetectionforallratsirrespec- tiveoftreatment, andonlyone outofsixpristane-injectedrats displayedalow,butdetectablelevelofIL-22(datanotshown).

3.1.3. Mesentericlymphnodes

Inthepilotstudy,alsowetweightsofthemesentericlymph nodesweremeasured,butnodifferenceswereobservedbetween thoseratsreceivingi.d.injectionofpristaneandthecontrols(data notshown).Mesentericlymphnodeswerethereforenotcollected fromratsinthemainexperiment.

3.2. Mainexperiment

3.2.1. Feedconsumptionandbodyweightgain

Feedconsumptionwasmonitoredattwotimepoints,weektwo andweektenof theexperiment. Inweek2and10, therewere nosigniﬁcantdifferencesinfoodconsumptionbetweentreatment groups.Thefeedconsumptionwassigniﬁcantlyhigherinmales thanfemalesatbothtimepoints(twowayANOVA,p=0.011and p=<0.001,respectively;Fig.2a).The MOSHintake viafeedare presentedinTable2;thehighestexposuregroupconsumed317 and 309mg/kg b.w./MOSH perday (group medians,weektwo) forthefemaleandmalerats,respectively.Bothmaleandfemale ratsdisplayedasteadyincreaseinbodyweightthatdidnotdif- ferbetweengroupstreatedorally(Fig.2c).Theratsinjectedwith pristaneshowedadecreasedbodyweightalongwithdiseasedevel- opment(Fig.2b).The two males in the positive control group havingremissionofarthritis symptomsafterpeakingatscore6 or7regainedweightincrease,althoughnotatthesamerateasthe animalsintheothertreatmentsgroups.

3.2.2. Arthritisincidence,severityanddayofonset

Inthepositivecontrolgroupasingleintradermalinjectionof 200␮lpristaneonday0inducedarthriticsymptomsfromday7 inall10rats.Allfemales and3malesdisplayedsigniﬁcantand severesymptoms(maximumscore12)andwerethereforeeuth- anizedduringtheexperiment. Twomalesshowedremission of

symptomsafter18daysandweresacriﬁcedonday40(Fig.3a).

Overall,inthemainexperimenttheratsinjectedwith200␮lpristane,displayedarthritissymptomswithinthesametimespanand degreeofseverityasthoseinthepilotexperiment.

TheratsinthegroupsreceivingpristaneandMOSHinthefeed aswellasthegroupreceivingnotreatmenthadarthritisscoresﬂuc- tuatingbetweenzeroandfourduringthe90daysofobservation.

Themaximumarthritisscoreperanimalduringthe90daysarepre- sentedinFig.3b.Therewerenosigniﬁcantdifferencesbetweenthe treatmentgroups.However,atalldosesofMOSHmaleshadasignif- icanthighermaximumscorethanfemales(p=<0.001;4000mg/kg pristane, p=0.007; 0mg/kg MOSH, p=0.019; 400mg/kg MOSH, p=0.048;40mg/kgMOSHand4000mg/kgMOSH).

3.2.3. Serumanalyses

IgGrheumatoidfactor(IgG-RF)weremeasuredinserumfrom bloodcollectedatday90,oratthedayofterminationinthepris- taneinjectiongroup.Althoughthere wasanoverall statistically signiﬁcantdifferenceduetotreatment(twowayANOVA,p=0.028), therewerenosigniﬁcantdifferencesintheposthoctestbetween thegroupsexposedviafeedorbetweensexeswithinthedifferent groups(Fig.4a).

3.2.4. Splenocytes

3.2.4.1. TLR2andTLR3expression. Neither%TLR2+cellsnorTLR2 expressionper cellshowed anysigniﬁcantdifferences between treatmentgroupsorsexes(Fig.4bandc).Thesameisapplicablefor TLR3expression,whichwaslowforallsamples(datanotshown).

3.2.4.2. Spontaneouscytokinesecretionfromsplenocytes.Inunstim- ulated cells there was anoverall effect of treatment (twoway ANOVA,factor‘treatment’p=0.049)onIL-17secretion,butthere wereno statisticaldifferences between thedifferent treatment groupsintheposthoctest.Therewasalsoanoverallsigniﬁcant differencebetweenmalesandfemalesrelatedtosecretionofIL- 17(twowaysANOVA,factor‘sex’p<0.001)(Fig.5a).LevelsofIL-2, TNF␣orIFN␥measuredinsplenocytesweremostlybelowthelimit oftheassays(datanotshown).

3.2.4.3. CytokinereleasefromLPSstimulatedsplenocytes.LPSstim- ulated cells showedno signiﬁcantdifferencein IL-17 secretion betweentreatmentgroups.TherewasanoveralllowerIL-17secre- tionincellsfrommales(twowaysANOVAp<0.001)comparedto females(Fig.5b).Therewerenosigniﬁcantdifferencesbetween treatments groups inlevels of IL-2,TNF␣or IFN␥ measuredin splenocytes(datanotshown).

(6)

Fig.2.Feedconsumptionandanimalweights.DA-ratswereexposedtoasingleintradermalpristaneinjection(200␮l),feedcontaining4000mg/kgpristane,orfeed containing0(control),40,400and4000mg/kgofabroadMOSHmixture.(a)Feedconsumption(gram;g)weektwoandweek10forthedietaryexposedmale(n=5)and female(n=5)rats.(b)Bodyweightcurve(gram;g)forindividualmale(n=5)andfemale(n=5)DA-ratsintradermallyinjectedday0with200␮lpristaneandweighed weeklyforamaximumof40days.(c)Groupmedianbodyweight(g)formale(n=5)andfemale(n=5)DA-rats,weighedweeklyfor90days.Filledsymbols=females, opensymbols=males.Injection;circularsymbols,pristaneinfeed;diamonds,0mg/kgMOSH;downwardtriangle,40mg/kgMOSH;hexagon,400mg/kgMOSH;square, 4000mg/kgMOSHupwardtriangle.Two-wayANOVAresults(treatmentandsexasthetwofactors)arestatedintheﬁguresifstatisticallysigniﬁcant(p<0.05).

Fig.3.Arthritisscores.(a)ArthriticscoresinindividualDA-ratsintradermallyinjectedwith200␮lpristaneonday0andmonitoredtwiceaweekforupto40days.(b) maximumarthriticscoresperanimalduringthe90days,infemale(closedsymbols)andmale(opensymbols)DA-ratsexposedtoasingleintradermalpristaneinjection (200␮l),feedcontaining4000mg/kgpristane,orfeedcontaining0(control),40,400and4000mg/kgofabroadMOSHmixture.Thesymbolsrepresentthevalueforthe individualanimalswhilethelinesrepresentthegroupmedianvalue.Two-wayANOVAresults(treatmentandsexasthetwofactors)arestatedinthefiguresifstatistically significant(p<0.05),andstatisticallysignificantgroupdifferences(p<0.05)fromthepost-hoctestareindicatedwithbrackets.

Fig.4. RheumatoidfactorandTLRexpression.(a)IgGrheumatoidfactor(IgG-RF)measuredinserumcollectedattermination,(b)TLR2(percent)and(c)TLR2(mean fluorescenceintensity(MFI))measuredinsplenocytes,fromfemale(closedsymbols,n=5)andmale(opensymbols,n=5)DA-ratsexposedtoasingleintradermalpristane injection(200␮l),feedcontaining4000mg/kgpristane,orfeedcontaining0(control),40,400and4000mg/kgofabroadMOSHmixture.Thesymbolsrepresentthevalue fortheindividualanimalswhilethelinesrepresentthegroupmedianvalue.Two-wayANOVAresults(treatmentandsexasthetwofactors)arestatedinthefiguresif statisticallysignificant(p<0.05).

(7)

Fig.5.Cytokinesecretionfromsplenocytes.ConcentrationsofsecretedIL-17ainsupernatantsof(a)unstimulated(b)LPSand(c)ConAstimulatedsplenocytes,and(d)IL-10 insupernatantsofConAstimulatedsplenocytes,fromfemale(closedsymbols,n=5)andmale(opensymbols,n=5)DA-ratsexposedtofeedcontaining4000mg/kgpristane, orfeedcontaining0(control),40,400and4000mg/kgofabroadMOSHmixture.Thesymbolsrepresentthevaluefortheindividualanimalswhilethelinesrepresentthe groupmedianvalue.Two-wayANOVAresults(treatmentandsexasthetwofactors)arestatedintheﬁguresifstatisticallysigniﬁcant(p>0.05).

3.2.4.4. Cytokine release from ConA stimulated splenocytes. In splenocytesstimulatedwithConAtherewasnodifferenceinIL-17a levelsbetweenthedifferenttreatmentgroups(Fig.5c).

Nosigniﬁcant differencesin IL-2 secretionfrom splenocytes stimulatedwithConAwereobservedbetweentreatmentgroups.

There was an overall signiﬁcant difference between male and female(twowaysANOVAp<0.001).Therewerenosigniﬁcantdif- ferencesbetweenthedifferenttreatmentgroupsinIFN␥(p=0.196, datanotshown)or IL-10(p=0.057) secretionfromsplenocytes (Fig.5d).IL-4secretionwasbelowthedetectionlimit(datanot shown).

4. Discussion

Whereassomemineralhydrocarbonsgivenintradermallymay induceautoimmunity,experimentaldataonthepotentialoforal MOSHexposuretocauseautoimmunityiscurrentlylacking[18].

OurstudydemonstratesthatdietaryexposuretoabroadMOSH mixtureor pristanefor 90days neither causesclinical signs of

autoimmune arthritis norinduce biological markers associated withimmuneresponses,inthesusceptible,arthritis-proneDArats.

Inaccordancewithpreviousmurinearthritisstudies[20,10]and thepilotstudyperformedtoestablishthepositivearthritiscontrol, allratsinjectedwithpristanedisplayedseverearthriticsymptoms interms ofredandswollenpaws/anklescorrespondingtohigh arthritisscores.Ratsexposedtodietarypristaneorthreedoselev- elsofMOSH,ontheotherhand,didnot showsigns ofarthritis andelevatedarthritisscorescompared withratsfedthecontrol feed(0mg/kgMOSH).Thearthritisscoreisasubjectivemeasure- ment,basedonavisualinspectionandgradingofthefourlimbs ofeachanimal.Precautionsweretakentoreducethesubjective impactontheresults,byallowingthesametechniciantoperform thescoringthroughouttheexperimentandblindingthetechnician withregardtoexposuregroups.Inspiteofsuchprecautions,some subjectivitywillinﬂuencethebackgroundlevelofscores.Notably, malesingeneralhadsigniﬁcantlyhighermaximumarthritisscores thanfemales.We presumethatthisgenderdifferencemightbe explainedbyanobservationbias,asthemaleratshavelargerpaws

(8)

thanfemales.Sincethescoredidnotdifferfromthecontrolgroup (0mg/kgMOSH)andnoneoftheotherarthritisparameterswere elevatedformalerats,there isnoevidencesuggestingthatthe slightly higherscoresreportedfor maleratswasrelated tothe higherintakeofMOSH-containingfeedcomparedtofemalerats.

Inadditiontothesubjectivearthritisscore,weincludedmea- surementsofaselectionofobjectivebiologicalmarkersassociated withimmunereactions.Inmurinemodels,parenteralexposures tocertainmineraloilsmayinduceautoimmunesymptomsassoci- atedwithrheumatoidarthritisandsystemiclupuserythematous (SLE)[21,10,11]andthesemodelshavebeenemployedtostudy thepathogenesisofautoimmunediseases.Inadditiontoclinical signsintermsofaffected/swollenlimbs/joints,severalbiological markerssuggestedtobeinvolved inboththedevelopmentand themaintenanceofautoimmunediseases,havebeeninvestigated.

Increasedsystemic(serum)levelsofrheumatoidfactor(RF)[11],IL- 17[22,23],IL-6,TNF␣[23]andIL-1␤[9,23]havebeenreportedin animalmodelsofarthritis.Furthermore,ithasbeenreportedthat tolllikereceptor3(TLR3)up-regulationinsplenicmacrophages participateinboththeinitiationandthemaintenanceofpristane- inducedarthritisinDArats[24,25]andalsotheproportionofTLR2⁺ dendriticcellsandmacrophages wereconsiderablyincreasedin thespleen[26].Moreover,murinearthritismodelshavefrequently beenappliedtoinvestigatetherapeuticeffectsofawiderangeof treatments.Decreasedseverityofarthritissymptomsinratshave beenassociatedwithreducedsplenocytesecretionofcytokinesIL- 6[27],TNF␣[28,29,27],IFN␥[28,27],IL-1␤[28,29]andaugmented levelsofIL-10[29].Ingeneral,majorlimitationsofpreviousstudies weretheapplicationofmineraloilinonehighdosewiththelack ofdose-responsestudies,andfurthermorethatonlyinjectionsand percutaneousrouteshavebeenemployed[18].Itisthereforenot knowniftheresultsfromthesestudieshaverelevanceforchronic exposuretomineraloilsviatheoralroute.Consequently,thebio- logicalmarkers(aswellasclinicalsymptoms/signs)wereselected due totheirpreviousassociationwitharthritis developmentin murinemodelsandincludedinthepresentstudytoinvestigate potentialeffectsofdietaryMOSHexposureinDA-rats.

Itshouldbeemphasizedthattheserumanalyseswerecarried outinseracollectedattermination.Thus,sincethepositivecon- trolratswereterminatedearly(≤40days)duetotheseverityof theirsymptomsandthefeedtreatmentsgroupswereterminated at day90, theserum levels cannot be directly compared. Fur- thermore,asalloftheobjectivemarkersexceptIL-17(discussed below)weremeasuredattermination,earlyandmild effectsof thedietaryMOSHmayhavebeenoverlooked.[10]demonstrated that repeated percutaneousapplication of baby oilon abraded ratskingave earlyandmild arthritissymptomslastingforonly 4-6days;howevertheseresultswerebasedsolelyonsubjective arthritisscores,asnobiologicalmarkerswerereported.Inthepilot experiment,pristaneinjectedratshad,attermination, asigniﬁ- canthigherlevelofserumIL-17thanthenegativecontrolanimals, whileotherarthritis-associatedcytokineswerenotaffected(IL-1␤, TNF␣).Therefore,IL-17waschosenasthemarkertomeasureinthe seracollectedatseveraltimepointsduringthemainexperiment (day10,30,60and90).SerumIL-17,however,wasonlydetectable inthe90-dayserumsamples,showingnodifferencesbetweenthe feedtreatmentgroups.Astheonlyserummarkermeasuredduring thewholetimecourseofthestudy,theresultsupportstheobser- vationthatMOSHandpristaneinfeeddidnotinduceanyarthritis symptomsintheseratsatanytimeperiodduringthe90days.In agreement,alsoserumlevelsofIL-1␤andTNF␣atday90didnot differbetweenthegroups.

Patientswithrheumatoidarthritis(RA)haveincreasedtitersof rheumatoidfactor(RF)insera.OverproductionofRFmayberelated toapathogeniceffectobservedinarthritis,althoughthemecha- nismshavenotbeenrevealed[11].RFisalsospeculatedtobea

predisposingfactorforthedevelopmentofRAasincreasedlev- elsofRFinsymptomfreeindividualshavebeenassociatedwith anincreasedriskofRA[30].Inaccordancewiththehumandata, increasedtitersofbothRF-IgMandRF-IgGwasobservedduring thedevelopmentofpristane-inducedarthritisinanmurinemodel usingDArats,howeveronlyRFIgMwassigniﬁcantlyelevated[11].

Inthepresentstudy,anincreasedserumlevelofRF-IgGbutnot RF-IgMwasdetectedinthepositivecontrol(atday40orearlier)in thepilotexperiment.Inthemainexperiment,inagreementwith theotherendpoints,theserumlevelsofRF-IgGwerenotaffected bydietaryMOSHorpristaneexposures.Inthepresentstudy,the RF-IgGlevelsinpristaneinjectedratsdidnotdifferfromthelev- elsintheratsexposedviafeedfor90days,butasdiscussedabove, thismaybearesultofRFanalysisperformedinseracollectedat differentagesandstagesofthedisease.

TofurtherinvestigatethepossibleinfluenceofMOSHingestion on immune modulation or function; splenocytes were investi- gated withregard tosurface markers and cytokine production asindicated bythe previouslypublishedliterature.Splenocytes werecultivatedinmedium,orinmediumwithLPSorConAto stimulateBorTcells,respectively.However,regardlessofstimu- lation,weobservednosignificantdifferencesincytokinesecretion betweenthefeedtreatmentgroups.Furthermore,noeffects on TLR2expressionfromsplenocytes(expressedas%TLR2+cellsand TLR2expressionperpositivecell)wereobservedafterMOSHexpo- sureinfeed.AlsoTLR3expressiondidnotdifferbetweengroups, butthelevelswerelowcompared towhathaspreviously been reportedinDArats.Thismaybeduetolackofanamplifyingstep inourdetectionprocedureaspreviouslydescribed[24].Thegen- erallackofsplenocyteeffectsinourstudyisinagreementwith ourrecentobservationsthatdietaryMOSHexposuresfor120days didnotinducesignificantimmunosuppressiveeffectsinFisher-433 rats(Nygaardetal.,inpreparation).Inthatstudy,weusedthesame feedpreparationsandthusthesamedosesofthebroadMOSHmix- tureasinthepresentstudy,andanotherarmofimmunefunction wasassessedbydetectionofKLH-specificIgMantibodies5days afteraKLHinjection.

AdditionofMOSHorpristanetothefeedapparentlydidnot affectthefeedintakeasweobservednodifferencesinconsump- tionoffeedbetweenthetreatmentgroupsregardlessofdiet;thus itispresumedthatamountofMOSHorpristaneconsumedispro- portionaltotheconcentrationsinthefeed.Thepresentdoselevels thusareabout10,100and1000timeshigherthanthemeanhuman exposurefromfood,asestimatedbyEFSA[14].

Toourbestknowledge,nocomparativepapersexistonthetox- icokineticsofpristaneadministeredviaintradermalororalroutes;

however wecould speculatethat intradermal injectionofpris- tanewouldresultinsubstantialdifferencescomparedwithdietary exposure.Bycomparingradioactivitylevelsintissueofratsdosed either intraperitoneally or per oswith the same quantity of a tritiated MOSH mixture, Ebert et al. [31] foundthat 24h after administration, theconcentrationsmeasured inliver andfat of injectedanimalswereapproximately20and1000higher,respectively,than thoseobservedin ratsreceivingthemixtureorally.

Similardifferencescouldoccurbetweenintradermallyanddietary exposedanimals,resultinginhigherpristaneconcentrationsintar- gettissuesinvolvedintheautoimmunearthritisprocess.

5. Conclusion

Thepresentstudyistoourbestknowledgetheﬁrstsub-chronic oralstudytoinvestigatethepotentialofdietarymineraloilsto induceautoimmunity,usingasusceptibleexperimentalratmodel, and several doselevels of a broad MOSH mixture relevant for humandietaryexposure.Theabsenceofclinicalarthritissymp-

(9)

tomsaswellaslackofchanges incommonarthritis-associated biologicalmarkersinseraandspleenisinsupportofthenotion thatMOSHexposureviathedietaryroutehavealowcapacityto promotedevelopmentofautoimmunity,inthiscasearthritis.

Acknowledgements

This research was part of a grant project co-funded by the European Food Safety Authority (EFSA), grant agreement GP/EFSA/BIOCONTAM/2013/01(“Bioaccumulationandtoxicityof mineraloilhydrocarbons in rats − speciﬁcity of different sub- classesofabroadmixturerelevantforhumandietaryexposures”).

Theconclusions,findings,and opinionsexpressedin this scien- tificmanuscriptreflectonlytheviewoftheauthorsandnotthe officialpositionoftheEuropeanFoodSafetyAuthority.Wethank FlorenceBlas-Y-Estrada,XavierBlanc(INRA),AstriGrestad,Tone Rasmussen,TrudeK.OlsenandKariG.Løken(NIPH)forexcellent technicalassistance.

References

[1]J.Aucouturier,L.Dupuis,V.Ganne,Adjuvantsdesignedforveterinaryand humanvaccines,Vaccine19(17–19)(2001)2666–2672.

[2]J.C.Cox,A.R.Coulter,Adjuvants–aclassiﬁcationandreviewoftheirmodesof action,Vaccine15(3)(1997)248–256.

[3]G.DelGiudice,R.Rappuoli,Inactivatedandadjuvantedinﬂuenzavaccines, Curr.Top.Microbiol.Immunol.386(2015)151–180,http://dx.doi.org/10.

1007/822014406.

[4]L.Barp,C.Kornauth,T.Wuerger,M.Rudas,M.Biedermann,A.Reiner,K.Grob, Mineraloilinhumantissues,partI:concentrationsandmolecularmass distributions,FoodChem.Toxicol.72(2014)312–321,http://dx.doi.org/10.

1016/j.fct.2014.04.029.

[5]M.Biedermann,L.Barp,C.Kornauth,T.Wurger,M.Rudas,A.Reiner,K.Grob, Mineraloilinhumantissues,partII:characterizationoftheaccumulated hydrocarbonsbycomprehensivetwo-dimensionalgaschromatography,Sci.

TotalEnviron.506–507(2015)644–655,http://dx.doi.org/10.1016/j.

scitotenv.2014.07.038.

[6]J.M.Firriolo,C.F.Morris,G.W.Trimmer,L.D.Twitty,J.H.Smith,J.J.Freeman, Comparative90-dayfeedingstudywithlow-viscositywhitemineraloilin Fischer-344andSprague-Dawley-derivedCRL:CDrats,Toxicol.Pathol.23(1) (1995)26–33.

[7]L.Barp,M.Biedermann,K.Grob,Y.E.F.Blas,U.C.Nygaard,J.Alexander,J.P.

Cravedi,Accumulationofmineraloilsaturatedhydrocarbons(MOSH)in femaleFischer344rats:comparisonwithhumandataandconsequencesfor riskassessment,Sci.TotalEnviron.575(2017)1263–1278,http://dx.doi.org/

10.1016/j.scitotenv.2016.09.203.

[8]G.W.Trimmer,J.J.Freeman,R.A.Priston,J.Urbanus,Resultsofchronicdietary toxicitystudiesofhighviscosity(P70HandP100H)whitemineraloilsin Fischer344rats,Toxicol.Pathol.32(4)(2004)439–447,http://dx.doi.org/10.

1080/01926230490465865.

[9]S.Herman,A.Kny,C.Schorn,J.Pfatschbacher,B.Niederreiter,M.Herrmann, M.H.Hoffmann,Celldeathandcytokineproductioninducedby

autoimmunogenichydrocarbonoils,Autoimmunity45(8)(2012)602–611, http://dx.doi.org/10.3109/08916934.2012.719948.

[10]B.Sverdrup,L.Klareskog,S.Kleinau,Commoncommercialcosmeticproducts inducearthritisintheDArat,Environ.HealthPerspect.106(1)(1998)27–32.

[11]P.Wernhoff,P.Olofsson,R.Holmdahl,Thegeneticcontrolofrheumatoid factorproductioninaratmodelofrheumatoidarthritis,ArthritisRheum.48 (12)(2003)3584–3596,http://dx.doi.org/10.1002/art.11342.

[12]B.Sverdrup,H.Kallberg,C.Bengtsson,I.Lundberg,L.Padyukov,L.Alfredsson, L.Klareskog,Associationbetweenoccupationalexposuretomineraloiland rheumatoidarthritis:resultsfromtheSwedishEIRAcase-controlstudy, ArthritisRes.Ther.7(6)(2005)R1296–R1303,http://dx.doi.org/10.1186/

ar1824.

[13]J.Dahlgren,H.Takhar,P.Anderson-Mahoney,J.Kotlerman,J.Tarr,R.

Warshaw,Clusterofsystemiclupuserythematosus(SLE)associatedwithan oilﬁeldwastesite:acrosssectionalstudy,Environ.Health6(2007)8,http://

dx.doi.org/10.1186/1476-069x-6-8.

[14]EFSA(2012).ScientiﬁcOpiniononMineralOilHydrocarbonsinFood.

Retrievedfrom.

[15]K.Grob,M.Biedermann,A.Artho,J.Egli,Foodcontaminationbyhydrocarbons frompackagingmaterialsdeterminedbycoupledLC-GC,Z.Lebensm.Unters.

Forsch.193(3)(1991)213–219.

[16]M.Niederer,T.Stebler,K.Grob,Mineraloilandsynthetichydrocarbonsin cosmeticlipproducts,Int.J.Cosmet.Sci.38(2)(2016)194–200,http://dx.doi.

org/10.1111/ics.12276.

[17]A.Noti,K.Grob,M.Biedermann,U.Deiss,B.J.Bruschweiler,Exposureof babiestoC15-C45mineralparafﬁnsfromhumanmilkandbreastsalves, Regul.Toxicol.Pharmacol.38(3)(2003)317–325.

[18]I.Kimber,J.C.Carrillo,Oralexposuretomineraloils:isthereanassociation withimmuneperturbationandautoimmunity?Toxicology344–346(2016) 19–25,http://dx.doi.org/10.1016/j.tox.2016.01.008.

[19]L.Zhang,M.Y.Mia,C.L.Zheng,M.A.Hossain,F.Yamasaki,O.Tokunaga,O.

Kohashi,ThepreventiveeffectsofincompleteFreund’sadjuvantandother vehiclesonthedevelopmentofadjuvant-inducedarthritisinLewisrats, Immunology98(2)(1999)267–272.

[20]C.B.Schmidt-Weber,D.Pohlers,A.Siegling,H.Schadlich,E.Buchner,H.D.

Volk,R.W.Kinne,Cytokinegeneactivationinsynovialmembrane,regional lymphnodes,andspleenduringthecourseofratadjuvantarthritis,Cell.

Immunol.195(1)(1999)53–65,http://dx.doi.org/10.1006/cimm.1999.1509.

[21]W.H.Reeves,P.Y.Lee,J.S.Weinstein,M.Satoh,L.Lu,Inductionof autoimmunitybypristaneandothernaturallyoccurringhydrocarbons, TrendsImmunol.30(9)(2009)455–464,http://dx.doi.org/10.1016/j.it.2009.

06.003.

[22]B.Astry,E.Harberts,K.D.Moudgil,Acytokine-centricviewofthe pathogenesisandtreatmentofautoimmunearthritis,J.InterferonCytokine Res.31(12)(2011)927–940,http://dx.doi.org/10.1089/jir.2011.0094.

[23]M.Stolina,B.Bolon,S.Middleton,D.Dwyer,H.Brown,D.Duryea,D.Zack,The evolvingsystemicandlocalbiomarkermilieuatdifferentstagesofdisease progressioninratadjuvant-inducedarthritis,J.Clin.Immunol.29(2)(2009) 158–174,http://dx.doi.org/10.1007/s10875-008-9238-8.

[24]L.Meng,W.Zhu,C.Jiang,X.He,W.Hou,F.Zheng,Lu,Toll-likereceptor3 upregulationinmacrophagesparticipatesintheinitiationandmaintenance ofpristane-inducedarthritisinrats,ArthritisRes.Ther.12(3)(2010)R103, http://dx.doi.org/10.1186/ar3034.

[25]W.Zhu,L.Meng,C.Jiang,J.Xu,B.Wang,Y.Han,S.Lu,Overexpressionof toll-likereceptor3inspleenisassociatedwithexperimentalarthritisinrats, Scand.J.Immunol.76(3)(2012)263–270,http://dx.doi.org/10.1111/j.1365- 3083.2012.02724.x.

[26]W.Zhu,L.Meng,C.Jiang,W.Hou,J.Xu,B.Wang,S.Lu,Inductionoftoll-like receptor2positiveantigen-presentingcellsinspleenofpristane-induced arthritisinrats,Mol.Biol.Rep.39(4)(2012)3667–3673,http://dx.doi.org/10.

1007/s11033-011-1141-3.

[27]R.Ulmansky,K.Turjeman,M.Baru,G.Katzavian,M.Harel,A.Sigal,Y.

Barenholz,Glucocorticoidsinnano-liposomesadministeredintravenously andsubcutaneouslytoadjuvantarthritisratsaresuperiortothefreedrugsin suppressingarthritisandinﬂammatorycytokines,J.Control.Release160(2) (2012)299–305,http://dx.doi.org/10.1016/j.jconrel.2011.12.024.

[28]S.FayazAhmad,P.Sultan,A.E.Ashour,T.H.Khan,S.M.Attia,S.A.Bakheet,A.R.

Abd-Allah,ModulationofTh1cytokinesandinﬂammatorymediatorsby Euphorbiahirtainanimalmodelofadjuvant-inducedarthritis,

Inﬂammopharmacology21(5)(2013)365–375,http://dx.doi.org/10.1007/

s10787-012-0161-5.

[29]M.Liu,W.Mao,H.Guan,L.Li,B.Wei,P.Li,Effectsoftaurochenodeoxycholic acidonadjuvantarthritisinrats,Int.Immunopharmacol.11(12)(2011) 2150–2158,http://dx.doi.org/10.1016/j.intimp.2011.09.011.

[30]H.D.Halldorsdottir,T.Jonsson,J.Thorsteinsson,H.Valdimarsson,A prospectivestudyontheincidenceofrheumatoidarthritisamongpeople withpersistentincreaseofrheumatoidfactor,Ann.Rheum.Dis.59(2)(2000) 149–151.

[31]A.G.Ebert,C.R.Schleifer,S.M.Hess,Absorption,disposition,andexcretionof 3H-mineraloilinrats,J.Pharm.Sci.55(9)(1966)923–929.