<%BANNER%>

UFIR



Induction of Cytoplasmic Rods and Rings Structures by Inhibition of the CTP and GTP Synthetic Pathway in Mammalian Cells
http://www.ncbi.nlm.nih.gov/pubmed/22220215 ( Publisher's URL )
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001289/00001
 Material Information
Title: Induction of Cytoplasmic Rods and Rings Structures by Inhibition of the CTP and GTP Synthetic Pathway in Mammalian Cells
Series Title: PLoS ONE 6:e29690
Physical Description: Journal Article
Creator: Chan, Edward
Carcamo, Wendy C.
Satoh, Minoru
Kasahara, Hideko
Terada, Naohiro
Hamazaki, Takashi
Chan, Jason Y. F.
Yao, Bing
Tamayo, Stephanie
Covini, Giovanni
von Muhlen, Carlos A.
Publisher: PLoS ONE
Place of Publication: PLoS ONE
Publication Date: December 29, 2011
 Subjects
Genre:
Spatial Coverage:
 Notes
Abstract: BACKGROUND: Cytoplasmic filamentous rods and rings (RR) structures were identified using human autoantibodies as probes. In the present study, the formation of these conserved structures in mammalian cells and functions linked to these structures were examined. METHODOLOGY/PRINCIPAL FINDINGS: Distinct cytoplasmic rods (∼3-10 µm in length) and rings (∼2-5 µm in diameter) in HEp-2 cells were initially observed in immunofluorescence using human autoantibodies. Co-localization studies revealed that, although RR had filament-like features, they were not enriched in actin, tubulin, or vimentin, and not associated with centrosomes or other known cytoplasmic structures. Further independent studies revealed that two key enzymes in the nucleotide synthetic pathway cytidine triphosphate synthase 1 (CTPS1) and inosine monophosphate dehydrogenase 2 (IMPDH2) were highly enriched in RR. CTPS1 enzyme inhibitors 6-diazo-5-oxo-L-norleucine and Acivicin as well as the IMPDH2 inhibitor Ribavirin exhibited dose-dependent induction of RR in >95% of cells in all cancer cell lines tested as well as mouse primary cells. RR formation by lower concentration of Ribavirin was enhanced in IMPDH2-knockdown HeLa cells whereas it was inhibited in GFP-IMPDH2 overexpressed HeLa cells. Interestingly, RR were detected readily in untreated mouse embryonic stem cells (>95%); upon retinoic acid differentiation, RR disassembled in these cells but reformed when treated with Acivicin. CONCLUSIONS/SIGNIFICANCE: RR formation represented response to disturbances in the CTP or GTP synthetic pathways in cancer cell lines and mouse primary cells and RR are the convergence physical structures in these pathways. The availability of specific markers for these conserved structures and the ability to induce formation in vitro will allow further investigations in structure and function of RR in many biological systems in health and diseases.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Edward Chan.
Publication Status: Published
Funding: This work was supported in part by a grant from the National Institutes of Health grant Al47859. WCC is supported by Bridges to Doctorate, University of Florida Graduate School, and University of Florida Alumni Graduate Fellowship. ST is supported by the University Scholars Program. Publication of this article was funded in part by the University of Florida Open-Access Publishing Fund. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. No additional external funding was received for this study
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management: All rights reserved by the submitter.
System ID: IR00001289:00001

Downloads
Full Text

PAGE 1

InductionofCytoplasmicRodsandRingsStructuresby InhibitionoftheCTPandGTPSyntheticPathwayin MammalianCellsWendyC.Carcamo1,MinoruSatoh2,3,HidekoKasahara4,NaohiroTerada3,TakashiHamazaki3,JasonY.F. Chan2,BingYao1,StephanieTamayo1,GiovanniCovini5,CarlosA.vonMu hlen6,EdwardK.L.Chan1*1 DepartmentofOralBiology,UniversityofFlorida,Gainesville,Florida,UnitedStatesofAmerica, 2 DepartmentofMedicine,UniversityofFlorida,Gainesville,Florida, UnitedStatesofAmerica, 3 DepartmentofPathology,Immunology,andLaboratoryMedicine,UniversityofFlorida,Gainesville,Florida,UnitedStatesofAmerica, 4 DepartmentofPhysiologyandFunctionalGenomics,UniversityofFlorida,Gainesville,Florida,UnitedStatesofAmerica, 5 DepartmentofGastroenterology,Istituto ClinicoHumanitas,Rozzano,Milan,Italy, 6 RheumaClinicforRheumaticDiseases,PortoAlegre,RioGrandedoSul,BrazilAbstractBackground:Cytoplasmicfilamentousrodsandrings(RR)structureswereidentifiedusinghumanautoantibodiesasprobes. Inthepresentstudy,theformationoftheseconservedstructuresinmammaliancellsandfunctionslinkedtothese structureswereexamined.Methodology/PrincipalFindings:Distinctcytoplasmicrods( 3mminlength)andrings( 2mmindiameter)inHEp2cellswereinitiallyobservedinimmunofluorescenceusinghumanautoantibodies.Co-localizationstudiesrevealedthat, althoughRRhadfilament-likefeatures,theywerenotenrichedinactin,tubulin,orvimentin,andnotassociatedwith centrosomesorotherknowncytoplasmicstructures.Furtherindependentstudiesrevealedthattwokeyenzymesinthe nucleotidesyntheticpathwaycytidinetriphosphatesynthase1(CTPS1)andinosinemonophosphatedehydrogenase2 (IMPDH2)werehighlyenrichedinRR.CTPS1enzymeinhibitors6-diazo-5-oxo-L-norleucineandAcivicinaswellasthe IMPDH2inhibitorRibavirinexhibiteddose-dependentinductionofRRin 95%ofcellsinallcancercelllinestestedaswell asmouseprimarycells.RRformationbylowerconcentrationofRibavirinwasenhancedinIMPDH2-knockdownHeLacells whereasitwasinhibitedinGFP-IMPDH2overexpressedHeLacells.Interestingly,RRweredetectedreadilyinuntreated mouseembryonicstemcells( 95%);uponretinoicaciddifferentiation,RRdisassembledinthesecellsbutreformedwhen treatedwithAcivicin.Conclusions/Significance:RRformationrepresentedresponsetodisturbancesintheCTPorGTPsyntheticpathwaysin cancercelllinesandmouseprimarycellsandRRaretheconvergencephysicalstructuresinthesepathways.Theavailability ofspecificmarkersfortheseconservedstructuresandtheabilitytoinduceformation invitro willallowfurtherinvestigations instructureandfunctionofRRinmanybiologicalsystemsinhealthanddiseases.Citation: CarcamoWC,SatohM,KasaharaH,TeradaN,HamazakiT,etal.(2011)InductionofCytoplasmicRodsandRingsStructuresbyInhibitionoftheCTPand GTPSyntheticPathwayinMammalianCells.PLoSONE6(12):e29690.doi:10.1371/journal.pone.0029690 Editor: WaelEl-Rifai,VanderbiltUniversityMedicalCenter,UnitedStatesofAmerica Received June9,2011; Accepted December2,2011; Published December,2011 Copyright: 2011Carcamoetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Funding: ThisworkwassupportedinpartbyagrantfromtheNationalInstitutesofHealthgrantAl47859.WCCissupportedbyBridgestoDoctorate,University ofFloridaGraduateSchool,andUniversityofFloridaAlumniGraduateFellowship.STissupportedbytheUniversityScholarsProgram.Publicationofthisarticle wasfundedinpartbytheUniversityofFloridaOpen-AccessPublishingFund.Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisionto publish,orpreparationofthemanuscript.Noadditionalexternalfundingwasreceivedforthisstudy. CompetingInterests: Theauthorshavedeclaredthatnocompetinginterestsexist. *E-mail:echan@ufl.eduIntroductionOverthepastfewdecadesmanyhumanautoantibodieshave emergedassignificantdisease-specificmarkersforsystemic rheumaticdiseases[1].Theseautoantibodiesaremainlydirected againstintracellularmacromolecularcomplexesorparticles,such asnucleosomesandsmallnuclear/cytoplasmicribonucleoproteins [1].Thus,humanautoantibodieshavealsoservedasusefulprobes forexploringsubcellularstructuresandfunctionsbecauseoftheir unexpectedspecificitytonovelself-antigens.Examplesofsignificantusesofhumanautoantibodiesinfurthercharacterizationof novelsubcellularstructuresincludedtheidentificationofp80coilininCajalbodies(formerlyknownascoiledbodies)[2]and GW182inGWbodies[3].Withinthepastfewyears,our laboratoriesidentifiednovelhumanautoantibodiesthatrecognizeduniquecytoplasmicstructuresdescribedprovisionallyasrods andrings.ThecurrentstudyreportstheidentificationofCTPS1 andIMPDH2ascomponentsassociatedwiththesemammalian RR. CTPS1andIMPDH2arekeyenzymesinthebiosynthetic pathwayforcytidinetriphosphate(CTP)andguanosinetriphosphate(GTP),respectively.CTPS1catalyzestherate-limitingstep ingeneratingCTPfromuridinetriphosphate.CTPisinvolvedin nucleicacidandphospholipidbiosynthesisandplaysanimportant roleincontrollingcellularproliferation[4].Twoisoforms,CTPS1 andCTPS2,havebeenidentifiedwith74%aminoacidsimilarity. PLoSONE|www.plosone.org 1 December2011|Volume6|Issue12|e2969029

PAGE 2

The67kDaCTPS1isatargetforantiviral,antineoplastic,and antiparasiticdrugdevelopment.Currentlyavailablespecific inhibitorsofCTPS1includeAcivicinand6-diazo-5-oxo-Lnorleucine(DON)[5]. Theoxidationofinosinemonophosphatetoxanthosine monophosphateistherate-limitingstepinthe denovo guanine syntheticpathwaycatalyzedbyIMPDH2,whichisknowntohave aroleinregulatingcellproliferation.Duetoitsrate-limiting property,ithasbeenamajortargetforimmunosuppressive, antiviral,andcancerchemotherapy.TwoisoformsofIMPDH, IMPDH1andIMPDH2,sharing84.0%sequenceidentityand 95.3%similarityhavebeenidentifiedandbothare56kDa proteinsthatfunctionastetramers[6].Astudyontheregulationof IMPDHwithmycophenolicacid(MPA)showedthatMPAbinds toIMPDHandcausesaconformationalchangeresultinginthe formationofinactiveangularaggregates[7].IMPDH2inhibitors suchasMPA,Ribavirin,andTiazofurinarecurrentlyusedfor variousmedicalconditions[8]. Inthecurrentreport,inhibitingeitherCTPS1orIMPDH2 usingdifferentcompoundsdemonstratedinductionofRRina varietyofcelltypes.Moreimportantly,RRwereidentifiedin mouseembryonicstemcells(ESCs)andtheirdisassemblywas observeduponinduceddifferentiation.Results RodsandringsinthecytoplasmAdistinctcytoplasmicpatternwasfirstidentifiedinaroutine anti nuclear antibody(ANA)testusingHEp-2cellslidesfrom INOVADiagnosticsInc.TheseANAslidesaregenerallyusedin clinicallaboratoriesforpatientsdevelopingantibodiestoselfproteins/antigensasadiagnosticscreeningassayforautoimmune diseases.Thestructuresrecognizedbyaprototypehumanserum 604weredistinctcytoplasmicrods( 3mminlength)andrings ( 2mmindiameter,Figure1).Onaverage,thereareoneto tworodsand/orringspercellincludingsomeinterestingapparent intermediatestructuressuchasafigure,anelongatedring, twistedring,rodswithpinloopsatoneend,aswellassomethat appearedtobetransitioningfromrodstothispeculiarform (Figure2).Immunofluorescencewasperformedtovisualizethe RRstructureandlocationwithrespecttothenucleusandthe Golgicomplex(Figure1A).Somerodsoftenalignadjacenttothe nucleus(shortarrows, 40%frequency)orperpendiculartothe nucleus(longarrows, 40%frequency),andringsmaybefound inthecytoplasm(arrowheads,5%frequency).RRare expressedinsomemitoticcells;however,co-stainingstudieswith anti-centromereproteinF(CENP-F),showedthattheexpression ofeitherrodsorringswereindependentofthecellcycle,asthey appearinallstages(Figure3).CENP-Fisacellcyclemarkerthat haslittleornoexpressionattheG1phase,withincreasinglevel duringtheSphase,andreachinghighestexpressionduringlateS andG2phases[12,13]. ToexaminetherelationshipofRRwithotherknown cytoplasmicorganelles,co-localizationstudieswereperformed usingknownmarkersforcytoplasmicstructures.Co-staining studieswithantibodiestotheGolgicomplexshowedthatRRwere independentoftheGolgicomplex(Figure1A).Althoughmany rodsappearedtobeinthesamevicinityastheGolgicomplex, therearealsomanyclear-cutexamplesofrodsunassociatedwith thissubcellularstructure.Cytoplasmicstructures,referredtoas actinrockets,werepreviouslydescribedinbacteria-infectedcells [14].TodeterminewhetherRRarerelatedtoactinrockets,costainingwithmouseanti-actinantibodywasperformedtoshow thatRRarenotenrichedinactin(datanotshown).Furthermore, uponcarefulexaminationathighermagnification,theHEp-2cells inourlaboratorywerenotcontaminatedwithbacteriaandcell viabilitywas 99%.SinceRRarefilamentous,co-staining experimentswerealsoperformedtodeterminewhethertubulin orvimentin,wasenrichedinRR.CostainingshowedthatRRare notenrichedintubulin(Figure1BD).Co-stainingofhuman prototypeserum604andanti-vimentinantibodyruledout enrichmentofRRwithvimentin(Figure1EG). Anotherinitialconsiderationwasthatprimarycilia,which functionasnonmotilesensoryorganellesandhavearod-like appearance[15],couldbeapossiblecandidate.ToruleoutRRas primarycilia,aco-stainingstudywithanti-pericentrin,which Figure1.ThedistributionofcytoplasmicrodsandringswasindependentoftheGolgicomplexandcentrosomes,andthese structureswerenotenrichedintubulinorvimentin. (A)MergedimageofHEp-2co-stainedwithhumananti-RRprototypeserum604/Alexa 488goatanti-humanIg(green)andrabbitanti-giantin(Golgimarker)/Alexa568goatanti-rabbitIg(red).Rodsareoftenpresentedadjacent(short arrows)orperpendicular(longarrows)tothenucleuswhilerings(arrowheads)arefoundeither1or2toacell.M,mitoticcell.HEp-2cellswerealso co-stainedwithserum604/Alexa488goatanti-humanIg(green,B,E,H)anddifferentcytoplasmicmarkersusingmouseanti-tubulin(C),anti-vimentin (F),anti-pericentrin(I),followedbyAlexa568goatanti-mouseIg(red).NucleiwerecounterstainedwithDAPI(blue).Bar,10mm. doi:10.1371/journal.pone.0029690.g001 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 2 December2011|Volume6|Issue12|e29690

PAGE 3

recognizescentrosomes,wasconducted.Centrosomesarethe precursorsofprimaryciliaandareoftenlocatedattheendsofthe primarycilia.Sincecentrosomeswerenotlocatedattheendsof rodsrecognizedbyhumanprototypeserum604(Figure1HJ), theserodsarenotprimarycilia.Infact,thereisnonoticeable associationofeitherrodsorringswithcentrosomes.Furthermore, uponcloserexamination,thefactthatring-typestructureshave neverbeenobservedforprimaryciliaisconsistentwiththe conclusion. ThesefindingsarehighlyreproducibleusingHEp-2cell substratefromINOVA,aswellashome-madecellslidesprepared withinhibitor-treatedcells(seebelow)usingvariouscommonly usedfixativesinbiologicallaboratoriesincludingacetoneat 2 20 u Cfor5minutes,acetone/methanol(1:1)for10minutes,or 3%paraformaldehydeand0.1%triton-X.Co-stainingexperimentswithknownsubcellularmarkersshowednoassociationwith knowncytoplasmicorganellessuchasGWbodies(datanot shown).IdentificationofCTPS1andIMPDH2asRRcomponentsDuringtheinitialphaseofourstudy,CTPS1becamea candidateRRproteinbecause,intheCarnegieproteintrapstudy, itwasnotedthatsimilarrod-likestructureswereobservedin Drosophila embryoscarryingGFP-CTPS1[16,17].Co-staining studieswithrabbitantiDrosophila CTPSantiserumshowed localizationofCTPS1toRR,identifiedwithprototypehuman anti-RRserumIt2006(Figure4A).Asthereare75.0%sequence identityand89.8%similaritybetweenhumanCTPS1and CTPS2,antiserumtoCTPS1probablycrossreactswithCTPS2 andbothmightbecomponentsofRR. AsecondcandidateRRproteinwaspostulatedinpartbecause ofanisolatedring-likestructureinthecytoplasmidentifiedbya Figure2.ExamplesofvariousintermediateRRstructuresinHEp-2cellsstainedwithhumanserum604(green)andcounterstained withDAPI(blue). (A)Figurestructure(arrowhead)withacurvedrodadjacenttothenuclearenvelope;(B)elongatedring;(C)elongated,twisted ring(shortarrow);(D)rodwithapin-loop(longarrow);(E)Figurestructure(arrowhead);(F)anotherrodwithpin-loop(longarrow).Nucleiwere counterstainedwithDAPI(blue).Bar,5mm. doi:10.1371/journal.pone.0029690.g002 Figure3.Theexpressionofrodsversusringswasnotcorrelatedwiththecellcycle. HEp-2cellswereco-stainedwithrabbitanti-CENP-F/ Alexa568goatanti-rabbitIgG(A)andhumananti-RRserum604/Alexa488goatanti-humanIgG(B).G1cellshadlittleornoCENP-Fstaining,whereas lateS/G2andmitotic(M)cellsshowedstrongstainingforCENP-F.NucleicounterstainedwithDAPI(blue).Bar,10mm. doi:10.1371/journal.pone.0029690.g003 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 3 December2011|Volume6|Issue12|e29690

PAGE 4

singleautoimmuneserumrecognizingtheenzymeIMPDH2[18] andtheconsiderationthatCTPS1andIMPDH2areenzymesin closelylinkedpathwaysasdiscussedbelow.Inthefirstexperiment todemonstratewhetherIMPDH2waslocalizedtoRR,a commerciallyavailablerabbitanti-IMPDH2antibodywasused toco-stainHEp-2cells.Anti-IMPDH2stainedbothrodsandrings (Figure4B,arrows).Immunoprecipitationanalysisusinganextract of[35S]-methionine-labeledK562cellsshowedthatrabbitantiIMPDH2andIt2006bothpulleddowna55kDaprotein (Figure4C).Furthermore,IP-Westernanalysisconfirmedthat the55kDaproteinimmunoprecipitatedbyIt2006wasIMPDH2, sinceitwasrecognizedbybothmousemonoclonalanti-IMPDH2 andrabbitanti-IMPDH2(Figure4D).SincehumanIMPDH1and IMPDH2have95.3%sequencesimilarity,ourdatacannot excludeIMPDH1isalsoenrichedinRR.ExpressionofrodsandringsinculturedcellsHEp-2cellslidesfromINOVADiagnosticshavebeen consistentlypositiveforRR,testedfromlottolot,sincethisstudy wasinitiatedoverfiveyearsago.ThesamewasnottrueforHEp-2 cellANAslidesfromothermanufacturersexaminedtodate,with stainingvaryingfromcompletelynegativetogranularordiffused cytoplasmicstaining.Thiscouldbetheexplanationforwhythe subcellularstainingobservedbyBluthner etal [18]waslargely differentfromours,exceptforshowingoneortwoisolatedring structures.Thesedifferencesamongproductsfromseveral manufacturersremainsaquestionnotcompletelyresolved,but maystemfromdifferencesincultureconditions,proprietary sampleprocessing,orboth.HomegrownHEp-2cellsoncoverslips or8-chamberedslideshavegenerallynotbeenpositiveforRR stainingunlesscellswerekeptinrapidlydividinglogphaseor undersubconfluentcelldensity.Confluentculturesweregenerally negativeforRR.Variabilityfromexperimenttoexperimentwas clearlyobserved,andonaverageonly5%ofcellswerepositive forRRdetectablebyserum604orIt2006.Cellswerealwayscostainedwithbothhumananti-RRandrabbitanti-giantin(Golgi marker)antibodiestoensurethatvariableswerenotintroducedin thefixationandimmunostainingstepsintheseexperiments. ExactlyhowINOVADiagnosticscanproducehighlyconsistent HEp-2cellslideswith 95%cellspositiveforRRstainingremains unresolved,butclearlythedifferencemuststemfromtheirunique manufacturingmethods.Thus,HEp-2cellsdonotexpressRR underourstandardcultureconditions.Cellsthatdonotordinarily expressRRhaveadiffusedcytoplasmicpatternwhenstainedwith prototypehumanserum(Figure5A).However,itwasnotedthat cellsrecentlythawedfromliquidnitrogenexpressRRinabout 15%ofthecells.Treatingculturedcellswith0.015%to0.5%of DMSOovernighttosimulatefreezingmediumcondition(often with10%DMSO)didnotinduceformationofRR.InductionofrodsandringsinculturedHEp-2cellsWhenCTPS1andIMPDH2wereidentifiedasRRcomponents,thenextseriesofexperimentsfocusedontheeffectsof inhibitingtheseenzymes.Effectsofeachdrugtreatmentwere analyzedbyimmunofluorescence(Figure5).Thefirstappliedwas theCTPS1inhibitor,DON,onHEp-2cellsculturedat2mM,a functionalconcentrationreportedinliterature[5,19].Asurprising findingwasthatbothrodsandringswerereadilydetectedin 96.7%ofcellsafterovernightincubation(Figure5B)comparedto 0%inuntreatedcontrolcellsculturedinparallel(Figure5A).The RRdetectedinDON-treatedcellswereidenticalinfrequencyto thoseobservedintheINOVAHEp-2slides.AsecondindependentCTPS1inhibitorAcivicin[4,5],aswellastheIMPDH2 inhibitorRibavirin,alsoshowedtheabilitytoinduceRR formationundersimilarconditions(Figure5CandD).Inaddition, RRinducedby2mMDON,knowntoinhibitCTPS1butnot IMPDH2,weredetectablewiththeanti-IMPDH2antibodyalong withthehumananti-RRserum.Therefore,CTPS1andIMPDH2 appearedtoaggregateafterinductionwitheitherinhibitortoform RR. InhibitionofCTPS1orIMPDH2inducedRRformationina concentration-dependentmanner(Table1).Thethreeinhibitors haddifferentefficiencyintheinductionofRRformationinHEp-2 cells.RibavirinwasthemostpotentinducerofRR;evenat concentrationslowerthan7.81mM,RibavirininducedRRin nearlyallcells,whereasDONandAcivicindidnotinduceRR efficientlyattheselowconcentrations(Table1,p 0.0001). Figure4.CTPS1andIMPDH2werehighlyenrichedinRRand humananti-RRprototypeserumIt2006recognizedIMPDH2. (A)EnrichmentofCTPS1,detectedbyrabbitanti-CTPS1(green),toRR (arrows)identifiedbyIt2006(red).Nucleiwerecounterstainedwith DAPI(blue).(B)IMPDH2stainedbyrabbitanti-IMPDH2(red)localizedto RR(arrows)detectedbyIt2006(green).Bar,10mm.(C)Immunoprecipitation(IP)analysisusinganextractof[35S]-methionine-labeledK562 cellsandrabbitanti-IMPDH2,It2006,andasecondhumananti-RR serum609.It2006recognizeda55kDaproteinbandthatco-migrated withIMPDH2immunoprecipitatedbyrabbitanti-IMPDH2.Serum609 didnotimmunoprecipitatethe55kDaprotein.(D)IP-Westernblot demonstratedthatserumIt2006immunoprecipitatedIMPDH2,which wasrecognizedbybothmousemonoclonalandrabbitanti-IMPDH2 antibodies.NHS,controlnormalhumanserum. doi:10.1371/journal.pone.0029690.g004 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 4 December2011|Volume6|Issue12|e29690

PAGE 5

SignificantinductionofRRbyRibavirinvsDONwasobservedat aconcentrationlowerthan0.5mM(p 0.0001)andvsAcivicinat aconcentrationlowerthan15.6mM(p 0.0001).InductionofRR byAcivicinvsDONwassignificantbetween0.25mMand 15.6mM(p 0.0001). TherearealsodifferentialtimerequirementsforRRinduction inHEp-2cellstreatedwithinhibitorsatafixed2mM concentration(Table2).RibavirininducedRRasearlyas 30minandbetween1and24h,nearlyallcellshadRR. InductionofRRbyRibavirinwassignificantlyfasterthanDON, whichinturnwassignificantlyfasterthanAcivicin(Table2). Onaverage,cellstreatedwithDONorAcivicinhadtwoRR percell(Figure5B,5C),whilecellstreatedwithRibavirinhad morethantwoRRpercell(Figure5D).Theadditiveeffectoftwo inhibitorsonRRformationwasalsoinvestigated.Additionofboth DONandRibavirinat2mMdidnothaveanoticeableeffecton RRformationsincebothaloneinducedformationofRRin 95%oftreatedcells.However,wheninhibitorswereappliedat lowerconcentrationsandforshorterperiods,obviousadditive effectswereobserved.Forexample,treatmentofHEp-2cellswith 1mMDONfor3hinducedRRin52%ofcells;however,when thiswascombinedwith1mMRibavirinfor30min,whichalone inducedRRin13%ofcells,RRwasdetectedin91%ofcells.In addition,whenHEp-2cellswereexposedtolowerconcentrations ofRibavirin(2mM,3h),theRRstructuresweresmalland difficulttovisualize;howeverwhencombinedwithDON(1mM, 3h),typicalRRstructureswereobserved.Thus,inhibitionofboth CTPS1andIMDPH2hadapparentadditiveeffectsonRR formation,supportingthefindingthattheyaggregatedinthesame structures.Anunrelatedcompound,enoxacin,usedasanoral broad-spectrumfluoroquinoloneantibacterialagentobservedto enhancemiRNAactivity[20]servedasanadditionalnegative controlasitdidnotinduceRRafter24htreatmentinvarious concentrations(10nM).Thus,theinhibitionofeitherCTPS1 orIMPDH2appearedtobespecificfortheinductionofRRin HEp-2cells.InductionofRRinothercancerandprimarycellsToaddresswhethertheinductionofRRisconserved,multiple celllineswereanalyzed,includingotheradherentmonolayercell lineslikeHeLa(Figure6A),CAL27(Figure6B),andHCT116 (Figure6D),aswellasasuspensionmonocyticcelllineTHP-1 (Figure6C).CellswereincubatedwithDONfor24handcostainedwithserumIt2006andanti-giantin.Allcelllinesexpressed RRatahighfrequency( 98%forHeLa,CAL27,andTHP-1), comparabletothefrequencyobservedinHEp-2cellsexcept HCT116cells,withonly35.8%ofcellspositiveforRR.Untreated cellsnormallydidnotshowexpressionofRR,butisolatedHeLa cellsshowedsomecytoplasmicgranularstainingandTHP-1 producedsomeshortimmaturerods.Todeterminewhetherthe expressionofRRcouldbeobservedinspeciesotherthanhuman, mouse3T3fibroblastsandratNRKcellswerestainedwith It2006.Freshlydefrosted3T3andNRKoftenshowedexpression ofRR(5to25%ofcells)inthefirstoneortwopassages,buttheir expressiondisappearedcompletelyinsubsequentpassages.The abilityoffreshlythawedcellstoexpressRRwasnotexamined furtherduetothelackofconsistencybetweenbatchesoffrozen cells.Mouseprimarycardiomyocytes,aswellastheoftencopurifiedendothelialcellsandfibroblastsisolatedfromneonatal heart,didnotexpressRRuntilinducedwith2mMDONin overnightculture(Figure7).Overall88.4%ofcellshaddetectable RR,butcardiomyocytesdetectedbyco-stainingofactininhadup to92.6%showinginducedRR(Figure7B).Thus,RRexpression appearedtobeconservedastheycouldbeinducedinmultiple immortalizedcells,aswellasinprimaryculturedcells.InductionofRRisquantitativelysensitivetothelevelof cellularIMPDH2proteinTomonitortheformationofRR,theuseofGFPfusion constructswasexplored.HeLacellsweretransfectedwithGFPIMPDH2(Figure8AC)ortheGFPvectoralone(Figure8DF). Followingtransfection,cellsweretreatedwith2mMofRibavirin for24h,andthenfixedandstainedwithrabbitanti-IMPDH2. GFP-IMPDH2transfectedcellshadbrighterfluorescencecomparedtountransfectedcells,confirmingtheexpressionof recombinantIMPDH2(Figure8A).Surprisingly,GFP-IMPDH2 transfectedcellsdidnotshowRRafterRibavirininduction, whereasneighboringuntransfectedcellsshowedRRasexpected (Figure8C).Toruleouttransfection-inducedartifactsinterfering withtheinductionofRR,theinductionofRRincontrolGFPvectortransfectedcellswasnotaffectedbythetransfection (Figure8F).Seriesofexperimentswereperformedtoexplore whetherthelowestexpressionoftransfectedGFP-IMPDH2could producecellswithGFP-labeledRRbytitratingtheamountof transfectedplasmidandbyreducingthetimeoftransfection.None Figure5.InhibitionofCTPS1-orIMPDH2-inducedformationof rodsandrings. UntreatedHEp-2cells(A)andcellstreatedwith inhibitors2mMDON(CTPSinhibitor,B),2mMAcivicin(CTPSinhibitor, C),or2mMRibavirin(IMPDH2inhibitor,D)for24hwereco-stained withhumananti-RRserumIt2006(green)andrabbitanti-giantin(red). NucleiwerecounterstainedwithDAPI(blue).Thepercentageofcells withRRdisplayedforeachexperimentisshowninthelowerright cornerwiththetotalnumberofcellscountedindicatedinparentheses. Bar,10mm. doi:10.1371/journal.pone.0029690.g005 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 5 December2011|Volume6|Issue12|e29690

PAGE 6

oftheseexperimentsyieldedlivecellslabeledwithGFP-RR. Therefore,theoverexpressionofIMPDH2appearedtoinhibit formationofRR,evenathighconcentrationsofRibavirin.When theseGFP-IMPDH2transfectedcellswereexposedtohigher concentrationsofRibavirinupto8mM,onlysmallaggregates wereobservedinsomeofthetransfectedcells. SinceoverexpressionofIMPDH2inhibitedRRinduction,the nextseriesofexperimentsweretoaddresswhetherknockdownof CTPS1orIMPDH2wouldaffecttheinductionofRR.The adoptedstrategywastoexaminecellstransfectedwitheither CTPS1siRNA(siCTPS1)orIMPDH2siRNA(siIMPDH2).Thus, HeLacellswereeitheruntransfectedortransfectedwithsiCTPS1 orsiIMPDH2for44h,followedbyanadditional4hof incubationwiththreedifferentconcentrationsofRibavirinat 2mM,0.2mM,or0.05mM.The2mMRibavirinconcentration wasselectedasthehighestconcentrationtestedand0.2mMwas selectedasthelowestbasedonthedatapresentedinTable1.Not showninTable1forHEp-2cells,0.05mMRibavirinwastestedin HeLacellsandwasshowntobeineffectiveintheinductionofRR. TheinductionofRRwasmonitoredbydoubleimmunofluorescenceusingrabbitanti-IMPDH2antibodyaswellasIt2006.After siIMPDH2knockdown,Westernblotanalysisshowedreductionof Table1. Concentration-dependentinductionofRRinHEp-2cellsusingdifferentCTPS1andIMPDH2inhibitorsfor24h.PercentHEp-2cellswithRR(totalcounted);numberofexperiments DrugConcentrationRibavirin1DON2Acivicin32mM498.2%(231);n=3 94.8%(98);n=5 97.3%(226);n=4 1mM498.9%(194);n=3 87.9%(108);n=3 98.5%(69);n=1 0.5mM499.2%(142);n=2 87.2%(102);n=2 94.3%(172);n=3 0.25mM598.5%(135);n=2 28.4%(116);n=2 94.5%(109);n=3 125mM594.4%(108);n=2 29.1%(79);n=2 98.3%(121);n=3 62.5mM590.8%(120);n=2 22.9%(96);n=2 98.1%(53);n=2 31.3mM598.9%(183);n=2 23.3%(124);n=2 95.2%(62);n=2 15.6mM696.8%(160);n=2 11.6%(155);n=2 30.4%(23);n=2 7.8mM794.1%(202);n=2 5.1%(118);n=2 0%(45);n=2 0.2mM795.6%(249);n=3 2.8%(105);n=2 0%(82);n=1 0mM40%(300);n=3 0%(300);n=3 0%(300);n=31RibavirininducedRRinnearlyallcellsregardlessofthetestedconcentration(0.2mMmM).Thereisnostatisticalsignificance(ns)betweenconcentrationsafter Bonferronisadjustmentexceptwithuntreatedcontrol(0mM;p 0.005);2DONinducedRRinmostcellsatconcentrationshigherthan0.5mM(p 0.005,0.5mMorhighervsanylowerconcentration);3AcivicinstartedtoinduceRRat15.6mMandconcentrationshigherthan31.3mM,nearlyallcellshadRR(p 0.005vslowerconcentration)andnodifferencebetween concentrationswasobserved;4nsbetweenRibavirin,DON,orAcivicin;5RibavirinvsDON,p 0.0001,RibavirinvsAcivicin,ns,DONvsAcivicin,p 0.0001;6RibavirinvsDON,p 0.0001,RibavirinvsAcivicin,p 0.0001,DONvsAcivicin,p=0.024;7RibavirinvsDON,p 0.0001,RibavirinvsAcivicin,p 0.0001,DONvsAcivicin,ns. doi:10.1371/journal.pone.0029690.t001 Table2. DifferentialtimerequirementsforRRinductioninHEp-2cellstreatedwithdifferentCTPS1andIMPDH2inhibitorsat 2mMconcentration.PercentHEp-2cellswithRR(totalcounted);numberofexperiments TimeafteradditionRibavirin1DON2Acivicin30min 0%(300);n=3 0%(300);n=3 0%(300);n=3 30min415.2%(67);n=2 8.3%(120);n=1 5.2%(102);n=1 1h598.1%(53);n=2 13.7%(80);n=3 11.5%(87);n=3 2h595.2%(42);n=2 20.2%(89);n=2 19.3%(119);n=2 3h698.1%(55);n=2 70.5%(88);n=3 39.4%(109);n=2 24h798.2%(231);n=3 94.8%(98);n=5 97.3%(226);n=41RibavirininducedRRasearlyas30minandbetween1to24h,nearlyallcellshadRR(p 0.0001,0minvsalltimepointsand30minvs1to24h).Thereisnostatistical significancebetween1to24hafterBonferronisadjustment;2InductionofRRbyDONwasslowerthanRibavirin(p 0.0001,1,2,and3h;ns,30minand24h)andthepercentageofRRpositivecellsgraduallyincreasedafter1h andwasnearlyallcellsat24h.3AcivicinhadsimilarpatterntoDONinwhichRRpositivecellsgraduallyincreasedafter1handnearlyallcellsat24h.4RibavirinvsDONandDONvsAcivicin,ns;RibavirinvsAcivicin,p=0.03;5RibavirinvsDONandRibavirinvsAcivicin,p 0.0001;DONvsAcivicin,ns;6RibavirinvsDON,DONvsAcivicin,RibavirinvsAcivicin,p 0.0001;7nsbetweenRibavirin,DON,andAcivicin. doi:10.1371/journal.pone.0029690.t002 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 6 December2011|Volume6|Issue12|e29690

PAGE 7

IMPDH2levelsto5%inuntreatedcells,siCTPS1hadlittleor noeffectonthelevelofIMPDH2asexpected,althoughsiCTPS1 wasconfirmedbyreal-timePCRtobefunctionalwith62% CTPS1mRNAknockdown(Figure9A).Thenumberofcellswith RRwerecountedandplottedasapercentage(Figure9B).While siCTPS1transfectiondidnotaffectthepercentageofcellspositive forRRinallthreeconcentrationsofRibavirinexamined, knockdownofIMPDH2showedsignificantincreaseinthe percentageofcellswithdetectableRRformationatboth 0.2mMand0.05mMofRibavirin(p=0.0001for0.2mMand 0.05mMcomparedtountransfectedandsiCTPS1transfectedcells byFishersexacttest).Forthe2mMRibavirintreatment, althoughRRwereinducedin100%ofcellstransfectedwith siIMPDHorsiCTPS1,knockdownofIMPDH2showeda decreaseinthesizeofRR(withtheaveragelengthofrodat 4.6mminsiIMPDH2transfectedversus6.9mminuntransfected) alongwiththenumberofRRpercell(4.5RR/cellinsiIMPDH2 transfectedversus7.1RR/cellinuntransfected,p=0.0024byttest)whencomparedtountransfectedcellsat2mMRibavirin.At 0.2mMRibavirin,increasedformationofRRwasobservedin cellswithreducedlevelsofIMPDH2(42.9%ofcellsinsiIMPDH2 transfectedvs21.6%forsiCTPS1,p=0.0048).At0.05mM Ribavirin,siIMPDH2transfectedcellshad22.4%cellswith shorterrods(Figure9D)comparedtosiCTPS1transfectedcells with3.6%(Figure9C,p=0.0001).Thus,thedatashowthatthe formationofRRbyRibavirinwassensitivetothetotalIMPDH2 levelandexcessintracellularIMPDH2madecellsresistanttothe formationofRR.Therefore,RRformationappearedtodepend onthemolecularratioofRibavirintoIMPDH2andcellswitha loweredIMPDH2levelcouldbeinducedtoformRRatlow concentrationofRibavirin.RRexpressioninembryonicstemcellsFromthedatapresentedintheprevioussections,exceptforthe uniquecaseoftheINOVAHEp-2cells,variousculturedcelltypes didnotconsistentlyexpressRRunlessinducedwithoneofthe CTPS1/IMPDH2inhibitors.Thenextquestionwaswhether certaincelltypeswereuniquelycapableofexpressingRRinthe absenceofinducers.ESCsarehighlyproliferativecellsthatcan self-renewandarepluripotent.Undifferentiated,uninducedESCs werefoundtoexpressRR(Figure10B)similartothosedescribed inHEp-2cells(Figure1A)with1or2RRpercell.Interestingly, unlikeHEp-2cellsormouse3T3cells(Figure10A),whereonly 10%ofcellshadrings, 90%ofESCexpressedrings,and 5% ofcellshadrods(Figure10B).Incontrast,ESCsthatwere differentiatedwithretinoicacid(RA)for4daysnolonger expressedRR(Figure10C),butthosedifferentiatedESCs remainedcapableofRRformationupontreatmentof2mM Acivicinfor24handtheRRphenotypewassimilartotheoriginal undifferentiatedESC(Figure10D).RRexpressioncouldbe inducedinmultiplecancerousandprimarycells,butRRwere foundtobeexpressedinuntreatedstemcells.DiscussionInthisstudy,anovelcytoplasmicstructurewasidentifiedby humanautoantibodiesproducedbypatientswithchronichepatitis Cvirus(HCV)infectionthroughANAtesting.Ourpreliminary Figure6.InhibitionofCTPS1-inducedformationofRRin severalhumancancercelllines. InductionofRRobservedinhuman cancercelllinesHeLa(A),CAL27(B),THP-1(C),andHCT116(D)when treatedwith2mMDONinculturefor24h,fixed,andco-stainedwith humananti-RRserumIt2006(green)andrabbitanti-giantin(red). UntreatedcontrolsshowedfewornoRR.Thepercentageofcellswith RRdisplayedforeachconditionisshownintheupperrightcornerwith thetotalnumberofcellscountedindicatedinparentheses.Bar,5mm. doi:10.1371/journal.pone.0029690.g006 Figure7.InhibitionofCTPS1-inducedformationofRRin mouseprimarycardiomyocytes,fibroblasts,andendothelial cells. Mouseprimarycardiomyocytespreparedtogetherwithfibroblastsandendothelialcells,weretreatedwith2mMDONandcultured for24h.Cellswereco-stainedwithhumananti-RRserumIT2006 (green)andmouseanti-actininmonoclonalantibody(red).Nuclei counterstainedwithDAPI(blue).Actinin-positivecardiomyocytes(A,B, red)aswellasactinin-negativefibroblastorendothelialcells(A,C)all showdistinctrods.ThepercentageofcellswithRRdisplayedisshown inthelowerrightcornerwiththetotalnumberofcellscounted indicatedinparentheses(A,allcells;B,actinin-positivecardiomyocytes only;C;actinin-negativefibroblastandendothelialcells).Bar,10mm. doi:10.1371/journal.pone.0029690.g007 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 7 December2011|Volume6|Issue12|e29690

PAGE 8

Figure8.OverexpressionofIMPDH2preventedtheinductionofRR. HeLacellsweretransfectedwitheitherGFP-IMPDH2(AC)oraGFP vector(DF)for24handfollowedby2mMRibavirinincubationfor24h.Cellswerestainedwithrabbitanti-IMPDH2(red)andcounterstainedwith DAPI(Blue).GFP-IMPDH2transfectedcells(T)didnotexpressRRwhileuntransfectedcellsshowedrods(arrows)andrings(arrowheads).RRwere detectedinbothGFPvectortransfectedanduntransfectedcells.Bar,10mm. doi:10.1371/journal.pone.0029690.g008 Figure9.RRformationincreasinglysensitivetoRibavirinwhenIMPDHproteinlevelisreducedbyspecificsiRNAknockdown. HeLa cellswereeitheruntransfected,transfectedwithsiCTPS1,orsiIMPDH2for44handfollowedbyincubationofRibavirinatvariousconcentrationsfor anadditional4h.(A)IMPDH2expressionwasmonitoredbyWesternblotusingrabbitanti-IMPDH2antibody.SerialdilutionsofuntreatedHeLacell lysates(100%,50%,and25%)wereusedtohelptocomparativelyquantitatethedegreeofproteinknockdown.Thelevelofactinwasusedasa loadingcontrol.RelativeIMPDH2proteinlevelswerequantifiedusingimageJandnormalizedtotheserialdilutionstandards.(B)Percentageofcells withRRwerequantifiedusingMayachitraImagowitharangeof131to298cellscountedperdatapoint.CellstreatedwithsiIMPDH2hada significantincreaseinpercentageofcellswithRRat0.2mMand0.05mMofRibavirinanalyzedbyFishersexacttest.(CD)Representativeimageof cellsatthelowestconcentrationofRibavirintransfectedwithsiCTPS1orsiIMPDH2.Arrows,rods;arrowhead,ring.Bar,5mm. doi:10.1371/journal.pone.0029690.g009 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 8 December2011|Volume6|Issue12|e29690

PAGE 9

datashowedthatonlyasubsetofHCVpatientshadanti-RR antibodies.Only 25%ofHCVpatientstreatedwithboth interferona andRibavirin,themostcommoncombination therapyforHCVinfection,hadanti-RRantibodieswhilenone oftheuntreatedHCVpatientsdevelopedananti-RRresponse. BasedonourpresentdatathatprimarycellsexposedtoRibavirin couldformRRinvitro,onecanspeculatethatcellsfrompatients exposedtoRibavirinwillformRR.Aggregated/modifiedIMPDH (andCTPS)inRRintheseRibavirintreatedpatientsmaybecome antigenicandtriggeranautoimmuneresponse. Mostsubcellularorganellesdisassemblepriortomitosisand reformaftercelldivision.Althoughsomemitoticcellsweredevoid ofRRandthussuggestingRRmightdisassemblepriortomitosis, eitherthedisassemblyofRRwasincompleteorthedisassembly didnotoccuratallinmanycells,since 50%ofmitoticcells, includingallstages,clearlyretainedRR(datanotshown).Atleast oneautoantigen,IMPDH2,recognizedbythehumananti-RR prototypeserumIt2006wasdeterminedthroughIP-Western analysis.Co-localizationstudiesdeterminedCTPS1tobeasecond componentofRR.Inhibitorsofthetwoenzymesinduced formationofRRincellsthatdidnotexpressRRduring continuousculture.RRinductioncouldbeaccomplishedwith lownanomolarinhibitorconcentrations.Overexpressionof IMPDH2inhibitedformationofRRwhileknockdownof IMPDH2proteinallowedformoresensitiveinductionofRR withultralowconcentrationsofRibavirinthatnormallydidnot induceRR.Furthermore,inHeLacellsoverexpressingIMPDH2, higherconcentrationsofRibavirinwerenoteffectiveandcould onlyinduceremnantofRR.Therefore,theinductionofRRis inhibitor-concentrationdependentwithlowerlevelsofIMPDH2 favoringRRformationinthepresenceoflowlevelofRibavirin. SinceCTPS1andIMPDH2aremajorenzymescontrollingthe biosyntheticpathwaysforCTPandGTP,respectively,the formationofRRusingmultipleindependentinhibitorsofthese enzymesimpliesthatthiscytoplasmicstructureislinkedtothese cellularmetabolicpathways.Itis,however,acknowledgedthat DONandAcivicinarenotcompletelyspecificforCTPS1butare notexpectedtoaffectactivityofIMPDH[21].Theobservations thattheformationofRRissensitivetotheconcentrationof IMPDH2(Figure9)andthatRRarepresentinsomecasesin untreatedcellssuggestthattheformationofRRislinkedtothe functionalleveloftheseenzymesincellularregulationrelatedto ribonucleotide(CTPorGTP)synthesis.Sincetheinhibitionof CTPS1causesformationofRRthatcanbedetectedusingan IMPDH2antibody,CTPS1andIMPDH2mustaggregate togethertoformRR,althoughdemonstrationoftheirdirect molecularinteractionislacking.Duetothelargesizeofthese structures,onemayspeculatethattherearemanyother componentsinvolvedinRRformationandsomeofthesemay serveasmolecularlinkerstofacilitateinteractionbetweenCTPS andIMPDH.Futurebiochemicalanalysesareneededtoexplore whetherthereisdirectinteractionofCTPSandIMPDH. Figure10.Uninducedmouseembryonicstemcellsexpressedpredominantlycytoplasmicringsthatweredisassembledduring retinoicacid-induceddifferentiationandreassembledwhentreatedwithAcivicin. (A)Mouse3T3cellsforcomparisonshownwithboth rods(80%,arrowhead)andrings(10%,arrows).(B)UndifferentiatedmouseESCsshownwith 90%rings(arrows)andfewrods.(C)ESCs treatedwithRAfor4daysshowednoRR.(D)RRwereinducedintheseRA-differentiatedcellsbytreatmentwith2mMAcivicinfor24h.Cellswere stainedwithhumananti-RRserumIt2006(green,AD)andco-stainedwithrabbitanti-giantin(red,A,Bonly).NucleicounterstainedwithDAPI(blue). ThepercentageofESCswithRRdisplayedforeachconditionisshowninthelowerrightcornerwiththetotalnumberofcellscountedindicatedin parentheses.Bar,5mm. doi:10.1371/journal.pone.0029690.g010 RodsandRingsinMammalianCells PLoSONE|www.plosone.org 9 December2011|Volume6|Issue12|e29690

PAGE 10

RR-likestructuresareevolutionarilyconservedSinceourinitialobservationofRRoverfouryearsago, morphologicallysimilarstructureswereidentifiedinfiverecent reports[17,22]inadditiontotwopreviousreports[7,26].The firstreportofRR-likestructuresmaybethe1987studyby Willingham etal. [26].Theyreportedacytoplasmicfibrillar structureidentifiedbyanIgG3monoclonalantibodygenerated fromaBalb/cmouseimmunizedwithmousetumorcells.These structureswereidentifiedusingtransmissionelectronmicroscopy andimmuno-electronmicroscopyaslargesingleparacrystalline arraysofindividualfilamentsnotenclosedbyalipidbilayer membrane.Theirstructuresweresimilarinappearanceand numberpercelltotheRRinthisstudy.Thetargetrecognizedby themonoclonalantibody,anunknown58kDaproteinin immunoprecipitationof[35S]-methionineradiolabeledcelllysate, wasreferredtoasnematin(fromnematodes).Theidentityofthe antigenwasnotelucidated[26].Similartoourresults,they reportedthatthesestructureswereunrelatedtotubulinor vimentinandevolutionarilyconservedinhuman,mouse,and rat.Itwasalsoreportedthatthenematinstructureoccurred followingthawingofstoredfrozencells[26]suchasHEp-2,3T3, andNRKcells,alsoobservedinourlaboratory. ThesecondstudyreportingRR-likestructureswasbyJi etal. identifyingIMPDH2proteinaggregatesformedfrominhibitionof enzymaticactivitythroughMPAtreatment[7].Thestructures wereidentifiedbyimmunofluorescenceandelectronmicroscopy analysis[26].Furtheranalysesrevealedthattheaggregateswere reversiblewiththeadditionofguanosineanditsderivatives.The structuresinthatstudyandinourswereinducedwithtwo independentIMPDH2inhibitors,MPAversusRibavirin.Both inducersinhibitIMPDH2,butMPAisanoncompetitiveand reversibleinhibitor[27],whileRibavirinisanonreversible inhibitor. Ingerson-Mahar etal. demonstratedthatCTPSformed filamentsin C.crescentus [23].Identificationofthestructurewas throughelectroncryotomographyalongtheinnercurvatureof C. crescentus .ThemetabolicenzymeCTPShasabifunctionalrole,as afilamentthatregulatescurvatureandperformsitscatalytic function.AkeydifferencewasthattreatmentofDONinthisstudy dissociatedtheCTPSfilamentsinbacteria,whileDONinduced theformationofRRinmammaliancellsinourstudy.Thelatter findingontheeffectofDONonmammaliancellsiscompletely consistentwithdatainarecentreportbyChen etal [25]. OtherrecentreportsidentifiedRR-likestructuresinbothyeast and Drosophila .Noree etal. identifiedfournovelfilamentsthatwere identifiedbyscreeningayeastGFPstraincollection[24].Through inhibitionofCTPS,theydeterminedthatfilamentswere comprisedofinhibitedCTPS,similartoourresults.Inaddition, theywereabletovisualizeCTPSfilamentsinthe Drosophila egg chamber,asubsetofgutcellsandneurons.Anotherreportalso identifiedanintracellularstructurecontainingCTPSin Drosophila cells[17].Thestructures,referredtoascytoophidia,were identifiedinovarycellsandothertissues[25,28].GFPprotein trapflystocksfromtheCarnegieProteinTrapLibraryidentified CTPSasthecomponentofcytoophidia.RRandcytoophidiamay bethesameconservedstructuresamongdifferentspecies[25]. Ramer etal. identifiedalargeintracellularneuronalorganelle thatoccursinasubsetofadultratsympatheticganglionneurons [22].Thestructure,identifiedbyimmunohistochemistryplus confocalandelectronmicroscopy,occurredmainlyinatoroidal shape,butalsoastwistsorrods.Thestructures,referredtoas loukoumasomesordonut-likestructures,wererecognizedbya monoclonalantibodythatwasraisedagainstneuron-specific b IIItubulin,butnotrecognizedbyother b III-tubulinantibodies.One differencebetweenthisandourstudywastheincreasedtoroidal (ring)structuresinneurons,whileweobservedmorerodsthan ringsinallinducedcancercells,butmoreringsthanrodsinESCs.SpontaneousRRversusinhibitor-inducedRRAlthoughRRmayappeartobehighlyconservedasdiscussed above,ourpreliminaryscreeninginmouseliver,kidney,and spleenusingindirectimmunofluorescencedidnotdetectthese structures.UndifferentiatedESCsexpressRRwithouttheuseof inhibitorswhiledifferentiatedcellsdonotexpressRRand, interestingly,differentiatedcellswerestillcapableofRRformation uponinductionwithaCTPS1inhibitor.However,thisshouldnot beinterpretedasde-differentiationbacktostemcells.Anoteof cautionisneeded,asitisnotclearwhetherESCsRRare structurallyandfunctionallythesameasthoseofcancercells treatedwithinhibitors.Oneclearphenotypicdifferenceisinthe ratioofrodstoringsinESCs(rod:ring=1:9)comparedtoallother celltypeswithinducedRRandincludingtheRRdetectedin INOVAHEp-2slides(rod:ring=9:1).Itisunclearwhetherthe spontaneousformationofRRinESCsisrelatedtothefunctional activityofIMPDH2orCTPS1butonecanspeculatethat,since ESCsaremetabolicallyhighlyactive,thespontaneousformation ofRRislikelytosuggestthatRR-associatedIMPDH2inESCsare functionallyactive.Theabilitytoself-renewandpluripotencyare twouniquepropertiesofstemcells,andthesepropertiesmayhave aneffectonIMPDH2(andCTPS1)activity.Oncestemcells differentiate,thecellularrequirementofIMPDH2(and/or CTPS1)maychangeandtheRRisnolongerneeded.One biologicalfunctionofRR,asdemonstratedhere,isrelatedto CTP/GTPsyntheticpathways. SimilartoESCs,theINOVAHEp-2cellswereactivelydividing cells(beforefixation),basedontherelativelyhighnumberof detectablemitoticcells.Thus,RRinINOVAHEp-2cellsarealso likelytobefunctionalintermsoftheircomponentCTPS1and IMPDH2.OnecanspeculatethattherequirementofCTPand GTPinrapidlydividingcellsfavorstheaggregationofthese nucleotidessyntheticenzymestopossiblyenhancetheirfunctional activities.Thisisconsistentwithourexperiencethat,whilefreshly thawedcellscandemonstrateupto25%cellswithRRinfirstor secondcellpassages,cellconfluenceisoftenassociatedwiththe completelossofRRexpressionIMPDH2/CTPS1nolonger assembleinthecytoplasm.ThisarguesfortheformationofRRto beassociatedwithahighrequirementofintracellularCTP/GTP andtheassociatedbiosyntheticenzymesareassembledintoRR, whichcanenhancetheirsyntheticactivity. Incontrast,inducibleRRincancercells,aswellasprimary cells,arecomposedofinactivatedenzymesbasedonthepresence ofaninhibitoraddedtoinducethesestructures.Higherexpression ofIMPDH2inhibitedformationofRRbycomparablelevelsof inhibitors.LowerexpressionofIMPDH2,asinthepartial knockdownexperiment,promotedformationofRRwithalower concentrationofinhibitor.TheapparentdifferencebetweenESCs withlikelyactiveRR-associatedIMPDH2/CTPS1versusthe inhibitor-inducedRRwithinactiveenzymesmaybeexplainedas follows.WhenHEp-2cellsareexposedtoaninhibitor,suchas Ribavirin,thelowerproductionofGTPmaytriggertheresponse forthebiosyntheticenzymetoassembleintooligomer/polymer configurationtofavoranincreaseinenzymeactivity,consistent withthespeculationabovethatESCshaveastrongGTP/CTP requirementtomaintainactivecellproliferation.OnceRRis formed,Ribavirinanirreversibleinhibitor,locksthe configurationofRR,makingtheenzymesinactive. Insummary,RRformationisregulatedbytheinhibitionof CTPS1andIMPDH2,importantinpurineandpyrimidineRodsandRingsinMammalianCells PLoSONE|www.plosone.org 10 December2011|Volume6|Issue12|e29690

PAGE 11

biosynthesisandcriticalforRNAandDNAsynthesis.The importanceoftheseenzymesinRNAandDNAsynthesissuggests thatRRmayplayaroleincelldivision.Inaddition,bothCTPS1 andIMPDH2aggregatetogethertoformtheRRstructureand inhibitionofonewillcausetheaggregationoftheother componentwithRR.StudieshavedemonstratedthatCTPS1 andIMPDH2areaberrantlyexpressedindifferentcancers[4,29]. However,asCTPS1andIMPDH2aredifferentiallyexpressed, theformationofRRincellsmayhelpregulatetheiroverall activities.FurthercharacterizationandunderstandingofRRmay leadtoanewbiomarkerfordiseasesornovelmarkersforstem cells.MaterialsandMethods CellcultureHeLa(humancervicalcancer),HEp-2(humanepidermoidlarynx carcinoma),andHCT116(humancoloncancer)wereculturedin DMEMcontaining10%FetalBovineSerum(FBS).Humanoral cancercellsCAL27wereculturedinDMEMcontaining10%FBS and1.5g/Lsodiumbicarbonate.THP-1,humanmonocytes,and K562,humanerythroleukemiacells,wereculturedinRPMI1640 with25mMHEPES,and10%FBS.Allcellswereculturedina 37 u Cincubatorwith5%CO2.Adherentcelllinesweremaintained at50%confluence.Suspensioncellsweremaintainedatanoptimum concentrationof106cells/mL.Allmediacontained100I.U./ml penicillinand100mg/mlstreptomycin.EmbryonicstemcellcultureanddifferentiationThemurineR1ESCswereobtainedfromDr.AndreNagy (UniversityofToronto)andmaintainedaccordingtoHamazaki etal. [30].ESCsweredifferentiatedwith1mMretinoicacidfor4 consecutivedaysasdescribed[31].MousecardiomyocytecollectionandpreparationThemousecardiomyocyteswereharvestedandprepared accordingtoChan etal. [32].HumanautoantibodiesAutoantibodieswithreactivitytoRRwereselectedbyauthors GCandCAvMduringroutineANAscreeningassay.Allthree humananti-RRserausedinthisstudywerefrompatientswith chronicHCVinfectionandwerede-identifiedsamples.Thesesera wereselectedlargelyduetotheirrelativespecificityinRRstaining andinavailablequantitytoallowfurtheranalyses.SerumIt2006 providedbyGC,andserum604and609providedbyCAvM,are consideredprototypeseraforanti-RR.Theclinicalcharacterizationofthesepatientsandotherswithanti-RRwillbesummarized inasubsequentreport[33].ImmunoprecipitationImmunoprecipitationof[35S]-methionine-labeledK562cells fortheanalysisoftheproteinsrecognizedbyhumanautoimmune serawasperformedasdescribed[34].IndirectimmunofluorescenceHEp-2cellslides(INOVADiagnostics,SanDiego,CA)or variousadherentcelltypesculturedin8-wellchamberslides(BD Falcon)wereusedforindirectimmunofluorescenceasdescribed [35,36].The8-wellchamberslideswerepreparedusing 60,000 cellsperwellin500mLofmedia.Non-adherentcellswere preparedusingthecytospinmethodtotransfercellsontoglass slidesasdescribed[37].Fixationmethodsincludedacetoneat 2 20 u Cfor5min,acetone/methanol(3:1)at 2 20 u Cfor10min, and3%paraformaldehydeinPBSatroomtemperaturefor 10minfollowedby0.1%triton-X/PBSforanother10min.For co-stainingstudiesthefollowingprimaryantibodieswereused: rabbitanti-giantin[38],anti-pericentrin(agiftfromDr.Marvin Fritzler,UniversityofCalgary),anti-DrosophilaCTPS1andpreimmuneserum(agiftfromDr.JimWilhelm,Universityof California,SanDiego),anti-IMPDH2(Abcam,Cambridge,MA: ab75790;Proteintech,Chicago,IL:12948-1-AP),mouseantiactin,anti-actinin(Sigma-Aldrich,St.Louis,MO:A7811),antitubulin,andanti-vimentin,andhumananti-RRserumIt2006, 604,and609.Tovalidatethattherabbitanti-DrosophilaCTPS1 serumrecognizedhumanCTPS1,afull-lengthhumanCTPS1 cDNA(IMAGEid3355881,OpenBiosystems,ThermoScientific) wasusedasatemplateinan invitro transcriptionandtranslation reaction(TNT,Promega)togenerate[35S]-methioninelabeled translationproducts.Thelatterwasshowntobeimmunoprecipitatedbytherabbitanti-DrosophilaCTPS1serumandnotbythe pre-immuneserumfromthesamerabbit.InductionofRRformationusingCTPSorIMPDH inhibitorsAcivicin(EnzoLifeSciences,AnnArbor,MI;BML-El1130010)andRibavirin(Sigma-Aldrich;R9644)weresolubilizedin waterto100mMand50mMstockconcentration,respectively.6diazo-5-oxo-L-norleucine(DON,Sigma-Aldrich;D2141)was solubilizedinDMEMtoastockconcentrationof100mM.Cells wereseededasmonolayersandallowedtoattachfor24h. Subsequently,CTPS1orIMPDH2inhibitorswereaddedto variousfinalconcentrationsfrom0.2mMto2mMandwerekept inacellincubatorfortimesrangingfrom15minto48h.siRNAknockdownAllsiRNAsusedinthisstudywereobtainedfromDharmacon RNATechnologies(Lafayette,CO).ThesiRNAsweredissolvedin molecularbiologygradewaterandtheresulting20mMstockwas storedinaliquotsat 2 80 u Cbeforeuse.Thetwopredesigned siRNAswere:ON-TARGETplusSMARTpoolsiRNAHuman CTPS1(targetsequences:UAACAUUGAUGCAGGAACA,AGACUAAACCUACCCAGAA,GGUGUUAUAUCAGGAAUUG,andGUAUGCAGGUGCUCAAAUC),siGENOMESMARTpoolsiRNAIMPDH2(targetsequences:GGACAGACCUGAAGAAGAA,GCACGGCGCUUUGGUGUUC,GGAAAGUUGCCCAUUGUAA,andCUAAAGAAAUAUCGCGGUA). TransfectionwasperformedusingLipofectamine2000(Invitrogen,Carlsbad,CA)perthemanufacturersinstruction.ThesiRNA wasusedatafinalconcentrationof100nM.Knockdown efficiencywasdeterminedbyreal-timePCRanalysisusingCTPS primers(AppliedBiosystems,Hs01041851_m1)orIMPDH2 primers(AppliedBiosystems,Hs00168418_m1)plusWesternblot forproteinexpression.TransfectionofGFPplasmidHeLacellswereseededin8wellchambersat50% confluency.Cellsweretransfectedwith0.35mgGFP-IMPDH2 (RG202977,OriGene,Rockville,MD)oraGFPvectoralone using0.53mgofLipofectamine2000.Thetransfectedcellswere examinedafter24hdirectlyaslivecellsorwereprocessedfor indirectimmunofluorescence.WesternblottingProteinlevelswereanalyzedasdescribed[39]usingrabbitantiCTPS(Abcam,1:1000)orrabbitanti-IMPDH(SantaCruzRodsandRingsinMammalianCells PLoSONE|www.plosone.org 11 December2011|Volume6|Issue12|e29690

PAGE 12

Biotechnology,PasoRobles,CA,1:1000).ForIP-Western,IgG wascrosslinkedtoproteinASepharosebeadsasdescribed[40] andhumanIt2006wasusedforimmunoprecipitationandmouse monoclonalandrabbitanti-IMPDHfromSantaCruzBiotechnologywereusedfordetection.Mouseanti-actinantibody31G9 wasusedat1:5,000or 1mg/mlasanormalizertocontrolfor loading.StatisticalanalysisCellcountingwasperformedusingMayachitraImago(Mayachitra,SantaBarbara,CA)todetectnucleicounterstainedwith DAPI.Unpaired,two-tailedStudentst-testandFishersexacttest wereusedtocompareindependentgroups.Whenappropriate, Bonferronisadjustmentformultiplecomparisonswasapplied.For allstatisticalanalysis,PrismforWindows,version5.0(GraphPad Software,SanDiego,CA)wasused.AcknowledgmentsErwinLam,NatashaDeming,andPabinaDhawanareacknowledgedfor theearlyworkhelpingtodevelopthisstudy.Dr.JosephG.Gall, DepartmentofEmbryology,CarnegieInstitutionforScience,Baltimore, MD,isgratefullyacknowledgedfortheveryearlyconsultationonthe similarityofRRtostructuresheobservedinGFP-CTPS Drosophila ovaries.AuthorContributionsConceivedanddesignedtheexperiments:WCCMSEKLC.Performed theexperiments:WCCMSTHJYFCBYSTEKLC.Analyzedthedata: WCCMSEKLC.Contributedreagents/materials/analysistools:HKNT CAvMGC.Wrotethepaper:WCCMSEKLC.References1.TanEM(1991)Autoantibodiesinpathologyandcellbiology.Cell67:841. 2.AndradeLE,ChanEKL,RaskaI,PeeblesCL,RoosG,etal.(1991)Human autoantibodytoanovelproteinofthenuclearcoiledbody:immunological characterizationandcDNAcloningofp80-coilin.JExpMed173:1407. 3.EystathioyT,ChanEKL,TenenbaumSA,KeeneJD,GriffithK,etal.(2002)A phosphorylatedcytoplasmicautoantigen,GW182,associateswithaunique populationofhumanmRNAswithinnovelcytoplasmicspeckles.MolBiolCell 13:1338. 4.KursulaP,FlodinS,EhnM,HammarstromM,SchulerH,etal.(2006) StructureofthesynthetasedomainofhumanCTPsynthetase,atargetfor anticancertherapy.ActaCrystallogrSectFStructBiolCrystCommun62: 613. 5.HoferA,SteverdingD,ChabesA,BrunR,ThelanderL(2001)Trypanosoma bruceiCTPsynthetase:atargetforthetreatmentofAfricansleepingsickness. ProcNatlAcadSciUSA98:641216. 6.CarrSF,PappE,WuJC,NatsumedaY(1993)CharacterizationofhumantypeI andtypeIIIMPdehydrogenases.JBiolChem268:272867290. 7.JiY,GuJ,MakhovAM,GriffithJD,MitchellBS(2006)Regulationofthe interactionofinosinemonophosphatedehydrogenasewithmycophenolicAcid byGTP.JBiolChem281:206. 8.FranchettiP,GrifantiniM(1999)Nucleosideandnon-nucleosideIMP dehydrogenaseinhibitorsasantitumorandantiviralagents.CurrMedChem 6:599. 9.KaurR,KlichkoV,MargolisD(2005)Exvivomodelingoftheeffectsof mycophenolicacidonHIVinfection:considerationsforantiviraltherapy.AIDS ResHumRetroviruses21:116. 10.RobertsonCM,HermannLL,CoombsKM(2004)Mycophenolicacidinhibits avianreovirusreplication.AntiviralRes64:55. 11.ZhouS,LiuR,BaroudyBM,MalcolmBA,ReyesGR(2003)Theeffectof ribavirinandIMPDHinhibitorsonhepatitisCvirussubgenomicrepliconRNA. Virology310:333. 12.CasianoCA,HumbelRL,PeeblesCL,CoviniG,TanEM(1995) Autoimmunitytothecellcycle-dependentcentromereproteinp330d/CENPFindisordersassociatedwithcellproliferation.JAutoimmun8:575. 13.CasianoCA,LandbergG,OchsRL,TanEM(1993)Autoantibodiestoanovel cellcycle-regulatedproteinthataccumulatesinthenuclearmatrixduringS phaseandislocalizedinthekinetochoresandspindlemidzoneduringmitosis. JCellSci106:1045. 14.SouthwickFS,LiW,ZhangF,ZeileWL,PurichDL(2003)Actin-based endosomeandphagosomerocketinginmacrophages:activationbythe secretagogueantagonistslanthanumandzinc.CellMotilCytoskeleton54: 41. 15.ChristensenST,PedersenSF,SatirP,VelandIR,SchneiderL(2008)The primaryciliumcoordinatessignalingpathwaysincellcyclecontroland migrationduringdevelopmentandtissuerepair.CurrTopDevBiol85: 261. 16.BuszczakM,PaternoS,LighthouseD,BachmanJ,PlanckJ,etal.(2007)The Carnegieproteintraplibrary:aversatiletoolforDrosophiladevelopmental studies.Genetics175:150531. 17.LiuJL(2010)IntracellularcompartmentationofCTPsynthaseinDrosophila. JGenetGenomics37:281. 18.BluthnerM,AppelhansH,MoosbruggerI,WiemannC,LenzD,etal.(2009) Identificationofinosine-1,2-monoph osphate-dehydrogenase-2asanovel autoantigen.In:ConradK,ChanEKL,FritzlerMJ,HumbelRL,von LandenbergP,etal.(2009)Proceedings9thDresdenSymposiumon Autoantibodies:PabstSciencePublishers,Lengerich,Germany.pp125. 19.WuF,LukiniusA,BergstromM,ErikssonB,WatanabeY,etal.(1999)A mechanismbehindtheantitumoureffectof6-diazo-5-oxo-L-norleucine(DON): disruptionofmitochondria.EurJCancer35:1155. 20.ShanG,LiY,ZhangJ,LiW,SzulwachKE,etal.(2008)Asmallmolecule enhancesRNAinterferenceandpr omotesmicroRNAprocessing.Nat Biotechnol26:933. 21.PinkusLM(1977)Glutaminebindingsites.MethodsinEnzymology46: 414. 22.RamerMS,CruzCabreraMA,AlanN,ScottAL,InskipJA(2010)Anew organellarcomplexinratsympatheticneurons.PLoSOne5:e10872. 23.Ingerson-MaharM,BriegelA,WernerJN,JensenGJ,GitaiZ(2010)The metabolicenzymeCTPsynthaseformscytoskeletalfilaments.NatCellBiol12: 739. 24.NoreeC,SatoBK,BroyerRM,WilhelmJE(2010)Identificationofnovel filament-formingproteinsinSaccharomycescerevisiaeandDrosophilamelanogaster.JCellBiol190:541. 25.ChenK,ZhangJ,TastanOY,DeussenZA,SiswickMY,etal.(2011)Glutamine analogspromotecytoophidiumassemblyinhumanandDrosophilacells.JGenet Genomics38:391. 26.WillinghamMC,RichertND,RutherfordAV(1987)Anovelfibrillarstructure inculturedcellsdetectedbyamonoclonalantibody.ExpCellRes171:284. 27.AllisonAC,KowalskiWJ,MullerCD,EuguiEM(1993)Mechanismsofaction ofmycophenolicacid.AnnNYAcadSci696:63. 28.LiuJL(2011)Theenigmaticcytoophidium:compartmentationofCTPsynthase viafilamentformation.BioEssays33:159. 29.HeY,MouZ,LiW,LiuB,FuT,etal.(2009)IdentificationofIMPDH2asa tumor-associatedantigenincolorectalcancerusingimmunoproteomicsanalysis. IntJColorectalDis24:1271. 30.HamazakiT,OkaM,YamanakaS,TeradaN(2004)Aggregationofembryonic stemcellsinducesNanogrepressionandprimitiveendodermdifferentiation. JCellSci117:5681. 31.GuP,LeMenuetD,ChungAC,ManciniM,WheelerDA,etal.(2005)Orphan nuclearreceptorGCNFisrequiredfortherepressionofpluripotencygenes duringretinoicacid-inducedembryonicstemcelldifferentiation.MolCellBiol 25:8507. 32.ChanJY,TakedaM,BriggsLE,GrahamML,LuJT,etal.(2008)Identification ofcardiac-specificmyosinlightchainkinase.CircRes102:571. 33.CoviniG,CarcamoWC,BrediE,vonMuhlenCA,ColomboM,etal.(2012) Cytoplasmicrodsandringsautoantibodiesdevelopedduringpegylated interferonandribavirintherapyinpatientswithchronichepatitisC.Antiviral Therapy,inpress. 34.SatohM,LangdonJJ,HamiltonKJ,RichardsHB,PankaD,etal.(1996) Distinctiveimmuneresponsepatternsofhumanandmurineautoimmuneserato U1smallnuclearribonucleoproteinCprotein.JClinInvest97:2619. 35.FritzlerMJ,HamelJC,OchsRL,ChanEKL(1993)Molecularcharacterization oftwohumanautoantigens:uniquecDNAsencoding95-and160-kDproteinsof aputativefamilyintheGolgicomplex.JExpMed178:49. 36. JakymiwA,IkedaK,FritzlerMJ,ReevesWH,SatohM,etal.(2006) AutoimmunetargetingofkeycomponentsofRNAinterference.Arthritis ResTher8:R87. 37.PauleyKM,SatohM,PauleyBA,Dominguez-GutierrezPR,WalletSM,etal. (2010)FormationofGW/PbodiesasmarkerformicroRNA-mediatedregulation ofinnateimmunesignalinginTHP-1cells.ImmunolCellBiol88:205212. 38.NozawaK,CasianoCA,HamelJC,MolinaroC,FritzlerMJ,etal.(2002) FragmentationofGolgicomplexandGolgiautoantigensduringapoptosisand necrosis.ArthritisRes4:R3. 39.LianS,FritzlerMJ,KatzJ,HamazakiT,TeradaN,etal.(2007)Small interferingRNA-mediatedsilencinginducestarget-dependentassemblyofGW/ Pbodies.MolBiolCell18:3375. 40.MoserJJ,ChanEKL,FritzlerMJ(2009)OptimizationofimmunoprecipitationwesternblotanalysisindetectingGW182-associatedcomponentsofGW/P bodies.NatProtoc4:674.RodsandRingsinMammalianCells PLoSONE|www.plosone.org 12 December2011|Volume6|Issue12|e29690