Fast voltammetric studies of the kinetics and energetics of coupled electron-transfer reactions in proteins

A wealth of information on the reactions of redox-active sites in proteins can be obtained by voltammetric studies in which the protein sample is arranged as a layer on an electrode surface. By carrying out cyclic voltammetry over a wide range of scan rates and exploiting the ability to poise or pulse the electrode potential between cycles, data are obtained that are conveniently (albeit simplistically) analysed in terms of plots of peak potentials against scan rate. A simple reversible electron-transfer process gives rise to a 'trumpet'-shaped plot because the oxidation and reduction peaks separate increasingly at high scan rate; the electrochemical kinetics are then determined by fitting to Butler-Volmer or Marcus models. Much more interesting though are the ways in which this 'trumpet plot' is altered, often dramatically, when electron transfer is coupled to biologically important processes such as proton transfer, ligand exchange, or a change in conformation. It is then possible to derive particularly detailed information on the kinetics, energetics and mechanism of reactions that may not revealed clearly or even at all by other methods. In order to interpret the voltammetry of coupled systems, it is important to be able to define 'ideal behaviour' for systems that are expected to show simple and uncoupled electron transfer. Accordingly, this paper describes results we have obtained for several proteins that are expected to show such behaviour, and compares these results with theoretical predictions.

Backbone dynamics of azurin in solution: slow conformational change associated with deprotonation of histidine 35

15N relaxation measurements have been performed on the type Iota blue copper protein azurin from Pseudomonas aeruginosa. The relaxation times show that one loop (residues 103-108) and one turn (residues 74-77) display fast internal motions. The rest of the protein is rigid with an average order parameter S(2) of 0.85 +/- 0. 05. The copper binding site shows the same degree of rigidity even though is it composed of several loops and lies outside the beta-sheet sandwich. Substantial exchange broadening was found for a number of residues surrounding the side chain of His-35. The average exchange rate has been determined from NMR exchange spectroscopy experiments and is 45 +/- 6 s(-)(1) at 41 degrees C. The exchange broadening is caused by the protonation/deprotonation equilibrium of His-35. The NMR results indicate that the two structures of azurin observed by X-ray diffraction of crystals at pH 5.5 and 9.0 [Nar, H., Messerschmidt, A., Huber, R., Van de Kamp, M., Canters, G. W. (1991) J. Mol. Biol. 221, 765-772] are present in solution and that they interconvert slowly.

The structural role of the copper-coordinating and surface-exposed histidine residue in the blue copper protein azurin

Copper K-edge extended X-ray absorption fine structure (EXAFS) spectroscopy and (15)N NMR relaxation studies were performed on samples of a variant azurin in which the surface-exposed histidine ligand of the copper atom (His117) has been replaced by glycine. The experiments were performed to probe the structure of the active site and the protein dynamics. The cavity in the protein structure created by the His-->Gly replacement could be filled by external ligands, which can either restore the spectroscopic properties of the original type-1 copper site or create a new type-2 copper site. The binding of external ligands occurs only when the copper atom is in its oxidised state. In the reduced form, the binding is abolished. From the EXAFS experiments, it is concluded that for the oxidised type-1 copper sites the protein plus external ligand (L) provide an NSS*L donor set deriving from His46, Cys112, Met121 and the external ligand. The type-2 copper site features an S(N/O)(3) donor set in which the S-donor derives from Cys112, one N-donor from His46 and the remaining two N or O donors from one or more external ligands. Upon reduction of the type-1 as well as the type-2 site, the external ligand drops out of the copper site and the coordination reduces to 3-fold with an SS*N donor set deriving from His46, Cys112 and Met121. The Cu-S(delta)(Met) distance is reduced from about 3.2 to 2.3 A. Analysis of the NMR data shows that the hydrophobic patch around His117 has gained fluxionality when compared to wild-type azurin, which may explain why the His117Gly variant is able to accommodate a variety of external ligands of different sizes and with different chelating properties. On the other hand, the structure and dynamics of the beta-sandwich, which comprises the main body of the protein, is only slightly affected by the mutation. The unusually high reduction potential of the His117Gly azurin is discussed in light of the present results.

The effect of pH and ligand exchange on the redox properties of blue copper proteins

A study of the structure and redox properties of the copper site in azurins by means of EXAFS, NMR, redox titrations, potentiometry, equilibrium cyclic voltammetry and rapid scan voltammetry on protein films is reported. The results are discussed in light of existing theories on structure and function of type-1 copper sites. The exit and entry of electrons take place through the C-terminal histidine ligand of the copper. The hydrophobic patch through which this residue penetrates the protein surface plays an important role in partner docking (cf. The rim of the porphyrin ring sticking through the surface of the cytochromes-c). We find no experimental evidence for strain around the metal site. The active centre is able to maintain ET activity even in the presence of fairly gross disturbances of the site structure. The analysis of the thermodynamics of the redox reaction shows that the protein matrix and the solvent play an important role in 'tuning' the redox potential around a "design" value of around 300 mV at room temperature. The metal site appears "designed" to stabilise the Cu(II) instead of the Cu(I) form. The remarkable evolutionary success of the blue copper proteins is ascribed to the sturdy overall beta-sandwich structure of the protein in combination with a metal site that is structurally adaptable because three of its four ligands are located on a loop. The electronic "gate" that occurs in the middle of a hydrophobic patch allows for fine tuning of the docking patch for recognition purposes.

An amicyanin C-terminal loop mutant where the active-site histidine donor cannot be protonated

A novel blue copper protein was constructed by replacing the C-terminal loop of amicyanin (Paracoccus versutus) by the homologous loop of rusticyanin. The C-terminal loop of both amicyanin and rusticyanin contains three (His, Cys, Met) of the four copper ligands. The amicyanin mutant exhibits all spectroscopic properties normally encountered for blue copper sites. The midpoint potential (369 mV) is the highest reported value for an amicyanin mutant. Cyclic voltammetry and NMR studies of the reduced form indicate that, in contrast to wild-type amicyanin and all amicyanin mutants described so far, the C-terminal histidine ligand does not protonate in the accessible pH range (pKa<4.5).

Effects of dimerization on protein electron transfer

In order to investigate the relationship between the rate of protein-protein electron transfer and the structure of the association complex, a dimer of the blue copper protein azurin was constructed and its electron exchange properties were determined. For this purpose, a site for covalent cross-linking was engineered by replacing the surface-exposed asparagine 42 with a cysteine. This mutation enabled the formation of disulfide-linked homo-dimers of azurin. Based on NMR line-broadening experiments, the electron self-exchange (e.s.e.) rate constant for this dimer was determined to be 4.2(+/-0.7) x 10(5)M(-1)s(-1), which is a seven-fold decrease relative to wild-type azurin. This difference is ascribed to a less accessible hydrophobic patch in the dimer. To discriminate between intramolecular electron transfer within a dimer and intermolecular electron transfer between two dimers, the e.s.e. rate constant of (Cu-Cu)-N42C dimers was compared with that of (Zn-Cu)- and (Ag-Cu)-N42C dimers. As Zn and Ag are redox inactive, the intramolecular electron transfer reaction in these latter dimers can be eliminated. The e.s.e. rate constants of the three dimers are the same and an upper limit for the intramolecular electron transfer rate of 10 s(-1) could be determined. This rate is compatible with a Cu-Cu distance of 18 A or more, which is larger than the Cu - Cu distance of 15 A observed in the wild-type crystal structure that shows two monomers that face each other with opposing hydrophobic patches. Modelling of the dimer shows that the Cu-Cu distance should be in the range of 17 A < rCu-Cu < 28 A, which is in agreement with the experimental findings. For efficient electron transfer, it appears crucial that the two molecules interact in the proper orientation. Direct cross-linking may disturb the formation of such an optimal electron transfer complex.

Role of ligand substitution on long-range electron transfer in azurins

Azurin contains two potential redox sites, a copper centre and, at the opposite end of the molecule, a cystine disulfide (RSSR). Intramolecular electron transfer between a pulse radiolytically produced RSSR- radical anion and the blue Cu(II) ion was studied in a series of azurins in which single-site mutations were introduced into the copper ligand sphere. In the Met121His mutant, the rate constant for intramolecular electron transfer is half that of the corresponding wild-type azurin. In the His46Gly and His117Gly mutants, a water molecule is co-ordinated to the copper ion when no external ligands are added. Both these mutants also exhibit slower intramolecular electron transfer than the corresponding wild-type azurin. However, for the His117Gly mutant in the presence of excess imidazole, an azurin-imidazole complex is formed and the intramolecular electron-transfer rate increases considerably, becoming threefold faster than that observed in the native protein. Activation parameters for all these electron-transfer processes were determined and combined with data from earlier studies on intramolecular electron transfer in wild-type and single-site-mutated azurins. A linear relationship between activation enthalpy and activation entropy was observed. These results are discussed in terms of reorganization energies, driving force and possible electron-transfer pathways.

The binding of imidazole in an azurin-like blue-copper site

Frozen solutions of the azurin mutant His117Gly in the presence of excess of methyl-substituted imidazoles have been investigated by electron spin-echo envelope modulation (ESEEM) spectroscopy at 9 GHz. The addition of imidazole is known to reconstitute a blue-copper site and variation of the non-protein bound ligand [N-methyl-, 2-methyl-, 4(5)-methylimidazole] has allowed the study of the copper-imidazole binding as a model for histidine binding in such sites. Quadrupole and hyperfine tensors of the remote nitrogen of the imidazoles have been determined. The quadrupole tensors indicate that the methyl-substituted imidazoles in the mutant adopt the same orientation relative to copper as the histidine-117 in the wild-type protein. Analysis of the hyperfine tensors in terms of spin densities reveals that the spin density on the remote nitrogen of the substituted imidazole has sigma and a variable pi character, depending on the position of the methyl group. For azurin the corresponding spin density is of virtually pure sigma character. In conclusion, blue-copper sites show subtle variations as regards the histidine/imidazole centred part of the wavefunction of the unpaired electron.