A simplified sample preparation for hair and skin proteins towards the application of archaeological fur and leather

Abstract

Conventional protocols for proteomics analysis usually start by extracting or solubilizing the proteins from their substrates. This step can be challenging for archaeological proteins, when they are heavily contaminated or decayed. The remains of animal fur/leather objects from an early medieval burial in Trossingen (580 CE) from Southwest Germany were submitted to proteomics analysis for species identification. One leather sample (TS3) yielded enough proteins to be identified as cow using a urea-based extraction (method “U”), confirming the microscopic identification. But two other samples (TS1 and TS2), compacted in a greyish brittle matrix with embedded hair visible only under microscope, could not be characterized with that method. A series of tests was performed using reduction/alkylation with tris(2-carboxyethyl)phosphine/chloroacetamide at 95 °C directly on the matrix (method “95C”), with or without the use of paramagnetic beads as cleaning procedure (from the single-pot solid-phase-enhanced sample preparation or SP3). Hair keratins were best recovered in the fur samples when digestion was performed directly on the insoluble fraction after reduction/alkylation. For both samples TS1 and TS2, an ovicaprine species was identified, with TS1 firmly identified as sheep due to the exceptional preservation of keratins and keratin-associated proteins. The simplified protocol also showed improvements on the identification of collagen in the leather sample TS3. SignificanceNorth European burials had a strong tradition of bodies wrapped or covered in animal skins; textiles, furs, items of leather and other organic materials were essential parts of grave furnishings (as part of the deceased's clothing as well as grave goods) but are mostly only preserved as residues, uncharacterized layers or stains. Even well preserved finds like the waterlogged organic remains from Trossingen show strong limitations for visual identification. Because the traditional protocol was unable to extract proteins efficiently from the soil matrix in which the samples were embedded, a new method was devised that enabled the determination of the sampled fur remains as sheep and the leather fragments as cow leather. Analyses showed that the key step for accessing the proteins in the soiled archaeological samples was heating for 10 min at 95 °C with a solution of tris(2-carboxyethyl)phosphine/chloroacetamide (TCEP/CAA). The protocol proposed in this study offers to work on minute samples (1 mg of sample or less) and overcame the challenge of separating the proteins from their archaeological matrix. It offers interesting perspectives for archaeological sites or objects where clothing are suspected but hardly detectable, such as burial sites.