L1, L2 & L3 haplogroups: mtDNA Analysis of Nile River Valley Populations

These studies may perhaps be forgiven for their shortcomings, when considering the dates of publication and the extant knowledge [in the field in question] at the time of publication. Nonetheless, for those who might be in the little know, a few things needed to be straightened out:

C. Fox, 1997:

mtDNA analysis in ancient Nubians supports the existence of gene flow between sub-Sahara and North Africa in the Nile valley

Abstract:

The Hpal (np3,592) mitochondrial DNA marker is a selectively neutral mutation that is very common in sub-Saharan Africa and is almost absent in North African and European populations. It has been screened in a Meroitic sample from ancient Nubia through PCR amplification and posterior enzyme digestion, to evaluate the sub-Saharan genetic influences in this population. From 29 individuals analysed, only 15 yield positive amplifications, four of them (26·7%) displaying the sub-Saharan African marker. Hpa I (np3,592) marker is present in the sub-Saharan populations at a frequency of 68·7 on average. Thus, the frequency of genes from this area in the Merotic Nubian population can be estimated at around 39% (with a confidence interval from 22% to 55%). The frequency obtained fits in a south-north decreasing gradient of Hpa I (np3,592) along the African continent. Results suggest that morphological changes observed historically in the Nubian populations are more likely to be due to the existence of south-north gene flow through the Nile Valley than to in-situ evolution.

Krings et al study, 1999:

mtDNA Analysis of Nile River Valley Populations: A Genetic Corridor or aBarrier to Migration?

To assess the extent to which the Nile River Valley has been a corridor for human migrations between Egypt and sub-Saharan Africa, we analyzed mtDNA variation in 224 individuals from various locations along the river. Sequences of the first hypervariable segment (HV1) of the mtDNA control region and a polymorphic HpaI site at position 3592 allowed us to designate each mtDNA as being of "northern" or "southern" affiliation. Proportions of northern and southern mtDNA differed significantly between Egypt, Nubia, and the southern Sudan. At slowly evolving sites within HV1, northern-mtDNA diversity was highest in Egypt and lowest in the southern Sudan, and southern-mtDNA diversity was highest in the southern Sudan and lowest in Egypt, indicating that migrations had occurred bidirectionally along the Nile River Valley. Egypt and Nubia have low and similar amounts of divergence for both mtDNA types, which is consistent with historical evidence for long-term interactions between Egypt and Nubia. Spatial autocorrelation analysis demonstrates a smooth gradient of decreasing genetic similarity of mtDNA types as geographic distance between sampling localities increases, strongly suggesting gene flow along the Nile, with no evident barriers. We conclude that these migrations probably occurred within the past few hundred to few thousand years and that the migration from north to south was either earlier or lesser in the extent of gene flow than the migration from south to north.

A Response to C. Fox and Krings et al.

The authors rely on just the hypervariable region, which as they acknowledge is known to be limited on the very probable potential of harboring "parallel mutations", and the absence or presence of the restriction enzyme identified site of HpaI site. They say:

Utilizing three sites in this manner should minimize incorrect classification of mtDNA types; however, because the two sites in HV1 are subject to repeated mutations (Hasegawa et al. 1993), we were concerned that some incorrect classification might nevertheless occur....

Approximately one-third of the Nile River Valley mtDNA types could be unambiguously classified on the basis of this database comparison; the results were nearly completely concordant with the classification based on the three sites, with the single discrepancy involving an Egyptian mtDNA that, on the basis of the three sites, was classified as northern but, on the basis of the database comparison, was classified as southern because it was identical to sequences found in two Songhai from Mali and two Kikuyu from Kenya (Watson et al. 1996). Because alteration of the classification of this one sequence does not significantly change any of the results that follow, this Egyptian mtDNA was still classified as northern, in accordance with the results from use of the three sites.

But what do we know of L3 based lineages, do they all have what the authors call?...

In addition, it has been proposed that the HpaI site at 3592 has a single origin in sub-Saharan Africa

Would M1 for instance have this site detected as positive?

The present author sees the method used herein, almost akin to using RFLP in Y chromosomes and microsatellite motifs, without having details on binary markers that could clearly define the monophyletic units themselves, thereby pooling otherwise different lineages based on absence or presence of certain restriction sites. We've seen this in the case of Y chromosomes, wherein E-M78, E-M81 and some other yet-to-be identified lineage were pooled together based on certain RFLP sequences, but when binary markers were tested, these related but distinct lineages came to the fore. Exclusively relying on three hypervariable segment sites for analysis, has definitely got to be one of the weakest aspects of this study, for reasons just mentioned.

In relation to the questions just asked above, we have:

The main sub-Saharan African haplotypes, thus, are characterized by a combination of 10394DdeI(+)/10397AluI(-)/3592HpaI(+) markers (haplogroup L, comprising the LI and L2 lineages) (Chen et al. 1995, 2000). A less frequent group of haplotypes lacks the African-specific 3592 HpaI marker [10394DdeI(+)/ 10397AluI(-)/3592HpaI(-)] (Chen et al. 1995, 2000) and has been designated as haplogroup L3 (Watson et al. 1997). A minority of African haplotypes (2.3% of Africans) lack all three of these mutations [10394DdeI(-)/10397AluI(-)/ 3592HpalI(-)]. Some align with the European lineage U (Chen et al. 2000), but a number of the mtDNAs belong to branches of the African haplogroup L3, itself derived from African haplogroup L1 (Watson et al 1997). — Clemencia Rodas et al., Mitochondrial DNA studies show asymmetrical Amerindian admixture in Afro-Colombian and Mestizo populations, 2003.

Note that even in the above piece, even though restriction enzyme markers allow for grouping into superphyletic units [in the case of those which were deemed to belong to the L3 supergroup, it isn't even that clear-cut, as there are those which lack all the restriction markers tested, and then there are those which harbor 10394 DdeI (+). L1 and L2 lineages on the other hand, seemed to have been pooled under that which tested positive for 3592 HpaI—as such, without further information, we are hardly told about which ones specifically fall into Hgs L1 and L2 respectively], these alone don't tell us much about the specific designated haplogroups or else subclades actually involved. *Distinct* subclades would have just been pooled under specific restriction enzyme identified sites, pending more information on subclade-specific characteristic or definitive markers.

At this point, the present author doesn't question the African origin of Hg M1, and by extension, the possibility of such for the *basic* characteristic motifs for the M macro-haplogroup; The present author merely brought Hg M1 up, as an example of one of the notable L3 derived lineages in East Africa, extending from the sub-Saharan region to the Northeast end. Hg M1 happens to occur predominantly in Africa. The present author's motive here is to get the reader to start thinking about the L3 lineage—which is just as African in origin as L1 and L2 lineages, but lack the so-called "African-specific" site. So, the fact that the term "African-specific" was used, might mislead someone to think that the other mtDNA lineages under study couldn't have also been African.