The distribution of mitochondrial DNA haplogroup H in southern Iberia indicates ancient human genetic exchanges along the western edge of the Mediterranean 2016
Candela L. Hernández1, Jean M. Dugoujon2, Andrea Novelletto3, Juan N. Rodríguez4, Pedro Cuesta5 and Rosario Calderón1
[excerpt]
Abstract
Background
The structure of haplogroup H reveals significant differences between the western and eastern edges of the Mediterranean, as well as between the northern and southern regions. Human populations along the westernmost Mediterranean coasts, which were settled by individuals from two continents separated by a relatively narrow body of water, show the highest frequencies of mitochondrial haplogroup H. These characteristics permit the analysis of ancient migrations between both shores, which may have occurred via primitive sea crafts and early seafaring. We collected a sample of 750 autochthonous people from the southern Iberian Peninsula [Andalusians from Huelva and Granada provinces]. We performed a high-resolution analysis of haplogroup H by control region sequencing and coding SNP screening of the 337 individuals harboring this maternal marker. Our results were compared with those of a wide panel of populations, including individuals from Iberia, the Maghreb, and other regions around the Mediterranean, collected from the literature.
Results
Both Andalusian subpopulations showed a typical western European profile for the internal composition of clade H, but eastern Andalusians from Granada also revealed interesting traces from the eastern Mediterranean. The basal nodes of the most frequent H sub-haplogroups, H1 and H3, harbored many individuals of Iberian and Maghrebian origins. Derived haplotypes were found in both regions; haplotypes were shared far more frequently between Andalusia and Morocco than between Andalusia and the rest of the Maghreb. These and previous results indicate intense, ancient and sustained contact among populations on both sides of the Mediterranean.
Conclusions
Our genetic data on mtDNA diversity, combined with corresponding archaeological similarities, provide support for arguments favoring prehistoric bonds with a genetic legacy traceable in extant populations. Furthermore, the results presented here indicate that the Strait of Gibraltar and the adjacent Alboran Sea, which have often been assumed to be an insurmountable geographic barrier in prehistory, served as a frequently traveled route between continents.
Keywords
Gene flow Phylogeography Population structure Iberian Peninsula North Africa Human evolution Background
The particular spatial distribution patterns of specific mitochondrial lineages in contemporary human populations offer insights into human origins, past migration events and gene flow with defined directions and demographic consequences [1, 2, 3, 4, 5]. Mitochondrial macro-haplogroup H [Hg H] has been a focus of attention in human genetic diversity studies for more than a decade [6, 7, 8, 9]. Examining the spatial distribution of H lineages and other features associated with its evolutionary history have been pivotal in understanding the formation of the western European gene pool. Early works revealed many features of this genetic cluster, which comprises a star-like phylogeny composed of a central, major node with many rarer variants that arise from it [10, 11]. The estimated coalescence time for Hg H [~21,000 years ago, ya] has led to the proposal that the clade was involved in a post-glacial population re-expansion from southwestern Europe to the rest of the continent.
Hg H [the native Euroasiatic marker par excellence] clearly dominates the mitochondrial DNA [mtDNA] gene pool of Europeans [~40-45% on average] [8, 9]. Hg H has an internally complex structure, with regional geographic specificities across Europe and the Mediterranean Basin. The patterns of variation revealed by H lineages [and sub-lineages] were better characterized as more refined molecular technologies were developed, which enabled, for example, screening of coding region Single Nucleotide Polymorphisms [SNPs] [7, 8] and complete sequencing [6]. This technical progress increased phylogenetic resolution, thus demonstrating that i] the number of internal branches shaping clade H is significantly greater than in other mtDNA Hgs widespread in Europe [7], and ii] the observed Hg H variation in eastern regions [e.g., Near/Middle East and Caucasus] shows marked differences to that found in western Europe [9]. Moreover, the classification of H mtDNA samples in sub-lineages with only control region variants has proven in most cases to be unreliable due to the recurrence of some polymorphisms and the absence of diagnostic sites [7].
Basques and other neighboring populations from the northern Iberian Peninsula have been excellent candidates for studying Hg H composition in western Europe [12, 13, 14]. These populations are presumed to be a source of the European post-glacial peopling signaled by some H sub-Hgs. For example, frequency peaks of mtDNA lineages H1 and H3 characterize Cantabrian/Iberian populations. In this line, Basques are presumably the native population of derived sub-branches emerging from H1 and H3, including H1j1, H1t1 and H3c2a. These, together with H2a5, compose ~40% of Hg H and are absent in other populations [13].
Hg H lineages are also dispersed outside of Europe. A pertinent example is found in North Africa, where Hg H is the main marker of the European influence across the Maghreb [15, 16]. Most North African H sequences belong to sub-Hgs H1 [42%] and H3 [13%] [17]; thus Hg H seems to be structured here in much the same way as in Iberia. Nevertheless, some surveys [16] have shown that great genetic diversity could be hidden in Hg H profiles among northern Africans. It has also been assumed that a portion of H lineages in the region were transferred by a post-glacial wave of expansion from the Franco-Cantabrian/Iberian region southward. Thus, a detailed analysis of this clade in populations located along this supposed migration route is crucial. We present here for the first time a comprehensive, high-resolution phylogenetic portrait of Hg H in Andalusia; this region would have been midway between the area of departure and the southernmost limit of mitochondrial H sequences in North Africa. It is thus possible to explore whether the southernmost region of Spain served as a stopover for Hg H descendant lineages in their intercontinental route or if other maternal genes underwent direct, non-mediated migration to North Africa, either through the Strait of Gibraltar or the surrounding maritime region. For this analysis, we selected two autochthonous populations from Huelva and Granada provinces, which are geographically placed at the western and eastern ends of Andalusia, respectively. The global mtDNA variation in the two territories exhibits significant genetic differentiation [18]. The population structuring seems to be primarily caused by the differential weight of African lineages U6, M1 and L, which are far more represented in the western than in eastern Andalusia. The complete sequencing of African mtDNA lineages found in the Andalusian gene pool [19] have interestingly revealed the occurrence of ancient trans-continental contact between northwestern Africa and Iberia, with Andalusia being the Atlantic side of the Peninsula where most African maternal traces are concentrated when compared to the rest of Europe.
This study attempts to provide an accurate picture of the distinctiveness of the Andalusian matrilineal gene pool in comparison to other Iberian and Mediterranean populations. The essential role played by Iberia in disseminating specific lineages into North Africa through migrations since prehistoric times is supported here by mtDNA Hg H.
Fig. 1 Map of the region. a Location of the Andalusian territory in the context of the Mediterranean region. b Municipalities of sampling in Huelva and Granada provinces:
Results
Dissection of Hg H in Andalusia
The observed frequencies of Hg H and its sub-Hgs in Andalusians are shown in Table 1. The overall frequency [44%] was similar to other Iberian populations [42% on average]. Among the 337 Hg H samples, 41 internal branches were observed, of which 24 had a frequency < 0.01. Hg diversity estimates [±SD] varied between 0.847 ± 0.027 in the west [Huelva] and 0.888 ± 0.013 in the east [Granada]. In comparison, the Basques had the highest frequency of Hg H [∼55%] and the lowest genetic diversity [0.646 ± 0.041, estimated from data in [13]]. This finding can be explained because most Basques carrying Hg H belonged either to the H1* paragroup or to some autochthonous H1 sub-branches [amounting to ~60% of mtDNA clade H variation].
Phylogeography of Hg H on both sides of the Strait of Gibraltar
Figure 3 shows surface maps based on frequency distributions of clades H, H1 and H3 in Europe and surrounding areas. The maps make more than apparent the presence of spatial structure, with frequencies decreasing from west to east across both the northern and southern sides of the Mediterranean [Moran’s correlograms Bonferroni corrected significance of 0.0000]. A pattern characterized by positive [and significant] autocorrelation values for short distances and significant negative autocorrelations at long distances [further details in [38]emerged for H1 when considering only the European population points; the correlogram exhibited a linear gradient [Additional file 10 B] with no inflection points.
In connection with the above findings, major differences for Hg H composition were also detected between populations located in the west and east of the Mediterranean Basin [F CT = 0.065, P-value <0.001], with H1 contributing more prominently to that structure. Roostalu et al. [9] interpreted these distinctive profiles as the result of a limited maternal gene flow after the Last Glacial Maximum [LGM] between the geographic extremes of the Mediterranean.
Figure 4 shows the networks of 708 [a] and 264 [b.] mtDNA sequences/haplotypes found in 12 Iberian and 5 North African populations for sub-Hgs H1 and H3, respectively. Both networks are strongly star-shaped. Interestingly, the H1 basal node [Fig. 4a] contains the entire spectrum of population samples used in the analysis, representing 46.5% of the total H1 sequences. The same is true for H3 [56.8% of the total H3 sequences]. Maternal relationships between Iberia and North Africa rely on a concomitant nodal presence of sequences from those regions in most of the sub-Hgs, with Iberians being the main contributors to those nodes [e.g. H1b initial radiations, Fig. 4a]. Most tip branches stemming directly from basal nodes [see the multiple radiations around them defined by single mutations] were unique and belonged in many cases to individuals of North African origin. In Hg H1 in particular, branches indicate a distinct radiation in North Africa compared to Iberia [see upper-left corner in Fig. 4a].
Discussion
The Iberian Peninsula, given its geographic position and role as a human refuge during the last glaciation, is an especially appropriate territory to study bidirectional migrations from Europe to North Africa and back following the LGM. The analysis of continent-specific mtDNA Hgs, such as H in Europe, the high frequencies of some H sub-Hgs in the western extreme of the Mediterranean and the estimated coalescence dates, mostly concentrated ~10,000 ya, make mtDNA Hg H a suitable candidate to investigate these migrations. Our study on Hg H with high phylogenetic resolution in southern Iberia highlights how intertwined and strongly rooted the evolutionary history among western Mediterranean regions has been, based on maternal heritage. Signals of ancient migrations across the Strait of Gibraltar and its surrounding maritime area have also been observed using other genomic markers [39].
Today, there is empirical evidence that the contribution of European and African mtDNA lineages is not equivalent on the opposite coasts of the western Mediterranean. The traces left by African mtDNA clades U6, M1 and L in the European gene pool are greatest in the Iberian Peninsula, with the highest frequencies of occurrence [~15%] in western Andalusia [18]. In contrast, frequencies of Hg H in northwestern Africa were comparable to those found in southwestern Europe [e.g., Iberian Peninsula, 40-45%]. The described scenario suggest that the Iberian Peninsula served as both a sink [recipient of genes] and source [donor of genes] population in relationship to Africa. The role of Iberia in the spread of genes from Africa to the rest of Europe would also be paralleled by the transmission of other genetic traces southward along the Mediterranean coast and across the Strait of Gibraltar. The lower demographic size in northwestern Africa in relation to Iberia during past millennia would have favored processes increasing the frequency of Hg H in the former region.
As the climate warmed after the end of the LGM, wide ranging movements of populations through Europe became possible [40]. Those post-LGM human movements from the three major refugia in the Mediterranean, Iberia, Italy and the Balkans, also reached the North African fringe [41]. In a broader context, the genetic structuring of Europe revealed by genome-wide [GW] studies primarily exhibits a latitudinal trend [42, 43], which coincides well with post-glacial processes and other later demographic events. The general GW autosomal pattern is further enriched by interesting regional variations. For example, the Italian and Iberian peninsulas share little recent common ancestry with other European populations, a fact that has been linked to old substructuring [44]. Our Iberian mtDNA data [18, 19] point to another possible interpretation for this local specificity: the genetic influence from the neighboring African continent.
Phylogeographic and phylogenetic analyses performed in the present study suggest that the main H sub-Hg indicators of post-glacial population expansions from the Iberian refuge were H1 and H3, the most frequent maternal lineages in Iberia. The time depths calculated for these sub-clades in Iberia and the Maghreb, together with the presence of specific H variants in some North African populations –revealed at a deep phylogenetic resolution [e.g. H1v1, H1w, and others in Tuareg from Libya [16]]– it would support scenarios of ancient radiations in the direction Iberia-to-North Africa. Accordingly, the genetic structure analysis of Mediterranean populations by AMOVA demonstrated that Italy was significantly different from Iberia [see Table 4] and that this dissimilarity was not as visible between Iberian and Maghrebian populations. These findings suggest that gene flow between Europe and northwestern Africa, involving Hg H, would have occurred primarily through Iberia. The HCA also provided strong support for this assertion.
Interestingly, the patterns of variation displayed by Hgs H1, H3 and H5 [see surface maps in Fig. 3 and Additional file 6], together with other evidence emerging from the HCA, suggest more than one post-glacial migration from the Franco-Cantabrian refuge. In a first, more geographically restricted migration, Basque populations with high levels of H1 would have been the primary participants. A second migration would have harbored the most recent sub-Hg H3 and reached the Maghreb. A probable third migration, with a center of origin in the eastern Mediterranean, would have carried the H5 mtDNA lineage into Italy and, to a lesser extent, the western Mediterranean, with a reduced impact in northern and western Iberia. These human movements could have been conducted, at least in part, by sea. It is interesting to note that the native population from Sardinia exhibits high frequencies of H1 and H3, a fact that would indicate early maritime relationships with the near mainland Europe.
An interesting result of the H1 and H3 star-like networks of the Iberian and Maghrebian populations [top left in Fig. 4a] was the number of haplotypes found two or more times in North Africa. This may represent an incipient molecular radiation, indicative of early human migrations between the western Mediterranean shores. Other derived variants from larger nodes occurred at variable frequencies [e.g., the most predominant in the H1 skeleton network occurs 74 times] and African individuals contributing to these nodes were scarce. This suggests that historical migratory flow between Iberia and the Maghreb had a lower genetic influence than ancient human movements. As shown in Table 2, the proportion of shared haplotypes between Andalusia and Morocco was high, but were considerably lower when comparing Andalusia against the Maghreb. Thus, migrations would have shown an expansion within a reduced geographic range.
The necessity of crossing a body of water to reach the African continent allows us to analyze the opportunities for navigation in early times. Broodbank [45, 46] considered any possible maritime activity in the Mediterranean before c. 12,000 ya to be episodic and of limited evolutionary significance. However, the presence of Mesolithic seafaring is now quite firmly established based on recent finds from Crete, the Aegean, Sardinia and Corsica. The unambiguous evidence for human presence on Mediterranean islands still dates to no more than c. 16,000 ya [47]. Thus, maritime contact between the two continents could have begun before 10,000 ya, the time depth for which the most frequent H Hgs can be used to detect migrations. This conclusion is strongly supported by similarities between lithic industries dominated by microliths and backed bladelets of the Iberomaurusian and Magdalenian and by Taforalt harpoon [northeastern Morocco], which has three short barbs on one edge and is contemporaneous with the Final Magdalenian in Mediterranean Spain [48]. Harpoons detected in prehistoric excavations from southern Catalonia to Malaga Bay [the Spanish Mediterranean corridor] are also characterized by having a single row of teeth. Barbed harpoons were the most common implement in the Upper Magdalenian [49] and could have been introduced to North Africa by the advancing ice. Archaeological dates confirm the simultaneity of the Epipalaeolithic-Neolithic transition in southern Spain and northern Morocco.
The rapid dispersal of innovations suggests that they were circulated through already existing networks [50]. The earliest presence of Neolithic industries in southern Iberia were dated at least 7500 calibrated ya [51]. All major elements of the Neolithic package arrived in southeastern Spain from the central Mediterranean and reached North Africa through west Mediterranean networks. However, some elements, such as pointed-based vessels, Almagra decoration, and lentils, were subsequently modified in North Africa before being dispersed to Iberia [52].
The first maritime contact would have been associated with fishing in waters more distant from the coast and with increasingly larger watercrafts used for the capture of large fish. In Nerja cave [Malaga], the remains of large sea mammals [e.g., monk seal, harbor seal, dolphin and large cetaceans of the Delphinidae family] have been found [49]. The route stopping on the small island of Alboran, between eastern Morocco and southeastern Andalusia, is particularly interesting for crossing the Alboran Sea, with land permanently in sight [19]. A bidirectional movement of women could occur as part of these dynamic journeys. The transport of the Neolithic package, including farmers, domesticated animals, seeds and tools, must have required sea crafts of considerable size, and this could increase the intensity of potential admixture. Given the high frequencies of Hg H on both shores, this gene flow would have been intense for thousands of years.
Reed rafts tied with green stems were the probable primitive watercrafts used in the Alboran Sea. The use of leather sails is also possible, since leather treatment technology is ancient. The Libyan Desert, which is adjacent to the coast, would have been a stronger barrier to human movement than maritime travel. Therefore, prehistoric bidirectional seafaring in the westernmost Mediterranean seems quite well evidenced by the archaeological record and it is also supported by genetic data. Protohistoric and historic episodes reinforced the connections between the Maghreb and Iberia across the administrative and politic unities established during the Roman Empire and, later, with the Muslim expansion, with relevant sociocultural and economic consequences. Thus, the westernmost extreme of the Mediterranean likely did not represent a true physical barrier to gene flow between both continents.
The patterns of variation in the Y-chromosome between western and eastern Andalusians, based on 416 males, have also been investigated for a set of Y-Short Tandem Repeats [Y-STRs] and Y-SNPs [53, 54, 55], Calderón et al., unpublished data] in combination to mtDNA analyses [[18, 19] and present study]. In general, for both uniparental makers, Andalusians exhibit a typical western European genetic background, with peak frequencies of mtDNA Hg H and Y-chromosome Hg R1b1b2-M269 [45% and 60%, respectively]. Interestingly, our results have further revealed that the influence of African female input is far more significant when compared to male influence in contemporary Andalusians. The lack of correspondence between the maternal and paternal genetic profiles of human populations reflects intrinsic differences in migratory behavior related to sex-biased processes and admixture, as well as differences in male and female effective population sizes related to the variance in reproductive success affected, for example, by polygyny [56, 57].
Conclusion
Here we present arguments that the western Mediterranean has not been a barrier to human gene flow and, more specifically, that the Strait of Gibraltar and adjacent areas acted as an active bridge between Africa and Europe. A pertinent example is found in Andalusia and its autochthonous contemporaneous populations, whose genetic composition is notably influenced both by the close proximity to North Africa and by intense involvement in the history of the Mediterranean.
Posts: 42918 | From: , | Registered: Jan 2010
| IP: Logged |
posted
"Here we present arguments that the western Mediterranean has not been a barrier to human gene flow and, more specifically, that the Strait of Gibraltar and adjacent areas acted as an active bridge between Africa and Europe."
Posts: 22234 | From: האם אינכם כילדי הכרית אלי בני ישראל | Registered: Nov 2010
| IP: Logged |
UBBFriend: Email this page to someone!
![[Big Grin]](biggrin.gif)
Printer-friendly view of this topic