Quantifying Bone Collagen Fingerprint Variation Between Species

Collagen is the most ubiquitous protein in the animal kingdom and one of the most abundant proteins on Earth. Despite having a relatively repetitive amino acid sequence motif that enables its triple helical structure, in type 1 collagen, that dominates skin and bone, there is enough variation for its increasing use for the biomolecular species identification of animal tissues processed or degraded beyond the amenability of DNA-based analyses. In recent years, this has been most commonly achieved through the technique of collagen peptide mass fingerprinting (PMF) known as ZooMS (Zooarchaeology by Mass Spectrometry), applied to the analysis of tens of thousands of samples across over one hundred studies in the past decade alone. However, a robust means to quantify variation between these fingerprints remains elusive, despite being increasingly required due to the shift towards a wider range of wild fauna and those that are more distantly related from currently known sequences. This is particularly problematic in fish due to their greater sequence variation. Here we evaluate the quantification of the relative closeness of collagen fingerprints between families using ANOSIM and a modified SIMPER analysis, incorporating relative peak intensity. Our results show a clear correlation between sequence differentiation and statistical distance of PMFs, indicating that the additional complexity of type 1 collagen in fish could directly affect the efficacy of biomolecular techniques such as ZooMS. Furthermore, this multivariate statistical analysis demonstrates that PMFs in fish are substantively more distinct than those of mammalian or amphibian taxa.