Evaluation of Group Genetic Ancestry of Populations From Philadelphia and Dakar in the Context

12 Pages·2016·0.74 MB·English

UUnniivveerrssiittyy ooff PPeennnnssyyllvvaanniiaa
SScchhoollaarrllyyCCoommmmoonnss
Department of Anthropology Papers Department of Anthropology
11-25-2009
EEvvaalluuaattiioonn ooff GGrroouupp GGeenneettiicc AAnncceessttrryy ooff PPooppuullaattiioonnss FFrroomm
PPhhiillaaddeellpphhiiaa aanndd DDaakkaarr iinn tthhee CCoonntteexxtt ooff SSeexx--BBiiaasseedd AAddmmiixxttuurree iinn
tthhee AAmmeerriiccaass
Klara Stefflova
University of Pennsylvania
Matthew C. Dulik
University of Pennsylvania, [email protected]
Athma A. Pai
University of Pennsylvania
Amy H. Walker
University of Pennsylvania
Charnita M. Zeigler-Johnson
University of Pennsylvania
See next page for additional authors
Follow this and additional works at: https://repository.upenn.edu/anthro_papers
Part of the Biological and Physical Anthropology Commons, and the Genetics Commons
RReeccoommmmeennddeedd CCiittaattiioonn
Stefflova, K., Dulik, M. C., Pai, A. A., Walker, A. H., Zeigler-Johnson, C. M., Gueye, S. M., Schurr, T. G., &
Rebbeck, T. R. (2009). Evaluation of Group Genetic Ancestry of Populations From Philadelphia and Dakar
in the Context of Sex-Biased Admixture in the Americas. PLoS ONE, 4 (11), 1-10. https://doi.org/10.1371/
journal.pone.0007842
This paper is posted at ScholarlyCommons. https://repository.upenn.edu/anthro_papers/23
For more information, please contact [email protected].
EEvvaalluuaattiioonn ooff GGrroouupp GGeenneettiicc AAnncceessttrryy ooff PPooppuullaattiioonnss FFrroomm PPhhiillaaddeellpphhiiaa aanndd
DDaakkaarr iinn tthhee CCoonntteexxtt ooff SSeexx--BBiiaasseedd AAddmmiixxttuurree iinn tthhee AAmmeerriiccaass
AAbbssttrraacctt
Background
Population history can be reflected in group genetic ancestry, where genomic variation captured by the
mitochondrial DNA (mtDNA) and non-recombining portion of the Y chromosome (NRY) can separate
female- and male-specific admixture processes. Genetic ancestry may influence genetic association
studies due to differences in individual admixture within recently admixed populations like African
Americans.
Principal Findings
We evaluated the genetic ancestry of Senegalese as well as European Americans and African Americans
from Philadelphia. Senegalese mtDNA consisted of ~12% U haplotypes (U6 and U5b1b haplotypes,
common in North Africa) while the NRY haplotypes belonged solely to haplogroup E. In Philadelphia, we
observed varying degrees of admixture. While African Americans have 9–10% mtDNAs and ~31% NRYs
of European origin, these results are not mirrored in the mtDNA/NRY pools of European Americans: they
have less than 7% mtDNAs and less than 2% NRYs from non-European sources. Additionally, there is <2%
Native American contribution to Philadelphian African American ancestry and the admixture from
combined mtDNA/NRY estimates is consistent with the admixture derived from autosomal genetic data.
To further dissect these estimates, we have analyzed our samples in the context of different demographic
groups in the Americas.
Conclusions
We found that sex-biased admixture in African-derived populations is present throughout the Americas,
with continual influence of European males, while Native American females contribute mainly to
populations of the Caribbean and South America. The high non-European female contribution to the pool
of European-derived populations is consistently characteristic of Iberian colonization. These data suggest
that genomic data correlate well with historical records of colonization in the Americas.
DDiisscciipplliinneess
Biological and Physical Anthropology | Genetics
AAuutthhoorr((ss))
Klara Stefflova, Matthew C. Dulik, Athma A. Pai, Amy H. Walker, Charnita M. Zeigler-Johnson, Serigne M.
Gueye, Theodore G. Schurr, and Timothy R. Rebbeck
This journal article is available at ScholarlyCommons: https://repository.upenn.edu/anthro_papers/23
Evaluation of Group Genetic Ancestry of Populations
from Philadelphia and Dakar in the Context of
Sex-Biased Admixture in the Americas
Klara Stefflova1*, Matthew C. Dulik2, Athma A. Pai2, Amy H. Walker1, Charnita M. Zeigler-Johnson1,
Serigne M. Gueye3, Theodore G. Schurr2, Timothy R. Rebbeck1,4
1DepartmentofBiostatisticsandEpidemiology,SchoolofMedicine,UniversityofPennsylvania,Philadelphia,Pennsylvania,UnitedStatesofAmerica,2Departmentof
Anthropology,UniversityofPennsylvania,Philadelphia,Pennsylvania,UnitedStatesofAmerica,3HoˆpitalGe´ne´raldeGrandYoffandUniversite´CheikhAntaDiop,Dakar,
Senegal,4AbramsonCancerCenter,SchoolofMedicine,UniversityofPennsylvania,Philadelphia,Pennsylvania,UnitedStatesofAmerica
Abstract
Background:Population history can be reflected in group genetic ancestry, where genomic variation captured by the
mitochondrial DNA (mtDNA) and non-recombining portion of the Y chromosome (NRY) can separate female- and male-
specific admixture processes. Genetic ancestry may influence genetic association studies due to differences in individual
admixture withinrecently admixedpopulations likeAfrican Americans.
PrincipalFindings:WeevaluatedthegeneticancestryofSenegaleseaswellasEuropeanAmericansandAfricanAmericans
from Philadelphia. Senegalese mtDNA consisted of ,12% U haplotypes (U6 and U5b1b haplotypes, common in North
Africa) while the NRY haplotypes belonged solely to haplogroup E. In Philadelphia, we observed varying degrees of
admixture.WhileAfricanAmericanshave9–10%mtDNAsand,31%NRYsofEuropeanorigin,theseresultsarenotmirrored
in the mtDNA/NRY pools of European Americans: they have less than 7% mtDNAs and less than 2% NRYs from non-
Europeansources.Additionally,thereis,2%NativeAmericancontributiontoPhiladelphianAfricanAmericanancestryand
theadmixturefromcombinedmtDNA/NRYestimatesisconsistentwiththeadmixturederivedfromautosomalgeneticdata.
To further dissect these estimates, we have analyzed our samples in the context of different demographic groups in the
Americas.
Conclusions:Wefoundthatsex-biasedadmixtureinAfrican-derivedpopulationsispresentthroughouttheAmericas,with
continualinfluenceofEuropeanmales,whileNativeAmericanfemalescontributemainlytopopulationsoftheCaribbean
and South America. The high non-European female contribution to the pool of European-derived populations is
consistently characteristic of Iberian colonization. These data suggest that genomic data correlate well with historical
records ofcolonizationin theAmericas.
Citation:StefflovaK,DulikMC,PaiAA,WalkerAH,Zeigler-JohnsonCM,etal.(2009)EvaluationofGroupGeneticAncestryofPopulationsfromPhiladelphiaand
DakarintheContextofSex-BiasedAdmixtureintheAmericas.PLoSONE4(11):e7842.doi:10.1371/journal.pone.0007842
Editor:JohnRelethford,StateUniversityofNewYorkCollegeatOneonta,UnitedStatesofAmerica
ReceivedJuly9,2009;AcceptedOctober15,2009;PublishedNovember25,2009
Copyright:(cid:1)2009Stefflovaetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits
unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited.
Funding:ThisstudywassupportedbygrantsfromthePublicHealthService(R29-ES08031,R01-CA85074,andP50-CA105641)toTRRandFacultyResearchFunds
fromPenntoTGS.Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofthemanuscript.
CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist.
*E-mail:[email protected]
Introduction each group, yet fail to predict the extent of the contribution from
each parental populationtodifferent SIREgroups [1].
Populationsofthepresent-dayAmericaswereshapedbydiverse One example where SIRE may inadequately represent this
incoming groups and their intermixing. Although the number of diverse contribution from parental populations is population
the Native Americans was greatly reduced due to conflict and stratification bias in molecular epidemiology case-control associa-
disease they, together with the early arriving Europeans and tionstudies[2].Whileancestryinformationinmolecularepidemi-
surviving Africans brought to the Americas during the massive ologyisusuallystudiedintermsofindividualancestry,maternaland
AfricanDiaspora,alllefttheirgeneticimprintinmultipleadmixed paternalgroupancestrycanprovideinformationabouttheancestry
populations. Later, several other immigrant groups from around of populations without making inferences about individuals,
the World (e.g. Asian populations) and increasingly common accountingforallsignificantlycontributingpopulations.Uniparen-
admixture among the existing groups further amplified the tally inherited mitochondrial DNA (mtDNA) and Y-chromosome
admixed character of this continent. This complicated ancestry (NRY) behave as uninterrupted single loci that are often used in
and admixture is reflected in an individual’s genetic background. estimatinggroupancestryandpredictinggender-specificpopulation
While recognizing that each individual is genetically unique, it is demographicprocesses[3,4].Together,theycanaccuratelyreflect
stillcommoninepidemiologytocategorizepeopleintoafewself- the average autosomal group ancestry [5]. Individually, they can
identifiedraces(SIRE)thatpartlyreflectthecomplicatedhistoryof helptoseparategender-biasedadmixtureprocesses[6,7].
PLoSONE | www.plosone.org 1 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
Because they have a higher incidence of several common theQiagen9604BrobotwiththeQIAamp96DNABuccalSwab
diseases [8] and a complicated history, African Americans have BiorobotKit(Valencia,CA).Priortotyping,thewholegenomeof
been studied in a variety of epidemiological and population these samples was amplified using the GenomePlex Complete
geneticsettings.AsaSIREgroup,theyrepresentthedescendants Whole Genome Amplification kit (Sigma,St.Louis, MO).
of Africans brought mainly during the African Diaspora from
W/WC/SW/SE Africa, admixed with Europeans and possibly mtDNA, NRY, and AIMs Typing
NativeAmericans.Still,everyregionalgroupofAfricanAmericans Thefirstandsecondhypervariablesegments(HVSIandHVSII)
may have been drawn from different African sources or have a ofmtDNAwereamplifiedandbothstrandssequencedbetweenbp
unique history that will influence the extent and pattern of 16,030–16,490 and bp 50–710 (Table S1) using a BigDyeTM
admixture and make them a unique group. Similar regionally- Terminatorv3.1(AppliedBiosystems)afterpurifyingwithExoSAP-
specific admixture most likely influences other groups in the IT(USB,Cleveland,OH).AfterpurificationwiththeQIAquick96
Americas. PCR Purification kit (Qiagen), the sequences were read using an
Based on reports of low resolution maternal and paternal ABI Prism 3130xl Genetic Analyzer (Applied Biosystems) and
ancestry, mtDNA and NRY reveal a sex-biased gene flow from analyzedusingSequencherv4.7software(GeneCodesCorp.,Ann
European males to US/Jamaican African Americans, but the Arbor, MI). Each sample was then hierarchically typed for
extent of group maternal and paternal European admixture mutationsinthemtDNAcodingregionusingRestrictionFragment
greatly varies [5,9] (e.g. 0–15% for mtDNA and 8.6–46.9% for Length Polymorphism (RFLP) assays to correctly assign each of
NRY in 9 US populations in Parra et al.). This sex-biased themintoaparticularhaplogroup(TableS2).Aphylogenetictree
admixture was also reported in African-descended populations in (Figure1)wasdrawnmanually,basedonthemedian-joiningtree
Uruguay [10] and both White and African Brazilians [11–13], constructed using Network 4.5.0.2 [17], listing the diagnostic
whileastudydealingwiththeFBImtDNAdatabaseshowslimited mutations RFLP-typed in coding and present in sequenced
gene flow of non-Europeans to the pool of US European hypervariableregions.
Americans[14].DetailedassessmentofAfrican,NativeAmerican, ForNRY,thesamplesweretypedusingpre-designedTaqMan
and European admixture in both European- and African- assays in combination with multiplex fragment analysis and
descended groups in different populations of the Americas may RFLP (Figure 2). The SNPs are listed in Table S3 (see also
improveourunderstandingofvariationwithinandbetweeneach References S1) together with the haplogroup designation
SIRE group from different regions. This combined analysis can established by the Y-Chromosome Consortium in 2002 [18] and
help to approximate the parental populations for any unchar- revisedinKarafetetal.2008[19].BothmtDNA,NRYvariation
acterizedgroupineachregionduringearlyepidemiologicalstudy andethnicity informationislistedin FileS1.
design and aid in understanding if African- and European- For a small subset of the reported samples (31 African
descended American SIRE means the same in different parts of Americans, and 6 European Americans), we also estimated the
theAmericas. autosomalancestrybygenotypingthesamplesonacommercially
Therefore, we estimated maternal and paternal continental available Illumina Golden Gate 1509 AIMs chip. The resulting
admixture proportions and report on the group ancestry of three genotypes were combined with available genotypes from an
populations: admixed Philadelphian African Americans and Illuminaadmixturepanel(YOR,CEU,JPT+CHB)thatrepresent
European Americans, as well as a control sample of Senegalese the ancestral African, European and Asian populations. The
fromDakarasoneofthepossiblesourcepopulations.Further,we corresponding admixture proportions were estimated using the
have mined the published literature for raw mtDNA/NRY data program STRUCTURE running 10,000/50,000 burn-in/repeti-
(Brazil and Cuba) or ancestry estimates (Caribbean, Colombia, tions, assuming an admixture model with correlated allele
Uruguay) as well as census data in order to interpret our results frequencies, running K=1–5 and reporting estimates for K=3
withinthecontext oftheAmericas. founding populations, where the posterior probability ln P(D)
plateaus (TableS4).
Methods
Phylogenetic and Statistical Analysis
Sampling, DNA Handling WeusedArlequin3.11[20]toestimategeneticdistancesbased
The Philadelphia samples consisted of 217 self-identified on Slatkin’s linearized F to construct multidimensional scaling
ST
African Americans and 204 self-identified European Americans. (MDS) plots to assess ‘‘between group’’, ‘‘within-population’’ and
Theseindividualswereascertainedbetween1995and2007aspart ‘‘between population within group’’ variation via the analysis of
of a prostate cancer case-control study, with cases identified molecularvariance(AMOVA)[21].Weincludedthephylogenetic
through Urologic Oncology Clinics at multiple hospitals of the relationshipofmtDNAhaplotypes/NRYhaplogroupsintheform
University of Pennsylvania Health System (UPHS) and controls of haplotypes (mtDNA) or distance matrix (NRY) and assumed
beingmenattendingUPHSgeneralmedicineclinics.Additionally, Tamura and Nei’s [22] model for nucleotide substitution for
49 subjects from Senegal (all cancer-free controls) were identified mtDNAsequences.TheMDSplotswereconstructedusingSPSS
andascertainedfromuniversityandhospitalpopulationsinDakar, with input data in the form of an Arlequin-generated matrix of
Senegal.AllstudysubjectsfromUSandSenegalprovidedwritten Slatkin’s linearized F distances [23]. For each MDS plot, we
ST
informed consentforparticipationinthisresearch.IRBapproval report the stress and RSQ statistics, which summarize the
for this study has been provided by the Committee on Studies goodness of fit of multidimensional data in two dimensions.
Involving Human Beings of the University of Pennsylvania Additionally, AMOVAwas reported fortheparental populations
(Protocol#3614-2)andbytheCommissionEthiqueetEvaluation (indicatedineachMDSplotinFigure3)showingthepercentage
attheHopitalGeneraldeGrandYoffinDakar(FWA00002772). of variationcaptured by definingthecontinental groups.
Genomic DNA was obtained from buccal swabs (Cyto-Pak
Cytosoft Brush, Medical Packaging Corporation, Camarillo, CA) Admixture Analysis
processed using either a protocol modified from Richards et al. AdmixturewasestimatedusingtheADMIX2.0[24]software.
[15]asdescribedpreviously[16]orusingamodifiedprotocolon We ran 50,000/100,000 (mtDNA/NRY) bootstrap simulations
PLoSONE | www.plosone.org 2 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
Figure1.MitochondrialDNAphylogenetictree.TreeofmtDNAhaplotypesbasedonmedianjoiningnetworkwithAfricanAmericancases
(yellow), African American controls (orange), Senegalese (white), European American cases (light green), and European American controls (dark
green).Nodesizesareproportionaltothesamplesizes,indicatedbynumberswithinthenode,withtheexceptionofhaplogroupsHandKlabeledby
numbers in red. Variable positions typed for these samples in coding, HVS I, and HVS II region are distinguished by red, black and grey font,
respectively.ThemaincontinentallocationisindicatedbythebackgroundcolorwithochreindicatingpredominantlyAfrican,greenWestEurasian,
greyAsian,andpinkAmerindianhaplogroups.TherawdatacanbefoundinFileS1.
doi:10.1371/journal.pone.0007842.g001
PLoSONE | www.plosone.org 3 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
Figure2.Ychromosomephylogenetictree.NRYtreehaplogroupsobservedinPhiladelphiaandSenegaldatawithtypedSNPsindicatedon
eachbranch.AssociatedwitheachbranchisthenumberofsamplesobservedforeachhaplogroupinthepoolofAfricanAmericans(AA,n=199),
European Americans (EA, n=190), and Senegalese (Af, n=33). We have omitted from the NRY tree M148 SNP designatingE1b1b1a3a (formerly
E3b1a).OneEAsamplebelongingtothishaplogroupwasaddedtoE1b1b1(#).OneEAsamplewas12f2a*(HgJ)butwasgroupedwithJ1(*)
becauseofspaceconstraints.TherawdatacanbefoundinFileS1.
doi:10.1371/journal.pone.0007842.g002
and report the estimates as a percent of contribution from a analyzed for admixture using ADMIX. The genetic variation
particular parental population, along with an estimate of the combinedwithphylogeneticdistanceswascapturedby9–18NRY
sampling error (SD). These calculations incorporate molecular haplogroups and 335 mtDNA haplotypes, defined by distance
divergence and haplotype frequencies, with both mtDNA and matrix (NRY) and sequences of HVS I/II and part of HVS III/
NRYbeingtreatedasasinglelocus.WeinitiallyexploredK=2–4 codingregion(mtDNA).
offoundingpopulationsandbothbasedontheSDoftheestimates
and previous reports, we only pursued K=3 where the Native Results
American andAsianfounding populations seem tobecombined.
ForthemtDNAanalysis,weusedpreviouslypublishedsetsdrawn First,wedeeplytypedbothmtDNAandNRYinPhiladelphian
from: West Africa (n=819, represented by Guinea Bissau [25], African Americans and European Americans. We also typed the
Senegal (our and published [26] data), and Sierra Leone [27]); samemarkersinagroupofSenegaleseinordertogaininsightinto
Europe [28] (n=3532); and the Americas [29] (n=58) as the thedetailedcompositionofoneofthefoundingpopulations(e.g.the
parental populations (note that the estimate represented by the presence of ‘‘Eurasian’’ haplotypes) instead of relying on less well
Americas will partially overlap with the possible admixture from characterized published data. We then proceeded to phylogenet-
SE Asia). For the NRY analysis, our parental populations were ically analyze these haplotypes in parallel withadmixture analysis
defined asfollows:West/West-Central Africa(n=834,represent- usingADMIX,exploringthepossiblefoundingpopulations.Based
ed by Guinea Bissau [30], Mali, Ghana, Benin, Senegambia [7] on these analyses, we focused further on analysis using three
(adding Senegalese reported in this publication) and Cameroon founding populations (K=3 was also corroborated by our
[7,31]); West Eurasia [32] (n=481, represented by Germany, STRUCTUREanalysisusingautosomalAIMs).
Denmark,Galicia,andTurkey);andtheAmericas[33](n=398).
ThemtDNAandNRYvariationinBrazilianadmixedpopulations Mitochondrial DNA (mtDNA)
wasfirstminedfromtheliterature:Afro-BrazilianmtDNA[34–36] We assessed the contribution of European, African and Native
(n=277) and NRY [12,34,35] (n=380) and White Brazilian American populations to the pool of Senegalese, Philadelphian
mtDNA [11,12] (n=247) and NRY [12] (n=180) and mtDNA/ AfricanAmericansandPhiladelphianEuropeanAmericansbytwo
NRY n=245/132 of the general population of Cuba [37]. The means:1) by admixture analysis usingADMIX [24](deriving the
combined datasets for each marker and group were subsequently possibleparental(ancestral)regionsfrompublishedsetscomprising
PLoSONE | www.plosone.org 4 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
Figure3.ContributionofEuropean,African,andNativeAmericanfemaleandmalelineagestothepopulationsofPhiladelphia,
Brazil,andCuba.MtDNA(1)andNRY(2)profileofAfricanAmericansorAfricanBrazilians(AA)andEuropeanAmericansorWhiteBrazilians(EA)
from USA (Philadelphia, PHL) and Brazil (BRZ), as well as general the population of Cuba (mix) projected either (A) as their position in the
multidimensionalscaleplots(MDS)depictingthegeneticdistanceswithrespecttoAfrica,Europe,andAmericaprojectedontothetwodimensional
planeor(B)aspiecharts,showingtherelativecontributionsfromAfrican(yellow),European(green)andNativeAmerican/Asian(pink)populations
calculatedbyADMIX.Thesecomplementaryanalysesshowgender-biasedadmixtureinSouthandNorthAmericansofprimarilyAfricanorEuropean
descent.AMOVAF representsthevariationcapturedbetweenthethreeparentalcontinents.
ST
doi:10.1371/journal.pone.0007842.g003
West/West Central Africa, Europe/West Eurasia, and the was comprised of other haplogroups, including R9a (East Asian),
Americas), and 2) by counting haplotypes assigned to be of West E1a(Melanesian)andahaplogroupMsequenceofunknownorigin.
Eurasian, Native American, Southeast Asian and African origin The ancestral composition of 204 Philadelphian European
basedonpublishedliteratureanddeepphylogeneticanalysisofour Americans was estimated to be 93% European (SD: 7%), with a
samples.ThemtDNAvariationinallthreepopulationssampledis small (although not significant) contribution from Native Amer-
shown in Figure 1 using a phylogenetic tree, adapted from its ican (1.6%, SD: 2%) and African (5.5%, SD: 5%) populations
median-joiningoutline[17]. (Table S5). Further analysis of these mtDNA haplotypes
Among 49 Senegalese, ancestry was composed mainly of confirmed that theancestry of nearly allPhiladelphian European
African haplogroups, except for six individuals with haplotypes Americans is of European origin. We observed mainly West
of Eurasian origin that are commonly found in North Africa Eurasian haplotypes (Table 1), with the exception of five
(U6 and U5b1b [38,39], Table 1). The ancestral composition haplotypes that can beconsidered East Asian (G3,M7b, andD4
of 217 Philadelphian African Americans was estimated by [40]) and North African (U6), accounting for the non-zero
ADMIXtobe9.1%European(SD:3%),1.7%NativeAmerican ADMIX estimates of Native American (here overlapping with
(SD:0.9%),and89.2%African(SD:3%).Additionaldetailsofthis SEAsia)and Africanadmixture inthispopulation.
distributioncanbefoundinTableS5(seealsoReferencesS1).
The breakdown of major mtDNA haplogroups and haplotypes is Y Chromosome (NRY)
presented in Table 1. These data suggest that our admixture Theancestryof33Senegaleseindividualswascomposedsolely
estimates are almost identical to the ancestry frequencies based of African haplogroup E, reflecting the typical pool of West
simplyoncountingtheknownWestEurasian/NativeAmerican/SE African NRY chromosomes [30]. The ancestral composition of
Asian/African haplotypes. From these estimates, the ancestral 199PhiladelphiaAfricanAmericans,basedonadmixtureanalysis
contributions were counted to be 10.1% West Eurasian, 1.4% estimates, was 31.2% West Eurasian (SD: 4%), 1.3% Native
NativeAmericanand87.1%African.Inaddition,1.4%ofancestry American (SD: 1.5%), and 67.5% African (SD: 4%) (Table S5).
PLoSONE | www.plosone.org 5 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
Table1. MitochondrialDNAvariation inthe Senegalese andPhiladelphian populations.
AAControls AACases SenegalControls EAControls EACases
Geogr.origin Majorhaplogr. Sub-haplogr. n=90 n=127 n=49 n=78 n=126
America A/B A2/B4 1/0 1/1 - - -
Asia R/E/M R9/E1/M - 1/1/1 - - -
G/N/D/M G/N1b/D4/M7b - - - - 1/1/1/1
WEurasia H/pre-HV H/pre-HV 7/0 2/1 - 35 44
HV/V HV/V - - - 2/1 5/3
J/T J/T 0/1 1/2 - 6/8 12/9
U U/K 1/0 0/5 - 8/12 20/17
I/W/X I/W/X - 1/1/0 - 3/0/2 7/2/2
Africa U6/U5* U6/U5b1 1/0 - 4/2 1/0 0/1
L0 L0a/L0k 3/1 10/0 - - -
L1 L1b/L1c 8/9 9/10 8/2 - -
L2 L2a 17 21 5 - -
L2b 1 4 1 - -
L2c 2 3 6 - -
L2d 3 1 - - -
L3 L3b 6 10 10 - -
L3d 5 11 5 - -
L3e 20 20 3 - -
L3f/L3h 1/1 9/1 2/1 - -
L4 L4h/L4g 1/1 - - - -
MajormtDNAhaplogroupspresentinPhiladelphianAfricanAmerican(AA),EuropeanAmerican(EA),andSenegalesesamples.Thephylogeneticdetailsaredepictedin
thetreeinFigure1andthesequencespresentedinFileS1.
(*Note:AlthoughhaplogroupUisofEurasianorigin,weplacedtheU6andU5b1haplotypesunder‘‘Africa’’sincetheyarecommonlyfoundinNorthAfrican
populations).
doi:10.1371/journal.pone.0007842.t001
For the NRY haplogroups, summarized in Figure 2, we see South America (Brazil) and North America (Philadelphia). As
nearly identical estimates: 31.5% West Eurasian, 1.5% Native shown in Figure 3, wepresent Multidimensional Scaling (MDS)
American,and 67%African. plots (based on Arlequin-derived Slatkin’s F genetic distances
ST
In contrast, the ancestral composition of 190 Philadelphia [20]) and admixture estimates depicted as pie charts. These
European Americans was estimated by admixture analysis to be admixtureestimateswerecalculatedusingADMIX[24],defining
almost 100% European (98.3% West Eurasian (SD: 3%), 1.1% thethree possibleancestral regions asWest/West Central Africa,
Native American (SD: 1.5%), and 0.6% African (SD: 1.4%)). Europe/West Eurasia, and the Americas (and SE Asia, in our
Theseestimateswereconsistentwithanalysisofcontinent-specific initial analysis considering four populations) and using phyloge-
NRY haplogroups: 98.5% West Eurasian, 1% Native American neticrelationshipsbetweentheobservedhaplotypes.Weobserved
and 0.5% African haplogroups. The NRY variation set into a astrikingdifferenceintheextentofadmixturebetweenNorthand
phylogenetic tree isdepicted inFigure 2. South America in both populations, with Brazilians having in
general higher admixture with the exception of White Brazilian
Autosomal Ancestry Informative Markers (AIMs) NRYs.Additionally,NRYandmtDNAprofiles,reflectinggender-
In addition to the maternal and paternal ancestry, we have specific admixture patterns, suggest diverging patterns of admix-
assessed theautosomal ancestryfor a subset of African American ture in male and female populations. This gender-biased
(n=31) and European American (n=6) samples by genotyping admixture is clearly identifiable both by the position in MDS
these on the commercially available Illumina 1509 AIMs chip, plotsandancestryproportionsofthegeneralpopulationofCuba
followed by estimating the admixture proportions using the (Figure3).ThesedatasuggestthatitisprimarilyEuropeanmales
program STRUCTURE (K=3) (Table S4). These estimates and African/Native American females that contributed to the
show.20%EuropeanancestryinAfricanAmericans(23.7%)and genetic ancestry of admixed populations of the Caribbean/South
a small African component in European Americans (2.5%) with
America.
NativeAmerican/SEAsianpopulationscontributinglessthan2%
toboth.Theseestimateswerecomparedtotheputativeautosomal
Discussion
group ancestry proportions calculated from the maternal and
paternal admixture estimates. We have characterized the mitochondrial DNA (mtDNA) and
non-recombining portionof Y-chromosome(NRY)variationin a
Gender-Biased Admixture in the Americas sample from Senegal as well as two major groups of Philadel-
We compared admixture patterns in two populations self- phians: self-identified European Americans and African Ameri-
identifiedaseitherprimarilyofAfricanorEuropeanancestryfrom cans. These two groups comprise over 88% of the Philadelphian
PLoSONE | www.plosone.org 6 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
population (45% and 43.2%, respectively, according to the 2000 represent the country. First, we analyzed admixture in published
U.S. Census). We found mainly African haplogroups in the reports that contained mtDNA and NRY data from White and
Senegalese sample, with the exception of 12.2% of Senegalese African Brazilians that were comparable to the data we collected
(3Wolof,2Fulbe,and1Sahalle)carryingU6andU5b1bmtDNA in our Philadelphia sample (Figure 3). This analysis revealed
haplogroups that, although haplogroup U is of Eurasian origin, directionaladmixturepatterns.First,separatingthematernaland
can be found throughout North Africa as a result of an ancient paternaladmixtureshowsclearlythatEuropeanmalescontributed
migrationbacktoAfrica.InPhiladelphianAfricanAmericans,we tothepopulationsofAmericatoagreaterdegreethanEuropean
observed a significant European admixture (mtDNA.9% and females. This is true for both African- and European-derived
NRY.31%) as well as a small (,2%) contribution from Native Americans, although less pronounced in the case of the
Americans.Tocalculatethecorrespondingautosomalancestryof Philadelphian European American sample. The admixture data
self-identified African Americans, accounting for both maternal inthegeneral population ofCuba support thistrend.
and paternal contributions, we used our data to compute Therefore, while male admixture is dominated by European
m =K m +K m [5], which was estimated to Y-chromosomes, the female admixture shows a remarkable
AUTO mtDNA NRY
be:78.4%African,20.1%European,and1.5%NativeAmerican. influence of African and Native American female ancestors, the
These calculated estimates seem to accurately reflect the latter prominent mainly in the South American/Caribbean pool,
autosomal group admixture, based on typing a small subset of as seen in Brazilians and Cubans. For example, both African
samples using autosomal AIMs (n=31, 74.4% African, 23.7% Americans and African Brazilians have a high percentage of
European, and 1.9% SE Asian/Native American, Table S4). admixture from European NRYs and some non-African mtDNA
Also, these estimates parallel previous reports, although our admixture that is drawn mainly from European or Native
estimates suggest a higher European contribution, especially American mtDNA pools in North and South America, respec-
compared to the 12.7–13.8% autosomal and 2.8–11% low tively. On the other hand, both European Americans from
resolution maternal European ancestry found in a sample from PhiladelphiaandWhiteBrazilians[11]donotshowadmixturein
PhiladelphiareportedbyParraetal.[9].Forexample,European their paternal gene pool (NRY being almost 100% European in
contributiontoNRY,autosomes,andmtDNAwasestimatedtobe both cases), while, as in the case of African-derived populations,
28.46%, 19.99%, and 8.51%, respectively, in African Americans the African and Native American mtDNAs contribute greatly to
from Pittsburg, Chicago, Baltimore and North Carolina [5], or the maternal pool of White Brazilians. This is in contrast to the
autosomal ancestry of African Americans from NY state was maternal pool of Philadelphian European Americans that shows
estimated to be 83% African, 15% European and 2% Native ,7% admixture from non-European sources, consistent with the
American [1]. EuropeanAmericansfromtheFBImtDNAdatabase[14].Thus,
In contrast to the admixed nature of African Americans, we therearedistinctdifferencesbetweenNorthandSouthAmericain
observedlittleadmixtureintheEuropeanAmericansample(,7% theextentofadmixturefromthethreefoundingpopulationsinthe
in mtDNA and ,1.7% in NRY). Group ancestry or uniparental pool of New World individuals who self-identify as ‘‘black’’ and
admixtureinEuropeanAmericanshasnotbeenwidelyreported. ‘‘white’’ (TableS5).
However, reports of admixture using the FBI mtDNA database To further investigate whether the patterns we observed in
[14]orautosomallocihavepresentedestimatesthatareconsistent Philadelphian, Brazilian, and Cuban populations have a similar
withourfindingsthatEuropeanpopulationshavecontributedthe impactinothercountriesintheAmericas,wecomparedpublished
vastmajorityofancestryofEuropeanAmericans.Inourcase,the mtDNA and NRY frequencies of African-derived and general
calculatedautosomaladmixtureis95.8%European,withAfrican populations, considering the demographics of the investigated
and Native American contributing less than 5% (2.8% African, countries. Focusing first on the published mtDNA and NRY
1.4%SEAsian/NativeAmerican),consistentwithpublishedwork admixture of African-derived populations of the Caribbean,
(1.6%and1.2%,respectively,intheUS[41]),aswellasourown Colombia, Brazil, and Uruguay, it is clear that they show the
autosomal estimates from a subset of the samples (95.7% same trend as the African American and African Brazilian
European). populations analyzed in this paper (Table 2). Namely, from
To further characterize the ancestry of Philadelphian popula- North to South, there is a decrease in the contribution of both
tions within the global context, we mined the literature for maternalandpaternalAfricanancestry,mainlyduetoadmixture
published reportsofmtDNAandNRYvariation, selectingBrazil with Native American females and European males. Also, we
andCubaasrepresentativeofSouthAmericaandCaribbeanthat detectthesamesex-biasedadmixture,wheremoreAfricanfemales
have sufficient resolution, sampling range, and sample size to thanmalescontributedtothepoolofAfrican-derivedpopulations
Table2. Published mtDNAand NRY profiles ofAfrican-descendedAmericans.
Africa-derivedpopulations mtDNA NRY Citations
Africa Eurasia America Africa Eurasia America
USA 94.7% 3.7% 1.6% 73.6% 26.4% 0% [33,49]
Caribbean* 90.3% 4.3% 5.3% 68% 32% 0% [50]
Colombia 78.8% 1.6% 19.7% 63% 36% 1% [51,52]
Brazil 76.7% 3% 20.3% 52.7% 44.3% 3% [12,13,34–36,53,54]
Uruguay 52.3% 19% 28.7% 30.2% 64.1% 5.1% [10]
TheboldnumbersindicatetheNorth-SouthgradientofdecreasingorincreasingtrendsinAfricanandNativeAmerican/WestEurasianadmixtureformtDNAandNRY.
(*Note:Afro-CaribbeansrepresentthepopulationsfromtheislandsofDominica,Grenada,Jamaica,St.Kitts,St.Lucia,St.Thomas,St.Vincent,andTrinidad).
doi:10.1371/journal.pone.0007842.t002
PLoSONE | www.plosone.org 7 November2009 | Volume 4 | Issue 11 | e7842
PhiladelphiaGeneticAncestry
acrosstheAmericas.WhiletheobservedNorth-Southtrendsseem of using uninterrupted single locus-like information to trace
to be consistent in African American populations, in order to maternal and paternal ancestry, the use of uniparental markers
dissectingreaterdepththeprocessesthatshapedthepopulations is limited to group ancestry estimates, bearing only very limited
of North and South America differently, we turned our focus to information about the ancestry of an individual. Care should
European-derived populations. therefore be exercised when interpreting our results on anything
Since previously published data on maternal and paternal other than thegroupancestry level.
ancestry of European-derived populations are scarce [11,14], we We have shown that estimates of group ancestry derived from
studiedthisgroupindirectlybycorrelatingtheknownmtDNAand combinedmtDNAandNRYadmixtureestimatespredictaverage
NRY ancestry of African-derived and general populations with autosomal ancestry. When separated, these estimates mirror
demographic information. First, although only 5% of the African gender-specific admixture processes, reflecting diverse socio-
slavetradearrivedinNorthAmerica[42],theUShasthehighest historical demographic processes. Also, groups sharing the label
proportion of self-identified African Americans (,13%) out of the of self-identified race across the Americas are often shaped by
regionsstudied(withtheexceptionofsomeoftheCaribbeanislands, differentsocialpressuresandthiswillbereflectedintheirgenome.
such as Jamaica, which has up to 91% self-identified African- This may add to the complexity of the population stratification
descended individuals). This implies that a significant portion of issue in molecular epidemiology, which strives for enhancing the
African parental variation in South America (and parts of the analysis by increasing the number of individuals. In the future,
Caribbean)existsaspartofadmixedpopulations(e.g.,Mulatoand characterization of source European, American and, more
Pardo, although little African admixture was reported in Mestizos importantly, genetically diverse African populations that contrib-
[43]).Toevaluatewhetherthegeneticdataareconsistentwiththis uted to the admixed pool of the Americas would enhance the
hypothesis,wecalculatedtheproportionofAfricanmaleandfemale present analysis.
lineages that were contributed to the general population solely by
individuals who self-identify as ‘‘black’’ or African (Table S6). Supporting Information
NearlyallofAfricanY-chromosomesarefoundinindividualsthat
File S1 mtDNA sequence data, NRY marker data and ethnic
self-identify as ‘‘black’’, whereas less than 50% of the African
information.
mtDNAsarefoundintheseindividuals.Inotherwords,asignificant
Found at: doi:10.1371/journal.pone.0007842.s001 (0.14 MB
fraction of African mtDNAs are found in groups that do not self-
XLS)
identify as ‘‘black’’. To determine which other populations these
African maternal lineages significantly contribute to, we estimated Table S1 Primers used for sequencing mtDNA HVS I and II.
the possible admixture in European-derived populations. While Sequence pairs and the annealing temperatures used for
theseestimatesareonlyapproximate,theproportionofcontribution amplification and sequencingof HVSI andHVS IIregions [1].
of European females to the pool of ‘‘white’’ individuals in the Found at: doi:10.1371/journal.pone.0007842.s002 (0.03 MB
Iberian-foundedCaribbeanandSouthAmericaisclearlylowerthan DOC)
to European Americans in the United States, with variable
TableS2 ListofmtDNARFLPassays.RFLPassaysofmtDNA
proportions of African and Native American females contributing
coding region that were used to ascertain the correct placement
toeachofthesepopulations(TableS7).
into aparticular mtDNA haplogroup.
Sex-biased admixture is not a process unique to the Americas.
Found at: doi:10.1371/journal.pone.0007842.s003 (0.05 MB
The pattern of NRY variation documents this phenomenon on
DOC)
every continent. For example,the uniqueY chromosome lineage
spreadby(malesrelatedto)GenghisKhanoverthevaststeppesof TableS3 ListofSNPstypedforNRY.Fromlefttoright:SNP,
Asia [44] or uni-directional mating of Bantu males and Pygmy nomenclature published by YCC in 2002 [2], nomenclature
females [45–47] can both serve as examples of the history of a published by YCC in 2008 [3], SNP rs number. Rs numbers
population being reflected in Y chromosome phylogeography in ending with the symbol # were typed using multiplex fragment
the Asian and African continents. In the Americas, European analysis. In case rs numbers are not established yet, these SNPs
males contributed significantly to all admixed populations [43]. canbetypedusingdetailsinHammer1998[4](*),Underhill2001
However, the difference between North and Caribbean/South [5] (**),or Hammer2001[6](***).
Americaliesbothinthediverseculturalhistoriesthatcategorized Found at: doi:10.1371/journal.pone.0007842.s004 (0.06 MB
peopleofadmixedancestryeitherbydescentorcolor[48],aswell DOC)
astheavailabilityofEuropeanwomen.Whileindividualswithany
Table S4 Autosomal AIMs ancestry estimates. Estimated
amountofAfricanancestrywereconsidered‘‘AfricanAmerican’’
admixture for subset of African American (n=31) and European
intheUnitedStates(the‘‘onedroprule’’),inBrazil,wheremostof
American(n=6)samplesgenotypedusingcommerciallyavailable
the first settlers were male, unions between European males and
Illuminaancestrypanelofautosomalancestryinformativemarkers
Native American/African females were common and the ‘‘skin
(AIMs). Estimates are reported by STRUCTURE software as
tone’’ of offspring was used to define an individual’s ‘‘race’’
average estimated membership in African, European and SE
[14,36]. Therefore, in contrast to European North Americans,
Asian/Native American clusters and average span of 90%
who have relatively low levels of non-European admixture from
probabilityinterval(PI)thatcanbetransposedtopseudo-standard
both male and female predecessors, individuals categorized as
error (pSE) by pSE=1/2 (PI/1.645). Posterior probabilities (Ln
‘‘White’’ Brazilians show higher levels of African and Native
P(D)) for K=1–5 were the following: 2731,387 (K1), 2491,546
American admixture. This non-European ancestry is almost
(K2), 2469,725(K3), 2467,914(K4), 2466,163(K5).
entirely derivedfrom maternallineages.
Found at: doi:10.1371/journal.pone.0007842.s005 (0.03 MB
Thereareseverallimitationsofourstudy.First,ourestimatesof
DOC)
admixture in European-derived populations in the Americas
should serve only as approximations, since this information was Table S5 MtDNA and NRY ancestry of populations from
mined from indirect sources (genetic data from complementary Philadelphia,BrazilandCuba.ProportionsofAfrican,European,
populationsanddemographics).Second,inspiteoftheadvantages andNativeAmericanancestryinthepopulations of Philadelphia,
PLoSONE | www.plosone.org 8 November2009 | Volume 4 | Issue 11 | e7842