Theoretical studies of blue copper proteins

Sammanfattning: In this thesis, a critical investigation is presented about the role of mechanical strain for the electron-transfer properties of the blue copper proteins. It has been found that the structure of a realistic model complex of the oxidised blue-copper active site, optimised in vacuum with the B3LYP method, has the same trigonal cupric geometry as in the protein. Thus, the protein does not strain the active site significantly. Instead, the protein presents a soft and polarisable cysteinate ligand that can form covalent bonds with the copper ion. This reduces the charge of the Cu(II) ion giving it properties more similar to a Cu(I) ion. The trigonal cupric geometry arises due to a covalent Cu(II)-S(Cys) pi bond, formed from the singly occupied Cu 3d orbital and a lone-pair on the sulphur atom. This way, it occupies two positions in a square coordination and forces the methionine ligand to an axial position at a longer distance. A second minimum similar in energy to the trigonal structure has been found. Its properties are comparable to the rhombic type I copper proteins (compared to the axial type I copper proteins), such as a longer Cu-Cys, shorter Cu-Met distance, and a very different spectrum. The electronic structure differs from the trigonal state in that it has an ordinary sigma interaction between the copper and sulphur atoms. However, because of the charge donation to the copper ion, the structure is almost tetrahedral and far from the expected square-planar geometry of normal Cu(II) complexes. Further, we have studied the suggestion that the protein can modulate the reduction potential by constraining the axial ligation. However, we find that this leads only to minor changes, less than 140 mV, whereas the natural variation among the proteins span a range of 800 mV. Thus, it is concluded that this would not be an appropriate way of fine-tuning the reduction potential. A similar investigation has been performed for the binuclear CuA site in cytochrome c oxidase. Finally, we have studied the inner-sphere reorganisation energy of several metal complexes, ranging from inorganic complexes to models of active sites for different proteins. This has shown that the blue-copper active site has a 120 kJ/mole smaller reorganisation energy than a copper ion with four water ligands. Thus, there does not seem to be any need for strain to achieve a low reorganisation energy. Instead, it is ensured by structures that are similar for both redox active forms and shallow metal-ligand potentials.