over the sands and far away: interpreting an iberian mitochondrial lineage with ancient western...

Download Over the sands and far away: Interpreting an Iberian mitochondrial lineage with ancient Western African origins

Post on 14-Feb-2017




7 download

Embed Size (px)


  • Original Research Article

    Over the Sands and Far Away: Interpreting an Iberian Mitochondrial Lineagewith Ancient Western African Origins

    ANTONIO F. PARDI ~NAS,1 JOSE LUIS MARTINEZ,2 AGUSTIN ROCA,3 EVA GARCIA-VAZQUEZ,3 AND BELEN LOPEZ1*1Departamento de Biologa de Organismos y Sistemas, Universidad de Oviedo, Asturias 33071, Spain2Unidad de Ensayos Biotecnologicos y Biomedicos, Servicios Cientfico-Tecnicos, Universidad de Oviedo, Asturias 33006, Spain3Departamento de Biologa Funcional, Universidad de Oviedo, Asturias 33006, Spain

    Objectives: There is an ongoing effort to characterize the genetic links between Africa and Europe, mostly using lin-eages and haplotypes that are specific to one continent but had an ancient origin in the other. Mitochondrial DNA hasbeen proven to be a very useful tool for this purpose since a high number of putatively European-specific variants of theAfrican L* lineages have been defined over the years. Due to their geographic locations, Spain and Portugal seem to beideal places for searching for these lineages.

    Methods: Five members of a minor branch of haplogroup L3f were found in recent DNA samplings in the region ofAsturias (Northern Spain), which is known for its historical isolation. The frequency of L3f in this population (1%) isunexpectedly high in comparison with other related lineages in Europe. Complete mitochondrial DNA sequencing ofthese L3f lineages, as well phylogenetic and phylogeographic comparative analyses have been performed.

    Results: The L3f variant found in Asturias seems to constitute an Iberian-specific haplogroup, distantly related tolineages in Northern Africa and with a deep ancestry in Western Africa. Coalescent algorithms estimate the minimumarrival time as 8,000 years ago, and a possible route through the Gibraltar Strait.

    Conclusions: Results are concordant with a previously proposed Neolithic connection between Southern Europeand Western Africa, which might be key to the proper understanding of the ancient links between these two continents.Am. J. Hum. Biol. 00:000000, 2014. VC 2014 Wiley Periodicals, Inc.

    Since the first high-resolution population genetics studiesit is known that all mitochondrial DNA (mtDNA) lineagesof the human species have a deep origin in Africa (Chenet al., 1995; Salas et al., 2002; Scheinfeldt et al., 2010).Many questions underlying that origin have been a matterof great interest for the fields of anthropology and genetics,including the inference of prehistoric population structure(Fadhlaoui-Zid et al., 2011; Rosenberg et al., 2002), regionaldemographic evolution (Atkinson et al., 2009; Gignouxet al., 2011), ancient intracontinental dispersals and migra-tions (Harich et al., 2010; Quintana-Murci et al., 2008), andthe occurrence and dating of the Out-of-Africa event (OOA;Fernandes et al., 2012). The relationship between themtDNA diversity in Africa and other continents is thus asubject for continuous research, as lineages with a recentAfrican ancestry, mostly included in paragroup L*, can befound worldwide (Behar et al., 2007).

    As expected, regions where these lineages are more com-mon include the Iberian Peninsula and the Middle Eastdue to their geographical position (Amorim et al., 2005;Gonder et al., 2007), as well as many American countriesdue to the recent Trans-Atlantic slave trade (Wilson et al.,2012; Zakharia et al., 2009). Regarding the Iberian Penin-sula, the Gibraltar Strait, being only 13 Km of width, is ageographic passage between Africa and Europe that hasbeen permeable at least since the Last Glacial Maximum ofaround 20,000 years ago (Anderung et al., 2005; Curratet al., 2010; Plaza et al., 2003). This might explain the fre-quencies of some African mtDNA haplogroups at as highas 6% in Iberia (Pino-Yanes et al., 2011), as well as theirappearance as far as in Eastern Europe (Malyarchuk andCzarny, 2005; Malyarchuk et al., 2008).

    Setting a timeframe for the arrival of these lineages iscomplicated, even when molecular dating is possible(Pereira et al., 2010). From complete mtDNA sequences,

    and using a global database for comparison, a recent studyby Cerezo et al. (2012) determined that the majority (65%)of putatively European-specific L* haplogroups were theconsequence of recent introgressions, likely following theroutes settled during the Roman Empire and up to theAtlantic slave-trade era. The remaining haplotypes, mostof them grouped in uncommon branches of the L1 and L2haplogroups, would have truly arisen in Europe as sug-gested by their very specific diagnostic mutations, withtheir ancestors arriving at the Peninsula at the beginningof the Neolithic, as early as 11,000 years ago.

    In the Iberian Peninsula, as described in other Euro-pean countries such as Italy (Brisighelli et al., 2012) orRussia (Morozova et al., 2012), the spatial distribution ofL* haplogroups is not homogenous (Supporting Informa-tion Table S1), being probably affected by geographical,social and historical patterns (Hernandez et al., 2014).Intuitive expectations are that regions in the Southshould have a relatively high number of African lineages(Casas et al., 2006), while those in the North should showonly minor contributions of these haplogroups (Maca-Meyer et al., 2003) or none at all (Lopez-Parra et al.,

    Additional Supporting Information may be found in the online versionof this article at the publishers website.

    Contract grant sponsor: Severo Ochoa FICYT-PCTI Grant, the Astu-rias Regional Government; Contract grant number: BP09038.

    *Correspondence to: Belen Lopez Martinez, Universidad de Oviedo,Area de Antropologa Fsica, Facultad de Biologa. Dpto. BOS, C/ Cat-edratico Rodrigo Ura s/n, 33071 Oviedo, Asturias, Spain. E-mail:lopezbelen@uniovi.es

    Received 6 December 2013; Revision received 11 July 2014; Accepted 17July 2014

    DOI: 10.1002/ajhb.22601Published online 00 Month 2014 in Wiley Online Library


    VC 2014 Wiley Periodicals, Inc.


  • 2009). However, African lineages do occur in northern Ibe-ria, some of them being shared with southern regions(Alvarez et al., 2010), while others appearing as rare out-liers (Cerezo et al., 2012).

    A remarkable case is the north-western coastal regionof Asturias, considered by historical sources to have suf-fered a certain demographic isolation until the 19th20thcenturies (Infiesta, 2005) and a low overall input of medie-val Arab emigration (Ruiz de la Pe~na, 1979). Recently,both facts were suggested in an analysis of the mitochon-drial hypervariable segment 1 (HVS1) of an ancestry-controlled sample of 429 volunteers (Pardi~nas et al.,2012b). Among the seven individuals from L* lineagesthat were found, five could be ascribed to the uncommonhaplogroup L3f1b4a, which was later supported by fullcontrol-region sequencing (Pardi~nas et al., 2012a).

    This haplogroup was first described in a South Africanindividual, with related haplotypes scattered across theAfrican continent reaching even the Middle East (Beharet al., 2008; Cerny et al., 2009). Its ancestor, L3f, seems tohave arisen in Eastern Africa around 50,000 years ago,and shows three branches well-defined in terms of theirdistribution inside the African continent, where they canamount to 10% of the local mtDNA frequency (Soareset al., 2012). By contrast, they are extremely rare outsideAfrica with the exception of African-Americans (Cernyet al., 2009). So far, and to our knowledge, L3f lineageshave been only sporadically reported in European popula-tions, most of them based in control-region motifs andaccounting to one or two samples per study (Casas et al.,2006; Irwin et al., 2007; Karachanak et al., 2012; Prietoet al., 2011; Turchi et al., 2008). Thus, the five members ofa very particular branch of this haplogroup found in Astu-rias provide a quite exceptional opportunity to delve deeperinto the molecular characteristics of this mtDNA lineage.This examination will elucidate the details and timing ofits entry into Iberia, and expand our knowledge of the his-torical genetic exchanges between Africa and Europe.


    This study was approved by the Research and EthicsCommittees of the Central University Hospital of Astu-rias and the University of Leon (Spain). All sample donorsgave written informed consent prior to their recruitment,according to the Spanish Law for Biomedical Research(Law 14/2007July 3rd).

    Full mtDNA amplification and sequencing

    Samples used for the study were collected for a previousstudy using buccal-swabs and a Chelex-extraction proto-col (Pardi~nas et al., 2012a,b), after which they werenamed AST022, AST227, AST233, AST347 and AST577.All five had been ascribed to haplogroup L3f1b4a onthe basis of their control-region motifs, which showedtwo different haplotypes separated by two mutations:16209C-16223T-16311C-16355T-16519C-73G-189G-200G-263G-309.1C-315.1C-523delA-524delC (AST022) and16209C-16223T-16311C-16519C-73G-189G-200G-263G-315.1C-523delA-524delC (all the others). The five sampledonors came from different towns, although two of themwere from the same region inside Asturias. Nevertheless,ancestor birthplaces and surnames reported during thesampling process showed no family connections betweenany of them up to two generations ago.

    In this study, the five samples were chosen for completemtDNA sequencing using the revised protocol by Ramoset al. (2009, 2011). In this procedure, the mtDNA genomeis divided in nine overlapping fragments, using oligonu-cleotides specifically designed to prevent the co-amplification of nuclear insertions of mitochondrial origin(NUMTs). This step, coupled with double-strand amplifi-cation for each fragment using internal and flanking pri-mers, minimizes several sources of incorrect base-calls onthe resulting data (Parson and Bandelt, 2007; Yao et al.,2008)