By Monica Heger
Using a new method to remove deaminated cytosines and repair ancient DNA, researchers from the Max Planck Institute were able to sequence the nuclear genome of an ancient hominin found two years ago in the Denisova Cave in Siberia.
Svante Pääbo, who led the team, announced the preliminary results of the sequencing project at Cold Spring Harbor Laboratory's Biology of Genomes meeting earlier this month. The ancient human appears to represent a new hominin species, Pääbo said.
The researchers had previously reported sequencing the mitochondrial DNA of the hominin (GWDN 3/24/2010), but had been unable to sequence the nuclear genome because of degradation and extensive chemical modification.
However, earlier this year, they reported a DNA repair method, published in Nucleic Acids Research, which allowed them to sequence the nuclear genome of the individual. The repair process also has the added benefit of revealing the methylome, so researchers will now be able to evaluate the epigenome as well as the genome simultaneously.
At the meeting, Paabo reported that he and his team sequenced the so-called X-woman to 2.1-fold coverage on the Illumina Genome Analyzer, generating 6.6 gigabases of sequence data.
He estimated that the Denisova species diverged from modern humans about 1 million years ago, and said that the X-woman was about 13 percent divergent from modern humans, or slightly more divergent than Neandertals are from humans, which Pääbo's team estimated was 12.7 percent (IS 5/11/2010). Unlike the Neandertals, though, there appeared to be no gene flow from that ancient species into modern humans.
The team is now in the process of analyzing the Denisova data and Pääbo said that they hope to sequence additional ancient humans.
To obtain the nuclear genome, the team had to figure out a way to repair the DNA, which degraded over time and harbored numerous modifications. "The major form of chemical damage is the deamination of cytosine to uracil, which is read" as a thymine, said Pääbo. That damage is seen particularly at the 5' ends of the DNA, he added.
In the NAR paper, the authors reported that even though the total number of chemical modifications is relatively small, they still pose a problem when sequencing ancient DNA because the coverage is frequently very low.
The researchers devised a way to remove the uracil residues and repair the resulting abasic sites prior to sequencing. They used the enzyme uracil-DNA-glycosylase to remove the uracil base, a separate enzyme to cleave the DNA at the resulting abasic site, and then DNA polymerase and adapters to repair the ends. While the removal of the uracil base results in a lower library yield, the following treatments successfully repair the fragments.
When they tested their method on Neandertal DNA, they found that the non-treated DNA had error frequencies of around 40 to 50 percent at the ends of fragments, but by using their repair process, they were able to greatly reduce those errors. They calculated that they reduced the error rate 22-fold in mtDNA using their method combined with deep sequencing.
"The treatment alone results in a 5.5-fold reduction in error rates, while deep sequencing results in an additional 4.4-fold reduction," the authors wrote. They observed similar results in nuclear DNA for the Neandertal, although "the true error rate at these low levels cannot be accurately calculated due to genuine sequence differences between this Neandertal and the human reference."
Pääbo said at the conference that they used this method on the Densiova individual, and "after treatment, you see almost no misincorporations." When they looked to see where the few misincorporations were, they were all at CpG sites.
At CpG islands, the methylated cytosines are not affected by the treatment, so when sequenced, they show up as a C-to-T substitution. "So in fact, we have the methylome preserved of this individual," he said.
They added that detecting the methylation status of a specific CpG site using this approach would still require very deep sequencing, however, other technologies such as bisulfite treatment or single molecule sequencing may soon enable even better analysis of methylation in ancient DNA.
"It might therefore become possible to investigate the activity of genes and phenomena such as X chromosomal inactivation and genetic imprinting in extinct species, provided that the signals are manifested in cells present in bones," the authors concluded.
Pääbo said that they are now looking for evidence of another hominin species that may have shared genes with the Denisova hominins. "We will have, in the future, not just the Neandertal genome, but the genomes of other archaic humans, which will illuminate the history of human evolution," he said.