The Active Site Loop Modulates the Reorganization Energy of Blue

Jan 28, 2013 - Department of Life Sciences, University of Modena and Reggio Emilia, ... to a deeper understanding of redox processes in living organis...
0 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/JPCL

The Active Site Loop Modulates the Reorganization Energy of Blue Copper Proteins by Controlling the Dynamic Interplay with Solvent Licia Paltrinieri,† Marco Borsari,† Antonio Ranieri,‡ Gianantonio Battistuzzi,† Stefano Corni,*,¶ and Carlo Augusto Bortolotti*,‡,¶ †

Department of Chemical and Geological Sciences and ‡Department of Life Sciences, University of Modena and Reggio Emilia, via Campi 183, 41125 Modena, Italy ¶ CNR-Nano Institute of Nanoscience, via Campi 213/A, 41125 Modena, Italy S Supporting Information *

ABSTRACT: Understanding the factors governing the rate of electron transfer processes in proteins is crucial not only to a deeper understanding of redox processes in living organisms but also for the design of efficient devices featuring biological molecules. Here, molecular dynamics simulations performed on native azurin and four chimeric cupredoxins allow for the calculation of the reorganization energy and of structure-related quantities that were used to clarify the molecular determinants to the dynamics/function relationship in blue copper proteins. We find that the dynamics of the small, metal-binding loop region controls the outer-sphere reorganization energy not only by determining the exposure of the active site to solvent but also through the modulation of the redox-dependent rearrangement of the whole protein scaffold and of the surrounding water molecules.

SECTION: Biophysical Chemistry and Biomolecules geometrical rearrangement of the first coordination sphere of the active site, while the contributions from the protein matrix and from the solvent are included in the so-called outer-sphere reorganization energy λout. For heme proteins like cytochrome c, which were thouroughly investigated as paradigmatic systems for ET processes, one of the determining factors to the λout was proved to be the accessibility of the active site to the solvent.12,13 A similar result was obtained for azurin, the most studied member of cupredoxins, for which the interface between protein and water in the hydrophobic region surrounding copper-binding His117 was found to be the main contribution to λout.14 Here, we performed classical molecular dynamics (MD) simulations to investigate native azurin and four of its mutants, in which the ligand-containing loop has been replaced either with that of other members of the blue copper family (amicyanin and plastocyanin) or with synthetic sequences featuring only Ala residues. The comparison of five distinct proteins that differ only in the loop composition allowed us to focus exclusively on the contribution of this region on the ET properties. Moreover, through MD simulation, we could take into account the contribution of thermal fluctuations and, in general, of the dynamic relationship between protein and

C

upredoxins are copper-containing proteins involved in electron transfer (ET) processes in both plants and bacteria.1 Their structure features a rigid β-barrel structure, whose C-terminal strands are linked by a loop that shields the copper atom from direct contact with the solvent. The length and sequence of the loop region can vary for different species, apart from the conserved presence of three copper ligands (His, Cys, and Met). Recent studies proved the central role played by the loop in determining the structure and the physiological properties of cupredoxins,2−4 but the molecular details of how it tunes the redox functionality were not yet fully unveiled. The elucidation of the role of the hydrophobic, ligand-containing loop in the ET processes is central not only for a deeper understanding of the structure/function relationships throughout the whole family of cupredoxins but also for tailoring the redox properties of biomolecules to be used in a number of applications such as bioelectronic devices, nanobiosensors, or artificial photosynthetic centers.5−7 Among the most deeply investigated issues in biological ET is the complex relationship that links the rate of the process to the structural features of the biomolecule8−10 and to their dynamic interplay with the solvent. One of the key factors in determining the ET rate is the reorganization energy λ, which represents the free energy required for the reactants to move from the equilibrium reactant structural ensemble to the equilibrium product ensemble, without transferring electrons.11 The reorganization energy can be conveniently divided in two terms; for a metalcontaining protein, the first (λin) is mainly determined by the © 2013 American Chemical Society

Received: December 20, 2012 Accepted: January 28, 2013 Published: January 28, 2013 710

dx.doi.org/10.1021/jz302125k | J. Phys. Chem. Lett. 2013, 4, 710−715

The Journal of Physical Chemistry Letters

Letter

Figure 1. Active site structures of P. aeruginosa azurin (AZU) and of its four copper-binding loop mutants AZAMI, AZPC, 4A3A, and 4A4A. The copper atom is represented as a red sphere, the metal ligands as licorice, and the hydrophobic loop as a gray ribbon. The sequence of the loops for all of the species is also shown. The image was prepared with VMD software22 using the crystallographic structures of the species (PDB ID codes: 4AZU,23 2FTA,4 2HX7,2 3FSW,24 and 3FSZ24).

solvent, which were suggested to be crucial for the redox functionality of a protein.15−21 In particular, we calculated the outer-sphere reorganization energy for all of the investigated proteins, and we monitored changes in the loop region over time and upon reduction. The relatively small but significant changes in λout for the five loop mutants were found to be related to differences in the extent of both short- and long-range solvent reorganization and to structural rearrangement of the whole protein scaffold following reduction/oxidation. We believe that our most important finding is that the nature and length of the relatively small loop region can tune the ET functionality by inducing sizable effects on the dynamics/function relationship of the cupredoxin as a whole. The loop region of native P. aeruginosa azurin (AZU hereafter) and that of four loop mutants is shown in Figure 1. The protein scaffold of AZU was conserved for all of the mutants, but different copper-binding loop sequences were introduced to yield the variants named AZAMI (azurin scaffold and amycianin loop region, CTPHPFM), AZPC (featuring the plastocyanin loop, CSPHQGAGM), and 4A3A and 4A4A mutants, whose metal-binding sequences are CAAAAHAAAM and CAAAAHAAAAM, respectively. The crystal structures of all of these cupredoxins are available,2,4,23,24 and they were taken as starting points to perform 100 ns long MD simulations for both the reduced and oxidized states, from which the reorganization energy for all of the five cupredoxins was obtained. Several methods were suggested in the literature for the calculation of reorganization free energies at different levels of the theory.13,14,25−30 In this Letter, we are comparing proteins where the ligands of the copper atom are conserved; therefore, it is resonable to infer that λin does not undergo dramatic changes between the different loop mutants. Moreover, for native azurin, the contribution of the λin term to the total reorganization free energy was found to be significantly smaller than the λout one.14 We therefore decided to focus only on λout, which was calculated with the energy gap fluctuations method.31 The values of λout are listed in Table 1. Experimental estimates for total reorganization energy are available only for azurin, whose λ has been fixed between 0.6 and 0.8 eV.30,32 Our calculated value for azurin is higher than the experimental ones; this is most likely due to the use of a nonpolarizable force field for the calculations, which is known to introduce some overestimation (which accounts to roughly 30%) of the reorganization energies values.13 Taking this effect into account would lead to a calculated λout value for azurin of approximately 0.73 eV, which would therefore be in line with the aformentioned range, also including the previously estimated value for λin of 0.15 eV.33 Calculated λ closer to

Table 1. Calculated Outer-Sphere Reorganization Energy λout, Solvent-Accessible Surface Area (SASA) of the Active Site Averaged over Both Reduced and Oxidized Ensembles, Changes (Reduced − Oxidized State) in the Total Dipole of the 260 Water Molecules Closest to the Cu Atom, and RootMean-Square Deviation (RMSD) between the Average Structures Obtained from MD in the Oxidized and Reduced Ensemblesa λout (eV) AZU AZAMI AZPC 4A3A 4A4A a

1.05 1.13 1.26 1.18 0.97

(0.03) (0.02) (0.01) (0.03) (0.02)

⟨SASA⟩ active site 0.198 0.565 0.442 0.476 0.261

(0.048) (0.110) (0.084) (0.035) (0.015)

|Δμ⃗water| 12.12 13.10 8.32 19.52 3.26

(0.91) (1.32) (0.01) (0.65) (0.02)

RMSD (ox. vs red.) 0.842 0.962 1.724 1.110 0.456

Estimated errors are given in brackets.

experimental results could have been obtained by choosing more refined theoretical approaches; nevertheless, because our main goal is the comparison of the outer-sphere reorganization free energies for different species, we chose a classical method over a computationally more demanding, quantum mechanical approach. Moreover, as our goal is to investigate the effects of the dynamics of the mutants on their ET kinetic properties, MD allows for a much broader sampling of conformations than QM-based methods presently do. The reliability of our calculations could be assessed by a direct comparison with the values of activation enthalpies (ΔH#) for heterogeneous protein−electrode ET, recently obtained by electrochemical measurements performed on the same set of proteins, immobilized on a gold electrode functionalized with a self-assembled monolayer (SAM).34 Because the binding orientation on the SAM is the same for the five cupredoxins,34 it is reasonable to assume a constant contribution of the adsorption process to the activation enthalpy of the five species with respect to what could be measured for freely diffusing proteins; therefore, the experimental data collected for the surface-immobilized loop mutants can serve as a test for our calculations. Figure 2 displays the plot of calculated λout versus the experimental ΔH# values; a very good correlation can be observed, suggesting that our simulations can be confidently used to calculate observables related to the molecular determinants to the ET kinetics. AZPC displays the highest value for λout (1.26 eV), followed by 4A3A and AZAMI (1.18 and 1.13 eV, respectively). AZU (1.05 eV) and 4A4A (0.97 eV) feature the smallest reorganization energies. The difference displayed by 4A3A and 4A4A is quite surprising as there is just a slight difference in the loop composition between the two variants, with 4A4A 711

dx.doi.org/10.1021/jz302125k | J. Phys. Chem. Lett. 2013, 4, 710−715

The Journal of Physical Chemistry Letters

Letter

predictions of the Marcus theory, according to which structurally similar reduced and oxidized forms would display small reorganization energy values. In order to determine which regions of the proteins provided the most significant contribution to the overall structural rearrangement, we calculated the RMSD between oxidized and reduced forms for each secondary structure element of the cupredoxins. The corresponding data are reported in the Supporting Information (SI). As expected, the β-sheets regions for all of the mutants are only slightly affected by a change in oxidation states, while the loop regions generally feature higher RMSD values. Interestingly, for all five proteins, the loop spanning residues 9−18 and 36−44 and the region encompassing the α-helix and the loop that follows display RMSD values among the highest. Although far from each other in the primary structure, these structural elements carry the residues that form the extended hydrophobic area surrounding the copper atom, together with the loop that was mutated in the present work (see Figure 3). This hydrophobic patch was previously assigned a central role in determining the reorganization energy for native azurin,14 based on the assumption that this portion of the protein would form fewer hydrogen bonds with the surrounding solvent molecules, which would then be greatly affected by changes in the copper oxidation state. The present results allow one to extend this hypothesys to the whole family of cupredoxins because the residues forming the hydrophobic region close to the Cu atom are invariably among the most significant contributors to the overall RMSD, which was proved to be directly correlated to λout. As the regions displaying the highest RMSDs are those surrounding the metal center and because they are mostly composed of hydrophobic residues, the most significant structural rearrangements following reduction must be tightly intertwined to changes in the solvent reorganization. In fact, it was previously demonstrated both through experiments35 and calculations14,30 that the water rearrangement caused by a change in the copper oxidation state is the major determinant of the total reorganization energy for azurin. In order to check whether this counclusion could be extended to the whole set of

Figure 2. Plot of the calculated values for the outer-sphere reorganization energy versus experimental activation enthalpy values obtained by electrochemical means. The experimental values were taken from ref 34.

featuring one additional alanine residue in the hydrophobic loop. Mere comparison of the crystal structures of the two proteins could not provide an explanation for this discrepancy. The reorganization energy is defined as the free energy accompanying structural rearrangements upon reduction/ oxidation; therefore, the reorganization of the protein scaffold is a natural quantity to be analyzed. We extracted the average structures from the MD simulations for both ensembles, and we calculated the root-mean-square deviation (RMSD) between the backbone atoms of the average oxidized structure and the average reduced one. The results are listed in Table 1. Higher values for the RMSD were found for the species displaying the largest λout values; on the contrary, the two loop mutants with the smallest RMSD (0.842 and 0.456 Å for AZU and 4A4A, respectively) are also those with the lowest reorganization energy. The high correlation between λout and the RMSD between the average structures obtained in the two ensembles is plotted in Figure 3. This finding is consistent with the

Figure 3. (A) Plot of the calculated values for the outer-sphere reorganization energy versus RMSD between the average structures of the oxidized and reduced ensembles. (B) Surface (left) and cartoon (right) representations of native azurin, prepared starting from PDB file 4AZU.pdb. (Left) Hydrophobic residues are colored white, polar residues are green, basic residues are blue, and acidic residues are red. (Right) The loops displaying the highest RMSD values for all of the mutants are colored blue (loop spanning residues 9−18), green (36−44), pink (72−79), and yellow (112− 121). 712

dx.doi.org/10.1021/jz302125k | J. Phys. Chem. Lett. 2013, 4, 710−715

The Journal of Physical Chemistry Letters

Letter

number of water molecules in a sphere centered on the copper atom and featuring a 0.9 nm radius. Figure 4 displays the probability that NW takes a particular value along the whole simulation, calculated for each protein and for both the reduced and oxidized ensembles (please note that while Figure 4 depicts the changes in NW between the two oxidation states, the values for the SASA in Table 1 were averaged between the two ensembles). The species featuring the most dramatic rearrangements as a function of the Cu oxidation states is AZPC, followed by AZAMI and 4A3A, as indicated by the shifts of the maxima and average values of the normalized distribution of configurations with a given NW. Changes in NW are smaller for AZU and almost absent for 4A4A, thus strengthening the hypothesis that the hydration shell has dramatic effects on the overall outer-sphere reorganization energy. The NW distribution plots only reflects changes in the number of solvent molecules in close proximity to the active site, but the reorganization energy can be affected by the water molecules within up to roughly 1.6 nm from the metal ion, as previously suggested14 and as further proved by the radial distribution function g(r) of the water oxygen atoms with respect to Cu (see the SI). To further investigate this longrange effect, we decided to follow reduction-induced changes of the dipole moment of the solvent |Δμ⃗ water|, which is directly connected to the energy of the system and therefore is a promising candidate for monitoring the outer-sphere reorganization energy of our proteins. In particular, we calculated the module of the total dipole moment μ⃗water of the 260 water molecules closest to the copper atom, that is, the number of solvent molecules that are comprised within a shell of 1.6 nm centered on Cu, which was previously chosen14 as representative of the solvent region perturbed by the electric field of the metal ion embedded in the protein matrix. We then monitored changes in μ⃗ water, induced by the reduction of Cu(II) to Cu(I), obtaining the module of the difference of the solvent dipole between the two oxidation states, |Δμ⃗ water|. The calculated values are listed in Table 1; large |Δμ⃗ water| values indicate significant reorientation of the solvent molecules upon copper reduction, while for species displaying smaller |Δμ⃗water|, it is reasonable to infer a less extended reorganization of hydration water molecules. Figure 5 shows the plot of the outer-sphere reorganization energy versus the module of the difference in the dipole moment of the hydration water molecules; it is interesting to see how an increase in the differences between the solvent dipole moment going from the oxidized ensemble to the reduced one |Δμ⃗ water| corresponds to larger reorganization energy values. The only species that does not follow this trend is AZPC. Although it is hard to provide an explanation for its behavior, it is possible that for this mutant, λout is dominated by the solvent rearrangements in the proximity of the Cu, which are high enough to account for the biggest reorganization energy value among the loop mutants. The peculiar behavior of AZPC (and the overall trend) was proved to be independent of the number of water molecules taken into account for the calculation of the total dipole moment of the solvent (see SI, Figure S3). These results provide evidence for the correlation, valid for all of the investigated cupredoxins, between changes in the structure of the water molecules within the hydration shell of the active site and the outer-sphere reorganization energy. Most importantly, it is evident how the solvent reorganization is not limited to a reorientation of the total dipole moment but is also

cupredoxins, we investigated the dynamic interplay between the protein matrix and the surrounding water molecules, starting from the analysis of the extent of exposure to water of the protein active site along the whole MD; in particular, we first calculated the solvent-accessible surface area (SASA) of the active site, defined as the Cu atom and the five metalcoordinating residues. The calculated SASA values, averaged over both Cu(I) and Cu(II) oxidation states, are listed in Table 1. The three cupredoxin mutants displaying the highest SASA values, AZAMI, 4A3A, and AZPC, are those for which the biggest λout was also calculated, while the SASAs are significantly lower for the mutants with the smallest reorganization energy values, namely, AZU and 4A4A. Similarly to what was found for cytochrome c, our results show that the solvent accessibility to the protein active site is clearly (and unsurprisingly) a major contribution to λout also for cupredoxins; nevertheless, this structural parameter does not completely justify smaller but meaningful differences in the reorganization energy between AZU and 4A4A or between AZAMI, 4A3A, and AZPC. For example, although the SASA of the active site for 4A4A is bigger than that for the corresponding azurin value, the latter protein displays a significantly higher reorganization energy. In the case of the blue-copper species under investigation, the finer regulation of λout seems to be achieved through the combined contributions of additional factors. We therefore monitored in deeper detail the interaction between the protein and solvent, with a particular focus on the relationship between λout and the reorganization of the solvent upon reduction. We quantified the reduction-driven modification of the hydration of the active site by determining NW in both oxidation states (see Figure 4); NW can be defined as the

Figure 4. Normalized distribution of configurations with a given number of water molecules NW within 0.9 nm from the Cu atom in the reduced (red) and oxidized (green) ensembles, respectively. Dashed lines correspond to average NW values for the two ensembles. 713

dx.doi.org/10.1021/jz302125k | J. Phys. Chem. Lett. 2013, 4, 710−715

The Journal of Physical Chemistry Letters

■

Letter

ASSOCIATED CONTENT

S Supporting Information *

Calculation of the radial distribution function g(r), calculation of the contribution to the overall RMSD for secondary structure elements, plots of λout versus |Δμ⃗ water| as a function of the number of water molecules, and further details on the computational methods. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (S.C.); carloaugusto. [email protected] (C.A.B.). Notes

The authors declare no competing financial interest.

■

Figure 5. Calculated values for the outer-sphere reorganization energy plotted against the module of the difference between the dipole moments of the 260 water molecules closest to the Cu atom calculated in the reduced and oxidized states.

ACKNOWLEDGMENTS This work was supported by CINECA, as part of the ISCRA Grant PSIBET “Investigation of the effects of protein/solvent interaction on biological electron transfer”. This work was also supported by a grant from the Ministero dell’Universitá e della Ricerca of Italy (PRIN 2009 Prot. n. 20098Z4M5E_002). C.A.B. thanks Carabel s.r.l. for financial support. S.C. gratefully acknowledges IIT Platform Computational for funding through the IIT Seed Project MOPROSURF and the MIUR through the FIRB project Italnanonet.

composed of significant variations of the number of solvent molecules within a given shell. In conclusion, the present work unveils the role of the hydrophobic loop region in determining the outer-sphere reorganization energy in cupredoxins. Similarly to what was reported for heme-containing ET proteins, λout turns out to be dependent on the accessibility to the solvent of the active site. Nevertheless, the active site SASA is not the only factor that controls the outer-sphere reorganization energy. Indeed, changes in length and composition of the relatively small hydrophobic loop region result in different dynamics of the whole molecule, leading to reduction-dependent displacements in the protein scaffold, especially in the loop-containing portion of the molecule close to the copper. These structural variations affect the extent of reorganization of the water molecules close to the active site, as well as long-range solvent reorientation. The combined effects of these molecular determinants contribute to the protein/solvent interaction, leading to a fine regulation of λout.

■

REFERENCES

(1) Dennison, C. Investigating the Structure and Function of Cupredoxins. Coord. Chem. Rev. 2005, 249, 3025−3054. (2) Li, C.; Banfield, M. J.; Dennison, C. Engineering Copper Sites in Proteins: Loops Confer Native Structures and Properties to Chimeric Cupredoxins. J. Am. Chem. Soc. 2007, 129, 709−718. (3) Battistuzzi, G.; Borsari, M.; Dennison, C.; Li, C.; Ranieri, A.; Sola, M.; Yanagisawa, S. Active Site Loop Dictates the Thermodynamics of Reduction and Ligand Protonation in Cupredoxins. Biochim. Biophys. Acta 2009, 1794, 995−1000. (4) Li, C.; Yanagisawa, S.; Martins, B. M.; Messerschmidt, A.; Banfield, M. J.; Dennison, C. Basic Requirements for a Metal-Binding Site in a Protein: The Influence of Loop Shortening on the Cupredoxin Azurin. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 7258− 7263. (5) Marshall, N. M.; Garner, D. K.; Wilson, T. D.; Gao, Y.-G.; Robinson, H.; Nilges, M. J.; Lu, Y. Rationally Tuning the Reduction Potential of a Single Cupredoxin Beyond the Natural Range. Nature 2009, 462, 113−116. (6) Li, W.; Sepunaru, L.; Amdursky, N.; Cohen, S. R.; Pecht, I.; Sheves, M.; Cahen, D. Temperature and Force Dependence of Nanoscale Electron Transport via the Cu Protein Azurin. ACS Nano 2012, 6, 10816−10824. (7) Ron, I.; Sepunaru, L.; Itzhakov, S.; Belenkova, T.; Friedman, N.; Pecht, I.; Sheves, M.; Cahen, D. Proteins as Electronic Materials: Electron Transport through Solid-State Protein Monolayer Junctions. J. Am. Chem. Soc. 2010, 132, 4131−4140. (8) Page, C.; Moser, C.; Chen, X.; Dutton, L. Natural Engineering Principles of Electron Tunnelling in Biological Oxidation−Reduction. Nature 1999, 402, 47−52. (9) Gray, H. B.; Winkler, J. R. Electron Tunneling Through Proteins. Q. Rev. Biophys. 2003, 36, 341−372. (10) Bortolotti, C. A.; Borsari, M.; Sola, M.; Chertkova, R.; Dolgikh, D.; Kotlyar, A.; Facci, P. Orientation-Dependent Kinetics of Heterogeneous Electron Transfer for Cytochrome c Immobilized on Gold: Electrochemical Determination and Theoretical Prediction. J. Phys. Chem. C 2007, 111, 12100−12105.

■

COMPUTATIONAL METHODS The MD simulations of azurin and its loop mutants were performed in water using the GROMACS software package36 and the AMBER99 force field.37 As the starting point of the simulations, oxidized structures of azurin and loop mutants AZAMI, AZPC, 4A3A, and 4A4A (PDB id codes: 4AZU, 2FTA, 2HX7, 3FSW, and 3FSZ) were taken as starting conformations. Water was represented with the TIP3P model.38 For each starting structure, one protein molecule was solvated in a periodic rhombic dodecahedral box. The simulations were performed in the NVT ensemble at the experimental temperature of 300 K. The temperature was kept constant by the isokinetic temperature coupling.39 MD simulations were performed for both reduced and oxidized forms of all of the variants; the two oxidation states were obtained in our simulations by changing the atomic charges of the protein active site (i.e., the Cu metal atom and the atoms belonging to the five copper-coordinating residues). The atomic charges calculated at the DFT level of the theory were taken from the literature.40 No other changes at the force field have been made. Further details are given in the SI. 714

dx.doi.org/10.1021/jz302125k | J. Phys. Chem. Lett. 2013, 4, 710−715

The Journal of Physical Chemistry Letters

Letter

(11) Bendall, D. Protein Electron Transfer; BIOS Scientific Publisher Ltd: Oxford, U.K., 1996. (12) Bortolotti, C. A.; Siwko, M. E.; Castellini, E.; Ranieri, A.; Sola, M.; Corni, S. The Reorganization Energy in Cytochrome c is Controlled by the Accessibility of the Heme to the Solvent. J. Phys. Chem. Lett. 2011, 2, 1761−1765. (13) Tipmanee, V.; Oberhofer, H.; Park, M.; Kim, K.; Blumberger, J. Prediction of Reorganization Free Energies for Biological Electron Transfer: A Comparative Study of Ru-Modified Cytochromes and a 4Helix Bundle Protein. J. Am. Chem. Soc. 2010, 132, 17032−17040. (14) Cascella, M.; Magistrato, A.; Tavernelli, I.; Carloni, P.; Rothlisberger, U. Role of Protein Frame and Solvent for the Redox Properties of Azurin from Pseudomonas aeruginosa. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 19641−19646. (15) Bortolotti, C. A.; Amadei, A.; Aschi, M.; Borsari, M.; Corni, S.; Sola, M.; Daidone, I. The Reversible Opening of Water Channels in Cytochrome c Modulates the Heme Iron Reduction Potential. J. Am. Chem. Soc. 2012, 134, 13670−13678. (16) Stock, A. Relating Dynamics to Function. Nature 1999, 400, 221−222. (17) Sagle, L. B.; Zimmermann, J.; Matsuda, S.; Dawson, P. E.; Romesberg, F. E. Redox-Coupled Dynamics and Folding in Cytochrome c. J. Am. Chem. Soc. 2006, 128, 7909−7915. (18) Henzler-Wildman, K.; Kern, D. Dynamic Personalities of Proteins. Nature 2007, 450, 964−972. (19) Skourtis, S.; Waldeck, D.; Beratan, D. Fluctuations in Biological and Bioinspired Electron-Transfer Reactions. Annu. Rev. Phys. Chem. 2010, 61, 461−485. (20) Zimmermann, J.; Thielges, M. C.; Yu, W.; Dawson, P. E.; Romesberg, F. E. Carbon−Deuterium Bonds as Site-Specific and Nonperturbative Probes for Time-Resolved Studies of Protein Dynamics and Folding. J. Phys. Chem. Lett. 2011, 2, 412−416. (21) Ly, H. K.; Marti, M. A.; Martin, D. F.; Alvarez-Paggi, D.; Meister, W.; Kranich, A.; Weidinger, I. M.; Hildebrandt, P.; Murgida, D. H. Thermal Fluctuations Determine the Electron-Transfer Rates of Cytochrome c in Electrostatic and Covalent Complexes. ChemPhysChem 2010, 11, 1225−1235. (22) Humphrey, W.; Dalke, A.; Schulten, K. VMD  Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33−38. (23) Nar, H.; Messerschmidt, A.; Huber, R.; van de Kamp, M.; Canters, G. Crystal Structure Analysis of Oxidized Pseudomonas aeruginosa Azurin at pH 5.5 and pH 9.0. A pH-Induced Conformational Transition Involves a Peptide Bond Flip. J. Mol. Biol. 1991, 221, 765−772. (24) Sato, K.; Li, C.; Salard, I.; Thompson, A. J.; Banfield, M. J.; Dennison, C. Metal-Binding Loop Length and Not Sequence Dictates Structure. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 5616−5621. (25) Heck, A.; Woiczikowski, P. B.; Kubar, T.; Giese, B.; Elstner, M.; Steinbrecher, T. B. Charge Transfer in Model Peptides: Obtaining Marcus Parameters From Molecular Simulation. J. Phys. Chem. B 2012, 116, 2284−2293. (26) Blumberger, J. Free Energies for Biological Electron Transfer from QM/MM Calculation: Method, Application and Critical Assessment. Phys. Chem. Chem. Phys. 2008, 10, 5651−67. (27) Simonson, T. Gaussian Fluctuations and Linear Response in an Electron Transfer Protein. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 6544− 6549. (28) Muegge, I.; Qi, P. X.; Wand, A. J.; Chu, Z. T.; Warshel, A. The Reorganization Energy of Cytochrome c Revisited. J. Phys. Chem. B 1997, 101, 825−836. (29) Blumberger, J.; Klein, M. L. Reorganization Free Energies for Long-Range Electron Transfer in a Porphyrin-Binding Four-Helix Bundle Protein. J. Am. Chem. Soc. 2006, 128, 13854−13867. (30) Corni, S. The Reorganization Energy of Azurin in Bulk Solution and in the Electrochemical Scanning Tunneling Microscopy Setup. J. Phys. Chem. B 2005, 109, 3423−3430. (31) Simonson, T. Electrostatics and Dynamics of Proteins. Rep. Prog. Phys. 2003, 66, 737−787.

(32) Gray, H. B.; Malmströ m, B. G.; Williams, R. Copper Coordination in Blue Proteins. J. Biol. Inorg. Chem. 2000, 5, 551−559. (33) Ryde, U. L. F.; Olsson, M. H. M. Energy of Blue Copper Models in the Protein. Int. J. Quantum Chem. 2000, 81, 335−347. (34) Monari, S.; Battistuzzi, G.; Bortolotti, C. A.; Yanagisawa, S.; Sato, K.; Li, C.; Salard, I.; Kostrz, D.; Borsari, M.; Ranieri, A.; et al. Understanding the Mechanism of Short-Range Electron Transfer Using an Immobilized Cupredoxin. J. Am. Chem. Soc. 2012, 134, 11848−11851. (35) Crane, B.; Di Bilio, A.; Winkler, J.; Gray, H. Electron Tunneling in Single Crystal of Pseudomonas aeruginosa Azurins. J. Am. Chem. Soc. 2001, 123, 11623−11631. (36) Berendsen, H.; van der Spoel, D.; van Drunen, R. GROMACS: A Message-Passing Parallel Molecular Dynamics Implementation. Comput. Phys. Commun. 1995, 91, 43−56. (37) Wang, J.; Cieplak, P.; Kollman, P. A. How Well Does a Restrained Electrostatic Potential (RESP) Model Perform in Calculating Conformational Energies of Organic and Biological Molecules? J. Comput. Chem. 2000, 21, 1049−1074. (38) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of Simple Potential Functions for Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926−936. (39) Brown, D.; Clarke, J. H. R. A Comparison of Constant Energy, Constant Temperature, and Constant Pressure Ensembles in Molecular Dynamics Simulations of Atomic Liquids. Mol. Phys. 1984, 51, 1243−1252. (40) van den Bosch, M.; Swart, M.; Snijders, J. G.; Berendsen, H. J. C.; Mark, A. E.; Oostenbrink, C.; van Gunsteren, W. F.; Canters, G. W. Calculation of the Redox Potential of the Protein Azurin and Some Mutants. ChemBioChem 2005, 6, 738−746.

715

dx.doi.org/10.1021/jz302125k | J. Phys. Chem. Lett. 2013, 4, 710−715