Electron Coupled Proton Transfer in Ferredoxins

Ferredoxins are proteins that contain one or several iron-sulphur clusters. This makes such proteins eminent in electron transfer reactions. FdI is a particularly intensely investigated iron-sulphur protein for which also mutants have been crysallized and spectroscopically characterized. This allows to compare theoretical investigations with a variety of experimental data and validate the approach taken.

Because Fd I (PDB code 7FDR) contains a [3Fe-4S] and a [4Fe-4S] cluster, first parameters for describing the electrostatics and the bond-, angle-, and dihedral-terms in the molecular dynamics force field have to be derived. This is done using ab initio calculations. Density functional theory has shown to be very suitable for such investigations. We used UB3LYP with the 6-31G** basis set to calculate optimized structures of the isolated [3Fe-4S](0/+) and the biologically more relevant [3Fe-4S](0/+)(SCH3)3 cluster which is closer to the situation in the protein. Charges are calculated using standard Mulliken, the ESP and NBO method. All three charge sets are used in the molecular dynamics calculations.

Analysis of the experimental results suggested that the protonation occurs directly from the protein matrix to the FeS cluster. However, in our MD simulations, we observed a number of other candidate-hydrogen atoms that could possibly serve as proton donors. In particular, the suggested proton at Asp15 is generally too far away from the FeS cluster to serve as an ideal candidate. The other protons from surrounding side chains are usually strongly bound. A possible reaction mechanism involves the additional protonation of the protein side chain from solvent water which would make the NH of Asp15 less acidic. However, an attractive alternative to direct proton transfer from the protein matrix is the diffusion of water into the active site. We have shown that such a water molecule is dynamically stable and indeed could serve as a possible proton donor.[1] Also, differences between a fast (native protein) and slow (D15E mutant) variant of the protein were observed in the MD simulations that could account for the observed slowdown of an order or magnitude upon mutating Asp15-->Glu15.[2,3]

[1] M. Meuwly and M. Karplus, Farad. Disc. Royal. Soc. 124, 297 (2003)
[2] K. Chen, J. Hirst, R. Camba, C. A. Bonagura, C. D. Stout, B. K. Burgess and F. A. Armstrong, Nature, 405, 814-817 (2000)
[3] M. Meuwly and M. Karplus, Biophys. J., in print (2004)