Elucidation of the Role of Redox-Active Disulfide Switches in Redox Signalling

Significant Achievements, Anticipated Outcomes and Future Work

The reactivity (and thus redox activity) of a disulfide bond is critically dependant on how much it is strained. Most of this strain will be accommodated through torsion of the dihedral angles. It is therefore important to have an accurate measure of how the stability cystine bridges changes as they are twisted. Previously this has been estimated using an empirical function from the Amber force field. A more accurate estimate was desired. The first step in this project was therefore to generate a full potential energy surface (PES) for the torsion of a disulfide bond.
Using diethyl disulfide as a model system, benchmark calculations were performed with the G3X method on the minima and important saddle points of the torsional PES. After comparison with calculations performed at lower levels of theory, MP2(full)/6-31G* was chosen as the most reliable method for mapping the PES. This was then calculated for the torsion around the three important dihedral angles of diethyl disulfide (corresponding to the three central dihedral angles in cystine). Energies were calculated at 10 degree increments for each of the critical parameters.
The resulting PES could then be used to predict the relative stabilities of the disulfide bonds found in the Protein Data Bank, using only their crystallographic torsional angles. Although it would have been preferable to fit a functional form to the PES, the surface was deemed to be too complex. Neverthless, the small grid size used produced a surface that was sufficiently smooth that it was possible to use simple linear interpolation to predict the relative disulfide stabilities. This revealed several conformations which were much more highly populated than would have been expected based on their stabilities, making them likely candidates for being involved in redox processes.
Further investigations on larger model systems have also been performed in order to measure the effect of the final two dihedral angles on the disulfide stability as well as the effects of chirality introduced by the protein backbone.
This work is currently on hold as Dr Haworth is on sabbatical leave in Germany during 2006. When she returns we plan to start investigating the redox potentials and dissociation mechanisms for some of the interesting disulfides identified from our PES. This will be done using semi-empirical and QM/MM methods.

Computational Techniques Used

Thus far, work on this project has exclusively employed the Gaussian03 suite of programmes. Most calculations have used relatively low level ab initio methods (MP2 and B3LYP) however much higher level calculations (eg. using the composite G3X method) have also been performed for benchmarking purposes. In the future we will continue to use Gaussian03, moving into modelling much larger systems with methods such as ONIOM and exploring the effects of solvating media on our systems using solvation schemes such as COSMO and IPCM. It is also likely that we will want to perform calculations using other software packages such as Amber and Vamp.

Publications, Awards and External Funding

External Funding and Awards

None.

Publications

N. L. Haworth, L. L. Feng, M. A. Wouters, High Torsional Energy Disulfides: Relationship Between Cross-Strand Disulfides And Right-Handed Staples, Journal of Bioinformatics and Computational Biology, 4, 2006, 155-168.