Another case of early human interbreeding confirmed in Siberia

It's been a busy year in research on recent human ancestry. Back in the spring, scientists completed a draft of the Neanderthal genome, which provided clear evidence that these now-extinct humans left some of their genes behind by interbreeding with some human ancestors. A bit earlier in the year, DNA sequencing revealed an even larger surprise: there seems to have been another population of premodern humans present in Asia that were genetically distinct from modern humans and Neanderthals. Now, the team behind both of these discoveries is back with a draft genome of this population that suggests it was genetically distinct from both humans and Neanderthals, and a single tooth that suggests it was physically distinct. And that it also interbred with the ancestors of a modern human population.

The new population was identified based on sequence from a single bone found in a cave called Denisova. Sequencing the genome of its mitochondria indicated it had branched off from the ancestor of both humans and Neanderthals roughly a million years ago, making it a relatively archaic lineage. But mitochondrial DNA is prone to rapid sequence changes as well as founder and bottleneck effects, which could exaggerate the divergence. The bone it came from didn't differ significantly from either of these human populations, meaning there was no physical indication that the Denisova remains represented a new population.

In the new paper, which will be published in Nature, the researchers have gone back and corrected both of these issues.

They've used a new technique to limit the errors in the sequencing of ancient DNA. These arise primarily from sites where a base has been either lost or chemically modified so that it base pairs improperly. The authors treated their DNA with an enzyme that recognizes the modified base and chops it off, turning it into the equivalent of a missing base. They then used a second enzyme that cuts the DNA backbone wherever there are bases missing. Combined, these two treatments significantly cut down on the errors.

Despite the single source, the authors were able to obtain nearly two-fold coverage of the Denisova genome, meaning that, on average, each base was sequenced twice. Over the billions of base pairs in a human genome, however, that average means we'll have both areas that are well sequenced and gaps that don't appear at all.

Placing Denisova in the human family tree

Nevertheless, this was enough to align Denisova with the human, Neanderthal, and chimp genomes, and figure out who was ancestral to whom. In contrast to the results obtained from mitochondrial DNA, the genomic data clearly grouped Denisova as closer to the Neanderthals, but not especially close. "DNA sequences of Neanderthals and the Denisova individual diverged on average 640,000 years ago, and from present-day Africans 804,000 years ago," the authors conclude. So, Denisova remains a distinct population, but one most closely related to Neanderthals.

How did the mitochondrial DNA look so different? One possibility is simply that it's chance: a mitochondrial lineage might have survived in the common ancestor of Neanderthals and Denisova, and then have been lost after their split. The other possibility is that Denisova interbred with a different, earlier Homo lineage that we've not identified yet. Right now, we simply can't distinguish between these possibilities.

But there is evidence that the Denisova population interbred with another Homo lineage: our own. The authors performed a similar analysis to the one done with Neanderthals. Since Denisova individuals should only have interbred with our lineage after it left Africa, the authors looked for affinities to individual non-African populations. For the most part, these came up blank, with just a small percentage of commonality arising due to the interbreeding done by Neanderthals. 

But there was a big exception: Melanesians, as represented by individuals from Papua New Guinea. Here, the shared genetic ancestry numbers rose to close to four percent, similar to that shared by Neanderthals and Europeans. So, it seems that, on their way to the Pacific, Melanesians and the Denisova population interbred briefly. Since that isn't true of the groups that are geographically closest to Denisova, like the Han Chinese, the authors suggest that Denisova individuals probably ranged much more widely than their single site suggests.

Putting some tooth into Denisova

With the paper, it's no longer a single sample from a single site. With the Denisova sequence in hand, they started genotyping other bones, and came up with a tooth from a separate individual in the same cave. And, unlike the hand bone they had previously found, the tooth (a molar) does have some distinctive features. It's unusually large compared to many samples from modern humans and Neanderthals, and shares a size and some features with molars from Homo erectus and Homo Habilis. But it's clearly distinct from their teeth as well.

It's not a lot to go on, and it's doubtful that the tooth alone would have caused much of a fuss without the accompanying DNA evidence. The authors simply conclude that the tooth supports the DNA in indicating that the Denisova remains represent a distinct population. But they don't go beyond that, choosing to "refrain from any formal Linnaean taxonomic designations that would indicate species or subspecies status for either Neanderthals or Denisovans."

Where does that leave us? The big picture of recent decades—that modern humans evolved in Africa and spread from there, displacing all other populations—is still largely accurate. But the details are looking much more complex than they were just last year. Those other populations are suddenly seeming a lot more diverse, and they didn't go away without contributing a bit to the genetic diversity of the modern human population.

Nature, 2010. DOI: 10.1038/nature09710  (About DOIs).