DNA in charred wheat grains from the Iron Age hillfort at Danebury, England.

Allaby, Robin G. ; Jones, Martin K. ; Brown, Terence A. 等

The genetic history of wheat is the story of the world's temperate staple food. Archaeologically, charred grains are the common way wheat is preserved. Study of burnt spelt wheat from the British Iron Age shows DNA is present, and begins to shows the wheat's character.

DNA and ancient wheat

Genetic change, always central to studies of prehistoric agriculture, still operates as the principal criterion for designating certain plant and animal taxa as 'domesticated' (Harris & Hillman 1991) rather than wild. To this end, much bio-archaeological research into early agriculture has examined the gross morphology of plant and animal remains for phenotypic features that on the one hand seem to correspond to agricultural selection pressures, and on the other possess evidence of a direct genetic basis. However, it has frequently proved difficult to consolidate these two requirements. Features relating to larger, more accessible fruits and seeds, or to smaller, more docile animals, have often been judged as resulting from selection, but in many cases the genetic basis to the trait is uncertain. In contrast, phenotypes with a clear genetic foundation, such as the ploidy-dependent features of wheat chaff (Kislev 1984; Gordon Hillman pers. comm.), are often difficult to account for in terms of agricultural selection pressures. In other words, we have clear evidence of past phenotypic change consistent with what we assume to have been agricultural selection pressures, and we know in broad terms that many of those changes are due, directly or indirectly, to genetic events, but our existing methods constrain us from tying these two strands strongly together.

The discovery of preserved DNA in a range of archaeological materials could transform our approach. If the obstacles of survival, degradation, contamination and taphonomy can be overcome, then such DNA as survives in ancient plants and animals has the potential to provide a direct link between genotype, phenotype and the cultural context. Not only would this open the way to precision, it could also liberate our thinking from the gross 'genetic events' of 'domestication' to the perspective of a continuous evolutionary dynamic, in which the constant restructuring in human society through space and time is reflected in an equally continuous process of phylogenetic response, sometimes minuscule, sometimes substantial, in the plants and animals with which humans have been most closely associated.

We are attempting to start towards this goal with one of the most abundant archaeological resources of agricultural relevance: preserved wheat seeds. In archaeological contexts, plants are preserved principally by one of four mechanisms: the partial or complete reduction to carbon by heat; the exclusion of water in desiccating environments; the exclusion of oxygen in anoxic, often waterlogged environments; and partial mineralization by calcium phosphate, calcium carbonate or, less commonly with archaeological remains, iron sulphide (pyrites). We do not know which types of preservation might be compatible with retention of DNA, but intuitively one might expect biomolecular decay to be retarded in dry and/or anoxic settings. This conjecture is supported by reports of ancient DNA in anoxically preserved plant remains from Miocene deposits at Clarkia, Idaho (Golenberg et al. 1990; Golenberg 1991; Soltis et al. 1992), and in maize cobs preserved through various combinations of carbonization and desiccation (Rollo et al. 1991; Goloubinoff et al. 1993), the latter papers also demonstrating the phylogenetic inferences that are possible. Wheat and other Old World cereals have been encountered in all four preservation states, but the different states vary in their geographical and temporal coverages. Anoxically preserved and dry preserved cereals have sometimes been encountered in rich and impressive assemblages (e.g. Jacomet & Schlichterle 1984; Rowley-Conwy 1991), but the fullest spatial and temporal coverage is associated with carbonized remains: carbonized remains are therefore the principal source material on which an ancient DNA approach to wheat bio-archaeology must be based. We have examined charred seeds of spelt wheat from the Iron Age hillfort at Danebury, England (Jones 1984; Nye & Jones 1991), directly dated by the conventional radio-carbon method to the second half of the 1st millennium BC (Cunliffe 1984), for the presence of ancient DNA.

Procedures and results

Ancient DNA studies are almost entirely dependent on the polymerase chain reaction (PCR; Saiki et al. 1988), a biochemical technique that results in amplification of ||micro~gram~ quantities of DNA from minute traces of starting material. The relatively large amounts of DNA provided by PCR are sufficient for genetic analysis by, for example, nucleotide sequencing (for a review of ancient DNA techniques see Brown & Brown 1992). In a PCR experiment the genetic region to be amplified is selected by two short, synthetic DNA molecules that are added to the reaction mixture. These molecules attach to the substrate DNA at either side of the target region and prime the amplification. In our experiments we targeted a 246 base pair segment of DNA from a polymorphic region, linked to the genes for the high-molecular-weight (HMW) glutenin proteins, chosen partly on technical and partly on scientific grounds. The technical advantage is that there are two pairs of HMW glutenin genes per genome in wheat, so hexaploid spelt wheat (with three genomes) has a total of 12 targets. Multiple targets increase the chances of a successful amplification from small amounts of starting DNA. In scientific terms the region is attractive because of its polymorphic nature. Sixteen different allelic forms of the HMW glutenin genes have been identified in modern European cultivars of wheat, various combinations occurring within a single genotype (Payne 1987), and many additional alleles are known in wild populations (Ciaffi et al. 1993). Six alleles have been characterized by nucleotide sequencing, each recognizable by diagnostic sequence features within the region targeted for amplification. Sequence analysis of the molecules obtained by PCR of ancient DNA might lead to identification of the HMW glutenin alleles present. Complete allelic typing of a single specimen would be extremely time-consuming, but even a partial allele set might allow broad phylogenetic comparisons to be made between different specimens. In addition, the role of the HMW glutenins in determining the viscoelastic properties of dough prepared from the grain (Flavell et al. 1989) has led to identification of modern alleles that confer 'good' or 'poor' bread-making qualities (Payne 1987). There is a certain appeal to the possibility that alleles conferring defined bread-making qualities might be identifiable in archaeological grain.

We subjected the charred grain to a standard nucleic acid extraction procedure, previously shown to be applicable to preserved material (Rogers & Bendich 1985). Initially the extracts were analysed by hybridization probing, the results suggesting that nucleic acid might be present (results not shown). Subsequently a PCR with the Danebury extract gave rise to amplification products of the expected size. DNA resulting from the PCR was cloned, and two nucleotide sequences obtained, both clearly derived from wheat HMW glutenin genes. Neither sequence was an exact match with any of the known alleles, though one (DANEB1) displayed just a single difference to the 5(X) allele. Single-position sequence changes can occur during PCR, this effect being more frequent when the substrate DNA is chemically-damaged, which is almost certainly the case with ancient DNA (Paabo 1989; Lindahl 1993). It is therefore possible that DANEB1 and 5(X) are identical. The second sequence, DANEB2, displayed eight nucleotide differences to the most closely similar of the known sequences. DANEB2 therefore appears to represent a novel HMW glutenin allele.

The major problem with ancient DNA studies lies with the difficulty in establishing that amplification products derive from DNA preserved in the specimens and not from modern contamination (reviewed by Brown & Brown 1992). PCR is so sensitive that DNA present in the air and working materials can easily give false-positive results. With archaeological material the problem is magnified by the possibility of contamination during the relatively uncontrolled conditions of the excavation. There is no mechanism for proving beyond doubt that an amplification product from ancient material is authentic; all that can be done is to take suitable precautions to minimize the chances of contamination and to carry out control experiments. Our project included precautions and controls equivalent to those adopted by other groups working with ancient DNA. We have also taken the added precaution in this pilot study of working with particularly rich and well-sealed assemblages from Danebury that were removed en bloc and immediately enclosed within aluminium foil, rather than being put through sieving or flotation. The only potential contamination on site would be from modern wheat pollen, which the collection regime was designed to avoid. In fact, the extraction procedure does not give significant yields of DNA from pollen, probably because of difficulties in disrupting the tough sporopollenin coat. Our laboratory technique incorporated standard precautions for avoiding contamination in PCR experiments, including the use of specialized equipment, autoclaved solutions, and a geographical separation between the rooms used for DNA preparation and PCR analysis. Our basic technique has proved reliable in projects involving PCR of human material, where contamination with personal DNA can be a major problem. The results of two control experiments are shown in FIGURE 2. With the 'water blank', a PCR with water rather than DNA extract gave no amplification product, showing that cross-contamination of samples did not occur during the PCR experiment. In the second control, no amplification product was seen when an 'extract blank' was used in the PCR. The extract blank was prepared by carrying out the entire extraction procedure without seeds, providing a rigorous test for poor technique and contamination of working solutions.

Discussion

Given these precautions and the results of the control experiments, we believe that an acceptable degree of confidence can be assigned to the results with the Danebury seeds. Not every extraction has succeeded, suggesting perhaps that not all seeds in the sample contain amplifiable DNA. In contrast, a single extraction with a second sample of charred grain -- 3300-year-old emmer wheat from Assiros Toumba, Greece (Jones et al. 1986; Jones 1987) -- has provided an amplification product of the correct size (results not shown). We are modifying the extraction and PCR techniques in attempts to improve the success rate with the Danebury seeds, and are extending the work to a brooder range of archaeological wheats in order to assess the overall potential of ancient DNA analysis. To obtain information from which phylogenetic inferences can be made, it will be necessary to study polymorphic sequences that enable different populations of archaeological wheats to be distinguished. The glutenin alleles may be suitable, though complete allelic typing of just one specimen would require a substantial number of cloning and sequencing experiments. Our current work will determine whether a phylogenetic analysis based on the glutenin genes is technically feasible.

Despite its archaeological importance, our knowledge of the chemical composition of carbonized plant tissue remains poor, and probably inferior to that of its geological counterpart, fusain (Scott 1989). While transformation by heat is clearly instrumental in creating a microenvironment hostile to soil biota and therefore perhaps permitting DNA preservation, visual inspection of carbonized specimens indicates a range of transformation levels, from burst, distorted specimens through to specimens more or less retaining their original shape and cellular structure, and with several deviations from the soot-black colour of full carbonization. As a working hypothesis we assume that such DNA as persists in carbonized specimens is present in uncarbonized domains that survive encased in transformed and thus protective tissue. An important step for future work is to corroborate, or otherwise, this hypothesis. A better understanding of the material may enable us to use the appearance of specimens as a means of preselecting individual seeds likely to contain ancient DNA. Analysis of individual seeds is desirable as most archaeological assemblages are probably not genetically homogeneous. The PCR technique is sensitive enough to amplify ancient DNA from a single seed, but this approach will be extremely time-consuming if there is no way of preselecting suitable specimens.

Acknowledgements. A preliminary report of this work has been published as part of a review article (Brown et al. 1993). We are very grateful to Glynis Jones, University of Sheffield, for supplying the Assiros Toumba grain. We also thank Gordon Hillman (University of London), Alan Clapham (University of Cambridge), Chris Howe (University of Cambridge), Paul Sims (UMIST), Robert Sallares (UMIST) and Keri Brown (UMIST) for helpful discussions. This work was begun with funding provided by the Biomolecular Palaeontology Special Topic of the Natural Environment Research Council, and subsequently supported by a research grant from the Science-Based Archaeology Committee of the Science and Engineering Research Council.

References

ANDERSON, O.D., F.C. GREENE, R.E. YIP, N.G. HALFORD, P.R. SHEWRY & J.-M. MALPICA-ROMERO. 1989. Nucleotide sequences of two high-molecular-weight glutenin genes from the D-genome of a hexaploid bread wheat, Triticum aestivum L. cv Cheyenne, Nucleic Acids Research 17: 461-2.

BIRNBOIM, H.C. & J. DOLY. 1979. A rapid alkaline extraction method for screening recombinant plasmid DNA, Nucleic Acids Research 7: 1513-23.

BROWN, T.A. & K.A. BROWN. 1992. Ancient DNA and the archaeologist, Antiquity 66: 10-23.

BROWN, T.A., R.G. ALLABY, K.A. BROWN & M.K. JONES. 1993, Biomolecular archaeology of wheat: past, present and future, World Archaeology 25: 64-73.

CIAFFI, M., D. LAFIANDRA, E. PORCEDDU & S. BENEDETTELLI. 1993. Storage-protein variation in wild emmer wheat (Triticum turgidum ssp. dicoccoides) from Jordan and Turkey. I. Electrophoretic characterization of genotypes, Theoretical and Applied Genetics 86: 474-80.

CUNLIFFE, B. (ed.). 1984. Danebury: an Iron Age hillfort in Hampshire, London: Council for British Archaeology. Research report 52.

FLAVELL, R.B., A.P. GOLDSBROUGH, L.S. ROBERT, D. SCHNICK & R.D. THOMPSON. 1989. Genetic variation in wheat HMW glutenin subunits and the molecular basis of bread making quality, Biotechnology 7: 1281-5.

FORDE, J., J.-M. MALPICA, N.G. HALFORD, P.R. SHEWRY, O.D. ANDERSON, F.C. GREENE & B.F. MIFLIN. 1985. Nucleotide sequence of a HMW glutenin subunit gene located on chromosome 1A of wheat (Triticum aestivum L.), Nucleic Acids Research 13: 6817-32.

GOLENBERG, E.M. 1991. Amplification and analysis of Miocene plant fossil DNA, Philosophical Transactions of the Royal Society of London (series B) 333: 419-27.

GOLENBERG, E.M., D.E. GIANNASI, M.T. CLEGG, C.J. SMILEY, M. DURBIN, D. HENDERSON & G. ZURAWSKI. 1990. Chloroplast DNA sequence from a Miocene Magnolia species, Nature 344: 656-8.

GOLOUBINOFF, P., S. PAABO & A.C. WILSON. 1993. Evolution of maize inferred from sequence diversity of an adh2 gene segment from archaeological specimens, Proceedings of the National Academy of Sciences, USA 90: 1997-2001.

HALFORD, N.G., J. FORDE, O.D. ANDERSON, F.C. GREENE & P.R. SHEWRY. 1987. The nucleotide and deduced amino acid sequence of an HMW glutenin subunit gene from chromosome 1B of bread wheat (Triticum aestivum L.) and comparison with those of genes from chromosomes 1A and 1D, Theoretical and Applied Genetics 75: 117-26.

HARRIS, D.R. & G.C. HILLMAN. 1991. Foraging and farming: the evolution of plant exploitation. London: Unwin Hyman.

JACOMET, S. & H. SCHLICHTERLE. 1984. Der kleine Pfahlbauweizen Oswald Heer's -- Neue Untersuchungen zur Morphologie neolithischer Nackweizenahren, in van Zeist & Caspari (1984): 153-76.

JONES, G. 1987. Agricultural practice in Greek history, Annual of the British School at Athens 82: 115-23.

JONES, G., K.A. WARDLE, P. HALSTEAD & D. WARDLE. 1986. Crop storage at Assiros, Scientific American 254(3): 95-103.

JONES, M.K. 1984. The plant remains, in Cunliffe (1984): 483-95.

KISLEV, M.E. 1984. Botanical evidence for ancient naked wheats, in van Zeist & Caspari (1984): 141-52.

LINDAHL, T. 1993. Instability and decay of the primary structure of DNA, Nature 362: 709-15.

NYE, S. & M.K. JONES. 1991. The plant remains: a quantitative analysis of crop debris, in B. Cunliffe (ed.), Danebury: an Iron Age hillfort in Hampshire: 439-47. London: Council for British Archaeology. Research report 73.

PAABO, S. 1989. Ancient DNA: extraction, characterization, molecular cloning, and enzymatic amplification, Proceedings of the National Academy of Sciences, USA 86: 1939-43.

PAYNE, P.I. 1987. Genetics of wheat storage proteins and the effect of allelic variation on bread-making quality, Annual Review of Plant Physiology 38: 141-53.

ROGERS, S.O. & A.J. BENDICH. 1985. Extraction of DNA from milligram amounts of fresh herbarium and mummified plant tissues, Plant Molecular Biology 5: 69-76.

ROLLO, F., F.M. VENANZI & A. AMICI. 1991. Nucleic acids in mummified plant seeds: biochemistry and molecular genetics of pre-Columbian maize, Genetical Research 58: 193-201.

ROWLEY-CONWY, P. 1991. Sorghum from Qasr Ibrim, Egyptian Nubia, c. 800 BC-AD 1811: a preliminary study, in J.M. Renfrew (ed.), New light on early farming: 191-212. Edinburgh: Edinburgh University Press.

SAIKI, R.K., D.H. GELFAND, S. STOFFEL, S.J. SCHARF, R. HIGUCHI, G.T. HORN, K.B. MULLIS & H.A. ERLICH. 1988. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase, Science 239: 487-91.

SANGER, F., S. NICKLEN & A.R. COULSON. 1977. DNA sequencing with chain-terminating inhibitors, Proceedings of the National Academy of Sciences, USA 74: 5463-7.

SCOTT, A.C. 1989. Observations on the nature and origin of fusain, International Journal of Coal Geology 12: 443-75.

SOLTIS, P.S., D.E. SOLTIS & C.J. SMILEY. 1992. An rbcL sequence from a Miocene Taxodium (bald cypress), Proceedings of the National Academy of Sciences, USA 89: 499-501.

THOMPSON, R.D., D. BARTELS & N.P. HARBERD. 1985. Nucleotide sequence of a gene from chromosome 1D of wheat encoding a HMW-glutenin subunit, Nucleic Acids Research 13: 6833-46.

VAN ZEIST, W. & W.A. CASPARI (ed.). 1984. Plants and ancient man: studies in palaeoethnobotany. Rotterdam: A.A. Balkema.