Chemicaltypecho

INFECTIONANDIMMUNITY,Sept.
2005,p.
5395–5401Vol.
73,No.
90019-9567/05/$08.
000doi:10.
1128/IAI.
73.
9.
5395–5401.
2005Copyright2005,AmericanSocietyforMicrobiology.
AllRightsReserved.
CellSurfaceHeparanSulfatePromotesReplicationofToxoplasmagondiiJosephR.
Bishop,BrettE.
Crawford,andJeffreyD.
Esko*DepartmentofCellularandMolecularMedicine,UniversityofCalifornia,SanDiego,LaJolla,California92093-0687Received15February2005/Returnedformodication31March2005/Accepted15April2005PreviousworksuggeststhatcellsurfaceheparansulfateactsasareceptorfortheApicomplexanparasiteToxoplasmagondii.
UsingChinesehamsterovarycellmutantsdefectiveinheparansulfatebiosynthesis,weshowthatheparansulfateisnecessaryandsufcientforinfectivity.
Further,wedemonstratethattheparasiterequiresNsulfationofheparansulfateinitiatedbyN-deacetylase/N-sulfotransferase-1,but2-Osulfationand6-Osulfationappeartobedispensable.
Inordertostudytheroleofheparansulfateinothercelltypes,wecreatedaconditionalalleleforN-deacetylase/N-sulfotransferase-1byusingCre-loxPtechnology.
MammarytumorcellslackingN-deacetylase/N-sulfotransferase-1exhibitedreducedtoxoplasmainfectivitylikeChinesehamsterovarycellmutants.
Surprisingly,heparin,chemicallymodiedheparinoids,andmonoclonalantibod-iestoheparansulfatehadnoeffectontoxoplasmainfection.
T.
gondiiattachmentandinvasionwereunchangedinN-deacetylase/N-sulfotransferase-1-inactivatedcellsaswell,butreplicationwasreduced.
Thus,heparansulfatedoesnotappeartofunctionasareceptorforT.
gondiibutinsteadfacilitatesparasitereplicationpostinvasion.
HeparansulfateisaglycosaminoglycanpolymerizedfromN-acetylglucosamine(GlcNAc)andglucuronicacid(GlcA)at-tachedtoaproteincoreofaproteoglycan(reviewedinrefer-ences15,24,and27).
Heparansulfateproteoglycanscanoccurastransmembrane,whileglycosylphosphatidylinositol(GPI)-linked,orsecretedproteins.
IntheGolgi,thechainsundergoaseriesofmodicationsofthesugarbackbone,includingNdeacetylation/NsulfationofGlcNAcresidues,epimerizationofD-GlcAtoL-iduronicacid(IdoA),andOsulfation.
Ndeacety-lationandNsulfationofGlcNAcresiduesbytheenzymeN-deacetylase/N-sulfotransferase(NDST)istheinitiatingstepinheparansulfatemodication.
SubsequentsulfationoccursinstretchesnearN-sulfatedglucosamineresidues,resultinginclustersofsulfatedsugarunits.
MammaliancellsexpressfourNDSTs(20).
NDST1andNDST2areubiquitouslyexpressed,whereasNDST3andNDST4areexpressedonlyduringdevel-opmentandintheadultbrain(1,2).
NDST1andNDST2appeartocontributealmostequallytoNsulfationbasedoncompositionstudiesofcellsderivedfrommutantmice(23,43a).
Viruses,bacteria,andparasitesutilizeheparansulfateasreceptorsforinfection,invasion,andcolonization(37,42).
Heparansulfatemayactasacoreceptor,inwhichtheinitialinteractionwithheparansulfatefacilitatesasecondinteractionwithaproteinreceptororglycolipid.
Forexample,herpessimplexvirusinteractswithheparansulfateviaglycoproteinC,followedbybindingtooneormoreHvereceptors(39).
Alter-natively,heparansulfatecouldactasaprimaryreceptor,sincecellscontinuouslyendocytoseheparansulfateproteoglycans(45).
Thus,someparasitesmight"piggy-back"intocellsboundtoheparansulfateproteoglycans.
ToxoplasmagondiiisamemberoftheApicomplexafamilythatincludesPlasmodium,Cryptosporidium,Eimeria,andSar-cocystis(10,40).
Toxoplasma,anobligateintracellularparasitewithacomplexlifecycle,isonethemostwidespreadparasitesinfectinghumansandanimals.
Almostallinfectedindividualsareasymptomatic;however,thediseasecanbeseriousinim-munocompromisedpatientsandininfantsfromcongenitalin-fections(43,44).
ThetachyzoiteformofT.
gondii,arapidlygrowingformoftheparasitefoundinacuteinfections,canefcientlyinvadeandrapidlymultiplyinalmosteverytypeofnucleatedmammaliancell,suggestingthatacommoninvasionmechanismmayexist(40).
Recently,T.
gondiihasbeensuggestedtoutilizeheparansulfateforattachmenttomammaliancells(12,33).
Speci-cally,Chinesehamsterovary(CHO)cellslackingheparansul-fatewere60%reducedforT.
gondiiinfection(33).
Further-more,theCHOcellmutantpgsE-606,decientinNDST1,hadsimilarlyreducedinfectivity,implyingthatsulfationofthechainplayedaroleintoxoplasmaattachment.
Inanotherstudy,solubleglycosaminoglycansexhibitedadose-dependentinhibitionofgliding,suggestingaroleforheparansulfateinglidingmotilityandtoxoplasmamigrationacrossextracellularmatrix(12).
Inthepresentstudy,wehavefurtheranalyzedhowthestructureofheparansulfateaffectsT.
gondiiinfectioninculturedcells,usingamorecompletesetofmutantcelllinesandmammaryepithelialcellsderivedfrommutantmice.
Incontrasttopreviousndings,wereportthattheroleofhepa-ransulfateininfectionisnotrelatedtoattachmentbutrathertoreplicationintheparasitophorousvacuole.
MATERIALSANDMETHODSCelllinesandculture.
Wild-typeCHO,pgsG-224(defectiveinallglycosami-noglycansduetoadeciencyinglucuronosyltransferaseI[GlcATI]),pgsE-606(defectiveinGlcNAcNsulfationduetoamutationinGlcNAcN-deacetylase/N-sulfotransferase[NDST1]),pgsA-677(defectiveinheparansulfateduetoa*Correspondingauthor.
Mailingaddress:9500GilmanDr.
,CMM-East,Rm.
1055,LaJolla,CA92093-0687.
Phone:(858)822-1100.
Fax:(858)534-5611.
E-mail:jesko@ucsd.
edu.
5395mutationinEXT1),pgsF-17(defectiveinuronyl2-Osulfation),andHS-6OST(60%reducedin6-Osulfation)cellswereculturedinHamF-12mediumsup-plementedwith10%fetalbovineserum(FBS;HyClone),penicillinG(100U/ml),andstreptomycinsulfate(100g/ml)(3–5,29,46).
Allcellswereculturedat37°Cinanatmosphereof5%CO2–95%airand100%humidity.
AllcelllineswerefreeofmycoplasmabasedonaPCRtest(Stratagene).
GenerationofNDST1-inactivatedcelllines.
TransgenicC57BL/6micebearingloxPsitesinsertedaroundexon2ofNDST1(NDST1f/f)werecrossedtomicebearingPolyomamiddleTantigendrivenfromthelongterminalrepeatse-quencesinmousemammarytumorvirus(MMTV)toinitiatetumorformationinthemammarygland(22).
Mammarytumorcells(MTCs)werepuriedfromtumorsaccordingtotheprotocolofPullanetal.
(36).
MTCswereculturedbyserialpassageinDulbeccomodiedEaglemedium(DMEM)supplementedwith10%FBS,penicillinG(100U/ml),andstreptomycinsulfate(100g/ml).
InordertoinactivateNDST1,cellswereinfectedwithaCrerecombinase-express-ingadenovirusaccordingtothemethodofLietal.
(28).
E1andE3defectiveadenovirusvectorAd-CMV-Cre(expressingCregenefromhCMVpromoter)wasprovidedbytheUCSDPrograminHumanGeneTherapy.
Cellsweretreatedtwiceover4daysfor90mininmediumwith108PFU/ml.
PCRanalysisveriedthedeletionof99%oftheloxP-ankedexonofNDST1.
Theresultingcellline,MTC-NDST1/,wasculturedinDMEMasdescribedabove.
cDNArescueofGlcATIandNDST1mutants.
RescueofpgsG-224mutantswithGlcATIcDNAwasdescribedpreviously(4).
AnNDST1cDNAexpressionvectorwaspreparedasdescribedbyGrobeandEsko(18).
MTC-NDST1/cellsweretransfectedwithNDST1cDNAusingLipofectaminebystandardmethods(Invitrogen).
CloneswereisolatedanddeterminedtohaveregainedNDST1activitybyrestorationofwild-typelevelsofFGF2binding(4).
Toxoplasmacultureandpreparation.
ToxoplasmastrainRH(ATCC50174)wasculturedbyserialpassageonhumanforeskinbroblastmonolayers(HFF;ATCCCRL-1634)maintainedinDMEMsupplementedwith10%FBS.
TachyzoiteswereisolatedfromHFFsbypassinginfectedcellsthrougha25-gaugeneedle,lteringthelysatesthrougha3-m-pore-sizelter,andcentrifu-gation.
ThecellswerewashedtwiceinDMEMwith1%FBS(invasionmedia)at4°Candusedimmediately.
Toxoplasmainfectivityassay.
ToxoplasmainfectivitywasdeterminedbythemethodofPfefferkornandPfefferkorn(34).
Briey,hostcellswereplatedon24-wellplatesatadensityof105perwell.
Thenextdaythecellswerewashedtwicewithphosphate-bufferedsaline(PBS),and4105tachyzoiteswereaddedin250lofinvasionmedium.
After1h,themonolayerswerewashedthreetimeswithPBSandincubatedat37°CinDMEMwith10%FBS(growthmedium)and2.
5Ciof[5,6-3H]uracilperwell(NEN;30to50Ci[1.
1to1.
8TBq]/mmol).
Attheindicatedtimes,monolayerswerewashedthreetimeswithPBS,lysedwith0.
1%sodiumdodecylsulfate,andcountedforradioactivitybyliquidscintillationspectrometry.
Allassaysweredoneintriplicateandrepeatedmultipletimes.
Chemicalandenzymaticmodicationofheparansulfate.
Toreducepharma-cologicallyNsulfationofheparansulfate,monolayersofMTCsorCHOcellsweregrownin24-wellplatesandtreatedovernightwith30mMsodiumchlorateat37°Cinnormalgrowthmedium.
Thechlorate-containingmediumwasre-moved,andinfectivityassayswereperformedasdescribedabove.
Heparansulfatewasremovedfromlivecellsbydigestiontwicewith1mUeachofheparinlyasesI,II,andIII(Calbiochem)for4hingrowthmedium.
Todeterminetheextentofremovalofheparansulfate,thecellswerexedwith3.
7%formalde-hydeingrowthmediumfor10minatroomtemperatureandstainedwithMAb10E4(1:200)inPBSwith1%FBSfor1hatroomtemperature(13).
AntibodybindingwasdetectedbyimmunouorescencewithAlexa488-labeleddonkeyanti-mouseantibodies(MolecularProbes).
AssayswithmetabolicallylabeledT.
gondii.
TachyzoiteswereculturedonHFFmonolayersinmediumsupplementedwith5Ciof[3H]uracil/mlfor2days.
Parasiteswerepuriedandincubatedonmonolayersofwild-typeCHO,mutantpgsE-606,MTCs,orMTC-NDST1/ata10:1ratioofparasitestocellsforthetimesindicated.
Monolayerswerewashedthreetimeswithice-coldPBStoremoveunassociatedparasites,lysedin0.
1%sodiumdodecylsulfate,andcountedbyliquidscintillation.
Underthesewashingconditions,90%oftheparasitesassociatedwithcellswereintracellular.
ImmunouorescencedetectionofT.
gondiibindingorinvasion.
Tachyzoitebindingorinvasionwasdeterminedbydifferentialantibodystainingbyaslightmodicationofpreviouslydescribedmethods(14).
Rabbitpolyclonalanti-SAG1andmouseMAbtoSAG-1(DG52)werekindlyprovidedbyJohnBoothroyd(StanfordUniversity).
Hostcellsweregrownonglasscoverslipstoconuency(5106cells).
Tachyzoiteswereaddedataratioof4:1parasitestocells.
After30or60min,cellswerewashedthreetimeswithice-coldPBSandxedwith3.
7%formaldehydefor10minatroomtemperature.
AfterwashingandblockinginPBScontaining5%FBS,thecoverslipswerestainedwithrabbitanti-SAG1antibodieslabeledwithAlexa488(green)ata1:2,000dilutionfor1hinPBScontaining1%FBS.
CoverslipswerethenwashedinPBS,andcellswereper-meabilizedbyincubationwithacetonefor5minat20°C.
Afterwashingandblockingasbefore,coverslipswerestainedwithMAbDG52labeledwithAlexa544(red)(1:1,000)inPBScontaining1%FBS.
Finally,thecoverslipswerestainedwithDAPI(4,6-diamidino-2-phenylindole;MolecularProbes),mountedonslides,andviewedbyepiuorescencemicroscopy.
Greenparasiteswerescoredasextracellular,andthosethatwerered,butnotgreenwerescoredasintracellular.
Fewerthan10%ofthetachyzoiteswereextracellular.
Experi-mentsweredoneintriplicateandrepeatedatleasttwice.
Tachyzoitereplicationwasmeasuredafter4and16hbyusingtheimmuno-uorescenceassaydescribedaboveexceptthecellswerepermeabilizedimme-diatelyandonlyMAbDG52labeledwithAlexa488wasusedtostainintracel-lularparasites.
RESULTST.
gondiiinfectsvirtuallyeveryknownavianandmammaliancelltype,suggestingacommonmechanismofinfection.
Inthepresentstudy,wedeneinfectionasathree-stepprocess:(i)adhesionoftheparasitetothehostcell;(ii)invasion,inwhichtheparasitephysicallyentersthecellsandformsaparasito-phorousvacuole;and(iii)replicationoftheparasiteinthevacuole.
PreviouslypublisheddatahassuggestedthatheparansulfatemightbeareceptorforT.
gondiiadhesion(12,33).
Inordertoexaminethespecicityofthisinteractiontoagreaterextent,weanalyzedtheoverallprocessoftachyzoiteinfectioninseveralCHOcellheparansulfate-decientmutantsbyusinganassayinwhich[3H]uracilincorporationintoreplicatingtachyzoiteswasmeasured(Fig.
1).
Thisassaydependsonallthreestepsasdenedabove.
pgsD-677cellslackheparansul-fateduetoamutationinthecopolymeraseEXT1,andpgsG-224cellslackallglycosaminoglycansduetoamutationintheinitiatingenzymeglucuronosyltransferaseI(GlcATI).
Thein-FIG.
1.
HeparansulfateisrequiredforefcientinfectionofCHOcellsbyT.
gondii.
VariousCHOmutantsdefectiveinheparansulfatebiosynthesisweretestedfortheabilitytosupporttachyzoiteinfectionbytheselectiveincorporationof[3H]uracil.
pgsG-224lacksallglycos-aminoglycans,pgsD-677lacksheparansulfate,pgsE-606lacksNDST1andmakesundersulfatedheparansulfate,pgsF-17lacksuronicacid2-Osulfation,andHS-6OSTdemonstrates60%lessGlcNAc6-Osulfation.
pgsG-224RcontainsthecDNAforGlcATIandmakesgly-cosaminoglycansnormally.
Eachbarrepresentstheaverageoftripli-catedeterminationsthestandarddeviationandisrepresentativeofmultipleexperiments.
5396BISHOPETAL.
INFECT.
IMMUN.
corporationof[3H]uracilwaslesseffectiveinbothmutantscomparedtowild-typecells(reduced50%),andrestorationofheparansulfatesynthesisinpgsG-224cellsbytransfectionwithavectorcontainingGlcATIcDNArestoredincorporationtonormallevels(Fig.
1)(4).
Thus,thedecreaseininfectionspecicallydependedonheparansulfate.
Thebiologyofhepa-ransulfatedependsonmodicationsofthecarbohydrateback-bone,includingNdeacetylation/NsulfationofsubsetsofGlcNAcresidues,6-Osulfationofglucosamineresidues,and2-O-sulfationofuronicacids.
Todeterminewhichmodica-tionsmightberequiredfortoxoplasmainfection,wetestedvariousCHOcellmutantsdefectiveinthesulfotransferaseenzymes.
T.
gondii-infectedmutantswithreduced6-Osulfa-tion(HS-6OST)orentirelylacking2-Osulfation(pgsF-17)tothesameextentaswild-typecells(Fig.
1).
InfectionofNDST1-decientcells(pgsE-606)wasreducedtothesameextentasmutantsentirelylackingheparansulfatecomparedtowildtype,supportingpreviousresults(33).
ThesendingssuggestthatT.
gondiirequiresonlytheactionofNDST1forefcientinfection.
Nsulfationisrequiredforinfectivityinepithelialcells.
InordertoexaminewhetherT.
gondiiinfectioninothercelltypesdependsonNDST1,wegeneratedmiceinwhichtheNDST1genecouldbeinactivatedinatissue-specicmannerbyusingCre-loxPrecombination(NDST1f/f)(19).
TheseanimalswerethencrossedtomiceharboringthepolyomamiddleT-antigenunderthecontrolofMMTVtranscriptionalelements.
Mam-maryepithelialtumorcells(i.
e.
,MTCs)wereisolatedfromtumorsarisinginNDST1f/fMMTV-T-antigenmice.
First,NDST1f/fMTCsweretreatedwithheparinlyasesI,II,andIIItoremoveheparansulfate,andlossofheparansulfatewasdeterminedbythemeasuringthedeclineof10E4reactivity,anMAbthatbindstoacommonheparansulfateepitope.
Both10E4reactivity(datanotshown)andT.
gondiiinfectionwerereducedby50%inheparinase-treatedMTCs(Fig.
2).
Phar-macologicalreductionofsulfationbytreatmentwithsodiumchloratealsoblockedT.
gondiiinfection(Fig.
2),asseenpre-viouslyinHFFcells(33).
NDST1wasinactivatedinNDST1f/fMTCsbyinfectionwithadenovirusexpressingCre-recombinase(28).
NDST1-decientcells(MTC-NDST1/)exhibited50%reductioninNsul-fation,avalueconsistentwiththatseeninpgsE-606cellswhichalsolackNDST1(5).
InactivationofNDST1alsoreducedinfectionbyT.
gondiiby50to70%(Fig.
2).
IntroductionofanexpressionvectorcontainingNDST1cDNArestoredinfectionnearlytothatseeninthewildtype.
Similarresultswereob-tainedinhepatocyteslackingNDST1(9),demonstratingthatNsulfationofheparansulfatebyNDST1isnecessaryandsufcientforT.
gondiiinfectioninmultiplecelltypes.
HeparinandheparansulfateblockingantibodieshavenoaffectonT.
gondiiinfection.
SincesulfationofheparansulfateappearedtoberequiredforT.
gondiiinfectivity,wehypothe-sizedthatheparansulfateservedasareceptorforparasiteadhesion.
Thus,wepredictedthatattachmentwouldbedis-ruptedbysolubleheparin,ahighlysulfatedformofheparansulfate.
Heparincanblockmanyprotein-heparansulfateinter-actionswhenaddedatlowconcentrations(10g/ml)(8).
However,addingupto1mgofheparin/mldidnotaffectinfectionofCHOorMTCcells,nordidlowerdosesrestoreinfectivitytoNDST1/cells(datanotshown).
Chemicallymodiedformsofheparinlacking2-O-sulfate,6-O-sulfate,orN-sulfategroupsalsodidnotblockinfection(datanotshown).
Heparinpreparationssuchasthesewilldecreaseinvasionofhepatocytesbymalariasporozoites,butonlyunderconditionsthatmimicbloodow(35).
SinceT.
gondiifailedtoinvadecellsunderowconditions,wecouldnottestthispossibility.
Finally,MAb10E4,specictoheparansulfate,didnotinhibittoxo-plasmainfectionofMTCs(datanotshown).
ThisantibodybindstoacommonepitopeonmostheparansulfatechainsandboundwelltoMTCs(13).
LossofNDST1doesnotaffectattachmentandinvasionofT.
gondii.
ThelackofeffectbyheparinandMAb10E4suggestedthatcellsurfaceheparansulfatemightnotaffectadhesionpersebutratheralaterstepininfectionsuchasinvasionorreplication.
Toexamineinvasion,weincubatedmetabolicallylabeledtachyzoiteswithCHOandMTCcellsfor15,30,or60minandmeasuredcountsassociatedwiththecells.
Surpris-ingly,tachyzoiteswereabletoattachandinvadeNDST1/cellsatlevelscomparabletothoseofwild-typecells(Fig.
3).
Toconrmthisresult,weinfectedcellsfor15or30minwithT.
gondiiandthendirectlyvisualizedintracellularparasiteswithantibodiestosurfaceantigen1(SAG1)ofT.
gondii(Table1).
Nodifferenceinthenumberofintracellularparasiteswasobserved0.
25and0.
5hpostinfection.
Toensurethatthemaximumextentofinvasionoccurred,toxoplasmawasaddedtohostcellsfor0.
5handthenremoved,andthecellswereincubatedforanother4h.
Again,nodifferencewasobservedinthenumberofinfectedcells4hpostinvasion.
FIG.
2.
Nsulfationofheparansulfateisrequiredforefcientin-fectionofmousemammarytumorcellsbyT.
gondii.
MTCswerepuriedfromNDST1f/fmice.
Cellsweretreatedwitheither30mMsodiumchlorate(ClO3)toinhibitsulfation;0.
1mUheparinasesI,II,andIIItoremoveheparansulfate;oradenovirusexpressingCre-recombinasetoinactivateNDST1(MTC-NDST1/).
Tachyzoitein-fectionwasmeasuredbytheselectiveincorporationof[3H]uracil.
[3H]uracilincorporationwasrescuedinMTC-NDST1/bystabletransfectionofNDST1cDNA(MTC-NDST1/R).
Eachbarrepre-sentstheaverageoftriplicatedeterminationsthestandarddeviationandisrepresentativeofmultipleexperiments.
VOL.
73,2005HEPARANSULFATEPROMOTESREPLICATIONOFT.
GONDII5397Heparansulfateaffectstherateofreplicationforintracel-lulartachyzoites.
Takentogether,thesedatasuggestedthatheparansulfatedoesnotactasareceptorforT.
gondiioraffectinvasionbutinsteadaffectsparasitesurvivalorreplication.
Totestthisidea,weculturedRHtachyzoitesbyserialpassageonpgsE-606orMTC-NDST1/cells.
Parasitessurvivedthroughmultiplepassagesbutseemedtolysemutantcellsataslowerratethanthatforwild-typecells(72hversus48h,respec-tively).
ThisndingsuggestedthatlackofNDST1mightaffecttherateoftachyzoitereplication.
Sinceincorporationofuracilbyintracellularparasitesreectstheirrateofdivision,wemea-sured[3H]uracilincorporationatvarioustimespostinvasion.
Nosignicantdifferencesin3Hcountswereobservedinwild-typecells,NDST1/CHOcells,andNDST1/MTCs4hpostinvasion(Fig.
4AandB).
However,after8hthelevelofincorporationof[3H]uracilwasreducedinthemutants,result-inginatwofolddifferenceafter16h.
SimilarresultswereFIG.
3.
N-sulfationofheparansulfatedoesnotaffectattachment/invasionofT.
gondii.
[3H]uracil-labeledT.
gondiiwasincubatedwithCHO(v)orpgsE-606cells()forthetimesindicated.
Thecellswerewashedextensivelywithice-coldPBStoremoveunassociatedpara-sites.
Theamountofradioactivityassociatedwiththecellswasmea-suredbyliquidscintillationcounting.
Eachpointrepresentstheaver-ageoftriplicatedeterminationsthestandarddeviationandisrepresentativeofmultipleexperiments.
FIG.
4.
Replicationoftachyzoitesincellswithalteredheparansulfate.
CHOwild-typeandpgsE-606(A),MTCsandMTC-NDST1/(B)orHFF,wild-type,andchloratetreated(C)wereincubatedwithtachyzoitesfor1hat37°C.
Thecellswerewashedextensivelyandincubatedat37°Cinthepresenceof[3H]uracilforthetimesindicated.
Eachpointrepresentstheaverageoftriplicatedeter-minationsthestandarddeviationandisrepresentativeofmultipleexperiments.
TABLE1.
T.
gondiiinvadeN-sulfationmutantsatthesamerateasthewildtypeaCelltypeTime(h)Meanno.
oftachyzoites/eldSDCHOwildtype0.
25720.
51640.
5/4204pgsE-6060.
25820.
51620.
5/4225MTCwildtype0.
25910.
52140.
5/4188MTC-NDST1/0.
25920.
52150.
5/4167aWildtypeandNDST1/CHOorMTCcellsweregrowntoconuencyonglasscoverslipsandincubatedwithRHtachyzoites(4:1;parasites/cell)for15or30minat37°Cininvasionmedia.
Themonolayerswerewashedextensivelywithice-coldPBStoremoveextracellularparasites.
OnesetofcellswaschallengedwithT.
gondiifor0.
5h,washed,andincubatedfor4hat37°C(0.
5/4).
Monolayerswereimmediatelyxedandstainedforintracellularandextracellularparasites(90%oftheparasiteswereintracellular).
Thenumberoftachyzoitespereldwasanaverageoffoureldsfromthreeseparateassays(20magnication).
Thenumberofhostcellspereldwascomparable,andexperimentswithsodiumchlorate-treatedCHOandHFFcellsyieldedsimilarresults.
5398BISHOPETAL.
INFECT.
IMMUN.
obtainedwithHFFwithorwithoutchloratetreatment(Fig.
4C).
TheseresultssuggestedthatT.
gondiimightdividefasterincellswithfullysulfatedheparansulfate.
Tachyzoitesdonotdivideforseveralhoursafterinvasion.
Thus,at4hmutantandwild-typecellshad20intracellularparasites/eld(Table1).
After18h,theaveragenumberofintracellularparasites/eldincreasedduetoreplication,andtheincreasewasgreaterinwild-typecellscomparedtoNDST1/cells(Table2,115versus81,respectively).
Representativeeldsfromwild-typeandknockoutcellsareshowninFig.
5.
Interestingly,thenum-berofparasitophorousvacuoleswasnotaltered,butthenum-berofvacuolescontainingeightparasites(threedivisions,23)versusfourparasites(twodivisions,22)wasgreaterinwild-typecellscomparedtothemutant(Table2).
SimilarresultswereobservedwithMTCseventhoughT.
gondiididnotreplicateaswellinMTCscomparedtoCHOcells(Table2).
DISCUSSIONT.
gondiiinvadesawidevarietyofmammalianspeciesandcelltypes,suggestingacommonmechanismofentry.
Theubiq-uitousnatureofthecellsurfaceglycan,heparansulfate,ledotherstohypothesizeitasareceptorforT.
gondii.
Indeed,severalstudieshavedemonstratedthatalteringheparansulfatealtersinfection(12,33).
Inthisreport,wehavequaliedandextendedthesestudiestoothercelltypesbyusingbothgeneticandpharmacologicalapproachestoalterheparansulfatestruc-ture.
Ourmajorndingsshowthat(i)heparansulfateisbothnecessaryandsufcientforrobustinfectionofseveralcelltypes,(ii)efcientinfectiondependsonNsulfationofthechains,and(iii)heparansulfatefacilitatesreplicationofT.
gondiiintheparasitophorousvacuole.
Importantly,ourdatashowthatheparansulfatedoesnotserveasareceptoronthecellsurfacebutinsteadfacilitatesparasitereplication.
Numerousbiologicalactivitieshavebeenascribedtohepa-ransulfatechains.
Ingeneral,theseactivitiesdependonthearrangementofvariouslymodiedsugarresiduesthatmakeupspecicbindingsitesforproteins.
Oneofthebest-studiedexamplesisthegrowthfactorFGF-2,whichbindstoapen-tasaccharidecontainingN-sulfatedglucosamine(GlcNS)resi-duesandatleastone2-O-sulfatedIdoA.
TheFGFreceptorbindstoasequencerichinGlcNS,6-O-sulfatedglucosamineresidues,and2-O-sulfatedIdoA(21,30).
Asexpected,altering2-Osulfationor6-OsulfationinCHOcellsdiminishesFGF-2bindingandsignaling(4,46).
IncontrastthesechangeshadnoeffectonT.
gondiiinfection,butpartialreductionofNsulfa-tiondiminishedinfectivityby50%.
InstudiesofcellslackingallNsulfationduetocombineddeciencyinNDST1andNDST2,nofurtherdiminutionofT.
gondiiinfectionhasbeenobserved(datanotshown).
TheseresultsshowthatNDST1generatesaspecicpatternofsulfationrequiredforefcientFIG.
5.
T.
gondiidivideslowerincellswithreducedsulfationofheparansulfate.
Cellswereinfectedwithtachyzoitesfor60min,washedextensively,andincubatedfor18hat37°C.
ThecellswerestainedwithMAbDG52(green,tachyzoites)andDAPI(nuclei)show-ingthattoxoplasmareplicatedtoalesserextentincellslackingNDST1.
(A)Wild-typeCHOcells;(B)pgsE-606cells.
TABLE2.
SulfationbyNDST1stimulatesT.
gondiireplicationaCelltypeTotalno.
oftachyzoites/eldSDNo.
ofvacuolesSDcontaining:OnetachyzoiteTwotachyzoitesFourtachyzoitesEighttachyzoitesCHOwildtype115113*2274103pgsE-60681101*339553MTCwildtype40413323231MTC-NDST1/2956251321*aWildtypeandNDST1knockoutCHOorMTCcellsweregrowntoconuencyonglasscoverslipsandincubatedwithRHtachyzoites(4:1;parasites/cell)for60minat37°Cininvasionmedia.
Themonolayerswerewashedextensivelyandincubatedingrowthmediaat37°Cfor18h.
Monolayerswerexedandstainedforintracellularparasites.
Thetotalnumberoftachyzoites/eldwasanaverageofeighteldsfromtwoseparateassays.
Thenumberofvacuolescontainingone,two,four,oreighttachyzoitesisshown.
Thenumberofhostcellspereldwascomparable.
*,Totalnumberin16elds.
VOL.
73,2005HEPARANSULFATEPROMOTESREPLICATIONOFT.
GONDII5399infection.
AdditionalstudiesareneededtodenethenatureoftheoligosaccharidesequenceactiveinT.
gondiiinfection.
T.
gondiiexpressesasurfaceantigen(SAG3)thatbindstoheparansulfate,whichoriginallysuggestedthatitmightactasaligandinvolvedinheparansulfate-dependentattachment(25).
MutantslackingSAG3exhibitedreducedinfectivity,andingconsistentwiththisidea.
However,ourdataindicatethatchangesinheparansulfatestructuredonotaffectattach-mentoftheparasitetohostcells.
Monteiroetal.
havereportedthatterminalsialicacidresiduesfacilitateinfection(31).
Per-hapsSAG3bindstosialylatedglycoconjugates,whichareasabundantonthecellsurfaceasheparansulfate.
Ourdatashowthatcellsurfaceheparansulfatedoesnotactasareceptorforattachmentorinvasion.
Consistentwiththisnding,antibodiestoheparansulfateandexogenousheparindidnotblockattachmentorinfection(33).
ByradiolabelingstudiesandbyimmunouorescenceweshowedthatT.
gondiiboundandinvadedwild-typeandNDST1/cellstothesameextent.
Thus,thedifferenceininfectionrelatestolaterstepsininfection.
Basedonuracilincorporationanddirectcountingofcellsinparasitophorousvacuoles,weconcludedthatparasitesreplicatedslowerincellswithalteredheparansulfate.
Severalpossibilitiesexisttoexplainthisnding.
First,toxoplasmamaytakeupheparansulfateduringinvasionanduseitasanutri-ent.
Thus,thechangeinheparansulfatestructureinthemu-tantscouldalteritsnutritionalvalue.
Inmammaliancellsin-ternalizedheparansulfateisdegradedinlysosomesandthecomponentsugarsandsulfatearesalvaged(32).
ExogenousheparinistakenupbyT.
gondii,possiblybywayofSAG3(11,12,16).
However,thefateofheparininternalizedbytoxo-plasmaisunknown.
Second,heparansulfatecouldbeinvolvedintheuptakeofnutrientsrequiredbytoxoplasma.
Forexam-ple,toxoplasmarequirespolyaminesforpropergrowth,andrecentstudiesindicatethatheparansulfatefacilitatessalvageofextracellularpolyamines(6,7,38).
However,attemptstorestoregrowthbytheadditionofpolyaminestothegrowthmediumwereunsuccessful.
Third,heparansulfateproteogly-cansmightbeinternalizedandplayaprotectiveroleintheparasitophorousvacuole.
Effortstovisualizeheparansulfateinvacuolesbyantibodystaining(10E4MAb)anddeconvolutionmicroscopywereunsuccessful.
However,wecannotexcludethepossibilitythatasmallsubpopulationofheparansulfateproteoglycanswereincorporatedintotheparasitophorousvac-uole.
Finally,heparansulfatecouldbeinvolvedinsignalingeventsrequiredbyT.
gondiiforproliferation.
Erk,SAP/JNK,andp38arephosphorylateduponT.
gondiiattachment(26,41).
Heparansulfatefacilitatessignalingviathesepathways,andalteringitsstructurecouldaffecttheresponsetovariousgrowthfactorsorchangetherestingstateofphosphorylationofthesekinases.
Severalmoleculeshavesurfacedaspotentialreceptorsorcoreceptorsfortoxoplasma,includingglycoconjugatestermi-natinginsialicacid(31).
Furtadoetal.
showedthatT.
gondiiboundtocellsurface1integrinsandtheirligands,lamininandtypeIVcollagen,enhancedattachment(16).
Grimwoodetal.
reportedthatparasitebindingcouldbegreatlyincreasedashostcellsproceededfromG1phasetothemid-Sphase,po-tentiallybyupregulationofacellsurfacereceptor(17).
Con-ceivably,T.
gondiimayexploittheseandotherreceptorsforattachment,invasion,andreplicationdependentonthehostcell.
Althoughheparansulfatedoesnotappeartoserveasareceptor,itclearlyplaysaroleinitslifecycle.
Furtherstudiesareneededtoexaminethisprocessingreaterdetail.
ACKNOWLEDGMENTSWethankDavidSibleyandAntonioBarraganforadviceandhelponthisproject.
WethankRobertRosenbergandLijuanZhangforprovidingthe6-O-sulfotransferase-decientCHOcellsandJohnBoo-throydforanti-SAG1antibodies.
J.
R.
B.
wassupportedbyaRuthL.
KirschsteinNRSAFellowship05292-02.
ThisstudywassupportedbyNationalInstitutesofHealthgrantsHL57345andGM33063toJ.
D.
E.
REFERENCES1.
Aikawa,J.
,andJ.
D.
Esko.
1999.
Molecularcloningandexpressionofathirdmemberoftheheparansulfate/heparinGlcNAcN-deacetylase/N-sulfotrans-ferasefamily.
J.
Biol.
Chem.
274:2690–2695.
2.
Aikawa,J.
,K.
Grobe,M.
Tsujimoto,andJ.
D.
Esko.
2001.
Multipleisozymesofheparansulfate/heparinGlcNAcN-deacetylase/N-sulfotransferase:struc-tureandactivityofthefourthmember,NDST4.
J.
Biol.
Chem.
276:5876–5882.
3.
Bai,X.
M.
,andJ.
D.
Esko.
1996.
Ananimalcellmutantdefectiveinheparansulfatehexuronicacid2-O-sulfation.
J.
Biol.
Chem.
271:17711–17717.
4.
Bai,X.
M.
,G.
Wei,A.
Sinha,andJ.
D.
Esko.
1999.
Chinesehamsterovarycellmutantsdefectiveinglycosaminoglycanassemblyandglucuronosyltrans-feraseI.
J.
Biol.
Chem.
274:13017–13024.
5.
Bame,K.
J.
,andJ.
D.
Esko.
1989.
UndersulfatedheparansulfateinaChinesehamsterovarycellmutantdefectiveinheparansulfateN-sulfotrans-ferase.
J.
Biol.
Chem.
264:8059–8065.
6.
Belting,M.
,L.
Borsig,M.
M.
Fuster,J.
R.
Brown,L.
Persson,L.
.
Frans-son,andJ.
D.
Esko.
2002.
Tumorattenuationbycombinedheparansulfateandpolyaminedepletion.
Proc.
Natl.
Acad.
Sci.
USA99:371–376.
7.
Belting,M.
,S.
Persson,andL.
.
Fransson.
1999.
Proteoglycaninvolvementinpolyamineuptake.
Biochem.
J.
338:317–323.
8.
Berneld,M.
,M.
Go¨tte,P.
W.
Park,O.
Reizes,M.
L.
Fitzgerald,J.
Lince-cum,andM.
Zako.
1999.
Functionsofcellsurfaceheparansulfateproteo-glycans.
Annu.
Rev.
Biochem.
68:729–777.
9.
Bishop,J.
R.
,andJ.
D.
Esko.
2005.
TheelusiveroleofheparansulfateinToxoplasmagondiiinfection.
Mol.
Biochem.
Parasitol.
139:267–269.
10.
Black,M.
W.
,andJ.
C.
Boothroyd.
2000.
LyticcycleofToxoplasmagondii.
Microbiol.
Mol.
Biol.
Rev.
64:607–623.
11.
Botero-Kleiven,S.
,V.
Fernandez,J.
Lindh,A.
Richter-Dahlfors,A.
vonEuler,andM.
Wahlgren.
2001.
Receptor-mediatedendocytosisinanapi-complexanparasite(Toxoplasmagondii).
Exp.
Parasitol.
98:134–144.
12.
Carruthers,V.
B.
,S.
Hkansson,O.
K.
Giddings,andL.
D.
Sibley.
2000.
Toxoplasmagondiiusessulfatedproteoglycansforsubstrateandhostcellattachment.
Infect.
Immun.
68:4005–4011.
13.
David,G.
,X.
M.
Bai,B.
VanderSchueren,J.
J.
Cassiman,andH.
VandenBerghe.
1992.
Developmentalchangesinheparansulfateexpression:insitudetectionwithMAbs.
J.
CellBiol.
119:961–975.
14.
Dobrowolski,J.
M.
,andL.
D.
Sibley.
1996.
Toxoplasmainvasionofmam-maliancellsispoweredbytheactincytoskeletonoftheparasite.
Cell84:933–939.
15.
Esko,J.
D.
,andS.
B.
Selleck.
2002.
Orderoutofchaos:assemblyofligandbindingsitesinheparansulfate.
Annu.
Rev.
Biochem.
71:435–471.
16.
Furtado,G.
C.
,Y.
Cao,andK.
A.
Joiner.
1992.
LamininonToxoplasmagondiimediatesparasitebindingtothe1integrinreceptor61onhumanforeskinbroblastsandChinesehamsterovarycells.
Infect.
Immun.
60:4925–4931.
17.
Grimwood,J.
,J.
R.
Mineo,andL.
H.
Kasper.
1996.
AttachmentofToxo-plasmagondiitohostcellsishostcellcycledependent.
Infect.
Immun.
64:4099–4104.
18.
Grobe,K.
,andJ.
D.
Esko.
2002.
RegulatedtranslationofheparansulfateN-acetylglucosamineN-deacetylase/N-sulfotransferaseisozymesbystruc-tured5-untranslatedregionsandinternalribosomeentrysites.
J.
Biol.
Chem.
277:30699–30706.
19.
Grobe,K.
,M.
Inatani,S.
Pallerla,J.
Castagnola,Y.
Yamaguchi,andJ.
D.
Esko.
CerebralhypoplasmaandcraniofacialdefectsinmicelackingheparansulfateNDST1geneexpression.
Development,inpress.
20.
Grobe,K.
,J.
Ledin,M.
Ringvall,K.
Holmborn,E.
Forsberg,J.
D.
Esko,andL.
Kjellen.
2002.
Heparansulfateanddevelopment:differentialrolesoftheN-acetylglucosamineN-deacetylase/N-sulfotransferaseisozymes.
Biochim.
Biophys.
ActaGen.
Subj.
1573:209–215.
21.
Guimond,S.
,M.
Maccarana,B.
B.
Olwin,U.
Lindahl,andA.
C.
Rapraeger.
1993.
ActivatingandinhibitoryheparinsequencesforFGF-2(basicFGF).
DistinctrequirementsforFGF-1,FGF-2,andFGF-4.
J.
Biol.
Chem.
268:23906–23914.
22.
Guy,C.
T.
,R.
D.
Cardiff,andW.
J.
Muller.
1992.
Inductionofmammary5400BISHOPETAL.
INFECT.
IMMUN.
tumorsbyexpressionofpolyomavirusmiddleToncogene:atransgenicmousemodelformetastaticdisease.
Mol.
Cell.
Biol.
12:954–961.
23.
Holmborn,K.
,J.
Ledin,E.
Smeds,I.
Eriksson,M.
Kusche-Gullberg,andL.
Kjellen.
2004.
HeparansulfatesynthesizedbymouseembryonicstemcellsdecientinNDST1andNDST2is6-O-sulfatedbutcontainsnoN-sulfategroups.
J.
Biol.
Chem.
279:42355–42358.
24.
Iozzo,R.
V.
1998.
Matrixproteoglycans:frommoleculardesigntocellularfunction.
Annu.
Rev.
Biochem.
67:609–652.
25.
Jacquet,A.
,L.
Coulon,J.
DeNe`ve,V.
Daminet,M.
Haumont,L.
Garcia,A.
Bollen,M.
Jurado,andR.
Biemans.
2001.
ThesurfaceantigenSAG3me-diatestheattachmentofToxoplasmagondiitocell-surfaceproteoglycans.
Mol.
Biochem.
Parasitol.
116:35–44.
26.
Kim,L.
,B.
A.
Butcher,andE.
Y.
Denkers.
2004.
Toxoplasmagondiiinter-fereswithlipopolysaccharide-inducedmitogen-activatedproteinkinaseac-tivationbyMechanismsdistinctfromendotoxintolerance.
J.
Immunol.
172:3003–3010.
27.
Kjellen,L.
,andU.
Lindahl.
1991.
Proteoglycans:structuresandinteractions.
Annu.
Rev.
Biochem.
60:443–475.
28.
Li,Z.
W.
,G.
Stark,J.
Gotz,T.
Rulicke,M.
Gschwind,G.
Huber,U.
Muller,andC.
Weissmann.
1996.
Generationofmicewitha200-kbamyloidpre-cursorproteingenedeletionbyCrerecombinase-mediatedsite-specicre-combinationinembryonicstemcells.
Proc.
Natl.
Acad.
Sci.
USA93:6158–6162.
29.
Lidholt,K.
,J.
L.
Weinke,C.
S.
Kiser,F.
N.
Lugemwa,K.
J.
Bame,S.
Cheifetz,J.
Massague,U.
Lindahl,andJ.
D.
Esko.
1992.
AsinglemutationaffectsbothN-acetylglucosaminyltransferaseandglucuronosyltransferaseac-tivitiesinaChinesehamsterovarycellmutantdefectiveinheparansulfatebiosynthesis.
Proc.
Natl.
Acad.
Sci.
USA89:2267–2271.
30.
Maccarana,M.
,B.
Casu,andU.
Lindahl.
1993.
Minimalsequenceinhep-arin/heparansulfaterequiredforbindingofbasicbroblastgrowthfactor.
J.
Biol.
Chem.
268:23898–23905.
31.
Monteiro,V.
G.
,C.
P.
Soares,andW.
deSouza.
1998.
HostcellsurfacesialicacidresiduesareinvolvedontheprocessofpenetrationofToxoplasmagondiiintomammaliancells.
FEMSMicrobiol.
Lett.
164:323–327.
32.
Neufeld,E.
F.
,T.
W.
Lim,andL.
J.
Shapiro.
1975.
Inheriteddisordersoflysosomalmetabolism.
Annu.
Rev.
Biochem.
44:357–376.
33.
Ortega-Barria,E.
,andJ.
C.
Boothroyd.
1999.
AToxoplasmalectin-likeactivityspecicforsulfatedinvolvedinhostcellinfection.
J.
Biol.
Chem.
274:1267–1276.
34.
Pfefferkorn,E.
R.
,andL.
C.
Pfefferkorn.
1977.
Speciclabelingofintracel-lularToxoplasmagondiiwithuracil.
J.
Protozool.
24:449–453.
35.
Pinzon-Ortiz,C.
,J.
Friedman,J.
Esko,andP.
Sinnis.
2001.
ThebindingofthecircumsporozoiteproteintocellsurfaceheparansulfateproteoglycansisrequiredforPlasmodiumsporozoiteattachmenttotargetcells.
J.
Biol.
Chem.
276:26784–26791.
36.
Pullan,S.
,J.
Wilson,A.
Metcalfe,G.
M.
Edwards,N.
Goberdhan,J.
Tilly,J.
A.
Hickman,C.
Dive,andC.
H.
Streuli.
1996.
Requirementofbasementmembraneforthesuppressionofprogrammedcelldeathinmammaryepi-thelium.
J.
Cell.
Sci.
109(Pt.
3):631–642.
37.
Rostand,K.
S.
,andJ.
D.
Esko.
1997.
Microbialadherencetoandinvasionthroughproteoglycans.
Infect.
Immun.
65:1–8.
38.
Seabra,S.
H.
,R.
A.
DaMatta,F.
G.
deMello,andW.
deSouza.
2004.
EndogenouspolyaminelevelsinmacrophagesissufcienttosupportgrowthofToxoplasmagondii.
J.
Parasitol.
90:455–460.
39.
Shukla,D.
,andP.
G.
Spear.
2001.
Herpesvirusesandheparansulfate:anintimaterelationshipinaidofviralentry.
J.
Clin.
Investig.
108:503–510.
40.
Sibley,L.
D.
,andN.
W.
Andrews.
2000.
Cellinvasionbyun-palatablepar-asites.
Trafc1:100–106.
41.
Valere,A.
,R.
Garnotel,I.
Villena,M.
Guenounou,J.
M.
Pinon,andD.
Aubert.
2003.
Activationofthecellularmitogen-activatedproteinkinasepathwaysERK,P38andJNKduringToxoplasmagondiiinvasion.
Parasite10:59–64.
42.
Wadstro¨m,T.
,and.
Ljungh.
1999.
Glycosaminoglycan-bindingmicrobialproteinsintissueadhesionandinvasion:keyeventsinmicrobialpathoge-nicity.
J.
Med.
Microbiol.
48:223–233.
43.
Wilson,C.
B.
,J.
S.
Remington,S.
Stagno,andD.
W.
Reynolds.
1980.
Developmentofadversesequelaeinchildrenbornwithsubclinicalcongen-italtoxoplasmainfection.
Pediatrics66:767–774.
43a.
Wang,L.
,M.
M.
Fuster,P.
Sriramarao,andJ.
Esko.
EndothelialdeciencyofheparinsulfateimpairsL-selectinandchemokinemediatedneutrophiltrafckingduringinammatoryresponses.
Nat.
Immunol.
,inpress.
44.
Wong,S.
Y.
,andJ.
S.
Remington.
1993.
BiologyofToxoplasmagondii.
AIDS7:299–316.
45.
Yanagishita,M.
,andV.
Hascall.
1992.
Cellsurfaceheparansulfateproteo-glycans.
J.
Biol.
Chem.
267:9451–9454.
46.
Zhang,L.
J.
,D.
L.
Beeler,R.
Lawrence,M.
Lech,J.
Liu,J.
C.
Davis,Z.
Shriver,R.
Sasisekharan,andR.
D.
Rosenberg.
2001.
6-O-sulfotransferase-1representsacriticalenzymeintheanticoagulantheparansulfatebiosyntheticpathway.
J.
Biol.
Chem.
276:42311–42321.
Editor:J.
B.
BliskaVOL.
73,2005HEPARANSULFATEPROMOTESREPLICATIONOFT.
GONDII5401