RESULTS

Sequential labelling of cysteines

In order to determine the relative accessibility of cysteines in wool the keratins and KAPs were subjected to a process of sequential alkylation as the concentration of DTT was raised in stepwise fashion from 0 to 15 mM.
A total of 38 peptides were detected and identified in the mass spectrum of protein extracts in the control run, involving wool that had not been treated with a reductant. Of these, two were from the epithelial keratin K10 and the rest were from either actin or serum albumin. A number of cysteines in the peptides were labelled with IAM but only in those from serum albumin.
In contrast cysteine labelling of keratins and KAPs was only detected when the wool was incubated with DTT. Among the labelled cysteines were a number identified in our earlier study 12, specifically: in the tail domain of K31, T-42 and T-45; in coil 1B of K31 and K33b, 1B-87; the head domain of K35, H-36; H-24 and H25 in K81 (H-29 and H30 in K83); and H61 and T-24 in K85.
In the first step, when a DTT concentration of 5 mM was used to reduce the proteins, eight peptides from four keratins and three KAPs containing IAM-labelled cysteines were identified in at least four out of the five biological repeats (Figures 2-4, Table 1). Interestingly, all the keratin peptides identified were labelled on either the head or tail domain. In the case of the type I keratins one peptide from the tail region of K31 was identified on which four cysteine residues were labelled. Peptides from three Type II keratins were also found to have IAM-labelled cysteines. For K83 this involved two peptides from the head domain of which three cysteines were found to be labelled. In the case of K85 a peptide containing one cysteine from its tail domain was labelled, while a peptide containing one cysteine from the tail domain of K86 was also labelled. IAM-labelled peptides were identified from three HSPs, specifically cysteines close to the N-terminus of KAP2.3 and the cysteines close to the C-terminus of KAP7.1, KAP11.1 and KAP13.1 (Figure 2, Table 1).
In the second step, proteins were unraveled further by reduction with 10 mM DTT and then labelled with acrylamide (Figures 2-4, Table 2). This resulted in the identification of a total of 62 cysteine-labelled peptides, of which 20 were from keratins and 42 from KAPs. Of the KAPs, 65% of were classed as HSPs, 16% UHSPs and the rest high glycine-tyrosine proteins (HGTPs). In the case of the type II keratins all the newly characterized peptides in this step (and hence cysteines) were either located in the head or tail domains (Figure 4). In contrast, in the type I keratins one acrylamide-labelled cysteine was identified in a peptide from coil 1B and another from coil 2, while an acrylamide-labelled cysteine was identified in a peptide from the L1 linker in K34 (Figure 3). No modified cysteines were found in the hendecad regions 3. As found previously12 around 90% of the labelled cysteines in the HSPs and UHSPs are found close to a proline residue in the sequence and over half within 20 residues of the N- and C-terminus of the proteins. Specifically these involved cysteines at the N-terminus of KAP3.3, KAP3.4 and KAP7.1, and cysteines close to the C-terminus of KAP31, KAP3.4, KAP4.2, KAP4.5, KAP9.2, KAP9.8, KAP10.11, KAP10.12, KAP12.2, KAP19.5 and KAP26.1 (Figure 2, Table 2). At the same time a number of cysteines were detected from the more central regions of some of the KAPs, among them KAP2.3, KAP3.4, KAP4.2, KAP6.1, KAP6.4, KAP10.4, KAP11.1,KAP13.1, KAP16.1-like, KAP19.3, KAP24.1 and KAP26.1.
In the third step very few IAA-labelled peptides were identified in the final alkylation step. Of the two keratin peptides identified the one from K31 was labelled in the coil 2 domain while that from K35 was labelled in the head domain (Figure 3, Table 3). Two KAPs were identified in this stage, one being KAP 6.1, an HGTP, and the second being from the HSP KAP13.1-like (Figure 2, Table 3).

Validation of the labelling experiments

To determine if the large increase in peptides identified in step 2 and the subsequent drop in peptide numbers in step 3 was due to the longer labelling time used with acrylamide the entire process was repeated using IAM in all steps. However, a similar pattern was observed when comparing the proportion of peptides identified between each of the steps. This led to the conclusion that the alkylation reagent used was not influencing the efficiency of the labelling and thus peptide identification.