Intramolecular Electron Transfer in Azurin
ContentsIntroductionAzurin is a bacterial protein that has been extensively studied by bioinorganic and biophysical chemists as a prototype of a Type 1 or "blue" copper protein. It contains a single copper ion that can be in the Cu+ or Cu2+ Cu state. The intensely blue color is due to a charge transfer transition from the cysteine thiolate ligand to the Cu in the Cu2+ state. It functions as an electron transfer (ET) mediator. The ET reactivity of azurin has been extensively studied, including studies of its reactivity with natural and artificial partners, and intramolecular ET from intrinsic and covalently attached electron transfer partners. The latter studies have been instrumental in defining and evaluating the factors influencing electron transfer reactivity through proteins. These factors include the electron transfer distance, the structure of the intervening peptide medium, the thermodynamic driving force, and the structure of the donor and acceptor. These studies have been instrumental in the iterative testing and advancing of electron transfer theory. One series of studies, delineated here, involves measurement of the rate constant for ET from a disulfide radical, produced by pulse radiolysis, to the Cu2+ ion [1]. This reaction can be made to occur because of particular structural features of azurin, the Cu2+ site is relatively buried and at the opposite end of the protein from the only disulfide, which is exposed to solvent and electron transfer reagents. StructureAzurin has 128 amino acids and a beta barrel structure, as exemplified by the structure of the Pseudomonas aeruginosa protein (4AZUA, chain A). The backbone is shown colored by the secondary structure assignment (pink for alpha helix, purple for a 310 helix, yellow for beta strands, blue for beta turns, and white for other structures). The copper ion (maroon) is at the top of this standard orientation. The primary ligands to the copper atom are the thiolate of Cys 112, and Nδ of His 117 and His 46, forming a trigonal planar site with 3 ligands. In addition there are two weaker ligands, the S of Met 121 and the carbonyl O of Gly 45, occupying axial positions to give an approximately trigonal bipyramidal site with 5 ligands. In addition to the Cu site, there is single disulfide at the at the "bottom" of the protein Exposure of the disulfide and the copper siteThe studies discussed here involve the intramolecular ET between a one electron reduced disulfide anion radical and the oxidized copper ion. In order to measure the rate constant for this process, the disulfide radical must be produced rapidly by a strong reductant in a bimolecular reaction. This bimolecular reaction must reduce the disulfide preferentially over the Cu2+ site. Azurin shows this preferential reactivity due to the lack of exposure of the copper site, with only part of the edge of the His 117exposed coupled with high exposure of the disulfide Cys 3-Cys 26. The reducing agent typically used is the CO2- anion radical, an especially strong reducing agent produced by pulse radiolysis of formate containing solutions. Thus the electrostatic interaction with the sites is also relevant. The copper site has no charges near the exposed His46 but the disulfide site disulfide site has both a positive residue (Lys 27) and a negative one (Glu 2) nearby. Electron transfer pathElectron transfer is slowed by increasing distance between the donor and acceptor, but this can be partially ameliorated by an appropriately constructed pathway of covalent bonds. Thus the distance between the electron donor and acceptor can be analyzed in terms of pathways involving covalent bonds, hydrogen bonds (less effective), and through-space jumps (even less effective)[2]. One pathway between the disulfide and copper involves a covalent path from Sγ of Cys 3, through the backbone Ser 4, Val 5, Asp 6, Ile 7, Gln 8, Gly 9 and Asn 10, then through a hydrogen bond from the O of Asn 10 to the proton on NE2 (Nε) of the ligand His 46. The second path branches through a hydrogen bond from the carbonyl O of Cys 3 to the peptide N of Thr 30, then through the backbone of Val 31 and then via a through-space jump from the side chain Cγ of Val 31 to Cγ of the side chain of Trp 48, then through the side chain and backbone of Trp 48 and Val 49, followed by a hydrogen bond from the backbone N of Val 49 to to carbonyl O of Phe 111 and then to the Cu via the ligand Cys 112. The orbital coupling provided by this path is sensitive to the distance of the through-space jump, and thus is influenced by the mobility of the structure but enhanced by the especially effective orbital overlap of the Cu-S bond compared to the Cu-N bond of the first path. Rate constants and activation parametersThe ET from the disulfide anion radical to the Cu in the native protein takes place, at pH 7, with a rate constant of 44 s -1 an enthalpy of activation (ΔH‡) of 47.5 kJ/mol and an entropy of activation (ΔS‡) of -56.5 J/mol K. An early question about the effect of the intervening residues on the ET reactivity concerned the single tryptophan residue in the core of the protein, with the concept that ET through delocalized π symmetry orbitals facilitates ET. Even replacing Trp 48 by a variety of nonpolar residues had little effect, but addition of a second tryptophan in place of Val 32(red) led to a significant increase in the ET rate constant to 285 s -1 with ΔH‡ of 47.2 kJ/mol and ΔS‡) of -39.7 J/mol K. There was essentially no change in the driving force for the ET reaction nor in the structure of the two ET partners, and the change in the entropy to a more favorable value is consistent with an improvement in the pathway. Another study sought to investigate the effect of changing the driving force without significantly changing the reorganization energy, the energy of the structural change coupled to ET. Yi Lu and coworkers developed a series of mutants that primarily involved the hydrogen bonding network around the Cu center[3]. One of these is N47S/M121L in which two residues near the Cu are mutated including the weakly interacting methionine. The native structure shows ASN 47 in this view. The structure of the mutant with Ser 47 at 5 o'clock and Leu 121 behind the Cu. This series of mutants varied the driving force by ~0.5 V and showed inverted behavior, as predicted by Marcus theory. That is, the rate constant increased with driving force to a maximum and then decreased at higher driving force. This analysis required that the reorganization energy of the mutants stay approximately constant, which was apparently the case. The most extreme modification of the Cu binding site of azurin, again by Yi Lu and coworkers, was to create the di-copper CuA site of Pseudomonas denitrificans cytochrome oxidase in azurin. In this transformation the Cu binding loop CSELCGINHAL replaced CTF_P_GHSAL of Pseudomonas aeruginosa azurin. The CuA site is also the site of reduction of cytochrome oxidase. The reactivity of this site, for single electron reduction, was established by the same method as used for the other azurin mutants [4]. Depending on pH, the site can have its single unpaired electron delocalized (mixed valence, pH 4) or localized on a single Cu (trapped valence, pH 8). The site is about a factor of 3 more reactive than the type 1 azurin site, when compared at pH 5, even though the driving force for the reaction is somewhat less. The estimate for the reorganization energy of the CuA site is slightly lower than that of a type 1 center.
| ||||||||||||
