Archive for the ‘Denisova’ Category

Expanding Hybrids And The Rise Of Our Genetic Common Denominator

September 29, 2012 10 comments

With the attestation of Neanderthal and Denisova DNA in the human genome, and strong indications for the genetic contribution of also other archaic hominines previously considered ‘extinct without issue’, the simple model of prefabricated, homogenous modern humans that moved ‘out-of-Africa’ to replace the human evolutionary residue all over the world in a single blow, failed dramatically. Now, the scientific community is literally forced to pay attention to decades of accumulated counter-evidence and criticism.

Average Joe replacing Average Joe: A new genetic common denominator rises up from the earth, combining the best of all available biological elements to adapt to a new way of life. Average Joe is a Hybrid!

One issue concerns the implied range expansions of a single ‘bottlenecked’, homogenous population that extended its African habitat to entirely new environments and climates. This should have been attested by selective sweeps in the genome – but it doesn’t.

While most people assumed that the out-of-Africa expansion had been characterized by a series of adaptations to new environments leading to recurrent selective sweeps, our genome actually contains little trace of recent complete sweeps and the genetic differentiation of human population has been very progressive over time, probably without major adaptive episodes.
[…] if some introgressed genes were really advantageous, they should have spread and fixed in the human population, but […] there is no widespread signature of strong selective sweeps in Eurasia. (Alves et al., 2012)

Selective sweep can be recognized by a large reduction of genetic variation near a favorable gene on the chromosome, caused by a quick expansion within a population of the gene by natural selection. Only a few complete sweeps and near-complete sweeps could be found, ‘suggesting that there was relatively little directional adaptive evolution associated with the “origin of modern humans.” Measuring by genetic change, agriculture was many times more important than the appearance of modern humans throughout the world’ (Hawks, 2012-07-20). Does this imply that genetic change of modern humans was predominantly not the result of sudden adaptive mutations? Possibly humans acquired their genetic adaptations to their respective extant environments in a different way:

[…] there are precious few genetic changes shared by all (or even most) humans today, that are not also shared with Neandertals, Denisovans, or plausible other archaic human groups (such as archaic Africans).
That of course follows from the fact that a fraction of today’s gene pool actually comes from those ancient groups. Their variation is (by and large) human variation. (Hawks, 2012-07-20)

Apparently, there was a host of archaic hominines out there, previously considered the evolutionary ‘dead ends’ from all over the world, whose traces can still be perceived as superimposed variability in the modern human genome. That is, up to now investigation on archaic admixture is mainly focused on the differences between modern populations, that increasingly emerge as the relicts of intense ‘archaic’ hybridization processes. ‘Neandertals and Denisovans fall within the variation observed for human nuclear sequences. Thus, only few fixed differences can be identified’ (Meyer et al., 2012). This way, out of 3.2 billion sequenced Neanderthal base pairs only about 600 Mb could be unambiguously attributed to Neanderthal introgression, what is low in comparison with Meyer’s estimation that 6.0% of the genomes of present-day Papuans derive from Denisovans. But, archaic hominines also shared a considerable genetic common denominator with modern humans, whose possible incorporation remains poorly analyzed. A variable portion of archaic DNA actually being shared with modern humans could affect the observed magnitude of introgression, but earlier assertions that Denisovans were indeed more divergent were never confirmed. Without clear traces of selective sweep, the true origin of ambiguously shared and distinctly archaic portions of the genome are impossible to tell apart. Reported examples of selective sweep remain rare:

We also identify over 100 Neandertal-derived alleles that are likely to have been the target of selection since introgression. One of these has a frequency of about 85% in Europe and overlaps CLOCK, a key gene in Circadian function in mammals. This gene has been found in other selection scans in Eurasian populations, but has never before been linked to Neandertal gene flow. (Sankararaman et al., 2012)

The circadian function refers to a chrono-biological adjustment to an external rhythm like daylight, what logically implies a genetic adaptation to the northern Neanderthal habitat with an exclusive advantage for northern populations. Such an introgressed innovation that apparently behaves like a new favorable mutation remains an exception, since hybrid incorporation and repatterning of whole chunks of introgressed DNA doesn’t require selective sweep.

These days many investigators try to reconstruct the past demography in their own way, though often the effort remains haunted by some remarkably conservative out-of-Africa assumptions. A wealth of newly published information on the subject is currently waiting for a proper interpretation to close the gap between our modern genome and the sequenced data retrieved from some of our ancestors. Conflicting perspectives often result in contradictory assertions that may be counter-intuitive and in need of reconciliation one with the other. The following investigation belongs to this category:

Science DOI: 10.1126/science.1224344

Meyer et al. – A High-Coverage Genome Sequence from an Archaic Denisovan Individual, 2012, link


We present a DNA library preparation method that has allowed us to reconstruct a high-coverage (30X) genome sequence of a Denisovan, an extinct relative of Neandertals. The quality of this genome allows a direct estimation of Denisovan heterozygosity, indicating that genetic diversity in these archaic hominines was extremely low. It also allows tentative dating of the specimen on the basis of “missing evolution” in its genome, detailed measurements of Denisovan and Neandertal admixture into present-day human populations, and the generation of a near-complete catalog of genetic changes that swept to high frequency in modern humans since their divergence from Denisovans.

This study has several interesting results worth mentioning: an extremely low genetic diversity of Denisova humans that can’t be observed in any modern population; the observation that Europeans have 24% less Neanderthal admixtures (“not being shared by Africans”) than Asians; and apparent indications of some hybridization event in the past, still noticeable in the chromosomes of all modern Denisovan descendents.

Low genetic diversity of Denisovans puts the asserted homogeneity of the modern human species in a new perspective. Despite all earlier speculation on an African bottleneck, designed to explain modern genetic homogeneity from a phylogenetic perspective, current genetic variability is now found to exceed the attested variability of ancient Denisovans for all modern ‘phyles’, or ethnicities:

Several methods indicate that the Denisovan hetero-zygosity is about 0.022%. This is ~20% of the heterozygosity seen in the Africans, ~26–33% of that in the Eurasians, and 36% of that in the Karitiana, a South American population with extremely low heterozygosity. Since we find no evidence for unusually long stretches of homozygosity in the Denisovan genome, this is not due to inbreeding among the immediate ancestors of the Denisovan individual. We thus conclude that genetic diversity of the population to which the Denisovan individual belonged was very low compared to present-day humans. (Meyer et al., 2012)

Nevertheless, Denisovans lack any relevant African affiliation. Their ‘phyle’ should have been separated long enough from the other branches of human evolution to have reached a genetic diversity comparable to Africans. Apparently, since this didn’t happen, genetic variability doesn’t simply translate to the age of an otherwise isolated population. Investigators may now dedicate their diligence to duplicate their calculations for a purported recent African bottleneck, and design a corresponding recent Denisovan bottleneck. Or they could just admit the geographic maxima of modern variability in Africa may rather represent archaic admixture than the age of a single human phyle.
Unfortunately, the Meyer study doesn’t mention the Neanderthal-like admixtures of Africans, except by saying that the ‘genetic contribution from Neandertal to the present-day human gene pool is present in all populations outside Africa’. It is important to keep this voluntary restriction in mind in reading their most remarkable assertion: ‘We estimate that the proportion of Neandertal ancestry in Europe is 24% lower than in eastern Asia and South America’ (Meyer et al., 2012).
This runs contrary to more detailed genetic analyses that previously revealed slightly higher levels of Neanderthal admixture in Europe. John Hawks counted derived SNP alleles of the 1000 Genomes Project being shared with the (Neanderthal) Vindija Vi33.16 genome, and found the surpluses in Europe and East Asia where rather comparable:

The Europeans average a bit more Neandertal than Asians. The within-population differences between individuals are large, and constitute noise as far as our comparisons between populations are concerned. At present, we can take as a hypothesis that Europeans have more Neandertal ancestry than Asians. If this is true, we can further guess that Europeans may have mixed with Neandertals as they moved into Europe, constituting a second process of population mixture beyond that shared by European and Asian ancestors. (Hawks, 2012-02-08 )

Unfortunately, the attested agreement between non-African and Neandertal genomes, and between Melanesian and Denisova genomes for that matter, didn’t result yet in the full identification of all specific genetic loci involved. Basically, the observed agreement was initially based on the differences between Africans and non-Africans in comparison with the archaic genome being investigated. Hence, the overall picture of archaic ‘differences’ may be distorted by shared components within the African reference group, that Meyer’s team didn’t include in their investigation and that Hawks didn’t quantify for his modern genomes that share derived SNP alleles with the (Neanderthal) Vindija Vi33.16 genome. In other words, this Neanderthal ancestry in Europe allegedly being 24% lower than Asia (according to Meyer et al.) is essentially meaningless without additional information that quantifies sharing:

My initial reaction to this difference is that it reflects the sharing of Neandertal genes in Africa. Meyer and colleagues filtered out alleles found in Africa, as a way of decreasing the effect of incomplete lineage sorting compared to introgression in their comparison. But if Africans have some gene flow from Neandertals, eliminating alleles found in Africans will create a bias in the comparison. If (as we think) some African populations have Neandertal gene flow, that probably came from West Asia or southern Europe. So as long as the present European and Asian (and Native American) samples have undergone a history of genetic drift, or if (as mentioned in the quote) they mixed with slightly different Neandertal populations, this bias will tend to make Asians look more Neandertal and Europeans less so.
Anyway, this demands further investigation. (Hawks, 2012-08-30)

Apparently, the legacy of the Out-of-Africa dogma caused Meyer et al. to take the African part for granted and just to look at the non-African part. We are lucky to have some additional information already at hand to more or less visualize how the Meyer et al. results could still be in tune with earlier results, that rather emphasized a closer match of Neanderthal admixtures with Europeans. The Austrian study of Hochreiter et al. (2012) actively incorporates the internationally shared Neanderthal and Denisova alleles in their calculations to measure the probability of uneven distribution (Fisher’s exact test) and to obtain the corresponding odds ratios, that give a symmetrical representation of the relative genetic enrichment for each type of admixture. From here on, all depends on how we perceive the human genome and what part of it we are willing to recognize as true Neanderthal or Denisova admixture, or something else.
Hochreiter’s study retrieved data from the Korean Personal Genome Project (KPGP) combined with those from the 1000-Genomes-Project:

Genotyping […] 1,131 individuals and 3.1 million single nucleotide variants (SNVs) on chromosome 1 […] identified 113,963 different rare haplotype clusters marked by tagSNVs that have a minor allele frequency of 5% or less. The rare haplotype clusters comprise 680,904 SNVs; that is 36.1% of the rare variants and 21.5% of all variants. The vast majority of 107,473 haplotype clusters contains Africans, while only 9,554 and 6,933 contain Europeans and Asians, respectively. (Hochreiter et al., 2012)

According to this data, only 6,490 (113,963 minus 107,473) of the rare haplotype clusters on chromosome 1 were exclusively non-African. The vast majority of all rare haplotypes, however, are shared with Africans one way or the other:

We characterized haplotypes by matching with archaic genomes. Haplotypes that match the Denisova or the Neandertal genome are significantly more often observed in Asians and Europeans. Interestingly, haplotypes matching the Denisova or the Neandertal genome are also found, in some cases exclusively, in Africans. Our findings indicate that the majority of rare haplotypes from chromosome 1 are ancient and are from times before humans migrated out of Africa. (Hochreiter et al., 2012)

The 9,554 and 6,933 European and Asian haplotypes thus per definition include a considerable overlap with extant African rare haplotypes. Moreover, the size of such an African overlap is proportional with the total count of shared Eurasian haplotypes. Mathematically it could be deduced that at the very least, 3,064 (ie. 9,554 minus 6,490) out of 9,554 ‘European’ haplotypes, and 443 out of 6,933 ‘Asian’ haplotypes should be also African. The maximum count of African rare haplotypes, however, that made it ‘Out-of Africa’ and are currently shared with non-Africans, remained well below 10%. Since over 90% of the African rare haplotypes are thus not shared with Neanderthals and Denisovans, in an Out-of-Africa scenario this would mean that a similar proportion of the European and Asian rare haplotypes could be expected to be non-Neanderthal and non-Denisova. Could we really rely on the ancestral origin of so many shared haplotypes? Just being shared African doesn’t make these haplotypes ancestral all of a sudden, and less without a proper quantification.
Let’s first try to quantify the potential Neanderthal admixtures a bit. Hu’s analyses of archaic segments should give an adequate peek inside the various admixtures for an educated guess:

Archaic hominin admixture with modern non-Africans was detected by genome wide analysis of Neanderthal and Denisovan individuals.
To gain better understanding in demographic and evolutionary significance of archaic hominin admixture, […] we identified 410,683 archaic segments in 909 non-African individuals with averaged segment length 83,460bp. In the genealogy of each archaic segment with Neanderthal, Denisovan, African and chimpanzee segments, 77~81% archaic segment coalesced first with Neanderthal, 4~8% coalesced first with Denisovan, and 14% coalesced first with neither (Hu et al., 2012)

Considering the above, apparently very few (or none?) of all the non-archaic haplotypes that made it out of Africa became rare. Such results, naturally, would imply one enormous problem about the construct Homo Sapiens Sapiens (HSS). Instead, the lack of rare haplotypes outside Africa that could be safely assigned unambiguously to what is generally considered the constituent forerunner of modern humans, indeed echoes much earlier claims of ancestral homogeneity. As already referred to above, population geneticist are very much acquainted with the concept of an early HSS bottleneck, since this was once designed to explain away all evidence of this kind. Hence, I appreciate the reasons why Hochreiter et al. prefer to consider the rare Denisova- and Neanderthal-like rare haplotypes in Africa ‘ancestral’, even those being exclusively African, but this preposition logically implies the existence of allegedly HSS ancestral haplotypes in Eurasia that are neither rare nor absent in Neanderthal and Denisova. Combined with the ever more unpopperian association of frequent haplotypes with HSS per definition, it has now all appearance Homo Sapiens Sapiens is nothing but the current genetic common denominator in disguise.
As for now, apparently the Neanderthal admixtures indeed account for the greater part of the Eurasian archaic components. The discovery of the Denisova component was just mere luck, and the odds are high that more archaic hominines contributed to the Eurasian admixtures. For all we know, on the eve of the transition to modern humans Europe was only inhabited by Neanderthal. However, the likelihood of additional archaic admixture in South East Asia are being widely discussed. Moreover, Hu’s results almost exclude the possibility that substantial African archaic admixtures, at least those not yet being fully incorporated in the ‘bottlenecked’ HSS population, expanded out of Africa. Altogether, it wouldn’t be farfetched to consider most of the 14% Eurasian remainder to be essentially archaic Asian. Actually, Hu’s 14% Eurasian admixtures currently unaccounted for could easily correspond to the genetic contribution of up to four Asian archaic hominines like Denisovan’s – wherever those may have existed in isolation from Denisovan-like populations that – as for now – potentially inhabited the large geographic stretch between their attested remains in the Altai mountains and their attested genetic contributions in Melanesia. Mendez et al. (2012) suggested ‘that the archaic ancestor contributing the deep lineage to Melanesians and the specimen from Denisova were members of genetically differentiated populations’, what indeed should make us wonder about the Asian location, or nature, of such unsampled hominine groups we are still missing from the record of potential archaic admixtures. Even locally admixted homo erectus have already been proposed.
Now, Hu’s fixed non-Neanderthal-non-Denisovan remainder and the ambiguous 4% apparently shared component between both sampled hominines, suggest ~40-50% ancestral overlap between Neanderthals and Denisovans for the admixtures attributed to Denisovans, against only ~4-5% ancestral overlap for Neanderthal-like admixtures. The unambiguous Denisovan component left may be considered fully Asian in origin, even though Meyer et al report an opposite effect on the current availability of Denisovan alleles all over the world:

Interestingly, we find that Denisovans share more alleles with the three populations from eastern Asia and South America (Dai, Han, and Karitiana) than with the two European populations (French and Sardinian) (Z = 5.3). However, this does not appear to be due to Denisovan gene flow into the ancestors of present-day Asians, since the excess archaic material is more closely related to Neandertals than to Denisovans (Meyer et al., 2012)

Indeed, the contribution from Denisovans is found ‘almost’ exclusively in island Southeast Asia and Oceania. Hence, Meyer’s assumption this effect is directly related to a higher proportion of archaic Neanderthal alleles in Asia justifies a ‘worse case’ scenario, where the ‘true’ Asian share could probably increase to 18-19%, against up to 81% rare archaic haplotypes that could now be tentatively attributed to essentially Eurasian Neanderthal admixtures. For now we are only interested in the counts of Neanderthal-like admixtures, so we could propose a conversion of Hochreiter’s rare haplotype counts results, that reduces the non-African count of rare haplotypes to ~5,224 Neanderthal non-African haplotypes, and that reduces the ‘non-exclusive Asian’ haplotypes to ~5,581 Neanderthal non-exclusive Asian haplotypes, while the same maximum of Hochreiter’s 9,554 haplotypes could still be assumed to be both ‘Neanderthal’ and ‘non-exclusive European’.
For sure, such an increased proportion for Neanderthal-like admixtures in Europe doesn’t make Meyer’s results more intuitive. All the contrary, Meyer’s 24% lower European contribution should make us wonder where the differences went to. Apparently, a changed proportion of non-African Neanderthal-like admixtures in Europe compared to Asia needs proportional compensation elsewhere. Unfortunately, this effect has not been illustrated in any of the studies that aim to quantify Neanderthal admixtures one way or the other.

Different scenarios based on Hochreiter’s rare haplotypes, Hu’s proportions of archaic contribution, and the following viable assumptions: Zero Eurasian and Afro-Asian components for scenario #1; Zero Eurasian and Afro-Eurasian components for scenario #2; Zero European and Afro-Eurasian components for scenario #7. The Meyer-scenarios assume a 24% lower non-African European component than Asian, while #1 and #2 were also calculated for equal shares. Lower Neanderthal-like proportions for Europe in comparison with Asian apparently imply a higher count for shared Afro-European haplotypes. The invariable high Afro-European component and less relevant Eurasian, Afro-Asian and Afro-European components are in support of an underpinning Eurasian substructure for the Neanderthal admixed population; and a massive expansion of already admixed European populations into Africa.

For a better comprehension I elaborated several possible solutions, combining the information of Hochreiter, Meyer and Hu. Hochreiter supplied values for three linear equations that involve six variables, representing the rare haplotype counts characterized as ‘exclusive European’, ‘exclusive Asian’, ‘Eurasian’, ‘Afro-Asian’, ‘Afro-European’ and ‘Afro-Eurasian’ . Meyer’s published proportion between European and Asian haplotypes introduces a fourth equation, that for comparison could be alternated with a more intuitive scenario that has non-African European and Asian rare haplotypes evenly distributed. However, a set of linear equations may only be solved (but not necessarily) if the number of equations is the same as the number of variables. Thus two variables remain undefined, what means that an array of solutions is still possible. I worked out a number of different scenarios, each based on two additional assumptions that are necessary to solve the equations. Thus, for scenario #1 I choose zero values for the Eurasian and Afro-Asian components; for scenario #2 I choose zero values for the Eurasian and Afro-Eurasian components; for scenario #3 I kept the Eurasian and Afro-European components on zero; for scenario #4 the same for the Afro-Asian and Afro-Eurasian components; for scenario #5 the Afro-European and Afro-Eurasian components were kept zero; for scenario #6 the same for the Afro-European and Afro-Asian components; and for scenario #7 zero values were assumed for the Afro-Eurasian and European component, the latter being valid only for the Meyer variant of the equations.
Scenarios #3-5 can’t be solved for natural values and scenario #6 is ambiguous. The remaining scenarios #1, #2 and #7 all show the predominance of shared Afro-European rare haplotypes, while Afro-Asian, Eurasian and Afro-Eurasian components are lower and not always required for a valid result. The effect of Meyer’s result can be illustrated for scenarios #1 and #2, where lower Neanderthal-like proportions for Europe in comparison with Asian apparently imply a higher count for shared Afro-European haplotypes and lower counts for Afro-Asian and Afro-Eurasian haplotypes.
These scenarios reveal the Afro-Asian component as fairly irrelevant, and the Afro-Eurasian component emerges as moderately weak. Only the Afro-European component remains definitely prominent in all scenarios. Remarkably, simulations that increase the Eurasian shared component are directly proportional to increases of the Afro-European component, while both are inversely proportional to the Afro-Eurasian component. This behavior supports the hypotheses that the Afro-Eurasian shared component is only moderately present; that at least the Asian Neanderthal admixtures don’t share any African origin or association in particular; and that Neanderthal haplotypes rather seem to have expanded proportionally into Africa and Asia alike from a European center. Especially the increased Afro-European component is remarkable, since an ancestral origin results problematic for rare haplogroups that feature a structural deficit in Asia.
At this stage it is impossible altogether to distinguish between haplotypes that introgressed through Neanderthal admixtures and those that may be ‘safely’ regarded ancestral to both Neanderthal and modern humans – so we should refrain from doing so beforehand. How ancestral the shared African haplotypes could possibly be? African substructure is no longer viable as a major explanation of Neanderthal admixtures in Eurasia. Actually, African substructure was already contradictory with the earlier Out-of-Africa bottleneck-and-homogeneity paradigms, and an additional west-to-east substructure to explain essentially different admixture patterns for Europe and Asia, is even less conceivable. Instead, ‘recent admixture with Neanderthals accounts for the greater similarity of Neanderthals to non-Africans than Africans’ (Yang et al., 2012). Less exclusive scenarios, that allow for early admixture events in the ancient Near Eastern contact zone, aren’t any less problematic for the discrepancy and leave the much lower Afro-Asian component without explanation. The most progressive and intuitive Out-of-Africa scenario, that considers the predominantly European distribution of Neanderthal haplotypes and predicts an increased admixture rate in Europe, now results falsified by this closer examination of Meyer’s 24% lower European rate. Apparently, the current distribution of admixtures only appeared to be in favor of any overall Out-of-Africa framework. The apparent lack of shared Afro-Asian haplotypes doesn’t indicate an African route for Asian admixtures, nor does the low count of shared Afro-Eurasian haplotypes advocate the importance of an ancestral component. Instead, an underpinning West-East dichotomy or Eurasian substructure already in place for the Neanderthal population before the attested admixture has already been proposed as a valid explanation:

Europeans and Asians could show distinct components of Neanderthal admixture if they had admixed with European and central Asian Neanderthals, respectively (Alves et al., 201)

The Afro-European shared haplotypes can’t be older than the long term genetic differentiation of Eurasian Neanderthals, what adds up to the already expounded rejection of African substructure in a recent Out-of-Africa scenario. A better explanation may be found in a massive expansion (or ‘backmigration’) of European populations into Africa, and a corresponding submersal of almost their full share of Neanderthal admixtures inside Africa subsequent to some late-Neanderthal admixture event.
Now the falsification of an important shared ancestral compenent in the African count of rare haplotypes becomes evident, Hochreiter’s data, reporting that ‘haplotypes matching the Denisova or the Neandertal genome are also found, in some cases exclusively, in Africans’, may be viewed in an entirely new perspective. If introgression of Denisovan admixtures was part of a rather ancient gene flow, albeit considerably younger than the Eurasian Neanderthal differentiation still noticeable in the strongly regionalized Neanderthal admixtures, some Denisovan alleles could have reached Africa contemporaneously with the ‘other’ archaic admixtures that arrived there through the European route, as displayed by the calculated haplotypes pattern above. Especially the world-wide distribution of shared Neanderthal-Denisova alleles raises some questions into this direction.
More detailed analysis on the immune gene OAS1, involved in ‘Denisovan’ introgression, revealed this gene was embedded in a very divergent string of DNA, referred at as the ‘deep lineage’ haplotype. Its divergence from all the other extant OAS1-related haplotypes was strong enough to exhibit the signature of archaic introgression, what means the haplotype ‘may have introgressed into the common ancestor of Denisova and Melanesians via admixture with an unsampled hominin group, such as Homo erectus’ (Mendez et al., 2012). The haplotype resembles the Denisovan haplotype ‘with the exception of one site (position 30504), at which the extant human carries the derived C and the Denisova specimen carries the ancestral T’, but even more striking is the current homogeneity of the deep lineage:

Broadly distributed throughout Melanesia, the deep lineage exhibits very low intraallelic diversity […], with an estimated TMRCA of ~25 kya (Mendez et al, 2012)

The attested Denisovan fossils in the Altai mountains had a slightly more ‘ancestral’ version of the gene, thus being different from the extant ‘deep lineage’. Actually, this unique signature boils down to a single hybridization event for this haplotype that involved one ancestral parent not unlike, but slightly different from the sampled specimen of Denisova Cave. Also Meyer’s observation that ‘Papuans share more alleles with the Denisovan genome on the autosomes than on the X chromosome’ and that eg. on chromosome 11 Denisovan ancestry is estimated to be lower in Papuans than in East Eurasians, corroborate to this hypothesis:

[…] there is significant variability in Denisovan ancestry proportion compared with the genome-wide average not just on chromosome X, but also on individual autosomes that have estimates that are also lower (or higher) than the genomewide average. (Meyer et al., 2012 sup)

Unfortunately, despite the negative evidence accumulated by Meyer et al. in their supplement against their own sex-biased modern population-history pet theories, their main article stopped short of dwelling on far more interesting factors such as hybrid chromosome repatterning that include ‘natural selection against hybrid incompatibility alleles, which are known to be concentrated on chromosome X’ and a marked uneven distribution of Denisovan ancestry also in the autosomes.

With hybrids, crossover events tend to compromise lineage specific regulatory regions on the chromosomes. Only favorable repatterning of the hybrid chromosomes results in viable offspring.

The disproportionate absence of Denisovan admixtures on the X chromosome virtually excludes a sex-biased demographic history in Oceania as an explanation and indeed, in their supplement Meyer et al. elaborated a potential rejection on logical grounds: also migrating males bring in their share of X chromosomes, so this way it can’t just disappear. A removal of Denisovan chromosome X by natural selection after the gene flow can be excluded as well: selection acting on genomic functional elements can be detected by its indirect effects on population diversity at linked neutral sites (McVickers et al., 2009), but Meyer’s team was right that they couldn’t establish that archaic ancestry was affected by the proximity to genes. However, natural selection against hybrid incompatibility alleles is still a poorly understood process – and especially if considered applicable just to the protein coding genes that constitute only about 3 percent of the human genome. This year the ENCODE Project Consortium confirmed actually over 80% of the genome to be involved in biochemical functions, in particular outside of the protein-coding regions. Genetic viability is most of all determined by the proper regulation of gene expression. Hence, much of the genome is considered constrained by biological constraints against evolutionary change. Of interest are the ‘large number of elements without mammalian constraint, between 17% and 90% for transcription-factor-binding regions as well as DHSs and FAIRE regions’ (Dunham et al., 2012), referring to regions linked to regulatory functions. But even here, the autors hold that depressed derived allele frequencies indicate ‘an appreciable proportion of the unconstrained elements are lineage-specific elements required for organismal function, consistent with long-standing views of recent evolution’.
It should be obvious that nature can’t expect much viable offspring from a fusion of gametes that brings together lineage specific regulatory regions in a random fashion. The deleterious effects of random hybrid recombination appear to be inversely proportional to chromosome crossover events during meiosis, that normally happens once for each generation. First generation hybrid offspring typically has enough directly inherited consistency of their regulatory regions on their genome left for being viable. But next generation chromosome crossover may already affect the regulatory processes of the haploid gametes being produced by meiosis. Initially, this seriously affects fertility and only the sheer scale of gamete production may compensate for the high probability of next-generation hybrid malfunction. This close relation between hybrid viability and a limited array of favorable crossover events, that shouldn’t compromise the regulatory functionality of the hybrid genome, apparently resulted also in a marked variability of ancestry proportions for each chromosome across the genome of Denisovan admixed populations.
For hybrids, selective processes are more efficient when directed at regulatory viability just before and during conception. Post-natal fitness, on the other hand, is most of all based on the ‘proven technology’ of coding genes whose selective advantage and usefulness were already attested in the parent species. Natural selection based on the success of coding genes thus may have been of less importance in recent hominine evolution than could be expected for the profound genetic change modern humans apparently went through. This detail can indeed be confirmed in the modern genome by the above mentioned lack of genetic sweep, despite important repatterning and recent genetic innovation due exactly to the occurrence of abundant hybridization in recent human evolution.
For the moment this issue should be considered isolated from the origin of the shared Neanderthal-Denisovan haplotypes, especially since this portion was already in place for the sampled Neanderthal and Denisovan specimen. Tentatively, this shared portion could be attributed to an earlier ‘bi-directional’ gene flow, leaving the specific Denisovan admixtures in Melanesia apparently to a subsequent hybridization event that only seems to have affected modern populations. The proposal above of a single hybridization event virtually excludes a scenario where the hybrid population could be considered firmly rooted in a local archaic population. Naturally, this runs counter to an array of earlier proposals that rather link Denisovan admixture events with a wide geographic range of Denisovan hominine presence between the Altai mountains and SE Asia (Reich); with different places during the migration of modern humans (Rasmussen); with distinct Denisovan admixture events in Oceanians and East Asians (Skoglund and Jakobsson); or with a process of continuous admixture where migration routes overlap with archaic hominine ranges (Currat and Excoffier).
However, a single late-Denisovan hybridization event doesn’t suffice as an exclusive scenario in the light of new evidence that posits Denisova Cave as a hotbed of Neanderthal contact. Abundant remains of Neanderthal were found nearby the cave:

The Chagyrskaya 6 mandible […] allows us now to link this material morphologically as well to the Neanderthals in Western Eurasia. Several questions remain: the timing and extent of Neanderthal expansion into the Altai, and especially the potential coexistence and interaction between Neanderthals and Denisovans. Based on availabe dates, the Neanderthals in Okladnikov cave and the possibly slightly earlier Chagyrskaya remains overlap with the wide range of dates for Layer 11 of Denisova cave. (Viola et al., 2012)

Both species even shared the same cave:

we have determined a high-quality nuclear genome from a pedal phalanx found in Denisova Cave in 2010. We show that the pedal phalanx derived from a Neandertal and thus that Neandertals as well as Denisovans have been present in the cave. (Sawyer et al., 2012)

Extensive contacts should at least have initiated a kind of fusion between the Neanderthal and Denisovan parent species into a single population where in time, due to multiple hybridization events, the variability of introgressed DNA would have been restored and integrated, into the Neanderthal genetic heritage and vice-versa. Hybrid repattering of admixed chromosomes probably wouldn’t have raised Denisovan heterozygosity beyond the elevated levels observed in modern populations, and less given the outstanding native homozygosity of the sampled Denisovans as a starting point. Indeed, the Denisovan sample has a reduced heterozygosity compared to any of the present-day humans analyzed by Meyer et al, though they reported the relative ratios of heterozygosity as fairly constant, what could be considered problematic for the assumption of archaic Neanderthal admixture already present in the shared DNA with Denisovans. However, 29 coding CCDS genes could be identified with more than one fixed non-synonymous SNC where ‘Denisova’ carries the ancestral allele, while in eight of these (OR2H1, MUC17, TNFRSF10D, MUC6, MUC5B, OR4A16, OR9G1, ERCC5), the Denisovan individual appeared heterozygous for all SNCs present in the gene. In table S44 it can be verified that 37% of this heterozygosity can be found in chromosome 11, 13% in chromosomes 6 and 7, and 11% in chromosome 8, while eleven chromosomes are homozygous for all investigated genes. Though Meyer’s team proposes this to be the result of duplications or repetitive regions, this heteromorph signature basically leaves the possibility of hybridization more than open. Since this results focus on the fixed non-synonymous SNC where Denisova carries the ancestral allele and modern humans the derived allele, the documented non-ancestral polymorphisms of Denisova even resemble modern-like admixtures.
According to Dienekes, within the group of polymorphic Eurasian SNPs there are less Denisovan than Neanderthal SNPs that are also monomorphic in the African Mbuti Pygmys. Because, actually the relation between Denisovans and Pygmy is ancestral and they share ancestral SNPs. This might reflect a lower penetration into Africa of the shared Denisovan-Neanderthal portion of archaic admixtures. If so, this could be partly due to a Denisovan origin of this shared portion, though some degree of Eurasian Neanderthal substructure may be involved as well. Interestingly, this also implies the tentative modern-like part of potentially admixtured (ie. polymorphic) SNPs in the Denisovan genome is actually less ‘African’ than the Out-of-Africa hypothesis should be happy with. If anything, despite some modern-like features, those admixtures should be assumed of eastern Neanderthal origin.

Proposed hybrid flows for various archaic components, superimposed on Denisovan related HLA-A gene distribution. Blue arrows carry Neanderthal admixtures; red-blue arrows carry mixed Neanderthal-Denisovan admixtures; the green arrow represents East African input; and brown arrows represent Asian admixture and reflux. Question marks represent unknown archaic hominines that may have contributed locally and possibly also to the common genetic denominator of modern humans.

Another modern-like feature of the potential Denisovan-Neanderthal hybridization is the above mentioned outstanding heterozygosity of chromosome 11, that almost screams for continuity with the hybrid signature of the same chromosome in Papuans, where Denisovan ancestry is strikingly low. I really wonder what could have been the impact of these hybrid changes for the now widespread Denisovan-like mtDNA segment inserted into this same chromosome 11 of modern human populations all over the world, as described in a previous post; and what could have been the link with the survival of insert-like mtDNA in the aboriginal DNA found in the Lake Mungo 3 remains (LM3) dated 40kya. But even a much lesser extend of any gradual continuation with respect to Denisova-related selective processes against hybrid incompatibility alleles would only make sense if modern populations are themselves the continuation of these same ancient hybridization processes. Apart from what this might imply for the very nature of the modern genome in general, we could at least incorporate this signature of a continuous hybridization process, that tentatively links the Altai Mountains with Oceania, in what we know about the current distribution of Denisovan admixtures. A northern route around the SE Asiatic habitat of probably very different archaic hominine populations, some of them possibly more erectus-like or even more habilis-like, such as Homo floresiensis (Argue et al., 2012), seems at present more likely than a straightforward direct southern route:

However, in contrast to a recent study proposing more allele sharing between Denisova and populations from southern China, such as the Dai, than with populations from northern China, such as the Han, we find less Denisovan allele sharing with the Dai than with the Han (Meyer et al., 2012)

Current evidence even seems to favor a specific Korean route before turning south along the Chinese coasts down to Oceania:

The enrichment of Neandertal haplotypes in Koreans (odds ratio 10.6 of Fisher’s exact test) is not as high as for Han Chinese from Beijing, Han Chinese from South, and Japanese (odds ratios 23.9, 19.1, 22.7 of Fisher’s exact test) – see also Figure 7. In contrast to these results, the enrichment of Denisova haplotypes in Koreans (odds ratio 36.7 of Fisher’s exact test) is is higher than for Han Chinese from Beijing, Han Chinese from South, and Japanese (odds ratios 7.6, 6.9, 7.0 of Fisher’s exact test) (hochreiter et al., 2012)

It has been suggested that ancient HLA-A genes of the primate immune system only survived on human chromosome 6 by balanced selection in the Denisovan lineage. Hence, the current geographical distribution of this genes is often taken as indicative for the wherabouts of Denisovan admixed descendents. As you can see on the attached map, the hybrid Denisovan trail described above corresponds fairly well with this view, except for Yunnan and Tibet where possible Denisovan-like admixtures are largely below detection level and certainly not derived from Melanesian arrivals. It can’t be excluded that here we may find the root origin of the archaic population whose remains were so far attested only in Denisova Cave. Interestingly, this hypothetized ultimate origin of the archaic Denisovan population is adjacent to Indo-China, where hominine evolution may have been as old and divergent as in Africa.
Thus, it becomes ever more difficult to identify with, or deny descendance of a particular hominine branch. Recent human evolution is like a snowball rolling down the hill. What we are is just everything what came down from the hill, and what didn’t stop rolling. We may question our ability to really take a turn since for all we know the ball just gathers more snow and increases momentum. We can’t even say our trail downhill was human all the way, or where it started, or define what sets the participants apart from everything else around it. But here we are, something completely new on the face of the earth. And most of it, we have in common.


  • Adcock et al. – Mitochondrial DNA sequences in ancient Australians: Implications for modern human origins, 2001, link
  • Alves et al. – Genomic Data Reveal a Complex Making of Humans, 2012, link
  • Argue et al. – An hypothesis for the phylogenetic position of Homo floresiensis, European Society for the study of Human Evolution 2012 meeting, link
  • Dienekes – A surprising link between Africans and Denisovans, Blog September 27, 2012, link
  • Dunham et al. – An integrated encyclopedia of DNA elements in the human genome, 2012, The ENCODE Project Consortium, link; Online review Universität Heidelberg: Allegedly Useless Parts of the Human Genome Fulfil Regulatory Tasks, 6 September 2012, link
  • Cox et al. – Testing for Archaic Hominin Admixture on the X Chromosome: Model Likelihoods for the Modern Human RRM2P4 Region From Summaries of Genealogical Topology Under the Structured Coalescent, 2008, link
  • Curnoe et al. – Human Remains from the Pleistocene-Holocene Transition of Southwest China Suggest a Complex Evolutionary History for East Asians, 2012, link
  • Hawks – Denisova at high coverage, Blog 2012-08-30, link
  • Hawks – Which population in the 1000 Genomes Project samples has the most Neandertal similarity?, Blog 2012-02-08, link
  • Hawks – Modern humans in with a whimper, Blog 2012-07-20, link
  • Hochreiter et al. – Rare Haplotypes in the Korean Population, at ASHG 2012, link
  • Hu et al. – Analysis of contributions of archaic genome and their functions in modern non-Africans, at ASHG 2012
  • McVicker et al. – Widespread Genomic Signatures of Natural Selection in Hominid Evolution, 2009, link
  • Mendez et al. – Global genetic variation at OAS1 provides evidence of archaic admixture in Melanesian populations, 2012, link, or try here
  • Meyer et al. – A High-Coverage Genome Sequence from an Archaic Denisovan Individual, 2012, link, supplement
  • Sankararaman et al. – A genomewide map of Neandertal ancestry in modern humans, at ASHG 2012
  • Sawyer et al. – Neandertal and Denisovan Genomes from the Altai, European Society for the study of Human Evolution 2012 meeting, link
  • Scally et al. – Revising the human mutation rate: implications for understanding human evolution, 2012, link
  • Viola et al. – A Neanderthal mandible fragment from Chagyrskaya Cave (Altai Mountains, Russian Federation), European Society for the study of Human Evolution 2012 meeting, link
  • Yang et al. – Ancient Structure in Africa Unlikely to Explain Neanderthal and Non-African Genetic Similarity, 2012, link
  • Yotova et al. – An X-linked haplotype of Neandertal origin is present among all non-African populations, 2011, link

Denisova Cave and the Mystery of the mtDNA Phylogenetic Tree

March 27, 2010 32 comments

Nobody expected a great surprise. Genetic testing of the little finger of an early hominin child found in the Siberian Denisova Cave, Kostenki, in the middle of archeological remains pertaining to Upper Paleolithic culture, would almost for sure confirm DNA similar to ours. There was a slim change that the pinky belonged to a Neanderthal from the neighborhood that got lost, but everything pointed at a an unequivocal member of the advanced group of hominins responsible for introducing symbolic art all over the world, the so-called anatomically modern humans (AMH).

The collection of personal adornments and artifacts suggestive of symbolic behavior from the Early Upper Paleolithic deposits of Denisova Cave, Altai, is one of the earliest and the most representative of the Upper Paleolithic assemblages from Northern and Central Asia. Especially important is a fragment of a bracelet of dark-green chloritolite, found near the entrance to the eastern gallery of the cave in the upper part of stratum 11. The estimated age of the associated deposits is ca 30 thousand years. According to use-wear and technological analysis, techniques applied for manufacturing the specimen included grinding on various abrasives, polishing with skin, and technologies that are unique for the Paleolithic – high-speed drilling and rasping. The high technological level evidences developed manual skills and advanced practices of the Upper Paleolithic cave dwellers. (Derevianko et al., 2008)

Humans spend more time gathering around the campfire to celebrate their victory on Nature, only to challenge evolution in an entirely new way.

Neanderthal were readily dismissed as potential authors of local Upper Paleolithic art, due to what boils down to a deep distrust against anything that would deem them capable of such a feat, and they were the only other early hominins around that we knew of – at least culturally speaking, since we don’t have much more than a little pinky after all. And indeed, the first genetic results showed the world was right about one thing: the little finger did not belong to a Neanderthal child. But nobody could have guessed how wrong the usual lot of junk scientists were about almost anything else. This was not the child from the same flesh and blood of modern humans, but a member of a previously unknown ancestral human subgroup.

Dr. Johannes Krause, of the Max Planck Institute in Germany, sequenced the entire mitochondrial DNA (mtDNA) genome and showed almost two times as many differences to modern human mtDNA as does Neanderthal mtDNA. You can find the genome at GenBank or EMBL using record ID FN673705 and check it out by yourself: Even Neanderthal was a close relative to modern humans compared to this hominin!

A phylogenetic analysis similarly shows that the Denisova hominin mtDNA lineage branches off well before the modern human and Neanderthal lineages (Fig. 3). Assuming an average divergence of human and chimpanzee mtDNAs of 6 million years ago, the date of the most recent common mtDNA ancestor shared by the Denisova hominin, Neanderthals and modern humans is approximately one million years ago (mean = 1,040,900 years ago; 779,300–1,313,500 years ago, 95% highest posterior density (HPD)), or twice as deep as the most recent common mtDNA ancestor of modern humans and Neanderthals (Krause, 2010)

Established paleo-anthropology is now faced with the challenge to rewrite the book of human evolution. And of course first things first, the dates were adjusted to make a better fit with pre-AMH cultures:

We note that the stratigraphy and indirect dates indicate that this individual lived between 30,000 and 50,000 years ago. At a similar time individuals carrying Neanderthal mtDNA were present less than 100 km away from Denisova Cave in the Altai Mountains, whereas the presence of an Upper Palaeolithic industry at some sites, such as Kara-Bom and Denisova, has been taken as evidence for the appearance of anatomically modern humans in the Altai before 40,000 years ago. (Krause et al., 2010)

Nobody has ever heard of pre-AMH bracelets, so let’s conveniently forget for a while about that fragment of a polished bracelet with a drilled hole, that was found earlier in the same layer that yielded the bone. Is it possible that here we have evidence that points to a third species, next to Neanderthal and AMH? A species, that might have been as civilized as a AMH, or a beast our ancestors didn’t breed with, or anything else that didn’t involve “us” so we can understand? The publication of Krause carefully omitted this pressing question and the word went out that for sure Krause had already access to autosomal data that could explain why. That Denisova child might have been anything but a Yeti.

Sure, mtDNA doesn’t make a species, no matter how different it may be from modern humans. There was no need for Krause to mention this. But divergence of mtDNA lineages has been taken as an indication of divergent hominin developments before. Explicitly with respect to Neanderthal, whose attested and validated mtDNA lineage was deemed sufficiently homogeneous and different from ours to provoke a definite ordeal. However, now we have the Denisova mtDNA sample to teach us modesty. After all, there are lots of things about mtDNA that need better understanding before we can even attempt to solve the question of how the modern forms spread, and how they evolved.

All conspires against the notion that paleogenetic mtDNA of Neanderthal, and now even more so the mtDNA from Denisova Cave, might be the precursor of modern mtDNA. It couldn’t have evolved so rapidly to modern mtDNA. A study on 44,000-year-old remains of Adelie penguins in Antarctica even confirmed the potential overestimation of the mutational change that is used for dating mtDNA of paleogenetic samples. This stems from a bias that is caused by nonsynonymous mutations, involving notable coding changes that are potentially deleterious and most likely won’t persist very long due to natural selection. Accordingly, only a portion of the mutational changes can be observed over a longer period of time:

Rates of evolution of the mitochondrial genome are two to six times greater than those estimated from phylogenetic comparisons. Subramanian et al., 2009)

The investigation showed that only the effect of synonymous mutations (“silent mutations”) in the mtDNA genome, that involve coding synonyms for the same proteins, remain stable. To retrieve the phylogenetic dates only these “silent mutations” should be measured, ie. changes on coding genes that produce coding synonyms that won’t affect the function of the gene. Mutations that effectively change the functionality of a gene and thus are most likely to be (slightly) harmful, get lost over time, since such mutations would finally bring about the extinction of a lineage and thus shouldn’t count for calculating the age of surviving lineages. The mtDNA “molecular clock” thus should only involve properly identified “silent mutations”.
This results were also important for interpreting the paleogenetic mtDNA samples of hominins.

Mildly deleterious mutations initially contribute to the diversity of a population, but later they are selected against at high frequency and are eliminated eventually. Using over 1,500 complete human mitochondrial genomes along with those of Neanderthal and Chimpanzee, I provide empirical evidence for this prediction by tracing the footprints of natural selection over time. The results show a highly significant inverse relationship between the ratio of nonsynonymous-to-synonymous divergence (dN/dS) and the age of human haplogroups. Furthermore, this study suggests that slightly deleterious mutations constitute up to 80% of the mitochondrial amino acid replacement mutations detected in human populations and that over the last 500,000 years these mutations have been gradually removed. (Subramanian, 2009)

Interestingly, this dN/dS ratio among Neanderthal was initially reported strikingly high.

These results suggest that slightly deleterious amino acid variants segregate within populations, and that differences in the intensity of purifying selection may affect mtDNA dN/dS ratios. Previous estimates based on mean pairwise differences (MPD) within the mtDNA HVRI suggested that Neandertals (MPD = 5.5) had an effective population size similar to that of modern Europeans (MPD = 4.0) or Asians (MPD = 6.3), but lower than that of modern Africans (MPD = 8.1) (Krause et al., 2007b). Recent population genetic analyses have revealed a higher mtDNA amino acid substitution rate (Elson et al., 2004) and relatively more deleterious autosomal nuclear variants (Lohmueller et al., 2008) in Europeans than in Africans, presumably due to the smaller effective population size of Europeans. Thus, it seems plausible that Neandertals had a long-term effective population size smaller than that of modern humans. (Green et al., 2008)

However, the new information supplied by the Denisova hominin reveals this assumed feature of Neanderthal mtDNA was actually a mistake:

The 12 proteins encoded by the Denisova hominin mtDNA (excluding ND6, Supplementary Information) show low per-site rates of amino acid replacements (dN) when compared to the per-site rates of silent substitutions (dS), consistent with strong purifying selection influencing the evolution of the mitochondrial proteins (dN/dS=0.056). Notably, when the evolution of mitochondrial protein-coding genes in modern humans, Neanderthals, chimpanzees and bonobos is gauged in conjunction with the Denisova hominin mtDNA, a previously described reduction of silent substitutions causing an increased dN/dS in Neanderthals is not observed. This is probably due to a more accurate reconstruction of substitutional events when the long evolutionary lineage leading to modern humans and Neanderthals is subdivided by the Denisova hominin mtDNA (see Supplementary Information) (Krause et al., 2010)

The immediate result of this new finds is that an earlier proposed reduction in length of the Neanderthal mtDNA lineage “about three times as large as would be expected if it was entirely due to the age of the fossil” (Green, 2008), resulting in an earlier common ancestor to modern humans, is wrong. The shrunken phylogenetic tree was accordingly corrected for by Krause: the mean age of the most recent mtDNA ancestor of modern humans and Neanderthal went down from 660.000 t0 465,700 years ago.

(mean = 465,700 years ago; 321,200–618,000 years ago, 95% HPD) (Krause et al., 2010)

The feature that contemporary dN/dS values of modern humans are high, especially among Europeans, also corresponds to current assumptions that concern a younger age or (in the case of Europeans) of a smaller effective population size. May this be another lousy interpretation of results that are barely understood? This could be another example of a solution that supplies an easy way out of a complex issue.

There might be more. COX2 is a coding gene located on mtDNA. According to Green et al.(2008):

COX2 has experienced four amino acid substitutions on the human mtDNA lineage after its divergence from the Neandertal lineage […]

Fixed mutations indeed tend to define both human lineages as mono-phyletic blocks. But the paper only mentions COX2 as a potential indication of divergent evolution, and due to the new information revealed by the Denisova hominin nothing remains of Green’s assertions that Neanderthal coding mtDNA is strikingly different from modern human mtDNA. The main argument why this would be irreconcilable with a continuous development can now be rejected:

The observation of four nonsynonymous substitutions on the modern human lineage, and no amino acid changes on the Neandertal lineage, stands in contrast to the overall trend of more nonsynonymous evolution in Neandertal protein-coding genes (Table 1), and deserves consideration. (Green, 2008)

There is NO overall trend among Neanderthal towards a more nonsynonymous evolution, hence the four new proteins that correspond to four nonsynonymous substitutions on the modern human lineage do not indicate a striking new tendency, since this kind of mutations happened all the time, also among Neanderthal.
The age calculations gain in reliability once the synonymous mutations involved are better identified and harbored on the phylogenetic tree, by comparing more hominins and branches. Quite considerable purifying selection has now been identified as applicable to both Denisova and Neanderthal mtDNA. However, the mtDNA of an old skeleton in Australia already showed us that neither of this leads us closer to the mtDNA of modern humans.

Whatever the nuance of details, that scream variety and continuity in human evolutionary development, we can’t deny a striking, almost exclusive unity of AMH mtDNA compared to the different forms that have been recovered from Neanderthal and – even more – Denisova:

The genealogies of mtDNA sequences in most human populations, including Aboriginal Australians, characteristically have very little hierarchical branching structure. This pattern of sequence variation is consistent with a population expansion following a population bottleneck and is generally taken as supporting the recent out of Africa model. Under this model, all contemporary sequences spread globally with an expanding population that replaced all other people and all other lineages. Africa has been postulated as the source of the expansion because some populations in Africa have more sequence diversity than populations anywhere else. (Adcock et al., 2001)

Almost, since the discovery of ancestral mtDNA of the gracile early human, found at Lake Mungo, Australia (code named LM3, age 62 kya), that is unmistakably an AMH, also attests the extinction of quite distinct outliers. There must have been a huge and progressive selective thrust towards modern mtDNA. The mtDNA of LM3 was kind of “modern” alright, but definitely the genetic distance fell outside the range of modern humans. The investigators observed this find poses a serious challenge to the “interpretation of contemporary human mtDNA variation as supporting the recent out of Africa model” (Adcock et al., 2001), effectively reducing Africa as a refuge for outgroups that have accumulated change and drifted apart rather than being a true indication of the source of all AMH related mtDNA. But even more so, the find strongly indicates that the current lack of hierarchical branching structure among humans can’t be understood as the direct result of a succession of AMH migrational waves alone. Some waves phased out and lost their origin from the record. Could it be possible that something about mtDNA triggered the worldwide substitution of extremely divergent older forms by the reduced array of current forms? Then how did this happen?

The mtDNA genome, modelled as a circle.

Let us regard the issue in a wider genetic perspective and forget about cheap scenarios of cannibal hominins exterminating each other, a view that conveniently ignores autosomal evidence of inter-hominin gene flow. One little segment of non-coding mtDNA can be found on the Displacement (D-) loop or control region, that is involved in repair activities. It has an analogy in the telomers of nuclear DNA, that are highly prone to insertion and deletion processes. This little region may be subject to the random change and stochastic speed-density that are necessary to infer a neutral “molecular clock”, but the location of this region on the mitochondria introduces a substantial bias in the basic assumption of overall neutrality. I will return at this issue.
Several studies have demonstrated the ongoing transfer and integration of mitochondrial DNA sequences into nuclear chromosomes. The evolutionary inclination of mtDNA genes to move from the D-loop control zone to the nuclear autosomal part of the DNA could be studied in more detail on the paleogenetic sample of an AMH fossil found near Lake Mungo, Australia, dated 40kya (Bowler et al.,2003):
“His mtDNA belonged to a lineage that only survives as a segment inserted into chromosome 11 of the nuclear genome, which is now widespread among human populations.” (Adcock et al., 2001)

This particular strand of early human (AMH) mtDNA vanished from the mitochondrial record ever since, all over the world, but the insertion in chromosome 11 flourished, especially outside Africa:

Overall, 39% of chromosomes tested carried the insertion. In four African populations, the frequency of chromosomes carrying the insertion ranges between 10 and 25%, whereas it varies between 38% and 78% in populations tested in Europe, Asia, Oceania, and South America.(Zischler et al., 1995)

Assuming a lower evolutionary rate in nuclear DNA, “these mitochondrial integrations might preserve the ancestral state of the mitochondrial sequence that existed at the time of transposition and could therefore be regarded as ‘‘molecular fossils.’’” (Zischler et al., 1998). Previous investigation on a similar, albeit much older Insert on chromosome 9 that “took place on the lineage leading to Hominoidea (gibbon, orangutan, gorilla, chimpanzee, and human) after the Old World monkey–Hominoidea split” (Zischler et al., 1998), that happened in the range of 17–30 MYA in a common ancestor of all hominoids, already established the value of nuclear insertions for reconstructing ancestral mitochondrial sequences of the Most Recent Common Ancestor (MRCA):

Thus, the MRCA sequence deduced from homologous integrations in different species will represent the ancestral mtDNA sequence more reliably and with less sequence ambiguities than an ancestral sequence deduced from the fast-evolving mtDNA sequences. (Zischler et al., 1998)

The remarkable affiliation of the autosomal Insert with both LM3 and hominoid mtDNA. The newly discovered, potentially ancestral affiliation with the Denosova hominin is not drawn.

The Insert on chromosome 11 definitely suggests fossil information of some early AMH individual, or at least of a hominin that interbred with early AMH. The closest match to the mtDNA of this particular individual was indeed an AMH, the gracile LM3 dated 40kya (Bowler et al., 2003) found in Australia at Lake Mungo. However, a simple comparison of the Insert to the current genome of modern human mtDNA reveals that this individual can’t possibly be the direct ancestor of modern human mtDNA. No close mtDNA matches of LM3 nor the Insert survived and the mtDNA of LM3 doesn’t indicate direct matrilinear inheritance of the original mtDNA source of this autosomal Insert either.

The LM3 Sequence Belongs to an Early Diverging mtDNA Lineage. The divergence of the LM3 sequence before the MRCA of contemporary human sequences is indicated by its grouping with the Insert sequence, which other reports have suggested diverged before the MRCA of sequences in living humans.
Although this analysis did not reliably establish an early divergence of the LM3/Insert lineage, it demonstrated that the lineage is unusually long. (Adcock et al., 2001)

This presentation of the Insert as a member of a single branch together with LM3 may be an oversimplification. The location of the Insert at the mtDNA phylogenetic tree of humans suggest an even more pronounced outlier:

Upon comparing 243 bp of a human-specific integration (Zischler et al. 1995) that corresponds to the conserved part of the mitochondrial D-loop of all available hominoid (n=14) and human (n=261) mtDNA sequences, only two insert-specific substitutions were traced, with both the ape mtDNA sequences and all human mtDNA sequences being identical at these positions. (Zischler et al. 1998)

Salient detail is that the two Insert specific substitutions (A on 16259 and C on 16288) are now covered by the mtDNA of the Denisova hominin. Even though the other differences with Denisova are big enough to exclude a close affiliation, this remarkable detail invites to the tentative proposal that the divergence of the Insert sequence could have happened long before the MRCA of human sequences that also include LM3.

Between 16,259-16,381 the mtDNA variation of the Insert nucleotides is covered by the corresponding nucleotides of apes, Lake Mungo 3, current aboriginal polymorphisms (not drawn) and the Denisova hominin.

This rare scope on a deep Eurasian affiliation, combined with extant aboriginal polymorphisms that echo the survival of Insert and LM3 features in the haplogroup N and M branches of modern mtDNA, suggest a much more complicated phylogenetic tree than the one currently in use. Aboriginal mtDNA polymorphisms drawn in the figure of Adcock et al., 2001 (above) are part of a mixture of the closely related haplogroups N (~P?) and M (Q?) that up to now define the earliest Out of Africa scenario. Together they could be closer to an extinct group of Eurasian outliers than African branches separately. Also typical East Asian loci of mtDNA show a remarkable similarity, making the case of African branches being ancestral to haplogroups N and M less straightforward. The establishment of any “reversed tree”, however, is hampered by the apparent extinction or extreme “pruning” of what might have been an enormous Eurasian mtDNA variability. Any scenario that reverses the tree should account for this low extant Eurasian variability in comparison with Africa.

Let’s return to the assumed “neutrality” of mitochondrial DNA inheritance. High variability of the control region might suggest otherwise. One of the prerequisites of fast evolution is a fast mechanism underneath genetic change, and the purpose of fast mtDNA mutations could be just that, to put the precondition of rapid evolutionary change. Anyhow, a similar observation was made concerning the massive STR of chimps on the Y-chromosome, that seem to be secondary to the incredible evolutionary changes on the Y-chromosome as observed in the recent study of Hughes et al. (2010) I already wrote about here.

A set of interesting differences of mtDNA between humans is located on the Hypervariable Region (HVR). Most strikingly, HVR is not highly variable per definition. For instance, investigations on the Ayu fish (Takeshima et al., 2005) revealed the Hypervariable region may also turn into a Hypovariable region, what suggests a special functionality of the property defining HVR (or general D-loop) variability. And a substantial susceptibility to damage.

The mitochondria continuously reproduce themselves at intervals averaging about 2 weeks, like bacteria by a process of binary fission. They generate most of the cell’s (chemical) energy supply and because mitochondria use oxygen as an electron acceptor, they produce harmful free radicals that may cause genetic damage, often deletion mutations. This free radical damage to mtDNA cannot be repaired, basically because the regular repair mechanisms of the cells can’t access the mitochondria and the mitochondrion has no repair mechanism of its own. Therefore, mitochondria accumulate damage at each mitochondrial generation, what gradually leads to malfunction and ultimately affects the health of the organism as a whole.

However, this dreary scenario must have some constraints, or else all life on earth would already have ended millions of years ago. Somehow the reproductory system must have been exempted from this process, or at certain circumstances, and also it seems the genetic damage to mitochondria can be slowed down by exercise, both physical and mental, but especially by consuming antioxidants like vitamin C or omega-3 fatty acids. These are abundant in fresh fruit, raw meat and fish, indispensable supplements to the species that lost the functionality of the L-gulonolactone oxidase (GULO) gene – amongst whom one of the two major primate suborders, the Anthropoidea (Haplorrhini), that happens to include human beings, together with tarsiers, monkeys and apes. Originally meant as a genetic “improvement” for getting rid of the old and weak when food shortages occurred, ie. those most badly in need of antioxidants to remain healthy, the loss of this gene also effectively confined this suborder of primates to subsistence in the tropics. Only humans succeeded in finding new habitats in colder climates. They left the hot places where fruits were available all year round and traditionally made up an important addition to the menu, because they could. Only humans evolved into great hunters, and developed the necessary skills to catch fish, in order to compensate for the irregular availability of fresh fruits. Notwithstanding unfavorable climates, they managed to keep their necessary supply of antioxidants at a save level. And they did, for hundreds of thousands of years. Until everything changed at the eve of Upper Paleolithic – when human cultural advance reached a critical level.

What went amiss when humans reached their first cultural highlights? Their success triggered important improvements in their living standards, that moved their prime focus away from the concerns of harsh survival, and towards the community around the fire. They spend more time preparing their meals, started to cook their meat and fish and thus destroyed their main antioxidant food supplies. Degenerative diseases made their introduction and invoked new selective pressures, that caused a steady gene flow from the south to rejuvenate the slowly degenerating mitochondrial lineages in the north. In the mean while females ceased to worry about the survival of the fittest and developed a preference for “feminine looking men over their more rugged counterparts” (DeBruine et al., 2010), triggering the most notorious changes in the human anatomy that resulted in Anatomically Modern Humans as a progressive tendency all over the world.

However, this does not fully explain the current low overall variability of mtDNA even in fruit-rich tropical territories in comparison to the attested mtDNA of early AMH such as Lake Mungo 3. Still, cultural level related natural selection might be a good trail to follow.

Booming AMH culture most probably also entailed a closer contact between different groups within a wider economical areas. For sure this new behavioral patterns would have initiated a catastrophic increase of contagious diseases as soon viruses and bacteria could circulate freely among newly interconnected communities. However, this also implies a strong relation between resistance against (new) infections and mtDNA, that vastly exceed the benign effects of Vitamin C. The relation between mtDNA, antioxidants and the development of new “genetic” cures may reach a lot further. At this point it is tempting to regress to the behaviour of mtDNA and its facility to travel to nuclear DNA, and evaluate the genetic potential of mtDNA as a genetic laboratory against new diseases. Indeed, the immune system is where human DNA might have evolved most and is where most human variability occur.

Despite the high homology between chimpanzee and human genes at the level of amino acid sequences, human genome contains 1418 genes that do not have direct orthologues in chimpanzee, many of which are related to immune defence.
A number of genome-wide scans for positive selection have recently been performed (Wagner, 2007). They confirm that many immune genes and their regulatory sequences have been the subjects of positive selection in humans.
Population genomics is still in its infancy and the specific predictions may vary among studies but this is where future discoveries are anticipated. (Danilova, 2008)

Then, survival of just one little branch of early human mtDNA must point directly to the main focus of Upper Paleolithic development. Of the early mtDNA strands only those that accumulated in Africa were safeguarded against the effects of progressive damage, due to the continuous availability antioxidants. But in the center of change the preconditions for rapid change were set, including the extinction of mtDNA that did not meet the new standards of natural selection against the inevitable pandemics of cultural cohabitation and coexistence. Relatively low population density prevented the accumulation of high haplotype diversity, and the surviving mtDNA haplogroups in Eurasia obliterated all traces of a long, rich and diverse hominin history. To the effect that the false positives of mtDNA lured the public opinion into thinking that a long list of pre-AMH hominins, that include famous names like Neanderthal, Peking Man, Rhodesian Man, Denisova hominin etc., became extinct.

We can’t solve the origin question with a narrow scope, since the only truth is that we still don’t know. However, the Denisova hominin shows us one important clue: the more we know, the more complicated the solution. And most probably, the more hominins involved.


  • Krause et al. – The complete mitochondrial DNA genome of an unknown hominin from southern Siberia, 2010, link (paysite): try here
  • Krause et al. – A complete mtDNA genome of an early modern human from Kostenki, Russia; 2010, link
  • Derevianko et al. – A Paleolithic Bracelet from Denisova Cave, 2008, link
  • Howell et al. – Molecular clock debate: Time dependency of molecular rate estimates for mtDNA: this is not the time for wishful thinking, 2008, link
  • Adcock et al. – Mitochondrial DNA sequences in ancient Australians: Implications for modern human origins, 2001, link
  • Ovchinnikov et al. – Molecular analysis of Neanderthal DNA from the northern Caucasus, 2000, link
  • Orlando et al. – Revisiting Neandertal diversity with a 100,000 year old mtDNA sequence, 2006, link
  • Green et al. – A Complete Neandertal Mitochondrial Genome Sequence Determined by High-Throughput Sequencing, 2008, link (paysite): try here
  • Takeshima et al. – Unexpected Ceiling of Genetic Differentiation in the Control Region of the Mitochondrial DNA between Different Subspecies of the Ayu Plecoglossus altivelis, 2005, link
  • Sankar Subramanian – Temporal Trails of Natural Selection in Human Mitogenomes, 2009, link
  • Subramanian et al. – High mitogenomic evolutionary rates and time dependency, 2009, link
  • Zischler et al. – A nuclear ‘fossil’ of the mitochondrial D-loop and the origin of modern humans, 1995, link
  • Zischler et al. – A Hominoid-Specific Nuclear Insertion of the Mitochondrial D-Loop: Implications for Reconstructing Ancestral Mitochondrial Sequences, 1998, link
  • DeBruine et al. – The health of a nation predicts their mate preferences: cross-cultural variation in women’s preferences
    for masculinized male faces, 2010, link
  • Nadia Danilova – Evolution of the Human Immune System Evolution of the Human Immune System, 2008, link
  • Allard et al. – Control region sequences for East Asian individuals in the Scientific Working Group on DNA Analysis Methods forensic mtDNA data set, 2004, link (paysite), try here
  • Bowler et al. – New ages for human occupation and climatic change at Lake Mungo, Australia, 2003, link
  • – Global human mtDNA phylogenetic tree, 2010, main

Recommended reading: