In the Altay Mountains of southern Siberia, some 200 miles from where Russia touches Mongolia, China, and Kazakhstan, nestled under a rock face about 30 yards above a little river called the Anuy, there is a cave called Denisova. It has long attracted visitors. The name comes from that of a hermit, Denis, who is said to have lived there in the 18th century. Long before that, Neolithic and later Turkic pastoralists took shelter in the cave, gathering their herds around them to ride out the Siberian winters. Thanks to them, the archaeologists who work in Denisova today, surrounded by walls spattered with recent graffiti, had to dig through deep layers of goat dung to get to the deposits that interested them. But the cave’s main chamber has a high, arched ceiling with a hole near the top that directs shimmering shafts of sunlight into the interior, so that the space feels holy, like a church.
In the back of the cave is a small side chamber, and it was there that a young Russian archaeologist named Alexander Tsybankov was digging one day in July 2008, in deposits believed to be 30,000 to 50,000 years old, when he came upon a tiny piece of bone. It was hardly promising: a rough nubbin about the size and shape of a pebble you might shake out of your shoe. Later, after news of the place had spread, a paleoanthropologist I met at Denisova described the bone to me as the “most unspectacular fossil I’ve ever seen. It’s practically depressing.” Still, it was a bone. Tsybankov bagged it and put it in his pocket to show a paleontologist back at camp.
The bone preserved just enough anatomy for the paleontologist to identify it as a chip from a primate fingertip—specifically the part that faces the last joint in the pinkie. Since there is no evidence for primates other than humans in Siberia 30,000 to 50,000 years ago—no apes or monkeys—the fossil was presumably from some kind of human. Judging by the incompletely fused joint surface, the human in question had died young, perhaps as young as eight years old.
Anatoly Derevianko, leader of the Altay excavations and director of the Institute of Archaeology and Ethnography in Novosibirsk, thought the bone might belong to a member of our own species, Homo sapiens. Sophisticated artifacts that could only be the work of modern humans, including a beautiful bracelet of polished green stone, had previously been found in the same deposits. But DNA from a fossil found earlier in a nearby cave had proved to be Neanderthal, so it was possible this bone was Neanderthal as well.
Derevianko decided to cut the bone in two. He sent one half to a genetics laboratory in California; so far he has not heard from that half again. He slipped the other half into an envelope and had it hand-delivered to Svante Pääbo, an evolutionary geneticist at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. It was there that the case of the Denisovan pinkie bone took a startling turn.
Pääbo, a transplanted Swede, is arguably the world’s leading expert in ancient DNA, especially human DNA. His milestones are many. In 1984 he became the first person to isolate DNA from an Egyptian mummy. In 1997 he accomplished the same feat for the first time with a Neanderthal, a kind of human that vanished more than 25,000 years before the Egyptian pharaohs. That secured his scientific reputation.
When Pääbo received the package from Derevianko, his team was hard at work producing the first sequence of the entire Neanderthal genome—another feat that had once seemed impossible and that was occupying most of his attention. His lab also had a backlog of other fossils to test from all parts of the globe. So it wasn’t until late 2009 that the little Russian finger bone drew the attention of Johannes Krause, at the time a senior member of Pääbo’s team. (He’s now at the University of Tübingen.) Like everyone else, Krause assumed the bone was from an early modern human. He had developed a method for distinguishing the DNA of such a fossil from that of the archaeologists, museum workers, and anyone else who might have handled and therefore contaminated it.
Krause and his student Qiaomei Fu extracted the finger bone’s mitochondrial DNA (mtDNA), a small bit of the genome that living cells have hundreds of copies of and that is therefore easier to find in ancient bone. They compared the DNA sequence with those of living humans and Neanderthals. Then they repeated the analysis, because they couldn’t believe the results they’d gotten the first time around.
On a Friday afternoon, with Pääbo away at a meeting at Cold Spring Harbor Laboratory on Long Island, Krause called a meeting of the lab staff and challenged anyone to come up with a different explanation for what he was seeing. No one could. Then he dialed Pääbo’s cell. “Johannes asked me if I was sitting down,” Pääbo remembers. “I said I wasn’t, and he replied that I had better find a chair.”
Krause himself recalls that Friday as “scientifically the most exciting day of my life.” The tiny chip of a finger bone, it seemed, was not from a modern human at all. But it wasn’t from a Neanderthal either. It belonged to a new kind of human being, never before seen.
In July 2011, three years after Tsybankov unearthed the bone chip, Anatoly Derevianko organized a scientific symposium at the archaeological camp a few hundred yards from Denisova cave. At an opening night dinner punctuated with frequent toasts of vodka, Derevianko welcomed the 50 researchers, including Pääbo, who had come to see the cave and share their views on how the mysterious new human fit into the fossil and archaeological record for human evolution in Asia.
The year before, two other fossils had been found to contain DNA similar to that of the finger bone, both of them molars. The first tooth had turned up among the specimens from Denisova housed at Derevianko’s institute in Novosibirsk. It was bigger than either a modern human or a Neanderthal tooth, in size and shape resembling the teeth of much more primitive members of the genus Homo who lived in Africa millions of years ago. The second molar had been found in 2010 in the same cave chamber that had yielded the finger bone—indeed, near the bottom of the same 30,000-to-50,000-year-old deposits, called Layer 11.
Remarkably, that tooth was even bigger than the first, with a chewing surface twice that of a typical human molar. It was so large that Max Planck paleoanthropologist Bence Viola mistook it for a cave bear tooth. Only when its DNA was tested was it confirmed to be human—specifically, Denisovan, as the scientists had taken to calling the new ancestors. “It shows you how weird these guys are,” Viola told me at the symposium. “At least their teeth are just very strange.”
Pääbo’s team could extract only a tiny amount of DNA from the teeth—just enough to prove they came from the same population as the finger, though not from the same individual. But the finger bone had been spectacularly generous.
DNA degrades over time, so usually very little remains in a bone tens of thousands of years old. Moreover, the DNA from the bone itself—called endogenous DNA—is typically just a tiny fraction of the total DNA in a specimen, most of which comes from soil bacteria and other contaminants. None of the Neanderthal fossils Pääbo and his colleagues had ever tested contained even 5 percent endogenous DNA, and most had less than one percent. To their amazement, the DNA in the finger bone was some 70 percent endogenous. Apparently, the cold cave had preserved it well.
Given so much DNA, the scientists easily ascertained that there was no sign of a male Y chromosome in the specimen. The fingertip had belonged to a little girl who had died in or near Denisova cave tens of thousands of years before. The scientists had no idea, at first, what she looked like—just that she was radically different from anything else they had ever seen.
For a while they thought they might have her toe too. In the summer of 2010 a human toe bone had emerged, along with the enormous tooth, from Layer 11. In Leipzig a graduate student named Susanna Sawyer analyzed its DNA. At the symposium in 2011 she presented her results for the first time. To everyone’s shock, the toe bone had turned out to be Neanderthal, deepening the mystery of the place.
The green stone bracelet found earlier in Layer 11 had almost surely been made by modern humans. The toe bone was Neanderthal. And the finger bone was something else entirely. One cave, three kinds of human being. “Denisova is magical,” said Pääbo. “It’s the one spot on Earth that we know of where Neanderthals, Denisovans, and modern humans all lived.” All week, during breaks in the conference, he kept returning alone to the cave. It was as if he thought he might find clues by standing where the little girl may have stood and touching the cool stone walls she too may have touched.
Pääbo grew up in Stockholm with his single mother, a chemist, and on certain days with his father, a biochemist named Sune Bergström, who had another, legitimate family and would later win a Nobel Prize. Pääbo’s own first passion was Egyptology, but he switched to molecular biology, then fused the two interests in 1984 with his work on mummy DNA. Once anchored in the study of the past, he never let go. He is 58 now, tall and lanky, with large ears, a long, narrow head, and pronounced eyebrows that arch up and down animatedly when he’s excited—about Denisova, for instance.
How had all three kinds of human ended up there? How were Neanderthals and Denisovans related to each other and to the sole kind of human that inhabits the planet today? Did their ancestors have sex with ours? Pääbo had a history with that kind of question.
The Neanderthal DNA he had made headlines with in 1997 was utterly different from that of any person now alive on Earth. It seemed to suggest that Neanderthals had been a separate species from us that had gone extinct—suspiciously soon after our ancestors first migrated out of Africa into the Neanderthals’ range in western Asia and Europe. But that DNA, like Krause’s first extract from the Denisovan finger, was mtDNA: It came from the mitochondria, the energy-producing organelles inside the cell, and not from the cell nucleus, where the vast bulk of our genome resides. Mitochondrial DNA includes only 37 genes, and it’s inherited only from the mother. It’s a limited record of a population’s history, like a single page torn from a book.
By the time of the Denisova symposium, Pääbo and his colleagues had published first drafts of the entire Neanderthal and Denisovan genomes. Reading so many more pages allowed Pääbo and his colleagues, including David Reich at Harvard University and Montgomery Slatkin at the University of California, Berkeley, to discover that human genomes today actually contain a small but significant amount of Neanderthal code—on average about 2.5 percent. The Neanderthals still may have been swept into extinction by the strange, high-browed new people who followed them out of Africa, but not before some commingling that left a little Neanderthal in most of us, 50,000 years later. Only one group of modern humans escaped that influence: Africans, because the commingling happened outside that continent.
Although the Denisovans’ genome showed that they were more closely related to the Neanderthals, they too had left their mark on us. But the geographic pattern of that legacy was odd. When the researchers compared the Denisovan genome with those of various modern human populations, they found no trace of it in Russia or nearby China, or anywhere else, for that matter—except in the genomes of New Guineans, other people from islands in Melanesia, and Australian Aborigines. On average their genomes are about 5 percent Denisovan. Negritos in the Philippines have as much as 2.5 percent.
Putting all the data together, Pääbo and his colleagues came up with a scenario to explain what might have occurred. Sometime before 500,000 years ago, probably in Africa, the ancestors of modern humans split off from the lineage that would give rise to Neanderthals and Denisovans. (The most likely progenitor of all three types was a species called Homo heidelbergensis.) While our ancestors stayed in Africa, the common ancestor of Neanderthals and Denisovans migrated out. Those two lineages later diverged, with the Neanderthals initially moving west into Europe and the Denisovans spreading east, perhaps eventually populating large parts of the Asian continent.
Later still, when modern humans ventured out of Africa themselves, they encountered Neanderthals in the Middle East and Central Asia, and to a limited extent interbred with them. According to evidence presented by David Reich at the Denisova symposium, this mixing most likely occurred between 67,000 and 46,000 years ago. One population of modern humans then continued east into Southeast Asia, where, sometime around 40,000 years ago, they encountered Denisovans. The moderns interbred with them as well and then moved into Australasia, carrying Denisovan DNA.
This scenario might explain why the only evidence so far that the Denisovans even existed is three fossils from a cave in Siberia and a 5 percent stake in the genomes of people living today thousands of miles to the southeast. But it left a lot of questions unanswered. If the Denisovans were so widespread, why was there no trace of them in the genomes of Han Chinese or of any other Asian people between Siberia and Melanesia? Why had they left no mark in the archaeological record—no distinctive tools, say? Who were they really? What did they look like? “Clearly we need much more work,” Pääbo acknowledged at the Denisova symposium.
The best of all possible developments would be to find Denisovan DNA in a skull or other fossil with distinctive morphological features, one that could serve as a Rosetta stone for reexamining the whole fossil record of Asia. There are some intriguing candidates, most from China, and three skulls in particular, dated between 250,000 and 100,000 years ago. Pääbo is working closely with scientists at the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing and has set up a DNA testing lab there. Unfortunately DNA does not preserve well in warmer climates. To date, no other fossil has been identified as Denisovan by the only way Denisovans can be known: their DNA.
In 2012 Pääbo’s group published a new version of the finger bone’s genome—astonishingly, one that in accuracy and completeness rivals any living human’s genome that has been sequenced. The breakthrough came from a German postdoc in Pääbo’s lab named Matthias Meyer. DNA consists of two interlocking strands—the familiar double helix. Previous methods for retrieving DNA from fossil bone could read out sequences only when both strands were preserved. Meyer had developed a technique for recovering short, single-stranded fragments of DNA as well, greatly increasing the amount of raw material to work with. The method produced a version of the Denisovan girl’s genome so precise that the team could discriminate between genetic information inherited from her mother and that from her father. In effect, they now had two highly accurate Denisovan genomes, one from each parent. These in turn opened a window on the entire history of their population.
One immediate revelation was how little variation there was between the parents’ genomes—about a third as much as there is between any two living humans. The differences were sprinkled across the genomes, which ruled out inbreeding: If the girl’s parents had simply been closely related, they would have had huge chunks of exactly matched DNA. The pattern indicated instead that the Denisovan population represented by the fossil had never been large enough to have developed much genetic diversity. Worse, it seemed to have suffered a drastic decline sometime before 125,000 years ago—the little girl in the cave may have been among the last of her kind.
Meanwhile the ancestral population of modern humans was expanding. Myriad fossils, libraries full of books, and the DNA of seven billion people are available to document our subsequent population history. Pääbo’s team discovered a completely different one inside a single bone chip. The thought tickles him. “It’s incredibly cool that there is no one walking around today with a population history like that,” Pääbo told me, his eyebrows shooting up.
And yet the Denisovans also have something to say about our own kind. With virtually every letter of the Denisovan genetic code in hand, Pääbo and his colleagues were able to take aim at one of the profoundest mysteries: In our own genomes, what is it that makes us us? What defining changes in the genetic code took place after we separated from our most recent ancestor? Looking at the places where all living humans share a novel genetic signature but the Denisovan genome retains a primitive, more apelike pattern, the researchers came up with a surprisingly short list. Pääbo has called it the “genetic recipe for being a modern human.” The list includes just 25 changes that would alter the function of a particular protein.
Intriguingly, five of these proteins are known to affect brain function and development of the nervous system. Among them are two genes where mutations have been implicated in autism and another that’s involved in language and speech. Just what those genes actually do to make us think, act, or talk differently than Denisovans, or any other creature that has walked the Earth, remains to be seen. The lasting contribution of studying Denisovan DNA, Pääbo says, “will be in finding what is exclusively human.”
But what of the little girl herself? The tiny bit of bone that is all we ever had of her—or at least the half that went to Leipzig—is gone now. In pulling DNA from it, Johannes Krause and Qiaomei Fu eventually used it all up. The little girl has been reduced to a “library” of DNA fragments that can be exactly copied again and again forever. In the scientific paper discussing the history of her population, Pääbo and his colleagues did mention, almost in passing, a few facts about her that they had gleaned from that library: She probably had dark hair, dark eyes, and dark skin. It isn’t much, but at least it sketches in broad strokes what she looked like. Just so we know whom to thank.