Received August 4, 2004; Accepted September 8, 2004.
To resolve the phylogeny of the autochthonous mitochondrial DNA (mtDNA) haplogroups of India and determine the relationship between the Indian and western Eurasian mtDNA pools more precisely, a diverse subset of 75 macrohaplogroup N lineages was chosen for complete sequencing from a collection of >800 control-region sequences sampled across India. We identified five new autochthonous haplogroups (R7, R8, R30, R31, and N5) and fully characterized the autochthonous haplogroups (R5, R6, N1d, U2a, U2b, and U2c) that were previously described only by first hypervariable segment (HVS-I) sequencing and coding-region restriction-fragment–length polymorphism analysis. Our findings demonstrate that the Indian mtDNA pool, even when restricted to macrohaplogroup N, harbors at least as many deepest-branching lineages as the western Eurasian mtDNA pool. Moreover, the distribution of the earliest branches within haplogroups M, N, and R across Eurasia and Oceania provides additional evidence for a three-founder-mtDNA scenario and a single migration route out of Africa.
South Asia as the Gate from Africa to Southeast Asia
In regard to the evolution and spread of modern humans, the genetic evidence obtained from high-resolution uniparental (i.e., mtDNA and Y chromosome) markers clearly supports the recent African origin of modern humans. Although this key feature of the Out-of-Africa scenario is widely accepted, the specific question of the routes used by modern humans to leave Africa is still being disputed. The traditional view, born out of the analysis of classic markers and the interpretation of population trees, is that there were two distinct exit routes: a southern route along the Asian coastline and a northern route through the Levant via Central Asia. The modernized variant of this model, however, suggests that both migrations stemmed from a single source in Africa (Cavalli-Sforza and Feldman 2003). Since the Eurasian/Oceanian mtDNA pool consists of two macrohaplogroups, a simplistic interpretation would tag one macrohaplogroup (M) to the southern root and the other (N) to the northern route (Maca-Meyer et al. 2001). This “one haplogroup–one migration” model, however, does not sit easily with the geographic distribution (fig. 2) and with estimated ages of the most basal branches of these macrohaplogroups and of R in particular. Moreover, the mtDNA record from Oceania indicates three autochthonous subhaplogroups of R and N (i.e., P, O, and S; see fig. 2) as well as two autochthonous M subhaplogroups (Ingman et al. 2000). Our analysis of the indigenous haplogroup R lineages in India points to a common first spread of the root haplotypes of M, N, and R along the southern route some 60–70 kya, since haplogroup M was estimated to be essentially of this age in India (Kivisild et al. 2003a) as well as in East Asia (Kong et al. 2003). The analysis further shows that the intrusion of haplogroup U2 (from the Near East ~50 kya) postdated the proto-settlement along the southern route, thus giving further support to the migration scenario proposed by Kivisild et al. (1999a, 1999b). Equating these two events, as Cann (2001) apparently suggested in her figure 1, is then difficult to reconcile with the mtDNA data. At the other extreme, a very early protomigration along the southern Asian coast, before the Toba event (~74 kya), as proposed by Oppenheimer (2003), does not receive support from the complete sequence data, at least given the employed calibration of the molecular clock. It must be noted, however, that this calibration is inevitably fraught with uncertainty; calibrating mtDNA founder events directly against well documented archeological records may provide more precise estimates in the future.
Well observed comment at the end. All the DNA dates are estimates, and quite frankly steps should be taken to calibrate them against known and well dated migration events. These guys need to talk to archaeologists more. Also, it no-one ever seems to acknowledge that mtDNA seems to under selection. There were humans in Morocco 162k ago, and still at 82k ago; which rather buggers up the nice ‘single exit’ scenario.