Mitochondrial Eve

Glenn R. Morton (
Thu, 13 May 1999 08:30:42 -0500

In March there were reports of two studies which showed that paternal
inheritance, and maternal/paternal mtDNA recombination took place. These
studies show that mtDNA is NOT solely inherited from the mother as all
mtDNA studies to date assume. I finally obtained those two articles.
They do indeed have profound implications for the way Christians handle
fossil man. These articles falsify the view that mankind is a recent
(circa 100,000 year) addition to our planet. They also falsify the
notion that Neanderthal is a separate species which was the conclusion
drawn by many after the first mtDNA study of Neanderthal (Krings,
1996), The first is Hagelberg et al, (1999). The second is Eyre-Walker
et al, (1999) [references below].

The Hagelberg study examined the mtDNA population structure in
Melanesia. They compared the mtDNA from 452 individuals finding three
major mtDNA lineages implying three different migrations of peoples into
the Pacific basin. Since the time of the peopling of the Pacific, these
three lineages had mutated into at least 175 different haplotypes of
mtDNA. But they found same mutation at one DNA site on 8 different
haplotypes which represent all three mtDNA lineages. This mutation is
unique to the people of the island of Nguna. Nguna is inhabited by
about 1000 people so to find the same mutation in three different
supposedly matrilineages is surprising. I will quote at length their

"There is strong statistical support for the existence of three
separate human mtDNA lineages in our data. Interestingly, we detected a
base substitution at mtDNA position 16076 in people belonging to eight
separate mtDNA types falling into the three separate Pacific lineages in
the small island of Nguna. The presence of a rare mutation in people of
different lineages is puzzling. There can be several explanations for
this observation:The 16076 mutations occurred several times by cahnce in
the island of Nguna if position 16076 was hypervariable. Alternatively,
the 16076 substitution might have occurred just once before the split of
the three separate lineages, although this would mean that several
subsequent reverse mutations at other sites must have happened to
account for the data observed. A third explanation is that the 16076
mutation occurred once in an ancestor of the present Nguna population
and was subsequently transferred to other lineages by paternal leakage
of mtDNA and subsequent recombination.
"Hypervariability in mtDNA is a well-recognized phenomenon. IN current
human mtDNA studies , the multiple occurrence of base substitutions in
separate branches of a phylogenetic tree is generally attributed to
independent mutations at hypervariable sites of the mtDNA control
region. According to this view, the 16076 mutation would have had to
have occurred independently several times in Nguna, but virtually
nowhere else in the world populations studied to dae, and in none of the
other locations of our study. This seems unlikely. It is also unlikely
that the 160076 mutation happened before the three lineages diverged, as
this would mean that several back-=mutations would have had to occur at
this site to explain the existence of the other mtDNA types without this
substitution in Nguna and throughout the western Pacific.
"Although at odds with current dogma on mtDNA inheritance, paternal
contribution and genetic recombination are possible explanations for the
phenomenon observed in Nguna. Views on the strict maternal inheritance
of mtDNA have been challenged. Paternal inheritance of mtDNA was
observed in mice progeny at a frequency of one in 10000. IN contrast to
popular belief, the sperm tail with its mitochondria is not excluded
from the embryo during human fertilization (Ankel-Simons & Cummins
1996[Ankel-Simons is Gordon Simons cousin in law--grm]), although
experimental evidence for an active exclusion mechanism ahs been
presented. If such a mechanism exists, it has the potential to fail on
occasions, which would enhance the rates of paternal mtDNA leakage.
"It seems liekly that paternal mtDNA molecules might make a small,
albeit significant, contribution to mtDNA lineages, particularly if
measured over prolonged time periods of evolutionary history." (
Hagelberg et al, 1999, p 489-490)

They go on to state,

"If our conclusion is correct, genetic recombination probably occurred
at least twice (and possibly as many as eight times) in Nguna and it
therefore a relatively frequent event in human evolution. This has
important implications for evolutionary and phylogenetic studies of
human mtDNA. First, recombinations would perturb estimates of the time
of divergence of mtDNA types, raising questions about the suggested time
and mode of recent human evolution. Second, the occurrence of
recombination would cast doubts on the labelling of some mtDNA control
region nucleotide positions as mutation hotspots. It is just as likely
that the frequent occurrence of some substitutions in unrelated human
lineages might be the result of recombination, notably in populations
like those of Europe which have undergone high levels of genetic
admixture in recent millenia.
"In our data on Pacific populations, we detect little evidence for the
existence of hypervariable sites that could not be explained by the
occurrence of recombination events. For example, position 16129 is
considered to be one of the most hypervariable in human mtDNA, with a
rate approximately eight times higher than the background mutation rate
in the first hypervariable segment of the human mtDNA control region.. .
. If position 16129 was hypervariable, it is extraoridinary that it has
not changed in the western Pacific in the long time since the expansion
of people into New Guinea and island Melanesia, a period of 30 000 to 60
000 years. How likely is it for position 16129 to be fixed in the
Pacific but hypervariable in the rest of the world? . . .
"We would like to suggest that some of the so-called hypervariable
sites, for example 16129, 16223, 16311 and 16362, are in fact extremely
ancient substitutions in human mtDNA, which accounts for their
widespread distribution in human populations. The occurrence of these
substitutions in different mtDNA lineages, used as evidence of their
hypervariability, could simply be the result of their extreme antiquity,
couples with the effects of occasional recombination events. It is
interesting that the distribuiton of these substitutions is particularly
high in African populations, as well as in New Guinea highalnnders, who
are thought to descend from some of the earliest anatomically modern
humans to migrate out of Africa." (Hagelberg et al, 1999, p.490-491

The African population is the oldest population of humans on earth.

Hagelberg et al then address why human mtDNA has so little variance when
compared with the variance seen in other species like chimpanzees.

"One of the worldwide effects of recombination would be to blur the
differences between human mtDNA lineages, as recombination would create
hybrid haplotypes and eradicate ancient haplotypes. This would make it
harder to reconstruct the ancestral relationships of human populations.
If paternal mtDNA leakage and subsequent recombination or some analogous
gene conversion event are significant in human evolution, it will be
necessary to revise the conclusions of many current studies based on
mtDNA, including views on the relationship between Neanderthals and
modern humans. The occurrence of a mtDNA type which differs at 27
positions from an average modern human mtDNA sequence would not be
sufficient to conclude that a single Neanderthal individual was of a
species that did not interbreed with anatomically modern humans, as a
lineage of such antiquity could easily be lost in the intervening 30 000
or more years since the individual lived." (Hagelberg et al 1999, p.

The other study, by Eyre-Walker et al, is a statistical study of
variations in mtDNA among chimps, gorillas and humans. Basically this
study showed that there are too many back-mutations (homoplasies)to be
accounted for by matrilineally inherited mtDNA. They studied the third
site in nucleotide codons. The third site is the third nucleotide in the
coding which occurs for proteins. They state,

"An analysis of 3628 synonymous third sites in the protein-coding
regions of 29 human mitochondrial sequences revelaed 126 polymorphisms.
If all third sites were equally liekly to change, the expected number of
homoplasies would be 2.2, whereas the observed number is 22; the
probability of such an excess is effectively zero. It follows that there
there has either been recombination or that all third sites are not
equally likely to change." (Eyre-Walker et al, 1999, p. 481)

The idea that mutation probabilities are not equal is addressed by
Eyre-Walker, as it was in Hagelberg et al's paper. Eyre-Walker et al

"There is no evidence of variation in the mutation rate. If some sites
have an elevated mutation rate in both directions, this would cause
sites that are variable in humans to also be variable in other
primates. In fact, no such tendency exists. The humber of polymorphic
sites and of apparently homoplastic sites in humans that are variable in
other primates both agree rather closely with the numbers expected if
all third sites are equally likely to change. (Eyre-Walker et al, 1999,
p. 481)

Eyre-Waler et al (p. 481) then argue that recombination is a real

"Finally, two steps which are required for recombination between
mitochondrial lineages are known to exist. Paternal mitochondria enter
the egg (Kaneda et al, 1995; Ankel-Simmons & Cummins 1996) and they
contain the enzymes necessary for homologous recombination (Thyagarajan
et al, 1996). The only barriers to recombination are the fusion of
mitochondria and the time for which paternal mitochondria survive once
they are in the egg. It remains cunclear whether mitochondria fuse
frequently and in mice there are efficient mechanisms for eliminate
paternal mitochondria, so that, within several hours of fertilization,
paternal mtDNA can no longer be detected. However, only relatively low
levels of recombination are probably required to generate the patterns
of homoplasy we observe."

They conclude,

"Mitochondrial DNA has been used extensively in the study of human
evolution. In many of these analyses the clonality of mitochondria has
been either explicitly or implicitly assumed. It is clear that many of
these conclusion will have to be treated with caution or reassessed. It
certainly seems dangerous to assume that mitochondria are clonal when
there is evidence against and no evidence in favour of such a
conjecture." (Eyre-Walker et al, 1999, p. 482.

This data, along with the possible Neanderthal/Human hybrid reported
earlier, screams out for a new apologetical treatment. Christians who
have argued for a humanity that was less than 100,000 years old and
genetically separate from archaic Homo sapiens such as Neanderthal and
Homo erectus are in danger of following the path traveled by the
young-earth creationists. That is, they are beginning to ignore
important pieces of data in order to support their preferred theological
interpretation. This is a too often occurrence among many christian


Eyre-Walker, Noel H. Smith and John Maynard Smith, "How Clonal are Human
Mitochondria?", Proc. Royal Soc. Lond. B (1999) 266:477-483

E. Hagelberg et al, "Evidence for Mitochondrial DNA Rcombination in a
Human Population of Island Mwelanesia," Proc. Royal Soc. Lond. B (1999)

Matthias Krings, et al., "Neandertal DNA Sequences and the Origin of
Modern Humans," Cell


Foundation, Fall and Flood Adam, Apes and Anthropology