0.15b.faloo.com

TissueAnrigens1996:48;192-198PrintedinDenmark.
A//rig110resewedCopyrighrQMunksgaard1996TISSUEANTIGENSISSN0001-2815ResultsofExpedicionHurnana11.
AnalysisofHLAclassI1allelesinthreeAfricanAmericanpopulationsfromColombiausingthePCWSSOP:identificationofanovelDQB1*02(*0203)alleleE.
A.
Trachtenberg,G.
Keyeux,J.
Bernal,J.
A.
Noble,H.
A.
Erlich.
ResultsofExpedicionHumana.
11.
AnalysisofHLAclass11allelesinthreeAfricanAmericanpopulationsfromColombiausingthePCWSSOP:identificationofanovelDQB1*02(*0203)allele.
TissueAntigens1996:48:192-198.
0Munksgaard,1996PCWSSOPtypingmethodswereusedtoanalyzetheHLAClassI1DRB1,DQAl,DQB1andDPB1lociofsamplesfromthreeAfricanAmericanpopu-lationsofColombia.
FortysamplesfromtheCauca(Pacific),andtwentysam.
pleseachfromtheChoco(NorthPacificCoast)andtheProvidencia(Caribbeanisland)populations,werecollectedandtheClass11locianalyzedundertheauspicesoftheExpedicionHumana.
Despitethelimitednumberofsamplesanalyzed,theAfricanColombianpopulationsexhibitaveryhighdegreeofclassIIpolymorphism.
AgreatdiversityofDRB1alleleswasfound,withrepresentativesfromallserologicalclasses,including19DRB1allelesintheProvidencia,16intheCaucaand14intheChocogroups.
Inaddition,anovelDQB1*02allele(*0203)wasfoundintwoindividualsfromtheCaucapopulationofthePacificCoast.
ThesequenceoftheDQB1*0203allele,asso.
ciatedwithDR3,differsfromDQB1*0201byonlyonenucleotidesubstitu-tion(C+A)inthesecondpositionofcodon57,resultinginanAlatoAspchange.
TheadditionofDQB1*0203bringsthetotalnumberofDQB1allelesidentifiedtodateto26.
HLAclassI1diversityismuchgreaterintheseAfricanColombianpopulationsthanthatseeninnearbyAmerindianpopulations.
AnalysisofregionalColombianAfricanAmericanHLApopulationgeneticsisdiscussedwithrespecttotheColombianAmerindianHLAgeneticsIdescribedinanaccompanyingpaper.
ThestudyofHLAclassI1variationhasincreaseddramaticallysincetheadventofthepolymerasechainreaction(1-3)whichhasgreatlysimplifiedthemethodsofallelicidentification,andhasledtoasteadyinfluxofHLApopulationgeneticdata(forexample,seeref.
4).
TheidentificationandanalysisofHLAclassI1(DRB1-DQA1-DQB1-DPB1)di-versityinuniqueandisolatedpopulationsissignifi-canttomanyfieldsofmedicineandresearch,in-cludingtissuetransplantation(9,forensics(6).
aswellastostudiesofdiseasesusceptibilityandresis-E.
A.
Trachtenberg',G.
Keyeux2,J.
Bernal',J.
A.
Noble'andH.
A.
Erlich''DepartmehtofHumanGenetics,RocheMolecularSystems,AlamedaandChildren'sHospitalOaklandResearchInstitute,Oakland,California,USA,'FacultaddeMedicina,PontificiaUniversidadJaveriana,Bogota.
ColombiaKeywords:African-Colombian-HLA-DO-MHC-PCR-populationgeneticsReceived23April,revised,acceptedforpublication14May1996tance(7-9),andstudiesofmolecularevolutionandanthropology.
Mostofthesequence-definedpoly-morphismintheclassI1genesislocalizedtothesecondexon,whichencodesthepeptidebindinggrooveofthecell-surfaceheterodimeronantigenpresentingcells.
Recognitionofpeptides,derivedfromforeignandself-antigens,presentedbytheap-propriateclassI1moleculeresultsintheactivationofCD4+cells.
HerewereportHLAclassI1allelicfrequenciesfromthreeisolatedAfricanAmericanpopulations192HLAclass11allelesinAfricanColombiansouttheworld-100,000-200,000yearsago(re-viewedin11).
TheSouthAmericanIndiansaremuchyoungerpopulationswithestimatesofthepeoplingofSouthAmericarangingfrom-10,000-40,000yearsago(12-16).
AlthoughtheAfricanAmericanandAmerindianpopulationshaveverydifferentevolutionaryhistoriesbeforesettlinginColombia,thesepopulationshave,sincetheirarrivalinColombia,livedinidenticalenvironmentsforcenturies.
ThiscreatesauniqueopportunitytostudytheevolutionoftheHLAclass11systemintwoverydisparatepopulationswithinaknownvicinityandtimeframe.
MaterialsandmethodsSamplesBloodsamplesfromthreeregionallydistinctCo-lombianAfricanAmericanpopulationswerecol-lected,includingtheCauca(N=20;Pacific),Choco(N=20;NorthPacificCoast)andtheProvidencia(N=30;Caribbeanisland).
Inthisstudy,noneofthesamplesanalyzedarepresumedtobefromfirst-de-greerelatives.
GenomicDNAwasextractedfromperipheralbloodleukocytesusingapreviouslypub-lishedmethod(17).
Figure1.
LocationoftheAfricanAmericanpopulationsfromForthisstudy,samplesfromthreeenvironmentallydistinctre-gionsofColombiawerecollectedundertheauspicesoftheEx-pedicionHumanaandanalyzedattheDRB1,DQAI.
andDQBllociusingPCWSSOPtypingmethods.
ColombianAfricanAmericanpopulationsstudiedincludetheCauca(Pacific),Choco(NorthPacificCoast),andtheProvidencia(Caribbeanisland)populations.
ColombiaanalyzedattheHLAclassIIDR-DQloci.
HLAClassIItypingGenomicDNAwasamp1ifiedusingthepolymerasechainaction(PCR)(23,181andtypedusingn&-radioactivesequenceandallele-specificoligonucle-otideprobemethodsforDRB1,DQB1,DQAlandDpB1(19-25;andpersonalcomunication~~fromthreegeographicallydistinctregionsofCo-lombia(Figure1).
TheancestorsoftheColombianAfricanpopulationsanalyzedinthisstudyarrivedviaslaveshipsfromtheWestCoastofAfrica,broughtbytheSpaniardsfromtheGuinea,SenegalandMaliareasinparticular(lo),duringthelast-500years.
TheanalysisofHLAclass11polymor-phismmaybehelpfulindeterminingtheoriginsoftheAfricanAmericanpopulations.
TheAfricanCo-lombiangroupsanalyzedherehavelivedinclosecontactwithparticularAmerindiangroupsformanygenerations,andanalysisoftheHLAallelesandhaplotypesofthesepopulationscanservetoidentifyanyAfricanadmixtureintheColombianIndians;conversely,theAmerindianHLAdatamayalsoservetoidentifyAmerindianadmixtureintheAfri-canAmericanpopulations.
Recentmitochondria1DNAdataisconsistentwithanAfricanoriginofmodemhumanswithAfricansdispersingthrough-SequencingthenovelDQBl*O2alleleUnusualDQBlprobehybridizationpatternsfortwoindividualswithintheCaucagroupofAfricanAmericanssignalledthepresenceofapotentialnovelDQB1*02allele.
GenomicDNAwasampli-fiedforDQB1,clonedintoeitherM13mp18(26)orpCRScript(Stratagene)andsixcloneswerese-quencedusingABIPRISM6dyeterminatorcyclesequencingforeachindividual.
ResultsanddiscussionDRBlTheanalysisofthesethreeColombianAfricanAmericanpop.
ulations(Providencia,Cauca,andChoco)revealedagreatdiversityofDRBlalleleswith28allelesintotal,containingallelesfromeachserologicalclass(Table1).
Inspiteoftherelativelysmallnumberofsamplesanalyzed,19DRB1alleleswerefoundintheProvidencia(30samplesma-193Trachtenbergetal.
Table1.
DRB1inAfricanColombianpopulationsDR4DR5DR6Tabla3.
DBO1inAfricanColombianpopulations040104030405040604071101110211031104120113011302130314011402ProvidenciaCaucaChocoProvidenciaCaucaChoco0.
070.
020.
030.
020.
180.
150.
070.
08[0.
02][0.
03]0.
020.
030.
020.
070.
030.
080.
030.
03[0.
02]608,9,100.
050.
15[0.
075]0.
100.
2250.
0250.
075[0.
05]0.
0250.
0250.
0250.
0250.
05[0.
05]0.
0250.
0254000.
0250.
100.
13[0.
025]0.
10[0.
05]0.
0250.
0250.
050.
100.
220.
080.
050.
025400[1Amerindianadmixturepossible.
Table2.
DQAlinAfricanColombianpopulations~~ProvidenciaCaucaChoco010101020103020103-04010501N0.
270.
280.
030.
030.
130.
080.
18600.
0750.
2750.
0250.
050.
1750.
1750.
22540.
000.
150.
200.
0250.
20.
1250.
10.
240.
00lyzed),16intheCauca(20samplesanalyzed),and14intheChoco(20samplesanalyzed).
IntheseAf-ricanColombianpopulations,oneortwoDRB1al-lelesareatrelativelyhigherfrequenciesthantheothers,forexampleDRB1*1503(f=O.
18)andDRBl"0301(f=O.
15)intheProvidenciagroup,DRB1*0302(f=0.
225)andDRBI*1503(f=O.
15)intheCauca,andDRB1*0701(f=0.
225)andDRB1*1503(f=O.
13)intheChoco.
TheremainingDRB1allelesarefoundatrelativelyequalandlow1940201New0203030103020402050105020503060206030604N0.
130.
140.
080.
050.
30.
030.
030.
20.
0360-0.
20.
050.
050.
0750.
250.
0750.
1250.
150.
02540~~0.
330.
200.
050.
050.
180.
150.
020.
0240frequencies.
(f=-0.
025T0.
05)ineachgroup.
ThethreeAfricanColombianpopulationseachhaveex-amplesofpossibleAmerindianadmixture,basedonthedetectionofcharacteristicAmerindianalleles.
(Itshouldbenotedthat,giventhegreatdiversityofallelesinAfrica,itismoredifficulttodemonstrateAmerindianadmixtureintheseAfricanpopulationsthatthereverse.
)IntheCaucapopulation,thechar-acteristicAmerindianallelesfoundincludedtheDRBlalleles*1602(f=0.
075),*0802(f=0.
05)and*0407(f=0.
05).
TheseallelesareallabsentorrareamongAfricans(4).
TheDRB1*1602inAfricansiscoupledtoDQB1*05,whileinAmerindiansandintheChoco,itisassociatedwithDQB1*0301.
TheseDRBlalleleswerefoundintheneighboringAmer-indianpopulation,theWaunana,atrelativelyhigh(DRB1*1602;f=0.
33;*0407;f=O.
l5)tolow(DRB1*0802;f=0.
03)frequencies.
IntheChocopopulation,whichneighborstheTuletribe,theAm-erindianDRB1allelesidentifiedincludeDRB1*1602(f=0.
025)andDRB1*0407(f=0.
05);theseallelesarealsofoundatmoderatefrequencies(DRB1*1602;f=O.
16,DRB1*0407;f=O.
12)intheTulepopulation.
Althoughwefoundverylowfre-quenciesofinferredCaucasianadmixtureintheTuleandWaunana,noAfricanColombianadmix-turewasidentifiedinthesetwoAmerindianpopula-tionsbutwasfoundinothersatlowfrequencies.
DQAlandDQBlThedistributionofDQAlandDQB1intheColom-bianAfricanAmericanpopulationsstudiedarelistedinTables2and3,respectively.
IntheColorn-bianAfricanAmericanpopulations,theDQAlal-lelicfrequenciesareevenlydistributedwith*0101,*0102,"0103,"0201,*0301,*0401and0501alle-lesinallthreegroups.
ThedistributionoftheDQB1allelesintheAfricanColombiansampleswascon-sistent,givenDR-DQlinkagedisequilibriumpat-HLAclassIIallelesinAfricanColombiansEEYVRFDSDVCVYRAVTPQGRPVAEYWNSQKCVL15405060GAGGAGTACOTGCGCWB1'0501GACACCGACGlCGGGOn:TACCV:GCAGTGAffiCCCCAGGGGCCCCCTCTTGCCGAGTAC"52MCACCCAGAAGGAAGTCnCECARASVDRVCRHNYEVAYRGILQRRVEPTVTISCACCGCGCCCCGGCGTCGOn:GACKCGTGn;CAGACACMCTACGAGGTGGCGTACCGCaX;ATCCn:CAGAGGAGAOn:GAGCCCACAtX-2ACCAXTCC708090100CQBl*05011.
GenebankAccessionNumberDQBOZ433329Figure2.
ThenovelDQB1*0203sequenceincomparisonwithDQB1*0501andDQB1*0201.
CAWX254Haplotype:IllDP~1'0102-WA1~0501-~B1~0201CR91~0101-~A1'0401-~B1*~4~2Or(21DR91'0301-~1'OS0l-~B1~0201CRB1'0302-DQA1'0401-~Bl~O4!
J2CLUCA288Haplocy~:IllORa:'0302-DaA1.
0501-481.
020)CR~l'1104-~A1'0102-DPBl'050~Figure3.
PossibleDR-DQhaplotypesofindividualswiththenovelDQB1*0203allele.
terns(4,27),withtheDRBldistributionforthethreegroups(Table3).
DQB1*0203:anovelDQB1*02alleleinanAfricanAmericanpopulationAnovelDQB1*02allele(DQB1*0203;GenebankAccession#DQBO2433329)'wasdiscoveredintwoindividualsfromtheCaucapopulation,anAfri-canColombiangrouplivingonthePacificcoast.
UnusualDQB1probehybridizationpatternsforthesetwoindividualssignalledthepresenceofapo-tentiallynovelDQBl*02allele.
Thenewallelewasrecognizedbecausetheprobefortheaminoacidse-quenceLGLPA(DB53:TGCTGGGGCTGCCT-GCC)(22)atposition52-57inexon2oftheDQBllocus,didnothybridizeasexpectedtoDQB1*0201,butallother*0201-specificprobeshybridizedtothesamples.
TheseresultsfurtherdemonstrateboththepowerofPCNSSOPtypingsystemstoidentifynovelallelesandtheplasticityofthesystemforpo-tentialupdatingbythesimpleadditionofnewprobes.
ThenewDQB1*0203allelebringsthetotalnumberofDQB1allelesto26,includingsevenalle-leswithsilentpolymorphisms(28).
'ThenameDQB1*0203hasbeenofficiallyassignedbytheWHONomenclatureCommitteeinNovember1995.
DNAsequencingofthesampleswiththeunusualprobereactivitypatternrevealedthatthesecondexonofthenewDQB1*0203isidenticaltoDQB1*0201fromcodonposition10through80,exceptatcodonposition57,whereasecondposi-tiontransversion(C+A)createsanaminoacidchangefromalanine(GCC)(neutralcharge)toas-particacid(GAC)(negativecharge)(Figure2).
ThenewDQB1*0203allelewasfoundcoupledtoDQA1*0501inthetwounrelatedCauca;DQA1*0501-DQB1*0201istheDQhaplotypeusu-allyassociatedwithDRB1*0301.
ThepossibleDR-DQhaplotypesoftheindividualsfromtheCaucapopulationcarryingthenovelDQB1*0203alleleareillustratedinFigure3.
ThenovelallelemaybeonthehaplotypeDRB1*0302-DQA1*0501-DQB1*0203;thishaplotypeprobablyarosefromtherecombinationofDQAl*O501-DQB1*02-,whichismorecommonlyfoundcoupledtoDRB1*0301thanwithDRB1*0302.
Asecondpos-sibilityisthatthenovelalleleisontwodifferentDRB1*03haplotypes,includingDRB1*0302-DQA1*0501-DQB1*0203,aswellasDRB1*0301-DQA1*050l-DQB1*0203haplotypes(Figure3).
ThehaplotypeassignmentsmadeforthesampleCauca-288includetheknownhaplotypeDRBl*1104-DQA1*0102-DQB1*0502,andthenovelDRB1*0302-DQA1*0501-DQB1*0203hap-lotype.
ThesampleCauca-254couldhavethehap-lotypesDRB1*0301-DQAl*0501-DQB1*0203andDRB1*0302-DQA1*0401-DQB1*0402,acom-monDRB1*0302haplotype.
Inthiscase,themorecommonDRB1*0301-DQA1*0501-DQB1*02-haplotypewasperhapstheancestralhaplotypeonwhichthenovelDQB1*0203allelewasgeneratedbypointmutationgrageneconversion-likeevent,andarecombinationeventwithDRB1*0302togive195Trachtenbergetal.
Table4.
OPE1inAfricanColombianpopulationsProvidenciaCausaChoco010102010301040104020901100111011301140115011601170118012301270130014001N0.
170.
10.
020.
150.
20.
030.
030.
030.
050.
080.
080.
020.
020.
02600.
4250.
0250.
0750.
050.
1750.
080.
050.
050.
0250.
05400.
50.
0250.
0250.
10.
2750.
0250.
0250.
02540thehaplotypefoundinsample288.
IftheDQB1*0203alleleisontwodifferenthaplotypes,thiswouldsuggestthat*0203wasnotgeneratedveryrecently.
Thegenerationof*0203from*0201couldeitherbetheresultofapointmutation(C+Aatthesecondpositionofcodon57)oragenecon-version-likeevent.
IntheCaucapopulation,apoten-tialdonoralleleforageneconversion-likeevent,couldbeDQB1*0402,whichhastheGACcodon(asparticacid)atposition57.
DQB1*0402,themostcommonDQB1allele(f=0.
25)intheCaucapopula-tion,isassociatedwithDRB1*0302.
ThenewDQB1*0203allelebringsthenumberofsetsofDQBlchainsthatdifferonlyatposition57tofive,includingDQB1*0201(Ala)vs.
*0203(Asp),DQB1*0301(Asp)vs.
*0304(Ala),DQB1*0302(Ala)vs.
*03032(Asp),andDQB1*0501(Val)vs.
*0502(Ser),vs.
*0503(Asp).
Atposition57oftheDQBP-chain,onlyAspandthreeotherunchargedresidues,Ala,ValorSer,areobservedinallprimates(29).
TheAspresidueatpo-sition57isbelievedtoparticipateinasaltbridgewithaninvariantArgresidueatposition80ofthea-chain(30)andhasbeenimplicatedindiseasesus-ceptibility(31-33).
ItislikelythatmoleculeswiththenegativelychargedresidueAsp-57maybestruc-turallyandfunctionallydifferentfromoneswiththeunchargedresidues,andmaythereforebindandpresentadifferentrepertoireofpeptides.
Selectivepressuresuchasoverdominancemayberesponsibleforthemaintenanceoftheseposition57variants.
BalancingselectionforthepresenceofbothAsp-57andAla,ValorSer-57allelesmayberesponsibleforthepatchworkpatternofpolymorphismatthisposi-tion.
Accumulatingdataonthedistributionandlink-ageofthenewDQB1*0203allelewillhelptore-solveremainingquestionsastotheevolutionofthisalleleanditsrelatedhaplotypes.
DPBlTheDPBlallelicdistributionsinthethreeColom-bianAfricanAmericanpopulationsarelistedinTa-ble4.
Inthesepopulations,18DPBlalleleswerepresent,theprevalentallelesbeingDPB1*O101,"0201,*0301,*0401and*0402inallthreegroups.
IntheCaucaandChocopopulations,DPB1*0101predominates(f=0.
425andf=0.
5,respectively),withDPB1*0402thesecondmostcommonalleleinbothgroups(f=0.
175andf=0.
275respectively).
IntheProvidenciapopulation,DPB1*0402isthemostprevalentallele(f=0.
20).
TwootherDPBlalleles,however,arefoundintheProvidenciawithfrequen-ciesclosetothatfoundforDPB1*0402,includingDPB1*0101(f=0.
15)andDPB1*0401(f=0.
15),withDPB1*0201(f=O.
l),DPB1*1701(f=0.
08),and"1801(f=0.
08)foundatslightlylesscommonfrequencies.
ThedistributionofDPBlallelesintheProvidenciaisunusualinthisregard,asgenerallyoneortwoDPBlallelesarefoundathigherfre-quenciesinapopulation(4,27).
Differencesbe-tweenthedistributionsofDPB1allelesintheProvi-dencia,ChocoandCaucapopulationsmayreflecttheiroriginalAfricanethnicitiesandmaybehelpfulindeterminingtheirpreciseAfricangeographicori-gins.
SummaryAnalysisofHLAclassI1alleles(n=140)amongthreeAfricanColombianpopulationsfromgeo-graphicallydistinctregionsofColombiarevealsabroadassortmentofalleles.
Incontrasttothelim-itedHLAclassI1allelicdistributionfoundintheColombianAmerindians,describedinanaccompa-nyingpaper,theHLAclassI1allelesoftheseneigh-boringAfricanColombianpopulationsshowmuchgreaterdiversity,includinganovelDQB1*0203al-leleintheCaucapopulation.
ThenovelDQBlallelemayhaveevolvedfromtheancestralDQB1*0201allelebyapointmutationorfromaninterallelicgeneconversion-likeeventatposition57withDQB1*0402,whichalsohasAspatposition57andisthemostcommonDQBlalleleintheCaucapop-ulation.
Mechanismsofbalancingselection,suchasoverdominance,mayplayaroleinmaintainingthisnewvariantinthepopulation,asposition57resi-duesarethoughttoplayafunctionallysignificantroleinpeptidebindingandpresentation.
TheobserveddiversityofHLAclass11allelesin196HLAclassI1allelesinAfricanColombiansTiwariJL,TerasakiPI.
HLAanddiseaseassociations.
NewYork,Springer,1985.
HillAVS.
AllsoppCEM,KwiatkowskiDetal.
CommonwestAfricanHLAantigensareassociatedwithprotectionfromseveremalaria.
Nature1991:352:595-600.
NepomG,ErIichHE.
MHCclassIImoleculesandautoim-munity.
AnnuRevImmunol1991:9:493-525.
GutierrezI.
HistoriadelNegroenColombia.
Bogota:NuevaAmericah,1986:17-23.
StonekingM.
DNAandrecenthumanevolution.
Evolution-aryAnthropoll993:60-73.
GreenbergJH,TurnerCI,ZeguraSL.
ThesettlementoftheAmericas:acomparisonofthelinguistic,dentalandgeneticevidence.
CumAnthropol1986:27:477498.
SalzanoF,Callegari-JacquesS.
SouthAmericanIndians:acasestudyinevolution.
NewYork,OxfordUniversityPress,1988.
KiddJ,BlackF,WeissKM,BalmsJ,KiddKK.
StudiesofthreeAmerindianpopulationsusingnuclearDNApolymor-phism~.
HumBiol1991:63:775-794.
GibbonsA.
GeneticiststracethetrailofthefirstAmericans.
Science1993:259:312-313.
StathmaryEIE.
mtDNAandthepeoplingoftheAmericas.
AmJHumGenet1993:53:793-799.
KeyeuxG,LefrancM-P.
ChevaillerA,LefrancG.
Molecu-laranalysisoftheIGHandMHCclass111regiongenesinonefamilywithIgAandC4deficiencies.
EipClinImmu-nogenet19907170-180.
SaikiRK,GelfandDH,StoffelSetal.
Primer-directeden-zymaticamplificationofDNAwithathermostableDNApolymerase.
Science1988:239:487491.
ScharfSJ,GriffithRL,ErlichHA.
RapidtypingofDNAse-quencepolymorphismattheHLA-DRBIlocususingthepolymerasechainreactionandnon-radioactiveoligonucle-otideprobes.
HumImmunoll991:30:190-20I.
SaikiR,WalshPS,LevensonCH.
ErlichHA.
GeneticAnal-ysisofamplifiedDNAwithimmobilizedsequence-specificoligonucleotideprobes.
ProcNatlAcadSciUSA1989:86:6230-6234.
BugawanTL,ErlichHA.
RapidtypingofHLA-DQBIDNApolymorphismusingnon-radioactiveoligonucleotideprobesandamplifiedDNA.
Immunogenetics1991:33:Rtus-TrachtenbergEA,BugawanTL,ErlichHA.
Identifi-:ationofanovelDQBlallele(*0609)segregatinginanAshkenaziJewishfamily:ImplicationsforDQB1typingsystems.
TissueAntigens1994:44:120-124.
BugawanTL,BegovichAB,ErlichHA.
RapidHLA-DPBjpingusingenzymaticallyamplifiedDNAandnonradioac-ivesequencespecificoligonucleotideprobes.
Immimoge-ietics1990:32:231-241.
MoonsamyPV,SurajVC,BugawanTL.
GeneticdiversityMithintheHLAclassIIregion:tennewDPBlallelesandheirpopulationdistribution.
TissueAntigens1992:40:153-157.
HoonsamyPV,AldrichCL,PetersdorfEW,HillEVS,Be-covichAB.
SevennewDPBlallelesandtheirpopulationiistribution.
TissueAntigens1994:43:249-252.
3charfSJ.
HornGT,ErlichHA.
Directcloningandse-penceanalysisofenzymaticallyamplifiedgenomicse-pences.
Science1986:223:1076-1078.
3egovichAB,McClureGR,SurajVCetal.
Polymorphism,ecombinationandlinkagedisequilibriumwithintheHLA:lass11region.
JImmiinol1992:148:249-258.
3odmerJG,MarshSGEetal.
NomenclatureforfactorsofheHLAsystem,1995.
Humlmrnunol1995:41:149-164.
irlichHA,GyllenstenU.
ShmdepitopesamongHLAClass163-170.
AfricanColombiansissimilartotheextentofdiver-sityobservedfortheimmunoglobulinheavychain(IGH)"switch-alpha"regionofthegenomeinthesepopulations(34,35).
TheHLAdistributionsfortheColombianpopulationsstudiedarealsoconsistentwithmitochondria1DNAevidencethatAfricanAmericansarethemostgeneticallydiversegroupandhavetheoldesthistoryofmodemhumans(re-viewedin11).
Thecomprehensiveandevendistri-butionofclassI1allelesintheColombianAfricanAmericansisinmarkedcontrasttothedistributionofallelesobservedinneighboringAmerindiangroups,whichhavemuchmorerestrictedHLAdi-versity(seeaccompanyingpaper).
Inthisstudy,verylittlegeneticadmixturewasfoundbetweentheAfricanAmericanandAmerindiangroups.
UsinggenetictagssuchasHLAallelicdistributionforeachAfricanAmericanpopulation,itmaysoonbepossibletodeterminetheAfricanoriginsofthesepeopleswhowereoriginallyintroducedtoColom-biaasslavesduringtheSpanishcolonizationinthe16thcentury.
AcknowledgmentsTheauthorswouldliketothankthepeoplesofCo-lombiaforparticipatingintheExpedicionHumanaandC.
RodasforherassistanceinsamplecollectionandDNAextractions.
ManythankstoFredReichertandAgnesCavalliforsequencingandtoKathyLev-ensonandJeraWodehouseformanuscriptprepara-tion.
ThisworkwassupportedinpartbyagrantfromtheNIHtoHAE(#HL47170-02),andbytheExpedicionHumana,InstitutodeGeneticaHumana,PontificiaUniversidadJaveriana,Bogota,Colombia.
jReferencesI.
SaikiRK,ScharfS,FaloonaF,MullisK,HornG.
Erlich2.
3.
4.
5.
6.
HA,ArnheimN.
EnzymaticamplificationofP-globinge-nomicsequencesandrestrictionsiteanalysisfordiagnosisofsicklecellanemia.
Science1985:230:1350-1354.
SaikiRK,BugawanTL,HornGT,MullisKB,ErlichHA.
AnalysisofenzymaticallyamplifiedP-globinandHLA-DQaDNAwithallele-specificoligonucleotideprobes.
Na-ture1986:324:163-166.
MullisKB,FaloonaF.
SpecificsynthesisofDNAinvitroviaapolymerasecatalyzedchainreaction.
MethodsEnzy-moll987:55:335-350.
ImanishiT,TatsuyaA,KimuraA,TokunagaK,GojoboriT.
AlleleandhaplotypefrequenciesfortheHLAandcomple-mentlociinvariousethnicgroups.
TsujiK.
AizawaM,SasazukiMed.
HLA1991.
Oxford:OxfordUniversityPress,1992:1:1065-1220.
ParkMS,TerasakiPI,BarbettiA,HanH.
CeckaJM.
Sig-nificanceoftheHLAmolecularstructuretotransplanta-tion.
ClinTransplant1988:301-308.
ComeyC.
BudowleB.
ValidationstudiesontheanalysisoftheHLA-DQalphalocususingthepolyernsechainreac-tion.
JForensicSci1991:36:1633-1648.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
I1:11III1t1197Trachtenbergetal.
I1alleles:geneconversion,commonancestry,andbalancingselection.
ImmunolToday1991:No.
11:411-414.
30.
BrownJH,JardetzkyTS,GorgaJCetal.
Three-dimen-sionalstructureofthehumanclassI1histocompatibilityan-tigenHLA-DR1.
Nature1993:364:33-39.
31.
ToddJA,BellJI,McDevittHO.
HLA-DQgenecontributetosusceptibilityandresistancetoinsulindependentdiabe-tesmellitus.
Nature1987:329:599-604.
32.
HornGT,BugawanTL,LongC,ErlichHA.
Allelicse-quencevariationoftheHLA-DQloci:relationshiptoserol-ogyandinsulin-dependentdiabetessusceptibility.
ProcNatlAcadSciUSA1988:85:6012-6016.
33.
ScharfSJ,FriedmannA,BrautbarCetal.
HLAclassJJal-lelicvariationandsusceptibilitytoPernphigusvulgaris.
ProcNatlAcadSciUSA1988:85:3504-3508.
34.
KeyeuxG,BernalJE.
Newallelevariantsoftheimmuno-globulinswitch(SIX"alpha")regions.
HumGenet1996:97:695-696.
35.
KeyeuxG,RodasMC,BautistaS,CarvajalE,BernalJE.
TracingAfricanadmixtureinColombia.
Immunoglobulinswitchalpharegionallelesinvariouspopulations.
ResultsofExpedicionHumana.
Submitted.
Address:E.
A.
TrachtenbergChildren'sHospitalOaklandResearchInstitute74752ndStreetOakland,CA94609USA198