The sequencing of ancient DNA has enabled the reconstruction of speciation, migration and admixture events for extinct taxa1. However, the irreversible post-mortem degradation2 of ancient DNA has so far limited its recovery—outside permafrost areas—to specimens that are not older than approximately 0.5 million years (Myr)3. By contrast, tandem mass spectrometry has enabled the sequencing of approximately 1.5-Myr-old collagen type I4, and suggested the presence of protein residues in fossils of the Cretaceous period5—although with limited phylogenetic use6. In the absence of molecular evidence, the speciation of several extinct species of the Early and Middle Pleistocene epoch remains contentious. Here we address the phylogenetic relationships of the Eurasian Rhinocerotidae of the Pleistocene epoch7–9, using the proteome of dental enamel from a Stephanorhinus tooth that is approximately 1.77-Myr old, recovered from the archaeological site of Dmanisi (South Caucasus, Georgia)10. Molecular phylogenetic analyses place this Stephanorhinus as a sister group to the clade formed by the woolly rhinoceros (Coelodonta antiquitatis) and Merck’s rhinoceros (Stephanorhinus kirchbergensis). We show that Coelodonta evolved from an early Stephanorhinus lineage, and that this latter genus includes at least two distinct evolutionary lines. The genus Stephanorhinus is therefore currently paraphyletic, and its systematic revision is needed. We demonstrate that sequencing the proteome of Early Pleistocene dental enamel overcomes the limitations of phylogenetic inference based on ancient collagen or DNA. Our approach also provides additional information about the sex and taxonomic assignment of other specimens from Dmanisi. Our findings reveal that proteomic investigation of ancient dental enamel—which is the hardest tissue in vertebrates11, and is highly abundant in the fossil record—can push the reconstruction of molecular evolution further back into the Early Pleistocene epoch, beyond the currently known limits of ancient DNA preservation.
Bibliographical noteFunding Information:
Acknowledgements E.C. and F.W. are supported by the VILLUM FONDEN (grant number 17649) and by the European Commission through a Marie Skłodowska Curie (MSC) Individual Fellowship (grant number 795569). E.W. is supported by the Lundbeck Foundation, the Danish National Research Foundation, the Novo Nordisk Foundation, the Carlsberg Foundation, KU2016 and the Wellcome Trust. E.C., C.K., J.V.O., P.R. and D.S. are supported by the European Commission through the MSC European Training Network ‘TEMPERA’ (grant number 722606). M.M. and R.R.J.-C. are supported by the University of Copenhagen KU2016 (UCPH Excellence Programme) grant. M.M. is also supported by the Danish National Research Foundation award PROTEIOS (DNRF128). Work at the Novo Nordisk Foundation Center for Protein Research is funded in part by a donation from the Novo Nordisk Foundation (grant number NNF14CC0001). M.R.D. is supported by a PhD DTA studentship from NERC and the Natural History Museum (NE/K500987/1 & NE/L501761/1). K.P. is supported by the Leverhulme Trust (PLP -2012-116). L.R. and L.P. are supported by the Italian Ministry for Foreign Affairs (MAECI, DGSP-VI). L.P. was also supported by the EU-SYNTHESYS project (AT-TAF-2550, DE-TAF-3049, GB-TAF-2825, HU-TAF-3593 and ES-TAF-2997) funded by the European Commission. L.D. is supported by the Swedish Research Council (grant number 2017-04647) and FORMAS (grant number 2015-676). M.T.P.G. is supported by ERC Consolidator Grant ‘Extinction genomics’ (grant number 681396). L.O. is supported by the ERC Consolidator Grant ‘PEGASUS’ (grant agreement number 681605). B.S., J.K. and P.D.H. are supported by the Gordon and Betty Moore foundation. B.M.-N. is supported by the Spanish Ministry of Sciences (grant number CGL2016-80975-P) and the Generalitat de Catalunya, Spain (grant number 2017SGR 859). J.A. is supported by the Spanish Ministry of Sciences (grant number CGL2016-80000-P). R.F. is supported by National Science Foundation (grant number 1025245). The ancient DNA analysis was carried out using the facilities of the University of Luxembourg, the Swedish Museum of Natural History and UC Santa Cruz. We acknowledge support from the Science for Life Laboratory, the National Genomics Infrastructure (Sweden) and UPPMAX for providing assistance with massive parallel sequencing and computational infrastructure. Research at Dmanisi is supported by the John Templeton Foundation (grant number 52935), and the Shota Rustaveli Science Foundation (grant number 18-27262). We thank B. Triozzi and K. Murphy Gregersen for technical support.
© 2019, The Author(s), under exclusive licence to Springer Nature Limited.