Heritage Cases

In recent decades biomolecular analysis of ancient skeletal material has become integral to unravelling human prehistory and evolution in the Late Pleistocene. Sequencing of ancient DNA (aDNA) from bones/teeth has yielded revolutionary insights yet the scope of this approach is hampered by the poor preservation potential of DNA. Successful extraction of aDNA beyond 10,000 years old is extremely rare, and in tropical/subtropical regions DNA retrieval is challenging even 1,000 years into the past. In contrast, proteins contain rich evolutionarily informative variation and have a much higher preservation potential than aDNA so can provide insights deeper in time. The analysis of ancient proteins using mass spectrometry (‘palaeoproteomics’) has recently yielded important insights into the population history of extinct human and animal groups from a range of environments1,2,3. This emerging field allows the posing of questions not possible with aDNA and has the potential to revolutionise our understanding of human prehistory and palaeontology. The chances of successfully extracting proteomic data from ancient skeletal material can be increased by applying a range of techniques used to assess organic preservation prior to sampling5,6. However, these approaches generally require destructive sampling of small amounts of material (several milligrams) and exportation to a specialised lab. In contrast, in recent years near infrared spectroscopy (NIR) has been demonstrated to quickly quantify collagen content in bone, completely non-destructively7. NIR instruments can be easily transported to a museum or excavation, making the technique ideal for the pre-screening of palaeontological and archaeological materials. In the Ancient Genomics Lab at the Francis Crick Institute, we are developing NIR models as a non-destructive pre-screening protocol for palaeoproteomic analysis to identify well preserved material and limit destructive sampling of skeletal material. In order to use NIR to predict protein preservation in new skeletal samples, we have constructed calibration models from a wide range of skeletal material where we have characterised the protein content through NIR scanning and proteomics. Our dataset contains animal bone/dentine from a range of environments and time-periods to form the basis of the NIR models, aiming to cover the full range of proteome preservation. We have constructed prediction models from a range of faunal material from the past 1.3 Ma, predominantly from European contexts. We are expanding these models to include skeletal material from a wider geographic and temporal range to make these prediction models as widely applicable as possible. South Africa is a key region for understanding human evolution, with many crucial prehistoric archaeological and fossil sites. We would like to include data from South African skeletal material in the NIR prediction models to enable future non-destructive screening of protein content in South African material. Pre-screening of protein preservation in bone is widely applicable for multiple biomolecular analyses, including palaeoproteomics, 14C dating of bone and isotopic studies of bone collagen for dietary, environmental and migration studies. Screening identifies skeletal material with sufficient preservation for successful analysis, allows targeted sampling approaches and limits unnecessary destructive sampling. We have identified six South African sites in the National Museum, Bloemfontein collection (Curator: Dr Lloyd Rossouw, Head of Department at the Florisbad Quaternary Research Station), that represent a diversity of bone and fossil protein preservation in ancient skeletal material from South Africa. In particular, a sequence of material from five sites in the Free State will provide a time transect of protein preservation in one region from 3,000 BP to ~2.6 million years ago. The material is housed at the Florisbad Quaternary Research Station. I am applying for permission to export samples of bone, tooth dentine and tooth enamel from 4-12 faunal specimens per site from six sites: 1. Kareepan, Free State - ~3,000 years old 2. Mahemspan, Free State - ~36,000 years old 3. Florisbad, Free State - MSA Level ~120,000 BP, Spring Context ~260,000 years old 4. Cornelia-Uitzoek, Free State - ~ 1,000,000 years old 5. Matjhabeng, Free State - Late Pliocene ~2,600,000 years old 6. Besaansklip, Western Cape - ~20,000 years old Non-destructive NIR scanning and sampling using a rotary drill will be carried out at the Florisbad Quaternary Research Station and only the sub-samples will be exported for analysis. ~300-50 mg material will be sampled of each tissue type from each element, targeting broken ends or broken surfaces for sampling. In cases where a fragment has already broken off the specimen and is loose in the sample bag, these will be collected rather than performing further sampling on the original specimen. The samples will be hand-carried or shipped with a commercial courier to London, UK. The proteomic analysis will be carried out in dedicated facilities for ancient biomolecules in the Ancient Genomics Lab at the Francis Crick Institute in London. Protein extraction of samples will be carried out using established protocols and proteomes will be sequenced using liquid chromatography tandem mass spectrometry (LC-MS/MS). Small sub-samples of the material will be analysed with HPLC (~1 mg) for amino acid analyses and FTIR at the University of York, UK. NIR and proteomic data will be generated through the study, allowing us to: 1) Test and develop NIR protein prediction models. We will explore the utility of a regional South African NIR prediction model compared to the general models. 2) Test protein preservation between skeletal elements. 3) Proteomic data will be related to the temporal and environmental conditions at each site to explore the time transect of protein preservation in the Free State of South Africa. This will inform on the potential of the region for future proteomic analyses to address archaeological, palaeoanthropological and palaeontological questions.