Optimized DNA sampling of ancient bones using Computed Tomography scans

Abstract

The prevalence of contaminant microbial DNA in ancient bone samples represents the principal limiting factor for palaeogenomic studies, as it may comprise more than 99% of DNA molecules obtained. Efforts to exclude or reduce this contaminant fraction have been numerous but also variable in their success. Here, we present a simple but highly effective method to increase the relative proportion of endogenous molecules obtained from ancient bones. Using computed tomography (CT) scanning, we identify the densest region of a bone as optimal for sampling. This approach accurately identifies the densest internal regions of petrous bones, which are known to be a source of high-purity ancient DNA. For ancient long bones, CT scans reveal a high-density outermost layer, which has been routinely removed and discarded prior to DNA extraction. For almost all long bones investigated, we find that targeted sampling of this outermost layer provides an increase in endogenous DNA content over that obtained from softer, trabecular bone. This targeted sampling can produce as much as 50-fold increase in the proportion of endogenous DNA, providing a directly proportional reduction in sequencing costs for shotgun sequencing experiments. The observed increases in endogenous DNA proportion are not associated with any reduction in absolute endogenous molecule recovery. Although sampling the outermost layer can result in higher levels of human contamination, some bones were found to have more contamination associated with the internal bone structures. Our method is highly consistent, reproducible and applicable across a wide range of bone types, ages and species. We predict that this discovery will greatly extend the potential to study ancient populations and species in the genomics era.

The preservation of ancient DNA in archaeological fish bone

“…We observe no significant differences in either library preparation success or endogenous DNA levels amongst different bone element groups (i.e., cranial, pectoral girdle or postcranial) in archaeological fish bones. This observation is strikingly different from ancient DNA results obtained from mammalian bones, where high endogenous DNA preservation is localized, either in the petrous bone (Gamba et al, 2014; Pinhasi et al, 2015) or in the dense, recently deposited circumferential lamellae of long bones (Alberti et al, 2018). Moreover, our findings support a recent alternative hypothesis for DNA preservation in vertebrate bone.…”

“…Moreover, our findings support a recent alternative hypothesis for DNA preservation in vertebrate bone. The localized DNA preservation in mammals has usually been explained by the observed high density of bones or bone regions (Bollongino et al, 2008; Geigl & Grange, 2018; Kendall et al, 2018; Alberti et al, 2018) that may be more resistant to exogenous microbial colonization or taphonomic degradation (Campos et al, 2012; Gamba et al, 2014). Recently, however, it has been suggested that it is the absence of bone remodeling in these bones that helps promote DNA preservation (Kontopoulos et al, 2019; Sirak et al, 2020), following observations that the petrous bone (Kontopoulos et al, 2019), the auditory ossicle (Sirak et al, 2020) and the circumferential lamellae of long bones (Treuting et al, 2017) experience little or no bone remodeling.…”

Historical biogeography of the leopard (Panthera pardus) and its extinct Eurasian populations

“…Tracing the potential contribution of the European leopard to modern Caucasian or other modern leopard populations will require the analysis of nuclear DNA from Pleistocene leopards. Unfortunately, the Pleistocene samples analysed in the present study were not suitable for the analysis of nuclear DNA due to their low endogenous DNA content, despite multiple re-sampling attempts [47].…”

Consensify: a method for generating pseudohaploid genome sequences from palaeogenomic datasets with reduced error rates

“…4a). Recent discoveries such as the mammalian petrous bone as a source of high purity ancient DNA (Pinhasi et al, 2015), and improved knowledge of the distribution of contaminant DNA across different bone structures (Alberti et al, 2018; Damgaard et al, 2015) mean that achieving levels of genome coverage suitable for Consensify is increasingly possible. For example, each single-stranded ancient dataset analysed here was generated using relatively modest sequencing effort, being approximate to a single lane of sequencing on a current Illumina HiSeq platform (Barlow et al, 2018).…”