(1) Institute of Enzymology, Biological Research Center,
Hungarian Academy of Sciences,
Budapest H-1518, PO Box 7, Hungary
(2) University of Oxford, Laboratory of Molecular Biophysics,
The Rex Richards Building, South Parks Road,
Oxford OX1 3QU U.K.
The determination of the covalent state of Cys from the amino acid sequence of proteins is important from the view point of locating point mutation sites for engineered half Cystines, to enhance protein stability or to investigate folding kinetics, to locate the reactive sulphydril groups in enzyme reactions, or to reveal the full covalent structure which may help improving 3D structure predictions.
Due to the efficient DNA recombinant technics the sequence databases are approx. 100-fold bigger than the three dimensional ones and the gap rapidly grows, meanwhile ones more and more often faces with the luck of information on post translational covalent modifications as e.g. the oxidation of Cysteins forming disulphyde bridges or bonding prostethic groups. The information to locate disulphyde bridges can be revealed by time consuming experimental methods.
In this paper we present a new method to predict the covalent state of Cys residues from the sequence alone. The prediction is based on multiple sequence alignments, conservation analysis and on the observation that Cys tend to occur overwhelmingly in the same oxidation state in the same protein. The prediction performs above 82\% in the jack-knife test of 81 proteins. The results demonstrate that the multiple sequence alignment information is a very efficient tool to make distinction between the functionally important oxidised Cys and the reactive sulphidryl groups. Further, we provide an analysis on the Cys in different redox state considering the secondary structures, linked structures, buried positions.