Excavations at Denisovan Cave have yielded tiny bone fragments that have had too much of an impact on our understanding of human evolution.
For a short period of time in the history of our species, we shared our world with other intelligent humans, closely related to us but different. We don’t know much about how our ancestors interacted with these other now-extinct hominids, but we do know that at least some of those interactions were quite intimate, as many modern humans now carry traces of DNA from Neanderthals and another ancient hominin group called Denisovans.
Most modern people of European and Asian descent carry between one and three percent Neanderthal DNA, and most people of Asian and Oceanian descent carry up to five percent Denisovan DNA. Because Neanderthals and Denisovans originated outside of Africa, the ancestors of modern African humans would never have encountered them, although researchers have suggested that a hitherto unidentified hominin species in Africa mingled with our ancestors there so we could all bear tracks. from that distant relative too.
These were not isolated incidents. The genetic inheritance that many of us now carry with us is probably the sign of years of ongoing contact between two groups. Consistent with that, it now appears that people may have interacted with Denisovans not in one place and time, but in two.
Still in our genes
Global map of Denisovan gene frequency in modern human genomes.
Sharon Browning, a research professor of biostatistics at the University of Washington, Seattle, and her colleagues made the discovery while testing a new method to scan human DNA for exactly this kind of mixing with now-extinct hominids such as Neanderthals and Denisovans.
One way to do that is to simply compare the genomes side by side and look for points that match. We can make such a comparison with Neanderthals and Denisovans because scientists have sequenced the genomes of both species. But for researchers looking for traces of the unknown African hominins mentioned above (or any other species without a sequenced genome), that won’t work, of course.
Scientists can also use a statistical method to look for groups of genes that seem out of place. It works because genes naturally accumulate mutations over time; most are little things that don’t make any difference, although others can ultimately influence our properties through natural selection. Humans and Denisovans had long been separated from their last common ancestor when we met again, so our genomes were busy acquiring their own distinct sets of mutations. The statistical method scans a human genome to look for groups of genes that appear to have many mutations that are not found in the rest of the population. Those are marked as possibly coming from crossing with another hominin species.
Browning and her colleagues developed a new version that works on a larger scale than the method already used. They tested it on 5,639 human genome sequences from people from Europe, Asia, America and Oceania; in case you’re wondering where to get several thousand complete human genomes, Browning and her colleagues got these from the UK10K Project, 1000 Genomes Project and the Simons Genome Diversity Project.
When they found genetic sequences that potentially resembled Denisovan, they compared them to the only Denisovan genome sequence we have, which came from a 30,000 to 50,000-year-old fossil found in a cave in the Altai Mountains in Siberia.
Not once, but twice
People of South Asian and Oceanian descent had different sets of Denisovan genes from people of East Asian descent. The East Asian genomes in the study carried genes that were closely related to the Altai specimen, but they also carried some that were not. Meanwhile, the South Asian and Oceanian genomes only carried genes unrelated to the Altai Denisovan. The best explanation is that Denisovans and humans have reproduced twice in our history, not just once, and that the Denisovans had already split into distinct populations.
Here’s a possible scenario: Some time after the first Denisovans reached Eurasia, a group gradually moved north, eventually reaching Siberia. Another group moved south, through South Asia and to the islands of Oceania. Between 40,000 and 60,000 years ago, modern humans migrated to the same region, where they met the southern Denisovan branch and lived side by side long enough to interbreed. Modern people of South Asian and Papuan descent still carry the genetic traces of that interaction.
Modern humans of East Asian descent carry the same genetic traces, meaning their ancestors hadn’t split off from South Asian populations when we first encountered Denisovans. But they also have a set of Denisovan genes more closely related to the Altai fossil, which the South Asian and Oceanic genomes don’t have. At some point after the split, the ancestors of today’s East Asian populations may have met the northern branch of the Denisovans, which had been separated from its southern relatives long enough for its genes to look different. They, too, interacted and interbred for a while, and that too left its genetic mark for Browning and her colleagues to find.
We don’t know exactly when any of those events happened, and researchers can’t really be sure of the order in which they happened. Further research could someday shed some light on that, and that’s part of what Browning and her colleagues could achieve by applying their method to more human genome sequences from around the world. They also want to look for other cases of interbreeding, either with Neanderthals and Denisovans, or with potential African hominids.
“Previous studies have found suggestive evidence for admixture of unidentified hominins, but more work is needed to explore this more deeply,” Browning said. “Each Homo lineage overlapping in time with modern humans is a candidate.”
In other words, our family tree can be a lot more complicated than we think. If researchers find evidence of even more genetic admixture with other early hominins, “it would tell us about the complexities of human history and also tell us that maybe we weren’t all that different from the other hominins,” Browning said.
Cell2017. DOI: 10.116/j.cell.2018.02.031 (About DOIs).