Roles of glutamates and metal ions in a rationally designed nitric oxide reductase based on myoglobin

Abstract

A structural and functional model of bacterial nitric oxide reductase (NOR) has been designed by introducing two glutamates (Glu) and three histidines (His) in sperm whale myoglobin. X-ray structural data indicate that the three His and one Glu (V68E) residues bind iron, mimicking the putative Fe_B site in NOR, while the second Glu (I107E) interacts with a water molecule and forms a hydrogen bonding network in the designed protein. Unlike the first Glu (V68E), which lowered the heme reduction potential by ∼110 mV, the second Glu has little effect on the heme potential, suggesting that the negatively charged Glu has a different role in redox tuning. More importantly, introducing the second Glu resulted in a ∼100% increase in NOR activity, suggesting the importance of a hydrogen bonding network in facilitating proton delivery during NOR reactivity. In addition, EPR and X-ray structural studies indicate that the designed protein binds iron, copper, or zinc in the Fe_B site, each with different effects on the structures and NOR activities, suggesting that both redox activity and an intermediate five-coordinate heme-NO species are important for high NOR activity. The designed protein offers an excellent model for NOR and demonstrates the power of using designed proteins as a simpler and more well-defined system to address important chemical and biological issues.

Rational design of proteins that mimic both structure and function of more complex native enzymes has been a long sought-after goal, as the process is an ultimate test of our knowledge and an excellent means to develop advanced biocatalysts (1–3). Although designed proteins that model the structure of native enzymes have been known for a while (4–10), successful designs of proteins that mimic both the structure and function of native enzymes have been reported only recently (11–16). While being able to design such functional proteins is laudable, the impact of such an achievement would be greater if the designed proteins can be used to address fundamental issues in chemistry and biology that are difficult to tackle by other methods. One primary example is the roles of conserved glutamates and metal ions in bacterial nitric oxide reductase (NOR) (17–19).

NO is critical for all life (20). Bacterial denitrification is a crucial part of the nitrogen cycle in nature that involves a four-step, five-electron reduction of nitrate () to dinitrogen (N₂) (17, 19). Bacterial NOR is a membrane-bound protein that catalyzes one step of this process, namely, the two-electron reduction of NO to N₂O (17, 19). With no crystal or solution structure available for bacterial NOR to date, sequence alignments and homology modeling (21, 22) have indicated that NOR is structurally homologous to the largest subunit (subunit I) of heme-copper oxidases (HCOs) (23), enzymes that catalyze reduction of O₂ to water. The active sites of both NOR and HCO contain a proximal histidine-coordinated heme and a distal three histidine-coordinated metal center. However, the metal center in HCOs is occupied by a copper (called Cu_B), whereas a nonheme iron is present in NOR (called Fe_B) (23, 24). In addition, two conserved glutamates, shown by modeling to be close to the Fe_B site (21, 22), are found to be essential for NOR activity (24, 25). Some members of HCOs such as cytochrome cbb₃ oxidase display NOR activity (26–28), although the activity is ∼50-fold lower than native NOR (26). Therefore, it is important to elucidate the structural features, specifically the roles of the conserved glutamates close to the Fe_B site and metal ions (copper vs. iron), responsible for the reduction of NO to N₂O.

To address these issues, biochemical and biophysical studies of native NOR and its variants have been carried out (24, 25, 29–37). For example, Richardson and coworkers investigated the effects of amino acid substitutions of the five conserved glutamates (E122 and E125 presumed to face the periplasm and E198, E202, and E267 located in the interior of the membrane, close to the catalytic site) in the catalytic subunit of Paracoccus denitrificans, NorB. The E122A, E125A, E198A, and E267A variants were inactive, indicating that these four glutamates are crucial for NOR activity (24, 25, 32, 33). On the other hand, Reimann et al. constructed a 3D model of NorB using homology modeling with the structures of HCOs as templates and suggested a plausible pathway consisting of these conserved glutamates for proton delivery (22). Despite these successes, the roles of the conserved glutamates and metal ions still remain to be fully elucidated, partly because of the difficulty in obtaining native NOR in high yield and the lack of a 3D structure. Even if these problems are resolved, it is still difficult to replace iron in the native Fe_B site with other metal ions, and spectroscopic studies of native NOR are often complicated by the presence of other metal cofactors (e.g., low-spin heme).

To overcome these limitations, a number of synthetic models of NOR using small organic molecules as ligands, have been made in which the nonheme Fe_B site can be replaced by a copper ion (17, 38–45). In addition, since these model systems lack additional metal-binding sites, spectroscopic studies are often simplified. Therefore, studies of these synthetic models have offered many insights. For example, Collman et al. showed that a fully reduced heme/nonheme Fe_B compound can react with two equivalents of NO leading to the formation of one equivalent of N₂O and a bis-ferric product (41). On the other hand, Karlin and coworkers showed that a small heme/Cu complex can efficiently lead to reductive coupling of NO to N₂O (43). However, it is also difficult to obtain the synthetic models in high yield due to the multiple steps required in chemical synthesis. Because of this limitation, no synthetic NOR model containing the two key conserved Glu residues (E198 and E267 in NOR) has been reported. It is also difficult to substitute different metal ions in the same metal-binding site without perturbing the site geometry and distances to the heme iron, as most ligands are not as rigid as those in native enzymes and different metal ions have different geometric and ligand donor set preferences.

We have recently designed a structural and functional protein model of bacterial NOR by engineering three histidines and one glutamate into the distal pocket of sperm whale myoglobin (swMb, L29H, F43H, H64, and V68E, named Fe_BMb) (14). Like synthetic models, this “bottom-up” approach complements the “top-down” approach of the study of native NOR in that it provides insights into whether certain “necessary” structural elements are enough to impart enzyme function. Thanks in part to recent advances in computational, molecular, and structural biology, the designed myoglobin protein model is much easier to synthesize and to crystallize than either native NOR or synthetic models. Since myoglobin has often been used for the development and calibration of numerous spectroscopic techniques (46–48), it is an ideal choice for spectroscopic studies. More importantly, the rigid protein network allows precise placement of either glutamate or metal ions in myoglobin to address their roles in NOR activity. Toward this goal, we have demonstrated that both the histidines and one of the glutamates are essential for iron binding and NO reduction activity (14). However, the role of the second Glu close to the Fe_B site and the role of different metal ions in the Fe_B site have not been addressed.

To address these important issues and to design even closer protein models of NOR, we introduced herein the second Glu to the second coordination sphere of the Fe_B site by mutating an Ile to a Glu (named I107E Fe_BMb). We show that the second Glu results in a ∼100% increase in NOR activity through hydrogen bonding interactions and that the two glutamates have dramatically different effects on the heme reduction potential. Additionally, by comparing the EPR, electrochemistry, X-ray structures, and NOR activity of iron, copper, and zinc derivatives of the designed protein, we have obtained deeper insights into the roles of metal ions in NOR.

Results

Structure and Function of Fe(II)-I107E Fe_BMb.

The X-ray crystal structures of heme-containing I107E Fe_BMb without metal ion in the Fe_B site and with Fe²⁺ in the Fe_B site are solved at 1.42-Å and 1.65-Å resolution, respectively (Fig. 1 A and B and Table S1). In the absence of metal ions in the Fe_B site, the structure shows a water molecule in the Fe_B site, which forms hydrogen bonds with NE2 atoms of all three His residues, both OE1 and OE2 atoms of E68, and the OE2 atom of E107 (Fig. 1A). Upon binding Fe²⁺, the Fe(II)-I107E Fe_BMb structure shows that Fe²⁺ is coordinated by three His, the OE2 atom of E68, and one water molecule. Notably, a water molecule bridges Fe²⁺ in the Fe_B site and the second glutamate (E107) with a distance of 2.32 Å to the OE2 atom of E107 (Fig. 1B).

Fig. 1.

Crystal structures of I107E Fe_BMb (A) (PDB ID code 3M38), Fe(II)-I107E Fe_BMb (B) (PDB ID code 3M39), Cu(II)-I107E Fe_BMb (C) (PDB ID code 3M3A), and Zn(II)-I107E Fe_BMb (D) (PDB ID code 3M3B). Water molecules, Fe(II), Cu(II), and Zn(II) are represented by red, green, orange, and gray spheres, respectively.

To probe the conformational changes of introducing the second Glu (E107), we performed a structural alignment of Fe(II)-I107E Fe_BMb and the previously reported Fe(II)-Fe_BMb (14). The comparison, shown in Fig. 2, indicates that both the polypeptide chain and the active site overlap well with each other. In addition, the two nonheme irons are located at similar positions with a 0.36-Å separation from each other. In contrast, E68 underwent a significant conformational rearrangement in the presence of E107. These observations suggest that the active site of Fe_BMb can be tuned by the formation of an extended hydrogen bonding network, resulting from the introduction of a second glutamate residue.

Fig. 2.

Overlay of Fe(II)-I107E Fe_BMb (cyan) (PDB ID code 3M39) with Fe(II)-Fe_BMb (orange) (PDB ID code 3K9Z).

The binding of Fe²⁺ to deoxy I107E Fe_BMb was further monitored by EPR (Fig. 3A). Since deoxy myoglobin contains Fe(II) heme that exhibits no EPR signals in X-band EPR (14), we added blue copper Cu(II)-azurin (49), a redox partner of native NOR (19), to oxidize both the reduced heme and nonheme irons in Fe(II)-I107E Fe_BMb to EPR-active Fe(III). Upon addition of Cu(II)-azurin, the oxidation of deoxy I107E Fe_BMb resulted in EPR signals at g = 6.12 and 5.56, typical of a high-spin heme-Fe(III). Upon addition of Fe²⁺, however, a decrease of the heme-Fe(III) EPR signals was observed, indicating that the Fe²⁺, when bound to the Fe_B site and oxidized by Cu(II)-azurin, is spin-coupled to heme-Fe(III). Such a spin coupling mimics that in NOR (35, 50–53), suggesting that I107E Fe_BMb models NOR closely, at least in this respect.

Fig. 3.

EPR spectra of deoxy I107E Fe_BMb (0.5 mM protein in 50 mM Bis-Tris, pH 7.0) with increasing concentrations of Fe²⁺ in the presence of wild-type Cu(II)-azurin (A), and oxidized I107E Fe_BMb with Cu²⁺ (B) or Zn²⁺ (C). Spectra were collected at 4 K, 5 mW power, and 9.05 GHz.

To probe the role of the second Glu (E107) in NO reduction activity, we measured the yield of N₂O production by Fe(II)-I107E Fe_BMb with excess NO under one turnover conditions. We monitored N₂O formation in the headspace of the solution using GC/MS and compared this result to that of Fe(II)-Fe_BMb, which lacks the second Glu (Fig. 4). Remarkably, Fe(II)-I107E Fe_BMb displays higher activity than Fe(II)-Fe_BMb. After ∼20 hr, ∼24% N₂O was produced by Fe(II)-I107E Fe_BMb, in contrast to ∼10% yield for Fe(II)-Fe_BMb, strongly indicating that the second Glu plays an important role in NO reduction, likely facilitating proton uptake during NO reduction.

Fig. 4.

Time-dependent N₂O production by Fe(II)-I107E Fe_BMb (▴) and Fe(II)-Fe_BMb (●) with ∼50 eq. NO under single turnover conditions. The yield was determined by a comparison of the ratio of NO∶N₂O peaks from the GC/MS chromatograms.

Other Metal Ions Binding to I107E Fe_BMb.

To find out if the resting state of the protein, i.e., oxidized or met I107E Fe_BMb, can bind other metal ions, Cu²⁺ or Zn²⁺ was titrated into met I107E Fe_BMb and monitored by EPR spectroscopy (Fig. 3 B and C). In the absence of metal ions, met I107E Fe_BMb exhibited high-spin heme signals at g = 6.03, 5.08, and 1.98 (Fig. 3B, black line). Upon addition of 2 eq of Cu²⁺, the signals at g = 6.03 and 5.08 decreased and a broad peak around g = 2.95 increased, probably due to spin coupling between heme-Fe(III) and Cu²⁺ in the Fe_B site. In contrast, addition of Zn²⁺, a metal ion with no unpaired electrons [i.e., incapable of spin coupling to heme-Fe(III)], produced an increase in the high-spin heme signals at g = 5.88 and 5.60 (Fig. 3C), indicating that the interaction between E68 and heme iron was weakened after metal binding.

The X-ray crystal structures of I107E Fe_BMb with Cu²⁺ or Zn²⁺ in the Fe_B site were solved at 1.37-Å and 1.60-Å resolution, respectively (Fig. 1 C and D and Table S1). Compared to Fe(II)-I107E Fe_BMb (Fig. 1B), a similar binding site was observed for Cu(II)-I107E Fe_BMb (Fig. 1C), where H29, H43, and H64 coordinate to Cu²⁺ with distances of 2.09, 2.10, and 2.04 Å, respectively, slightly shorter than the corresponding distances in the Fe²⁺ structure. In comparison to Fe(II)-I107E Fe_BMb, the water bridging the Cu²⁺ and the second Glu (E107) is shifted toward Cu²⁺ in the Fe_B site (2.03 Å) with respect to E107 (3.04 Å). Interestingly, this bridging water molecule was not observed in Zn(II)-I107E Fe_BMb (Fig. 1D), but the two O atoms of E68 coordinate to Zn²⁺ with similar distances (2.26 Å for OE1 and 2.29 Å for OE2). The longer distance between OE1 of E68 and heme iron in the Zn-bound structure (2.68 Å) in comparison to the Cu- and Fe-bound structures, is likely the result of a weaker interaction, which is also supported by an observed increase of the high-spin heme signals in the EPR spectra upon Zn²⁺ binding (Fig. 3C). These results suggest I107E Fe_BMb is capable of incorporating different metal ions into its designed Fe_B site, offering an excellent opportunity to compare the role of these metal ions in the same protein scaffold.

Effect of Glutamates and Metal Ions on the Redox Potential of I107E Fe_BMb.

Since EPR and X-ray structural studies indicate metal binding to I107E Fe_BMb, we used spectroelectrochemistry to measure the effects of glutamates and metal ions on the heme reduction potential. When there is no metal ion in the Fe_B site, the I107E Fe_BMb displays a reduction potential of -134 ± 3 mV vs. the normal hydrogen electrode (NHE) (Fig. S1A), similar to that of Fe_BMb (-158 ± 4 mV) without the I107E mutation (14). In the presence of Cu²⁺, I107E Fe_BMb has a reduction potential (-137 ± 2 mV) (Fig. S1B) almost identical to that of the same protein in the absence of metal ions in the Fe_B site, indicating that copper binding to the Fe_B site has little effect on the reduction potential of the heme iron. This observation is similar to that observed for Cu²⁺ binding to Cu_BMb (Cu(II)-Cu_BMb, 80 mV vs. Cu_BMb, 77 mV) (54). On the other hand, the presence of Fe²⁺ and Zn²⁺ increased the reduction potential of I107E Fe_BMb from -134 ± 3 mV to -64 ± 3 mV vs. NHE (Fig. S1C) and -105 ± 2 mV vs. NHE (Fig. S1D), respectively. The different effects of Cu²⁺, Fe²⁺, and Zn²⁺ on the reduction potential of I107E-Fe_BMb indicate that these metal ions in the Fe_B site may play different roles through different coordination properties.

NOR Activity of I107E Fe_BMb in the Presence of Different Metal Ions.

The NO reduction activity of I107E Fe_BMb in the presence of Fe²⁺, Cu⁺, or Zn²⁺ was monitored by GC/MS under single turnover conditions. When Fe(II)-I107E Fe_BMb was exposed to excess NO, N₂O could be observed to form with increased yield over time (Fig. S2). Similarly, N₂O formation was observed for Cu(I)-I107E Fe_BMb, indicating that Fe or Cu binding to the Fe_B site results in comparable NOR activities. It should be noted that because of the high solubility of N₂O (∼25 mM in water at room temperature), GC/MS cannot be used to quantify the rates of NO reduction under these conditions. In contrast, no N₂O formation was observed with redox inactive Zn²⁺, which demonstrates that redox active Fe²⁺ or Cu⁺ in the Fe_B site plays a crucial role in NO reduction.

To gain deeper insight into the process of NO reduction, EPR studies were further performed to monitor the initial process of NO reduction. In the absence of metal ions, the EPR spectrum of ferrous I107E Fe_BMb-NO shows hyperfine splitting resulting from bound NO and the proximal histidine, indicating the formation of a six-coordinate ferrous heme-NO species (Fig. 5, top line). After incubation of Fe(II)-I107E Fe_BMb with excess NO, a distinct three-line hyperfine structure appears at 15 min (Fig. 5A), suggesting the formation of a five-coordinate ferrous heme-NO species as a result of cleavage of the proximal His-Fe heme bond (55). A three-line hyperfine structure was also observed for Cu⁺ and Zn²⁺, except that the signal intensity is lower with Cu(I)-I107E Fe_BMb-NO (Fig. 5B) and more pronounced in Zn(II)-I107E Fe_BMb-NO (Fig. 5C). The lower intensity of the three-line hyperfine structure for Cu(I)-I107E Fe_BMb-NO suggests the major species formed is a six-coordinate ferric heme-NO complex, which is EPR silent (41). These differences further suggest that the metal ion in the Fe_B site plays a key role in formation of the intermediates, thereby tuning NOR activity.

Fig. 5.

EPR spectra of deoxy I107E Fe_BMb (0.5 mM) with no metal bound in the presence of NO after 5 min (top line), with 2 eq Fe²⁺ (A), Cu⁺ (B), or Zn²⁺ (C) incubated with excess NO (∼200 eq) for 1, 5, and 15 min. Spectra were collected in 50 mM Bis-Tris pH 7.0 at 30 K, 0.2 mW power, and 9.05 GHz.

Discussion

Using Rationally Designed Proteins to Address Important Issues in Chemistry and Biology.

Important issues such as the roles of the conserved glutamates and nonheme Fe_B in NOR have been previously addressed using biochemical and biophysical studies or biomimetic modeling (24, 25, 27–37, 45, 56, 57). As a complementary approach, rational protein design, using small, easy-to-produce and well-characterized proteins such as myoglobin, offers a powerful method with which to gain insights into more complex native enzymes such as NOR (14). Similar to synthetic models (41, 43), the metal ion at the putative Fe_B site in the protein model can be substituted freely. Better yet, Glu residues can be placed at precise locations in the protein, including the secondary coordination sphere, due to its rigid network. By carefully choosing a suitable protein template, rational protein design could be generally applied to address other important issues in chemistry and biology.

The Roles of Glutamates.

Although two conserved glutamates (E198 and E267) are known to be crucial for NOR activity (24, 25), their roles are not well defined (18, 19). In a previous study (14), we demonstrated that one Glu, E68, is important for both iron binding and NOR activity of Fe_BMb. The crystal structures of both Fe(II)-Fe_BMb and Fe(II)-I107E Fe_BMb show that one O atom of E68 directly coordinates to Fe_B (Fig. 2). In synthetic models of NOR, it has also been found that the presence of a glutamic acid mimic significantly increases the stability of iron binding to the Fe_B site (40). Furthermore, a theoretical study by Blomberg et al. (58) showed that a model with an Fe_B coordinated by three histidines, one glutamate, and one water molecule provides an energetically feasible reaction mechanism of NO reduction. However, the structural model of NOR constructed recently by Reimann et al. (22) shows that the closest conserved Glu (E267) still has its carboxylate O atom 7 Å away from Fe_B, which suggests that Glu may not bind to Fe_B in native NOR. One interesting finding from our study is that the Glu (E68) underwent a significant conformational rearrangement in the presence of another Glu (E107) (Fig. 2). Therefore, the Fe_BMb provides a viable model of NOR that is consistent with Blomberg’s model, but cannot rule out Reimann’s model due to possible conformation changes.

While the role of the first Glu is still uncertain until a 3D structure of NOR in its active form is available, the role of the second Glu is even less defined. We address this question by introducing a second Glu (E107) to Fe_BMb. The crystal structures shown in Fig. 1 indicate that E107 interacts with a water molecule and forms a hydrogen bonding network in both Fe(II)-I107E Fe_BMb and Cu(II)-I107E Fe_BMb. Interestingly, although a similar water molecule was observed in the active site of Fe(II)-Fe_BMb (Fig. 2), activity assay data indicate that the presence of E107 in Fe(II)-I107E Fe_BMb increases NOR activity by ∼100% (Fig. 4), suggesting that the second Glu may potentionally play a role in providing one of the protons during reduction of NO to N₂O. Although free Glu outside a protein has a pK_a ∼ 4.3, the studies of native NOR showed that the pK_a of its Glu close to the active site has a value of ∼6.6 (22, 25, 33). The hydrogen binding network in our protein models may contribute to the fine-tuning of the Glu pK_a to be more neutral, similar to that in NOR. Moreover, it is interesting that Cu(I)-I107E Fe_BMb also shows NOR activity, which provides an interesting protein model of HCOs with NOR function (26–28), even though Glu residues are not conserved in native HCOs.

Additionally, spectroelectrochemical studies showed that the reduction potential of I107E Fe_BMb with no metal ion in the Fe_B site is similar to that of Fe_BMb, but much lower than that of Cu_BMb (77 mV) (54), which contains the same three His, but no Glu in the metal-binding site above the heme. Since both Fe_BMb and I107E Fe_BMb contain the V68E mutation that has been shown to decrease the heme reduction potential of native myoglobin from 59 mV to -137 mV (59), it is likely that the introduction of a negatively charged Glu close to the heme group is what is responsible for the dramatically lower heme redox potential. A conserved glutamate, predicted to be located near the catalytic heme b₃ in NOR, was proposed to be responsible for a ∼260-mV decrease in reduction potential (60 mV) in comparison to the other two heme centers, heme b (345 mV) and heme c (310 mV) (60). Our Fe_BMb and I107E Fe_BMb models mimic this feature of NOR. Notably, although introduction of the first Glu (E68) lowered the heme potential by ∼110 mV (14), introduction of the second Glu via the I107E mutation did not result in a significant difference in the heme reduction potential, suggesting that the effect of the two conserved Glu residues in NOR on heme reduction potential is not additive, with the effects highly dependent on the location of the Glu.

The Roles of Metal Ions.

The roles of metal ions in NOR are another important question as iron is found in the native Fe_B site and HCO employs copper at the corresponding Cu_B site. With different metal ions in the Fe_B site, the crystal structures clearly show the heme and nonheme dinuclear center existing in different local environments (Fig. 1). Although a similar hydrogen bond network is formed in both Fe(II)-I107E Fe_BMb and Cu(II)-I107E Fe_BMb, the conformation of E68 and E107 with respect to the nonheme metal center and heme iron is different from each other. Moreover, the coordination geometry differs significantly with Zn²⁺ in the Fe_B site. A hydrogen bond is absent from the Zn crystal structure, but both the O atoms of E68 act as metal-binding ligands. These observations demonstrate that the identity of the metal ion in the Fe_B site can tune the active site through their interactions with the His and Glu ligands, resulting in formation of different coordination geometries with different hydrogen bonds.

In addition to structural fine-tuning, the metal ion at the Fe_B site can also tune the heme iron reduction potential in I107E Fe_BMb. Spectroelectrochemical studies showed that the binding of Fe²⁺ or Zn²⁺ results in an increase in the heme reduction potential by ∼70 mV and ∼30 mV, respectively (Fig. S1). In the case of Cu(II)-I107E Fe_BMb, the crystal structure shows that OE1 of E68 is closer to the heme iron (2.07 Å) (Fig. 1C) than its metal-free form (2.15 Å) (Fig. 1A). The stronger interaction from the negatively charged E68 could offset the effect of positively charged Cu²⁺ binding, resulting in similar reduction potentials observed for Cu(II)-I107E Fe_BMb and I107E Fe_BMb.

In a previous study (61), EPR data showed that during NO reduction, the binding of Cu⁺ to the Cu_B site of Cu_BMb can weaken the proximal heme Fe-His bond, while complete cleavage of the heme Fe-His bond occurred when Zn²⁺ was bound to Cu_BMb-NO. In this study, we observed that a five-coordinate heme-NO species was formed with Fe²⁺, Cu⁺, or Zn²⁺ bound to the Fe_B site of I107E Fe_BMb (Fig. 5). Significantly, a five-coordinate heme-NO species has also been observed for both NOR (30, 31, 35) and the member of the HCO family with the highest NO reduction activity, cytochrome cbb₃ oxidases (26, 62). However, this species was not observed for Fe(II)-Fe_BMb-NO and Cu(I)-Fe_BMb-NO, which lack the second Glu (E107). In both these cases, the proximal heme Fe-His bond was only weakened, as indicated by a decrease of the nine-line hyperfine splitting signals in the EPR spectra (Fig. S3). These observations suggest that formation of a five-coordinate heme-NO species may play an important role in NOR reactivity.

Conclusions

We have successfully designed a structural and functional model of NOR, by introducing a second glutamate in the vicinity of the Fe_B site, named I107E Fe_BMb. This protein model mimics native NOR more closely by bearing the structural feature of three histidines and two glutamates in the Fe_B site, as predicted for native NOR. We have demonstrated that the two glutamates can play different roles in NO reduction activity; namely, one acts as a ligand to Fe_B (E68), and the other acts as a proton transfer group (E107). Furthermore, by substituting different metal ions into the nonheme metal site, we have demonstrated that Fe_B plays crucial roles in fine-tuning the active site by donating electrons and by mediating the formation of a five-coordinate heme-NO intermediate during NO reduction. In the absence of a crystal structure for native NOR, this study offers an ideal protein model and provides valuable structural as well as mechanistic information for native NOR.

Materials and Methods

Protein Preparation.

I107E Fe_BMb (swMb L29H/F43H/V68E/I107E) was constructed, expressed, and purified using the procedure described previously (14). The purity and identity were confirmed by SDS-PAGE and electrospray ionized MS: observed: 17,392 ± 1 Da; calculated: 17,391 Da.

EPR Spectroscopy.

EPR spectra were recorded on a Bruker ESP 300 equipped with an Oxford liquid helium cryostat and an ITC4 temperature controller. The samples of met I107E Fe_BMb, Cu(II)-, or Zn(II)-I107E Fe_BMb were prepared as described previously (14). The samples of NO-bound deoxy I107E Fe_BMb, Cu(I)-, Fe(II)-, or Zn(II)-I107E Fe_BMb were prepared by injecting 1 mL of purified NO gas into the EPR tube containing 300 μL of protein (0.5 mM, 10% glycerol, in 50 mM Bis-Tris pH 7.0). The samples were then flash frozen in liquid N₂ after incubation for 1, 5, or 15 min. The molar extinction coefficient of the Soret band of I107E Fe_BMb at 406 nm (175 mM^-1·cm^-1), calculated using the standard hemochromagen method (63), was used to determine protein concentration. The metal sources of Cu(I), Cu(II), Zn(II), and Fe(II) were [(CH₃CN)₄Cu]PF₆, CuSO₄, ZnSO₄·7H₂O, and FeCl₂, respectively.

Spectroelectrochemical Measurements.

Protein reduction potentials were measured using an optically transparent thin layer electrode as previously described (64). The potential of the working electrode was applied in the positive direction for deoxy I107E Fe_BMb with Fe²⁺ and in the negative direction for metal free and with Cu²⁺ or Zn²⁺. Other procedures are the same as described previously (54).

X-Ray Crystallographic Studies.

Fe(II)-I107E Fe_BMb was crystallized anaerobically in a glove box at room temperature using the conditions described for Fe(II)-Fe_BMb (14). I107E Fe_BMb, Cu(II)-, or Zn(II)-I107E Fe_BMb were crystallized aerobically. Diffraction-quality crystals were soaked in a cryoprotectant solution of 30% PEG 400 and flash frozen in liquid nitrogen. Diffraction data were collected at the Brookhaven National Lab Synchrotron Light Source X12C beamline. The crystal structure was solved using the same method as for Fe(II)-Fe_BMb (14).

NOR Activity Assay.

NO reduction was monitored by GC/MS. The protein was reduced to the deoxy form by excess dithionite that was removed with a size-exclusion column (PD-10). Then 2 eq metal, Cu(I), Fe(II), or Zn(II), was added to the protein solution (0.6 mM, 3 mL in 50 mM Bis-Tris buffer, pH 7.0). The samples were prepared anaerobically in a glove box. Purified NO gas was injected into the head space of the reaction flask with the molar ratio of NO∶protein ≈ 50∶1. Other procedures are the same as described previously (14, 61).