DNA diagenesis and palaeogenetic analysis: Critical assessment and methodological progress

Abstract

Palaeogenetic data obtained from fossilizing or fossil bones and teeth are of great importance to studies of vertebrate evolution, human biological and cultural evolution, plant and animal domestication and reconstructions of palaeoenvironment and palaeoecology. These studies are based on the retrieval of DNA preserved in fossilizing bones and teeth. DNA is present in fossils, if at all, in only very small amounts, which makes its amplification with PCR necessary for detailed sequence analysis. Erroneous nucleotides can be incorporated during in vitro amplification either because of post-mortem base damage of the original DNA template or simply because the fidelity of DNA polymerases is not absolute and can be decreased by suboptimal buffer conditions or possibly by compounds in the fossil extracts. These erroneously introduced nucleotides can be mistaken for authentic mutations of the ancient sequence compared to the closest extant sequence. Moreover, contamination by modern DNA, which is not chemically modified and therefore a better substrate for the Taq polymerase, can also lead to erroneous results.

Here, we will present the procedures that we have developed in order to (i) ensure negligible mutagenicity of the PCR reaction, (ii) eliminate contamination by DNA molecules originating from previous PCR reactions and cloning procedures, (iii) prevent contamination with modern DNA of fossil bones and teeth during and after their excavation, and (iv) prevent degradation of ancient DNA after excavation. Finally, we will discuss our results on DNA preservation as a function of the taphonomy of the skeletal part that is analyzed and of the depositional context of preservation.

Introduction

The study of DNA preserved in fossilizing bones and teeth (palaeogenetics) started some 20 years ago with the advent of the Polymerase Chain Reaction PCR (Saiki, 1985, Pääbo, 1989, Pääbo et al., 1989), which made it possible to amplify minute amounts of DNA to a level that allowed its analysis by the techniques of molecular biology. Expectations ran high in the research fields dealing with the deciphering of information preserved in fossilizing hard tissue and spectacular palaeogenetic results were obtained, (e.g., DeSalle et al., 1992, DeSalle et al., 1993, Woodward et al., 1994). These later turned out to be derived from modern DNA contamination, (e.g., Wang et al., 1997, Guttiérez and Marin, 1998) and it was only after strict authentication criteria had been established (Austin et al., 1997, Cooper and Poinar, 2000) that reliable data were produced and the field of palaeogenetics joined the community of the biological sciences. The authentication of palaeogenetic data, however, remains a major concern.

In fact, since DNA integrity in living tissue is maintained by permanent and precise DNA repair processes that cease immediately when the organism dies, DNA in a dead body is degraded by various mechanisms. Biological processes such as autolysis and putrefaction, as well as chemical processes such as hydrolysis and oxidation, degrade DNA leading to modifications of bases, which no longer encode the genetic information of the original molecules, and to severe fragmentation of the DNA macromolecules. Thus DNA is present in fossilizing hard tissue such as bones and teeth (which we will hereafter simply call “fossils”), when at all, in only minute amounts and as small fragments. This makes its amplification necessary, which in turn causes serious contamination problems. Indeed, PCR amplification of ancient DNA is considerably less efficient than amplification of modern DNA. As a consequence, a few contaminating modern DNA molecules are sufficient to mask the presence of authentic, endogenous DNA and the DNA sequences obtained may be mistaken for those of ancient DNA (e.g., Hofreiter et al., 2001b). This constitutes a major problem, especially when remains from humans and domestic animals are being analyzed. We therefore set out to make the palaeogenetic approach more reliable by optimizing the analytical conditions in such a way that the DNA sequences produced were authentic reflections of the DNA sequences preserved in fossils.

We critically analyzed the chain of events that constitute the procedures to which fossils are subjected from the moment they are excavated to their final analysis, and studied their potential to degrade preserved DNA and to contaminate the samples with modern DNA. It quickly became clear that several of the procedures involved in the analysis of both fossils and their DNA can have disastrous effects. These need to be considered in order to minimize the production of palaeogenetic results that are either false-positive (due to contamination and error-prone analytical methods) or false-negative (due to post-depositional DNA degradation). We have therefore developed methods of analysis that minimize the identified pitfalls.