Modeling Deamidation in Sheep α-Keratin Peptides and Application to Archeological Wool Textiles

Abstract

Deamidation of glutamine (Q) and asparagine (N) has been recognized as a marker of degradation and aging in ancient proteins. Using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) to study deamidation in wool textiles, we identified eight peptides from α-keratin proteins in sheep wool that could potentially be used to assess the level of degradation. For each chosen peptide, the extent of deamidation was determined by comparing the calculated theoretical distribution with the measured distribution using a genetic algorithm that gives the best fit to the measured distribution. Variations in the levels of deamidation were observed between peptides and in modern wool samples buried for up to 8 years in which deamidation levels were relatively low under short-term burial. In contrast, deamidation was higher in archeological textile fragments from medieval sites ranging from the 9th to 13th century in York (United Kingdom) and Newcastle (United Kingdom) and from the 13th to 16th century in Reykholt (Iceland). Major differences were observed between the British and the Icelandic samples, showing a negative correlation between age of samples and levels of deamidation, but highlighting the effect of local environment. In addition, nanoscale liquid chromatography-electrospray ionization tandem mass spectrometry (nanoLC-ESI-MS/MS) data indicated that deamidation in wool's α-keratin was influenced by primary and higher-order structures. Predominance of deamidation on glutamine rather than asparagine in the archeological samples was attributed to a higher abundance of Q in the α-helical core domain of keratins, neighboring residues and steric hindrance preventing deamidation of N.

Wet degradation of keratin proteins: linking amino acid, elemental and isotopic composition

“…Pigmentation and dyeing/mordanting had no significant effect on the AA composition of the control samples. This finding was not consistent with the proteomic observation that dyed samples were more oxidised than undyed, due to the dyeing/mordanting process itself . It is likely that the sum of these changes was too small to be detected using the quantitative methods in the present study.…”

“…The AA concentration, % content and D/L values were significantly dependent on the site of burial, but not on the dyeing methods (with the exception of [Glx]% and [Ser]% only). This contrasted with proteomic characterisation of these samples, where the presence of alum was associated with higher extents of deamidation . The effect of duration of burial was not tested in the present study, but proteomic analysis showed that this also had significant effects .…”

“…). These changes were consistent with: (1) an increase in the proportion of hydrophobic AAs in the residual fibre; (2) deamidation and oxidation of AA side chains (Asn, Gln, Thr, His, Arg, Tyr, Phe); (3) the elimination of S from Cys under conditions of moist heat; and (4) the survival of the melanin fraction of the fibre while proteins are lost (Table ).…”

“…However, other processes occur in which different parts of the fibre degrade at different rates . Given the predominance of protein molecules in hair fibres, diagenesis is likely to be controlled by hydrolytic and oxidative degradation of proteins …”

“…The latter artificially accelerate protein degradation reactions to simulate longer‐term archaeological diagenetic processes . Both groups of samples have also been characterised proteomically …”