MAPPING THE NEANDERTHAL GENOME
[SVANTE PÄÄBO:] We are now in the process of analyzing the Neanderthal genome, putting together all the little DNA pieces we have extracted from the sequence of this fossil and starting to compare it to humans and chimps. One question that we are struggling with and thinking about is, What is the relationship of the Neanderthals to us?
We try to address that in different ways. One thing that we're beginning to see is that we are extremely closely related to the Neanderthals. They're our relatives. In a way, they're like a human ancestor 300,000 years ago. Which is something that leads you to think: what about the Neanderthals? What if they had survived a little longer and were with us today? After all, they disappeared only around 30,000 years ago, or, 2,000 generations ago. Had they survived, where would they be today? Would they be in a zoo? Or would they live in suburbia? These are the questions I like to think about.
And it's the question and not the answer that's interesting because these questions have no answers. We will never know. But they are interesting questions to think about because they somehow reflect how we think about differences between us and our ancient ancestors.
For example, if the Neanderthals were here today, they would certainly be different from us. Would we experience racism against Neanderthals much worse than the racism we experience today amongst ourselves? What if they were only a bit different from us, but similar in many ways — in terms of language, technology, social groups? Would we still have this enormous division that we make today between humans and non-humans? Between animals and ourselves? Would we still have distanced ourselves from animals and made this dichotomy that is so strong in our thinking today? These things we will never know, right? But they are fascinating things to thnk about.
Truthfully I don't know how it would play out if they were alive today. I truly don't know. It's interesting to speculate about the mutual genetic influence of humans and Neanderthals, in both directions. Probably both sides of this are true to some extent. It's probably true that these tendencies in ourselves that create racism today would be at play in a world with Neanderthals equally or even more. It's probably also true that the dichotomy between humans and all other mammals would not be experienced as such a strong chasm as in the way we think about the differences today.
Of course it is very difficult to find evidence in the absence of something. Today there is no unequivocal evidence that Neanderthals contributed genes to people living today. But that's not to say that it doesn't exist. The only thing we know is that they did not contribute mitochondrial DNA. That's the one thing we can say.
But since just a few weeks ago, we have discovered a large part of the entire genome of Neanderthals: 66 to 70 percent of it. So we can now address the question of Neanderthal's genetic contribution to modern humans much more rigorously. And we can ask about both directions of genetic influence because of course mixture goes two ways. So far we could only study modern humans that live today and we couldn't find any evidence that Neanderthals contributed to humans. But now that we have the Neanderthals, we can ask about the other direction. So one thing we're asking is: Is there any evidence that early human ancestors interbred with Neanderthals and contributed genes to them? That's something we are struggling with at the moment.
We are analyzing this and it's a very difficult analysis, because if there are tiny parts of a genome that were contributed to humans, you have to assume that they are not due to some error in our analysis, or contamination of the DNA of modern humans, or any little bias in the algorithms we use.
In the public media we're pretty much depicted as saying there was absolutely no mixture. It's very hard to convey these subtle messages to the public. If you read our papers, we say very carefully that there is absolute proof that they didn't contribute mitochondrial DNA. That doesn't mean that they couldn't have contributed other parts of their genome.
It's still clear that any genetic contribution from the Neanderthals to modern humans has to be pretty small. There is a lot of evidence in the sense that Africa is more genetically diverse; there's more variation in Africa than for all humans outside Africa, although there are a lot fewer people there — only 800 million people or so.
Everything we find outside Africa has close relatives inside Africa in terms of genetic variance. But that's why I often like to say that if we look at our genome, our DNA, we are all Africans today. Either we live in rather recent exile, maybe since 50,000 years or so, or we live inside Africa. If Neanderthals had contributed a lot to Europeans today, we would see that Europeans would have genetic variance not found in Africa today. Or that would differ a lot from people in Asia, for example, which we don't find.
As an outsider to paleontologists, I'm often rather surprised about how much scientists fight in paleontology. And I am thinking about why that is the case. Why do we have less vicious fights in molecular biology, for example? I suppose the reason is that paleontology is a rather data-poor science. There are probably more paleontologists than there are important fossils in the world. To make a name for yourself is to find a new interpretation for those fossils that are extant. This always goes against some earlier person's interpretation, who will not like it very much.
There are many other areas of science where we can agree to disagree, but at least we often generally agree on what data we need to go out and collect to resolve the issue and no one wants to come out too strongly on one side or the other because the data could, in a year or two, prove you are wrong.
But in paleontology you can't decide what you will find. You can not in most cases go out and test your hypothesis in a directed way. It's almost like social anthropology or politics — you can only win by somehow yelling louder than the other person or sounding more convincing. That's perhaps the reason why paleontologists get so heated in these fights.
Of course in the question of modern human origins the fight is between the multi-regionalist versus the out-of-Africa hypothesis. Genetic data from studies of the mitochondrial DNA in the 80's and then data from lots of other parts of our genome strongly favored the out of Africa hypothesis, which Stringer had long been one of the main proponents of in paleontology.
It's very clear from the genetic evidence that the big picture is the out of Africa one. That's not to say that there couldn't have been a small contribution from earlier archaic forms such as Neanderthals in Europe to present-day Europeans or erectus forms in Asia to Asians, but it has to be very small. The other camp, which follows Wolpoff, is in a minority today.
But the Neanderthal genome will be a chance to address that. And it depends on what you're interested in. As a geneticist I'm not so interested with who had sex with whom 30,000 years ago. The question is, as a geneticist today: did Neanderthals contribute significantly to our gene pool today? Did the Neanderthals have an impact on the variance we carry? And that's got to be small.
But for understanding what happened when modern humans met Neanderthals, this question of how we interacted with each other is rather important. It would be quite interesting, for example, if we were to find that there was a gene flow, but in the other direction, mainly from modern humans into Neanderthals. Because after all when two groups meet each other and there is some social inequality, one almost always mixes, but it's generally rather directional — the offspring of the humans will generally stay with one or the other group.
Generally with the non-dominant group interface with a dominant group, it would very often be a reverse gene flow. When modern humans are in this situation, most often the males from the dominant group have offspring with females from a non dominant group, and the kids stay with the non dominant group. If something like that happened when we met the Neanderthals, we would perhaps have had an inflow from modern humans into Neanderthals. And we would not see it in mitochondrial DNA in modern humans.
But I have a feeling about how we speculate about the Neanderthals. I often like to say that it's more about our world view than anything that happened back then. If you are a racist, you could play it either way. One could say that if the Neanderthals contributed to current day Europeans and if the Erectus in Asia did the same to Asians, there must be old variants adapted to living in Europe which had been there for hundreds of thousands of years. This means there is this group that was adapted to living in Europe, say, which was living there and then started to move around the world. You can start telling stories like that.
But you can equally do it some other way, saying that the people who left Africa were the more innovative advanced people who exploited new territories. That they were able to go out and do these things and that this was somehow in the genetic subset of what existed in Africa. You can spin it either way you like. I don't think that there's any scientific knowledge or insight that will convince people to change their ingrained ideas about this. One is often asked, why do you think Neanderthals disappeared? Did we kill them all? If you like to see modern humans as very violent, you would say this was obviously our first big genocide. Just look at how we behave today.
On the other hand, what if I say the earliest modern human fossils in the Middle East are 93,000 years old. The latest Neanderthals in the Middle East are 60,000 years old. So we have 30,000 years of peaceful coexistence in the Middle East. If you could have that today, it would be wonderful. You can say, well, that's probably because modern humans came and certainly the Neanderthals disappeared and modern humans disappeared again, so they never interacted with each other. But who knows? It just reflects how we think of ourselves and what stories we make up about what happened back there.
The very first attempts to extract DNA from old remains go back to the first half of the '80s, when I started working with ancient Egyptian mummies for example. But it was the invention of polymerase chain reaction by Kary Mullis that made it possible to target and go in and get a piece of DNA that you were interested in from a fossil and reproduce your results and show that you can do it again and again so it was reliable. Others could repeat it. That was in the late '80s, the beginning of the '90s.
And then this was applied to the Neanderthals for the first time in '97, which was when we started to have an impact on paleontology and paleoanthropology. But what we have experienced over the past two years is that there is new technology: high-throughput DNA sequencing. There are machines now where you can just take the DNA you've extracted from a fossil and sequence random pieces of DNA in it so efficiently that you can, without targeting anything special, just see everything that's in there and then look what on this DNA molecule looks similar to humans or chimps.
And it's generally just a few percent — two, three, four percent — but there's such a tremendous throughput in these technologies you can afford to do that. You can throw away 95, 98 percent of your data and just look at the rest. That has changed the ball game so that we can now begin to look at the total genome of extinct forms, such as the Neanderthals or the mammoths or other extinct animals.
As a kid, I wanted to be an archeologist and Egyptologist and wanted to excavate in Egypt and things like that. I guess, as is often the case, I had a far too romantic idea about these things. When I got to university I started studying Egyptology and found that it was not at all like Indiana Jones, as I had imagined. At least as it was taught in Sweden, it was very much linguistics, it was thinking about ancient Egyptian verb forms and things like that. I got disenchanted with it and didn't know what to do. I was also influenced by my father, so I decided to study medicine and do research.
But that was when DNA technologies and cloning and DNA was coming of age. I knew that there were thousands of mummies in collections in Egyptological museums and that hundreds of new mummies were found every year in Egypt. No one seemed to have applied the new technologies to the remains, which was rather the obvious thing to do — that is, say, take a sample of an ancient Egyptian mummy, extract the DNA and clone it in bacteria and study the DNA from it.
I started doing that as a hobby activity on the side of other things. I was a bit scared of my thesis advisor, who was a rather dominant person, so I did it secretly at night and on the weekends. It was a successful endeavor. We showed that you could stain DNA in the cell nuclei in the tissue samples from a few Egyptian mummies. You could also extract the DNA and show that there were human DNA in there. But it was then hampered by the impossibility of retrieving any specific things, so small were the traces of DNA. It was changed two years later when the polymerase chain reaction came about. Then I ended up going to Berkeley to a lab that also had interest in this and that's where I developed it.
When I started out in '84/'85, intent on studying the genomes of ancient civilizations, I was, as is often the case in this kind of situation, driven by delusions of grandeur. I thought that I would be able to easily study the ancient genomes. I dreamt of addressing questions in Egyptology. For example, how do historico-political events that we read about impact the population? When Alexander the Great comes to Egypt, what is the influence on the population? Is it just a political change? The Arab Conquest: does that mean that a large part of the population is replaced? Or is it mainly a cultural change? There's no way we can answer this question from historical records. But my dream was to address questions like this. Then, after some initial success, I realized the real limitations on what I wanted to do.
There was then a much longer period where we concentrated on extinct animals. We did the first DNA sequences from a mammoth, from marsupial wolves, moa's in New Zealand, because we didn't have to struggle with the contamination issue there, where modern human DNA can easily be distinguished from these things.
In the last few years, with a few new technological possibilities, one can begin realizing the dreams I had in the '80s.
One big dream is to address questions about things that are specific to humans relative to other life forms. Such as language. So it was extremely exciting a couple of years ago the when the FOXP2 gene was identified. A mutation in that gene in humans resulted in a specific speech problem. And it seems to be a speech problem that has to do with articulation. The primary problem concerns muscle control in the oral thorax — a millisecond of control you need of what your vocal chords, your tongue, your lips do to produce articulate speech.
We studied the evolution of that gene. The protein that's made from it is a protein whose function is to turn on and turn off the activity of other genes in the body. And the protein carries two amino acid substitutions that are unique to humans, that are seen in no other primates and that happened on the human lineage.
There are also patterns of variation in the FOXP2 gene among humans today that suggest that there had been positive selection acting on it and that only one variant spread rapidly to all humans on the planet today. It was very tantalizing to speculate that those variants were those amino acid changes, and that they had an impact on our ability to produce articulate speech.
There are two things we are exploring since we addressed the variation question and showed a singular positive selection. One is to look at the Neanderthals and to resolve, what were the Neanderthals like? Did they have these amino changes or not? And it turns out that they do share these amino acid changes with us — which I found surprising. To the extent that these changes in amino acids have something to do with articulate speech, we share with the Neanderthals. This is of course one gene out of many others that are still unknown having to do with language and speech. There could have been some difference, but from the little we know there is no reason to assume a difference.
The other thing we are trying to address is whether these amino acid differences are of importance/ That's a difficult thing to address. What we have done is construct a laboratory mouse that makes not the mouse version of this protein, but the human version of the protein from its endogenous FOXP2 gene (because mice like every other vertebrate have a FOXP2 gene).
We've analyzed that mouse extensively over the last two years and run all kinds of tests. We and our collaborators have looked at over 300 different traits in this mouse, always comparing the knock-in mice that are changed — the humanized mice — to litter mates born together of the same mother that are wild-type, producing the mouse protein. We always compare them directly to each other, so they have had the same birth experience, the same environment. There are only two of over 300 traits that we've looked at where the humanized mice differ significantly from the wild-type mice.
One thing I don't understand is that the humanized mice are slightly more cautious in a new environment than the wild mice. If they are, for example, in a social group and entered an open area that they are to explore, the humanized mice stay along the walls for the first few minutes, whereas the wild-type mice are more bold and enter the open area where a mouse feels more exposed and vulnerable. But that's a difference initially, for the first four or five minutes, and then there is no difference. I don't know what to think about that.
The other mystery is tantalizing and shocked me: the mice vocalize differently. We measure that by taking the pups away from their mother when they're two weeks old, and they peep so that the mother comes and brings them back into the nest. We can record that in the ultrasound area. We work with people who are experts in analyzing sonograms of vocalizations, and there is a clear — subtle, but clear — difference in vocalization. This supports my belief that these changes have something to do with muscle control in the oropharynx or so, probably something to do with articulation in humans. But that's an earlier change on the human lineage than the divergence to the Neanderthals; it's something we share with the Neanderthals.
The major thing that we realized rather early on when we started applying PCR in ancient DNA is that contamination is a very serious issue. Particularly that pertaining to the contamination of modern human DNA because it's all around us. The dust in this room is to a big extent skin fragments from our bodies that contain DNA and can land in experiments or be in the chemicals we use. That's why we decided to stay away from studying modern human DNA early on, such as ancient Egyptian human mummies, for example, because it was almost impossible to show that you had the right thing. With Neanderthals, it's a bit different because we're lucky that for the mitochondrial DNA, we have clear differences.
When we applied high-throughput sequencing technology through the Neanderthal remains, we tried different technologies. We made extracts from the best Neanderthal we had from Croatia (we made two extracts) and we sent one extract to Eddie Rubin's group at Berkeley, who cloned it in bacteria. They used technology that has been around from the '80s, but is now more efficient, sequenced clones and retrieved like 60,000 to 70,000 base pairs from the Neanderthal.
We sent another extract from the very same bone to 454 Life Sciences in Bradford where Jonathan Rothberg, who founded the company, and bought the rights to pyrosequencing from Sweden (where I knew the people who had developed it). Then we applied their technology to it. And that turned out to be much more efficient. They sequenced around almost a million base pairs — around 750,000 base pairs.
We analyzed these different data sets and there were two papers published in the same week; one in Science, which Eddie Rubin produced, and one in Nature, produced with 454. There was some tension with Eddie Rubin's group because they were very intent on continuing with the cloning approach in bacteria. But it seemed obvious to me that it wasn't efficient enough; they had produced ten times less data from what we had sent them than 454 produced.
We had sent them a DNA extract that we had produced in our clean room from the bones. When you make it into a library, in bacteria you can then propagate the bacteria and they are there forever if you like. But what they sequenced was everything that was in that library. There was nothing more to be had and there are so many losses in this process of getting the DNA into the bacteria that most of it is lost. When you sequence with this high throughput sequencing technology, you put synthetic pieces of DNA on the ends of your DNA, and you can by PCR amplify that and also in principle multiply this and use it to exhaustion.
It turned out that when we looked at the data, we realized that there were quite a few differences between these two data sets. But as the data sets from Berkeley were so much smaller, it was very hard to gauge these things. And then there was a group in San Francisco, Jeff Wall, a theoretician, who compared these two data sets and pointed out that there were features in the high throughput 454 data set that were significantly different from what had come from the clones from Berkeley, indicating that there was contamination in the 454 data set and that were differences in divergence from the human genome or the frequencies in which you see the new mutated variance of things that vary among humans to-date.
However, we also realized there were other arguments in that paper that were incorrect. Jeff Wall's paper, for example, said that longer sequences differ from short sequences indicating that the longer ones are probably contamination, but all of that probably has to do with the efficiency with which you can map it to the genome as the sequences get shorter and have errors in them due to damage in the ancient DNA.
They also get much harder to map, so you require a much more perfect match. But anyway, they were right; there was contamination in that data set; we still don't know where it entered since both extracts came from our clean room, prepared from the same bone. Probably the most likely thing was that this was introduced at 454 at the same time they were sequencing Jim Watson — I wouldn't be totally surprised if we tried to investigate this and there was some slight mixture of Jim Watson with the Neanderthal sequence there.
We have afterwards been able to go back to that data set now that we have the complete Neanderthal mitochondrial genome and we can look much more carefully at sites all across the genome. And we do have our best point estimate as to the percentage of contamination in the data set that was published.
In 2008 we published a paper in Cell with the complete mitochondrial genome where we then discussed this point. Prior to the publication of the Jeff Wall paper, we had already published the measures to avoid contamination that were implemented before then. We stated that we would tag the sequences in the clean room so that when they left the clean room for sequencing, every sequence would start with this little tag. That way we would know that the sequence came from our clean room. Because we do see that when we use small amounts of DNA, there is some carry-over in the sequencing machines. But that was already published before Jeff Wall's paper came out and then they never cited it, which I was also sad about. We had pretty much solved the problem before that paper came out.
The next big thing as far as announcements coming out of my lab is the scientific publication of the draft version of the Neanderthal genome — the first draft version of it. The only finished genomes among mammals are the human and the mouse. Everything else are drafts. The Neanderthal will be a very, very rough draft — it's just 1.5-fold, meaning that statistically every nucleotide in the genome has a 1.5-fold chance to have been hit. But that means there are a lot of parts of the genome that are not hit at all, something like 30 percent we have not seen. But we can get a first overview of the genome. We can make windows of say 100,000 base pairs and out of those we will get on the average 60,000, 70,000 and go over the genome.
What's involved in getting to a 100 percent is sequencing much deeper in more samples. We will be doing this in the next few years.
Some people ask me why I ended up in Germany. Like so much that happens in life, it's just by chance. When I was at Berkeley, I wanted to go back to Europe and I had a girlfriend who was a graduate student in Munich, so I went there and visited. Her professor, a very good geneticist, asked me to give a seminar, said there would be a position open in a year if I wanted to apply.
Things move slowly in university politics. By the time I got an offer for that position, I didn't have that girlfriend anymore. But it ended up being a very good offer compared to what was available to me in Cambridge or Sweden. I said, well, it's not that I ever dreamt about living in Germany for the lifestyle, but I can go there a few years, do good work and then I can move on. And I learned that's a very good attitude, i.e. to encounter places that you don't expect to be paradise. My experience of Germany is that it's a lot nicer than my preconceptions.
I had a pretty good time there and then six, seven years later came this possibility to be part of the process of opening a new Max Planck institute that would be in Eastern Germany where there was a political decision to build research institutes to the same density as in West Germany after reunification.
The Germans had the willingness to ask in what areas of research is Germany particularly weak? And obviously one area where Germany was very weak was genetics, and anthropology in particular, and for a good reason. The Max Planck Society, for example, the predecessor of that was called the Kaiser Wilhelm Society, which existed before the war, and they had an Institute of Anthropology in Berlin, where people like Mengele, who was heavily involved in the Holocaust, had appointments. So since May 1945, the Max Planck society — its successor — of course stayed away totally from anthropology. And that's also reflected in the university landscape. Anthropology, after the war, did not have any high prestige at all as an academic discipline and the quality was correspondingly low.
The creationists are a much smaller issue in Germany than in America. Most of the e-mails and letters that I get in that vein come from America — much more than from Germany. But the little critical comments we have received have come from that quarter. They are objecting to the idea that there would be a common ancestor with Neanderthals or with chimpanzees. I have been able to work with religious fundamentalists, mostly a graduate students in my lab from the Middle East. In the end, I can't argue with someone who says, "God is almighty and anything I perceive or think can be put in my head by God because God can play some game and make me think there is evidence there for evolution." I can't argue against that. I couldn't understand why a God would deceive me in such a way. But if that's the case, then we study God's deception, right? And then we agreed on that with this graduate student, and then we could work in the framework that we started the evidence for this.
It's much more the case in the US that when you give talks or are at evolutionary conferences, there can be organized opposition to what's happening.
There's no place for religion in scientific conversation. Science has to stay science, and religion religion. Obviously religion has an important role in the lives of probably eighty percent of the people on the planet today. And it's an important one. Science need not go out and fight this, but there is no role for religion in science. It's two different realms. I must admit that if I'm confronted with existential questions or life-and-death things, I tend to become irrational and also think in magical ways. Often when we are confronted with things that threaten the foundations of our life, we tend to go for religious or magical or non-scientific ways of dealing with it. That's part of being human.