Protein disulphide isomerase (PDI) has been known for many years to assist in the folding of proteins containing disulphide bonds, but the exact mechanism by which it achieves this is only now becoming clear. The active site of PDI closely resembles that of the redox protein thioredoxin, and cDNA cloning has revealed a superfamily of proteins with related active-site sequences, in organisms ranging from bacteria to higher animals and plants. Recent mutagenesis studies are now helping to unravel the catalytic mechanism of PDI, and work in yeast and other systems is clarifying the physiological roles of the multiple PDI-related proteins.(Fig. I). It comprises approximately 500 amino acid residues, after the removal of an amino-terminal signal sequence that targets the protein to the ER lumen. Regions of sequence showing internal repetition or homology to other known proteins suggest that PDI is a modular protein. This is supported by the intron-exon pattern of the human PDI gene, which shows that the boundaries between the sequence modules correspond to intron-exon junctions, with each of the modules being encoded by one or two exons. In addition, we have shown recently that proteolysis of native bovine PDI occurs preferentially at the junctions between the sequence modules, which strengthens