6.1. Region c of PDI is not crucial for the main functions of the protein, but the end of domain a’ contains an area sensitive to mutations

The proposed peptide-binding site of PDI, as identified by the use of a tripeptide probe and photoaffinity labelling, in a single tryptic peptide that corresponded to a 27 amino acid sequence beginning at the end of a’ and extending over the highly acidic C-terminal region c (Noiva et al. 1993) was studied here for its role in C-P4H subunit function and for the isomerase and chaperone functions of PDI. The results presented here indicate that while the c region is totally expendable for these functions, the end of domain a’ plays a role since these functions are sensitive to changes in this region, as studied by a range of small and larger deletions and single amino acid mutations.

The region in a’ that was identified as being sensitive for mutations in the C-P4H function and the chaperone and isomerase functions consists, according to secondary structure predictions, of its most C-terminal α helix and the preceding loop region while the β -strand preceding these was not identified as critical. Whether the mutations introduced in this region caused major structural changes was assessed by recording the CD spectra of three mutants. While some differences were seen in all of these compared to the wild type PDI, the most significant differences were seen in Δ449-451 mutant. The nature of these alterations cannot be assessed directly by CD; the differences indicating small changes in secondary structure, but giving no indication of possible changes in

tertiary structure or in domain-domain interaction. Subsequent studies showed that many of these mutants were indeed more sensitive to proteolytic cleavage due to conformational disruption (Klappa et al. 2000). Peptide binding studies using chemical cross-linking showed that mutants that were unable to form functional C- P4H tetramer were also less efficient in peptide binding. These mutants were more protease sensitive and it was also shown that proteolytic cleavage of F449R mutant resulted in a fragment that co-migrated with the abb’ domain construct of PDI and N-terminal sequencing revealed an intact N-terminus indicating that the whole of the a’ domain had been cleaved. This fragment was also able to bind peptides. It could therefore be that the mutations in the a’ domain indirectly inhibit peptide binding to domain b’, possibly by occupying the peptide binding site through intramolecular association or by causing subtle structural changes in the b’ domain that prevent the association (Klappa et al. 2000).