Unlocking the Past

by Martin Jones

  This remains a most unusual find. The brain was one of the organs the ancient Egyptians removed, as it was subject to autolysis, the self-triggering breakdown that initiates bodily decay. The soft tissues that most readily survive in the absence of water or air are skin, hair, gut and tendon. Muscle is not infrequently found, but brain is an unusual survival, although not confined to this woman alone. As the team dug further, another thirty-nine bodies came to light. Eight of these also had brain-like material within them, which the team was now in a position to subject to special scrutiny. Each excavated skull was subjected to a range of non-invasive scanning procedures, prior to close examination under a microscope. It transpired that the actual brains lay within the soft mass, and had shrunk to a quarter of their original size. The shrinking of these soft, granular, tan-grey organs had not, however, obscured what they were. Within the brain could be detected thalamus, basal nerve ganglia and the ventricular system. Under a microscope, individual cells could be made out, and through ethydium bromide staining the remains of ancient DNA visualized. Its concentration was estimated at about one part per million, that is 1 percent of the original concentration of DNA in the living brain.

  Pääbo was struck by the exceptional preservation within these Florida sinkholes. He made contact with the Miami anthropologist John Gilford in order to get hold of a similar brain from another sinkhole at Little Salt Spring. Working with his 7,000-year-old sample, he amplified and sequenced enough of the DNA to recognize not only that it was human DNA, but also that it was of a particular genetic lineage not hitherto encountered in the New World. Here was clear evidence that the uses of ancient DNA could extend beyond recognition to the uncovering of new genetic information, not available from living individuals.

  Over the following four years, Svante Pääbo and Allan Wilson applied the Polymerase Chain Reaction to a new set of ancient tissues, including Egyptian mummies, two extinct mammals, a marsupial wolf and a giant ground sloth, and, in addition, to a relatively young piece of dried pork. The last two yielded what were at the time the oldest and youngest specimens of ancient DNA. The ground sloth was 13,000 years old, and the dried pork four years old. There was a range of dates in between: the 140-year-old quagga, mummified humans between 300 and 4,000 years old, and an 8,000-year-old brain. The apparent state of survival followed no clear pattern. The youngest and the oldest, the sloth and the pork, were in much the same state of disrepair. Such DNA as survived was broken up into very short lengths, often of no more than 100 or so base-pairs. All the damage was apparently done in the first few years before full desiccation had taken effect. A very ancient body desiccated within days or weeks might retain more intact DNA than a very recent body dried over a period of months or years. In general, the time taken for even a rapid dehydration of something the size of a human body was sufficient to reduce ancient DNA to very short fragments, whatever its age. The lengths of ancient DNA that survive suggest that we can expect length fragmentations of the source DNA by a factor between a hundred and a million. The analyst of ancient DNA is looking at small, individual pieces of a very, very large jigsaw, with at most a partial glimpse of the lid of the box.

  Pääbo and Wilson’s work with PCR also cast some light on why the cloning had sometimes presented a different picture. What they noticed was that, although the PCR itself indicated very small quantities of highly fragmented DNA, the overall survival of DNA in dry tissue was significantly greater. There was some reason why a lot of the DNA was not responding to the PCR. When Pääbo looked more closely at the material, he noted a marked contrast in the pattern of bases that formed the cross-links in the double helix. The bases fall into two types. The ‘pyrimidines’ have a single ring structure, and include cytosine, thymine, and the latter’s RNA counterpart, uracil. The ‘purines’ have a double ring structure and include the remaining two bases, adenine and guanine. In Pääbo’s tissues, the pyrimidines had fared far worse than the purines. Judging from how much had survived, the pyrimidines were at least twenty times more susceptible to damage in mummified tissue than the purines. As we shall see in due course, different patterns prevail in different environments. What their comparisons between PCR and cloning led the researchers to suspect was that the latter was doing rather more than just amplifying the ancient DNA. A living organism has within its molecular toolkit a series ofrepair mechanisms. To stay alive, it has to maintain vast lengths of DNA sequence in good condition, and the living cell is continually repairing minor sequence flaws. What Pääbo concluded was that the host used to clone the damaged ancient DNA was doing something similar and creating cloning artefacts that gave a rather better impression of DNA survival than was actually the case. By contrast, the use of PCR had trimmed the amplification process down to its raw chemical mechanics. It was much more likely to provide an accurate reflection of what survived. Even with this approach, there are amplification artefacts and certain simple forms of repair, but PCR both gave a realistic reflection of ancient DNA, and pushed the sample size limits to their extreme. It was, in principle at least, possible to amplify a single surviving DNA molecule to quantities large enough to study. This enormous potential opened up the vistas of analysis, and the aspirations of a growing band of ancient DNA scientists. By the end of the 1980s the race was on for the oldest DNA, and journals such as Nature and Science were poised at the finishing line.

  into the distant past

  Things moved very fast. By 1990, attention had turned to the unmistakable outline of a leaf, darkened by long-term burial but retaining much of its original physical form. Botanists had given this extinct species the name Magnolia latahensis. The leaf came from a basin in Idaho popular with bikers and holiday-makers, at a place called Clarkia. The basin is made up of deposits of clay and ash that rapidly accumulated within an ancient lake 17-20 million years ago. The Clarkia site had been visited by fossil hunters for years, and plant tissue and fish remains found there had been the subjects of much study. Karl Niklas had looked at leaves from the site a few years earlier. He demonstrated that cell structures were discernible, such as the chloroplasts, the tiny green bodies within the cell that capture light and turn it into biochemical energy for the plant’s use. He also found that certain of the molecules that had endowed the leaf with its original colour remained intact. Now Edward Golenberg at Detroit, Michigan, went one step further and detected within those chloroplasts the remains of their DNA, thousands of times more ancient than any DNA Pääbo and Wilson had amplified.

  These leaves were compression fossils, caught up and compacted after some change in the drainage pattern accelerated the lake’s infill. Their excellent preservation resulted from the absence of oxygen at the lake bottom, an extreme version of what led to the preservation of peat discussed in the previous chapter. It was oxygen that was damaging the pyrimidine bases in the mummified specimens, and its absence from the Clarkia leaves might explain why, as well as being older by far than other ancient DNA specimens, the amplified sequence was also longer. The chloroplast DNA sequence was 790 base-pairs long. This Magnolia leaf had raised the stakes for ancient DNA. Clarkia in Idaho became a magnet for the newly emerging band of molecule hunters.

  Among these were the members of the Extinct DNA Study Group. These Berkeley palaeontologists and biologists had for some time been interested in tracking down DNA that was millions, rather than thousands, of years old. From the early 1980s they had been toying with the idea of a quite different source of ancient DNA than the mummified tissues so far examined. This source was to be found on the islands of the Dominican Republic where, 30 million years ago, lagoons were fringed by a ring of now extinct resinous trees. Beneath the canopy of the trees, a group of bees engaged in the sometimes dangerous task of gathering the sticky fresh resin that exuded from the trees. From time to time, these diminutive resin-collectors would get caught up in the object of their desire, and their struggle to escape would only drag them further into their viscous tomb. Millions of years later, the resin had fos
silized to amber, but the whole body, and even the delicate wings, had been captured and frozen in time. Not only had the amber matrix become a permanent prison, but it also formed an effective barrier against oxygen and other gases, and served as a mummifying agent. Among its constituents are sugars and compounds called terpenes which contribute to preservation in two ways. They dehydrate the entrapped specimen, and inhibit microbial breakdown. The Extinct DNA Study Group had their eye on this seemingly ideal molecular time-capsule. Among the group was the entomologist George Poinar Jr. His early speculations about these insects in amber inspired the author Michael Crichton to start work on Jurassic Park. A few years later, and spurred on by the Magnolia result, George Poinar worked with his son Hendrik and another California scientist, Raul Cano, to give substance to those speculations.

  After the 17-20-million-year-old fossil leaf, it was only a question of time before the Berkeley group managed to amplify the bee’s DNA. Others were on the same track. At the American Museum of Natural History in New York, Rob de Salle and David Grimaldi were looking at termites from similar Dominican amber. By 1992, both groups had succeeded in pushing back the age of the oldest DNA to 30 million years. By 1993, Hendrik Poinar and Raul Cano would have pushed the time range even further back, with an amber-embedded weevil, to 130 million years old. The time constraints were falling away. Brian Farrell at Colorado had a preliminary positive result from a 200-million-year-old fish, and Noreen Tuross from the Smithsonian Institute was attempting to amplify DNA from a 400-million-year-old brachiopod. It seemed that life’s fundamental code could be recalled from the depths of geological time, and if that code could be recalled, whole living worlds could be contacted and viewed in meticulous detail. But all was not plain sailing–there were clouds on the horizon.

  a time for caution

  Back in 1990, the Extinct DNA Study Group had not been the only ones to follow in Edward Golenberg’s footsteps. Up until the Magnolia publication, the front runner in the race for ancient DNA was emerging as Svante Pääbo, the Egyptologist-turned-molecular-biologist, now with Allan Wilson’s renowned lab in Berkeley. His sequences from Egyptian mummies and pickled brain tissue had caught everyone’s imagination, and he was enjoying the limelight in this novel and exciting field. However, as he travelled to Idaho in August of 1990, it was with the knowledge that there was a new kid on the block, whose DNA targets were thousands of times older than his. Pääbo needed to have a look for himself, and he did so with a certain amount of scepticism.

  His scepticism had a sound scientific basis, arising from DNA research in a quite separate field. He had read the papers of a cancer research scientist working in London. Tomas Lindahl’s research initially had nothing to do with fossils. He was concerned with the stability of DNA, a topic of great importance for understanding human health, for example in relation to ageing and cancer. A decade before anyone had thought of extracting ancient DNA, he had published estimates of the rates of breakdown of various parts of the DNA molecule. A process to which he paid particular attention was that of ‘depurination’. The purines were the two DNA bases, adenine and guanine, that proved to be more resilient to decay than their pyrimidine counterparts in the presence of oxygen but without water. In contact with water, however, the purine residues break down, and this process of depurination also weakens the sugar phosphate backbone to which the bases are attached. Working from Lindahl’s estimates for the rate of depurination, Pääbo was able to gauge how long a fragment of ancient DNA might survive when in contact with water. The sequence amplified by Golenberg was long by ancient DNA standards, measuring 790 base-pairs. Pääbo estimated that the survival of such a length could be measured, not in millions of years, but in thousands. From the consideration of chemical kinetics alone, the strand amplified from the 17-20-million-year-old fossil was at most 8,000 years old. Pääbo took his own samples for study in the Berkeley lab, and by early 1991 was ready to challenge Golenberg’s findings.

  In March 1991 the two came together to put their cases at the meeting rooms of London’s Royal Society. The audience was made up of the diverse collection of researchers who had been drawn into the molecule hunt. There were geologists, archaeologists, chemists, geneticists, taxonomists and many others. Among them were the senior scholars who had laid the foundations of the new field. They listened as the two young scientists stood up in turn to make their cases.

  Pääbo and Golenberg shared a rather attractive aura of enthusiasm common to researchers on the cusp of some new scientific departure, a discreet discomfort with the lapel mike, laser pointer and conference clothes, and an urge to get back to the field or the lab, where the real action was. Pääbo however had gently moved on from this position. He was no longer simply the bright young star of the field, but was getting used to a new role as traffic policeman in a convoy moving with rather too much momentum for its own safety. His counterpart was aware of this; he had already read a ‘cautionary note’ placed in a recent issue of Current Biology by Pääbo and Wilson. In a slightly nervous presentation, Golenberg acknowledged the need for care and control, but also pointed out there were now several drivers in the fast lane. From the same ancient Clarkia beds, researchers at Washington State University claimed to have sequenced DNA from a Taxodium leaf. An Idaho group had reported amplifying a sequence from oak of the same age. Then of course there were the amber specimens that went even further back.

  Pääbo was not to be deflected. He reiterated the kinetic arguments drawn from Lindahl’s work, and outlined with care his own group’s attempts to replicate Magnolia DNA. They had found no plant DNA within the leaf, but bacterial DNA was present. The inference was that Golenberg’s result arose from contamination. This was something of a turning point in the quest for ancient DNA. The excitement about new results had infected everyone, including the referees used by high-profile scientific journals. There was never any doubt that a method like PCR, that could theoretically find a single surviving molecule, was extremely vulnerable to contamination, even in the cleanest of labs. However interesting and sincere Golenberg’s arguments were, there was a palpable sense that the convoy of ancient DNA hunters should adopt a more cautious respect for the highway.

  That more cautious approach came naturally to the meetings’ organizer, Geoff Eglinton. He was an established figure in the science community and had no need to race for the pages of Nature and Science. He also had a lifetime’s experience of the problems of research into biomolecules, having looked for them everywhere from the oldest rocks in the ground to the first samples of moon dust brought back with the Apollo mission. This experience gave him the sense that the quest for ancient DNA was proceeding in a somewhat myopic manner. Different DNA scientists were asserting different things, and there was much discussion of whether the results of this or that lab could be believed or not, rather than how best their results could be explained. The key to scrutinizing the results was to broaden the analysis from DNA to other molecules. The reason something as delicate as a leaf survives at all, even during its lifetime, is because it wears a tough waterproof coat. Its waxy cuticle contains molecules that are among the most persistent in nature. If there is exceptional survival of the less durable molecules such as DNA within the leaf, then the persistence of that waxy protective exterior should presumably reflect that.

  Eglinton was another visitor to the Clarkia site, from which he returned with a sample of the darkened ancient leaves. Back in the lab, he took a scalpel to detach the outer layer of one of the fossil leaves from the Clarkia bed, and used a special solvent to dissolve what lipids remained. This solution was subjected to a highly sensitive method of organic analysis called GC/MS (gas chromatography/mass spectrometry). What the analysis demonstrated was that the waterproofing lipid molecules in the Clarkia leaves were in reasonably good shape. There may indeed have been pockets of cellular tissue protected from water, in which case Lindahl’s rate estimates need not apply. His studies had been on DNA in an aqueous medium. However, there was
something else rather odd about the Clarkia lipids.

  Eglinton did not stop there; he also wanted to contrast the lipid profile in what was visibly discernible as the leaf’s waterproof envelope and the lipid content elsewhere in the deposit. What one might expect is a sharp contrast between the two, confirming that the lipid signal received corresponded to the envelope itself. Instead, what he found was that the leaf lipids were not confined to the leaf itself. They had dispersed into the surrounding sediment. Rather like the Somerset Levels peats, the exceptional appearance of the leaves, buried so long ago under the Clarkia muds, betrays how speedily so many decay processes were arrested. However, just as in the more recent peat, they weren’t arrested completely. The Clarkia leaves changed colour, and we can guess from peat processes how that happened. The lipids also intermingled with the sediments, and we are less clear what was going on there. At some stage, living bacteria came on to the scene, as is clear from Pääbo’s results. When they came is less clear. They might represent a quite recent contamination. Could they, on the other hand, not have lived within the sediments for much longer, and could this be linked to the reworking of the lipids? These are questions that remain open.

  Scientific funding is fickle. As with economic predictions, it depends as much on ‘community confidence’ as on cool rationality. There were undoubtedly questions left unanswered by the Clarkia DNA papers, but the answers would variously be negative, cautious or complex, and perhaps lack the immediate impact a future front page of Nature would require. There was a sense of a downturn in the prospects for Magnolia DNA, and the science funding bodies chose to invest elsewhere. In Michigan, Edward Golenberg shifted of necessity to new areas of research without having gained the means to resolve the Magnolia issue once and for all.