Role of Conserved Residues - American Chemical Society

J. Phys. Chem. B 2009, 113, 1919–1932

1919

Tuning of Copper-Loop Flexibility in Bacillus subtilis CopZ Copper Chaperone: Role of Conserved Residues Agustina Rodriguez-Granillo† and Pernilla Wittung-Stafshede*,†,‡ Department of Biochemistry and Cell Biology, Rice UniVersity, Houston, Texas 77251, and Department of Chemistry, Umeå UniVersity, 901 87 Umeå, Sweden ReceiVed: August 26, 2008; ReVised Manuscript ReceiVed: December 05, 2008

Bacillus subtilis CopZ is a copper (Cu) chaperone that binds and delivers Cu to intracellular targets to maintain cellular Cu homeostasis. Like Cu chaperones from other organisms, including the human homologue Atox1, CopZ has the ferredoxin-like fold and binds Cu(I) via two Cys in a conserved M11X12C13X14X15C16 motif located in a solvent-exposed loop. Here, we have performed extensive molecular dynamics simulations on strategic CopZ variants to reveal structural and dynamic roles of three residues near and in the Cu loop (i.e., Met11, Ser12, and Tyr65). Met11 is conserved in all Cu chaperones, whereas Ser12 and Tyr65 are exchanged for Thr and Lys in eukaryotes like Atox1. Therefore, our simulations included apo and holo forms of Met11Ala, Ser12Ala, and Tyr65Ala, as well as Ser12Thr and Tyr65Lys, CopZ variants. We have discovered that the conserved Met is solvent exposed and important for optimal Cu-loop flexibility in the apo form of CopZ but is buried in the core and aids in packing of the fold in holo-CopZ. Ser12 and Tyr65 are important for assuring Cu-loop flexibility in the apo form; in the Cu-bound form, these residues participate in stabilizing electrostatic networks. The two eukaryotic residues tested are not good substitutes for the prokaryotic counterparts in CopZ. By comparisons to data for Atox1, we conclude that common residues (like Met) and unique residues (like Ser12 and Tyr65 in CopZ) have evolved differentially in prokaryotic and eukaryotic Cu chaperones to tune the flexibility of the Cu loop of the apo form and to provide electrostatic Cu-site stabilization of the holo form. Introduction Copper (Cu) is an essential trace metal that plays a central role in cellular metabolism.1 It serves as a cofactor in a number of enzymes involved in many important celullar processes, such as ceruloplasmin, cythochrome c oxidase, and Cu/Zn superoxidase dismutase, among others.2-5 At the same time, free Cu is toxic to the cell, as it can catalyze redox reactions resulting in free radical species that can cause oxidative damage to biological macromolecules. Therefore, organisms have learned how to restrict the amount of free Cu to less than one atom per cell.6 Cellular Cu homeostasis is maintained by a family of proteins called Cu chaperones, which are conserved from unicellular organisms to mammals.7 These small, soluble proteins, bind, guide, and protect the Cu ions within the cell, delivering them to the appropriate functional targets.5,8-11 In humans, the Cu(I) chaperone Atox1 delivers Cu to metal-binding domains (MBDs) of to two P1B-type ATPases in the secretory pathway: the Menkes (or ATP7A) and the Wilson (or ATP7B) disease proteins.5,9,12 In the Gram positive bacterium Bacillus subtilis, CopZ delivers Cu(I) to CopA, the corresponding P1B-type ATPase in this organism.13-15 CopZ is a 69-residue protein that, like Atox1 and the MBDs of P-type ATPases, has a ferredoxinlike fold with a βRββRβ arrangement and a conserved metalbinding motif MX1CX2X3C, located in the solvent-exposed β1-R1 loop, which binds a single Cu(I)16,17 (Figure 1). Structural work has demonstrated that Cu chaperones and target MBDs from various organisms possess the same fold and * To whom correspondence should be addressed. E-mail: pernilla.wittung@ chem.umu.se. † Rice University. ‡ Umeå University.

coordinate Cu(I) via the two Cys in the MX1CX2X3C motif with a proposed S-Cu-S linear coordination geometry.18-23 In some cases, such as for Bacillus subtilis CopZ16 and the yeast Atx1,24 the S-Cu-S angle significantly deviates from linear, suggesting three- rather than two-ligand coordination for the Cu. The third ligand was proposed to be an exogenous thiol, such as DTT present in the buffer.16,24 However, we recently showed that CopZ coordinates Cu with a distorted linear coordination that deviates by ∼26° from linear even in the absence of a third ligand, suggesting that the protein structure imposes the nonlinear Cu coordination (regardless of the presence of a third ligand or not).25 The distorted linear coordination may favor Cu(I)-mediated CopZ dimerization, which has been observed in solution.13,26 We previously showed that the conserved Met at the beginning of the Cu-binding motif is not directly involved in metal ligation.21,25 This is in accord with the suggestion that this residue acts as a tether that modulates Cu-binding loop structure.21,27 There are few important sequence differences between the eukaryotic and bacterial Cu chaperones: the amino acids at position 65 of CopZ (60 in Atox1) and X1 in the metal-binding motif. Although distant in sequence, residue 65 is situated in close proximity to the metal binding site in the tertiary structure (Figure 1). In eukaryotic Cu chaperones, including Atox1, the residue that aligns with residue 65 in CopZ is an invariant Lys, which is proposed to neutralize the overall negative charge of the Cu-thiolate center in the holo form.21,28 However, in prokaryotes, including CopZ, residue 65 is always a Tyr.21 On the other hand, residue X1 is a Ser in prokaryotic Cu chaperones, including CopZ (Figure 1), but a Thr in eukaryotic Cu chaperones.21 Also, we recently showed that, whereas Met10

10.1021/jp807594q CCC: $40.75  2009 American Chemical Society Published on Web 01/26/2009

1920 J. Phys. Chem. B, Vol. 113, No. 7, 2009

Rodriguez-Granillo and Wittung-Stafshede

Figure 1. Left: Solution structure of the holo form of WT CopZ (1K0V.pdb) revealing its ferredoxin-like fold. The Cys (sticks) and Cu (blue) are indicated. Right: Blow-up of the Cu-binding loop area: Met11, Ser12, Cys13, Cys16, Tyr65, and Cu are labeled.

is always buried in Atox1 regardless of Cu, it is exposed to solvent in apo-CopZ but becomes hidden in the core upon Cu binding.25 Recently, we reported an in vitro unfolding study of apo and holo forms of wild-type (WT) Atox1 and CopZ.29 We found that, although Cu stabilizes the folded state toward chemical and thermal perturbations in both cases, the human protein was more stable than the bacterial one.29 We further performed classical molecular dynamics (MD) simulations on WT Atox1 and CopZ with and without Cu,25 in order to gain insight into the differences found in vitro. Furthermore, we recently extended our computational work to address the roles of key residues in Atox1 (i.e., Met10, Thr11 and Lys60) in protein structure, stability, and dynamics.42 For this, Met10Ala, Thr11Ala, and Lys60Ala variants were generated computationally and then subjected to MD simulations in apo and Cu(I)-bound forms. We also included Thr11Ser and Lys60Tyr variants in our study to assess how residues conserved in prokaryotic species would behave in the corresponding positions in Atox1. Here, we have performed corresponding MD simulations on CopZ, by simulating Met11Ala, Ser12Ala, Ser12Thr, Tyr65Ala, and Tyr65Lys in apo and holo forms. The aim of this work was not only to determine the roles of three key residues in CopZ structural dynamics but also to obtain insights into the reasons of why Met is completely conserved but Ser and Tyr are only preserved in bacteria. We find that the conserved Met is essential for a proper Cu-loop flexibility in the apo form and for proper packing of the fold in the holo form, similarly to what we found for Atox1. On the other hand, Ser12 and Tyr65 are important for a sufficient Cu-loop flexibility in the apo form and these residues participate in an electrostatic network with the Cu-coordinating Cys in the holo form. Finally, insertions of eukaryotic residues in CopZ are not good replacements for the bacterial residues. Thus, it appears that specific residues have evolved differentially in Cu chaperones of prokaryotes and eukaryotes to tune the flexibility of the Cu loop. Computational Methods The best method to explore the potential energy surface of large complex systems (such as proteins) is classical MD simulations. This method generates a trajectory of the system, from which time averages of properties can be obtained.30 This approach allows structural, thermodynamic and dynamic infor-

mation to be obtained that may not be accessible by means of experimental techniques. The starting point for the simulations is generally a high-resolution structure obtained by X-ray or NMR. Because in a classical MD simulation atoms are described with models, the root-mean-square deviation (rmsd) with respect to the initial structure first increases until it reaches equilibrium. The total simulation is therefore divided into the equilibration phase and the production phase, the latter being used for data analysis. In the case of CopZ, due to its high flexibility, we have previously reported that its equilibration phase is ∼60 ns.25 Here, because high-resolution structures of the CopZ mutants are not available, the starting points for the simulations in these cases are structures that we generated in silico using the reported WT apo and holo solution structures of CopZ as templates. Therefore, and plus the fact that the introduced mutations may have structural and/or dynamic effects on the CopZ structure, we predicted that the mutant systems would require longer equilibration phases. Protein structures were obtained from the PDB: 1P8G17 and 1K0V16 for apo- and holo-CopZ, respectively. Both proteins contain four additional non-native residues at the C-terminus as a result of how the proteins were cleaved during purification;16,17 however, this probably has little effect on the fold and biophysical parameters. Although there is experimental evidence suggesting that CopZ may bind Cu(I) in a three-coordinated fashion, via an exogenous nonprotein ligand present in the media16 or by the result of protein dimerization,13,26 the nature of this third ligand has not yet been identified. Therefore, the effect of the presence of a third Cu(I) ligand originating from an exogenous donor or as a result of protein dimerization was not taken into account. Also, the lack of a CopZ high-resolution structure in which Cu(I) is coordinated via more than two ligands makes it difficult to address these issues computationally. Most importantly, to compare the mutant results with our previously reported WT results,25 it is crucial to maintain the parametrization that we originally used, especially for the Cu center in the holo protein. Therefore, we decided to analyze the twocoordinated Cu(I) center. The following mutations were generated in CopZ in silico with the LEaP module of Amber9:31-34 Met11Ala, Ser12Ala, Ser12Thr, Tyr65Ala, and Tyr65Lys. Classical MD simulations were performed for the different protein variants using Amber9.31-34 The initial structures were immersed in a pre-

Role of Conserved Residues in CopZ Dynamics equilibrated truncated octahedral cell of TIP3P explicit water molecules35 and Na+ ions were added to neutralize the systems.31-34 Protein atoms were described with the parm99SB force field parametrization.36 Parameters of the coordinating Cys and Cu in the holo forms were taken from ref 25. Briefly, the geometry of holo-CopZ was optimized at the QM(PBE/DZP)MM(Amber99) level (Cu plus two cysteinates) followed by HF/ 6-31G(d)/RESP charge parametrization as described in Amber standard protocol. Water molecules extended at least 9 Å from the surface of the proteins. Simulations were performed in the NPT ensemble (constant pressure of 1 atm and temperature of 300 K was maintained using the Berendsen coupling scheme37), employing periodic boundary conditions. A SHAKE algorithm was employed to keep bonds involving hydrogen atoms at their equilibrium length.38 The optimized systems were heated to 300 K and equilibrated for 200 ps. The structures were simulated until the backbone rmsd as a function of time was stable in the last 20 ns. These converged last 20 ns were used for data analysis. Rmsd, rms fluctuations (rmsf) per residue, and protein/ solvent radial distribution functions g(r) were calculated for each of the systems using the ptraj module of Amber9.31-34 Results and Discussion Analysis of WT CopZ. In our first MD work of Cu chaperones, we found that CopZ, in both apo and holo forms, has increased conformational flexibility compared to the human counterpart Atox1.25 Therefore, instead of equilibrating in ∼20 ns like Atox1, CopZ was simulated ∼100 ns to allow proper equilibration, and the last 20 ns were analyzed. In the first study, we focused our analysis in residues Met11, Cys13, and Cys16, which are part of the Cu binding motif MX1CX2X3C. Here, we extend the WT analysis of CopZ to two residues that have been proposed to also be important for Cu binding: residue X1 (Ser12) and Tyr65 in CopZ (Figure 1). Ser12, located in the Cu loop, is only conserved as such in prokaryotic Cu chaperones, including Bacillus subtilis CopZ.21 This position is filled with a Thr in eukaryotic Cu chaperones (including Atox1) and MBDs (except in MBD3 that is a His) and in prokaryotic MBDs.21 In eukaryotic Cu chaperones, this residue was proposed to hydrogen bond (HB) with the first Cu coordinating Cys of the partner MBD during Cu transfer.28 However, its role in prokaryotic Cu chaperones has not been explored. In our MD simulations, residue flexibility is represented by backbone rmsf, and solvation by the radial distribution function g(r) between the residue (a certain side chain heavy atom) and water (O). In the absence of Cu, Ser12 is very floppy25 and solvent exposed and does not maintain any stable interaction with Tyr65 or Cys13 (Figure 2A-C). Instead, it points toward the β2-β3 loop, and interacts with Met11 and Glu38. On the other hand, in the holo form, Ser12 is still solvent exposed as in the apo form (Figure 2A), but is much more restricted25 and no longer interacts with the β2-β3 loop or Met11, which is completely buried. In this case, it points toward the Cu site and interacts with Tyr65 and Cys13 (Figure 2B, C). Therefore, it appears that Ser12 helps to stabilize the Cu-bound state by facilitating interactions near the Cu site. Far away in sequence but spatially close to the metal binding motif, Tyr65 in CopZ, located in the R2-β4 loop, is conserved in prokaryotic Cu chaperones and MBDs.21 In eukaryotes, this position corresponds to Lys in Cu chaperones (including Atox1) or Phe in MBDs (except MBD3 that has Pro).21 The role of a Tyr here is not clear, as this residue lacks the positive charge of Lys, proposed to contribute to the stabilization of the -1 net charge of the Cu bis-thiolate center in holo-Atox1.21,28 As

J. Phys. Chem. B, Vol. 113, No. 7, 2009 1921 Tyr65 is located in a surface-exposed loop, it is always solvent exposed (Figure 2D). In apo-CopZ, Tyr65 is flexible,25 far away from the functional Cys (Figure 2E), and interacts with Glu9 and Gln63 located in the β1-R1 and R2-β4 loops, respectively. Similarly to Lys60 in holo-Atox1,42 in holo-CopZ, Tyr65 points to the Cu site and interacts extensively with Ser12 and Cys16 (Figure 2C,E). Therefore, it appears that Tyr65 undergoes a conformational change upon Cu binding, which stabilizes the Cu-bound state. Met11Ala CopZ. Although Met11 in CopZ is conserved in all Cu chaperones and MBDs, the behavior of this residue differs between Atox1 and CopZ. By NMR, Met11 was found to be solvent exposed,16,17 as opposed to Met10 in Atox1, which is buried.39 However, we previously identified a significant change of the dynamics and solvent exposure of Met11 depending on the presence of Cu.25 In apo-CopZ, Met11 is entirely solvent exposed and floppy, whereas it is hidden in the protein core in the holo form, similarly to what was found for Atox1. It appears that Met11 serves as a Cu-dependent switch in CopZ, and its role as a hydrophobic anchor as appears to be the case in Atox1 is less obvious. We therefore decided to mutate this residue to Ala in CopZ. apo- and holo-Met11Ala CopZ were simulated for ∼118 and ∼120 ns with mean rmsd (with respect to the first structure) and standard deviation of 2.4 ( 0.2 and 3.1 ( 0.6 Å, respectively (Figure 3). The holo form of Met11Ala undergoes significant conformational changes and has increased flexibility as compared to apo-Met11Ala. In fact, after ∼80 ns, holoMet11Ala undergoes a conformational change which is not present in the apo form. This is demonstrated by a “jump” in the rmsd time evolution and is consistent with the fact that Met11 is exposed in the apo form, not making any stabilizing interactions whereas in the holo form it significantly stabilizes the core. Removal of this residue is thus likely more tolerated in the absence of Cu. During the last 20 ns of simulation, which were used for data analysis, however, both forms are stable with mean rmsd (with respect to the average structure of the last 20 ns) of 0.7 ( 0.1 and 0.7 ( 0.2 Å for apo- and holo-Met11Ala, suggesting no detectable conformational changes. Whereas the apo form of Met11Ala has significant more restricted backbone than apo-WT, the mutant is more floppy than WT in the holo form (Figure 4A,B). The net effect is that instead of becoming more structured, as in WT, Met11Ala gains flexibility upon Cu binding. The most affected regions are residues 10 to 40, which include the Cu loop, helix R1, strand β2, and connecting loops (Figure 4A,B). Interestingly, even though Met11 is solvent exposed in the apo form of WT, it still has a significant effect on the backbone dynamics of the whole protein. An overlay of the equilibrated apo structures of WT and Met11Ala revealed an rmsd of 3.4 Å and significant structural differences (Figure 5A). It appears that solvent exposure of the bulky hydrophobic Met11 side chain in apo-CopZ is important to ensure flexibility of the Cu loop. Replacement of Met11 by Ala rigidifies the protein because the entropic effect is absent. In apo-WT, Met11 and Ser12, although highly mobile, interact with Ala39 and Glu38, located in the β2-β3 loop. This interaction assures that apo-CopZ is flexible enough to bind Cu but that the fold is not too destabilized. On the other hand, in Met11Ala mutant, presence of the small Ala11 results in lack of an interaction with the β2-β3 loop, which disrupts the interface between these loops. This loop-loop interaction helps maintain the ferredoxin-like fold in WT. In its absence, the Cu loop in Met11Ala forms a turn that folds



Figure 2. (A, D) Protein-solvent radial distribution functions g(r) of Ser12 (A) and Tyr65 (D) of WT CopZ with (red) and without Cu (blue). (B, C, E) Histograms of distance distribution (in Å) between heavy atoms of Ser12 (O) and Cys13 (S) (B), Ser12 (O) and Tyr65 (O) (C), and between Tyr65 (O) and Cys16 (S) (E) of WT CopZ with (red) and without Cu (blue). Analysis based on earlier simulation data.25

Figure 3. Rmsd (in Å, with respect to the first structure) of the backbone heavy atoms (N, CR, C) as a function of simulation time for the different variants in apo (A) and holo (B) forms. Blue, Met11Ala; green, Ser12Ala; orange, Ser12Thr; red, Tyr65Ala; cyan, Tyr65Lys.

Role of Conserved Residues in CopZ Dynamics

J. Phys. Chem. B, Vol. 113, No. 7, 2009 1923

Figure 4. Average fluctuations (rmsf in Å) of backbone heavy atoms (N, CR, C) per residue for WT and the different variants Met11Ala (blue), Ser12Ala (green) and Ser12Thr (orange), and Tyr65Ala (red) and Tyr65Lys (cyan) in apo (A, C, E) and holo (B, D, F) forms. The secondary structure elements are indicated (R1, R2; β1-β4).

into the protein, markedly restricting the motion of the Cucoordinating Cys and completely burying Cys13 (Figure 5A). holo-Met11Ala exhibits important structural changes comparing to holo-WT. In the mutant, strand β1 and helix R1 are shortened, so the Cu loop extends from residue 8 to 15 versus residue 10 to 13 in WT. This change increases the flexibility of the coordinating Cys. The mutant is destabilized because important hydrophobic interactions in the protein core involving Met11 are absent. In holo-CopZ, Met11 interacts extensively with residues in strands β2 and β3, including Val35, Asn36, Leu37, Gly40, Lys41, and Val42, but most strongly with Val35, Leu37, and Val42. Because of the lack of these interactions in holo-Met11Ala, after ∼80 ns of simulation the protein undergoes a large conformational change in which the interaction interface between strands β2 and β3 and helix R1 is lost, opening up the whole structure. Overlay of the final equilibrated structures shows these structural differences and reveals a rmsd of 4.1 Å

between WT and mutant structures (Figure 5B). Altogether, the destabilization observed in Met11Ala demonstrates the importance of Met11 in maintaining the integrity of the ferredoxinlike fold in holo-CopZ. The structural and dynamical defects observed in apoMet11Ala CopZ result in a significant alteration of the S(Cys)-S(Cys) distance distribution. Whereas WT exhibits two populations of distances centered at ∼4.5 and ∼7 Å, Met11Ala mutant has its Cys far from each other, most of the time at ∼9 Å (Figure 6A). This alteration will most probably affect the ability of Cu uptake, as the Cys will have to rearrange significantly to be able to coordinate Cu. Similarly to WT, in Met11Ala Ser12 interacts with Cys13 only in the holo form (Figure 6B,C). The interaction is more stable in the holo mutant, which may counterbalance the increased destabilization of the Cu loop in this mutant. On the other hand, the electrostatic stabilization of Cys16 by Tyr65 in the holo form is lost in



Figure 5. Superimposition of the final equilibrated structures of WT (black) and Met11Ala (blue) in the apo (A) and holo (B) forms. Left: overall structure. Center: blow-up of residues 7 to 15 and 35 to 43; Met/Ala11 and Ser12 (A) and Met/Ala11 (B) for each variant are labeled. Right: blow-up of residues 10 to 19 and 62 to 68 (A), and 8 to 17 and 62 to 68 (B); Met/Ala11, Ser12, Cys13, Cys16, Tyr65 and Cu for each variant are labeled.

Met11Ala: the Tyr65 fluctuates further away from Cys16 in the presence than in the absence of Cu in the mutant (Figure 6D,E). The gain of structure in the Cu loop in apo-Met11Ala is accompanied by complete burial of Cys13, whereas Cys16 is more exposed (Figure 7A,C). If Cys13 is buried it may not be able to receive Cu from the solvent or a donor. It is possible that in this case Cys16, which is more exposed, will be the first ligand to interact with Cu, and Cys13 will then have to change conformation to accommodate Cu coordination. In the presence of Cu, whereas Cys16 mobility and exposure is not affected by the mutation, Cys13 is more flexible and becomes significantly more exposed in Met11Ala (Figure 7B,D). Increased accessibility of Cys13 to the solvent in the holo form may increase the probability of Cu release to the solvent or Cu chelators. The average proximity to solvent of residues 11, 12, and 65 is shown in Figure 8. Ala11 in this mutant is more exposed than the corresponding Met11 in apo- and holo-WT (Figure 8A,B). Because Ala is a smaller residue it is able to move more freely than Met. In conclusion, because of its different conformation in apoand holo-CopZ, Met11 contributes to protein stability and

dynamics in different ways in the presence and absence of Cu. In apo-CopZ, Met11 is important for assuring flexibility of the Cu loop and the coordinating Cys, which may be requirements for proper Cu uptake. Met11 is essential for exposing Cys13 to solvent and maintaining an appropriate S(Cys)-S(Cys) distance that is optimal for Cu binding. In holo-CopZ, a buried Met11 is necessary to make key hydrophobic contacts within the protein core. This in turn protects the Cu site by not exposing Cys13 to the solvent. Ser12Ala and Ser12Thr CopZ Variants. As mentioned earlier, between residues Met11 and Cys13 there is a Ser12 in CopZ, which corresponds to a Thr in Atox1. To explore the role of this residue in CopZ and to directly compare it with the residue conserved at this position in Atox1, we mutated Ser12 to Ala and Thr. Ser12Thr has a greater negative effect than Ser12Ala in the apo form, but a smaller one in the holo form. Ser12Ala was simulated for ∼113 ns in both apo and holo forms, with mean rmsd (with respect to the first structure) and standard deviation of 3.1 ( 0.4 and 4.8 ( 0.4 Å, respectively, demonstrating greater conformational changes in the holo form (Figure 3). On the other hand, Ser12Thr was simulated ∼111 and ∼120 ns in apo



Figure 6. Histograms of distance distribution (in Å) between heavy atoms of Cys13 (S) and Cys16 (S) in the apo form (A), Ser/Thr12 (O) and Cys13 (S) in apo (B) and holo (C) forms, and between Tyr/Lys65 (O/N) and Cys16 (S) in apo (D) and holo (E) forms, for the different CopZ variants and WT. Black, WT; blue, Met11Ala; green, Ser12Ala; orange, Ser12Thr; red, Tyr65Ala; cyan, Tyr65Lys.

and holo forms, with mean rmsd (with respect to the first structure) and standard deviation of 5 ( 1 and 3.5 ( 0.4 Å, respectively, indicating in this case that the apo form has greater deviations and conformational dynamics than the holo form (Figure 3). Thus, it appears that apo-Ser12Thr and holoSer12Ala undergo significant conformational changes throughout the simulations, pointing to the importance of Ser12 in CopZ structure and dynamics. In the last 20 ns all variants are stable, with mean rmsd (with respect to the average structure of the last 20 ns) of 1.0 ( 0.2 Å and 1.2 ( 0.2 Å for apo- and holoSer12Ala, respectively, and 1.8 ( 0.5 Å and 1.2 ( 0.2 Å for apo- and holo-Ser12Thr, respectively. We notice that apoSer12Thr exhibits the greatest flexibility within all apo mutants. The two mutations have different effects on backbone dynamics in the two protein variants. In the absence of Cu,

Ser12Ala CopZ exhibits restricted backbone motion in the Cu loop, helix R1 and β2-β3 loop, but the rest of the protein is destabilized as compared to WT (Figure 4C). The most significant structural changes that occur in this mutant are an extension of helix R1 (residues 17 to 26 versus 20 to 24 in Ser12Ala and WT, respectively), a shortening of the Cu loop (residues 9 to 16 versus 10 to 19 in Ser12Ala and WT, respectively), and a complete loss of strand β4, which in WT corresponds to residues 67 to 72 (Figure 9A). Overlay of the equilibrated structures reveals a rmsd of 4.0 Å between Ser12Ala and WT. On the other hand, greater changes are observed for apo-Ser12Thr; an overlay of the equilibrated structures results in a rmsd of 5.7 Å with WT. This mutant is extremely flexible; the complete structure of this variant is destabilized as compared to WT (Figure 4C), and its structure is opened up (Figure 9A).



Figure 7. Protein-solvent radial distribution functions g(r) of Cys13 in apo (A) and holo (B) forms, and of Cys16 in apo (C) and holo (D) forms, for the different CopZ variants and WT. Black, WT; blue, Met11Ala; green, Ser12Ala; orange, Ser12Thr; red, Tyr65Ala; cyan, Tyr65Lys.

The increased flexibility found in this mutant is likely due to an entropic effect due to incorporation of an exposed methyl group at position 12. Because CopZ WT is very floppy to begin with, apparently it cannot tolerate a residue in position 12 that increases entropy further. In apo-CopZ WT, Ser12 interacts with Met11 and Glu38 in the β2-β3 loop. These interactions ensure a correct positioning of Met11 allowing it to interact with the β2-β3 loop, stabilizing the fold, as previously discussed. In Ser12Thr, because of its increased flexibility, Thr12 cannot interact with Met11 and in turn Met11 cannot interact with β2-β3 loop. This results in destabilization of this region and weakening of the loop-loop interface as also found in apoMet11Ala. As in apo-Ser12Ala, in Ser12Thr the secondary structure of strand β4 is completely lost. As opposed to the apo forms, the Ser12Ala change in CopZ has a greater effect than the Ser12Thr change in the presence of Cu. Both mutants are more flexible than WT, in particular in the Cu loop and helix R1; this is most dramatic in the Ser12Ala CopZ mutant (Figure 4D). In holo-WT, the interactions between Ser12 and Cys13 and Tyr65 allow for proper Culoop conformation allowing Met11 to be buried in the protein core. In Ser12Ala this interaction is abolished, Tyr65 moves away from the Cu loop, and consequently the loop becomes distorted, flexible, and extended (from residue 10 to 19); in addition, the entire fold is destabilized (Figure 9B). This conformational change prevents interactions between Met11 and the protein core, which adds to the destabilization. Also, helix R2 and strand β4 move away from the central ferredoxin-like fold, extending the overall protein size. The overlay of the final structures in this case results in a rmsd of 4.6 Å with respect to WT. In holo-Ser12Thr, Thr12 is also unable to interact with Cys13 and again, Met11 cannot interact with the protein core. However, in this case Met11 and Tyr65 interact with each other,

which assures proximity between the R2-β4 loop and the Cu loop, conserving in part the fold (Figure 9B). Superimposition of the final structures of holo forms of Ser12Thr and WT reveals a rmsd of 3.6 Å with respect to WT. Whereas in apo-Ser12Ala the S(Cys)-S(Cys) distance distribution is not greatly altered compared to WT (most of the time, ∼5 Å), in apo-Ser12Thr this distance increases dramatically, reaching values up to 14 Å, with two dominant peaks centered at 10 and 12 Å (Figure 6A). This longer distance will probably affect the efficiency of Cu uptake, similarly to that in the Met11Ala mutant. Formation of a HB between residue 12 and Cys13 in apo-Ser12Thr, which is absent in apo-WT (Figure 6B), may in part explain the altered Cys-Cys distance distribution. In holo-WT, an interaction between Ser12 and Cys13 serves as an electrostatic stabilization of the Cu center, but in holo-Ser12Thr this interaction is absent (Figure 6C). This property of Ser12Thr mimics the behavior of Ser12Ala, which also is unable to form this interaction. Also, in both mutants Tyr65 moves away from the Cu loop in the holo forms; this is more marked in Ser12Ala (Figures 6E and 9B). Therefore, it appears that both mutations cause destruction of electrostatic interactions of the Cu-bound state. Solvation of the functional Cys is not greatly affected in apoSer12Ala (Figure 7A,C). However, in apo-Ser12Thr solvent accessibility of both Cys residues is greatly altered, due to the increased flexibility of this region. Whereas Cys13 is more buried, Cys16 is much more solvent exposed, comparing to WT (Figure 7A,C). This difference further explains the longer S(Cys)-S(Cys) distances found in this mutant. In the presence of Cu, Cys13 is more buried in Ser12Ala whereas it is slightly more exposed in Ser12Thr (Figure 7B). Cys16 is more exposed in both mutants, to the largest degree in Ser12Ala CopZ (Figure



Figure 8. Protein-solvent radial distribution functions g(r) of Met/Ala11 in apo (A) and holo (B) forms, Ser/Ala/Thr12 in apo (C) and holo (D) forms, and Tyr/Ala/Lys65 in apo (E) and holo (F) forms for the different CopZ variants and WT. Black, WT; blue, Met11Ala; green, Ser12Ala; orange, Ser12Thr; red, Tyr65Ala; cyan, Tyr65Lys.

7D). Consequently, Cu is less protected in these mutants and, therefore, may be more susceptible to dissociation. Solvation of Met11 is affected differently in the two Ser12 mutants. In the apo form, whereas Met11 is more exposed in Ser12Ala, it becomes more buried in Ser12Thr, compared to apo-WT (Figure 8A). The structural changes that occur in the holo mutants mentioned above result in a complete exposure of Met11, as opposed to WT (Figure 8B). Exposure of Met11 in Ser12Ala mimics that of apo-WT, consistent with the capability of this residue to interact with the β2-β3 loop. In contrast, Ser12Thr has its Met11 even more exposed, without making any interaction with the rest of the protein. Based on our findings, it is clear that burial of Met11 in WT holo-CopZ is key to ensure a proper loop and Cys conformation. To conclude, Thr (the human residue) cannot substitute Ser in position 12 in CopZ. Ser12 is important for assuring a proper Cu-loop conformation and dynamics of the coordinating Cys

both in the apo and holo forms. Also, substitution of Ser12 by either an Ala or Thr results in a destabilization of the Cu-bound state due to loss of electrostatic stabilization of the Cu center provided by residues 12 and 65. Importantly, in these holo mutants Met11 is not buried, which is another source of destabilization. Tyr65Ala and Tyr65Lys CopZ Variants. Finally, the role of residue 65 in CopZ structure and dynamics was analyzed. This position is a Tyr in this protein and aligns with Lys60 in Atox1 and provides a HB to Cys16 in the presence of Cu. In this case, the presence of the eukaryotic residue in position 65 of CopZ is tolerated much better than an Ala substitution. Tyr65Lys, the most stable of all mutants, was simulated ∼107 and ∼120 ns with mean rmsd (with respect to the first structure) and standard deviation of 2.3 ( 0.3 and 2.6 ( 0.2 Å for apo and holo forms, respectively (Figure 3). In the last 20 ns, Tyr65Lys is very stable with mean rmsd (with respect to the



Figure 9. Superimposition of the final equilibrated structures of WT (black), Ser12Ala (green), and Ser12Thr (yellow) in the apo (A) and holo (B) forms. Left: overall structure. Center: blow-up of residues 7 to 14 and 35 to 43 (A), and 10 to 14 and 37 to 41 (B); Met11 and Ser/Ala/Thr12 (A) and Met11, Ser/Ala/Thr12 and Cys13 (B) for each variant are labeled. Right: blow-up of residues 9 to 19 and 64 to 66 (A), and 10 to 19 and 64 to 66 (B); Cys13, Cys16, Tyr65, and Cu for each variant are labeled.

average structure of the last 20 ns) of 0.7 ( 0.1 and 0.8 ( 0.1 Å for apo and holo forms, respectively. In contrast, Tyr65Ala is the least stable of all mutants, and thus, it was simulated longer times: ∼122 and ∼131 ns with mean rmsd (with respect to the first structure) and standard deviation of 3 ( 1 and 5 ( 2 Å for apo and holo forms, respectively (Figure 3). These values demonstrate great destabilization of this mutant in both forms, with significant structural dynamics. After ∼60 ns, it undergoes a conformational change in both apo and holo forms, as evidenced by a break point in the rmsd time evolution, in which the protein loses significantly secondary and tertiary structure. After this change, the rmsd (with respect to the average structure of the last 20 ns) is stable, with mean values of 1.0 ( 0.2 and 2.1 ( 0.6 Å for apo and holo forms, respectively. Both apo mutants have restricted motion of the Cu loop and helix R1 and, consequently, of the coordinating Cys (Figure 4E). Apart from this region, apo-Tyr65Lys has similar backbone fluctuations than WT. On the contrary, except for the loop and helix R1, apo-Tyr65Ala is more floppy than WT. In the holo forms, whereas the backbone of the Cu loop in Tyr65Lys is slightly more flexible than WT, the entire backbone of Tyr65Ala

is markedly destabilized (Figure 4F). Backbone fluctuations in holo-Tyr65Ala are on average 3 times greater than WT. Except for the Cu loop and helix R1, which gains structure, the structure of apo-Tyr65Lys is similar to WT, with a rmsd of 2.3 Å (Figure 10A). The presence of Lys in position 65 markedly stabilizes the Cu loop and helix R1, because this residue, instead of pointing to the solvent as Tyr65 in apo-WT, points toward the loop and interacts extensively with Cys16 (Figure 6D). This interaction also stabilizes the R2-β4 loop, in which residue 65 is located. On the contrary, apo-Tyr65Ala is largely destabilized and presents significantly less secondary structure than WT: helix R2 and strand β4 are unfolded (Figure 10A). Also, the Cu loop is shortened, going from residue 10 to 15, so that now Cys16 is part of helix R1, as opposed to WT in which it is part of the loop. Because of this, helix R1 adopts a different orientation and is shifted, extending from residue 16 to 19 versus 20 to 23 in Tyr65Ala and WT, respectively. After ∼60 ns the protein undergoes a large structural change in which it loses tertiary contacts: helix R2 and strand β4 no longer fold against strands β1-β3. In this way, the entire ferredoxin-like fold is destabilized. In apo-WT, Tyr65 forms a strong HB with



Figure 10. Superimposition of the final equilibrated structures of WT (black), Tyr60Ala (red), and Tyr65Lys (cyan) in the apo (A) and holo (B) forms. Left: overall structure. Center: blow-up of residues 8 to 13 and 36 to 42 (A), and 10 to 14 and 37 to 41 (B); Met11 and Ser12 for each variant are labeled. Right: blow-up of residues 9 to 19 and 64 to 66 (A), and 10 to 19 and 64 to 66 (B); Cys13, Cys16, Tyr/Ala/Lys65 and Cu for each variant are labeled.

Gln63, which positions the latter for a stable interaction with its consecutive residue Asp62. These residues are all located in the R2-β4 loop. The latter interaction allows interaction between Asp62 and Lys18 located at the end of the Cu loop. This network of interactions present in WT apo-CopZ preserves tertiary contacts and conserves the fold. In Tyr65Ala these contacts are not present and thus the stability and structure of the fold is compromised, with a final rmsd of 4.8 Å compared to WT. In the presence of Cu, the structural changes observed for the apo forms are attenuated in Tyr65Lys but are markedly intensified in Tyr65Ala. The equilibrated WT and Tyr65Lys holo structures are very similar, with a rmsd of 1.5 Å (Figure 10B). Differences include lengthening of the Cu loop by one residue (strand β1 is shorter) and loss of some secondary structure in strand β4. On the other hand, holo-Tyr65Ala is destabilized significantly, with a final rmsd with WT of 5.8 Å (Figure 10B). As in apo-Tyr65Ala, lack of important secondary and tertiary contacts because of lack of Tyr65 are the consequence of a similar conformational change that occurs after ∼60 ns. However, in this case the overall effect is greater, because Tyr65 (or Lys65 in case of Tyr65Lys) is an essential source of

electrostatic stabilization of the Cu bis-thiolate center. As a consequence, helix R2 and the R2-β4 loop completely lose their interface with the Cu loop and helix R1. Opening of the ferredoxin-fold exposes a great portion of the protein core to the solvent, which in turn destabilizes even more the backbone. As a consequence, strand β1 changes orientation and moves toward helix R1, protecting, in part, the hydrophobic core. Other important structural changes include the lengthening of the Cu loop (residues 9 to 19 in Tyr65Ala versus 10 to 13 in WT) and distortion of its conformation and destabilization and altered orientation of helix R1. All the structural changes described explain the ∼3-fold increment in backbone fluctuation of this mutant compared to WT, and probably suggest that this mutant will unfold and lose Cu in solution. The distance between the coordinating Cys is not greatly altered in apo-Tyr65Lys, but it is longer in Tyr65Ala (Figure 6A). In the latter, these values hardly reach 4 Å, as opposed to WT and Tyr65Lys. As mentioned above, Lys65 in apoTyr65Lys is close to Cys16 as opposed to WT (Figure 6D), and this appears to be key to prevent lost of tertiary contacts, as in Tyr65Ala. In the presence of Cu, as opposed to Tyr65Lys and WT, Ala65 in Tyr65Ala is not able to interact with Ser12,



TABLE 1: Summary of the Most Significant Changes in the Different Mutants with Respect to WT CopZa

a +, increase; -, decrease with respect to WT of backbone flexibility, length of Cys-Cys distance and Met/Cys solvent exposure. ( represents a mix of increased/decreased flexibility in different regions of the protein. WT means wild-type like behavior of that property. In apo-WT, there is no interaction between S12 and C13 and between Y65 and C16, and M11 is exposed. In holo-WT, there is an interaction between S12 and C13 and between Y65 and C16, and M11 is buried and interacts with the hydrophobic core.

and in turn Ser12 cannot interact with Cys13 (Figure 6C), which may contribute further to the destabilization of this holo mutant. As mentioned above, the same is true for the electrostatic stabilization of the Cu-bound state provided by Tyr65, which is completely absent in Tyr65Ala. Lys65 in Tyr65Lys is still able to provide this interaction (Figure 6E), and that is probably why this residue is tolerated in this position. Solvation of the Cys is unaltered in Tyr65Lys, regardless the presence of Cu (Figure 7), consistent with normal Cys-Cys distances and normal interactions involving the Cys in this mutant. This is not true for Tyr65Ala, in which both Cys are more buried in the apo form (Figure 7A,C), and this correlates with altered Cys-Cys distances. Solvation of Met11 is not greatly altered in apo- and holo-Tyr65 mutants, although this residue is slightly more exposed in apo-Tyr65Lys (Figure 8A,B). In summary, in the absence of Cu, Tyr65 is essential for CopZ structure and dynamics, assuring appropriate Cu-loop conformation and dynamics that will aid in Cu uptake. In this case, Lys may not be a good replacement for Tyr, as apo-Tyr65Lys exhibits a loop that is more structured and rigid. On the other hand, in the presence of Cu, Lys can substitute Tyr in this position. Presence of a residue capable of donating a HB to Cys16, as Tyr or Lys, in position 65 stabilizes the Cu-bound state and provides a platform for the interaction between helices R2 and R1. Comparison to Human Cu Chaperone Homologue. Recently, we reported the consequences of removing Met10, Thr11, and Lys60 in structure and dynamics of apo- and holo-Atox1.42 Here, we have assessed the consequences of removing the corresponding residues in the bacterial homologue CopZ, Met11, Ser12, and Tyr65. Although close to the Cu-binding loop, none

of the residues are interacting with the Cu. Met11 is conserved in all organisms, including Cu chaperones and MBDs, whereas Ser12 and Tyr65 are only conserved as such in prokaryotic Cu chaperones.21 We have simulated Met11Ala, Ser12Ala, Ser12Thr, Tyr65Ala, and Tyr65Lys variants of CopZ in apo and holo forms until structural equilibrium was reached, resulting in an impressive total simulation time of ∼1.2 µs. We have provided a detailed description of the effects of Cu-loop structure and dynamics, overall protein flexibility, S(Cys)-S(Cys) distance and solvent exposure of key residues as a function of these mutations. The most important differences with respect to CopZ WT are summarized in Table 1. CopZ mutants are much more unstable than the corresponding mutants in Atox1,42 and required longer simulations times to reach equilibrium. This is consistent with the fact that CopZ WT is in fact a more floppy protein than Atox1 WT as evidenced by MD simulations25 and in vitro protein stabilities.29 As opposed to in Atox1, where the changes are subtler and affect mainly the Cu-loop structure and dynamics,42 mutations in CopZ have greater effects and alter significantly the protein structure regardless the presence of Cu. This suggests that the human protein is able to tolerate mutations much better than the bacterial one, at least at the structural level. As in Atox1, each CopZ variant studied here has a unique behavior, which allows dissection of the role of each residue in CopZ structure and dynamics. Comparison with the corresponding mutations in Atox1 can further clarify the conservation of certain residues (like Met11) and the divergence of others (like Ser12 and Tyr65) in bacterial and human Cu chaperones. An important difference between CopZ and Atox1 is the behavior of Met11/10: always buried in Atox1 but only buried

Role of Conserved Residues in CopZ Dynamics in holo-CopZ. Similar to the role of Thr11 in apo-Atox1,42 solvent exposure of the hydrophobic side chain of Met11 in apo-CopZ may serve as an entropic effect that increases the flexibility of this protein, thus explaining in part the difference in floppiness between Atox1 and CopZ. This is evidenced by the fact that whereas replacement of Met11 for Ala in apoCopZ significantly reduces backbone motion, Met10Ala mutation in apo-Atox1 has the opposite effect.42 It thus appears that exposure of Met11 in apo-CopZ has evolved as a way of assuring protein and Cu-loop flexibility to facilitate Cu uptake. On the other hand, some similarities are found between the role of Met11 in CopZ and Met10 in Atox1. In both cases, Met is important for proper Cu-loop dynamics and Cys exposure. In the apo form, improper packing of the protein core as a consequence of Met removal alter the Cu-loop dynamics significantly (although in opposite ways), and consequently affect the S(Cys)-S(Cys) distance distribution for optimal Cu binding, as both Met mutants have their Cys much further away from each other. Because the Cys will have to significantly rearrange to bind Cu in these mutants, this is predicted to affect Cu uptake. Based on previous experimental results for Cu removal in Met10Ala Atox1,40 it appears that the net effect of Met mutations in Cu uptake will include a combination of kinetic and thermodynamic contributions. In the holo form, the behavior of the conserved Met is similar in both CopZ and Atox1. In both cases the Met assures stability of the fold by participating in several hydrophobic contacts with the protein core. In fact, several residues located in strands β2 and β3 that interact extensively with Met11, including Val35, Leu37, and Val42, are conserved between CopZ and Atox1 (Val35 in CopZ is replaced by another hydrophobic residue in Atox1, Ile33), pointing to the conservation of a network of interactions that contributes to the ferredoxin-like fold. This suggests that Met shares the same function in both holo proteins: to contribute to the stability of the fold. Thus, it appears that even if Met is conserved at the sequence level between prokaryotes and eukaryotes, its location and dynamics in the tertiary structure has adapted to the intrinsic “needs” of each system. apo-CopZ may need to have increased conformational flexibility than Atox1 to interact with its partner proteins in vivo, and exposure of Met11 may have evolved to provide this entropic resource. On the other hand, Atox1 may be flexible enough for interaction with its partners, and exposure of Met is therefore detrimental for its stability. Recently, we found that the Thr11 and Lys60 mutated Atox1 variants, particularly in the apo form, were more rigid than WT Atox1, suggesting that the presence of these two residues increase the entropy of WT Atox1.42 Except for apo-Ser12Thr, Ser12, and Tyr65 CopZ apo variants have also reduced flexibility of the Cu loop compared with apo-WT. The increased flexibility of the Cu loop due to the presence of Ser12 and Tyr65 in WT CopZ can be explained with similar arguments to those of Atox1.42 Ser12 and Tyr65 do not maintain any stable interaction in the apo form; instead their side chains are highly mobile that fluctuate close to the Cu loop. Proper plasticity of the Cu loop and thereby appropriate solvent exposure of the two Cys may be necessary for efficient Cu uptake in vivo. On the other hand, in agreement to what we observed between the apo forms of WT and Thr11Ser Atox1, addition of a methyl group in the side chain of residue 12 in CopZ (as in Ser12Thr variant) further destabilizes the backbone, as the exposed hydrophobic group serves as an entropic booster. Also and similarly to what we observed for Atox1, some of the Ser and Tyr apo mutants have Cys13 more buried than WT. This may

J. Phys. Chem. B, Vol. 113, No. 7, 2009 1931 affect the efficiency of Cu uptake, as Cys12 is the first Cys proposed to interact with Cu.41 Ser12 and Tyr65 are important stabilizing factors for holoCopZ, as Thr11 and Lys60 in holo-Atox1, since both residues participate in an electrostatic network of interactions that ultimately stabilize the Cu-bound state. This network further assures proper solvation of the Cu-coordinating Cys, as also seen for Atox1, which may facilitate Cu uptake and release. In CopZ, it appears that the presence of a polar residue (Tyr, or Lys as tested here) in position 65 is essential for maintaining the integrity of the holo protein. holo-Tyr65Ala variant is extremely unstable in our MD simulations, and this protein will probably unfold and lose the Cu in vitro. Surprisingly, Ser12 is essential to maintain Met11 buried in holo-CopZ, as mutation to either Ala or Thr completely exposes this residue. Burial of Met11 in holo-CopZ is the most significant change that occurs upon Cu binding and contributes to the stabilization of the Cubound state by providing core interactions. Similarly to what we observed for Atox1,42 Ser12Thr and Tyr65Lys variants of CopZ did not behave WT-like. The human residue, Lys, is tolerated better than Ala in position 65 in both apo- and holo-CopZ, as also seen for the prokaryotic residue in that position in Atox1.42 In the holo form, Lys is able to keep the electrostatic network, but since this residue interacts with Cys16 also in the apo-Tyr65Lys variant, it stiffens the Cu loop in the apo form. As noted above, the intrinsic flexibility of the Cu loop in apo-CopZ WT may be necessary for efficient Cu uptake and/or interaction with its partner proteins in vivo. Therefore, the reduction of loop flexibility in Tyr65Lys may affect these abilities. On the other hand, whereas Thr12 was tolerated better than Ala12 in holo-CopZ, Ala was a better replacement in the apo form, as observed for residue 11 in Atox1.42 In the apo form, Ser12Thr is the most flexible of all CopZ variants. These changes are the result of an entropic effect due to the exposure of the methyl group of Thr12 in apoSer12Thr, just as happens for apo-Atox1 WT.42 In Atox1 WT, in which Met10 is completely buried in the core, the presence of Thr in position 11 assures proper Cu-loop dynamics that would otherwise be too rigid for Cu uptake. In apo-CopZ, however, exposure of Met11 significantly destabilizes the Cu loop, so the presence of another residue that increases backbone entropy further in position 12 is not tolerated and the fold is compromised. Furthermore, similar to what was seen with the Thr11Ser mutant in apo-Atox1,42 the Ser12Thr mutant in apoCopZ gains a HB between residue 12 and Cys13 and has improper Cys solvation compared to the WT protein, which may explain the altered Cys(S)-Cys(S) in both mutants. In the holo form, in which Ser12 and Cys13 interact in WT, the Ser12Thr variant is able to provide such interaction (to a limited extent though), whereas the Ser12Ala is not. Thus, it appears that residues in position 12 and 65 in CopZ, and 11 and 60 in Atox1, have evolved to tune the flexibility of the Cu loop in the apo form and provide electrostatic stabilization of the holo form. Conclusions This work provides a detailed analysis of the roles of key residues located close to the Cu site in structure and dynamics of the bacterial Cu chaperone CopZ. Simulations of CopZ pointmutated variants together with a thorough comparison to the corresponding mutations in the human Cu chaperone Atox1 provide a clear understanding of the reasons behind the conservation of Met and divergence of residues X1 and 65/60 between CopZ and Atox1. We conclude that residues in Cu chaperones have evolved to maintain a delicate balance of the

1932 J. Phys. Chem. B, Vol. 113, No. 7, 2009 Cu-binding loop flexibility and core rigidity, which may likely be important for Cu uptake and release and for interaction with partners in vivo. Abbreviations: Cu, copper; MBD, metal binding domain; WT, wild-type; MD, molecular dynamics; rmsd, root-meansquare deviation; rmsf, root-mean-square fluctuation; HB, hydrogen bond. Acknowledgment. We thank Dr. Alejandro Crespo for useful discussions. Support for this project was provided by the Robert A. Welch Foundation (C-1588). This work was supported in part by the Rice Computational Research Cluster funded by NSF under Grant CNS-0421109, and a partnership between Rice University, AMD, and Cray; and by the Shared University Grid at Rice funded by NSF under Grant EIA-0216467, and a partnership between Rice University, Sun Microsystems, and Sigma Solutions, Inc. References and Notes (1) O‘Halloran, T. V.; Culotta, V. C. J. Biol. Chem. 2000, 275, 25057. (2) Huffman, D. L.; O’Halloran, T. V. Annu. ReV. Biochem. 2001, 70, 677. (3) Puig, S.; Rees, E. M.; Thiele, D. J. Structure 2002, 10, 1292. (4) Puig, S.; Thiele, D. J. Curr. Opin. Chem. Biol. 2002, 6, 171. (5) Harris, E. D. Crit. ReV. Clin. Lab. Sci. 2003, 40, 547. (6) Rae, T. D.; Schmidt, P. J.; Pufahl, R. A.; Culotta, V. C.; O‘Halloran, T. V. Science 1999, 284, 805. (7) Rosenzweig, A. C.; O’Halloran, T. V. Curr. Opin. Chem. Biol. 2000, 4, 140. (8) Harrison, M. D.; Jones, C. E.; Solioz, M.; Dameron, C. T. Trends Biochem. Sci. 2000, 25, 29. (9) Hamza, I.; Schaefer, M.; Klomp, L. W.; Gitlin, J. D. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 13363. (10) Hung, I. H.; Casareno, R. L.; Labesse, G.; Mathews, F. S.; Gitlin, J. D. J. Biol. Chem. 1998, 273, 1749. (11) Klomp, L. W.; Lin, S. J.; Yuan, D. S.; Klausner, R. D.; Culotta, V. C.; Gitlin, J. D. J. Biol. Chem. 1997, 272, 9221. (12) Hung, I. H.; Suzuki, M.; Yamaguchi, Y.; Yuan, D. S.; Klausner, R. D.; Gitlin, J. D. J. Biol. Chem. 1997, 272, 21461. (13) Kihlken, M. A.; Leech, A. P.; Le Brun, N. E. Biochem. J. 2002, 368, 729. (14) Radford, D. S.; Kihlken, M. A.; Borrelly, G. P.; Harwood, C. R.; Le Brun, N. E.; Cavet, J. S. FEMS Microbiol. Lett. 2003, 220, 105. (15) Banci, L.; Bertini, I.; Ciofi-Baffoni, S.; Del Conte, R.; Gonnelli, L. Biochemistry 2003, 42, 1939. (16) Banci, L.; Bertini, I.; Del Conte, R.; Markey, J.; Ruiz-Duenas, F. J. Biochemistry 2001, 40, 15660. (17) Banci, L.; Bertini, I.; Del Conte, R. Biochemistry 2003, 42, 13422. (18) Rosenzweig, A. C.; Huffman, D. L.; Hou, M. Y.; Wernimont, A. K.; Pufahl, R. A.; O’Halloran, T. V. Structure 1999, 7, 605. (19) Banci, L.; Bertini, I.; Ciofi-Baffoni, S.; Huffman, D. L.; O’Halloran, T. V. J. Biol. Chem. 2001, 276, 8415.

Rodriguez-Granillo and Wittung-Stafshede (20) Arnesano, F.; Banci, L.; Bertini, I.; Huffman, D. L.; O’Halloran, T. V. Biochemistry 2001, 40, 1528. (21) Arnesano, F.; Banci, L.; Bertini, I.; Ciofi-Baffoni, S.; Molteni, E.; Huffman, D. L.; O’Halloran, T. V. Genome Res. 2002, 12, 255. (22) Ralle, M.; Lutsenko, S.; Blackburn, N. J. J. Biol. Chem. 2003, 278, 23163. (23) DeSilva, T. M.; Veglia, G.; Opella, S. J. Proteins 2005, 61, 1038. (24) Pufahl, R. A.; Singer, C. P.; Peariso, K. L.; Lin, S. J.; Schmidt, P. J.; Fahrni, C. J.; Culotta, V. C.; Penner-Hahn, J. E.; O’Halloran, T. V. Science 1997, 278, 853. (25) Rodriguez-Granillo, A.; Wittung-Stafshede, P. J. Phys. Chem. B 2008, 112, 4583. (26) Zhou, L.; Singleton, C.; Le Brun, N. E. Biochem. J. 2008, 413, 459. (27) Poger, D.; Fuchs, J. F.; Nedev, H.; Ferrand, M.; Crouzy, S. FEBS Lett. 2005, 579, 5287. (28) Wernimont, A. K.; Huffman, D. L.; Lamb, A. L.; O’Halloran, T. V.; Rosenzweig, A. C. Nat. Struct. Biol. 2000, 7, 766. (29) Hussain, F.; Wittung-Stafshede, P. Biochim. Biophys. Acta 2007, 1774, 1316. (30) Leach, A. R. Molecular Modeling; Addison Wesley Longman: New York, 1997. (31) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995, 117, 5179. (32) Pearlman, D. A.; Case, D. A.; Caldwell, J. W.; Ross, W. S.; Cheatham III, T. E.; DeBolt, S.; Ferguson, D.; Seibel, G.; Kollman, P. Comput. Phys. Commun. 1995, 91, 1. (33) Case, D. A.; Cheatham, T. E., 3rd; Darden, T.; Gohlke, H.; Luo, R.; Merz, K. M., Jr.; Onufriev, A.; Simmerling, C.; Wang, B.; Woods, R. J. J. Comput. Chem. 2005, 26, 1668. (34) Case, D. A.; Darden, T. A.; Cheatham, T. E., 3rd; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Merz, K. M.; Pearlman, D. A.; Crowley, M.; Walker, R. C.; Zhang, W.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.; Wong, K. F.; Paesani, F.; Wu, X.; Brozell, S.; Tsui, V.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.; Hornak, V.; Cui, G.; Beroza, P.; Mathews, D. H.; Schafmeister, C.; Ross, W. S.; Kollman, P. A. AMBER 9; University of California: San Francisco, 2006. (35) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J.; Impey, R. W.; Klein, M. L. J. Chem. Phys. 1983, 79, 926. (36) Hornak, V.; Abel, R.; Okur, A.; Strockbine, B.; Roitberg, A.; Simmerling, C. Proteins 2006, 65, 712. (37) Berendsen, H. J.; Postma, J. P.; van Gunsteren, W. F.; Di Nola, A.; Haak, J. R. J. Chem. Phys. 1984, 81, 3684. (38) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. J. Comput. Phys. 1977, 23, 327. (39) Anastassopoulou, I.; Banci, L.; Bertini, I.; Cantini, F.; Katsari, E.; Rosato, A. Biochemistry 2004, 43, 13046. (40) Hussain, F.; Olson, J. S.; Wittung-Stafshede, P. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 11158. (41) Banci, L.; Bertini, I.; Cantini, F.; Felli, I. C.; Gonnelli, L.; Hadjiliadis, N.; Pierattelli, R.; Rosato, A.; Voulgaris, P. Nat. Chem. Biol. 2006, 2, 367. (42) Rodriguez-Granillo, A.; Wittung-Stafshede, P. Biochemistry 2009, in press.

JP807594Q

Role of Conserved Residues - American Chemical Society

Recommend Documents