truncatedwww.avtt3.com

BioMedCentralPage1of17(pagenumbernotforcitationpurposes)BMCEvolutionaryBiologyOpenAccessResearcharticleTheXenopusFcRfamilydemonstratescontinuallyhighdiversificationofpairedreceptorsinvertebrateevolutionSergeyVGuselnikov1,ThamindaRamanayake2,AleksandraYErilova1,LudmilaVMechetina1,AlexanderMNajakshin1,JacquesRobert2andAlexanderVTaranin*1Address:1InstituteofCytologyandGenetics,Novosibirsk,Russiaand2UniversityofRochesterMedicalCentre,Rochester,NY,USAEmail:SergeyVGuselnikov-sgus@bionet.
nsc.
ru;ThamindaRamanayake-Thaminda_Ramanayake@URMC.
Rochester.
edu;AleksandraYErilova-a_shatina@ngs.
ru;LudmilaVMechetina-lucie@bionet.
nsc.
ru;AlexanderMNajakshin-najakshi@bionet.
nsc.
ru;JacquesRobert-robert@mail.
rochester.
edu;AlexanderVTaranin*-taranin@bionet.
nsc.
ru*CorrespondingauthorAbstractBackground:RecentstudieshaverevealedanunexpecteddiversityofdomainarchitectureamongFcR-likereceptorsthatpresumablyfulfillregulatoryfunctionsintheimmunesystem.
Differentspeciesofmammals,aswellaschickenandcatfishhavebeenfoundtopossessstrikinglydifferentsetsofthesereceptors.
Tobetterunderstandtheevolutionaryhistoryofpairedreceptors,weextendedthestudyofFcR-likegenesinamphibianrepresentativesXenopustropicalisandXenopuslaevis.
Results:ThediploidgenomeofX.
tropicaliscontainsatleast75genesencodingpairedFcR-relatedreceptorsdesignatedXFLs.
TheallotetraploidX.
laevisdisplaysmanysimilargenesprimarilyexpressedinlymphoidtissues.
Upto35domainarchitecturesgeneratedbycombinatorialjoiningofsixIg-domainsubtypesandtwosubtypesofthetransmembraneregionswerefoundinXFLs.
NoneofthesevariantsaresharedbyFcR-relatedproteinsfromotherstudiedspecies.
PutativeactivatingXFLsassociatewiththeFcRγsubunit,andtheirtransmembranedomainsarehighlysimilartothoseofactivatingmammalianKIR-relatedreceptors.
ThisarguesinfavorofacommonoriginfortheFcRandtheKIRfamilies.
PhylogeneticanalysisshowsthattheentirerepertoiresoftheXenopusandmammalianFcR-relatedproteinshaveemergedaftertheamphibian-amniotessplit.
Conclusion:FcR-andKIR-relatedreceptorsevolvedthroughcontinualspecies-specificdiversification,mostlikelybyextensivedomainshufflingandbirth-and-deathprocesses.
Thismodeofevolutionraisesthepossibilitythattheancestralfunctionofthesepairedreceptorswasadirectinteractionwithpathogensandthatmanyphysiologicalfunctionsfoundinthemammalianreceptorsweresecondaryacquisitionsorspecializations.
BackgroundImmuneresponsesareregulatedbyabalanceofopposingsignalsdeliveredfromleukocytesurfacemolecules[1,2].
Inthemammalianimmunesystem,severalfamiliesofactivatingandinhibitoryreceptorsformanelaboratedregulatorynetworkthattightlyaffectsallstagesofPublished:16May2008BMCEvolutionaryBiology2008,8:148doi:10.
1186/1471-2148-8-148Received:26September2007Accepted:16May2008Thisarticleisavailablefrom:http://www.
biomedcentral.
com/1471-2148/8/1482008Guselnikovetal;licenseeBioMedCentralLtd.
ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.
org/licenses/by/2.
0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page2of17(pagenumbernotforcitationpurposes)immuneresponses.
Theevolutionaryhistoryofthisnet-workispoorlyunderstood.
While"pairing"ofreceptorsintotheinhibitoryandactivatingformsappearstohaveoccurredininvertebrates,thereisnoclearevolutionarycontinuitybetweeninvertebrateandvertebratereceptorsystems[3,4].
Furthermore,ambiguityofrelationshipsisoftenobservedforpairedreceptorsfromdifferentlineagesofvertebrates[5-8].
ClassicalFcreceptors(FcR)andkillercellimmunoglobu-linreceptors(KIR)constitutetwofamiliesthatareproto-typicfortheparadigmofimmuneregulationthroughintegrationofactivatingandinhibitorysignals.
Membersofeachfamilyfallintotwomainsignalingclasses.
TheinhibitoryreceptorscontainITIMsintheircytoplasmictails,whiletheactivatingreceptorsassociatewiththeITAM-bearingtransmembranesignalsubunits,suchasFcRγ(FcRs)orDAP12(KIRs).
FcRsarewidelyexpressedonvariousleukocytesubsets.
Theyregulatephagocytosis,cytokinesrelease,antibody-dependentcellmediatedcyto-toxicity,andantibodysynthesis[9,10].
KIRsplayacrucialroleinregulationofhumanNKcellcytotoxicityviarecog-nitionofMHCclassIantigensonthesurfaceoftargetcells[11-13].
Duringthelastdecade,ithasbeenrecognizedthatFcRsandKIRsbelongtolargefamiliescomprisedofstructur-allyrelatedyethighlydiverseproteins.
Thusfar,eighthumanandsixmouseFcR-like(FCRL)geneshavebeendescribed[14-23].
Twoofthem,designatedFCRLAandFCRLBaccordingtothenewnomenclature[24],areintra-cellularproteinscomposedofthreeIg-likedomainsandaC-terminalmucin-likedomain.
Sixhuman(FCRL1-FCRL6)andthreemouse(FCRL1,FCRL5,andFCRL6)genescodeforcellsurfacereceptorswiththeextracellularregions(EC)composedoftwotonineIg-likedomainsandintracellularregionsbearingdifferentpatternsoftheITIM-,ITSM-andITAM-likemotifs.
ApartfromtheFcR-characteristicD1,D2andD3subtypes,twonewstructuralIg-likedomainsubtypes,D4andD5,havebeenidentifiedintheseproteins.
Furthermore,oneofthenovelmousegenes,FCRLS,encodesasolublemosaicproteincontain-ingascavengerdomain[17,22].
StudiesoftheKIRfamilyhavealsorevealeditsconsidera-blestructuralandfunctionalheterogeneity.
HumanKIR-likeproteins(KIRL)includecellsurfacereceptorsoftheLILR(ILT/LIR/MIR)familyaswellasFcαR,GPVI,Nkp46,OSCAR,Lair1andLair2[25,26].
TheLILRfamilyconsistsofbothinhibitoryandactivatingforms:LAIR-1isaninhibitoryreceptor,LAIR2issoluble,theothersareacti-vating.
LikeFcRsbutunlikeKIRs,theactivatingLILRreceptors,aswellasFcαR,OSCAR,NKP46,andGPVIassociatewiththeFcRγsubunit[27-31].
However,thetransmembraneregions(TM)ofactivatingKIRLsarestructurallydifferentfromthoseofFcRs.
Intriguingly,therepertoiresoftheFcR-andKIR-relatedproteinsaredifferentfromonespeciestoanotherinhighervertebrates.
Forinstance,eachofthesixhumanandfourmouseextracellularFCRLshasauniquedomainarchitecture[22,24].
FunctionalequivalentsofKIRsinrodentsareC-typelectinreceptorsoftheLy49family[32].
ThemousealsolackscounterpartsofFcαRandLair-2andhasfewerLILRhomologuesdescribedasPIRs[25,26].
ProfounddifferencesintheKIRandFcRfamilieshavebeenalsorevealedbetweenmammalsandbirds.
Recentdatashowthatthechickengenomehasmorethanahun-dredgenesforKIR-likepairedreceptorsknownasCHIRs[33-35].
Atthesametime,asingleFcR-relatedgenehasbeendetectedinthisspecies[36,37].
Comparisonofthe3-Dstructureofmembrane-proximaldomainsofFcγRIIandKIR2Ddemonstratedtheirsimilarfoldingandpromptedasuggestionthatthetwofamiliesmayhavehadacommonorigin[33].
ThissuggestionwasmadebeforeidentificationofFCRLs.
Alaterphylogeneticanalysisdidnotprovidesolidsupportinfavorofhom-ologyoffiveFCRL-characteristicIg-domainsubtypeswiththetwomaindomainsubtypesofKIR-relatedreceptors[5].
Nevertheless,thisideaofthecommonancestryoftheFcRandKIRfamilieshasbeenrevivedinmodifiedformaftertherecentidentificationofafamilyofpairedrecep-torscalledleukocyteimmune-typereceptors(LITR)incat-fish[6].
LITRsarecomposedofseveraldomainsubtypessomeofwhichresembletheFcR-characteristicdomainsD1D2,whereasothersaremoresimilartoKIRLdomains.
SuchcompositionofIgdomainshasbeenproposedasancestralforthetetrapodpairedreceptors[6].
However,weaksequencesimilaritybetweenLITRandKIRLIgdomains,aswellastheabsenceoftheD3,D4andD5typedomainsinLITRsdidnotallowconclusionsaboutdefi-niterelationshipsoftheteleosteanproteinswiththehighervertebrateFcRandKIRfamilies.
ThefactthatallknownFcR-andKIR-relatedreceptorsareprimarilyexpressedincellsoftheimmunesystemiscon-sistentwiththeircontributiontotheimmuneregulation.
However,theexactfunctionsofallFCRLsandmanyKIRLsarestillunknown.
TheambiguityofthestructuralandfunctionalevolutionofFcR-andKIR-relatedreceptorscomplicatesourunderstandingofhowthisregulatorynet-workisorganized,whichfactorsdrivesitsspecies-specificchangesandultimatelyhowitmaybemanipulatedfortherapeuticpurposes.
Togaindeeperinsightintotheevolutionoftheimmu-noregulationthroughpairedreceptors,westudiedtheFcRfamilyintheamphibiansXenopuslaevisandXenopustrop-BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page3of17(pagenumbernotforcitationpurposes)icalis.
ThedataobtainedprovideevidenceinfavorofacommonoriginoftheFcRandKIRfamiliesandtheirceaselessdiversificationthatappearstobecausedbyverystrongnaturalselectionpressure.
ResultsTheFcRfamilyisexpandedinamphibiansDuringourstudiesofthehumanandmouseFcR-likegenesweobservedthatESTdatabasescontainnumerousX.
tropicalisandX.
laeviscDNAsencodingproteinsstruc-turallysimilartomammalianFcRsandFCRLs.
Thedegreeofaminoacidsequenceidentityrangedfrom25to43%fordifferentdomainsubtypes.
WedesignatedthesegenesXFL(XenopusFcR-Like).
TherecentsequencingoftheX.
tropicalisgenomeprovidedanopportunitytoexaminetheorganizationandstructureoftheXFLgenesinmoredetails.
Weusedinsilicoanalysisoftheversion2,3and4genomicsequencesdepositedattheJGIwebsite.
Wedidnotconsidertheconsortiumgenemodelsinthissurvey.
Directapplicationofgenepredictionprogramstogenomicsequencesoftenresultsinerroneousmodels.
Toovercomethispitfall,wefirstidentifiedexonscodingfortheXFLECandTMdomainsusingtheTBLASTNsearchwithaminoacidsequencesofthecorrespondingX.
tropi-calisandX.
laevisESTcDNAsormammalianFcR-likepro-teins.
TheidentifiedX.
tropicalissequenceswereusedinthesecondroundofthecomputationalscreeningtorevealexonsthatmighthavebeenoverlookedinthefirstround.
Theprocedurewasrepeateduntilnonovelexonswereidentified.
Theexonslackingframe-shiftmutationsorstopcodonswereexaminedforthepresenceoftheAGandGTsplicesignalsmatchingthephase1rule.
Thereaf-ter,thegenemodelsweregeneratedusingbothautomaticandmanualprocedures.
TheexonsfortheTMregionsservedasthegenedelimiters.
Thisapproachresultedinfindingseveralhundredexonson33scaffolds.
Ofthese,19scaffoldscontained1–2exonsthatmayeitherrepre-sentmisassembledgeneregions,genefragmentsorpseu-dogenes.
Theexonson14otherscaffoldscouldbearrangedinatleast75XFLgenes.
Fig.
1showsthepre-dictedorganizationandexon/intronarrangementofthesegenes.
Theexonsencodingthesignalpeptidesandcyto-plasmicregionsarenotshowninthisschemebecauseofpooraccuracyoftheirprediction.
Nevertheless,incertaincases,suchexonscouldbedelineatedonthebasisofalignmentofthegenomicsequenceswiththeESTcDNAs.
Atthetime,theESTdatabasescontainedX.
tropicaliscDNAscorrespondingto13XFLgenes.
Thesignalpep-tideswereinvariablyencodedbytwoexons,likeinthemammalianFcR-likegenes.
InafractionofXFLgenes,theexonsforsignalpeptidesmettherule30/21bp(thelengthofthefirst/thelengthofthesecondexon),whichischaracteristicofmammalianFCRL1-6,FcγRIandFCRLBgenes.
TheXFLcytoplasmicregionsareencodedbyonetofiveexons.
Inthemammaliangenomes,FcR-likegenesarelinkedtothegenesoftheCD2family.
Thus,theFCRL6geneisapartofaconservedsyntenicgroupthatincludestheSLAMF8(BLAME),IgSF9,DUSP23,TAGLN2andNGES1genes[37].
Wefoundasimilargroupinthescaffold626thatcontains16XFLgenes(Fig.
1).
TheXFLgene626_16islocatedbetweentwoCD2-likegenes,oneofwhichshowsthegreatestsimilaritytomammalianSLAMF8(BLAME).
XenopushomologsofthemammalianIgSF9,DUSP23,TAGLN2andNGES1genesarealsotightlylinked.
Thisconservedsyntenytakentogetherwiththeresultsofthesequencecomparisons(Seeadditionaldatafile1)andphylogeneticanalysis(seebelow)stronglysup-portstheassignmentoftheXFLgenesastrueamphibianhomologsofthemammalianFcRgenes.
TheXFLreceptorsaresubdividedintotwoclassesWithafewexceptions,thepredictedXFLgenescodefortypeIcellsurfacereceptors(Fig.
2).
TheirTMregionsfallintotwostructuraltypesthatwedesignatedTM1andTM2.
ThecharacteristicfeatureofTM1isthepresenceofaconservedNxxRmotifattheN-termini.
TM2lackschargedresidues.
Interestingly,TM1regionsarehighlyhomologoustotheTMregionsofsomeKIRLssuchasLILRA2,PIR-A,NCR1/NKp46,GPVI,OSCAR,andFcαR(Fig.
3).
AlltheseproteinsareknowntoassociatewiththeFcRγsignalsubunit[27-31].
ItisimportanttostressthatTMsofclassicalactivatingFcRsarequitedifferent,andbearatypicalconservedstructuralmotif(M/L)Fxx(D/N)TxL[38].
Despitetheextensivesearch,wedidnotfindexonsforTMregionswithsuchasignatureintheX.
tropi-calisgenome,orintheavailableXenopusESTcDNAs.
OurphylogeneticanalysissupportsacloserelationshipoftheTMregionsoftheXFLandKIRLproteins.
AsshowninFig3b,theNxxRmotif-containingTMsandTMsofactivatingclassicalFcRsformtwodistinctclusters.
TheTMregionofDAP12-associatingKIR2DS,amemberofthehumanKIRfamilyofMHCclassI-specificNKcellreceptors,isnotrelatedtoeitherofthesegroups.
ComparisonoftheX.
tropicalisgenomicandESTsequencesshowedthattheTM1-containingproteinslackacytoplasmictailorhaveaveryshortone.
Ineverysuchcase,theTMregionandthetailareencodedbythesameexon.
GeneswithaTM2havelongercytoplasmictails.
Thesetailscontainonetothreetyrosine-basedmotifsmatchingtheconsensusYxxL/I/VandareencodedbyexonsseparatedfromTMexons.
Allthesestructuralfea-turesarecompatiblewiththesubdivisionoftheXFLreceptorsintotwofunctionalclasses,activatingandinhib-itory.
Thenumberofgenesforeachclassisroughlysimi-larandtheyareintermingledinthegenome(Fig.
1).
BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page4of17(pagenumbernotforcitationpurposes)GenomicorganizationofthepredictedX.
tropicalisFcR-likegenesFigure1GenomicorganizationofthepredictedX.
tropicalisFcR-likegenes.
TheexonsforeachparticularsubtypeoftheIg-likedomains(D1-D6)aremarkedbyadifferentcolorasindicated.
ExonsforTMswiththeNxxRmotif(TM1)areinblackandthosefortheTMregionswithoutchargedresidues(TM2)areinwhite.
ThegenemodelssupportedbyX.
tropicalisESTcDNAsareboxed.
Arrowsindicatetranscriptionalorientation.
Thegenesaredesignatedbytheirscaffoldnumberandtheirconsecu-tivepositionatthecorrespondingscaffold(version4.
1).
Filledcirclesshowpositionofgapsintheassembly.
Toconservespace,onlyfractionsofscaffoldsareshown;theirbordersareindicatedinkbattherightandleftsides.
1234567141931943733745251234567891011121314151617123451214614728213456789101162443976464286978412345Scaffold6262019920037938012345678910111213141516CD2LCD2L500924NGES1TAGLN2IGSF9DUSP2396410kbScaffold658Scaffold946Scaffold362Scaffold4352934359304350115Scaffold1131Scaffold125644721Scaffold15242730-D1-D2-D3-D4-D5-D6-Tm1-Tm2-(n)NstretchCD2L-othergenes-orientation-DESIGNATIONS:12Scaffold7212191991Scaffold110711112012Scaffold14036110312Scaffold15496041Scaffold11573054012Scaffold14137348BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page5of17(pagenumbernotforcitationpurposes)Schematicrepresentationofdomainarchitectureofhuman,mouse,XenopuslaevisandX.
tropicalisFcR-likeproteinsFigure2Schematicrepresentationofdomainarchitectureofhuman,mouse,XenopuslaevisandX.
tropicalisFcR-likeproteins.
ThestructureofX.
laevismoleculesisdeducedfromcDNAsequences,whereasthestructureofX.
tropicalismole-culesispredictedbasedonthegenomicsequencesandconfirmedbytheESTcDNAsequences(markedwithasterisk).
TheIg-likedomainsbelongingtotheD1-D6structuralsubtypesareshownbycirclesandtheTMregionsbythicklines.
ThinlinesandrectanglesdesignatecytoplasmictailsandYxxV/L/Imotifs,respectively.
ThecolorpatternfortheIg-domainssubtypesandtransmembranetypesareasinFig.
1.
PairedreceptorswithsimilarextracellularregionsbutdistinctTMregionsareboxed.
X.
tropicalisH.
sapiensFcRFCRLAB612435M.
MusculusFcRfcrlab61s5-D1-D2-D3-D4-D5-D6-DS-DM-Tm1-Tm2-ITAM-ITIM***********XFLX.
laevisXFL1.
51.
41.
91.
21.
51.
81.
101.
123BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page6of17(pagenumbernotforcitationpurposes)TheECregionsofXFLsarehighlydiverseAremarkablefeatureofthepredictedXFLproteinsisanextraordinarydiversityoftheirdomainarchitectures.
Overall24differentcombinationsofsixstructuralsub-typesoftheIg-likedomainsintheECregionsofXFLswerefound.
FivesubtypeswereassignedtoD1-D5subtypespreviouslyidentifiedinthemammalianmembersoftheFcRfamily.
AsixthsubtypeappearstobeXenopus-specificasnocloserelativeswerefoundintheproteindatabases.
TheD3subtypedomainismostfrequentandmayberepeatedupto7timesinaprotein.
Althoughwecannotruleoutthatsomeofthegenepredictionsresultfromgenomeassemblyartifacts,wefoundhighsimilaritybetweentheexpressedandgenomicsequences.
TheEST-genomecomparisonsshowedtheabsenceoftwoexonsinthegenomicsequences.
Inbothcases,closeinspectiondemonstratedthepresenceofgapsinthecorrespondinggenomicregions.
Elevenof13cDNAsfullymatchedourgenemodels.
Thisfact,togetherwiththeabsenceofgapsinmanypredictedgenesandreiterationofcertaindomainarchitecturestwoormoretimessuggestahighdegreeofconfidenceintheproposedmodels.
AmongthepredictedproteinstherearetypicalpairswithidenticalectodomainsanddistinctTMsubtypes.
(Fig.
1and2).
TheECregionsof11XFLsarecomposedofD1,D2,andD3domains,likemammalianFcγRI.
ThisistheonlyECcompositionsharedbytheknownmammalianandXenopusFcR-likeproteins.
IfweconsiderstructuralsubtypesoftheTMregionsasdistinctdomains,upto36domainarchitec-turesmaybedistinguishedamongtheXFLproteins,noneofwhicharepresentamongmammalianmembersoftheFcRfamily.
Lineage-specificexpansionofXenopusandmammalianFcRfamiliesWhiletheattributionoftheXFLproteinstotheFcRfamilyisunequivocalaccordingtothereciprocalsequencecom-parisonsandproteindatabaseanalysis,itremainsunclearhowXenopusandmammalianproteinsarerelatedtoeachother.
Toassesssuchrelationships,wegeneratedaseriesofphylogenetictreeswiththeMEGA3softwarepackage[39].
Forthispurpose,aminoacidsequencesoftheIgdomainsubtypesfromallthepredictedXFLswerealigned(seeadditionaldatafile1).
TreesweregeneratedusingtheNJandMEmethods.
Tosimplifythetrees,thealignedblocksofD1,D2andD3domainswerereducedbyremovingredundantsequenceswithcloseassociation.
Thereafter,thesequencesforalltheXFLdomainsubtypesAlignment(A)andphylogeneticanalysis(B)ofthededucedTMregionsoftheXenopusandmammalianFcR-andKIR-likepro-teinsFigure3Alignment(A)andphylogeneticanalysis(B)ofthededucedTMregionsoftheXenopusandmammalianFcR-andKIR-likeproteins.
AllthedisplayedmammalianmembersoftheKIRfamilyassociatewithFcRγsubunit.
TheX.
tropicalisgenesaredesignatedaccordingtothescaffoldnumberandageneposition.
Identicalandsimilarresiduesareshownbywhitelettersonblackandgraybackgrounds,respectively.
TheNeighbor-JoiningtreeofthenucleotidesequencesoftheTMexonswasconstructedusingtheMEGA3software[39].
Thebootstrapvaluesareshown.
KIR2DS4946.
10626.
11413.
1658.
5658.
17626.
12626.
31256.
2435.
1362.
4hGPVIhOSCARhIGSF1LILRA2bRcRhFcRhNKP46rKIRLhFcRIhFcRImFcRImFcRIIIhFcRIII988197980.
1mFcRIVBmFcgRIIILVWYHTAFSLVMCLLFAVDTGLYFYVRRNLQThFcgRIIIPPGYQVSFCLVMVLLFAVDTGLYFSVKTNIRSmFcgRIVPPWHQITFCLLIGLLFAIDTVLYFSVRRGLQShFceRIKYWLQFFIPLLVVILFAVDTGLFISTQQQVTFhFcgRIPVWFHVLFYLAVGIMFLVNTVLWVTIRKELKRmFcgRIPVWFHILFYLSVGIMFSLNTVLYVKIHR-LQR435.
1DYTLQNLIRLIVSVCLSVLCLCFVINHLKKGKXFL2DYTLQNAIRLSLSVFLSVLCLCFVCNHLK-RE1131.
2NYTVQNTVRLTASAFIFLLASCLLFHHLKSPQ362.
3RYTIENTIRLALAVCILLSGSLLSYCHIKRQV626.
5RRTLENTLRLILAAVIFIVILCLLFYHIKTVDXFL1.
7FGSLQNTVRLALSAVIFIIILLIVFYHIKTME626.
3RTALENIVRLVFSVLILIIILWILFHQKKPRRNKP46DHTAQNLLRMGLAFLVLVALVWFLVEDWLSRKFcaRDYTTQNLIRMAVAGLVLVALLAILVENWHSHTPIRADHTMENLIRMGMAVLVLIVLSILATEAWRSHRLILRA2DYTVENLIRMGVAGLVLVVLGILLFEAQHSQROSCARDYTRGNLVRLGLAGLVLISLGALVTFDWRSQNGPVIYYTKGNLVRICLGAVILIILAGFLAEDWHSRRABMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page7of17(pagenumbernotforcitationpurposes)werealignedtogetherwiththedomainsequencesofthehumanproteins,andtheirrelationshipswereanalyzedbythesameprocedure.
ThefinaltreeisshownonFig4.
ThetreetopologysupportssubdivisionoftheXenopusandhumanIgdomainsintofivecommon(D1-D5)andoneXenopus-specific(D6)subtypes.
Mostimportantly,thetreeshowsseparateclusteringoftheXenopusandhumansequences.
Thisbranchingpatternsuggeststhatduplica-tionsoftheFcR-likegenesinamphibianandmammalianlineageswerelineage-specificandthatseparationofthemammaliangenesintoclassicalFcRsandFCRLsoccurredafterthesplitoftheamphibianandmammalianlineages.
TherelationshipsamongtheXFLgenesprovideevidenceforacomplexpatternoffamilyevolutioninamphibians.
Itsdetaileddescriptionisbeyondthescopeofthepresentpaperandwillbepublishedseparately.
Whatisrelevanttonotehere,isthatthestrongassociationoftheamphibiansequenceswitheachotherdoesnotnecessarilycorre-spondtohighlevelofsequencesimilarityamongthem.
ThecumulativetreeillustratessubdivisionoftheXFLD1,D2andD3domainsintostructuralvariantswhoserela-tionshipswitheachothercannotberesolved.
TheD1,D2andD3domainsubtypesfallinto15,13and21structuralgroups,respectively.
Thedegreeofsequencesimilarityamongthegrouprepresentatives(32–45%identicalresi-dues)isintherangeofsequencesimilaritybetweenXFLandmammalianFcRdomains.
InthecaseofD3,thediversityismainlyderivedfromasmallnumberofgeneslocatedinthescaffolds362,435,1131and1256.
Forinstance,theextracellularpartsofthepredictedproteins1131_3and1131_5arecomposedoffiveD3domainseach.
Thesedomainsaresubdividedintofourgroups.
EachoffiveD3domainsoftheXFL1256_2representsadistinctstructuralvariantproducingaseparatebranchinthetree.
Ontheotherhand,morethan50proteinsencodedbythegenesoftheotherscaffoldshaveonetosevenD3domainsbelongingtothesamegroup(group1).
Thisgroupmaybefurthersubdividedintothreemainsubgroups1.
1,1.
2and1.
3andanumberofindividualmembersbasedonthestructureoftheD2andD1domains.
Forinstance,theproteinsencodedbythegenesonthescaffold626haveverysimilarD3domainsbuttheirD1andD2domainsfallinto8structuralvariantswithpoorlyresolvedrelationships.
Thisfactstronglysug-geststhatintergenicexonrecombinationwasafrequenteventintheevolutionoftheXFLfamily.
Oftengenescon-tainingexonsfortheD4,D5orD6domainsubtypes,ninebelongtothesubgroup1.
1andonetothesubgroup1.
3.
Interestinglysubgroups1.
1and1.
2differinpatternsofaminoacidreplacementintheD1andD2domains.
TheD2domainsofsubgroup1.
2arecharacterizedbyexten-sivevariationinthelengthandsequenceoftheFGloop,theequivalentofCDR3oftheV-typedomains(Fig.
5).
TheregioncoveringthestrandsCtoFiswellconserved.
Incontrast,theD2domainsofsubgroup1.
1showmorevariationintheregionbetweentheC'andFstrands.
TheirF-Gregionisrelativelyconserved.
TheD1domainsofthe1.
1and1.
2subgroupsalsodisplayvariabilityatdifferentsites(notshown)suggestingtheexistenceofatleasttwoclassesofligandsforthegroupIreceptors.
TheD1andD2domainsareimplicatedinbindingtoIgGandIgEbyclas-sicalFcRs.
However,theresiduesknowntocontacttheFcportionofIgarenotconservedintheXFLssequences,ExperimentalsupportofXFLdiversityTogainadeeperinsightintostructureandexpressionoftheXFLgenes,westudiedthisfamilyinX.
laevis,acloserelativeofX.
tropicalis.
IncontrasttoX.
tropicalisthathasadiploidgenome,X.
laevisisanallotetraploidspecies.
TheimmunesystemofX.
laevisisoneofthemostthoroughlystudiedamonglowervertebrates[40-42].
FivedifferentX.
laeviscDNAsforXFLproteinswereobtainedfromtheIMAGEconsortiumandsequenced.
Morethan30cDNAs,11ofwhichwereunique,wereadditionallyclonedfromseveralX.
laeviscDNAlibrariesusingscreeningwithanexonencodingtheD3domainofgroup1asaprobe.
Of16distinctcDNAs,9werefull-length,theothersweretruncatedatthe5'end.
NinecDNAsencodedtypicalcellsurfaceproteinscontainingTM2-likeTMregionsandcytoplasmictailsofvaryinglengthwithonetothreetyro-sine-basedmotifs(Fig.
2).
Theaminoacidsequencesoftwocloneshadshortcytoplasmictails.
TheirTMregionscontainedtheNxxRmotifandwereassignedtotheTM1subtype.
OnecDNAcloneencodedproteinlackingTMbutcontainingatypicalcytoplasmictailwithtwotyro-sine-basedmotifs.
Finally,fourcDNAscodedforputativesecretedproteinscomposedoftheIg-likedomainsonly.
Atpresent,itisunclearwhetherornotthelatterfiveclonesrepresentalternativetranscriptsofgenesencodingcellsurfacereceptors.
Asexpected,theinitialsequencecomparisonsdemon-stratedthatmostofX.
laevisderivedXFLsmaybejoinedintogroup1.
TheirD3domainsshared65to95%identi-calresidueswitheachotherandwiththeX.
tropicalisD3domainsofthegroup1receptors.
TheseproteinsweredesignatedXFL1.
1–1.
14.
AsinthecaseofX.
tropicalis,theX.
laevisgroup1proteinsshowedvariabledegreeofiden-tity(35to85%)intheirD1andD2domains.
TwootherproteinsweredesignatedXFL2andXFL3.
TheirD3domainsshared35–45%identicalresidueswitheachotherandwiththeD3domainsofthegroup1proteins.
Accordingtothephylogeneticanalysis(notshown),theX.
laevisXFL2andXFL3genesweremostsimilartotheX.
tropicalis435_1and1131_1,2,3genesrespectively.
ToestimatethegenomiccomplexityoftheXFLfamilyinX.
laevis,weperformedSouthernblothybridizationusingBMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page8of17(pagenumbernotforcitationpurposes)Neighbor-JoiningtreebasedontheD1-D5nucleotidesequencesofX.
tropicalisXFLsandhumanFcRandFCRLgenesFigure4Neighbor-JoiningtreebasedontheD1-D5nucleotidesequencesofX.
tropicalisXFLsandhumanFcRandFCRLgenes.
X.
tropicalisgenesaredesignatedaccordingtoascaffoldnumberandtheirconsecutiveposition(SeeFig.
1).
Forgenescontainingmultipleexonsfordomainsofthesametypetheseexonsarenumberedaccordingtotheirposition(i.
e.
D3.
1-D3.
3).
ThetreewasconstructedusingMEGA3softwarewithp-distancesfornucleotidesequencesitesandpair-wisedeletionoption.
Thenumbersonthetreerepresentvaluesforthebootstrapandinteriorbranchtestsafter250replicates.
95/999986/999383/999485/9890100/99100100/99100100/9910099/99100100/99100100/9910092/989995/9998100/99100626.
8D1626.
9D1.
2362.
1D1626.
15D1.
2626.
4D1.
2435.
1D11131.
5D1435.
7D1626.
3D1.
2626.
12D1.
1626.
3D1.
1626.
4D1.
1658.
3D11256.
3D1658.
14D1658.
1D1FcgRIID1FcgRIIID1FcgRID1FceRID1FCRLBD1FCRL3D1FCRL5D1FCRL4D1658.
5D5658.
2D5658.
11D5658.
1D5FCRL6D5FCRL5D5.
1FCRL1D5FCRL5D5.
6FCRL2D5FCRL3D5.
1658.
5D6658.
11D6658.
12D6658.
13D6658.
7D6658.
2D6658.
1D6FCRL4D4FCRL2D4FCRL3D4FCRL1D4658.
1D4658.
11D4658.
12D4658.
3D4658.
2D4FcgRIIID2FcgRIIbD2FcgRID2FceRID2FCRL6D2FCRLBD2FCRLAD2FCRL2D2FCRL3D2FCRL4D2FCRL5D2626.
12D2.
2626.
16D2658.
6D21131.
4D2362.
1D2435.
1D2626.
15D2.
2435.
6D2626.
9D2.
1626.
3D2.
1626.
4D2.
1626.
4D2.
2626.
9D2.
2FCRL2D3FCRL3D3FCRL5D3FCRL1D3FCRL4D3FcgRID3FCRL6D3FCRLAD3FCRLBD31256.
2D3.
3435.
2D3.
11131.
6D3.
11131.
3D3.
21131.
1D31256.
1D3.
1435.
5D3.
1435.
3D3.
1435.
1D3435.
3D3.
2435.
4D3.
2435.
2D3.
21256.
2D3.
41256.
2D3.
2D21131.
4D3.
51256.
2D3.
51256.
2D3.
1361D32.
1131.
4D3.
36.
15D3.
626658.
19D3.
11131.
5D3.
577/959598/99960.
05D1D5D6D4D396/9998BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page9of17(pagenumbernotforcitationpurposes)theD1-exonofXFL2andD3-exonsoftheXFL1.
1andXFL3genesasprobesundernon-stringentconditions.
Multiple(upto25)hybridizingbandswererevealedontheblotsprobedwiththeD3-exonoftheXFL1.
1gene(Fig.
6).
TheXFL2-andXFL3-specificprobesrevealedoneandthreehybridizingbands,respectively.
TheseresultsdemonstratedthattheclonedXFLgenesconstituteonlyapartofthefamilyandthat,likeinX.
tropicalis,mostoftheX.
laevisXFLgenesappeartobelongtothegroup1.
X.
laevisXFLgenesareprimarilyexpressedinlymphoidtissuesToassesstheXFLexpressionpattern,weperformedNorthernblothybridizationoftotalRNAfromvarioustis-suesofadultfrogs.
TheexonforD3domainofthegroupIwasusedasaprobeatlowstringencyconditions.
Asexpected,Northernblottingrevealeddiffusebandsrepre-sentingmultiplegenetranscripts.
Thehighestsignalintensitywasobservedinthespleenandthymus(Fig.
7).
Toexamineexpressionpatternsoftheindividualgenes,wedesignedgene-specificprimersforsevendifferentXFLcDNA.
Toexcludepossiblecrossmatching,the3'-untrans-latedsequencesweremainlyusedforthereverseprimers.
TheRT-PCRanalysisdemonstratedthatthetissuedistri-butionofthecorrespondingmRNAisvariable(Table1).
Thetranscriptsofallthegenesweremainlydetectedinlymphoid(spleen,thymus)andnon-lymphoidtissuescontainingcellsofhaemopoieticorigin(e.
g.
,liver,intes-tine,lung).
ExpressionoftheXFL3,XFL1.
10andXFL1.
12geneswasdetectedinbrain.
ThetissuedistributionofmRNAintadpoleswasslightlybroader.
Inparticular,allXFLstestedexcept1.
8,weredetectedinthegillswhichisknowntobeaveryactiveimmunologicalsiteowingtointensebloodcirculationandhighexposuretoantigens[43].
Furthermore,expressionoftheXFL1.
8genewasfoundonlyintadpolespleen.
DifferentpatternoftheXFLtranscriptdistributioninadultsandlarvaesuggestthattheexpressionofatleastaproportionoftheXFLgenesisdevelopmentallyregulated.
TM1facilitatesXFLassociationwiththeFcRγsubunitThepresenceoftheNxxRmotif-bearingTMsinmanyXFLssuggestedthat,likethemammalianactivatingKIRLs,theseXenopusreceptorsmayrequireFcRγchainforcellsurfaceexpressionand/orsignaltransduction.
Toexaminewhetherthisisthecase,wegeneratedaseriesofconstructsAlignmentofdeducedaminoacidsequencesofX.
tropicalisD2domainsbelongingtosubgroups1.
1and1.
2Figure5AlignmentofdeducedaminoacidsequencesofX.
tropicalisD2domainsbelongingtosubgroups1.
1and1.
2.
Thedomainsaredesignatedaccordingtothescaffoldnumber(version4.
1)andconsecutivepositionofageneencodingthatparticulardomain(Fig.
1.
).
Identicalandsimilarresiduesareshownbywhitelettersonblackandgraybackgrounds,respec-tively.
Dashesrepresentgapsintroducedtomaximizesimilarity.
Grayarrowsindicatepredictedβ-strandsformingIg-likedomain(A-G).
626.
8GWLILQAPPAVHEGDSLSLRCHSRPEYR-AWNPVFYKDNKPIGSPVSGSELHIGRVGVTASGTYRCEKKMCYYC--YTTVTTLTADRTITVT626.
7DLLILQAPPAVHEGDSLSLRCHSQPGYD-TRNPVFYKDNKAIGSPVSGSELQIGRVNVTESGTYRCDKEMCYYC---QTFFNYTAYRTISVS946.
3DRLILQAPPAVHEGDSLFLRCHSWPGYG-TRKPVFYKDNKAIGSPVRGSVLQIGRVGVTASGTYKCEKGIYFG---YNNYRTHSDEKNISVS626.
2NLLIMQAPPAVHEGDSLSLRCHTWPGYYYTRNPVFYKDNEAIGAPVSGSELHIGRVNVSASGTYSCEKEIYIDR--INNYRTYSDEKTISVS946.
4DLLILQAPPAVHEGDSLSLRCHSRPGYV-TRNPVFYKDNKPIGPPVSGSELQIGRVNVTVSGTYGCEKDIYY----KYNYHTYSAKQYILVS946.
1DWLILQAPPAVHEGDSLSLRCHSRPGLD-AKKPVFYKDNKAIGPPVSGSELQIGRVGVTASGTYRCDKEIYFYYAFGNGYRTYKDEQYISVS946.
5AWLILQAPPAVHEGDSLSLRCHSRPGYD-TRDTIFYKDNEPIGSPVSDSELQIGRVGVTASGTYRCDKEIYFYYPYRNGYRIYVATQHISVS946.
6DWVILQAPPAVYEGDSLSLRCHSRPGFE-AGNSIFYKDNKAIGSPVSGSELQIGRVNVTASGTYKCEKEIYYYY----RYRSHGAEQHVRVQ946.
8GWVILQAPPAVHEGDSLSLKCYSRPGYD-TRNPVFYKDNKAIGSPVSGSELQIGRVDVTASGTYGCKKEIFFHYLVGNRYRSHGAEQYVRVQ1413.
1DPLILQAPPAVHEGDSLSLRCHSRPGYD-TWNPVFYKDNKPIGSPVRGSELHIGRVDVTASGTYRCEKELHYCNYFNPCSIASSDERYILVS1413.
2DPLILQAPPAVYEGDSLTLRCQRRPGYD-TRNPVFYKDNYAIGSPVSGSELHIGRVDVTASGTYLCEE--------KSSFKTLKAENYISVS946.
2DRLILQAPPAVHEGDSLSLRCHSRPGYD-SRNPVFYKDNKAIGFPVSGSELQIGRADVTASGTYRCQKVIRFNSG-RNYLILNTDEKYISVS946.
10DLLILQAPPAVHEGDFLSLRCHSQPGYD-TRDTVFYKGNKAIGSPVSGSELQIGRVDVTASGTYRCAKEIDFG----NLVYIAKAHHTISVS626.
15DPLILQAPPAVYEGDSLSLRCHSQPAYR-EKKLVFYKDNETIGPPVSGSELQIGRVNVTASGTYRCGKEITRYV--FSPVTPYTAHRNISVQ658.
6DYLSLKVPPFVFEGDNLQVSCAGYPGYYADTAKLYKGDN-VIDSS-GNGSFHIGRVTMATSGSYTCYRSVQYHNKYYNKESSAVISVK658.
9DYLSLKVPPFVFEGDNIQVSCAGYPGYKAGDAKLNKENN-FIGSS-GNGSFHIGRVTMATSGSYTCHRAVRYHSLYHNKESSAVISVK658.
7DYLSLKVPPFVFEGDNLQVSCAGYPGYEAGNAKLYKGNE-FIGFS-GNGSFHIGRVTMATNGSYTCYRLVRHHGLFYHQESSVYISIK658.
2GYLFLKVPPFVFEGDNIQVSCAGYPGYYAGDAKLNKGDQ-LIGSS-PSGSFHFGRVTMATSGPYTCYRPVWHHSMYYSQVSSVVISVK658.
8GYLTLKVPPFVFEGDNLEVSCAGYPGYYAGAAKLNKGDQ-FIGNL-SSGNFHIGRVTMATSGPYTCYRPVWHHSMYYSQVSSVYISVK658.
10GYLSLKVPPFVFEGDNLQFSCAGYPGYKADTAKLNKGDQ-LIGFS-SSGNFHIGRVTMATSGPYTCYRPVRHDGMYYNKVSSVVISVK658.
4DYLSLKVPPFVFEGDNLQVSCAGYPGYKAESAKLYKGDQ-LIGSS-PSGNFHIGRVTMATSGPYTCYRYVWHHSMYHYKESSVYISVK658.
12DYLSLKVPPFVFEGDNLQVSCAGYPGYKADTAKLNKGNQ-LIGSS-PSGNFNFGRVTMATSGSYICYRAVKHHLIYYNQKSSVYISVK658.
11GYLALKVPPYVFEGDNLQVSCAGYPGYYTDTAKLYKGNQ-LIGGPSSSVNFNFGRVTMATSGSYTCYRAVKHHFIYYNQESSVYISVK658.
3DYLSLKVPPFVFEGDNLQVSCAGYPGYYGGNAKLYKGYE-FMASS-GTGSFQIGRVTMATSGPYTCYRYVWHHSAYHSKWSSVVISVK658.
13GYLTLKVPPFVFEGDYLEVSCAGYPGYEAGTARLYKGNDTMIGSS-STGSLWIGAVAMATSGCYTCYRRVVHHGYSYEKKSDVYISVKGFA'ABC'ECGBMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page10of17(pagenumbernotforcitationpurposes)enablingexpressionofXFL2,XFL1.
7andtheX.
laevisFcRγsubunitasrecombinantepitope-taggedproteins.
XFL2andXFL1.
7wereexpressedin293Tcellsasrecombinanthemagglutinin(HA)-taggedproteinscontainingeithertheirownTM1regionsortheTMregionofPDGFR.
TheX.
laevisFcRγchainwastaggedwithc-mycepitope.
SingletransfectionofXFL2-HAdidnotinduceitssurfaceexpres-sion,andtheproteinaccumulatedintracellularlyasdeter-minedbyimmunofluorescentmicroscopyofpermeabilizedcells.
However,thesurfaceexpressionofXFL2wasrestoredwhenitwasco-transfectedwithFcRγ(Fig.
8).
IncontrasttoXFL2,XFL1.
7-HAwastargetedtothesurfaceofthetransfectedcellsintheabsenceoftheadaptermolecule,althoughitssurfaceexpressionwasincreasedtwo-foldinthepresenceofFcRγ.
BothXFL2andXFL1.
7proteinswerereadilyexpressedonthecellsurfacewhentheirECregionswerefusedwithaTMofPDGFR.
TheseresultsshowthatXFLmoleculescontainingaTM1regionwiththeNxxRmotifassociatewiththeFcRγchain.
ThesurfaceexpressionofXFL1.
7islessdependentonthepresenceofFcRγchain.
Thismaybeexplainedbydiver-gentstructureoftheirTMs.
SuchdifferencesalsohavebeenobservedamongtheFcRγ-associatingKIRLs;FcRγiscriticalforsurfacetargetingofLILRA2,butnotFcαRorOSCAR[27,29,31].
DiscussionComparativestudiesofmammalsandchickenhaverevealedanunexpectedstructuralandfunctionalvariabil-ityofthepairedreceptorfamiliessuchasFCRLsandKIRLs[22,25,26,34-37].
Studieshavealsoindicatedthattherep-ertoiresofthesefamilieshaveevolvedinaspecies-specificmanner.
Theevolutionaryfactorsresponsibleforsuchdiversityremainpoorlyunderstood.
Therecentdescrip-SouthernblotanalysisofXenopuslaevisgenomicDNAFigure6SouthernblotanalysisofXenopuslaevisgenomicDNA.
HybridizingprobescorrespondedtotheexonsforD1domainofXFL2orD3domainsofXFL1.
1andXFL3.
1.
06.
0PvuIIXFL2_D1kbXFL1_D30.
53.
06.
010.
0kbPstIBamHIPvuIIXFL3_D3PvuII10.
05.
01.
03.
0kb1.
0Table1:VariationsinXFLexpressioninX.
laevisadults(A)andtadpoles(T,stages46–58).
XFL1.
21.
71.
81.
101.
111.
123SpleenA/T+/+-/NTThymusA/TLiverA/T-/-+/NTNT-/-IntestineA/TLungA/TBrainAGillsTNT–nottested.
ResultsareindicativeofRT-PCRperformedontotalRNAextractsfromdifferenttissuesamples(n=3).
BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page11of17(pagenumbernotforcitationpurposes)tionofLITRsasputativeteleostcounterpartsofbothFCRLsandKIRLsleftmanyquestionsunanswered,sinceonlyaweaksimilarityofLITRstothemammalianandavianreceptorshasbeenfound[6,44].
Thepresentstudyfillsthegapbyextendingtheanalysisofpairedreceptorfamiliestoamphibians,themostprimitivebranchoftetrapods.
InthediploidX.
tropicaliswehaveidentifiedatleast75genescodingforpairedFcR-likecellsurfacereceptors.
Themerefactofthefamilyexpansionisnotunusual.
Rapidevolutionarychangeofagenecontentknownasthe"expansion-contraction"or"birth-and-death"processhasbeendocumentedinmanyfamiliesofimmunity-relatedreceptors[45].
WhatdistinguishestheFcRfamilyfrommanyotherpairedreceptorfamiliesistheextraordinarystructuraldiversityofitsmembers.
CombinatorialjoiningofsixIgdomainsubtypesgeneratesasmanyas24ECarchitectures.
WhenweconsidertheTMsubtypesasdis-tinctdomains,thenumberoftheXFLdomainarchitec-turesincreasesto35.
Noneofthesevariantsaresharedbyeitherhumanormousehomologs,althoughfiveIg-domainsubtypesarecommonfortheXenopusandmam-malianproteins.
Overall,50differentdomainarchitec-turescanbedefinedamonghuman,mouse,andX.
tropicalisFcR-relatedproteins.
RequirementsfortheexpressionofXFL1.
7andXFL2onthecellsurfaceFigure8RequirementsfortheexpressionofXFL1.
7andXFL2onthecellsurface.
Epitope-taggedXFL1.
7orXFL2wereectopicallyexpressedintheirnativeforms,orwiththeTMregionsreplacedbythatofPDGFRintransientlytrans-fected293Tcells.
Effectofco-transfectionwithFcRγchainwasalsostudied.
ImmunocytochemicalstainingoftheXFL2-transfectedcellsisshownatright.
Transfectionefficiencyisshownaspercentageofantigen-positivecells.
22141617614NorthernblotanalysisofXFLmRNAdistributioninX.
laevistissuesFigure7NorthernblotanalysisofXFLmRNAdistributioninX.
laevistissues.
PooledtotalRNAfromsix6-montholdfrogswerehybridizedunderlowstringencyconditionswiththeD3exonofXFL1.
5asanuniversalprobeforgroupIXFLgenes.
spleenthymuskidneylungmusclebrainintestine1.
53.
04.
0kbXFL1.
5D3-actinBMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page12of17(pagenumbernotforcitationpurposes)ExceptfortheD1-andD2-subtypedomainsnohomolo-gousstructuralelementswerefoundbetweenXFLsandcatfishLITRs(notshown).
OtherIg-domainsubtypescomposingECofLITRsseemtobemoresimilartotheKIRLdomains.
SuchmixeddomaincompositionhasbeensuggestedtopredatetheKIRandFcRfamilies[6].
TheXFLdataareconsistentwiththesuggestionsthattheFcRandKIRfamiliesshareevolutionaryroots[33,46].
AstrongargumentinfavorofthismodelisthefactthatactivatingFcRsuseapeculiarTMmodule(TM1)toassociatewithFcRγchainthatishomologoustoTMsofactivatingmam-malianKIRLs.
TM1appearstobeanancestralelementoftheprimordialFcR/KIRfamilythathasbeenlostbyCHIRs,classicalFcRsandKIRs,butretainedbyXFLsaswellasXenopusandmammalianKIRLs.
Althoughavailabledatadonotallowinferringthestruc-tureoftheKIRLandFCRLancestor,itisclearthatthefam-ilyevolvedbyinter-andintragenicrecombinationsinaspecies-specificway.
Theformermechanismgaverapidchangeinthenumberofgenesperfamily(birth-and-death),whereasthelatterwasresponsibleforextensivedomainshuffling.
TheFcR-relatedreceptorsindifferentvertebratespeciesaresimilarintheirsubdivisionintoacti-vatingandinhibitoryformsandpredominantexpressioninlymphoidtissues.
However,theratioofinhibitorytoactivatingmembers,thecellulardistribution,andtheexactamountandarchitectureofectodomainsareuniqueineachexaminedspecies.
WhatmightbetheevolutionaryforcesresponsibleforthisdegreeofdiversitycommonamongtheFcR-andKIR-relatedreceptors,andXFLsinparticularWecantrytoanswerthisquestionbyinferencefromtheattributedfunctionoftheactualmammalianreceptors,whichistoregulateimmuneresponses.
ClassicalFcRsregulateBcellresponsesbybindingtotheIgGandIgEimmunecom-plexes,whereasKIRs,oratleasttheirinhibitoryforms,regulateNKcellfunctionbybindingspecificallytoMHCclassImolecules.
However,Ig-bindingappearstobeasec-ondaryor(derived)specialization,sinceourprevious[37]andpresentdatatogetherwiththedefinitionofachickenIgYreceptorasamemberoftheCHIRfamily[47]stronglyarguethatclassicalFcRshaveemergedaftertheseparationofmammalsandbirds.
MHC-recognitionasapotentialancientfunctionofFCRLs/KIRLsismoreattractive.
Theabilitytointeractwithclassicalandnon-classicalMHCclassImoleculesisafeatureofsomeKIRLs[11-13,48-50],ithasbeenalsosuggestedformammalianFCRLs[51]andcatfishLITRs[46].
Fromthispointofviewthediversifica-tionofFCRL/KIRLsmayhavebeendrivenbythenecessitytomatchtherapidevolutionofMHClociunderpathogenpressureof,asithasbeensuggestedfortheKIRandLy49geneclusters[[32,52]and[53]].
ThismaybethecaseforXFLstooasthereareatleast20non-classicalMHCclassIinX.
laevis[54].
Therearehoweversomeinconsistencieswiththissce-nario.
First,thescopeofvariabilityofdomainarchitec-turesamongFCRLsseemstobeexcessivelyhighrelativetoMHCclassImolecularstructure.
Second,thefunctionsofmammalianKIRLsarenotlimitedtoMHCantigen-bind-ing.
Amongligandsofthesereceptors,therearealsocolla-gens(GPVIandLAIR1),IgA(humanFcαR),IgG(bovineFcγR2)andintegrins(mouseGp49B1)[55-59].
Humanα1-B-glycoprotein,adistantsecretedrelativeofKIRLs,bindstothecysteine-richsecretoryprotein3[60],whileitsopossumhomologisasnakevenommetalloprotein-ase-neutralizingfactor[61].
DuetotheindependentexpansionofKIR-relatedreceptorsinmammals,birds[34,35]andamphibians[ourunpublisheddata]itisdiffi-culttodeterminewhichofthesefunctionsaretrulyancientorofancestraltype.
Finally,mouseLy49isaclearexampleoftheself-MHCrecognitionservedbyreceptorsstructurallydifferentfromKIRLsandFCRLs.
AnalternativeexplanationfortheextraordinarydiversityofXFLsandotherFcR-andKIR-relatedreceptorsmaybethattheyaredirectlyinvolvedininnateimmunity.
Com-binatorialdiversityisahallmarkoftheimmunesystemanditisusuallyassociatedwithrecognitionofpathogens.
Thecapacityofpairedreceptorstodirectlybindtopatho-gensiswelldocumented[62-66].
InthelatestofthesestudiesithasbeenfoundthatmousePIR-B,anditshumanrelativesLILRB1andLILRB3recognizeStaphyloco-ccusaureusandmodulateTLR-mediatedinflammatoryresponsesagainstthisbacterium.
Thesefactsareusuallyinterpretedintermsofadaptivecoevolutionofthemicro-organisms,whichimpliesthatthepathogenrecognitionisasecondaryorderivedfunction.
However,theextensivevariabilityoftheFcR/KIRrelativesraisesthepossibilitythatthesereceptorsexpandedprimarilytofightpatho-gens,whereastheknownimmuno-regulatoryfunctionsmayrepresentsecondaryacquisitionsorspecializations.
Dependingonthenatureofthepathogenandthesignalpropertiesofthereceptors,itisclearthatpathogen-recep-torinteractionmaybeeitheradvantageousordetrimentalforthehost,andassuch,mayrapidlychangetheratioofactivatingversusinhibitoryreceptors,aswellastheirrespectiveamountandspecificities.
Inthisregard,aparal-lelmaybedrawnwiththespecies-specificexpansionofvariousreceptorfamiliesininvertebratesthatparticipateininnateimmuneresponses[67,68].
Whileitmaybedif-ficulttoobtaindirectevidencetosupportthisscenario,itclearlydeservesattention.
TheelucidationofthefactorsresponsibleforthediversificationoftheFcR-andKIR-relatedreceptorsmaycontributetobetterunderstandBMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page13of17(pagenumbernotforcitationpurposes)theirfunctionandultimatelydevelopnewtherapiesbasedontheirimmunoregulatoryproperties.
ConclusionOurstudyshowsthatintwoamphibianspeciesXenopustropicalisandX.
laevis,pairedreceptorshavediversifiedintoalargefamilyofgenes,XFLs,preferentiallyexpressedinlymphoidtissues.
TheextracellularregionsofthesereceptorsarecomposedofonetoelevenIg-likedomainsbelongingtosixstructuralsubtypes.
AfractionofXFLsuseaTMmodule(TM1)toassociatewithFcRγsignalingsub-unit.
TM1ishighlysimilartoTMsofactivatingFcRγ-asso-ciatingKIRLs.
ThisfactstronglyarguesinfavorofacommonevolutionaryoriginoftheFcRandKIRfamilies.
ThevariationinnumberandcompositionofdistinctIg-likeandTMdomainsubtypesgeneratesstrikingdiversityofdomainarchitecturesamongXFLs.
Phylogeneticanaly-sisshowsthatthisdiversityemergedinalineage-specificmanner.
ClassicalFcRsandotherknownmammalianFcR-relatedproteinsappeartobespecifictomammals.
ThecontinualandextensivediversificationofdomainarchitecturesintheFcRandKIRfamiliesindicatesastrongselectionpressurenotcompletelyconsistentwiththeusualassumptionthatpairedreceptorshavebeenprima-rilyselectedtoregulateimmuneresponses.
WeproposethatFcR/KIR-relatedreceptorsmighthaveprimarilyexpandedaspathogen-recognizingcomponentsofinnateimmunitywhiletheirknownphysiologicalfunctionshavebeenacquiredlaterinalineage-specificmanner.
MethodsExperimentalAnimalsAdultandlarvaloutbredXenopuslaeviswereobtainedfromtheX.
laevisResearchResourceforImmunobiologyattheUniversityofRochesterMedicalCenter[69].
LarvaldevelopmentstagesweredeterminedaccordingtoNieu-wkoopandFaber[70].
AllanimalswerehandledunderstrictlaboratoryandUCARregulations.
Adultsandlarvaewereeuthanizedwith0.
5%and0.
1%Tricainemeth-anesulfonate(TMS),respectively.
cDNAlibraryconstructionandscreeningcDNAlibrariesfrom2μgspleentotalRNAfromXenopuslaevisadultsorfroglets(stage60–62)wereconstructedusingSMARTcDNALibraryConstructionKit(Clontech).
FirststrandcDNAwasamplifiedusingAdvantage2RCREnzymeSystem(Clontech).
Sizefractionationwasachievedbyseparationonasepharosecolumnsand0.
5–4kbcDNAswereligatedintolambdaTriplEx2armsandthenpackedwithGigapackIIIGoldCloningKit(Strata-gene).
Librariescontaining106independentrecombinantcloneswereamplified.
cDNAlibrariesfromXenopuslaevis/gilliLG7hybrids(fromadultspleenortadpolespleenandliverRNA)werekindlyprovidedbyDr.
LouisDuPasquier(UniversityofBasel,Switzerland).
AllfourcDNAlibrariesdescribedwerescreenedusing32P-labeledPCRfragmentcodingforthefirstD3domainofXFL1.
2(279bp)asdescribedbySambrooketal.
[71].
PlasmidscontainingcDNAinsertswererecoveredfromisolatedpositivephagesbyinvivoexcision.
cDNAsweresequencedusinganautomatedfluorescentsequencerABI-Prizm310(AppliedBiosystems).
GenBankaccessionsofcDNAclonesXenopusESTcDNAclonesdc12e01,dai46h06,daa24c04,dab24g06andNISC_mp06d01wereobtainedfromtheI.
M.
A.
G.
E.
Consortium[72]throughATCC(USA)orResearchGeneticsInc(USA),sequencedasdescribedaboveandsubmittedtoGenBank.
AccessionnumbersforthesecDNAsare[GenBank:AY293300],[GenBank:AY293303],[GenBank:AY293305],[GenBank:AY297106],and[GenBank:EF591296],respectively.
[GenBank:AY293301],[GenBank:AY293302],[Gen-Bank:AY293304],[GenBank:AY297104],[GenBank:AY297105],[GenBank:DQ367411],[GenBank:DQ367415]and[GenBank:EF431890–EF431893]acces-sionnumberswereassignedtocDNAsequencesobtainedthroughcDNAlibraryscreening.
SouthernblotanalysisGenomicDNAfromXenopuserythrocyteswasisolatedasdescribedbySambrooketal.
[71]anddigestedtocomple-tionwithrestrictionendonucleasesBamHI,HindIIIorPvuII.
ThedigestedDNA(10μg/lane)wasseparatedon1%agarosegelandtransferredontoZeta-probenylonmembranes(BioRadLaboratories)bythevacuumblot-tingtechniquein0.
4MNaOH.
Hybridizationswith32P-labeledprobeswereperformedfollowingthemembranemanufacturer'srecommendations.
TheprobeswerePCRamplifiedfragmentscodingforthefirstD3domainofXFL1.
2(279bp),D1domainofXFL2(215bp)orD3domainofXFL3(239bp).
RNAextraction,cDNAsynthesis,andRT-PCRamplificationTissuesampleswerehomogenizedin0.
8mLofTrizolrea-gent(Invitrogen).
TotalRNAwasextractedaccordingtothemanufacturer'sprotocol.
AsampleRNApelletwasresuspendedinRNasefreewaterandquantifiedwithSmartSpecspectrophotometer(BioRad).
500ngofquan-tifiedtotalRNAwereusedtosynthesizecDNAwithiScriptfirststrandcDNAsynthesiskit(BioRad)accordingtothemanufacturer'sprotocol.
NegativeRTcontrolswererunforeachsampleatthesametime.
cDNAandnegativeRTcontrolsamplesweredilutedthreetimestoafinalvolumeof60μlbeforeproceedingtoPCRamplification.
ForeachPCRreaction(30μltotalvolume)3μlof2mMdNTPs,3μlof10*PCRbuffer,10pmolofeachprimer,2UofTaqDNApolymerase(LifeTechnologies),and1μlofcDNAwereused.
Thentubesweresetfor35cycles:45secatBMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page14of17(pagenumbernotforcitationpurposes)95°C,45secat56–64°Cand30–90secat72°C.
Thefol-lowingprimerswereusedforRT-PCR:XFL1.
2forward–5'GGAAGCTATCAGTGCCAAACA3',reverse–5'TGAGTCTCCTGGGAGGACAGA3',XFL1.
7forward–5'ACACCAAAGAGGCTGCAGTTC3',reverse–5'GATGAG-GAGCATCTTCATGGT3',XFL1.
8forward–5'ATCGCTATCGCTCTAATGGAGC3',reverse–5'CAGTCTCGTGAGATTCAGCCG3',XFL1.
10–forward5'GACCAAGTGGACATTGTTGTGC3',reverse–5'TTCTC-CGGCCTGTCCACCTC3',XFL1.
11forward–5'CTCAG-GATTCCATCCAAAGTG3'reverse–5'CTTGGTCCAGTCCCGCACTG3',XFL1.
12forward–5'AGATGCACCCGACAAGTGAAGA3',reverse–5'TCAG-GACAGCCAGTGCTACTG3',XFL3forward–5'CTA-CACAAGGATACAACCCTG3',reverse–5'TTCTTGGGCATCACCAGAGAG3',andasinternalcon-trol,β2M,forward–5'CCCTTGTGGTGTAACTGTGCTC3',reverse–5'GCACACACCAATCAGAAAAAGGAC3'.
Negative(RT)controlswerealsoperformedwithsameprimerstocontrolforgenomicDNAcontamination.
NorthernblotanalysisTotalRNA(10μg/lane)extractedandquantifiedasdescribedabove,wasseparatedon1%agarosegelwithformaldehyde[71]andtransferredontoZeta-probenylonmembranes(BioRadLaboratories)bycapillarytransferin20*SSC.
Hybridizationwith32P-labeledprobewasper-formedatnon-stringentconditionsfollowingthemem-branemanufacturer'srecommendations.
TheprobewasPCRamplifiedfragmentcodingforD3domainofXFL1.
5(282bp).
AsacontrolforRNAintegrityaprobeencodingX.
laevisβ-actinwasused.
ConstructionscDNAregionsencodinganextracellularorextracellularplustransmembranepartofXFL2(orXFL1.
7)wereclonedusingprimerswithXmaIandPstI(orSalI)sitesandligatedintopDisplay(Invitrogen)vectorwithanN-termi-nalHAepitopeandwithorwithoutaC-terminalPDGFRtransmembranedomain.
ThecDNAportionsusedwere45–809bpor45–938bpforXFL2cDNA[GenBank:AY293305]and225–806bpor225–938bpforXFL1.
7cDNA[GenBank:EF591296].
CompletecodingregionofX.
laevisFcRγcDNA[GenBank:EF431895]wasclonedusingprimerswithNheIandApaIsitesandligatedintopAP-Tag5(GenHunter)vectorwithaC-terminalc-mycepitope.
ImmunochemistryandflowcytometryConstructionsweretransientlytransfectedinto293TcellsusingUnifectin56(IBCH,Moscow,Russia)accordingtothemanufacturer'sprotocol.
Seventytwohoursaftertheyweretransfected,thecellswereusedforimmunocyto-chemistryandcytometricanalysis.
Forcellsurfacestain-ing,transfectedcellswerewashedtwicewithWashBuffer(PBS,containing1%FCSand0.
1%NaN3).
Thecellswerefirstincubatedwithrabbitanti-HA(Sigma)(anti-hemag-glutininprotein)inWashBufferfor30minonice.
CellswerethenwashedthreetimeswithcoldWashBufferandincubatedwithgoatanti-rabbitIg-FITC(BDBioscience)inWashBufferfor30minonice.
ThecellswerewashedthreetimeswithWashBufferandanalyzedusingamicro-scopeAxioscop2plusandFACSAriacytometer(BDBio-science).
Forintracellularstaining,transfectedcellsweresmearedonglassslides,fixedwithacetoneandstainedforFcRγsubunitwithanti-c-mycmonoclonalantibodies(Sigma)andgoatanti-mouseIgG-TexasRed(MolecularProbes).
BioinformaticstoolsNucleotideandaminoacidsequenceswereanalyzedusingutilitiesattheNCBI[73],EMBL[74]andBCM[75]websites.
AminoacidsequenceswerealignedusingClus-talutilitiesintheMEGA3software[39]andshadedwiththeBoxShadeprogram[76].
ThenucleotideandaminoacidsequencesofknowngeneswereretrievedfromtheGenBankusingENTREZattheNCBI[73].
ThegenomicsequenceswereretrievedfromandanalyzedattheEnsembl[77,78]orJGIwebsites[79].
HomologysearcheswereperformedusingTBLASTNandTFASTAprograms.
TheGeneScanprogram[80,81]andtheWebgenepro-grampackage[82,83]wereusedfortheautomatedgenestructureprediction.
TheXFL-surroundinggeneswereidentifiedusingtheEnsemblandJGIutilitiesandwereverifiedbyreciprocalsequencecomparisonsattheNCBIwebsiteusingtheBLASTPprogram.
PhylogeneticanalysiswasperformedwithMEGA3[39]fornucleotidesequencesofexonsandaminoacidsequencesofdomainsafteralignmentwiththeCLUSTALoption.
Incertaincases,theCLUSTALgeneratedalign-mentsweremanuallycorrected.
PhylogenetictreeswereconstructedusingthebootstrapandinteriorbranchtestsoftheNeighbor-joining(NJ)methodwithp-distances(proportionofdifferences).
MinimumEvolution(ME)treeswereessentiallythesameastheNJtreesinthemajorbranchingpatterns.
ListofabbreviationsIg:Immunoglobulin;FcR:classicalleukocyteFcReceptor;FCRL:FcR-Like;KIR:KillercellImmunoglobulinRecep-tor;KIRL:KIR-Like;XFL:XenopusFcR-Like;LITR:Leuko-cyteImmune-TypeReceptors;ITAM:ImmunoreceptorTyrosine-basedActivatingMotif;ITIM:ImmunoreceptorTyrosine-basedInhibitoryMotif;ITSM:ImmunoreceptorTyrosine-basedSwitchmotif;EST:ExpressedSequenceTag;EC:Extracellularregion;TM:Transmembraneregion.
BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page15of17(pagenumbernotforcitationpurposes)Authors'contributionsAVTandJRdesignedthestudy,SVG,TRandAMNper-formedmolecularstudies,AYEandAVTperformedgenomesearchandgenepredictions,madesequencealignmentsandphylogeneticanalysis,LVMcarriedoutcellstainingandflowcytometry,AVTwrotemanuscript,JRhelpedtodraftthemanuscript.
Allauthorsreadandapprovedthefinalmanuscript.
AdditionalmaterialAcknowledgementsTheexpertanimalhusbandryprovidedbyTinaMartinandDavidAlbrightisgratefullyappreciated.
WethankDrs.
NicolasCohenandAndreaBot-taroforcriticalreviewofthemanuscript.
WewouldlikealsotothankVadimKhaychukforhisactiveparticipationtothisstudy.
ResearchwassupportedbytheRussianFundforBasicResearch05-04-49268(A.
M.
N),U.
S.
CivilianResearchandDevelopmentFoundationRUB1-2837-NO-06(A.
V.
T,J.
R),RASProgram"Originandevolutionoflife"(A.
V.
T),NIHR01-CA-108982-02(T.
R.
),NIHR24-AI-059830andNSFMCB-0136536(J.
R.
).
References1.
HealyJI,GoodnowCC:Positiveversusnegativesignalingbylymphocyteantigenreceptors.
AnnuRevImmunol1998,16:645-670.
2.
RavetchJV,LanierLL:Immuneinhibitoryreceptors.
Science2000,290:84-89.
3.
AzumiK,DeSantisR,DeTomasoA,RigoutsosI,YoshizakiF,PintoMR,MarinoR,ShidaK,IkedaM,AraiM,InoueY,ShimizuT,SatohN,RokhsarDS,DuPasquierL,KasaharaM,SatakeM,NonakaM:GenomicanalysisofimmunityinaUrochordateandtheemergenceofthevertebrateimmunesystem:"waitingforGodot".
Immunogenetics2003,55:570-581.
4.
VogelC,TeichmannSA,ChothiaC:Theimmunoglobulinsuper-familyinDrosophilamelanogasterandCaenorhabditisele-gansandtheevolutionofcomplexity.
Development2003,130:6317-6328.
5.
NikolaidisN,KleinJ,NeiM:OriginandevolutionoftheIg-likedomainspresentinmammalianleukocytereceptors:insightsfromchicken,frog,andfishhomologues.
Immunoge-netics2005,57:151-157.
6.
StaffordJL,BengtenE,DuPasquierL,McIntoshRD,QuiniouSM,ClemLW,MillerNW,WilsonM:AnovelfamilyofdiversifiedimmunoregulatoryreceptorsinteleostsishomologoustobothmammalianFcreceptorsandmoleculesencodedwithintheleukocytereceptorcomplex.
Immunogenetics2006,58:758-773.
7.
CannonJP,HaireRN,MuellerMG,LitmanRT,EasonDD,TinnemoreD,AmemiyaCT,OtaT,LitmanGW:Ancientdivergenceofacomplexfamilyofimmune-typereceptorgenes.
Immunoge-netics2006,58:362-373.
8.
HawkeNA,YoderJA,HaireRN,MuellerMG,LitmanRT,MiracleAL,StugeT,ShenL,MillerN,LitmanGW:Extraordinaryvariationinadiversifiedfamilyofimmune-typereceptorgenes.
ProcNatlAcadSciUSA2001,98:13832-13837.
9.
DaeronM:Fcreceptorbiology.
AnnuRevImmunol1997,15:203-234.
10.
NimmerjahnF,RavetchJV:Fcγreceptors:oldfriendsandnewfamilymembers.
Immunity2006,24:19-28.
11.
CampbellKS,ColonnaM:Humannaturalkillercellreceptorsandsignaltransduction.
IntRevImmunol2001,20:333-370.
12.
MorettaL,MorettaA:Killerimmunoglobulin-likereceptors.
CurrOpinImmunol2004,16:626-633.
13.
LanierLL:NKcellrecognition.
AnnuRevImmunol2005,23:225-274.
14.
HatzivassiliouG,MillerI,TakizawaJ,PalanisamyN,RaoPH,IidaS,TagawaS,TaniwakiM,RussoJ,NeriA,CattorettiG,ClynesR,Men-delsohnC,ChagantiRS,Dalla-FaveraR:IRTA1andIRTA2,novelimmunoglobulinsuperfamilyreceptorsexpressedinBcellsandinvolvedinchromosome1q21abnormalitiesinBcellmalignancy.
Immunity2001,14:277-289.
15.
DavisRS,WangYH,KubagawaH,CooperMD:IdentificationofafamilyofFcreceptorhomologswithpreferentialBcellexpression.
ProcNatlAcadSciUSA2001,98:9772-9777.
16.
MechetinaLV,NajakshinAM,VolkovaOY,GuselnikovSV,FaizulinRZ,AlabyevBY,ChikaevNA,VinogradovaMS,TaraninAV:FCRL,anovelmemberoftheleukocyteFcreceptorfamilypossessesuniquestructuralfeatures.
EurJImmunol2002,32:87-96.
17.
GuselnikovSV,ErshovaSA,MechetinaLV,NajakshinAM,VolkovaOY,AlabyevBY,TaraninAV:AfamilyofhighlydiversehumanandmousegenesstructurallylinksleukocyteFcR,gp42andPECAM-1.
Immunogenetics2002,54:87-95.
18.
DavisRS,LiH,ChenCC,WangYH,CooperMD,BurrowsPD:Def-initionofanFcreceptor-relatedgene(FcRX)expressedinhumanandmouseBcells.
IntImmunol2002,14:1075-1083.
19.
DavisRS,StephanRP,ChenCC,DennisGJr,CooperMD:Differen-tialBcellexpressionofmouseFcreceptorhomologs.
IntImmunol2004,16:1343-1353.
20.
ChikaevNA,BykovaEA,NajakshinAM,MechetinaLV,VolkovaOY,PekloMM,ShevelevAY,VlasikTN,RoeschA,VogtT,TaraninAV:CloningandcharacterizationofthehumanFCRL2gene.
Genomics2005,85:264-272.
21.
WilsonTJ,PrestiRM,TassiI,OvertonET,CellaM,ColonnaM:FcRL6,anewITIM-bearingreceptoroncytolyticcells,isbroadlyexpressedbylymphocytesfollowingHIV-1infection.
Blood2007,109:3786-3793.
22.
DavisRS:Fcreceptor-likemolecules.
AnnuRevImmunol2007,25:525-560.
23.
EhrhardtGR,LeuCM,ZhangS,AksuG,JacksonT,HagaC,HsuJT,SchreederDM,DavisRS,CooperMD:Fcreceptor-likeproteins(FCRL):immunomodulatorsofBcellfunction.
AdvExpMedBiol2007,596:155-162.
24.
MaltaisLJ,LoveringRC,TaraninAV,ColonnaM,RavetchJV,Dalla-FaveraR,BurrowsPD,CooperMD,DavisRS:NewnomenclatureforFcreceptor-likemolecules.
NatImmunol2006,7:431-432.
25.
MartinAM,KulskiJK,WittC,PontarottiP,ChristiansenFT:Leuko-cyteIg-likereceptorcomplex(LRC)inmiceandmen.
TrendsImmunol2002,23:81-88.
26.
KelleyJ,WalterL,TrowsdaleJ:Comparativegenomicsofnatu-ralkillercellreceptorgeneclusters.
PLoSGenet2005,1:129-139.
27.
MortonHC,Herik-OudijkIEvanden,VossebeldP,SnijdersA,Ver-hoevenAJ,CapelPJ,WinkelJGvande:FunctionalassociationbetweenthehumanmyeloidimmunoglobulinAFcreceptor(CD89)andFcRγchain.
MolecularbasisforCD89/FcRγchainassociation.
JBiolChem1995,270:29781-29787.
28.
TaylorLS,McVicarDW:FunctionalassociationofFcεRIγwitharginine(632)ofpairedimmunoglobulin-likereceptor(PIR)-A3inmurinemacrophages.
Blood1999,94:1790-1796.
Additionalfile1AlignmentofdeducedaminoacidsequencesofD1-D5domainsofX.
tropicalis,X.
laevisXFLsandhumanFcR-relatedproteins.
X.
tropi-calisgenesaredesignatedaccordingtoascaffoldnumberandtheircon-secutivepositionatthecorrespondingscaffold(version4.
1).
Forproteinscontainingmultipledomainsofthesametypethesedomainsarenum-beredfromN-toC-terminus(i.
e.
D3.
1-D3.
3).
X.
laevisdomainsaredesignatedaccordingtothenameoftheclonedXFLcDNA(XFL1.
1-1.
12,XFL2andXFL3)orGenBankaccessionnumberoftheESTcDNA(i.
e.
BU903031).
Identicalandsimilarresiduesareshownbywhitelettersonblackandgraybackgrounds,respectively.
Dashesrepresentgapsintro-ducedtomaximizesimilarity.
Clickhereforfile[http://www.
biomedcentral.
com/content/supplementary/1471-2148-8-148-S1.
pdf]BMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page16of17(pagenumbernotforcitationpurposes)29.
NakajimaH,SamaridisJ,AngmanL,ColonnaM:HumanmyeloidcellsexpressanactivatingILTreceptor(ILT1)thatassoci-ateswithFcreceptorγ-chain.
JImmunol1999,162:5-8.
30.
BerlangaO,TulasneD,BoriT,SnellDC,MiuraY,JungS,MoroiM,FramptonJ,WatsonSP:TheFcreceptorγ-chainisnecessaryandsufficienttoinitiatesignallingthroughglycoproteinVIintransfectedcellsbythesnakeC-typelectin,convulxin.
EurJBiochem2002,269:2951-2960.
31.
MerckE,GaillardC,GormanDM,Montero-JulianF,DurandI,Zuraw-skiSM,Menetrier-CauxC,CarraG,LebecqueS,TrinchieriG,BatesEE:OSCARisanFcRγ-associatedreceptorthatisexpressedbymyeloidcellsandisinvolvedinantigenpresentationandactivationofhumandendriticcells.
Blood2004,104:1386-1395.
32.
YokoyamaWM,PlougastelBF:Immunefunctionsencodedbythenaturalkillergenecomplex.
NatRevImmunol2003,3:I304-316.
33.
DennisGJr,KubagawaH,CooperMD:PairedIg-likereceptorhomologsinbirdsandmammalsshareacommonancestorwithmammalianFcreceptors.
ProcNatlAcadSciUSA2000,97:13245-13250.
34.
ViertlboeckBC,HabermannFA,SchmittR,GroenenMA,DuPasquierL,GobelTW:Thechickenleukocytereceptorcom-plex:ahighlydiversemultigenefamilyencodingatleastsixstructurallydistinctreceptortypes.
JImmunol2005,175:385-393.
35.
LaunK,CoggillP,PalmerS,SimsS,NingZ,RagoussisJ,VolpiE,Wil-sonN,BeckS,ZieglerA,VolzA:Theleukocytereceptorcom-plexinchickenischaracterizedbymassiveexpansionanddiversificationofimmunoglobulin-likeLoci.
PLoSGenet2006,2:e73.
36.
TaylorAI,GouldHJ,SuttonBJ,CalvertRA:ThefirstavianIg-likeFcreceptorfamilymembercombinesfeaturesofmamma-lianFcRandFCRL.
Immunogenetics2007,59:323-328.
37.
FayngertsSA,NajakshinAM,TaraninAV:Species-specificevolu-tionoftheFcRfamilyinendothermicvertebrates.
Immunoge-netics2007,59:493-506.
38.
KimMK,HuangZY,HwangPH,JonesBA,SatoN,HunterS,Kim-HanTH,WorthRG,IndikZK,SchreiberAD:Fcγreceptortransmem-branedomains:roleincellsurfaceexpression,γchaininter-action,andphagocytosis.
Blood2003,101:4479-4484.
39.
KumarS,TamuraK,NeiM:MEGA3:IntegratedsoftwareforMolecularEvolutionaryGeneticsAnalysisandsequencealignment.
BriefBioinform2004,5:150-163.
40.
DuPasquierL,SchwagerJ,FlajnikMF:TheimmunesystemofXenopus.
AnnuRevImmunol1989,7:251-275.
41.
RobertJ,CohenN:Evolutionofimmunesurveillanceandtumorimmunity:studiesinXenopus.
ImmunolRev1998,166:231-243.
42.
DuPasquierL,RobertJ,CourtetM,MussmannR:B-celldevelop-mentintheamphibianXenopus.
ImmunolRev2000,175:201-213.
43.
CohenN,DimarzoS,Rollins-SmithL,BarlowE,Vanderschmidt-Par-sonS:Theontogenyofallotoleranceandself-toleranceinlar-valXenopuslaevis.
InMetamorphosisEditedby:BallsM,BownsM.
Oxford:ClarendonPress;1985:388.
44.
StaffordJL,WilsonM,NayakD,QuiniouSM,ClemLW,MillerNW,BengténE:IdentificationandcharacterizationofaFcRhomologinanectothermicvertebrate,thechannelcatfish(Ictaluruspunctatus).
JImmunol2006,177:2505-17.
45.
NeiM,RooneyAP:Concertedandbirth-and-deathevolutionofmultigenefamilies.
AnnuRevGenet2005,39:121-152.
46.
StaffordJL,BengtenE,DuPasquierL,MillerNW,WilsonM:Chan-nelcatfishleukocyteimmune-typereceptorscontainaputa-tiveMHCclassIbindingsite.
Immunogenetics2007,59:77-91.
47.
ViertlboeckBC,SchweinsbergS,HanczarukMA,SchmittR,DuPasquierL,HerbergFW,GbelTW:Thechickenleukocytereceptorcomplexencodesaprimordial,activating,high-affinityIgYFcreceptor.
ProcNatlAcadSciUSA2007,104:11718-23.
48.
ManavalanJS,RossiPC,VladG,PiazzaF,YarilinaA,CortesiniR,Man-ciniD,Suciu-FocaN:HighexpressionofILT3andILT4isagen-eralfeatureoftolerogenicdendriticcells.
TransplImmunol2003,11:245-258.
49.
BrownD,TrowsdaleJ,AllenR:TheLILRfamily:modulatorsofinnateandadaptiveimmunepathwaysinhealthanddisease.
TissueAntigens2004,64:215-225.
50.
TakaiT:ANovelRecognitionSystemforMHCClassIMole-culesConstitutedbyPIR.
AdvImmunol2005,88:161-192.
51.
DavisRS,DennisGJr,OdomMR,GibsonAW,KimberlyRP,BurrowsPD,CooperMD:Fcreceptorhomologs:newestmembersofaremarkablydiverseFcreceptorgenefamily.
ImmunolRev2002,190:123-36.
52.
TrowsdaleJ,ParhamP:Mini-review:defensestrategiesandimmunity-relatedgenes.
EurJImmunol2004,34:7-17.
53.
KhakooSI,RajalingamR,ShumBP,WeidenbachK,FlodinL,MuirDG,CanavezF,CooperSL,ValianteNM,LanierLL,ParhamP:Rapidevo-lutionofNKcellreceptorsystemsdemonstratedbycompar-isonofchimpanzeesandhumans.
Immunity2000,12:687-698.
54.
FlajnikMF,KasaharaM,ShumBP,Salter-CidL,TaylorE,DuPasquierL:AnoveltypeofclassIgeneorganizationinvertebrates:alargefamilyofnon-MHC-linkedclassIgenesisexpressedattheRNAlevelintheamphibianXenopus.
EmboJ1993,12:4385-4396.
55.
ClemetsonJM,PolgarJ,MagnenatE,WellsTN,ClemetsonKJ:TheplateletcollagenreceptorglycoproteinVIisamemberoftheimmunoglobulinsuperfamilycloselyrelatedtoFcαRandthenaturalkillerreceptors.
JBiolChem1999,274:29019-29024.
56.
LebbinkRJ,deRuiterT,AdelmeijerJ,BrenkmanAB,vanHelvoortJM,KochM,FarndaleRW,LismanT,SonnenbergA,LentingPJ,MeyaardL:Collagensarefunctional,highaffinityligandsfortheinhib-itoryimmunereceptorLAIR-1.
JExpMed2006,203:1419-1425.
57.
MonteiroRC,WinkelJGVanDe:IgAFcreceptors.
AnnuRevImmunol2003,21:177-204.
58.
ZhangG,YoungJR,TregaskesCA,SoppP,HowardCJ:Identifica-tionofanovelclassofmammalianFcγreceptor.
JImmunol1995,155:1534-1541.
59.
CastellsMC,KlicksteinLB,HassaniK,CumplidoJA,LacoutureME,AustenKF,KatzHR:gp49B1-α(v)β3interactioninhibitsanti-gen-inducedmastcellactivation.
NatImmunol2001,2:436-442.
60.
UdbyL,SorensenOE,PassJ,JohnsenAH,BehrendtN,BorregaardN,KjeldsenL:Cysteine-richsecretoryprotein3isaligandofalpha1B-glycoproteininhumanplasma.
Biochemistry2004,43:12877-12886.
61.
Neves-FerreiraAG,PeralesJ,FoxJW,ShannonJD,MakinoDL,Gar-rattRC,DomontGB:StructuralandfunctionalanalysesofDM43,asnakevenommetalloproteinaseinhibitorfromDidelphismarsupialisserum.
JBiolChem2002,277:13129-13137.
62.
AraseH,MocarskiES,CampbellAE,HillAB,LanierLL:Directrec-ognitionofcytomegalovirusbyactivatingandinhibitoryNKcellreceptors.
Science2002,296:1323-1326.
63.
ArnonTI,AchdoutH,LiebermanN,GazitR,Gonen-GrossT,KatzG,Bar-IlanA,BloushtainN,LevM,JosephA,KedarE,PorgadorA,MandelboimO:Themechanismscontrollingtherecognitionoftumor-andvirus-infectedcellsbyNKp46.
Blood2004,103:664-672.
64.
ZebhauserR,KammererR,EisenriedA,McLellanA,MooreT,Zim-mermannW:IdentificationofanovelgroupofevolutionarilyconservedmemberswithintherapidlydivergingmurineCeafamily.
Genomics2005,86:566-580.
65.
KuespertK,PilsS,HauckCR:CEACAMs:theirroleinphysiol-ogyandpathophysiology.
CurrOpinCellBiol2006,18:565-571.
66.
NakayamaM,UnderhillDM,PetersenTW,LiB,KitamuraT,TakaiT,AderemA:PairedIg-likereceptorsbindtobacteriaandshapeTLR-mediatedcytokineproduction.
JImmunol2007,178:4250-4259.
67.
FlajnikMF,DuPasquierL:Evolutionofinnateandadaptiveimmunity:canwedrawalineTrendsImmunol2004,25:640-644.
68.
SchulenburgH,BoehnischC,MichielsNK:HowdoinvertebratesgenerateahighlyspecificinnateimmuneresponseMolImmunol2007,44:3338-3344.
69.
XenopuslaevisResearchResourceforImmunobiology[http://www.
urmc.
rochester.
edu/smd/mbi/xenopus/index.
htm]70.
NieuwkoopPD,FaberJ:NormalTableofX.
laevislaevis(Daudin)NewYorkandLondon:GarlandPublishing,Inc;1994.
71.
SambrookT,FritschEF,ManiatisT:Molecularcloning:alaboratoryman-ualPlainview,NY:ColdSpringHarborLaboratoryPress;1989.
72.
LennonG,AuffrayC,PolymeropoulosM,SoaresMB:TheI.
M.
A.
G.
E.
Consortium:anintegratedmolecularanalysisofgenomesandtheirexpression.
Genomics1996,33:151-152.
73.
NationalCenterforBiotechnologyInformation[http://www.
ncbi.
nlm.
nih.
gov]PublishwithBioMedCentralandeveryscientistcanreadyourworkfreeofcharge"BioMedCentralwillbethemostsignificantdevelopmentfordisseminatingtheresultsofbiomedicalresearchinourlifetime.
"SirPaulNurse,CancerResearchUKYourresearchpaperswillbe:availablefreeofchargetotheentirebiomedicalcommunitypeerreviewedandpublishedimmediatelyuponacceptancecitedinPubMedandarchivedonPubMedCentralyours—youkeepthecopyrightSubmityourmanuscripthere:http://www.
biomedcentral.
com/info/publishing_adv.
aspBioMedcentralBMCEvolutionaryBiology2008,8:148http://www.
biomedcentral.
com/1471-2148/8/148Page17of17(pagenumbernotforcitationpurposes)74.
EuropeanBioinformaticsInstitute[http://www2.
ebi.
ac.
uk]75.
BaylorCollegeofMedicineHumanGenomeSequencingCenter[http://www.
hgsc.
bcm.
tmc.
edu]76.
Boxshadeserver[http://www.
ch.
embnet.
org/software/BOX_form.
html]77.
EnsemblGenomeBrowser[http://www.
ensembl.
org]78.
HubbardT,BarkerD,BirneyE,CameronG,ChenY,ClarkL,CoxT,CuffJ,CurwenV,DownT,DurbinR,EyrasE,GilbertJ,HammondM,HuminieckiL,KasprzykA,LehvaslaihoH,LijnzaadP,MelsoppC,MonginE,PettettR,PocockM,PotterS,RustA,SchmidtE,SearleS,SlaterG,SmithJ,SpoonerW,StabenauA,StalkerJ,StupkaE,Ureta-VidalA,VastrikI,ClampM:TheEnsemblgenomedatabaseproject.
NucleicAcidsRes2002,30:38-41.
79.
JointGenomeInstituteEukaryoticGenomics[http://genome.
jgi-psf.
org]80.
TheNewGENSCANWebServer[http://genes.
mit.
edu/GENSCAN.
html]81.
BurgeC,KarlinS:PredictionofcompletegenestructuresinhumangenomicDNA.
JMolBiol1997,268:78-94.
82.
WebGeneHomePage[http://www.
itb.
cnr.
it/sun/webgene/]83.
MilanesiL,D'AngeloD,RogozinIB:GeneBuilder:interactiveinsilicopredictionofgenestructure.
Bioinformatics1999,15:612-621.