of December 9, 2014.This information is current asImmune Privilege of the FetusExosome-Mediated Immune Cells, Suggesting Molecules and Convey Apoptosis in ActivatedCarry Functional Fas Ligand and TRAIL Exosomes Secreted by Human Placentaand Lucia Mincheva-NilssonAnn-Christin Stenqvist, Olga Nagaeva, Vladimir Baranovhttp://www.jimmunol.org/content/191/11/5515doi: 10.4049/jimmunol.1301885November 2013;2013; 191:5515-5523; Prepublished online 1 J ImmunolMaterialSupplementary5.DC1.htmlhttp://www.jimmunol.org/content/suppl/2013/11/01/jimmunol.130188Referenceshttp://www.jimmunol.org/content/191/11/5515.full#ref-list-1, 25 of which you can access for free at:cites 61 articles This article Subscriptionshttp://jimmunol.org/subscriptions is online at:The Journal of Immunology Information about subscribing to Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at: Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at: Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright 2013 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc., is published twice each month by The Journal of Immunology at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from TheJournalofImmunologyExosomes Secreted by Human Placenta Carry Functional FasLigandandTRAILMoleculesandConveyApoptosisinActivatedImmuneCells, SuggestingExosome-MediatedImmunePrivilegeoftheFetusAnn-ChristinStenqvist,OlgaNagaeva,VladimirBaranov,andLuciaMincheva-NilssonApoptosis is crucially important in mediating immune privilege of the fetus during pregnancy. We investigated the expression andin vitro apoptotic activity of two physiologically relevant death messengers, the TNF family members Fas ligand (FasL) and TRAILin human early and term placentas. Both molecules were intracellularly expressed, conned to the late endosomal compartment ofthe syncytiotrophoblast, and tightly associated to thegeneration and secretion of placental exosomes.Using immunoelectron mi-croscopy, we show that FasL and TRAIL are expressed on the limiting membrane of multivesicular bodies where, by membraneinvagination,intraluminalmicrovesiclescarryingmembranalbioactiveFasLandTRAILareformedandreleasedintheextra-cellularspaceasexosomes. Analyzingexosomessecretedfromplacental explant cultures, toourknowledge, wedemonstratefortherst timethat FasLandTRAILareclusteredontheexosomal membraneasoligomerizedaggregatesreadytoformdeath-inducingsignalingcomplex. Consistently, placental FasL- andTRAIL-carryingexosomestriggeredapoptosisinJurkatTcellsandactivatedPBMCinadose-dependentmanner. LimitingtheexpressionoffunctionalFasLandTRAILtoexosomescomprisea dual benet: 1) storage of exosomalFasL andTRAIL in multivesicularbodies is protected from proteolyticcleavageand 2) upon secretion, delivery of preformed membranal death molecules by exosomes rapidly triggers apoptosis. Our results sug-gestthatbioactiveFasL-andTRAIL-carryingexosomes, abletoconveyapoptosis, aresecretedbytheplacentaandtieuptheimmunomodulatoryandprotectiveroleofhumanplacentatoitsexosome-secretingability. TheJournalofImmunology, 2013,191:55155523.Mammalianreproductionhastofaceandsolvetheim-munological challengeofacceptingasemiallogeneicfetus and supporting its development and growth. Thehumanhemochorial placenta, knowntopreferentiallyexpresspaternal Ags, is in direct contact with the maternal blood, and thus,thesyncytiotrophoblast (STB) coveringthechorionic villi isphysically exposed to a risk of a potential attack from the maternalimmunesystem. Despitethis, thehumanplacentaandthefetalallograft evadeaharmful maternal immuneattackandenjoyanimmunologicprivilegeintheuterinecavity.Multiplematernal andfetal mechanismsof toleranceworkinconcerttopreventfetalrejection(reviewedinRef.1).Severalofthese mechanisms are mediated by the placenta that not onlysupplieshormonal, nutritional,andoxygensupportbutalsostandsout asanimportant immunomodulatoryorgan. Apoptosis, orpro-grammedcell death, isoneof theproposedmaternofetal mecha-nisms oftolerance thatplays an important roleinimplantationandpregnancy(2). Apoptosis is detectedinthe placenta throughoutpregnancy and has been shown to participate in trophoblast invasion,differentiation, and turnover during placental formation (3, 4) and inthecreationofmaternalimmunetolerancetowardthefetus(5,6).Apoptotic cell death can be initiated through an intrinsic pathway byintracellular signals sensed by mitochondria or an extrinsic pathwayvia death receptors, members of the TNF superfamily (7, 8). Of allmembers, thedeathmessengersFasligand(FasL)andtheTRAILarephysiologically themost relevant.FasL(apoptosis Ag[APO]-1LandCD95L), a homotrimeric37-kDatypeIImembraneprotein, signalscell deathbyengage-ment of its cognate receptor Fas (APO-1andCD95). Initially,FasL was described as an apoptosis-inducing protein expressed onactivated immune cells. Later studies revealed that it is involved inorgandevelopment andhomeostasisandisexpressedonvariousother cells andtissues includingcells associatedwithimmuneprivileged sites such as the eye, brain, testis, and placenta (9, 10).The intracellular and extracellular domains of FasL are located intheN- andC-terminal regions, respectively. Theproapoptoticactivityisassociatedwithmembranal expressionofFasLwhereoligomerization of the molecule is needed in the formation of thedeath-inducing signaling complex (DISC) (11). A cleavage of theextracellulardomainbythemetalloproteasemetralysinproducesa 26-kDa soluble form of FasL (12) that loses its apoptotic activityandcaneveninhibittheactionofmembrane-bound(m)FasL.TRAIL(APO-2LandCD253) was rst identiedas a TNFsuperfamilymember capableof inducingapoptoticcell deathDivision of ClinicalImmunology, Department of ClinicalMicrobiology, UmeaUni-versity,S-90185Umea,SwedenReceivedfor publicationJuly16, 2013. Acceptedfor publicationSeptember 25,2013.This work was supported by the Swedish National Research Foundation (Vetenskapsradet,Grant K2013-54X-22341-01-05); the Swedish National Cancer Research Foundation(CancerfondenGrant CAN2010/495); Central ALFfundingandSpjutspetsanslag,CountyofVasterbotten;andtheInsamlingsstiftelsenattheMedical Faculty, UmeaUniversity.Addresscorrespondenceandreprint requeststoDr. LuciaMincheva-Nilsson, Divi-sionofClinical Immunology, DepartmentofClinical Microbiology, Umea Univer-sity, S-90185 Umea, Sweden. E-mail address: lucia.mincheva-nilsson@climi.umu.seTheonlineversionofthisarticlecontainssupplementalmaterial.Abbreviationsusedinthisarticle: APO, apoptosisAg; CTB, cytotrophoblast; DAB,3,39-diaminobenzidine tetrahydrochloride; DISC, death-inducing signalingcomplex;EM, electron microscopy; FasL, Fas ligand; IEM, immunoelectron microscopy;mFasL, membrane-boundFas ligand; MMP,matrixmetalloproteinase;MVB,multi-vesicularbody;sFasL,solubleFasL;STB,syncytiotrophoblast.Copyright 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00www.jimmunol.org/cgi/doi/10.4049/jimmunol.1301885 at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from mainlyintumorcells.TRAIL, atypeIImembraneprotein, spon-taneouslytrimerizes aroundacentrallypositionedZnatomim-portant for its stability, solubility, and biological activity as a deathligand (13) and binds to four membrane-bound receptorsTRAIL-R1/death receptor 4, TRAIL-R2/death receptor 5, TRAIL-R3/decoy receptor 1, and TRAILR4/decoy receptor 2 (3); and thesoluble receptor osteoprotegerin (14). The ability of TRAIL to killtransformedcellsbyapoptosisiswell established, but theover-allphysiological roleofTRAILisnot fullydened; however,TRAIL-induced apoptosis has also been implied in other pro-cessessuchasactivation-inducedcell death, homeostasis, intra-thymic negative selection, suppression of autoimmunity, and immunesurveillance (14). Studies of the cytotoxic immune response showedthat both FasL and TRAIL could be recovered from the supernatantof cultured activated T cell blasts and leukemia cells in a bioactivemicrovesicular form (15). These ndings were supported by ourmorphological studies of the cytotoxic machinery of activateddecidual gdT and NK cells, which revealed that FasL and perforinwere preformed and stored in microvesicles in multivesicular body(MVB)likemembrane-boundcytotoxicgranules(5).Both FasL and TRAIL were found to be constitutively expressedbytheplacentaandproposedascooperatingplayersforinducingapoptosisin activatedlymphocytes(1618) thatmight posea po-tentialthreattothefetus. Asstatedabove, membrane-boundFasligand(mFasL)butnotsolubleFasL(sFasL)promotesapoptosisof Fas-bearing cells (19, 20). Convincing reports by us and othershave shown that the human placenta lacks cell surface membranalFasLexpression(21,22).Instead,theFasLexpressionislocatedinside the cytoplasm in the endosomal compartment of the STB ofthechorionicvilliandcouldbesecretedinactivemicrovesicularform(22).Exosomes are30- to100-nmsizedmembrane-boundmicro-vesicles formedandreleasedthroughthelateendosomal com-partment byavarietyof cells. Theycontainbothcytosolicandtransmembraneproteins, andmRNAandmicroRNAandcanbeviewedas awayof intercellular communicationwithout theneedforcellcell contact, thusconveyingapotential biologicactivitytomoredistanttargets. Dependingontheiroriginandcomposition, the exosomes can have either immunostimulatoryor immunosuppressiveproperties (2325). Lately, theroleofmicrovesicles/exosomesinthemultifactorial waysofpromot-ingimmunotolerancehas gainedelevatedfocus inmanydif-ferent areas, includingtumorescapeandfetal survival duringpregnancy(2325).Inpreviousimmunoelectronmicroscopy(IEM) studiesof thehuman placenta, we showed that FasL expression takes place in theSTBandis solelyconned to theendosomal compartmentwhereFasL-expressing microvesicles were revealed in the MVBs as wellassignsof MVBfusionwiththeplasmamembrane, suggestinga release of FasL-carrying exosomes (22). In the current study, wehave extended and deepened our investigation to include thebiogenesis, expression, andpreciseintracellular locationof theother major proapoptotic TNFsuperfamilymember TRAIL. Inthis study, weprovideevidencethat similarlytoFasL, TRAILexpressioninthehumanplacentais restrictedtoSTBandispresentinMVBsonmicrovesiclesreleasedintheextracellularspaceasexosomes.Furthermore,toourknowledge,weshowfortherst timethat FasLandTRAILwereexpressedontheexo-somal membrane inanoligomerizedcomplexof twohomotri-meric FasL and/or TRAIL molecules required for DISC formationandtriggeringapoptosis, andshowevidence for their biologicactivityinexosome-inducedapoptosisinJurkatTcellsandacti-vatedPBMCfromhealthydonors.MaterialsandMethodsSamplesandAbsEarlyandtermplacentalspecimensweredonatedbyhealthywomenun-dergoing elective termination of pregnancy (814 wk of gestation) ornormal deliveryafter ethical committeeapproval andinformedconsent.ClonesandspecicitiesofAbswereasfollows:mouseanti-humanFasL(G247-4; BD Pharmingen), goat anti-human TRAIL (K-18), mouse anti-human CD63 (MX-49.129.5; Santa Cruz Biotechnology), mouse anti-humanpanMIC(clone6D4; BDBiosciences), FITC-conjugatedmouseanti-humanCD63(Immunotech), mouseanti-humanCD95(EOS9.1;eBioscience), mAbs against TRAIL-R1 (clone number 69036) andTRAIL-R2 (clone number 71908) (R&DSystems), CD56 (My31; BDBiosciences),mAbsagainstCD4,CD8,CD19,andisotype-matchedcon-trol mAbs IgG1, IgG2a, and IgG2b (DakoCytomation), normal goat serum(Vector Laboratories), and HRP-conjugated secondary Abs (JacksonImmunoResearchLaboratories).RNAextraction,RT-PCR,andquantitativePCRfromisolatedandlaser-microdissectedSTBTrophoblast fromplacental villi (three early and three termplacental samples)was isolated using a mild optimized protocol as described in Ref. 26. STBwasseparatedandenrichedfromtheisolatedtrophoblast bypositiveselectionwithanti-humanpanMICcoatedDynabeadsaspreviouslydescribed(26),put immediately in dissolving solution and frozen for total RNA isolation. Inaddition, STBwasobtainedbylasermicrodissectionofchorionvilli (twoearly and two late placental samples) and dissolved for RNA extraction (27).Total RNA extraction and real-time quantitative RT-PCR were performed asdescribedpreviously(5,26, 27).Inbrief,RNA wasextractedaccordingtothe acid guanidinum thiocyanate-phenol-chloroformmethod,andreversetranscription was performed at 40C for 15 min using random hexamerprimers(LifeTechnologies/AppliedBiosystems)andmurineleukemiavirusreversetranscriptase(Promega). Assays-on-DemandGeneExpres-sion probes (Applied Biosystems) (FASLG, assay ID Hs00181226_g1; andTNFSF10,assayIDHs00921974_m1)wereused.Multiplexedreal-timePCR assay with 18S rRNA as an endogenous control (Life Technologies/AppliedBiosystems)wasperformedona7900HTinstrument withthefactorydefaultthermocycleconditionsfor40cycles,andrawdata wereanalyzed using a relative quantitation method. Gene expression levels arepresentedasarelativefoldchangebetweenthestudiedsamples. Eachgenetest includedPBMCstimulatedbyPMA/ionomycinasapositivecontrolandanegativecontrolomittingadditionoftemplate.ImmunohistochemistryChorionicvilliwere xed in4%paraformaldehydefor2h,washed inPBSwith3.5%sucroseand1%shgelatin, andsnap-frozeninliquidnitrogen.Eight-micrometer cryosections were stained as described previously (22, 26).FcgRswereblocked byamixture ofhorseserum(ImmPRESSreagentkit)and 10% normal human serum (AB+), aldehyde groups were blocked by PBSwith 0.1 M glycine, and 0.03% H2O2 in PBS was used for extinguishing theendogenous peroxidase activity. ImmPRESS anti-mouse or anti-goat Reagentkit (Vector Laboratories) wereusedfor enhanceddetectionwith3,39-dia-minobenzidinetetrahydrochloride(DAB)assubstrate. Serial sectionswereincluded as negative controls stained by exchanging the rst Ab with isotype-matched control mAbs or normal goat serum for the polyclonal primary Abs.IEMofchorionicvilliFreshlyisolatedplacental samples wereimmediatelyxedin4%para-formaldehydefor4h, washedin0.1Mphosphatebuffercontaining7%sucrose and 0.05% saponin at 4C overnight, and snap-frozen. Cryosectionswereprocessedbyindirectimmunoperoxidasemethod, aspreviouslyde-scribed(22, 27) usingmAbcloneG247-4against FasLandgoat anti-human Ab (K18) against TRAIL. ImmPRESSanti-mouse or anti-goatReagent kit (VectorLaboratories) wasusedforenhanceddetectionwithDABas substrate. One-percent normal humanserum(AB+) and0.5%BSAinPBSwasusedtoblockFcRsonSTBinallblockingsteps.Anti-goat andanti-rabbit ImmPRESSreagentsandDABwereusedtorevealspecicstaining. Ultrathinsections wereexaminedinaZeiss EM900electronmicroscope(CarlZeiss)withoutadditionalstaining.Sectionsin-cubatedwithisotype-matchedcontrol mAbs or afnitypuriednormalgoatserumwereusedasnegativecontrol.PlacentalexplantculturesExplantsoffreshlyobtainedearlyplacentaswereimmediatelyprocessedand put in culture within 1 h after extraction as described in Ref. 28. In brief,5516 PLACENTALEXOSOMESEXPRESSAPOPTOSIS-INDUCINGFasLANDTRAIL at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from 10 mg wet weight pieces of chorion villi were cultured in ultracentrifugedmedium of RPMI 1640 supplemented with 0.5% BSA (Sigma Aldrich) andantibioticsat 37Cinhumidatmospherewith5%CO2. Supernatant washarvestedafter 24 hin cultureandused forexosome isolation.IsolationofexosomesSupernatants fromshort-termplacental explant cultures fromindividualdonors wereclearedfromcell debris andlarger particles bysequentialcentrifugations at 4,000 3 g for 30 min and 17,000 3 g for 25 min. Afterltrationthrougha0.2-mmlter, thesupernatant wasultracentrifugedat110,0003gfor2h. Thepellet wasresuspendedinsterile, particle-freePBSandultracentrifugedthrougha20and40%discontinuous sucrosegradient; the exosomes werecollected and washed a couple oftimes withsterile PBS. The yield of exosomes was estimated with bicinchoninic acidprotein assay kit (Pierce). Isolated exosomes were stored at 280C in PBSor radioimmunoprecipitation assay buffer (Pierce) supplemented withproteaseinhibitormixture(RocheDiagnostics)untiluse.WesternblotanalysisIsolated placental exosomes, snap-frozen pieces of chorion villi, and culturedJurkat cells were solubilizedwithradioimmunoprecipitationassay buffer(Pierce). Afterseparationon12%SDS-PAGE, proteinsweretransferredtopolyvinylidene diuoride membranes (GE Healthcare). Membranes wereblocked for 1 h with 5% fat-free powdered milk in PBS with 0.05% Tween20 (PBST) before incubation overnight with the appropriate Abs. HRP-conjugatedsecondaryAbswereappliedfor1h.PBST wasusedinallwashing steps. Signals were detected with Amersham ECL plus Western blotdetectionsystemwithAmersham ECL developinglms(GEHealthcare).ImmunoowcytometryofproteinexpressiononthesurfaceofisolatedplacentalexosomesForanalyzingtheexpressionofproapoptoticmoleculesonthesurfaceofisolatedplacentalexosomes,surfactant-freeultraclean4-mmsulfatelatexmicrobeads (Interfacial Dynamics) were used as described previously (27,29). Thebeads werecoatedwithantiFasL-, antiTRAIL-, or isotype-matchedcontrol Abs. After blockingof uncoupledsites, isolatedexo-someswereaddedandincubated4Covernightwithend-to-endrotation.After washing, FITC-conjugatedCD63mAbwas incubatedfor 30minbefore nal washing. The beads were analyzedby owcytometry onFACScan(BDBiosciences)usingCellQuestsoftware.IEMofproteinexpressiononthesurfaceof isolatedplacentalexosomesIsolatedexosomeswereabsorbedonformvar/carbon-coatednickelgrids,washed with PBS, and xed with 2% paraformaldehyde in PBS for 10 min.Negative staining was performed on the grids using 1.9% methylcellulosecontaining 0.3% uranyl acetate.Excess uid wasremoved andallowed todry before examination with electron microscope. For IEM, after blockingin 0.1 M glycine and 0.3% BSA, grids with exosomes were incubated withappropriate monoclonal or polyclonal Abs, isotype-matched control Abs, orultracentrifuged goat serum for 1 h in wet chamber. After washing, the gridswereincubatedwith secondaryAbsconjugatedwith 5-or10-nmgold par-ticles for 1 h. After washing and additional xation with 2.5%glutaraldehyde,the grids were negatively stained as described previously (27). In double IEMstaining after completed rst staining with anti-Fas and after a blocking stepwith goat serum, the second IEMstaining for TRAILwas performed.Conjugation of the secondary Abs with 5- or 10-nm gold particles was usedto distinguishbetweenexosomalFasLandTRAIL staining.IsolationandactivationofPBMCandcultureofJurkatcellsPBMCfromhealthydonors were isolatedbyLymphoprep(Nycomed)gradient centrifugationandusedfor detectionof exosome-inducedapo-ptosis. OnemillionisolatedPBMCwereactivatedfor 24hwithPMA/ionomycinas previouslydescribed(27, 28), andthe PBMCactivationwasmeasuredbyowcytometryofFasandTRAIL-R1andTRAIL-R2expression.Jurkatcells,purchasedfromtheAmericanTypeCultureCol-lection, wereculturedinRPMI 1640medium, supplementedwith0.5%BSA(SigmaAldrich)andantibioticsat 37Cinatmosphereof5%CO2andhumidityandusedinapoptosisdetectiontests.AssessmentoftheapoptoticeffectofplacentalexosomesAnnexinV:PEApoptosisDetectionKitI(BDBiosciences)andimmuno-owcytometrywereusedfordetectionofexosome-inducedapoptosisonJurkatcells, accordingtothemanufacturersinstructions.Inbrief, Jurkatcells(1 3 106/ml)wereculturedina96-wellplate,treated withdifferentconcentrations of isolated placental exosomes for 24 h, harvested, washedwith PBS, stained with Annexin V/7-aminoactinomycin D according to themanufacturersdescription,andanalyzedbyowcytometryonFACScan(BDBiosciences). ActivatedFas-expressingPBMCwereincubatedwithexosomes for 24 h in RPMI 1640 medium, supplemented with 0.5% BSA(SigmaAldrich) andantibioticsandthereafter analyzedbyelectronmi-croscopy(EM)toassessapoptoticsignsintheirmorphology.StatisticalanalysisTheresults, appropriateforstatisticalanalysis, weresubjectedtotwostatisticaltestsStudentttestand two-wayANOVA performedwith Graph-PadPrism 5program.ResultsFasLandTRAILaretranscribedandexpressedinearlyandtermhumannormalplacentasFig. 1A shows the relative mRNA expression of FasL and TRAILinSTBofearlyandtermpregnancyplacentasassessedbyreal-timequantitativeRT-PCR. STBfromveindividual samplesofearly and term normal pregnancy placentas were analyzed. As canbe seen, mRNAfor both FasLand TRAILwas found in allsamples conrming these molecules expression in early and termplacentas. FasLshowedhigher mRNAexpressionlevel inearlypregnancyplacentas, whereas TRAILmRNAdominatedinex-pressionintermplacentas; however, thesedifferenceswerenotstatisticallysignicant (Fig. 1A). Fig. 1BillustratesFasLandTRAILexpressionat theproteinlevel, revealedbyimmunohis-tochemical stainingandlight microscopyofserial sectionsofchorionicvilli. Positivestainingforbothmoleculeswasseeninthe villoustrophoblast,preferentiallyintheSTBofthechorionicvilliandonsomeHofbauercells(22)inthevillousstroma. Thestainingwasstrongandseemedtobeintracellularlylocalized.Inlight microscopy, theDAB-reactionproductgavetheimpressionthatbothSTBandcytotrophoblast(CTB)werestained;however,the following ultrastructural analysis by IEM excluded staining ofCTB. Nostainingwas observedinnegativecontrolsperformedwithisotype-matchedcontrol Abs. Theseresultswereconsistentinall stainedsamples(n=4). Similar resultswereobtainedintermplacentas(n=3;datanotshown).Fromtheseexperiments,weconcludethatFasLandTRAILareconstitutivelytranscribedandintracellularlyexpressedas proteins inthe STBof humanearlyandtermplacentas.IEMrevealedFasLandTRAILexpressiononintraluminalmicrovesiclesintheMVBofSTB,consistentwiththeendosomalpathwayofexosomebiogenesisWeassessedthepreciselocalizationofFasLandTRAILinpla-centasontheultrastructural level byindirect immunoperoxidasestainingandEM. Four individual earlyplacental samples wereexamined, andnosignicant differencesinthestainingpatternswere observed. Representative photomicrographs, illustrating IEMstaining for FasL and TRAIL, are shown in Fig. 2. Both moleculesexhibited a similar intracytoplasmic staining pattern. The stainingofFasL(Fig. 2A, 2B)andTRAIL(Fig. 2C, 2D)intheplacentawas mainlyrestrictedtotheSTBof thechorionicvilli andinscarceindividual Hofbauercells(22)inthevillousstroma(datanotshown).NostainingwasobservedintheCTBcells.Further-more, the apical microvillous (Fig. 2A, arrowheads) and the basalmembranal surfaces of STBwere generallydevoidof stainingindicatingthatthere wasno expression ofthesemoleculeson theapical and basal syncytiotrophoblastic membrane. Instead, thelimitingmembraneofcytoplasmicvacuolesandnumerousmicro-vesiclesof406080nmindiameter tightlypackedinsidethevacuoles werepositivelystainedfor FasL(Fig. 2A, 2B) andTheJournalofImmunology 5517 at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from TRAIL2(Fig. 2C, 2D). The stainedcytoplasmic vacuoles dis-played the classical morphology of MVBs (Fig. 2A, 2B, 2C, 2D).In all identied MVBs, the number of intraluminal FasL- orTRAIL-stainedmicrovesicleswasbyfarhigherthan9,anumberdenedas thelower limit of vesiclecontent distinguishingbe-tweenearlyandlateendosomes(30).TheMVBswerelocatedatdifferent levelsinthesyncytioplasmandwerebetween600800nmandupto1.52mminsize.Severalofthemwerefusedwiththe apical microvillous membrane of the STB, as illustrated in Fig.2D, whereaMVB, positivelystainedforTRAILonitslimitingmembrane, isopeningtotheapical intervillousspaceandre-leasingits microvesicle content. The stained limiting membrane oftheMVBstogetherwiththestainedintraluminal vesiclessuggestvesiclegenerationbyinwardbuddingof theMVBs limitingmembraneconsistent withbiogenesisofexosomes. Insummary,our IEManalyses demonstratedthat bothFasLandTRAILaretypicallylocalizedonintraluminal vesicles withthesizeandmorphology of exosomes enclosed in the MVBs of the lateendosomal compartment. Moreover, therewerefrequent mor-phological signsof microvesiclesecretion/releaseintheinter-villousapicalspaceoftheSTB.SecretedplacentalexosomesexpressexclusivelybioactiveaggregatedformofFasLandTRAILTofurtherconrmtheIEMresultsofgenerationandsecretionofFasL- and TRAIL-bearing exosomes, 24-h cultures of placentalexplantswereestablishedinthepresenceofmetalloproteaseinhib-itors.Theshorttimeofculturewaschosentoensuretheexplantsgood condition throughout the culture avoiding microvesicle releasebecauseofnecrosis.Thus,thevastmajorityofnanovesiclesintheculturesupernatantemanatefromexosomesecretion.Theexo-somes, isolated and enriched by sucrose gradient ultracentrifugationfromtheculturesupernatant, weresubjectedtoanalysisforFasLandTRAILwiththreedifferent methodsWesternblot analysis,immunoowcytometry, andIEM. TheresultsfromrepresentativeexperimentsaresummarizedinFig.3.FIGURE 1. FasL and TRAIL are constitutivelyexpressedbyhumannormal placentathroughoutpregnancy. (A) Quantitative real-time RT-PCRmRNAexpressionof FasLandTRAILinSTB,isolated or laser microdissected from early (n = 5)andterm(n=5) placenta. TherelativemRNAexpressionwas determinedbycomparisonofmRNA expression in term pregnancy placenta tomRNAexpressioninearlypregnancyplacenta(relativeexpression=1). PMA/ionomycin-stim-ulatedPBMCservedas positive control. Allsamples were normalizedto18SrRNAas de-scribedinMaterialsandMethods.(B)Immuno-histochemical stainingof earlynormal placentashowing the presence ofFasLandTRAIL inthevillous trophoblast and in separate Hofbauer cellsin the stroma of the chorionic villi. Serial sectionsstained with isotype-matched irrelevant mouseanti-human mAb and normal goat serum served asnegative controls. Note the dotted staining of thetrophoblastthatseemstobeintracellularlylocal-ized. Arrows point to the chorionic villi coveredwith stained villous trophoblast; arrowheadspointtoHofbauercellsandstarstothestromaltissuecomposedoflooselyconnectedstromalcells. Originalmagnication 3400.V,Placentalvessel located in the stroma of a chorionic villous.5518 PLACENTALEXOSOMESEXPRESSAPOPTOSIS-INDUCINGFasLANDTRAIL at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from Isolated placental exosomes, placental tissue, and a positivecontrol of lysedJurkat Tcells were usedinthe Westernblotanalyses of FasL and TRAIL molecules (Fig. 3A). The tetraspaninCD63 was analyzed as an exosomal control. As can be seen, FasLwas expressedinall exosomal fractions exclusivelyas asingleprotein band of 75 kDa corresponding to a membranal hex-americformofFasL, composedoftwotrimericmolecules(11),whereasinplacental tissue, botha37-kDaandahexameric75-kDaFasLwererevealed(Fig.3A).Incontrast,lysedJurkatcellsshowedadominatingbandof37kDa. Similarly, TRAILwasenrichedinthe exosomal fractions andplacentas as a 70-kDaband, whereasthemajor amount of TRAILinJurkat wasex-pressed as a 34-kDa band and only a smaller amount as a 70-kDaproteinband.Toourknowledge,thisistherst demonstrationofFasL and TRAIL in human placental exosomes, showing that bothFasLandTRAIL,themajorapoptosis-inducingmoleculesoftheTNFsuperfamily, arenot onlysecretedbyexosomesbut exclu-sivelyenrichedasanoligomerizedmembranalbioactiveform.Tofurther provethat thelocalizationof FasLandTRAILisindeedontheexosomalmembrane, experimentswereperformedby immunouorescence staining and immunoow cytometry withexosomesloadedonlatexbeads (Fig. 3B) andbyimmunogoldstainingof exosomes andIEM(Fig. 3C). Representativehisto-grams of the ow cytometric analysis, shown in Fig. 3B, revealedmembranalFasLandTRAILexpression.InFig.3C,usingnega-tivecontraststaining, weillustratethetypicalexosomalsizeandcup-shapedmorphologyoftheisolatedmicrovesicles, thusprov-ingthat theyareexosomes. Besidesamorphological character-istic, the positive immunogoldstainingof CD63andplacentalalkalinephosphates further depict themas exosomes of humanplacental origin. IEMstainingofFasLandTRAILwasrevealedonthesurfaceoftheexosomes, conrmingthat thesemoleculeswereexpressedasmembranal proteins. WemadeadoubleIEMstaining with FasL and TRAIL using 5- and 10-nm gold particles,conjugatedtosecondaryAbs todistinguishbetweenFasLandTRAILlocalization. Interestingly, wefoundFasL(stainedby5-nm gold particles) and TRAIL (stained by 10-nm gold particles)ondifferentexosomes,indicatingthatatleasttoacertainextendtheymightbegeneratedinseparateMVBs.Insummary, withthreeindependent methodsWesternblot,immunoowcytometrywithlatexbeads, andIEMwedemon-strate that the placenta secretes exosomes that carryFasLandTRAIL, enrichedontheexosomal membraneas singleproteinbandswithsizesconsistentwiththeirbioactivemembranalform.FIGURE 2. Immunoelectron microscopic localization of FasL andTRAILinSTBofhumanearlyplacentas. (A)Low-powermicrographofSTB stained for FasL. The electron-dense reaction product labels MVB inthesyncytioplasm(arrow). Notethat theapical surfacemembraneisnotstained(arrowheads).(B)High-powermicrographofFasL-positiveMVBcontaining tightly packed and intensively stained intraluminal micro-vesicles (arrowheads). (C) Micrograph of TRAIL-positive MVB (arrow) inperinuclear locationshowingstainedlimitingmembraneandinternalmicrovesicles. (D)High-powermicrographofanapicallylocatedMVB,which exhibits staining of the limiting membrane and internal micro-vesicles (arrowheads) and forms an opening to the intervillous space (star).Note that apical plasma membrane is TRAIL negative (arrow). Scale bars,1mm(A);400nm(B);and200nm(C,D).N,Nucleus.FIGURE 3. FasL and TRAIL are enriched on the membrane of placentalexosomes.(A)WesternblotanalysesofFasLandTRAILproteinexpres-sionbyexosomesfromtwodonors, isolatedfromsupernatantsof earlyplacentaexplantcultures, comparedwithexpressioninearlyplacentatissueandJurkatcells.CD63wasusedtoconrmtheexosomaloriginofthe isolated microvesicles. Protein load, 40 mg/well. (B) Immunoowcytometryof puriedplacental exosomes capturedonlatexmicrobeadscoatedwithmAbsagainstFasL andTRAILandrevealedbyuorescencestainingwithmAbs against theexosomal marker CD63. Superimposedshadedhistogramrepresents negative controls withisotype-matchedmAbs. (C) IEMof whole-mount placental exosomes. The typical cup-shaped morphology andsizeof30100 nmisshownbynegative contraststainingseentotheleft.SpecicmAbsandimmunogoldlabelingwith5- or 10-nm gold particles were used to reveal surface protein expression.Anti-CD63stainingconrmstheexosomal natureoftheisolatedmicro-vesicles, andtheirplacentaloriginisrevealedbystainingforplacentalalkaline phosphatase (PLAP). FasL staining was performed with 5-nm goldparticles and TRAIL with 10-nm particles. Note that in the double staining,FasLandTRAILarepresentonseparateexosomes.Scalebars,100 nm.TheJournalofImmunology 5519 at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from FasL-andTRAIL-expressingplacentalexosomesinduceapoptosisin vitroinadose-dependentmannerWestudiedtheproapoptoticabilityof exosomes fromseparateplacental samples in vitro in two model systemsJurkat leukemiaTcellsandactivatedPBMCfromhealthydonors. Theapoptosiswas revealedbyAnnexinV:PEApoptosis DetectionKit I (BDBioscience) andbyEMlookingfor earlysigns of apoptosis inactivated PBMCs incubated with placental exosomes. Fig. 4A and4Bsummarizestheresultsof Jurkat cell experimentswithexo-somesisolatedfromculturesupernatantsofindividual placentalexplantsandFig. 4CGtheresultsfromexperimentswithacti-vatedPBMC. Ascanbeseen, theplacental exosomesinducedstatistically signicant apoptosis in Jurkat cells in a dose-dependentmanner (n=7; Fig. 4A). Incontrast, theexplant culturesuper-natantsdepletedof microvesicleaftergradient ultracentrifugationwereunabletotriggerapoptosisofthetargetcellsindicatingthattheapoptoticeffect wasexosomespecic(datanot shown). Thein vitro proapoptotic effect on the Jurkat cells could differ betweenindividual exosomesamples(Fig. 4B). Althoughmost exosomalsamples hadagoodproapoptoticcapacity, somehadlower in-tensity, andtwosampleshadveryloweffect. Thereasonforthevariation in the apoptotic capacity is at present not known but maybeexplainedbyseveralfactorsforexample, lessconcentrationofapoptoticmoleculesonthesurfaceofsomeisolatedexosomesbecauseofinsufcient protectiveeffect oftheproteaseinhibitormixtureduringisolationandstorage; variations becauseof dif-ferencesintheSTBor inthegestational stageof theplacentalsamples; variations because of the condition of the samples andof the explant cultures performance; and/or biological differencesbetweenthe donors. Exosomal samples, unprotectedbymatrixmetalloproteinase(MMP) inhibitors lost their FasLandTRAILexpressionandtheirapoptoticabilityasshowninSupplementalFig. 1. Thepresenceofmetalloproteaseinhibitorsintheculturemediumduringstorage andexperimental procedures withexo-somes is essential for the preservation of their proapoptotic effect.Despite a biological variation in the apoptotic potency in individualsamples, themeanvalue6SDofinducedapoptosiswasstatisti-callysignicantanddosedependent(Fig.4A).In addition, we studied the in vitro apoptotic effect of exosomeson PBMCfromhealthy donors (n = 5) activated by PMA/ionomycin. Theactivationwasassessedbymeasurement ofFasandTRAILreceptorsbyimmunoowcytometryasillustratedinatypicalexperimentinFig. 4CEwhereFasandTRAILrecep-torexpressionisshownondifferentsubpopulationsofactivatedPBMC from one donor. Apoptosis was assessed by Annexin V:PEApoptosisDetectionKit (Fig. 4F)andEM(Fig. 4G). AscanbeseeninarepresentativePBMCexperiment(Fig.4F),thenumberofAnnexinVexpressingapoptoticcellsisdoubledinthepres-enceof40mgexosomes.IntheEManalyses,wesawcellswithintactcellularmembranebutwithclearearlyapoptoticsignslikefragmentation of the nucleus, vacuolization of the cytoplasm, andmembraneblebbingasillustratedinFig.4G.Takentogether,ourresults indicate that human placenta secretes exosomes withmembranalexpressionofpreformedfunctionalFasLandTRAILmolecules that are able torapidlytrigger apoptosis invitroinJurkatTcellsandinactivatedPBMCfromhealthydonors.DiscussionThree main points summarize the results presented in this paper: 1)IntheSTBofhumanplacenta,theapoptosis-inducingmoleculesFasLandTRAILareconstitutivelytranscribedandexpressedasproteins on the MVBs membrane andon the membrane ofintraluminal vesicles that are released in the extracellular space asexosomes.TheproteinexpressionofFasLandTRAILis,thus,strictlyconnedtotheendosomal compartment ofSTB, whereastheapicalandbasalsyncytiotrophoblasticmembranesarecom-pletely devoid of FasL and TRAIL expression. 2) The STB-derivedexosomes enrich FasL and TRAIL on their membrane as bioactiveoligomerized molecules able to form DISCs. 3) FasL- and TRAIL-carrying placental exosomes trigger apoptosis in Jurkat T cells andactivated PBMC from healthy donors ina dose-dependent manner.Takentogether, theseresults indicatethat humanearlyandtermplacentaconstitutivelysecretefunctionalFasLandTRAILonex-osomes that are able to convey apoptosis and thus might contributetoprotectionofthefetoplacentalunitfromactivatedmaternalim-munecells.The intracellular localizationof FasLandTRAILinhumanplacenta is intimately bound to the biogenesis of exosomes. To ourknowledge, thisistherst demonstrationof TRAILproteinex-pression on the ultrastructural level and the rst report that TRAILissolelyexosomallyexpressedinSTBof humanplacenta. Fur-thermore, aconstitutivereleaseof FasL-andTRAIL-expressingexosomesfromtheapical microvilloussurfaceofSTBwasdem-onstrated. Our resultsareintunewithpreviousinvestigationsbyothers and us (22, 31, 32) showing that FasL is targeted to the MVBof secretorylysosomesandisexpressedonexosome-likemicro-vesicles(5,15,33,34).Twoindependentpathwaysthatcontrolthe internalization of FasL on microvesicles in MVBs have beenidentiedphosphorylationofFasLbybindingofitsproline-richdomaintoFgr, Fyn, andLyntyrosinekinasesandadirectubiquitinylationat the lysine residues ankingthe proline-richdomain(3537). TheextracellulardomainofFasL, TRAIL, andother members of the TNF family contain cleavage sites forMMPsthatgeneratesolubleproteinsbyproteolyticsheddingandsubstantially reduce the proapoptotic activity of the shed proteins.ThemFasListheproapoptoticmolecule, whereassFasLis1000timeslessactiveininducingapoptosis(38,39).TranslocationofFasLtotheplasmamembraneasawaytoprovidemFasLwillexposethemoleculetoproteolyticcleavagereducingapoptoticactivity. Moreover, FasLexpressed on the cellular membrane,insteadofdampeningtheimmuneresponses, inducesinamma-tionandpromotes allograft rejection(19, 3942) as comparedwith sFasL that was proposed to be anti-inammatory (40, 41, 43).Thus, different molecular formsof FasLdomediatedifferentfunctions.Theexactmechanismsarenotcompletelyunderstood,but compartmental expression, levelsof membranal and/or shedform, and the surrounding microenvironment where it is expressedseemtoact together toinuence its biological functions (10).SimilarlytoFasLtherearedifferencesinTRAILactivityde-pending on whether it is in a membrane-bound form or cleaved togenerateasolubleform. Initially, it was foundthat bothformswere equally effective (44), whereas later reports singled outmembrane-boundTRAILas the molecular formpreferablyin-ducingcelldeath(4547). Recently, itwasreportedthatTRAILproteininSTBculturescouldbeat least partlylocalizedtothecytoplasm(18). Previouslyproposedfunctionasaproteinre-pairingsyncytialdamageandmaintainingtheplacentalbarrierintegritycouldnot beconrmed. Instead, asimilar functionforTRAILasforFasLasacontributortoimmuneprivilege wassug-gested, andanintracellular compartmental restrictionof TRAILexpression in the placenta similar to that of FasL was assumed butnot documented (18). In line with this assumption, we show thatTRAILis restrictedtoSTBs endosomal compartment andex-pressedandreleasedonexosome-likemicrovesicles. Theprevi-ouslydescribedcellmembraneexpressionon CTBandSTB(18)couldnot be conrmeddespitedetailedIEManalyses. Similarresults of exosomal TRAIL and FasL expression have been showninothercellsandsitessuchasdendriticandTcells(5, 48, 49),5520 PLACENTALEXOSOMESEXPRESSAPOPTOSIS-INDUCINGFasLANDTRAIL at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from melanoma,andovariancancercells(50,51)andinjointuidofrheumatoidarthritispatients(52). Interestingly, ourdoubleim-munogoldstainingofFasLandTRAILrevealedthesemoleculesonseparateexosomes, suggestingthat inhumanplacentathesemolecules might be enriched and processed through separateMVBs. Thisisconsistent withapreviousndingofonlypartialcolocalizationofendosomal FasLandTRAILexpressioninhu-manmelanoma(46).FIGURE4. Humanplacenta-derivedexosomescarryfunctional FasLandTRAILontheir membrane. (AandB) FasL- andTRAIL-bearingplacentalexosomes trigger apoptosis of Jurkat T cells in a dose-dependent manner. (A) Staple diagram showing average percentage 61 SD of dose-dependent apoptosisin Jurkat cells induced by different concentrations of exosomes isolated from supernatant of placental explant cultures (n = 7). (B) Induction of apoptosis byindividualexosomalpreparationsderivedfromexplantculturesfromsevenconsecutivesamplesofhumanearlyplacentafromnormalpregnanciesshowingbiological variationinapoptoticactivity. (CG)FasL-andTRAIL-expressingexosomestriggerapoptosisinactivatedPBMCfromhealthydonors. (CE)ImmunoowcytometryshowingFas-andTRAIL-R1and-R2expressingcellsinsubpopulationsofPBMCfromhealthydonors, activatedbyPMA/ion-omycin. (F)ImmunoowcytometryanalysisofapoptoticcellsinactivatedPBMCbeforeandafteradditionof20mgplacental exosomes. Notethat thepercentage of the apoptotic PBMC is doubled in thepresence of FasL- and TRAIL-bearingplacental exosomes.(G) Electronmicrographof a representativeexperiment out of three, showing classical apoptotic signs of nuclear fragmentation, vacuolization of the cytoplasm and plasma membrane blebbing of activatedPBMCincubated withFasL- andTRAIL-bearing placentalexosomes.Originalmagnication 35000.***p=0.0001;**p=0.0016; *p=0.0498.TheJournalofImmunology 5521 at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from Soluble and membranal FasLand TRAILmolecules haveseparatebiologicactivities;thus,exosomallyexpressedFasLandTRAILmaybeviewedasasolubleformthatpreservesmem-branalexpressionandaproapoptoticbiologicactivity. Usingthehighlyspecicanti-FasLmAbG247-4(53),werevealedasingleFasLproteinbandof 75kDa inthe exosomal fractions anda single TRAIL protein band of 70 kDa. FasL and TRAIL, as wellasothermembersoftheTNFsuperfamily, aremembranallyex-pressed as homotrimeric molecules. However, aggregation by self-inducedoligomerizationisnecessaryforformationofDISCandrapidtriggering of apoptosis. Holler et al. (11) provedthat a hex-americFasL, consistingof twohomotrimersat closeproximity,represents the minimal ligand structure required to form DISC andsignal apoptosis. Both trimeric and hexameric FasL bound equallywelltoFasbuttrimericFasLfailedtoinduceaDISCandisthusinefcient intriggeringof apoptosis (11). AhexamericFasLwouldhaveam.w.between75and80kDadependingonglyco-sylationlevel. Wesuggest that thesingle75kDaFasLandthe70kDaTRAILproteinbandsintheplacentalexosomalfractionsrepresent twohomotrimericmembranal moleculesaggregatedina bioactive hexameric form of FasL (11) and TRAIL, respectively.Oursuggestionisbasedonthefactthatinourexperimentsexo-somes carrying the 75-kDa FasL and 70-kDa TRAIL were able totrigger exosomal dose-dependent apoptosis in Jurkat cells andactivatedPBMC,implicatingthat afunctionalDISCwas formed.When the expression of these oligomerized ligands was lost fromthe exosomal membrane, the proapoptotic ability was lost as wellasshowninSupplementalFig.1.Thefunctionalexperimentsarein-tunewithapoptosistriggeredbyexosomesfromothersourcessuchas activatedTcells andDCcells (5, 15, 34, 48, 49) andvarious cancer cells (31, 33, 50, 51, 54). To our knowledge, this isthe rst report of expression of preformed oligomerized FasL andTRAILonexosomal membrane. Thesphingolipid-, cholesterol-,andtetraspanin-richexosomalmembraneemanatesfromtheendo-somal compartment andresemblesinitscompositionthecellularmembraneslipidrafts, knowntopromoteattachment andaggre-gation of signaling proteins as well as recycling (55). Therefore, theexosomal membrane is exceptionally suitable for the oligo-merization and selective enrichment of biologically active TNFsuperfamilyligands, abletoformDISC(11). Limitingtheex-pressionofoligomerizedFasLandTRAILtoexosomesinMVBprotectsthesemoleculesfromMMPcleavageandkeepsthemina molecular form ready to trigger cell death in stringently denedevents such as delivering of cytotoxic hit to a target cell (5, 15,48)orprotectionofspecictissues/organsorsitesfromanimmu-nologicattack(9), renderingthesesitesanimmunologicprivilege.In contrast to our results, conrmed in other settings (5, 15, 3134, 4951, 54), Abrahams et al. (21) couldnot detect surfaceexpression of FasL on trophoblast cell linederived microvesiclesandcouldonlyinduceapoptoticactivitywhenthevesiclesweredisruptedbyTritonXtreatment, concludingthat FasLwas lo-calizedinsidethe microvesicles. Thereasonsforthisdiscrepancyarenot clear but might haveseveral explanations, for example,lossofapoptoticactivitybecause ofsheddingbyproteolyticcleav-age. Protectionagainstmetalloproteaseactivityisofcrucialimpor-tance,becausethedifferentformsofthesemoleculesmembranalhomotrimeric, oligomerized, and shedall have separate biologicalfunctions andfates. Theretrieval of a37-kDaproteinbandafterTritonXmembranedisruptiontreatment(21)cannotbethemem-branal exosomal FasLdescribedbyus inthis study. Maybe itemanatesfromcross-linkingortrimerizationoftheTritontreat-mentproduced sFasL. It has been reported that, if the cleavage ofmFasL occurs by MMP7 at a particular protease cleavage site, thegenerated sFasL may form functional trimers (11, 56, 57). Anotherexplanationcouldbethat wehavestudieddifferent microvesiclesbecausetheywereisolatedfromdifferentsourcesandbydifferentisolationprocedures.The physiological importance of apoptosis in normal pregnancyis associated with the fetal immune privilege in the pregnant uterusa phenomenon similarly regulated as other immune privileged sites,the eye being the one more thoroughly studied (9, 58, 59). We canspeculate that the release of proapoptotic placental exosomes is inhighest concentrationnear theplacentaandbuildsanexosomalgradient inthematernal bloodnearest theplacenta, whichfunc-tionsasaprotectiveshieldofthefetoplacental unit. Theresultspresented in this study suggest that the placenta secretes exosomesexpressingpreformedbioactiveFasLandTRAILmoleculesabletoconveyapoptosisandmight partlycontributetotheimmuneprivilege of the fetus. Thus, the immunomodulatory and protectiverole of the placenta seems at least partly associated to its exosome-secretingability.Several published reports (reviewed in Ref. 25), including this one,concern normal placenta and depict exosomal secretion as benecialtopregnancy. Alogical questionishowtheexosomal secretionisaffected in pathological pregnancies. So far, there are few conclusiveexosomestudies regardingsuchconditions. Inpretermpregnancy,exosomal concentration in the maternal peripheral blood was shownto be lower compared with that in women with normal term deliveryand nonpregnant women (60). Previous studies (61) of preeclampsiahighlightedanother type of extracellular microvesicles, the muchlarger shedding microvesicles generated from the apical microvilloussyncytial membrane, and showed that they are highly elevated in thecirculation of preeclamptic women and seemed to be involved in theenhancedinammation,cytokineproduction,andendothelialde-structionanddysfunctionthat arecardinal characteristicforthisenigmaticdisease. One can speculatethat there may be a balanceintheSTBbetweengeneratingsheddingvesiclesandgeneratingexosomes,andadisturbanceofthisbalancebyoverproducingorunderproducingeither typeof extracellular STBvesicles mightaffectthene-tuningoftheplacentalfunction. Manymorewell-designedfuturestudiesofsyncytiotrophoblastmicroparticlesandexosomesinpathological pregnancyconditionsareneededtoelucidatethisquestion.AcknowledgmentsWethankthedonorsandthecolleaguesat theDepartmentofObstetricsandGynecologyandthestaff of theoperationtheater CentralopIIatUmea UniversityHospital, Umea, Vasterbotten, Sweden, for providingplacentalsamples.DisclosuresTheauthorshavenonancialconictsofinterest.References1. Mincheva-Nilsson,L.2006.Immunecellsandmoleculesinpregnancy:friendsorfoestothefetus?ExpertRev.Clin.Immunol.2:457470.2. Sharp, A. N., A. E. Heazell, I. P. Crocker, and G. Mor. 2010. Placental apoptosisinhealthanddisease.Am.J.Reprod.Immunol.64:159169.3. Huppertz, B., H. G. Frank, J. C. Kingdom, F. Reister, andP. Kaufmann. 1998.Villouscytotrophoblastregulationofthesyncytialapoptoticcascadeinthehu-manplacenta.Histochem.CellBiol.110:495508.4. Mor, G., S. Straszewski, andM. Kamsteeg. 2002. Roleof theFas/Fasligandsysteminfemalereproductiveorgans: survival andapoptosis. Biochem. Phar-macol.64:13051315.5. Mincheva-Nilsson, L., O. Nagaeva, K. G. Sundqvist, M. L. Hammarstrom,S. Hammarstrom, andV. Baranov. 2000. gdTcellsofhumanearlypregnancydecidua:evidenceforcytotoxicpotency.Int.Immunol.12:585596.6. Clark,D.A.2005.Tolerancesignalingmolecules.Chem.Immunol.Allergy89:3648.7. Ashkenazi, A., andV. M. Dixit. 1998. Deathreceptors: signalingandmodula-tion.Science281:13051308.5522 PLACENTALEXOSOMESEXPRESSAPOPTOSIS-INDUCINGFasLANDTRAIL at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from 8. Russo, M., A. Mupo, C. Spagnuolo, andG. L. Russo. 2010. Exploringdeathreceptorpathwaysasselectivetargetsincancertherapy. Biochem. Pharmacol.80:674682.9. Green, D. R., andT. A. Ferguson. 2001. Theroleof Fas ligandinimmuneprivilege.Nat.Rev.Mol.CellBiol.2:917924.10. Lettau, M., M. Paulsen, H. Schmidt, andO. Janssen. 2011. Insights intothemolecularregulationofFasL(CD178)biology.Eur.J.CellBiol.90:456466.11. Holler, N., A. Tardivel, M. Kovacsovics-Bankowski, S. Hertig, O. Gaide,F. Martinon, A. Tinel, D. Deperthes, S. Calderara, T. Schulthess, et al. 2003. Twoadjacent trimericFasligandsarerequiredfor Fassignalingandformationofadeath-inducingsignalingcomplex.Mol.Cell.Biol.23:14281440.12. Vargo-Gogola, T., H. C. Crawford, B. Fingleton, andL. M. Matrisian. 2002.Identicationofnovelmatrixmetalloproteinase-7(matrilysin)cleavagesitesinmurineandhumanFasligand.Arch.Biochem.Biophys.408:155161.13. Hymowitz, S. G., M. P. OConnell, M. H. Ultsch, A. Hurst, K. Totpal,A. Ashkenazi, A. M. de Vos, and R. F. Kelley. 2000. A unique zinc-binding siterevealedbyahigh-resolutionX-raystructureofhomotrimericApo2L/TRAIL.Biochemistry39:633640.14. Martinez-Lostao, L., I. Marzo, A. Anel, and J. Naval. 2012. Targeting theApo2L/TRAILsystemfor the therapy of autoimmune diseases and cancer.Biochem.Pharmacol.83:14751483.15. Mart nez-Lorenzo, M. J., A. Anel, S. Gamen, I. Monlen, P. Lasierra, L. Larrad,A.Pineiro,M.A.Alava,andJ.Naval.1999.ActivatedhumanTcellsreleasebio-active Fas ligand and APO2 ligand in microvesicles. J. Immunol. 163: 12741281.16. Hammer,A.,andG. Dohr.2000.Expression ofFas-ligandin rst trimesterandtermhumanplacentalvilli.J.Reprod.Immunol.46:8390.17. Phillips, T. A., J. Ni, G. Pan, S. M. Ruben, Y. F. Wei, J. L. Pace, and J. S. Hunt.1999.TRAIL(Apo-2L)andTRAILreceptorsinhumanplacentas:implicationsforimmuneprivilege.J.Immunol.162:60536059.18. Bai, X., J. L. Williams, S. L. Greenwood, P. N. Baker, J. D. Aplin, andI. P. Crocker. 2009. Aplacental protectiverolefor trophoblast-derivedTNF-relatedapoptosis-inducingligand(TRAIL).Placenta30:855860.19. Allison,J.,H.M.Georgiou,A.Strasser,andD.L.Vaux.1997.Transgenicex-pression of CD95 ligand on islet beta cells induces a granulocytic inltration butdoesnot conferimmuneprivilegeuponislet allografts. Proc. Natl. Acad. Sci.USA94:39433947.20. Roh, C. R., J. W. Lee, B. H. Kang, S. H. Yang, B. G. Kim, D. S. Bae, J. H. Kim,andJ. H. Lee. 2002. Differential expressionsofFasandFasligandinhumanplacenta.J.KoreanMed.Sci.17:213216.21. Abrahams,V.M.,S.L.Straszewski-Chavez,S.Guller,andG.Mor.2004.Firsttrimestertrophoblast cellssecreteFasligandwhichinducesimmunecell apo-ptosis.Mol.Hum.Reprod.10:5563.22. Frangsmyr, L., V. Baranov, O. Nagaeva, U. Stendahl, L. Kjellberg, andL. Mincheva-Nilsson. 2005. CytoplasmicmicrovesicularformofFasligandinhumanearlyplacenta: switchingthetissueimmuneprivilegehypothesisfromcellulartovesicularlevel.Mol.Hum.Reprod.11:3541.23. Chaput, N., andC. Thery. 2011. Exosomes: immunepropertiesandpotentialclinicalimplementations.Semin.Immunopathol.33:419440.24. Clayton, A., andM. D. Mason. 2009. Exosomes intumour immunity. Curr.Oncol.16:4649.25. Mincheva-Nilsson, L., and V. Baranov. 2010. The role of placental exosomes inreproduction.Am.J.Reprod.Immunol.63:520533.26. Stenqvist, A. C., T. Chen, M. Hedlund, T. Dimova, O. Nagaeva, L. Kjellberg,E. Innala, andL. Mincheva-Nilsson. 2008. Anefcient optimizedmethodforisolation of villous trophoblast cells from human early pregnancy placenta suitablefor functional andmolecularstudies.Am.J.Reprod.Immunol. 60:3342.27. Hedlund,M.,A.C.Stenqvist,O. Nagaeva,L.Kjellberg,M.Wulff,V.Baranov,and L. Mincheva-Nilsson. 2009. Human placenta expresses and secretes NKG2Dligandsviaexosomesthatdown-modulatethecognatereceptorexpression:ev-idenceforimmunosuppressivefunction.J.Immunol.183:340351.28. Mincheva-Nilsson, L., O. Nagaeva, T. Chen, U. Stendahl, J. Antsiferova,I. Mogren, J. Hernestal, andV. Baranov. 2006. Placenta-derivedsolubleMHCclassI chain-relatedmoleculesdown-regulateNKG2Dreceptor onperipheralbloodmononuclear cells duringhumanpregnancy: apossiblenovel immuneescapemechanismforfetalsurvival.J.Immunol.176:35853592.29. Lamparski,H.G.,A.Metha-Damani,J.Y.Yao,S.Patel,D.H.Hsu,C.Ruegg,and J. B. Le Pecq. 2002. Production and characterization of clinical gradeexosomesderivedfromdendriticcells.J.Immunol.Methods270:211226.30. Mari, M. M., M. V. Bujny, D. Zeuschner, W. J. Geerts, J. Grifth, C. M. Petersen,P. J. Cullen, J. Klumperman, andH. J. Geuze. 2008. SNX1denesanearlyendosomal recycling exit for sortilin and mannose 6-phosphate receptors. Trafc9:380393.31. Andreola, G., L. Rivoltini, C. Castelli, V. Huber, P. Perego, P. Deho,P. Squarcina, P. Accornero, F. Lozupone, L. Lugini, et al. 2002. Inductionof lymphocyte apoptosis by tumor cell secretion of FasL-bearing microvesicles.J.Exp.Med.195:13031316.32. Smith, A. J., J. R. Pfeiffer, J. Zhang, A. M. Martinez, G. M. Grifths, andB. S. Wilson. 2003. Microtubule-dependent transport of secretoryvesicles inRBL-2H3cells.Trafc4:302312.33. Jodo, S., A.M. Hohlbaum, S. Xiao, D. Chan, D. Strehlow, D.H. Sherr,A. Marshak-Rothstein, and S.T. Ju. 2000. CD95 (Fas) ligand-expressing vesiclesdisplayantibody-mediated, FcR-dependentenhancementofcytotoxicity. J.Immunol. 165:54875494.34. Monleon, I., M. J. Mart nez-Lorenzo, L. Monteagudo, P. Lasierra, M. Taules,M. Iturralde, A. Pineiro, L. Larrad, M. A.Alava, J. Naval, andA. Anel. 2001.Differential secretionof Fas ligand- or APO2 ligand/TNF-relatedapoptosis-inducingligand-carryingmicrovesiclesduringactivation-induceddeathof hu-manTcells.J.Immunol.167:67366744.35. Bossi, G., andG. M. Grifths. 1999. Degranulationplaysanessential part inregulating cell surface expression of Fas ligand in T cells and natural killer cells.Nat.Med.5:9096.36. Blott, E. J., G. Bossi, R. Clark, M. Zvelebil, and G. M. Grifths. 2001. Fas ligandistargetedtosecretorylysosomesviaaproline-richdomaininitscytoplasmictail.J.CellSci.114:24052416.37. Zuccato, E., E. J. Blott, O. Holt, S. Sigismund, M. Shaw, G. Bossi, andG.M.Grifths. 2007.SortingofFas ligandtosecretorylysosomesisregulatedbymono-ubiquitylationandphosphorylation.J.CellSci.120:191199.38. Schneider, P., N. Holler, J. L. Bodmer, M. Hahne, K. Frei, A. Fontana, andJ. Tschopp. 1998. Conversion of membrane-bound Fas(CD95) ligand to itssolubleformisassociatedwithdownregulationofitsproapoptoticactivityandlossoflivertoxicity.J.Exp.Med.187:12051213.39. Shudo, K., K. Kinoshita, R. Imamura, H. Fan, K. Hasumoto, M. Tanaka,S. Nagata, and T. Suda. 2001. The membrane-bound but not the soluble form ofhumanFasligandisresponsibleforitsinammatoryactivity.Eur. J. Immunol.31:25042511.40. Hohlbaum, A. M., S. Moe, and A. Marshak-Rothstein. 2000. Opposing effects oftransmembraneandsolubleFasligandexpressiononinammationandtumorcellsurvival.J.Exp.Med.191:12091220.41. Gregory, M. S., A. C. Repp, A. M. Holhbaum, R. R. Saff, A. Marshak-Rothstein,andB. R. Ksander. 2002. MembraneFasligandactivatesinnateimmunityandterminatesocularimmuneprivilege.J.Immunol.169:27272735.42. Kang, S. M., D.B. Schneider, Z. Lin, D.Hanahan, D.A. Dichek, P. G. Stock,and S. Baekkeskov. 1997. Fas ligand expression in islets of Langerhans does notconfer immuneprivilegeandinsteadtargets themfor rapiddestruction. Nat.Med.3:738743.43. Suda, T., H. Hashimoto, M. Tanaka, T. Ochi, andS. Nagata. 1997. MembraneFasligandkillshumanperipheralbloodTlymphocytes,andsolubleFas ligandblocksthekilling.J.Exp.Med.186:20452050.44. Grifth,T.S.,S.R.Wiley,M.Z.Kubin, L.M.Sedger,C.R.Maliszewski,andN. A.Fanger. 1999.Monocyte-mediatedtumoricidalactivityviathetumorne-crosisfactor-relatedcytokine,TRAIL.J.Exp.Med.189:13431354.45. Lunemann,J.D.,S.Waiczies,S.Ehrlich,U.Wendling,B.Seeger,T.Kamradt,andF.Zipp.2002.DeathligandTRAILinducesnoapoptosisbutinhibitsacti-vationofhuman(auto)antigen-specicTcells.J.Immunol.168:48814888.46. Kelly, M. M., B. D. Hoel, andC. Voelkel-Johnson. 2002. Doxorubicinpre-treatment sensitizes prostate cancer cell lines to TRAIL induced apoptosis whichcorrelateswiththelossofc-FLIPexpression.CancerBiol.Ther.1:520527.47. Wajant, H., D. Moosmayer, T. Wuest, T. Bartke, E. Gerlach, U. Schonherr,N. Peters, P. Scheurich, andK. Pzenmaier. 2001. Differential activationofTRAIL-R1and-2bysolubleandmembraneTRAILallowsselectivesurfaceantigen-directedactivationof TRAIL-R2byasolubleTRAILderivative. On-cogene20:41014106.48. Munich, S., A. Sobo-Vujanovic, W. J. Buchser, D. Beer-Stolz, and N. L. Vujanovic.2012.DendriticcellexosomesdirectlykilltumorcellsandactivatenaturalkillercellsviaTNFsuperfamilyligands. OncoImmunology 1: 10741083.49. Alonso, R., M. C. Rodr guez, J. Pindado, E. Merino, I. Merida, and M. Izquierdo.2005. Diacylglycerol kinase a regulates the secretion of lethal exosomes bearingFas ligand during activation-induced cell death of T lymphocytes. J. Biol. Chem.280:2843928450.50. Mart nez-Lorenzo, M. J., A. Anel, M. A. Alava, A. Pineiro, J. Naval, P. Lasierra,and L. Larrad. 2004. The human melanoma cell line MelJuSo secretes bioactiveFasL and APO2L/TRAILon the surfaceof microvesicles: possible contributiontotumorcounterattack.Exp.CellRes.295:315329.51. Peng,P.,Y.Yan,andS.Keng.2011.Exosomesintheascitesofovariancancerpatients:originandeffectsonanti-tumorimmunity.Oncol.Rep.25:749762.52. Mart nez-Lorenzo, M. J., A. Anel, B. Saez-Gutierrez, M. Royo-Canas,A.Bosque,M.A. Alava,A. Pineiro,P. Lasierra,J. As n-Ungr a,andL. Larrad.2007. RheumatoidsynovialuidTcellsaresensitivetoAPO2L/TRAIL. Clin.Immunol.122:2840.53. Strater, J., H. Walczak, C. Hasel, I. Melzner, F. Leithauser, and P. Moller. 2001.CD95 ligand (CD95L) immunohistochemistry: a critical study on 12 antibodies.CellDeathDiffer.8:273278.54. Clayton, A. 2012. Cancercellsuseexosomesastoolstomanipulateimmunityandthemicroenvironment.OncoImmunology1:7880.55. Lingwood, D., andK. Simons. 2010. Lipidrafts as a membrane-organizingprinciple.Science327:4650.56. Locksley, R. M., N. Killeen, andM. J. Lenardo. 2001. TheTNFandTNFre-ceptorsuperfamilies:integratingmammalianbiology.Cell104:487501.57. Voss,M., M.Lettau,M. Paulsen, andO.Janssen.2008.Posttranslationalregu-lationofFasligandfunction.CellCommun.Signal.6:11.58. Benhar, I., A. London, andM. Schwartz. 2012. Theprivilegedimmunityofimmuneprivilegedorgans:thecaseoftheeye.FrontImmunol3:296.59. Lee, H. O., J. M. Herndon, R. Barreiro, T. S. Grifth, and T. A. Ferguson. 2002.TRAIL: amechanismof tumor surveillanceinanimmuneprivilegedsite. J.Immunol.169:47394744.60. Taylor, D. D., S. Akyol, and C. Gercel-Taylor. 2006. Pregnancy-associatedexosomes and their modulation of T cell signaling. J. Immunol. 176: 15341542.61. Knight, M., C. W. G. Redman, E. A. Linton, and I. L. Sargent. 1998. Shedding ofsyncytiotrophoblast microvilli into the maternal circulation in pre-eclampticpregnancies.Br.J.Obstet.Gynaecol.105:632640.TheJournalofImmunology 5523 at Cinvestav del IPN on December 9, 2014http://www.jimmunol.org/Downloaded from