The story behind the paper: Sedimentary DNA from a submerged site reveals wheat in the British Isles
Writing about web page http://www.sciencemag.org/content/347/6225/998
Late last month, I was proud to be joint last author on a paper in Science on the presence of wheat in the British Isles 8000 years ago. But how does a medical microbiologist come to be involved in a study on the intricacies of the Neolithic transition?
Well, like many of life’s greatest ventures, it all began in a bar…
I have to admit to a weakness for rounding the week off by a Friday evening trip to the bar. This started when I worked in Barts in 1980s and 1990s, where the Robin Brook Centre bar hosted many a lively conversation (and acted as a link to various melodramas, including an alleged murder, hostage taking and a police shoot-out: but that’s another story).
When I arrived at the University of Birmingham in 2001, I was delighted to discover the delights of the Bratby Bar, nestled within the university’s Staff House. During more than a decade of visits, I had the chance to chat to all sorts of people from across the University, from Pro-Vice-Chancellors to post-docs. Fortuitously, John Heath (formerly Head of Biosciences, latterly Birmingham’s PVC for Estates) introduced me to Vince Gaffney, a garrulous landscape archaeologist from Geordieland (below).
Having recently set up a next-generation sequencing service and also having picked up on the excitement of ancient DNA research, at intervals I suggested to Vince that he should let us have some archaeological material to play with, to see if we could get any sequences out of it. Imagining we could tread in the footsteps of Schliemann or Carter, I had in mind something glamorous like a mummified hand or a skeleton from a ritual burial. Instead, we ended up with some mud! But mud of a highly precious and productive sort.
Vince was interested in understanding how the Neolithic transition (the spread of farming after the domestication of plants and animals) arrived in northwest Europe. The arrival of farming in this part of the world coincided with rising sea levels following the end of the last Ice Age. Vince had a track record in studying the landscapes that were inundated during this time and he was convinced the earliest clues to the arrival of the Neolithic in this part of the world would be found in these now-submerged sites.
Vince pointed me in the direction of some pioneering studies on sedimentary ancient DNA, which had established that DNA from macroscopic plants and animals could be detected in sediments even in the absence of macrofossils and could be used to reconstruct past environments. Two studies in particular stood out: one on the Viking settlements in Greenland and the other on the detection of sheep and moa DNA from outside a cave in New Zealand. It struck me that this was an exciting emerging field, fertile with opportunity.
Vince suggested that we try to detect signs of Neolithisation by searching submerged sediments for DNA from domesticated species that had no natural relatives in North Western Europe. That ruled out cows (wild relative: the aurochs) and pigs (related to wild boar), but made sheep and goats an attractive target. I pointed out to Vince that although we had the wherewithal to do the high-throughput sequencing and bioinformatics, it would be a rather fraught process trying to devise and implement protocols for target-specific amplification of ancient DNA. Instead, buoyed up by recent success with metagenomics on human faecal samples, I suggested that we try simple shotgun metagenomics—in other words we just extract DNA from the sediment cores and sequence it directly without any attempt at target-specific amplification or capture.
And then a period of turbulence descended on our academic lives…
I was headhunted and recruited to a new position at the University of Warwick in April 2013, while Vince was preparing to leave the University of Birmingham and eventually ended up at the University of Bradford. This could have signalled the end of the proposed research, but Vince and I were determined to continue with the work.
In fact, as luck would have it, my move to Warwick breathed new life into the project, as I hooked up with Robin Allaby from Warwick’s School of Life Sciences. Robin, seen here in the guise of a modern-day Jesus of the barley field, not only had a track record in the evolution of domesticated species, particularly plants, but had also established a dedicated ancient DNA laboratory at Warwick, ideal for performing DNA extractions from sediment cores.
I quickly persuaded Robin of the merits of the project and, as I was preoccupied with establishing a new Division of Microbiology and Infection, passed over to him day-to-day supervision of the work. Fortunately, Robin was able to recruit his recently graduated PhD student, Oliver Smith to the study. Oliver was an ideal candidate in having experience with ancient DNA studies, while also being between projects. Funding for the work came from my start-up package from Warwick Medical School, which paid for a sequencing instrument (an Illumina MiSeq), sequencing reagents and a salary for Oliver for nine-ten months.
By the middle of 2013, Vince had tracked down the perfect samples for the project—some 8000-year-old submerged sediment cores that had been collected from the Solent by an maritime archaeologist Gary Momber. Oliver extracted DNA from four samples of sediment in the ancient DNA lab and then sequenced them on our MiSeq. He and Robin then analysed the metagenomic sequences. Robin soon recognised that naïve use of existing metagenomics analysis pipelines was likely to turn up spurious results because of biases in what was represented in the databases (see recent Ed Yong's blog post on “discovery” of platypus DNA in Virginia and plague on New York subway), so he devised an improved method that avoided the problem.
Contrary to our initial hopes, Robin and Oliver did not discover any sheep or goat DNA. Instead, they discovered sequences from wheat, a domesticated plant that originated in the Middle East, with no close wild relatives in Northern Europe. This represented a triumph for metagenomics in an ancient DNA research, confirming two advantages of this approach over target-specific assays:
- It is open-ended, not just targeting what you expect to find, but also revealing the unsuspected.
- It is probably more sensitive than target-based amplification in garnering relevant information from billions of base pairs of unamplified DNA rather than amplified copies of just a few hundred base pairs of a sequence barcode.
After that, Robin played a key role in co-ordinating the writing and submission of a manuscript, carefully steering our paper through the reviewing and editorial process. And so, finally, we ended up with every academic’s dream-come-true—a paper in Science magazine!
Of course, my account of things here is heavily biased towards the role of sequencing and bioinformatics in this project. It is also important to recognise the key role played by our archaeological collaborators in framing the right questions, gathering the right samples, performing the palaeo-environmental analyses and providing the relevant contextual interpretation of the findings.
And this success brings a new challenge: what on earth is Warwick Medical School going to do with this high-impact paper in Science for REF2020, as I cannot see it flying with the clinical medicine Unit of Assessment! But we have five years to work on that problem!
Let me close by raising a figurative glass to toast the role of Birmingham’s Staff House Bar in all this! A note to all PVCs for Estates: shouldn’t all universities be investing in similar drinking establishments to catalyse new projects and facilitate collegiality? And a note to the relevant promotion panel in Warwick: shouldn’t it soon be Professor Robin Allaby. I’ll drink to both points!
Robin and I celebrating success at the top of the Shard.
Commentary on the Paper in Science: http://www.sciencemag.org/content/347/6225/945.summary