Comparative analysis of bone collagen extraction and mass spectrometric techniques for ZooMS and application to worked bone artefacts

Introduction: Zooarchaeology by Mass Spectrometry (ZooMS) is rapidly becoming a staple technique in archaeology. Presently, this collagen peptide mass fingerprinting technique utilises MALDI-based mass spectrometric approaches to characterise unique species marker peptides for the identification of ancient bone samples. MALDI-MS is an effective and high-throughput MS technique, but its applications have declined in recent years. It is now typically used in clinical research, meaning many laboratories do not have access to such instrumentation. In this study, we compare three established collagen extraction techniques with both MALDI-TOF/TOF and nanoLC-MS/MS approaches to compare the efficacy and efficiency of both MS techniques for the analysis of diagnostic collagen peptide markers of known domesticated animal species.

Methods: This study analyses bone collagen from a collection of common modern domesticated animal species using three established collagen extraction protocols - ammonium bicarbonate gelatinisation, and acid demineralisation in 0.6M HCl followed by acid soluble and acid insoluble fractionation using 30kDa ultra centrifuge spin filters. Extracted proteins were digested with trypsin and analysed using both an Applied Biosystems 4800 MALDI-TOF/TOF and nanoLC-MS/MS on a Thermo Q Exactive Orbitrap coupled with a Thermo Easy-nLC1000. An initial shotgun proteomic approach was used for nanoLC-MS/MS, followed by a semi-targeted scan of theoretical peptide m/z values for CO1A1 and CO1A2 for the target species. Database searching was conducted using MaxQuant against the curated SwissProt database with additional collagen sequences added for domestic species.

Results: A summary of MaxQuant database searching results can be found in Table 1. Several identified collagen peptides within each sample were found to be homologous across taxa, however diagnostic COL1A1 or COL1A2 peptides were identified for most of the seven modern samples (For example: Figure 2), as previously reported in the literature [2]. Interestingly, the diagnostic COL1A1 peptide [3] for Sheep was only identified in the AC sample, resulting in ambiguous Sheep/Goat identifications for the AI and AM samples. The Dog bone samples were remnants after cremation which impacted the efficacy of collagen extraction, resulting in inconclusive species identification. The peptides identified in these samples were mainly belonging to keratins, and other common laboratory contaminants. The Whale AM sample was found to have significant signs of exogenous bacterial and fungal contamination that was not present in the other Whale bone samples, resulting in a significantly higher protein and peptide count however this did not interfere with the identification of Whale COL1A1 peptide markers.