Ancient Egyptian DNA, Cheddar Man and wolves as pets
Professor Brian Cox takes a deeper dive into the science of ancient DNA, with two returning experts answering more of your questions.
How accurate is ancient DNA when it comes to understanding our past? Why was diversity in DNA vital for the survival of early humans? Do we really share 60% of our genetic code with bananas?
More Questions of Science gives you extra episodes that continue the conversations started in A Question of Science.
From his home studio, Professor Brian Cox catches up with expert panellists from earlier in the series. Together, they tackle your follow-up questions, explore new angles and go deeper into the topics that sparked your curiosity.
Brian: Hello and welcome to this bonus episode of A Question of Science, the show where we tackle your questions on everything from Artificial Intelligence to the future of fertility, with the help of a panel of leading experts. The original debates were recorded in front of a live studio audience at the Francis Crick Institute in London, and this is a follow-up show and it's a chance to hear some of your feedback and answer some more of your questions.
Now, in today's conversation, I'll be travelling back in time to learn more about how ancient DNA can help us unlock the mysteries of modern life. And to do that, I'm joined by Pontus Skoglund, group leader of the Francis Crick Institute's Ancient Genomics Lab, and Dr. Tom Booth, an archaeologist and senior research scientist at the Francis Crick Institute.
Thank you both for coming back. I'd like to start with a fun question from Arianna Sophea, who wants to know: If you could speak to just one ancestor from our evolutionary past, who would be the most valuable to hear from and why? So, Pontus?
Pontus: Right now, I'm thinking a lot about the origins of ancient Egypt. We sequenced the first genome from ancient Egypt last year, and one of the earliest mentioned people in history is King Narmer who unified ancient Egypt. And I'm not sure he was such a sympathetic character, but I would be interested in talking to him and hearing what he has to say about how that happened.
Brian: So, so when is that? Is that, that's The Old Kingdom. Is that 3000 BC? Something like that?
Pontus: Yeah, 3000 BC, something like that. He founded the first dynasty. His reign was the formation of the early dynastic Egypt, and it was probably a process of territorial conquest that we see in our world today as well.
Brian: Going back to the original question, I suppose it's who would you like to speak to, but also scientifically speaking, who would you like to speak to?
Tom: Yeah, it's hard not to get dragged into sort of thinking more about who I would like to. I mean, you get dragged to these interesting cases like Pontus mentioned.
I mean, Pontus was also involved with this study of an ancient canine, 5,000 year old canine from Gotland, which is this island in the Baltic Sea, which turned out to be a wolf and Gotland, because it's an island, wouldn't naturally have had wolf populations on there. So, clearly the wolf had been taken across to Gotland by a human.
So what madness drove someone to get a wolf, put it in a boat and cross the Baltic Sea to get to Gotland and what they thought they were doing in doing that. I mean, clearly it's got something to do with the understanding of the domestication of dogs. But that would be interesting to understand the thought process.
The first person who decided it was a good idea to drink milk from an animal, because obviously humans are the only species that naturally drink the milk from another animal at the moment. At what point did someone think this is a good idea, you know, approach a horse or a sheep or a cow and thought, "oh, why don't we just have a go at drinking this?"
You know, because that did set particularly people in Northern Europe and in East Africa onto an evolutionary trajectory that led to lactase persistence, our ability to digest milk sugars beyond infancy. So again, that's another one where, you know, you could come up with a, do you realise that you are gonna sort of profoundly change the genetics of people in the future by doing this?
Brian: The answer is everyone, from every stage throughout all of human history. Nice specific answers. And let's turn now to ancient DNA and a question from listener Annabel Harrison, who was curious about the limits of studying mitochondrial DNA. And I thought it might be useful, Pontus, to start by explaining the difference between nuclear DNA and mitochondrial DNA and why mitochondrial DNA is only inherited through the maternal line.
Pontus: Mitochondria are a little compartment in our cells that millions and millions and millions of years back, it was a microorganism that came into our cells very early on in life. And it's just transmitted from the egg to the embryo, and so that means that it's only transmitted from mothers, not fathers, in general at least. And so while, with ancient DNA, we want to study the big family tree of everyone, the two parents of everyone, who ever lived thousands of years going back or get as close to that as possible, mitochondrial DNA only traces the maternal line. So, our mothers, their grandmothers, the great-grandmothers and their mothers, and their mothers and so on. But as everyone realises, the family tree is much bigger than that. In fact, our genomes, our full DNA trace tens of thousands of such paths that don't just go through the maternal line, and that's why it's much more useful.
In the past, people would study mitochondrial DNA for pragmatic reasons, I would say. There are 200 copies of mitochondrial DNA for every copy of our normal nuclear DNA. And so, it's just easier to find in ancient DNA, we find it in more abundance. It's also kind of smaller, it's at 16,000 letters of this ATGC - so kind of a constrained system to study for evolution. But we want to study the 3 billion base pairs in our full DNA to trace tens of thousands of lineages going back into history.
Tom: In the, kind of, 2000s, a lot of people tried to infer the population history of places by looking at mitochondrial DNA in modern populations, and looking at how certain populations in different parts of the world shared mitochondrial DNA. From that, they try to construct population histories. And what's clear, with new ancient DNA techniques, is a lot of those histories that were constructed in the noughties were wrong, because they were only looking at this tiny portion of the genome. But that doesn't mean that mitochondrial DNA can't be useful. So, for instance, one of the things we like to look at is in cemeteries, single cemeteries, the different, both mitochondrial DNA and Y-chromosome DNA - which refers to the paternal lineage - and how that's represented in particular cemeteries, in ancient cemeteries, it can tell us potentially about the social organisation of past societies. So for instance, if everybody in a particular cemetery has the same mitochondrial DNA, but different paternal lineages, it might suggest a matrifocal system, where people who were descended from particular women were being buried in these specific cemeteries. So, these kinds of things are still useful.
Brian: We do actually have a question from Abi Su, which is related to this question of the accuracy of these techniques, how far we can go back. So, Abi Su asked: How can we be confident that the Cheddar Man is a direct ancestor of a man living in Cheddar today,
given how much harder it is to establish familial connections from thousands of years ago using ancient DNA, than recent ones with modern DNA? So, maybe Pontus?
Pontus: Yeah, this meme goes around that the Cheddar Man is an ancestor, direct ancestor, to someone in a village nearby him, but it clearly has no scientific basis at all.
We talk in genetics about degrees of relationship, genetic relationship, so we have one degree to our parents, two degrees to our grandparents, three degrees to our great grandparents, but Cheddar Man lived nine or ten thousand years ago, and that's over 300 generations ago. We don't have the resolution to tell if someone has 300 degrees of relationship to him, and even if we did, it would be hard to say that he's the direct descendant, right?
It starts to become really difficult around ten degrees back into the past. And that's all about this individual descendant-ancestor relationship. And that's also usually not what we focus on in science. I think it's often more interesting to think about population level questions. For example, there is a real question, what proportion of someone living today's ancestry can be traced to the group that Cheddar Man was part of, for example, in Britain
10,000 years ago? In fact, the answer to that is probably very small on the order of 1% or maybe even less because more recent people came into Europe and Britain and changed the landscape, resulting in this group not contributing very much ancestry, I think, and that's a common feature we see in genetic history all over the world.
But in a broader set, if we ask the question, if someone in Europe today, what proportion of their ancestry is from people living in Europe 10,000 years ago, that Cheddar Man was also really similar to, it becomes more I think, maybe about 20% or on the order of that, that people can trace their ancestry to people in Europe 10,000 years ago. And 10,000 years ago, the remainder of their ancestry would be found on the Eurasian steppe, primarily, and in Anatolia and early farmers.
Tom: Yeah, this is a bit of a zombie factoid that refuses to die. I was involved with this study of Cheddar Man that came out in about 2019, and it implied what Pontus is saying, that there's very little genetic continuity between the people who lived in Britain back then and people who live in Britain today or even people who lived a few thousand years later.
But yeah, I just wanted to say that they share potentially a mitochondrial haplogroup, this maternal lineage, but it's much less likely that the guy who's lived in Cheddar is the direct maternal descendant of Cheddar Man. Most obviously, because Cheddar Man was a man, he couldn't pass on his mitochondrial DNA anyway.
But yeah, it's much more likely that they share a descendant that was quite far back, even in Cheddar Man's past. And that's how the mitochondrial DNA has been inherited. I mean, it is not even that specific to Cheddar. So, this haplogroup that they share is pretty common amongst sort of modern groups in Europe and Britain more generally. So it's not even particularly specific to Cheddar.
Brian: Continuing this theme of lineages, it is rather the complicated, isn't it? I mean, we often hear the statement that 2% of our genome traces back Neanderthals, but also we hear people say, but 60% of our genome we share in common with bananas. So, can you clear that up?
Tom: When you think of human DNA as being the sequence of 3 billion letters, you know, the vast majority of those letters, even between humans and Neanderthals are the same. So what we look at are these specific subset of those letters, which is a, a small proportion of those letters, which we know commonly vary between human groups, between human populations, between potentially humans and archaic hominids like Neanderthals, and it's sort of the 2% of those letters that we know that we can trace back that seem to be in common with Neanderthals that we've inherited from Neanderthals. But that has implications, that's a reflection of the whole genome to some extent of what we've
Pontus: inherited for our whole genome from those.
The banana number comes from lining up the DNA strings of a banana and a human, and if you line them up and 60% is the same, but that depends on a lot of things. If you line up the DNA of a modern human and a Neanderthal, 99.5 or something like that is the same, but that number depends on how much variation there was in our ancestor or common ancestor with Neanderthals and our common ancestor with bananas, but that's those numbers. The separate number 2% of many people on the planet,
of
Pontus: their ancestry can be traced to Neanderthals a hundred thousand years ago, but a hundred thousand years ago 0% of their ancestry can be traced to bananas.
Brian: Actually, we, we heard in the debate that, just to underline how complicated this is, it seems there's an unidentified species, neither Homo sapien, nor Neanderthal that shows up in ancient DNA of the Denisovan ancestors. Pontus, could you speak more about that discovery and what it means?
Pontus: It's a fascinating, I think, question quite technical, but when you line up and we model the history of people today, Homo sapiens, Neanderthals, and Denisovans, the big picture is that Neanderthals and Denisovans are closer related, they share the common ancestry that Homo sapiens don't share. That's the big picture when you compare the DNA. But actually, Neanderthals are a little bit closer to Homo sapiens, it doesn't just fit a simple tree. And then there are multiple ways to solve that. One of which is that Neanderthals and Homo sapiens had some more recent contact that Denisovans weren't part of.
And the other way is that Denisovans have deeper ancestry from perhaps Homo erectus, or something like that, in East Asia, for example. That has been the leading, I think, hypothesis for a while, but also the secondary contact between Neanderthals and Homo sapiens seems increasingly likely as well. And so perhaps both of the things are true.
Tom: Yeah, so the population that led to humans around the world today wasn't potentially the first group of modern Homo sapiens that migrated out of Africa. Increasingly, it looks like there was earlier migrations of modern Homo sapiens, which went into various different areas and essentially, to all intents and purposes, died off.
So, were unsuccessful and there's some indication that one of these migrations, they interacted with Neanderthals, and Neanderthals sort of absorbed some of their DNA and then they died off. Yet the Neanderthals persisted, so then when the sort of human migrations that led to people around the world today sort of spread out of Africa, they encounter Neanderthals that already had some admixture from these early humans. So it gets fiendishly complicated where you have all of these dispersals and mixing events and then populations which essentially died off. And as Pontus said at the beginning, we're talking about this world where you have multiple types of humans existing all around the same time, and people have talked about as being slightly Tolkien-esque fantasy of these different types of small and large humans that are in contact and interacting with each other. So, you know, that model to some extent is beyond what anyone really ever imagined before ancient DNA sequencing technology. And yet, even when you take into account potentially this mixing with earlier dispersals of humans, there's still this possibility of there being some admixture between Denisovans and what's called this super archaic species, but who that species is, it's still a compelling mystery at the moment as to what's going on.
Brian: And you mentioned there actually this, you called it a Tolkien-esque kind of picture. So we all have these images of what these ancestors look like, the Denisovans and the Neanderthals.
There's a question from a listener on YouTube actually, about how accurate our picture of what these species, subspecies, what they look like just given the DNA.
Pontus: There's very amazing work looking into how different mutations that they had might give rise to different biology, and it's a very fascinating research field.
But, you know, while we all know that our DNA encodes how we look like, for example identical twins look very similar as everyone knows. You know, science overall struggles. We can't exactly reproduce how people look like just from the DNA, but it might be possible in the future. Then there's an issue, of course, that these people lived in the past and might have slightly different makeup. So, I think fossils are still our best way to understand this. And actually, Denisovans, who for long were a mystery, they didn't have any big skeletal fossils associated with them, last year the first link between a skull and Denisovan DNA was made with this Harbin skull in China, where they got mitochondrial DNA, which in this case is probably enough to say that this big skull was a Denisovan. And it's really big, it's bigger than probably 99.9% of people on the planet. If you met this person, they would seem very large and and strong.
Tom: Yeah, I mean, the problem is as well is that we can look at, and when people do look at, variants that have been inherited from archaic species like this 2%, Neanderthal ancestry, and then look at the physical traits that these variants are associated with in living people.
But the problem with that approach is that we don't really know whether in living people those variants have the same kind of effect as they do in the ancient groups. So we can see that particular variants might be associated with things like cranial shape or pigmentation and things, but then we don't know whether those genes or those variants would've functioned in the same way in the ancient hominins.
Brian: So someone else asked about genetic bottlenecks, which we discussed on the programme I think, genetic bottlenecks and how they affect our ancient DNA. So Tom, maybe you could start by explaining what a genetic bottleneck is?
Tom: A genetic bottleneck is when the proportion of the population that are reproducing, so the number of people who are reproducing, drops to a low number. So it's shaped like a bottleneck. And this can be, or is most often to do with, what people infer to be broader population declines, although it can be other things.
For instance, if a society decides that a particular part of their society, segment of society, are gonna have children amongst themselves, that can produce a genetic bottleneck within that section of society, because you are lowering the kind of diversity of the breeding population, the people are having children.
So, there's lots of different things that could produce these bottlenecks, but particularly population decline is one of the main ones.
Brian: Pontus, I was also gonna ask about, we hear sometimes about real, sharp reductions in the population of humans or our ancestors, associated sometimes with volcanic eruptions and so on.
So could you give us some examples of how small the population of our ancestors became at various times in history?
Pontus: I think the main one we're still thinking about a lot is when people moved outside of Africa or expanded outside of Africa, people were still in Africa. There's a clear lower diversity in the individuals we see even 45,000 years ago outside of Africa.
In fact, a person with recent African ancestry, their two chromosomes that they have from their mother and father has more differences between them than if you compare the chromosomes of people from two different corners of for example, Europe and Oceania or the Americas. And we think that's due to probably some type of reduction, when people moved out of Africa. But there's a twist to that, because these bottlenecks are not the only thing that can create differences in diversity. Also mixing between groups that have been isolated for some time will result in diversity. So in the past, people used to think that the fact that in modern people it looked like diversity was going down stepwise, the further you got from East Africa, that that was a result of an expansion that was kind of serial
outside of Africa. But in fact, when we look at ancient DNA, the picture is not as clear. In fact, people in Europe 10,000 years ago, such as Cheddar Man, had lower diversity than people further away from Africa. And so this mixing really also can create diversity patterns. It's really not that straightforward to be really sure that they are due to bottlenecks.
Brian: We sometimes read, you know, that the population dropped to, I don't know, 10,000 individuals or breeding pairs or whatever, is that rooted in evidence or fact now, or is that just...?
Pontus: We take that with a very big grain of salt. I would not look much at those numbers at all, and they could easily be ten or a hundred times bigger. Even if it's partially true, just the difference between the census size and what you can possibly model with mathematics, but also they don't take into account different groups and structure between them.
Whenever we extrapolate far back into the past, further than we have DNA, it becomes really difficult. When we can be quite sure of things, it's when we have direct evidence, ancient DNA, and we can compare ancient DNA directly between each other.
Brian: We've almost run out of time, but I suppose language, it's one of the most, most human of attributes with its complex language.
And so the spread of complex language across our populations historically is extremely important. And so there is a question about how DNA evidence can help us understand the spread of language, if at all.
Tom: There's one particular language family that ancient DNA has really been brought to bear on, and that's the Indo-European language family.
And it's long been recognised, since the 19th century, that groups of languages from quite dispersed parts of Europe and Asia have some words in common with one another, suggesting that they have a common root. And language works a little bit like DNA in that it is inherited and, I mean, this is quite surprising in some ways, but there's aspects of language that change in a similar way to the way the mutations work in DNA, that that are predictable.
So even though the words, what they call cognates, these words that are essentially for the same things, but different words in different languages are different from each other. By looking at the sound changes that are likely to have occurred through time, you can reconstruct the original root word and linguists and archaeologists are trying to pinpoint the origin of where this root language was originally spoken, by looking at what these root words meant and linking that to the archaeological evidence. So one of the things that this language has is a lot of words for vehicles, so vehicles with wheels, and the earliest evidence we have for wheeled vehicles that vehicles is from about in Europe is from about 3,500 BC.
So we know that that puts like an earliest possible date for those languages occurring. And then it looks like that they have borrowed certain words from this other family of language, Uralic languages, which puts them somewhere kind of approximate to those groups early on in maybe somewhere in Eastern Europe.
A lot of their words for agriculture or agricultural product, cereals for instance, are borrowed from other languages. And their own words for cereals are quite vague, which suggest that they were familiar with cereals, but probably didn't plant cereals themselves. But they have lots of words for domesticated animals suggesting that they were probably pastoralists without any developed cereal agriculture themselves.
And all of these things brought together put their origins probably somewhere north of the Black Sea on the Pontic Caspian steppe with civilizations that lived there around you know, 3300 BC. This was a much debated topic, but the ancient DNA came along and found that there was evidence for ancestry from that region spreading across Europe and then eventually into South Asia from around 3000 BC.
So it perfectly fitted, or pretty perfectly, with this development of the distribution of Indo-European languages today. And even today, generally speaking, people who speak Indo-European languages generally have some portion of their ancestry that's derived from these people who lived on the steppe.
Brian: So it's a correlation between the population spread inferred from the DNA evidence and then the spread of words, essentially that you're looking for. Pontus, I wanted to pick up, it brings us really neatly back to where we started about your conversation with someone from the Old Kingdom of Egypt. So, is there a crossover here?
We're about the same time, right? We're about 3000 BC or so. So, would there be common words? Would there be a way of you beginning a conversation by pointing to a wheel, or something like that?
Pontus: Yeah, that would be a, probably a different language family. Ancient Egyptian was a different language family from Indo-European, but there's a very interesting thing there,
so there's spoken language, but there's also written language and the idea of human writing, which we have the first evidence from, almost simultaneously from in Mesopotamia - present day Iraq - and ancient Egypt, which are still quite far apart in geography. And so people have wondered how did that happen? Did they get the idea simultaneously?
And actually, in this first person, when we sequenced this first ancient genome from ancient Egypt, 4,700 years old, we saw a genetic signature that suggested a strong link with Mesopotamia in part of his ancestry. About 20% of his ancestry in this particular model we found fitted with Mesopotamia, but we didn't have any other individuals from earlier in Egypt, so that number could very well change. But there's definitely something indicating a link there. So could this be ancient DNA, adding another piece of evidence of contact between these two different places that could help us understand the spread of early writing alongside what archaeology and history has added?
And I think more ancient DNA will be very interesting to look at.
Brian: Yeah, thank you. Unfortunately, we've run out of time. This is fascinating, we could go on for another hour, easily. But actually before we go, I should say, if you're listening to this and are as interested as I am, then the Francis Crick Institute is launching an immersive exhibition on ancient DNA in July, that's July, 2026 because you might be listening to this in 2045 or something. You can find out more about the research that scientists like Pontus and Tom are doing, and discover how we're all connected to each other, and to the very earliest humans, on the Crick's website. But for now, goodbye.
