Six months left stranded in one of the Earth's coldest inhabited regions with fur-trappers from a remote Siberian tribe, near-death experiences with rampaging bears, and frequent encounters with deep-frozen mammoth remains and ancient archaeological artefacts sound like scenes from an Indiana Jones movie. But they are all par for the course for evolutionary biologist Eske Willerslev, who moves to the University of Oxford from Copenhagen this autumn thanks to a Wellcome Bioarchaeology Fellowship. Not that he will actually be in Oxford much: his next challenge is searching for ancient human DNA in the vast frozen wastes of Siberia, Greenland and North America.

The old and the new

Palaeontology and archaeology are being radically transformed by the study of ancient DNA, a field that has taken off over the past few decades. Researchers have managed to extract preserved fragments of genes from human remains many thousands of years old, as well as from the bones of extinct animals as diverse as mammoths, Siberian cave lions and bison, Mauritian dodo and New Zealand moa birds. This DNA can be used to study evolution in action by comparing ancient genes to modern ones, working out which modern relatives extinct species are related to, discovering where ancient humans settlers originally came from, and many other uses beside.

However, the field has been plagued by problems of contamination (see box below) from the DNA that is ever present on human hands, exhaled cells, bacteria and other sources. For example, a Viking tooth recently tested at Oxford yielded genetic material from at least 20 different people. To compound the problem, even relatively recent bones contain vanishingly small quantities of DNA, which is difficult to extract. Dr Willerslev's ground-breaking new research may allow bioarchaeologists to circumvent some of these complications.

Working with colleagues at the University of Copenhagen, the Ancient Biomolecules Centre at Oxford (which Dr Willerslev has just joined; see box entitled below) and elsewhere, Dr Willerslev has managed to find fragments of DNA from bacteria, fungi, plants and even mammals in frozen soil cores up to 350 000 years in age.

A paper published earlier in 2003 in Science detailed gene fragments from 19 different plants, and eight animals, including musk ox, mammoth, lemming, reindeer and horse. Though DNA appears to persist for longer in frozen soil, tests have also turned up DNA from ancient plants and several species of extinct moa bird from sediment samples taken in high altitude New Zealand caves. A recent paper in Nature detailed further ancient DNA analysis on these birds.

Perhaps the most significant aspect of this research, however, was the fact that the entire diversity of DNA from the Siberian sites came from just seven samples, each weighing less than two grams, implying that ancient DNA can be commonly found. "Permafrost covers 25 per cent of the earth’s land surface," says Dr Willerslev, suggesting that there is a lot of potential information about past ecosystems out there to find.

Plus, there is little vertical movement of sediment in permafrost, so each layer can be carbon dated to work out how old the preserved genetic material is. Contamination is minimized, as samples are removed from frozen soil cores under sterile lab conditions.

Cold start

Dr Willerslev first started looking at the possibility that permafrost cores might retain bacterial DNA for his doctorate in 2000, and soon moved on to search for plants. Bacteria are very abundant in soil, and so Dr Willerslev wasn't totally shocked to find preserved bacterial DNA. But the possibility that he would find traces of more complex organisms was much less certain. "Searching for DNA from vertebrates was really a long shot," he admits. Taking that chance was well worth the effort though.

The unexpected discovery - proving that DNA can persist frozen and separated from the cellular environment for great periods of time - opens up may new possibilities. It should allow experts to test the composition of environments, and the genetic diversity of long-gone people and animals, without searching for rare fossilized bones and teeth. Though he cannot be sure at this stage, Dr Willerslev believes the animal DNA may primarily come from dung and urine.

He now hopes to take this work one stage further at Oxford. His focus will be on studying the ancient people of the Arctic, by extracting soil cores from some of the oldest pre-Eskimo and pre-Inuit settlement sites, and searching for traces of these peoples' genetic heritage.

"I've always had an interest in ethnography," says Dr Willerslev, who, beginning at the age of 19, took time from his undergraduate course to complete a series of daring Siberian expeditions. He collected frozen animal bones and cultural artefacts from Siberian tribes in far-flung Yakutia, Magadan, and Kamchatka, which were later given to the Copenhagen Museum. Subsequent expeditions to gather permafrost cores have allowed him to experience equally harsh weather in Alaska, Canada and Greenland.

The search for ancient human DNA will now allow him to combine his interests in biology, archaeology and ethnography. The main problem is that although pre-Eskimo and pre-Inuit people of 5000 or more years ago are known to have been culturally different (their hunting strategies and artefacts, such as spear heads and dwellings, are totally different to Inuit equivalents, says Dr Willerslev) it is still unclear where these people came from. They may have been Inuit ancestors, or have come from different ethnic stock.

Intriguingly, adds Dr Willerslev, Inuit folklore records descriptions of a long-gone 'tall and very different-looking people', once known to share their land. However, though many archaeologically significant, ancient settlement sites are dotted throughout the Arctic regions, no bones have been found to shed light on the enigma. That may be because the dead were put out to sea, says Dr Willerslev. But they may still have left their trace on solid ground, and DNA from soil cores, though fragmented, might provide clues as to which modern people these early inhabitants were most closely related to.

In order to answer questions about the movement of ancient people in the Arctic and into the New World, Dr Willerslev has got permission to take soil cores from a range of archaeologically significant sites. These include the Bluefish caves in Northern Yukon, Canada (thought to be one of the earliest human settlement sites known in North America), and the several hundred thousand-year-old Diring hominid settlement site in Siberia.

Beyond humans

The methods could also be used to answer questions about domestication of animals, and Dr Willerslev also plans to look for dog DNA. Testing the diversity and distribution of modern bacterial and viral pathogens will also be high on the agenda. Many modern diseases are thought to have taken grip with the switch from the nomadic hunter-gatherer lifestyle, in favour of more permanent communities (bacteria responsible for syphilis and tuberculosis for example, and diseases caused by parasitic worms seem to have appeared at this time). Finding even small fragments of human viruses could provide very important data for scientists struggling to understand how these pathogens evolve, says Dr Willerslev.

Soil sampling avoids the thorny political issues sometimes surrounding the removal of buried human remains, or the difficult decisions required when destroying parts of museum specimens to search for ancient DNA, says Dr Willerslev, citing further advantages of the technique.

There are limits, however. No DNA has been found so far in sediments older than 300 000 to 400 000 years old, so for the time being scientists may have to rely on conventional palaeontological methodology to test hypotheses about the very distant past.

External links

• Eske Willerlev: Research interests
• Ancient Biomolecules Centre
• Science
• Nature