< Back to Articles
View Web Version
On a summer day in Siberia back in 2010, archaeologists at the Denisova Cave in the Altai Mountains painstakingly scraped layers of ancient dirt from the east gallery floor.
The scientists collected the precious soil and carefully transported it by cable car down to the nearby Anuy River. Using icy water to sieve the dirt from rocks or bone, they examined the leftover bits for remnants of cave hyena or cave lion or other ancient animals.
But what they found was far more interesting than old animal skeletons. They recovered a bit of bone that looked more hominin than animal, a single 26mm-long phalanx - a tiny toe.
Since the archaeologists had previously found a tiny finger bone in the cave from an ancient human relative “Denisovan” , they figured this new toe bone would be from the same creature. But early tests showed that, “to our surprise, the mitochondrial DNA was pure Neanderthal,” said Svante Paabo, lead author of the paper and director of the Department of Genetics at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany.
As luck would have it, Paabo’s group had just developed a new, sensitive protocol that could generate robust DNA libraries from ancient, fragmented DNA. Using the new approach, they sequenced and pieced together the most complete Neanderthal genome sequence to date: 52-fold sequence coverage, greater than many modern human DNA sequences.
But apart from the sequence itself, this wealth of genetic data has already given researchers insight into Neanderthal social patterns and breeding between different hominin groups of the Late Pleistocene. Could it also help describe our uniquely human spark, illuminating just what it means to be a person?
A History of Hominins
Archaeologists have excavated the Denisova Cave in Siberia every summer since the 1970s, when Soviet scientists first found the treasure trove within. Animal bones with chop marks fill the many layers, and scientists guess that ancient hominins dragged their food to the cave’s shelter over thousands of years. Because the cave’s internal temperature hovers around 32° F (0°C), the bones and DNA within are better preserved than if they had been exposed to the elements. But, since the cave was utilitarian and not a burial site, bones of the hominins that ate the animals were not abundant.
Very occasionally, however, amid the animal bones, archaeologists have discovered remains of an ancient hominin. “There are not many hominin remains at all,” said Paabo, and so “it’s really very, very skillful of the excavator to recognize these very small bones as potentially being human.” In 2008, a little hominin finger bone was found at the site. When it was first analyzed in 2010, Paabo’s group sequenced it to 1.9-fold coverage  and found that it was a new type of hominin, one more closely related to Neanderthals than humans. Based on the genome sequence, they called this new group “Denisovan.” Also in 2010, using three Neanderthal specimens from Croatia, Paabo’s group published a draft sequence of the Neanderthal genome to 1.3-fold coverage .
The New Ancient Protocol
In 2012, however, Paabo’s group finished developing their new protocol  geared specifically for ancient, fragmented DNA. With this protocol, the team was able to reconstruct an amazingly complete 30× coverage of the Denisovan genome from that finger bone .
So when the archaeologists discovered the new Neanderthal toe bone in 2010, Paabo’s team was ready. They scraped 38 mg of bone powder using a dentistry drill to create 100 µL of DNA extract and then used the new protocol to generate 4 DNA libraries. In the past, only methods that used double-stranded DNA were used to create DNA libraries for ancient genomes. But Paabo’s new method used a single-stranded library preparation, which works better in highly fragmented ancient DNA. Heat denatures the DNA, which separates the two strands, and then an adapter oligonucleotide is ligated to the 3´ end. The strands are immobilized on streptavidin-coated beads, which protect DNA from being lost during DNA purification. Next, the single strands are used as a template to produce their opposing strands. “Each double stranded molecule has two chances to make it into the library, one for each strand,” said Paabo. PCR amplifies the DNA and then it’s sequenced quite efficiently.
Once the DNA is sequenced, working with it is just about the same as working with any modern DNA. “The quality of the DNA sequence from the Altai Neanderthal was as good as the quality of the sequence from living humans,” said Montgomery Slatkin, co-author of the paper from the University of California, Berkeley, who worked on the population genetics. “That’s the real breakthrough in this paper and the previous Denisovan paper. In the earlier papers, we had to make special allowances for particular kinds of sequencing errors that appeared in ancient DNA that don’t appear in modern DNA. But now we don’t worry about that so much.” David Reich’s group at Harvard Medical School also worked on the population genetics. Hominin Promiscuity
Analyses by Slatkin’s team revealed that the Altai Neanderthal’s parents were probably half-siblings (though they also could have been uncle/niece, aunt/nephew, grandfather/granddaughter or grandmother/grandson). “It suggests a first little insight into social patterns: rearing children between close relatives was not that uncommon in Neanderthal groups,” said Paabo.
Then, to investigate interbreeding between Neanderthals, Denisovans, and early modern humans, the teams compared the Neanderthal genome to 25 present-day human genomes, as well as the Denisovan genome and another low-coverage genome of a Neanderthal infant that Paabo’s group sequenced for this paper. They found more evidence that Neanderthals and Denisovans contributed genes to some modern humans and that Neanderthals also mixed with Denisovans. Interestingly, they also found gene flow into Denisovans from a mystery hominin, possibly Homo erectus, which existed at the same time.
Are We So Different?
Such a complete Neanderthal genome provides fascinating insight into genetic differences between modern humans and an ancient, extinct relative. The researchers found more than 31,000 single nucleotide substitutions, and more than 4100 short insertions and deletions that differ between all present-day humans and the Neanderthal, Denisovan, and great ape genomes. This translates to only 96 amino acid changes in 87 proteins in modern humans compared to Neanderthals. “It’s a manageable list. You can actually look through it all,” said Paabo. “It is a sort of genetic recipe for what makes a modern human.”
“These proteins are unusually interesting. It gives you a list of candidates to look very hard at,” said Slatkin. “It wouldn’t surprise me to see them involved in some human genetic diseases.”
Three of these 87 proteins are expressed in the proliferative layers during mid-fetal brain development and are associated with the kinetochores of the mitotic spindles that pull the chromosomes apart. Interestingly, another paper published on Christmas Day, 2013, found that a type 2 diabetes risk allele present in Native Americans originally came from Neanderthals. 
Expressing Neanderthal Proteins
To really understand how we differ from our closest relatives, scientists will need to create both modern human and Neanderthal forms of these 87 proteins and look for any functional differences in animal models.
“Neanderthals and humans aren’t very different morphologically, but they appear to have been very different behaviorally because humans evolved culture and Neanderthals didn’t very much,” said Slatkin. “It would be nice to find genes that correlate with that evolution. Still, we can’t just point the finger and say, ah! Here’s the gene that caused humans to be humans.”
 Meyer M et al, “A high-coverage genome sequence form an archaic Denisovan individual,” Science, 338, 222-226 (2012).
 Reich D et al, “Genetic history of an archaic hominin group from Denisova Cave in Siberia,” Nature, 468, 1053-60 (2010).
 Green R et al, “A draft sequence of the Neandertal Genome,” Science, 328, 710-722 (2010).
 Gansauge M and Meyer M, “Single-stranded DNA library preparation for the sequencing of ancient of damaged DNA,” Nature Protocols, 8, 737-748 (2013).
 SIGMA Type 2 Diabetes Consortium, “Sequence variants in SLC16A11 are a common risk factor for type 2 diabetes in Mexico,” Nature, published online 25 December 2013.
< Back to Articles