Ancient protein
[1] Ancients proteins have been recovered from a wide range of archaeological materials, including bones,[2] teeth,[3] eggshells,[4] leathers,[5] parchments,[6] ceramics,[7] painting binders[8] and well-preserved soft tissues like gut intestines.
[9] These preserved proteins have provided valuable information about taxonomic identification, evolution history (phylogeny), diet, health, disease, technology and social dynamics in the past.
[10] The introduction of high-performance mass spectrometry (for example, Orbitrap) in 2000 has revolutionised the field, since the entire preserved sequences of complex proteomes can be characterised.
However, like the research of aDNA (ancient DNA preserved in archaeological remains), it has been limited by several challenges such as the coverage of reference databases, identification, contamination and authentication.
[19] This golden trio drew many talented biologists, geologists, chemists and physicists to the field, including Marilyn Fogel,[20] John Hedges[21] and Noreen Tuross.
[30] Understanding how ancient proteins are formed and incorporated into archaeological materials are essential in sampling, evaluating contamination and planning analyses.
[31][32][33] However, the formation of proteinaceous tissues is often complex, dynamic and affected by various factors such pH, metals, ion concentration, diet plus other biological, chemical and physical parameters.
Dental calculi are defined as calcified biofilms, created and mediated by interactions between calcium phosphate ions and a wide range of oral microbial, human, and food proteins during episodic biomineralisation.
[40] For example, experimental studies demonstrate that robust, fibrous and hydrophobic keratins such as feathers and woollen fabrics decay quickly at room temperature.
[50] It is resistant to heating and enzymatic degradation; structurally, it has a beta-barrel associated with binding to small hydrophobic molecules such as fatty acids, forming stable polymers.
[58] It is a slow conversion with a long half-time, depending on adjacent amino acids, secondary structures, 3D folding, pH, temperature and other factors.
Type-1 collagen protein fibrils of a permafrost-preserved woolly mammoth (Yukon, Canada) were directly imaged and shown to retain their characteristic banding pattern.
This also constitutes the first time that collagen banding, or the molecular structure for any ancient protein, has been directly imaged with scanning electron microscopy.
Mineralised archaeological remains such as bones, teeth, shells, dental calculi and ceramics require an extra demineralisation step to release proteins from mineral matrices.
[67] To make ancient proteins soluble, heat, sonication, chaotropic agents (urea/guanidine hydrochloride, GnHCl), detergents or other buffers can be used.
[68] After demineralisation, protein solubilisation, alkylation and reduction, buffer exchange is needed to ensure that extracts are compatible with downstream analysis.
Currently, there are three widely-used protocols for ancient proteins and gels (GASP),[69] filters (FASP)[70] and magnetic beads (SP3)[71] can be used for this purpose.
[13] Recently, open search engines such as MetaMorpheus, pFind and Fragpipe have received attention, because they make it possible to identify all modifications associated with peptide spectral matches (PSMs).
[86][87][88] However, it may be challenging to evaluate the outputs of de novo sequences and optimisation may be required for ancient proteins to minimise false positives and overfitting.
Mineral-binding seems to stabilise proteins, but this is a complex, dynamic process that has not been systematically investigated in different archaeological and burial contexts.
Although minimally-destructive or non-destructive sampling methods are being developed for parchments, bones, mummified tissues and leathers, it is unclear if they are suitable for other types of remains such as dental calculi, ceramics and food crusts.
[108] Reference databases are also biassed towards model organisms such as yeasts and mouses,[109] and current sequence data may not cover all archaeological materials.
[112][113] The lack of post-translational modifications and subsequent experimental studies demonstrate that these sequences may be derived from bacterial biofilms, the cross-contamination of control samples or modern laboratory procedures.
