AAlways on the lookout for "anomalies" in global distributions of genetic markers I came across another one which has a very high frequency in two areas and a void in between: The high frequency area is Europe, Southwestern Asia and North Africa, in other words, the homeland of Neanderthals and the areas where they or humans with Neanderthal admixture later migrated to. The other "high" (though it is quite low indeed) frequency region is South America. The void is in East Asia and Siberia.
The authors of the paper describing this gene (Donnelly et al., 2010)  mention the possibility of a Neanderthal admixture in modern humans.
The gene in question is the MAPT (for: microtubular associated protein tau) gene; it is classed into two families of chromosomes, the H1 and the H2 haplotypes:
- H1 has been widely studied because of its close relationship with several neural disorders such as Alzheimer and Parkinson diseases as well as a variety of palsy and sporadic fronto temporal dementia.   H1 is the most frequent variant among modern humans and is found in all human populations at frequencies ranging from 100% to 70%. It has evolved into several sub-haplotypes (H1a to x).
- H2, is the ancestral variant (and is found in gorillas and chimps); it is also known as "inversion haplotype". It is found at variable but low frequencies across the globe (0 to 30%), with a marked cline as you move away from Western Eurasia and North Africa. H2 is associated with a higher fecundity rate and is believed to be positively selected for. 
A discontinuous distribution
The highest global frequencies for H2 are found around the Mediterranean Sea (ranging from 13% in North African Mozabites to 37.5% in Sardinians), decreasing to between 15% and 24% in more distant European areas. Elsewhere it is a lower frequency haplotype: 4% in Mandenkas, 0.7% among Biaka Pygmies and between 2& and 12.2% in most of South West Asia.
It is found at "extremely low frequencies in three populations (Mongols, Taiwan Chinese, and Japanese) and could be the result of admixture or just a very low frequency in the region". Siberians do not carry the H2 haplotype nor do the North American Cheyenne or Pima (both from Arizona and Mexico).
In North America, where Y chromosome R is found among natives and this is explained as caused by admixture with English, Spanish and French colonization, this H2 haplogrop is absent. This clearly shows that it was not introduced by the Europeans otherwise it would have admixed just the same way as R haplogroup did (unless R hg is not European, and entered America as one of the founding Haplogroups during the peopling of the New World, via Asia).
Amerindians' frequencies are around 6.4 % (Mayan and Quechua) dropping to 1.1% (Rondonian Surui) and zero among Ticuna, Guihiba and Karitiana. The presence among South American Natives is, according to the paper, due to historical European admixture.
I differ: if Europeans (i.e. Portuguese or Spanish Conquistadors mixing with the natives) were the source, we would find it uniformly spread out across the former Spanish and Portuguese colonies in America. Instead it is absent in Mexico, SW USA, Mesoamerica (excluding the Maya), Colombia and Venezuela, Argentina and Chile, Brazil...
Furthermore, only certain native groups carry it, others do not. This is a clear indication of its ancient origin and survival among Native Americans prior to the arrival of Europeans.
The map shown below depicts the global range and frequencies. Notice the absence of H2 hg in Siberia, the purported ancestral home of Amerindians.
The Neanderthal connection
Donnelly's paper tries to identify the place where this H2 haplogroup inversion took place. The authors propose three scenarios:
- African scenario, with an Eastern or Central African origin and its dispersion in an OOA event.
- Southwest Asian origin, in this case it flowed back into Africa explaining its current presence there.
- Neanderthal origin. They point out that this hypothesis was first suggested by Hardy et al.,  and it explains a SW Eurasian origin since that was the homeland of Neanderthals and also the highest global frequencies which are found in the Middle East.
Hardy et al. had written: "We suggest that a plausible explanation for the ingress of the H2 haplotype is that an ancestral H. neanderthalensis allele, [...] entered the European H. sapiens genome during the period of co-existence, and has spread through selection pressure to its current allele frequency of approx. 25% in this population " .
However Donnelly et al. are cautious and do not favor this scenario:"Though we think that this scenario is unlikely, more evidence is needed with respect to Neanderthal’s genetic contribution to modern humans before any strong conclusion can be made. Once again, this scenario can best be addressed once the Neanderthal genome is completed." .
The recent sequencing of the Neanderthal genes has now allowed us to settle this issue definitively:
No Neanderthal or Denisovan H2
The remark about high Middle Eastern frequencies is interesting, especially its peak of 31% among the Druze who, incidentally are believed to be the source of the mtDNA X2 haplogroup mentioned in my previous post as a potentially Neanderthal mtNA, or, (As I will explain in another future post, probably the mtDNA of some even older hominin, below I also hint at this possibility).
The time frame proposed by Donelly et al., favors a possible Neanderthal admixture because it spans a wide period: 16,400 to 108,400 years BP, plenty of time for Neanderthal - human admixture. 
Fortunately for us all, two years later (2012), Setó-Salvia et al.,  used the Neanderthal and Denisovan gene sequences to study the matter and discovered that:
"both hominins harbored the H1j variant [which] represents ~ 1.7% of all possible subhaplotypes in present-day humans of European ancestry" 
This finding disqualifies both a Neanderthal or a Denisovan origin for the human H2 haplogroup and the authors conclude that the inversion to H2 arose within the modern human lineage.
Interestingly, the H1j that those hominins carry, has not been correlated to neurodegenerative diseases in modern humans and the paper suggests that H1j introgressed into us due to admixture with Neanderthals or Denisovans.
The paper points leaves the door open to finding H2 in our ancestors because the "data from both Neandertals and Denisova are scarce, we cannot discard the existence of H2 chromosome carriers" among both hominins.  I was disappointed, I really believed that our H2 came from Neanderthals.
How about an even more archaic Homo as the source of H2?
But I have not lost hope: Setó-Salvia et al., found "68 intronic variants within the MAPT gene, 23 exclusive to Denisova hominin, 6 limited to Neandertals, and 24 exclusive to present-day humans." . This is a lot of variation and that requires time to develop, H1 must have diverged from the ancestral H2 longer ago than the 16.4 - 108.4 kya suggested by Donnelly.  And this may mean that H2 is indeed an ancestral form, found in an even older ancestor that made it out of Africa into SW Asia and Europe: Homo erectus.
Both Donnelly and Setó-Salvia cite previous work (Stefansson et al., 2005)  which gave an even older date for the inversion of the ancestral H2 form to the H1 found currently in modern humans (and also in the othere extinct branch of Neanderthals and Denisovans).
Steffansson set the inversion H1 into H2 at ~2.3 million years ago. And this was long before there were any modern humans, Neanderthals or Denisovans around.
However Donnelly et al., dismissed the date because the H2 lineage found nowadays in humans is very homogeneous while H1 is very diverse, so they suggest that H2 is younger since and had less time to evolve and diversify. But this may not be the case since H2 offers some reproductive advantages (enhanced fertility) as well as none of the negative neurological aspects provoked by the H1 haplogroup. The fact that it is uniform may indicate positive natural selection at work, or a strong genetic drift among modern humans after they admixed with an H2 carrying ancestor.
The image below shows this new scenario of admixture with an even older ancestor:
Based on the date 2.3 Mya, there are not many options to choose our ancestor from: either H. habilis or H. erectus. Both left Africa, the former went on to the Caucasus and were probably ancestral to H. erectus in Asia and later Denisovans. The latter spread out across South East Asia and reached China where it has survived until quite recent times.
The lack of H2 among humans in the H. erectus territories in Indonesia, China and S.E. Asia may mean that our vector is H. habilis, from Dmanisi in Georgia. Or it may mean that the humans that introgressed with them did not later move on into Southeastern and Eastern Asia. But did march on into America or did the H. sapiens that reached America pick up the H2 haplotype there, from a first peopling wave of H. habilis?
This could be settled with a gene sequence from H. erectus or H. habilis. Perhaps one day it will be done.
 Michael P. Donnelly, et al., (2010). The Distribution and Most Recent Common Ancestor of the 17q21 Inversion in Humans. The American Journal of Human Genetics 86, 161–171, February 12, 2010. doi: 10.1016/j.ajhg.2010.01.007
 Hardy, J., et al., (2005). Evidence suggesting that Homo neanderthalensis contributed the H2 MAPT haplotype to Homo sapiens. Biochem. Soc. Trans. 33, 582–585.
 MAPT microtubule-associated protein tau [ Homo sapiens (human) ] Gene ID: 4137, updated on 8-Jun-2014
 Setó-Salvia, Núria et al., (2012). Using the Neandertal and Denisova Genetic Data to Understand the Common MAPT 17q21 Inversion in Modern Humans. Human Biology: Vol. 84: Iss. 6, Article 1
 Stefansson, H., et al., (2005). A common inversion under selection in Europeans. Nat. Genet. 37, 129–137
Patagonian Monsters - Cryptozoology, Myths & legends in Patagonia Copyright 2009-2014 by Austin Whittall ©