by Bryan Sykes
13
ADAM JOINS THE PARTY
The story I have narrated in this book is a history of the world recorded in the gene that is the easiest to read, mitochondrial DNA. So far, then, it is the gospel according to Eve. The beauty and simplicity of viewing the record of the past through mitochondrial DNA derive from its unique genetics, and in particular from the clear message that passes virtually unchanged from generation to generation, modified only by the slow ticking of the molecular clock as mutations gradually build up one at a time.
It would be strange indeed if a second, completely different, history were to be encrypted in the other genes that we carry. All these other genes are found on the chromosomes of the cell nucleus. According to the latest estimates, there are just under 30,000 of them. Are there 30,000 different versions of the human past waiting to be read? In one sense there are, because each of these genes could have a different history. Each of them might have a different common ancestor somewhere in the course of human evolution. However, while our nuclear genes have percolated down through time, it is quite impossible to trace all these lines back along a known pathway of descent in the way that we were able to do with mitochondrial DNA. The reason is that, unlike mitochondrial DNA, the nuclear genes are inherited equally from both parents. While you have only one mitochondrial ancestor in the last generation, your mother, you have two nuclear ancestors, your mother and your father. That doesn’t sound too complicated. But go back one more generation. Now you have four nuclear ancestors, your grandparents; but still only one mitochondrial ancestor, your mother’s mother. Go back another generation and there are eight nuclear ancestors, your great-grandparents; yet still only a single mitochondrial ancestor, your grandmother’s mother. At each generation the number of nuclear ancestors doubles. Go back twenty generations, to about AD 1500, and there could be, theoretically, over one million ancestors who could have contributed to your nuclear genes. In practice, many of these potential ancestors will actually be the same individuals, whose lines of descent have come down to you along different pathways, crossing between males and females through the generations in an unpredictable way.
Tracing the genealogy of all 30,000 genes through this maze of interconnections would be quite impossible. Add to that the confusion introduced by recombination, and the magnitude of the task becomes mind-numbing. The shuffling of chromosomes at each generation means that any one gene might itself be a combination of one part from one ancestor and another from someone else. Reading the different individual versions of human history from these genes, and bits of genes, in the cell nucleus is impossibly complicated at the moment. It will take a long time to advance beyond the kind of crude summaries of human history that we already have from the days of gene frequency comparisons.
However, one gene – or, more correctly, one chromosome – is immune from these ghastly complications. It is called the Y-chromosome, and it has only one purpose in life: to create men. By comparison with the other human chromosomes it is small and stunted, and it carries only one gene which really matters. This is the gene that stops all human embryos from turning into little girls. Without a Y-chromosome, the natural course of events is for the human embryo to develop into a female. If an embryo has a Y-chromosome, and if the gene, which has been given the undistinguished name SRY, is working properly, then it will trigger a number of other genes on different chromosomes to steer the development of the embryo away from becoming a female and towards becoming a male. The SRY gene activates genes on other chromosomes which suppress the development of ovaries and instead promote the growth of testes and the production of the male hormone testosterone.
Two observations pinpointed the key part played by the SRY gene in sex determination. Very rarely, in something like one in 20,000 births, a girl is born with a Y-chromosome. These girls look normal, they have normal intelligence and they develop normally, though they are usually slightly taller than average. But at puberty their ovaries and uterus do not develop properly, and they cannot have children. Genetic analysis of the Y-chromosomes of these girls shows that the SRY gene is either missing altogether or contains a mutation that stops it working properly. The other piece of graphic evidence that the SRY gene is itself sufficient to make a male came from research on mice. Male mice have Y-chromosomes too, and they carry the mouse equivalent of the human SRY gene – called, in a burst of imaginative classification, Sry. In a very elegant genetic engineering experiment, the Sry gene was cloned from a male mouse and transplanted into a fertilized mouse egg that would otherwise have turned into a female. Despite the fact that the mouse embryo had only the cloned gene to work on, rather than a complete Y-chromosome, it turned into a male.
So this is how the sex of a baby is determined. Fathers, being male, have a Y-chromosome. Half of their sperm contains his Y-chromosome, carrying the SRY gene, and the other half carries another chromosome – the X-chromosome – instead. The sex of the baby depends entirely on whether or not the particular sperm that fertilizes the mother’s egg contains an X-or a Y-chromosome. If the successful sperm carries an X-chromosome, then the child will be a girl. If it carries a Y-chromosome instead, the child will be a boy. The woman has no influence whatsoever on the sex of the child. How many women in past centuries would have loved to know this simple fact? How often was the ‘failure’ to produce sons attributed to a failure, deliberate or not, on the part of wives to conceive boys?
Just as mitochondrial DNA follows a maternal genealogy through the generations, the inheritance of Y-chromosomes by sons from their fathers should trace the mirror-image paternal pathway from one generation to the next. If the Y-chromosome could be genetically typed, and if it were not involved in recombination that would scramble the message, then there was good reason to believe that it would be the perfect complement to mitochondrial DNA in reading the history, not of women, but of men. The Y-chromosome, in common with all the chromosomes of the nucleus, is a very long, linear molecule of DNA. While mitochondrial DNA has just over sixteen and a half thousand bases in its DNA circle, the Y-chromosome stretches for about sixty million bases from one end to the other. It might be the runt among human chromosomes, but it still packs more than four thousand times as much DNA as mitochondria. Moreover, there is some gene shuffling within it. At the tips of each end of the Y-chromosome there is a section of DNA that recombines with the X-chromosome; but since these sections involve less than 10 per cent of the whole chromosome, this doesn’t present a great problem. Genes that are on the recombining part of the Y-chromosome will trace a mixed genealogy, swapping unpredictably from males to females just like all the other nuclear genes. However, the remaining 90 per cent of the Y-chromosome, between the recombining ends, is not scrambled. This long segment travels intact through the generations. But are Y-chromosomes different from one another, and if so how do they differ? Only if there were variety and diversity in the Y-chromosome would it be of any value at all for reading human history. If all Y-chromosomes were exactly the same, they would be no use for our purposes.
Chromosomes are intensively studied under the microscope by trained cytogeneticists in medical genetics laboratories who are on the lookout for abnormalities that can diagnose inherited diseases like Down’s syndrome or explain the cause of infertility. With all this activity going on, cytogeneticists had noticed that some Y-chromosomes stood out as being much longer than the average. This was promising; but it was not a very precise way of differentiating between Y-chromosomes on a large scale. Besides, the lengths were unstable and changed between one generation and the next. What was needed was the same kind of testing involving Y-chromosome DNA that had identified mitochondrial DNA as such a star. Then it would be straightforward to fingerprint Y-chromosomes from hundreds or thousands of volunteers easily and cheaply. But how were the segments of Y-chromosomes that were going to show the biggest differences among people to be found?
The rich diversity of the mitochondria is concentrated in a small DNA circle of only a f
ew thousand bases. Better still, the control region compresses about a third of the diversity of the whole mitochondria into just five hundred bases that can be sequenced in a single run on an automated sequencing machine. Would something similar be found in the Y-chromosome? The answer was not long in coming. Several labs, hoping for the best, began to look for differences between Y-chromosomes by sequencing through the same segment of Y-chromosome DNA from volunteers who were as distantly related to one another as possible. In one of the first studies, 14,000 bases were sequenced from the Y-chromosomes of twelve men from widely different geographical origins. Only a single mutation was ever found. If an equivalent 14,000 bases had been taken from mitochondrial DNA instead of the Y-chromosome, they would have shown dozens of mutations in the same number of people. Another lab sequenced a 700 base segment of one gene from the Y-chromosomes of thirty-eight different men without finding a single difference in any of them!
This was all rather depressing for the scientists involved (thankfully, I wasn’t one of them). There was a lot of head-scratching. Why were Y-chromosomes so similar all around the world? Since Y-chromosomes didn’t carry any genes to speak of, and were full of ‘junk’ DNA which had no obvious function, the expectation was that there should be more, not less, variation on the Y-chromosome than on regular, gene-rich chromosomes. Mutations are free to accumulate in ‘junk’ DNA because it doesn’t do anything, so its precise sequence doesn’t really matter. Most mutations that occur in genes which do have important functions interfere with their proper working and are soon eliminated by natural selection. It was certainly a puzzle to find that there were so few mutations on the Y-chromosome.
The most popular theory advanced to account for this lack of variation was that it had to do with the fact that, under the right circumstances, men can have a lot more children than women. If, in the past, only a few men had lots of children, and therefore lots of sons, their Y-chromosomes would spread around quickly at the expense of the Y-chromosomes of their unfortunate male contemporaries who had few children or none at all. If this had happened a lot, the theory went, there would be far fewer different Y-chromosomes around today than if all men had roughly the same number of children. It’s true that there have been some particularly prolific males. The world record holder is Moulay Ismail, Emperor of Morocco, who is alleged to have had 700 sons (so presumably as many daughters) by the time he was forty-nine in 1721. He died in 1727 – so there was another six years to have some more. The most prolific woman comes way behind this. She is Mrs Feodora Vassilyev, a Russian woman who produced sixty-nine children between 1725 and 1765. They were all multiple births – sixteen pairs of twins, seven sets of triplets and four lots of quadruplets – so she was a remarkable woman in that respect as well. The capacity of women to produce large numbers of children is limited by their biology, which restricts them to one pregnancy a year at most. Men, on the other hand, are not restricted by this timetable and can, in theory, have thousands of children. But the fantasy of enormously prolific males seeding the entire world, thereby reducing the diversity of Y-chromosomes by their prodigious feats of polygamy, turned out to be just that. A fantasy. A hard slog in laboratories around the world over the past ten years has found that there are plenty of mutations on the Y-chromosome after all.
These mutations come in two main types. The first is exactly the same as those we are already used to seeing in mitochondrial DNA: the simple change from one base to another. However, unlike in mitochondria, where they are neatly compressed into the control region, these mutations are spaced out at irregular intervals right along the length of the Y-chromosome. This is a practical nuisance because each one has to be tested individually, but it is not an insuperable obstacle. The other type of mutation is very uncommon in mitochondria, although we did encounter one example in the Polynesian samples. That is where there was a deletion of nine bases from the mitochondrial DNA circle. A careful look at the DNA sequence around that region revealed that in fact this wasn’t so much a deletion from the Polynesian mitochondrial DNA as a doubling up, a duplication, of that nine-base segment in the rest of us. This type of mutation, where short segments of DNA are repeated over and over again, is remarkably common in the nuclear chromosomes and, thank heavens, in this respect the Y-chromosome is no exception. Dozens of these repeated segments have been discovered on the Y-chromosome, and the difference between individuals lies in the number of repeats. Fortunately, this is an easy thing to measure. This rich source of variation suddenly revealed that there are thousands of different Y-chromosomes around that can be distinguished from one another on the basis of these two sorts of mutation. Genetic fingerprinting of Y-chromosomes has become a reality.
Because it has been such a struggle for the scientists involved to find the useful mutations, laboratories have been very careful about whom they tell when they find a new one. As a consequence, labs have organized themselves into rival cliques which have used different sets of mutations to fingerprint Y-chromosomes; there is not yet a common standard. This means that there are different evolutionary networks being produced by the different confederations of laboratories. This is only a temporary situation, and I hope and expect that in the near future these will be reconciled into a scheme which everyone can accept. But how is it looking up to now? In particular, does the history of Europe revealed by the Y-chromosome bear any resemblance to the one read from mitochondrial DNA which forms the basis for this book? Does the Y-chromosome version of events agree or disagree with the mitochondrial DNA in placing such a great emphasis on the Palaeolithic as the source of our genetic legacy? In other words, does the history of men agree with the history of women? The answer came in an article published in the 10 November 2000 edition of the journal Science.
‘The genetic legacy of Paleolithic Homo sapiens in extant Europeans: a Y-chromosome perspective’ was the culmination of a large collaboration between scientists from Italy, eastern Europe and the United States. I had been asked to comment on the paper by the BBC on the day it was published, and had a copy faxed through to the Royal Society in London where I was at a scientific meeting. As soon as the fax arrived I took it into one of the drawing rooms which overlooked St James’s Park and sat down. My heart sank as I went through the long list of authors at the beginning of the paper. There, second from the end, was the name L. Luca Cavalli-Sforza. After all the battles of the previous four years, I could hardly expect my old adversary to agree with me at last.
Reading through the article, I could see that it was constructed along generally similar lines to our mitochondrial paper of 1996. They had fingerprinted the Y-chromosomes of 1,007 males from twenty-five European and Middle Eastern locations. Then, just as we had, they had drawn an evolutionary framework and identified clusters. They discovered ten Y-chromosome clusters rather than the seven that we had found with mitochondria. Then they had estimated the ages of these clusters, as we had done, from the accumulated mutations within each one. I turned the pages with growing excitement. What were the ages of these clusters going to be? Would they be mostly in the Palaeolithic, like six of the seven mitochondrial clusters? Or would they be much more recent, in the time of the Neolithic and the early farmers? I certainly knew what I expected the paper to say, given Luca’s prominent position as an author and his well-known views on the magnitude of the genetic impact of agriculture. The paper was full of dense statistics but there, on the penultimate page, my eye went straight to the vital paragraph. It began: ‘Analyses of mitochondrial DNA sequence variation in European populations have been conducted,’ and it referenced our 1996 paper. ‘These data suggested’, it continued, ‘that the gene pool has about 80% Palaeolithic and 20% Neolithic ancestry.’ That was fair. I read on to the next sentence, expecting it to begin the demolition of our position. But, it did not. Instead, I read the words: ‘Our data support this conclusion.’
I couldn’t believe it. The tension drained from my body. The battle was over. We had been put through the wringer for four
and a half years. We had endured the panics about the mutation rate being wrong, about mitochondrial recombination messing everything up, and about the control region being completely unreliable. And now it was over. Mitochondrial DNA and the Y-chromosome told the same story. The history of men tallied with the history of women. Luca and I could finally agree. It had been a tough battle, but a fair one. The Neolithic farmers had certainly been important; but they had only contributed about one fifth of our genes. It was the hunters of the Palaeolithic that had created the main body of the modern European gene pool.
14
THE SEVEN DAUGHTERS
From the remains in Cheddar Gorge we had extracted direct proof of the genetic continuity between people living today and the hunters of the Upper Palaeolithic. We now knew that this unbroken thread, accurately and faithfully recorded in our DNA, stretched back beyond the beginnings of history, beyond the ages of iron, bronze and copper to an ancient world of ice, forest and tundra. Only the exceedingly slow beat of the molecular clock separated the DNA we found in Cheddar Man from the DNA in our two utterly modern descendants Adrian Targett and Cuthbert the butler. The evolutionary reconstruction we had done on the DNA from thousands of living Europeans had pointed us to that conclusion, and eventually we had found physical evidence to validate it. Now we also had the crucial endorsement from another genetic system altogether, the Y-chromosome, of the assertion that our genetic roots do indeed go back deep into the Palaeolithic.
Our reconstructions had identified seven major genetic clusters among the Europeans. Within each of these clusters, the DNA sequences were either identical or very similar to one another. Over 95 per cent of modern-day native Europeans fit into one or other of these seven groups. Our interpretation of European prehistory and the emphasis it placed on the Palaeolithic hunter–gatherers had depended on giving ages to these clusters, and we had worked these out by averaging the number of mutations we found in all the modern members of the seven different clans. This gave us a measure of how many times the molecular clock had chimed within each clan. Knowing the rate at which the clock ticked, we could then work out how old each clan really was. Old clusters had accumulated more changes over the millennia. The molecular clock, slow as it is, would have struck more often. Young clusters, on the other hand, would not have had as much time to accumulate as many changes, and the DNA sequences of people within a young cluster would be more alike.
