Pandora's Seed Page 2
GENETIC BEADS
I paid the cab driver and got out near the University of Chicago bookstore, taking in the surroundings. Gothic-style edifices, constructed during Chicago’s earlier building boom, toward the end of the nineteenth century, surrounded me on all sides. It had been a conscious attempt on the part of the new university—it was founded in 1890, with funds provided by the oil baron John D. Rockefeller—to connect with an older tradition of learning. I felt as though I were back among the gleaming spires of Oxford, running between undergraduate tutorials. My destination, however, was a much newer structure.
The Cummings Life Science Center was constructed in 1970; as befitted a structure meant to house scientists engaged in the advanced study of biology, then undergoing a revolution as a result of Watson and Crick’s elucidation of the structure of DNA, the building’s brick tower was bracingly modern, even a bit brutal. But I had come to talk to Jonathan Pritchard, who was using the most advanced techniques in genetics to look at the history of our species. The juxtaposition of this building amid a campus of older structures seemed fitting, given what I was here to discuss.
I located his office on one of the upper floors, and we chatted as he made me a cup of tea. An avid distance runner, with the intense, lanky look of a marathoner, he seemed somewhat surprised that I had made the trip just to talk to him. I asked him about his move from Oxford to Chicago, his personal life (one of his son’s drawings hung above his desk), and what it felt like to have been granted tenure at one of the world’s most prestigious universities at the precocious age of thirty-seven. He laughed, confident in his intellectual abilities, like so many of the mathematically gifted people I have known, and explained that his life was going well. We then moved on to the reason for my visit.
I wanted to talk shop. Or, rather, I wanted to get his take on the findings of his important research paper. In their PLoS publication, he and his colleagues had described a new method of detecting selection in the human genome. It made use of something called the HapMap, a collection of data on the so-called haplotype structure of the human genome. And to understand that we’ll need to delve into the science a little.
The long string of DNA that makes up your entire genome is broken into smaller strings called chromosomes—there are twenty-three pairs of them—containing the 23,000 or so genes that direct your body to do what it does. These genes code for things like sugar-digesting enzymes in your gut, or blood-clotting proteins, or the type of earwax you have—all of the physical traits that make you who you are. The chromosomes are linear strings of DNA, composed of four chemical building blocks known as nucleotides: A, C, G, and T. The sequence of these nucleotides—AGCCTAGG, and so on, along the entire length of the chromosome—encodes the information in your genome and determines what each gene will do in your body. The nucleotides are arrayed along the chromosomes like beads on a string, a linear orchestra of musicians, each playing their own part in the symphony that is you. You get one of each of your chromosome pairs from your mother and one from your father.
Something funny happens to these musical beads, though, as they are passed from your parents to you. They shuffle—like a deck of cards—partially mixing up the original linear strings of beads your parents had. That’s right: your parents’ chromosomes literally exchange genetic information along their lengths, breaking and reconnecting their paired strands to produce a completely new version of a chromosome to pass on to you. This is part of the reason why you don’t look identical to other members of your family, but we don’t know exactly why it occurs. The best theory going is that it’s probably a good thing to generate novel chromosomal arrangements of the musical beads in each generation so that your child’s DNA orchestra can play a different tune if times change—think about having to evolve quickly in times of intense climatic upheaval. As it’s pretty much ubiquitous in animals and plants, there’s almost certainly a very good reason it’s there.
Probably a few readers are wondering at this point, “If the chromosomes are paired, then why does shuffling change anything? Surely they are copies of the same beads, so wouldn’t shuffling them just produce the same combinations in each of the two new chromosomes?” The reason for the new combinations is that each member of a pair is actually a slightly flawed version of the other. As the chromosomes get passed down through the generations, they have to be copied by the cellular machinery for each new organism. Although this is done with great care, and there are proofreading mechanisms to make sure the copied beads look like those on the original strand, occasionally a mistake is made. By chance, one color of bead is substituted for another—a red for a green, for instance. It doesn’t happen very often—perhaps a couple of times for each chromosome in every generation—but when it does happen, these changes, which geneticists call mutations, get passed down through the generations. They serve to introduce additional variation into the gene pool. Over time the changes have accumulated to such an extent that, on average, one in every one thousand beads differs between the chromosome pairs. Thus, each chromosome that is passed on is a shuffled version of Mom’s and Dad’s chromosomes, with the shuffling detectable through the patterns of the variable beads. It sounds very complicated in theory, but if you think about it as beads on a string it is a bit easier to grasp.
FIGURE 1: RECOMBINATION CREATES “SHUFFLED” CHROMOSOMES OVER TIME.
What the HapMap project did was to assess the way the beads had been shuffled in different human populations. By looking at people from Africa, Europe, and Asia, it deduced that there was an average length to the sections of the string of beads that hadn’t been shuffled. The length was a function of how old the population was, the average size of the population over time, and other factors that helped to determine exactly where on the string recombination could have occurred. The math behind all of this gets pretty tricky, but the take-home message is that there is an average length of these recombined places on the string of beads. Over time, many, many generations of recombination had produced a kind of “signature” for the bead structure of a population—a pattern that served to distinguish one population’s strings of beads from another’s, since people living in the same geographic region tend to share more ancestors than people from different parts of the world.
Pritchard and his colleagues had developed a new statistical method to find regions of the chromosomes that seemed to have too little shuffling. In other words, they found parts of chromosomal bead strings that had long sections that seemed too similar to each other—as if everybody was wearing a uniquely patterned necklace, except that one long section of each person’s necklace was pretty much identical to everyone else’s. For segments like this it was possible to infer that something had happened to produce a long section of beads that seemed to be inherited like a block among many people, as though it had spread through their necklaces like a fashion accessory. One person liked the particular combination of beads they saw in part of someone else’s necklace, and copied it to include in theirs. Fashion tastes served to spread the bead pattern far and wide, and pretty soon lots of people were wearing it.
Of course, chromosome patterns can’t be recognized by looking at someone, and you can’t just take a section of someone’s chromosome and splice it into your own, so the explanation for this genetic “faddishness” had to lie somewhere else. Because the chromosomes carry genes, not beads, the inference was that the particular pattern in one person’s chromosome provided some sort of an evolutionary advantage, allowing it to spread through the population. When this happens in nature, the process is known not as fashion but as selection: Darwin’s force, the one that he got so excited about back in the nineteenth century, that served to create highly adapted organisms over many generations. The opposable thumb, color vision, our amazing brains—all had their origins in small changes that had been selected for in our DNA millions of years ago.
The section of the chromosomal beads that many people shared must have had a particular change in its gen
etic code that conferred an evolutionary advantage, and because of this the people carrying it were more likely to survive and pass on their DNA—and that popular section of beads. Studying such patterns gives us a way to peer back in time and ask how our genomes have been molded by past episodes of selection. By sifting through clues found in the DNA of people alive today, it is possible to see evidence of events that happened many generations ago, like a forensic detective trying to piece together details about a crime from evidence at the scene. And this is exactly what Pritchard and his colleagues did in their PLoS paper.
I asked Pritchard to explain the method, and he went to the whiteboard in his office and started to draw parallel horizontal lines in different colors, explaining how he and his colleagues had scanned the HapMap data for blocks of similar structure. He explained what they called the integrated haplotype score, or iHS, as a way of correcting for regional variation in the pattern of recombination in the human genome. The iHS accounts for the fact that some parts of the genome will have long regions that appear to be quite similar across all populations, simply due to the physical structure of the DNA in that region (and not necessarily to selection for a popular section of beads), while others will experience more recombination and will vary quite a bit among individuals. In their analysis, Pritchard and his colleagues were looking for chromosomal regions that should have been shuffled and diverse—and thus old—but weren’t. These sections were younger than they should be, suggesting that something had happened relatively recently to cause a change in the pattern of that particular region of the genome—in other words, a block of similar beads had spread throughout the population due to selection.
“We were looking at events that had not gone to fixation,” he explained. This meant cases where the short stretches of fashionable beads were not yet shared by everyone in the population. They did this so that they could estimate the expected level of genetic variation for each region independently, as a kind of internal control. This gave much greater power to their statistical analysis, and made it more likely that the chromosomal regions that returned a positive score really had been subject to selection.
Pritchard discussed the careful work they had done to try to account for any biases inherent in the analysis. He discussed the limitations of dealing with the three populations included in the HapMap data set, and his plans to look at data from additional populations. The HapMap was the product of an international consortium of biomedical researchers interested in finding genetic variants that could be associated with common diseases, like diabetes or hypertension. It was not planned with the primary intention of telling us more about our evolutionary history, but Pritchard’s small team of researchers had discovered a way to use it to do this.
Their analysis had revealed hundreds of regions of the genome, scattered across all of our twenty-three pairs of chromosomes, that had been strongly selected. There seemed to be smoking guns everywhere, all containing stories about the evolutionary history of our species. But the most incredible single discovery to come out of the study, and the reason I was talking to Pritchard, was how recently these selective events had occurred. All of them had happened in the past 10,000 years.
We usually think of natural selection as a long, slow process. Darwin and other evolutionary biologists typically thought in terms of selection happening over millions of years, as the slight advantage provided by a new trait slowly won out over its less successful rivals. “Survival of the fittest”—the term was actually coined by the nineteenth-century social scientist Herbert Spencer—was about the gradual accumulation of genetic variants that eventually led a species to become better adapted to its environment. In studies that had been done on model experimental organisms such as bacteria or fruit flies, the calculated selective advantage was typically a fraction of a percent, which meant that organisms with the trait took thousands of generations to show evidence of selection—in other words, to show that widespread bead pattern. Pritchard’s finding of hundreds of episodes of selection in the past 10,000 years—only around 350 human generations—implied that our species had been subjected to a very strong selection pressure during this time.
What could have caused this huge change in our genome? Pritchard was quick to point out that his method may have been biased toward these results, although he admitted that his continuing analyses reinforced his finding that there had been stronger selection during this period than there had before. Other researchers have since confirmed Pritchard’s results, suggesting that this time period had indeed produced a significant number of changes in our genome. To understand the timing of these events, we need to look elsewhere, outside our DNA, in the stones and bones of paleoanthropology.
THE IMPORTANCE OF INFLECTION
Our species is a relative newcomer on the biological scene. While horseshoe crabs and sharks are recognizable in the fossil record from over 100 million years ago, the hominid lineage—composed of apes that walk upright like us—doesn’t appear until around 5 million years ago. Our genus, Homo, appears even later, around 2.3 million years ago, with the first large-brained hominids to make stone tools, Homo habilis, and their descendants Homo erectus. Hominids with an even larger brain, looking more like us, appear around 500,000 years ago, but they still don’t belong to our species. In other words, we are rank newcomers on the evolutionary scene.
According to a recent reanalysis of fossils discovered in 1967 by Richard Leakey at the Kibish Formation on the Omo River in southern Ethiopia, our species—Homo sapiens—first appeared as a biologically recognizable entity around 195,000 years ago. These fossils were originally dated to 130,000 years ago, but newer methods have shown them to be 65,000 years older than previously thought. In the highly contentious field of paleoanthropology, where a single discovery can rewrite history, the dates seem pretty certain. Older human remains may be discovered in the future, but as far as we know these were the first humans ever to tread the earth.
The next oldest human fossil finds date to nearly 40,000 years later, at Herto in Ethiopia and Jebel Irhoud in Morocco, but it isn’t until around 120,000 years ago that significant numbers of Homo sapiens start to show up in the fossil record. The best-known finds are at Qafzeh and Skhul Caves in present-day Israel and the Klasies River in South Africa. The dearth of Homo sapiens fossils over a 75,000-year period could be due to low population densities, or perhaps it is simply a result of a poorly studied fossil record, but the implication is that humans were a relatively rare species limited to Africa and the Middle East.
Based on the number of known archaeological sites, along with clues from the human genome, we can estimate the ancient demography of our species since that time. If we plot this on a graph, an interesting pattern emerges (Figure 2). Our species had an unknown but probably fairly stable population size between the time when we originated around 200,000 years ago until around 80,000 years ago. Judging from the sparse fossil record of human remains, the population size was small and the species scattered throughout East and North Africa. Around 120,000 years ago, when we show up in the Middle East and South Africa, there was still no evidence of a significant change in the number of humans. Rather, these small, dispersed groups appear to have wandered into new territory. The Middle East at this time was basically a geographic extension of North Africa, with a similar climate, flora, and fauna, so these early humans did not venture far beyond their African home. There is no evidence of a human presence elsewhere in Asia or Europe at this time.
Between 80,000 and 50,000 years ago, however, something significant happened to the human population. The fossil and archaeological record runs dry, and there is little evidence for humans anywhere, including Africa. The settlements in the Middle East and South Africa were abandoned, as though we were retreating in the face of a catastrophic challenge. Taking the evidence at face value, the inference is that we went through a population crash. And according to recent genetic analyses, this is precisely what happened. By assessing the level of genetic
diversity in the present-day human population, which turns out to be remarkably low in comparison with that of our nearest cousins, the great apes, it is possible to calculate that the human population may have averaged no more than a mere two thousand people around 70,000 years ago. Our species was literally on the brink of extinction, at the nadir of our 200,000-year population curve. Then, around 60,000 years ago, something else happened—a change in the direction of the curve, known in mathematics as an inflection point. The human population actually started growing, and this seems to correlate with the first appearance of humans outside of Africa and the Middle East. Within 45,000 years we had spread to every continent (apart from Antarctica), increasing from the couple of thousand who survived the population crash to a few million hunter-gatherers, spread throughout the entire world. What led to this expansion will be discussed in Chapter 4.
FIGURE 2: THE VARIATION IN HUMAN POPULATION SIZE OVER THE PAST 100,000 YEARS. NOTE THE USE OF A LOGARITHMIC SCALE ON THE VERTICAL AXIS (103 = 1,000, 106 = 1 MILLION, ETC.).
At 10,000 years ago, something really momentous happens: we see another change in the curve, with a massive acceleration in the rate of population growth. Taking us from a few million to over six billion today, this was the true explosion of our species—the Big Bang that led to humans dominating the world stage. What set in motion this sudden growth spurt around 10,000 years ago? If you are an archaeologist, you will know the answer immediately. It was at this time that we settled down and made a conscious decision to change our relationship with nature. We developed agriculture. While hunter-gatherers had relied on finding their food sources, agriculturalists created theirs. This seemingly simple transition in the way we obtained nourishment set in motion a sea change in human history. Instead of being held captive by where we could find enough plants and animals to survive, we gained control of our food supply. The result was that food was no longer the limiting factor in determining how many people could live in one place. We’ll explore how this came about in the next chapter, but for now it is enough to say that by controlling the food supply, we gained the ability to choose how many people could live in a particular location. If the population increased in size, it was relatively easy to grow more food. This radical change in lifestyle set in motion the Big Bang in our population growth curve.