Genetics & Genealogy

Neanderthal DNA

By Alec Christensen, May 2000

This page was last updated on

Over the past few years, DNA has been very much in the news. While most research has been conducted on genetic material extracted from modern humans, geneticists and biological anthropologists have used such data to reconstruct patterns of ancient population history. Scientists have begun to extract DNA from ancient remains and report the successful extraction of mitochondrial deoxyribonucleic acid, or mtDNA, from the ribs of a 29,000 year old skeleton from Russia. This skeleton of an infant has been assigned to the subspecies Homo sapiens neanderthalensis with an mtDNA sequence very divergent from all of those reported from modern humans, and scientists calculate that our ancestors and this young child's ancestors separated between 853,000 and 365,000 years ago.

Over the past few years, DNA has been very much in the news, whether in accounts of the O.J. Simpson trial or news about medical breakthroughs. While most research has been conducted on genetic material extracted from modern humans, geneticists and biological anthropologists have used such data to reconstruct patterns of ancient population history. But there is only so much that you can say about past populations by looking at present ones. So scientists have begun to extract DNA from ancient remains. In last Thursday's issue of Nature, I.V. Ovchinnikov, A. Gotherstrom, G.P. Romanova, V.M. Kharitonov, K. Liden, and W. Goodwin published an article entitled "Molecular analysis of Neanderthal DNA from the northern Caucasus" in which they report the successful extraction of mitochondrial deoxyribonucleic acid, or mtDNA, from the ribs of a 29,000 year old skeleton from Russia. This skeleton of an infant from Mezmaiskaya Cave has been assigned to the subspecies Homo sapiens neanderthalensis, which some people prefer to call Homo neanderthalensis. Ovchinnikov et al. found an mtDNA sequence very divergent from all of those reported from modern humans, and calculate that our ancestors and this young child's ancestors separated between 853,000 and 365,000 years ago. This follows three years after M. Krings and colleagues reported in Cell that they had extracted mtDNA from the original Neanderthal, a middle aged man who lived between 30,000 and 50,000 years ago. They calculated that our last common ancestor lived between 690,000 and 550,000 years ago. In a second analysis of DNA from the same specimen, they placed that date within a range of 741-317,000 years ago.

These two reports have confirmed the belief of many scholars that Neanderthals were not our ancestors; rather, they were our distant cousins. Ever since the skeleton studied by Krings et al. was found in the Feldhofer cave in the Neander Valley (or Thal/Tal) near Dusseldorf in 1856, there has been debate over its relationship to us. Some German scholars argued that it belonged to a diseased Cossack; others recognized its antiquity and saw it as a primitive member of our own ancestral tree. Most anthropologists now agree that the Feldhofer skeleton and many others like it are representatives of a population of humans that adapted very well to life in glacial Europe between about 230,000 and 30,000 years ago. Where they disagree is over the question of this population's relation to us. Some, most vocally Milford Wolpoff, argue that Neanderthals became us - at least, those of us with European ancestry. They see a continuous evolutionary process visible, leading through Neanderthals and other "archaic Homo sapiens" to modern humans. Others, particularly Chris Stringer and Ian Tattersall, argue that our direct ancestors migrated out of Africa less than 100,000 years ago and replaced the Neanderthals. How did they replace them? Many believe that there were significant cognitive differences between them and us; that is, that modern humans thought about the world differently, and this allowed them to think about the world in a more effective way.

Genetic data provides a new way of resolving this debate. If the Neanderthals were closely related to us, this should be visible in their genes. If they were our great-uncles, they should be more distantly related. Also, if their descendants live anywhere today, it is presumably in Europe, so in that case modern Europeans would presumably be more closely related to them than populations from elsewhere.

Before we assess how these new analyses answer these questions, we need to understand what they are actually analyzing. What is mitochondrial DNA? Each human cell has mitochondria in it, small organelles that maintain a semi-independent existence within. Mitochondria carry their own DNA, rather than being controlled by the DNA found in the cell's nucleus. Your nuclear DNA comes from a fusion of your mother's and father's genes, but all the mitochondria in your body are descended from the mitochondria in your mother's ovum. So while you only share 50% of your genes with either of your parents, your mtDNA is identical to your mother's. Why do anthropologists and geneticists study mtDNA? Well, primarily because it's easier. Each cell has one copy of its nuclear DNA, but it may have up to 1,000 copies of its mtDNA. Also, the fact that mtDNA passes down from just one parent, without the recombination that occurs between the nuclear DNA of both parents, makes it more straightforward to reconstruct branching, phylogenetic trees. Finally, mtDNA mutates relatively quickly, so there is a high amount of variation between humans.

DNA is composed of a long series of base pairs, matched nucleotides whose sequence forms the genetic code, which tells cells how to develop. Out of thousands of base pairs in mtDNA, anthropologists usually focus on a few specific segments that are known to vary a lot between individuals. Hypervariable region I has been sequenced for both Neanderthal specimens - all of it for Feldhofer, and a large portion of it for Mezmaiskaya.

These two studies indicate that Neanderthals were quite distantly related to us. Looking at a specific segment of the mtDNA, Ovchinnikov et al. found that 12 out of the 345 base pairs they recovered differed between the two Neanderthal sequences. Very few modern human samples differ from each other by that much, which indicates that Neanderthals may have been more variable genetically than modern humans are. When compared to a standard modern human sequence, 22 of the 345 Mezmaiskaya base pairs differed and 27 out of 379 positions recovered from the Feldhofer specimen differed. This is roughly triple the number of differences found between any two modern human specimens.

What do the dates given in these papers mean? Well, assuming a constant mutation rate, you can calculate the number of generations it should have taken for two sequences to diverge as much as they have. But is that the date that the two populations separated? No. After all, there are plenty of people living around me, basically identical to me, who nonetheless share no maternal ancestors with me unless we go back several millennia. So dates of individual divergence always predate actual population separation. So what exactly does the range of 741-317,000 years ago mean? It could mean that our ancestors diverged from the proto-Neanderthals over 700,000 years ago, or less than 300,000 years ago. Obviously the evolutionary significance of those two dates is completely different. When the two Neanderthals are compared, it can be calculated that they diverged between 352,000 and 151,000 years ago, while modern humans diverged from each other between 246,000 and 106,000.

Because of the uncertainty about what these dates mean, the new findings have not settled the debate about our relationship to Neanderthals. Even after the new mtDNA study, there are still those who insist that Neanderthals and early modern humans may have interbred. After all, two individuals is not exactly a huge sample. And it is clear that the two groups coexisted in some areas of western and central Europe for at least several millennia. Given our own history, it seems improbable that any two human groups, no matter how different, could have lived side by side without some children resulting unless it was physically impossible for them to breed. Although mtDNA in and of itself does not speak to fertility, it is unlikely that Neanderthals and modern humans differed enough to make interbreeding biologically impossible. There is also skeletal morphological evidence, particularly from the Iberian peninsula, for admixture between Neanderthals and modern humans. Perhaps because such morphological evidence cannot be quantified as genetic data can, it has not received as much attention. But it does indicate that the picture is more complex than a simple mass replacement. In some areas, there was clearly more genetic contact between the two populations than in others.

So where do we stand today? After these two studies, we can be very confident that Neanderthals were quite different from us. But then again, we already knew that. We can also be fairly confident that they did not contribute a huge number of genes to the modern gene pool. But without more genetic data on both Neanderthals and their contemporary early modern humans, we cannot say much more than that. Which brings up an interesting point: why hasn't anybody isolated mtDNA from one of those early modern humans who lived at the same time as the Neanderthals, to see if they really are more closely related to us? I have wondered about this question myself. Even mitochondrial DNA does not preserve that well, and the older a specimen gets the less likely it is to contain intact DNA. One thing that hastens the breakdown of DNA is heat, so we may never be able to extract any from those early modern humans that lived in Africa and the Middle East. Both Neanderthal specimens were in fairly cool conditions. But why has no one extracted any from early modern humans of Europe? Perhaps we shall hear those results next year.

Neanderthal DNA Update

Two months ago, I reviewed a pair of recent studies of mitochondrial DNA from two Neanderthal skeletons. In the June 2000 issue of the American Journal of Human Genetics, Michael Scholz et al report the results of a different kind of analysis of two more Neanderthal specimens and one early modern human. Their results do not conflict with the earlier studies, but the methodological differences require differences of interpretation as well.

The earlier studies by Krings et al and Ovchinnikov et al derived exact sequences of the hypervariable region of mtDNA. This requires well preserved bones and very careful extraction. Scholz et al. used a simpler method called "Southern blot hybridization". Simply put, DNA was extracted from each specimen and tested against DNA from the other specimens to see the degree of similarity. Two identical specimens should hybridize completely, and two unrelated specimens should not hybridize at all. This method should work no matter what part of the genome is preserved.

Neanderthal specimens from Warendorf-Neuwarendorf, Germany, from before 50,000 years ago, and Krapina, Croatia, from 110 to 100,000 years ago, were compared with an anatomically modern human specimen from Stetten, Germany, dated to 35,000 years ago, a modern human, a modern chimpanzee, and an ancient mammoth and reindeer. The last two were included as controls: if they hybridized with any of the ancient human remains, it would indicate a methodological problem with the reaction.

In fact, the control specimens proved completely unrelated to the hominoid (human and chimpanzee) samples. This indicates that the other reactions are probably valid. The two Neanderthal samples proved quite similar to each other. They were also fairly similar to the Stetten sample. But the modern human sample was four times more similar to the Stetten sample than to either of the Neanderthal ones.

Based on this finding, the authors conclude that the Neanderthals may have been a distinct species from modern humans, while Stetten belonged to our species. This is certainly a plausible explanation. But the Stetten individual is also far closer to us in time than either of the Neanderthals. It is one third the age of Krapina, and perhaps half the age of Warendorf-Neuwarendorf. Given this temporal proximity, even if it and Krapina belonged to the same evolving lineage, we would expect Stetten to be more similar to modern humans.

So this study adds to the evidence that Neanderthals were genetically distinct from modern humans. But then again, based on morphological evidence, this was never in doubt. It also shows that anatomically modern humans from 35,000 years ago were more closely related to us than 100,000 year old Neanderthals. Again, this was never in doubt. Beyond this, the conclusions are less solid. As the authors admit, they have no way of knowing what portions of the genome are the cause of the differences, and they do not provide a comparison with any other case of specific vs. sub-specific distinction. The conclusion that Neanderthals were outside the modern pool of human variation does not in and of itself mean that they were a different species, since humans exhibit far less genetic variation than many other perfectly valid species. And as Relethford, Wolpoff, and others have pointed out in different venues, many different biological processes can produce similar genetic results. For the moment, I am still inclined to believe that Neanderthals did contribute to the modern gene pool, although their genetic contribution has been overwhelmed by the larger size of African-derived populations.