Probing the modelled structure of wheatwin1 by controlled proteolysis and sequence analysis of unfractionated digestion mixtures

We set up a method to get rapid information on the three-dimensional struct ure of peptide and proteins of known sequence. Both native and alkylated po lypeptide is hydrolyzed with a number of proteases at different digestion t imes and the resulting mixtures are compared by HPLC analysis to establish the differences in the hydrolysis pathways of the folded and unfolded molec ule. Then, the unfractionated digestion mixtures of the native polypeptide are submitted to automatic sequence analysis to identify the hydrolysis sit es. The sequence of each fragment present in the mixtures is reconstructed and its amount determined by quantitative data of the sequence analyses. We used this approach to determine the amino acid surface accessibility of wh eat-win1, a pathogenesis-related protein from wheat, and constructed a pred ictive three-dimensional model based on the knowledge of the tertiary struc ture of barwin, a highly homologous protein from barley. The procedure allo wed us to quickly identify and quantify the hydrolysis at the susceptible b onds which could be classified as exposed, partially hidden, or inaccessibl e, The results were useful to evidentiate and discuss concordances dances a nd differences between experimental and model predicted accessibilities of amino acid residues. (C) 1999 Wiley-Liss, Inc.