• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Apr 1, 2003; 100(7): 3796–3801.
Published online Mar 24, 2003. doi:  10.1073/pnas.0636858100
PMCID: PMC153001
Bioinorganic Chemistry Special Feature
Biochemistry

Nitric oxide binding at the mononuclear active site of reduced Pyrococcus furiosus superoxide reductase

Abstract

Nitric oxide (NO) has been used as a substrate analog to explore the structural and electronic determinants of enzymatic superoxide reduction at the mononuclear iron active site of Pyrococcus furiosus superoxide reductase (SOR) through the use of EPR, resonance Raman, Fourier transform IR, UV-visible absorption, and variable-temperature variable-field magnetic CD spectroscopies. The NO adduct of reduced SOR is shown to have a near-axial S = 3/2 ground state with E/D = 0.06 and D = 12 ± 2 cm−1 (where D and E are the axial and rhombic zero-field splitting parameters, respectively) and the UV-visible absorption and magnetic CD spectra are dominated by an out-of-plane NO(π*)-to-Fe3+(dπ) charge–transfer transition, polarized along the zero-field splitting axis. Resonance Raman studies indicate that the NO adduct is six-coordinate with NO ligated in a bent conformation trans to the cysteinyl S, as evidenced by the identification of ν(N–O) at 1,721 cm−1, ν(Fe–NO) at 475 cm−1, and ν(Fe–S(Cys), at 291 cm−1, via 34S and 15NO isotope shifts. The electronic and vibrational properties of the S = 3/2 {FeNO}7 unit are rationalized in terms of a limiting formulation involving a high-spin (S = 5/2) Fe3+ center antiferromagnetically coupled to a (S = 1) NO anion, with a highly covalent Fe3+−NO interaction. The results support a catalytic mechanism for SOR, with the first step involving oxidative addition of superoxide to form a ferric-peroxo intermediate, and indicate the important roles that the Fe spin state and the trans cysteinate ligand play in effecting superoxide reduction and peroxide release.

Over the past 4 yr, evidence has accumulated for a novel pathway for detoxification of reactive oxygen species that is specific for anaerobic and microaerophilic microorganisms (13). The key enzyme in this pathway is superoxide reductase (SOR), which catalyzes the reduction of superoxide to hydrogen peroxide (4), with rubredoxin as the probable physiological electron donor (2, 5). Although all SORs have a β-barrel domain containing a unique type of mononuclear Fe center that serves as the site for superoxide reduction (6, 7), some have an additional N-terminal domain that contains an intrinsic rubredoxin-like mononuclear Fe site, which is ligated by four cysteines in a distorted tetrahedral arrangement analogous to that found in desulforedoxin (6). These two classes are conveniently referred to as 1Fe- and 2Fe-SORs, but they are also known by their trivial names, neelaredoxins and desulfoferrodoxins, respectively.

Spectroscopic and crystallographic studies of the 1Fe-SOR from Pyrococcus furiosus (79) and the 2Fe-SOR from Desulfovibrio desulfuricans (6, 10, 11) have shown that the mononuclear iron active site is ligated by four equatorial histidines (3epsilonN and 1δN) and one axial cysteinate in a square-pyramidal arrangement. The sixth coordination site appears to be occupied by a monodentate glutamate in the oxidized state but is vacant or occupied by a weakly coordinated water molecule in the reduced state, thereby providing a site for superoxide binding and reduction. These structural studies, combined with recent mutagenesis and pulse radiolysis kinetic results (1216), have led to the proposal for the catalytic mechanism shown in Fig. Fig.1.1. The key steps involve oxidative addition of superoxide to the ferrous form of SOR to yield a ferric-peroxo intermediate, protonation possibly by a conserved lysine residue in the reduced active-site pocket to yield a ferric-hydroperoxo intermediate, and dissociation of the ferric-hydroperoxo intermediate, via Fe–O bond cleavage coupled with glutamate binding, to yield the H2O2 product. This mechanism is further supported by the recent characterization of a transient ferric-peroxo species in peroxide-treated samples of a variant of Desulfoarculus baarsii in which the active-site glutamate is replaced by alanine (17).

Figure 1
Proposed catalytic mechanism of SOR.

Characterization of enzymatic intermediates is required for both detailed understanding of the mechanism of SOR and addressing the key questions of how and why the SOR active site preferentially catalyzes reduction rather than dismutation of superoxide. However, these intermediates are short lived and difficult to study experimentally. Nitric oxide (NO) has been extensively used as a substrate analog of molecular oxygen to form stable nitrosyl derivatives that provide insight into oxygen transport and activation intermediates in many heme (1821) and nonheme (2229) iron enzymes. Hence, we report here the formation and spectroscopic characterization of a stable NO-bound derivative of the reduced 1Fe-SOR from P. furiosus using the combination of EPR, UV-visible absorption, and variable-temperature, variable-field magnetic CD (VTVH MCD), resonance Raman, and Fourier transform IR (FTIR) spectroscopies. The structural and electronic characterization of the NO adduct of SOR facilitates understanding of how the active site is tuned for oxidative addition of superoxide and release rather than intraligand cleavage of the peroxide product.

Materials and Methods

Biochemical Techniques and Sample Preparation.

Recombinant natural abundance and 34S globally enriched P. furiosus SOR was over-expressed in Escherichia coli and purified to homogeneity according to published procedures (8, 9). Unless otherwise indicated, the samples of P. furiosus SOR used in this work were in 100 mM Hepes buffer at pH 7.5 and were reduced by using 10 mM ascorbate. NO derivatives were prepared under strictly anaerobic conditions by adding an NO generator, methylamine hexamethylene methylamine (MAHMA) NONOate (Cayman Chemical, Ann Arbor, MI) or by exchanging the headspace above the SOR solution with NO gas. 14NO gas (>98%, Aldrich) and 15NO gas (99%, ICON Isotopes, Summit, NJ) were further purified before use by bubbling through a concentrated KOH solution to remove NO2. No changes in pH were detected on addition of MAHMA NONOate or exposure to NO gas. All samples were handled in a Vacuum Atmospheres (Hawthorne, CA) glove box (<1 ppm O2).

Spectroscopic Methods.

X-band (≈9.6 GHz) EPR spectra were recorded on a Bruker (Billerica, MA) ESP-300E EPR spectrometer with a dual-mode ER-4116 cavity and equipped with an Oxford ESR-9 flow cryostat (4.2–300 K). Spin quantitations were assessed as previously described (8), and EPR simulations were carried out by using the simfonia software package (Bruker). Resonance Raman spectra were recorded as described (9) on samples frozen at 17 K by using a scanning Instruments SA (Edison, NJ) Ramanor U1000 spectrometer fitted with a cooled RCA-31034 photomultiplier tube by using lines from a Coherent Sabre 100 10-W Argon Ion laser (Coherent Radiation, Palo Alto, CA). FTIR measurements were recorded by using a Bio-Rad FTS 575C spectrometer. Absorption spectra were recorded on Shimadzu UV301PC spectrophotometer. VTVH MCD measurements were recorded on samples containing 55% (vol/vol) glycerol by using a Jasco (Easton, MD) J-715 (180–1,000 nm) spectropolarimeter mated to an Oxford Instruments Spectromag 4000 (0–7 T) split-coil superconducting magnet. VTVH MCD saturation magnetization data of the SOR S = 3/2 iron-nitrosyl species were analyzed according to the published procedure (30), by using software supplied by Frank Neese and Edward I. Solomon (Stanford University, Stanford, CA).

Results

EPR.

No X-band EPR signals were observed for ascorbate-reduced samples of P. furiosus SOR, but intense resonances corresponding to two overlapping S = 3/2 species were observed in samples treated with NO generator, methylamine hexamethylene methylamine NONOate. For samples containing NO and SOR in an approximate 1:1 stoichiometry (Fig. (Fig.2),2), the major component is a rhombic resonance with g = 4.34, 3.76, 2.00 that is readily rationalized in terms of a conventional spin Hamiltonian as originating from the lower Ms = ±1/2 doublet of an S = 3/2 ground state with E/D = 0.06 and D > 0 parameters, where D and E are the axial and rhombic zero-field splitting parameters, respectively. The minor resonance exhibits a more axial resonance with g = 4.07, 4.00, 2.00 that originates from the lower Ms = ±1/2 doublet of an S = 3/2 ground state with E/D = 0.01 and D > 0. No resonance from the upper Ms = ±3/2 doublet was observed over the temperature range 4–60 K. Consequently, estimates of the separation between the Ms = ±1/2 and ±3/2 doublets (2D) were obtained by fitting plots of the EPR intensity vs. 1/T to a Boltzmann population distribution over a two-level system: D = 12 ± 2 cm−1 for the E/D = 0.06 component and D = 5 ± 1 cm−1 for the E/D = 0.01 component. Quantitation vs. a CuEDTA standard indicates that the entire resonance corresponds to 20% of the Fe containing SOR, and simulation of the spectra as the sum of two overlapping resonances revealed that the E/D = 0.06 and 0.01 species account for 84% and 16%, respectively. The substoichiometric spin quantitation is attributed to reversible binding of NO, because both resonances are progressively lost on repeated exchange of the headspace gas with Ar with continuous stirring. Moreover, redox cycling and exposure to up to a 3-fold excess of NO did not result in any significant loss of Fe from P. furiosus SOR, as judged by quantitative restoration of the oxidized absorption spectrum (8), after ascorbate reduction, NO treatment, and reoxidation with hexachloroiridate.

Figure 2
X-band EPR spectrum of the NO adduct of ascorbate-reduced P. furiosus SOR. The upper line represents the experimental spectrum. The lower line represents the simulated spectrum corresponding to 84% of a resonance with g1,2,3 = 4.343, 3.763, 1.997, ...

The origin of the heterogeneity in the ground-state properties of the NO adduct of reduced SOR was investigated by EPR studies as a function of pH, exposure to light, and the NO:SOR stoichiometry. The ratio of the E/D = 0.01 and 0.06 S = 3/2 species was not significantly perturbed by varying the pH over the range 5.5–9.0, and both resonances were lost at pH >9.5. In addition, neither resonance exhibited significant photolability, as evidenced by EPR studies at 10 K, while irradiating the sample in the cavity with a Xe arc lamp. However, the ratio of the intensities of the E/D = 0.01 and 0.06 S = 3/2 species was found to depend on the stoichiometry of NO to SOR. As the stoichiometry was increased from 0.3 to 3.0, the spin quantitation of the entire S = 3/2 resonance increased from 15% to 30% of the Fe-containing SOR, and the resonance changed from being almost exclusively that of the E/D = 0.06 species to being a 60:40 mixture of the E/D = 0.06 and 0.01 species.

Analogous near-axial S = 3/2 EPR signals to those observed for SOR have been observed for the nitrosyl adducts of a wide range of reduced mononuclear nonheme enzymes (2227) and ferrous nitrosyl inorganic complexes (28, 31, 32). Under Enemark/Feltham notation (33), these species are conveniently represented as {FeNO}7, which treats the iron-nitrosyl as a bracketed unit with a superscript to represent the total number of Fe d and NO π* electrons. Detailed spectroscopic studies and electronic structure calculations for complexes containing S = 3/2 {FeNO}7 units (31, 32) have shown that the ground- and excited-state properties are best rationalized in terms of a limiting formulation involving high-spin Fe3+ (S = 5/2) antiferromagnetically coupled to NO (S = 1). The EPR results for the nitrosyl adduct of ferrous SOR, therefore, support oxidative addition of NO to yield species with substantial Fe3+-NO character.

Absorption and VTVH MCD.

The room temperature absorption and 4.2 K MCD spectra for the NO adduct of ascorbate-reduced P. furiosus SOR are shown in Fig. Fig.3.3. The NO-treated samples were prepared by exposing samples containing 55% (vol/vol) glycerol to an atmosphere of NO for 30 sec before transfer under Ar into sealed cuvettes for the spectroscopic measurements. Parallel EPR studies on identical samples frozen at the same time as MCD samples exhibited EPR spectra very similar to those shown in Fig. Fig.2,2, with the total S = 3/2 resonance accounting for 20% of the Fe containing SOR and 90% of this resonance arising from the E/D = 0.06 component.

Figure 3
UV-visible absorption and MCD spectra of the NO adduct of ascorbate-reduced P. furiosus SOR. (Upper) Room-temperature absorption spectra. (Lower) MCD spectra at 4.22 K with an applied magnetic field of 6 T. The sample was 0.9 mM in SOR, the buffering ...

Ascorbate-reduced SOR is colorless and has no significant absorption or MCD bands in the 14,000–29,000 cm−1 region. On addition of NO, the solution turns a red–brown color, and the absorption spectrum is dominated by a well-resolved band centered at 21,050 cm−1 (475 nm; epsilon = 530 M−1[center dot]cm−1) and a shoulder centered near 28,000 cm−1 (357 nm; epsilon ≈ 900 M−1[center dot]cm−1; see Fig. Fig.3).3). Analogous absorption bands have been reported for the S = 3/2 NO adducts of a wide range of nonheme proteins and inorganic complexes (22, 28, 31, 34, 35) and have been attributed to charge transfer (CT) transitions associated with the{FeNO}7 unit. The 4.2 K MCD spectrum of the NO adduct of ascorbate-reduced P. furiosus SOR in the 14,000 – 29,000 cm−1 region correlates closely with the absorption spectrum and is dominated by a positive MCD C term centered at 21,050 cm−1, with less pronounced positive C terms apparent at 28,000 cm−1 (357 cm−1) and as a shoulder at ≈23,500 cm−1 (425 nm; see Fig. Fig.3).3). The negative C term centered at 12,400 cm−1 that is present in the samples before and after NO treatment corresponds to the 5T2g (dxy) → 5Eg (dx2−y2) ligand field transition of the square-pyramidal high-spin Fe2+ center in unreacted reduced SOR. Evidence for unreacted reduced SOR is also apparent by the characteristic pattern of (Cys)S-to-Fe2+ CT bands (8) that are observed in the low-temperature MCD spectrum in the 29,000–40,000 cm−1 region (data not shown). In agreement with EPR spin quantitations, which indicated that only 20% of the reduced SOR is in the NO-bound form, the intensity of (Cys)S-to-Fe2+ CT MCD bands in the NO-treated sample was ≈80% of that observed for samples before NO treatment.

Well-resolved absorption bands centered between 20,000 and 24,000 cm−1 that give rise to positive MCD C-terms centered at the same energies are a unifying attribute of the absorption and MCD spectra of all S = 3/2 {FeNO}7 species investigated thus far (28, 31, 34). This transition has been assigned to the out-of-plane NO(π*)-to-Fe3+(dπ) CT within the idealized Fe3+-NO unit, because these orbitals are predicted to have optimal overlap based on SCF-Xα-SW calculations (31). To test the validity of the proposed assignment, the transition polarization was assessed via VTVH MCD saturation magnetization studies on the MCD band centered at 21,050 cm−1 (475 nm) (see Fig. 7, which is published as supporting information on the PNAS web site, www.pnas.org). The results indicate a uniaxial transition polarized along the unique (z) axis of the predominantly axial zero-field splitting. This is in complete accord with assignment to an out-of-plane NO(π*)-to-Fe3+(dπ) CT transition, and the MCD magnetization data indicate that the Fe–NO bond is collinear with the zero-field splitting axis.

Resonance Raman and FTIR.

To assign vibrational bands based on 15NO isotope shifts, samples for resonance Raman and FTIR studies were prepared by incubating concentrated samples of ascorbate-reduced SOR, 4–5 mM, under an atmosphere of 14NO or 15NO gas. Parallel EPR studies of the samples used for resonance Raman and FTIR investigations showed resonances that were almost exclusively (>95%) from the S = 3/2 species with E/D = 0.06.

In accord with the assignment of the absorption band at 21,050 cm−1 (475 nm) to the out-of-plane NO(π*)-to-Fe3+(dπ) CT transition, resonance Raman studies using 476-nm excitation revealed enhancement of vibrational modes associated with the {FeNO}7 unit (see Fig. Fig.4).4). The high-frequency intraligand stretching region shows a band that shifts from 1,721 cm−1 in the 14NO adduct to a shoulder at 1,690 cm−1 in the 15NO adduct, in addition to modes at 1,552, 1,616, and 1,668 cm−1 that show no 15NO/14NO shift and are assigned to vibrations of the polypeptide backbone (9). On the basis of the 31-cm−1 isotope shift, the 1,721 cm−1 band is assigned to the ν(N–O) stretching mode, which is predicted to exhibit a 31-cm−1 14N/15N isotope shift based on a simple diatomic oscillator approximation. A band with a 32-cm−1 15NO/14NO shift was also observed in the FTIR spectrum at 1,728 cm−1 (see Fig. 8, which is published as supporting information on the PNAS web site). The difference in the frequencies of the ν(N–O) stretching mode as determined by resonance Raman (1,721 cm−1) and FTIR (1,728 cm−1) is attributed to the measurement conditions, i.e., frozen solution at 17 K for Raman and aqueous solution at room temperature for FTIR.

Figure 4
Resonance Raman spectra of the 14NO and 15NO adducts of ascorbate-reduced P. furiosus SOR in the N–O and Fe–N stretching regions. The samples were ≈4 mM in SOR and were prepared by reduction with 10 mM ascorbate followed by incubation ...

The low-frequency metal-ligand stretching region shows one 15NO/14NO-sensitive vibration that shifts from 475 cm−1 in the 14NO adduct to 468 cm−1 in the 15NO adduct (Fig. (Fig.4).4). Both ν(Fe–NO) stretching and δ(Fe–N–O) bending modes are expected in this region, and these modes are likely to be extensively mixed. Nevertheless, several lines of evidence favor assignment of the band at 475 cm−1 to a mode that predominantly involves ν(Fe–NO) stretching. First, in the absence of a specific enhancement mechanism, bending modes generally exhibit weaker resonance enhancement than stretching modes and hence are weaker in resonance Raman spectra. This general observation is borne out by resonance Raman studies of nitrosyl hemes in which the ν(Fe–NO) stretching and δ(Fe–N–O) bending modes have been rigorously assigned based on 15NO, N18O, and 15N18O isotope shifts and normal mode calculations (1921). The ν(Fe–NO) stretching modes are invariably more strongly enhanced than the δ(Fe–N–O) bending modes, and the enhancement of the bending mode generally decreases as the Fe–N–O angle approaches 180°. Second, the observed 15N downshift, 7 cm−1, is consistent with a bent FeNO unit with an angle closer to 180° than to 90°. A ν(Fe–NO) stretching mode at 475 cm−1 is predicted to have a 15N downshift in the range 5–13 cm−1 based on a simple two-body diatomic oscillator approximation, with the smallest shift predicted for Fe–N–O bond angles close to 180°, and the largest shift predicted for Fe–N–O bond angles close to 90°. The range of 15N downshifts for ν(Fe–NO) stretching modes predicted for this simple model is generally in good agreement with experimental data obtained for well characterized nitrosyl hemes, i.e., downshifts as low as 2 cm−1 for linear FeNO and as high as 14 cm−1 for bent FeNO with an Fe–N–O angle of 115° (1921).

In addition to the ν(Fe–NO) and ν(N–O) stretching modes, vibrational modes associated with the Fe-S(Cys) unit are also weakly enhanced in resonance with absorption band at 21,050 cm−1 (475 nm). A comparison of the resonance Raman spectra obtained for the NO adducts of natural abundance and 34S globally labeled SOR by using 476-nm excitation revealed two 34S-sensitive bands, located at 291 cm−1 and 758 cm−1, that exhibit 3-cm−1 downshifts in the 34S-labeled samples (see Fig. Fig.5).5). These two bands are readily assigned to the ν(Fe–S) and ν(S–C) stretching modes of the Fe–S(Cys) unit, respectively, on the basis of previous assignments for oxidized P. furiosus SOR (9). Enhancement of vibrations associated with the (Cys)S–Fe–NO unit was observed with 476-nm excitation, but not with 647-nm excitation [the optimal wavelength for enhancing vibrations of the Fe–S(Cys) unit in oxidized P. furiosus SOR (9)] (see Fig. Fig.5).5). Hence the observed vibrational modes associated with the Fe–S(Cys) unit are not the result of partial oxidation. Rather, they are enhanced via kinematic coupling with the ν(Fe–NO) stretching mode and/or an excited-state A-term mechanism involving the competing π interactions within the trans (Cys)S–Fe–NO unit. Irrespective of the enhancement mechanism, the observation of vibrational modes associated with the Fe–S(Cys) unit demonstrates that cysteinate is ligated trans to NO in the NO adduct of SOR that is responsible for the S = 3/2, E/D = 0.06 species.

Figure 5
Resonance Raman spectra of the NO adduct of natural abundance (NA) and 34S globally labeled samples of ascorbate-reduced P. furiosus SOR in the low-frequency region. (A) NO adduct of NA SOR by using 476-nm excitation. (B) NO adduct 34S globally labeled ...

Discussion

Interaction of ascorbate-reduced P. furiosus SOR with NO has been shown to result in the reversible formation of a stable six-coordinate derivative of the mononuclear Fe active site with NO bound trans to the cysteinate ligand. The ground- and excited-state electronic properties of this NO adduct, i.e., near-axial S = 3/2 ground-state with zero-field splitting parameters E/D = 0.06 and D = 12 ± 2 cm−1 and visible absorption and low-temperature MCD spectra dominated by a well-resolved band centered at 21,050 cm−1 (475 nm), are characteristic of nonheme iron proteins and inorganic complexes containing S = 3/2 {FeNO}7 units (22, 31, 34). Moreover, on the basis of the detailed spectroscopic and electronic structure calculations that are available for inorganic complexes containing S = 3/2 {FeNO}7 units (31, 32), the ground- and excited-state properties are best rationalized in terms of a limiting formulation involving a high-spin (S = 5/2) Fe3+ center antiferromagnetically coupled to a (S = 1) NO anion. The resulting Fe3+–NO interaction is highly covalent as a result of strong NO σ- and π-donor bonding to Fe3+, and the strong π overlap is manifest by the visible absorption band at 21,050 cm−1, which is attributed to the out-of-plane NO(π*)-to-Fe3+(dπ) CT transition that is shown to be polarized along the unique (z) axis of the predominantly axial zero-field splitting via VTVH MCD studies.

The resonance Raman and FTIR studies of the NO adduct of reduced SOR reported herein constitute the first vibrational characterization, to our knowledge, of a S = 3/2 {FeNO}7 unit in a mononuclear nonheme iron enzyme and provide assessment of the structure and bonding within the trans (Cys)S–Fe–NO unit. Using excitation at 476 nm into the out-of-plane NO(π*)-to-Fe3+(dπ) CT transition, resonance Raman studies have identified the ν(N–O) stretching mode at 1,721 cm−1 and ν(Fe–NO) stretching mode at 475 cm−1, on the basis of 15NO downshifts of 31 and 7 cm−1, respectively. In accord with the limiting Fe3+–NO formulation, the frequency of the intraligand ν(N–O) stretching mode is substantially reduced compared with NO gas, ν(N–O) at 1,876 cm−1 (36), as a result of electron transfer into the NO π* orbital. The only other vibrational data for a biological {FeNO}7 unit come from resonance Raman studies of the NO adduct of deoxyhemerythrin by using 647-nm excitation. Although the ν(N–O) stretching mode was not observed, a band at 433 cm−1 and a weak shoulder at 421 cm−1 were assigned to the ν(Fe–NO) and δ(Fe–N–O) modes, respectively, of the {FeNO}7 unit, based on 6–7 cm−1 15NO and N18O downshifts (29). The frequencies of the ν(Fe–NO) modes dictate a substantially weaker Fe–NO bond in hemerythrin than in SOR, and this is likely to be a consequence of a strong H-bonding interaction between the bent FeNO and the μ-hydroxo bridge of the diiron center (29).

Greater insight into the structure of the {FeNO}7 unit in the NO adduct of SOR is provided by comparison with the resonance Raman data for the S = 3/2 {FeNO}7 unit in the inorganic complexes, Fe(L)(NO)(N3)2 (L is N,N′,N′′-trimethyl-1,4,7-triazacyclonane) and Fe(EDTA)(NO), which have Fe–N–O bond angles of 156° (31, 37). Both complexes exhibit resonance Raman spectra similar to that of the NO adduct of SOR with bands at 1,712 and 1,776 cm−1, respectively, that were identified as ν(N–O) stretching modes based on 15NO downshifts of 31–32 cm−1, and strong bands at 497 and 496 cm−1, respectively, with downshifts of 6–8 cm−1 (31). However, the bands at 497 and 496 cm−1 were assigned to δ(Fe–N–O) bending rather ν(Fe–NO) stretching (31) on the basis of IR studies of metal nitrosyl complexes, which indicated larger 15NO isotope shifts for δ(M–N–O) modes (6–15 cm−1) than for ν(M–NO) modes (1–6 cm−1) (38, 39). Subsequent resonance Raman studies of nitrosyl hemes (1921) have clearly shown that these ranges are applicable only to near-linear MNO units. Moreover, the pattern of isotope shifts observed for the 497 and 496 cm−1 bands in these complexes, i.e., larger isotope shift for 15NO than N18O (31), is a characteristic of the ν(M–NO) modes of bent FeNO units in nitrosyl hemes (20). Taken together with the strong resonance enhancement of these bands, reassignment to modes predominantly involving ν(Fe–NO) stretching is clearly warranted.

The structure of the FeNO unit in the six-coordinate nitrosyl adduct of SOR is likely to be very similar to that found in the structurally characterized six-coordinate Fe(L)(NO)(N3)2 complex (Fe–N–O angle = 156°), on the basis of similar frequencies for the ν(N–O) and ν(Fe–NO) modes. The most significant difference is a weaker Fe–NO bond in SOR, as evidenced by the 21-cm−1 decrease in the frequency of the ν(Fe–NO) mode. This is attributed to the mutual trans influence of the cysteinate and NO ligands, which compete for the same set of Fe d-orbitals for both σ- and π-bonding interactions (see Fig. Fig.6).6). The principal bonding interactions within the {FeNO}7 involve π-donor interactions involving the half-filled π* orbitals on NO and the half-filled Fe3+ dxz and dyz orbitals and the σ-donor interaction between the filled σ* orbital on NO and the half-field Fe3+ dz2 orbital (31). For the Fe-S(Cys) fragment, the principal bonding interactions involve π-donor interactions involving the filled S pπ and ppseudo σ orbitals and the half-filled Fe3+ dxz and dyz orbitals and the σ-donor interaction between the filled S ppseudo σ orbital and the half-field Fe3+ dz2 orbital (8). Evidence for the weakening of the Fe–S(Cys) bond in the NO adducts of SOR relative to the six-coordinate high-spin Fe3+ site in the oxidized enzyme comes from the 32-cm−1 decrease in the frequency of the vibrational mode that primarily involves Fe-S(Cys) stretching [323 cm−1 in oxidized native SOR (9) compare with 291 cm−1 in the NO adducts of SOR]. The trans influence of a cysteinate ligand on the Fe–NO bonding interaction is also evident in the resonance Raman studies of six-coordinate nitrosyl hemes containing S = 1/2 {FeNO}7 units. Nitrosyl heme proteins with cysteinate trans to the bound NO exhibit ν(Fe–NO) frequencies 10–20 cm−1 lower than those with histidyl ligation.

Figure 6
Schematic depiction of the principal in-plane (Upper) and out-of-plane (Lower) bonding interactions for trans (Cys)S–Fe–NO with an S = 3/2 {FeNO}7 unit.

Thus far, it has not been possible to characterize the optical and vibrational properties of the S = 3/2 {FeNO}7 species with E/D = 0.01 that is progressively formed as the NO-to-SOR stoichiometry increases. Consequently, the origin of the heterogeneity in the NO-bound forms of SOR that is apparent in EPR studies is unclear at present. Mutagenesis experiments involving the glutamate and lysine residues that are in close proximity to the substrate-binding site are currently in progress to assess whether the observed heterogeneity is a consequence of differences in hydrogen-bonding interactions. An alternative possibility is that the more axial E/D = 0.01 species corresponds to a five-coordinate {FeNO}7 species in which the Fe-S(Cys) bond has been cleaved. Similar near-axial S = 3/2 {FeNO}7 EPR signals have been observed in several nonheme Fe proteins before binding substrate or substrate analogs to form six-coordinate S = 3/2 {FeNO}7 species, which exhibit more rhombic resonances similar to the E/D = 0.06 species observed in this work (22, 27). In addition, the trans influence of NO binding is responsible for cleavage of weak proximal Fe–His bonds in many ferrous heme proteins, particularly those with regulatory roles involving the sensing of small molecules such as NO (soluble guanylate cyclase), O2 (FixL), and CO (CooA) (40). Reversible NO-induced cleavage of the Fe–S(Cys) bond in SOR, coupled with reversible nitrosylation of the free thiolate, could therefore be invoked to explain the dependence of the heterogeneity on the NO concentration.

Comparison of the ν(N–O) and ν(Fe–NO) frequencies for the six-coordinate S = 3/2 {FeNO}7 species in the NO adduct of SOR and the Fe(L)(NO)(N3)2 complex with those of six-coordinate S = 1/2 in heme proteins reveals dramatic differences in the bonding in S = 1/2 and 3/2 {FeNO}7 species. Six-coordinate S = 1/2 {FeNO}7 species in heme proteins have much stronger Fe–NO bonds, as evidenced by ν(Fe–NO) frequencies in the range 536–558 cm−1 (20, 21, 41), and much weaker N–O bonds, as evidenced by ν(N–O) frequencies in the range 1,555–1,624 cm−1 (20, 21). Although no clear consensus has emerged concerning the most appropriate description of the electronic structure and bonding in S = 1/2 {FeNO}7 species (18, 32, 42), the increase in Fe–NO bond strength and concomitant decrease in the N–O bond strength, compared with S = 3/2 {FeNO}7 species, appear to be a direct consequence of increased σ-donation into the empty dz2 orbital that results from lowering the Fe spin state from high to low or intermediate spin.

Finally, we address the implications of the spectroscopic characterization of the NO adduct of SOR for understanding how the mononuclear Fe active site of SOR is tuned for superoxide reduction. Clearly, the formation of the NO adducts of reduced SOR by oxidative addition of NO at the vacant coordination site of the ferrous active site is consistent with an inner-sphere mechanism for superoxide reduction, with the first step involving oxidative addition of superoxide to yield a ferric-peroxo intermediate. Moreover, the oxidative addition of π-bonded diatomics is enhanced by the electron-donating ability of the trans cysteinate ligand. Absorption and VTVH MCD studies of the square-pyramidal ferrous active site have shown intense (Cys)S-to-Fe2+ CT bands (8), indicative of strong pπ-dπ S–Fe bonding. The resulting increased electron density in the Fe dπ orbitals therefore facilitates ligand reduction via Fe dπ to ligand π* electron transfer. The S = 3/2 {FeNO}7 species in SOR is proposed as a stable analog of the S = 5/2 {FeOO}9 high-spin ferric (hydro)peroxo intermediate, albeit with one rather than two electrons in each of the π* orbitals (see Fig. Fig.6).6). Filled π* orbitals, coupled with the trans influence of the cysteinate, would be expected to weaken the Fe–OO(H) bond relative to the Fe–NO bond as a result of decreased π*–dπ bonding. This accounts for the instability of the ferric (hydro)peroxo intermediate with respect to cleavage of the Fe–OO(H) bond. On the basis of the differences in the bonding that are apparent from the vibrational studies of S = 1/2 and S = 3/2 {FeNO}7, a high-spin ferric (hydro)peroxo intermediate is likely to be essential to promote breakdown via Fe–O rather O–O bond cleavage. This is in accord with recent studies by the Solomon and Que groups, which have shown that high-spin Fe3+-(alkyl)peroxo have much weaker Fe–O bonds and much stronger O–O bonds than their low-spin counterparts (43, 44). Overall, characterization of the NO adducts of SOR indicates that the Fe spin state and the trans cysteinate ligand play important roles in effecting superoxide reduction and peroxide release at the SOR active site. By providing a stable analog of the ferric (hydro)peroxo intermediate, the NO adduct of SOR is likely to be very useful for probing H-bonding interactions in the active-site pocket.

Supplementary Material

Supporting Figures:

Acknowledgments

This work was supported by a grant from the National Institutes of Health (GM60326 to M.W.W.A. and M.K.J.) and a National Science Foundation Research Training Group Award (DBI9413236) to the Center for Metalloenzyme Studies.

Abbreviations

SOR
superoxide reductase
VTVH MCD
variable-temperature variable-field magnetic CD
CT
charge transfer
FTIR
Fourier transform IR

Footnotes

This paper was submitted directly (Track II) to the PNAS office.

References

1. Liochev S I, Fridovich I. J Biol Chem. 1997;272:25573–25575. [PubMed]
2. Jenney F E, Verhagen M F J M, Cui X, Adams M W W. Science. 1999;286:306–309. [PubMed]
3. Lombard M, Fontecave M, Touati D, Nivière V. J Biol Chem. 2000;275:115–121. [PubMed]
4. Adams M W W, Jenney F E, Jr, Clay M D, Johnson M K. J Biol Inorg Chem. 2002;7:647–652. [PubMed]
5. Coulter E D, Kurtz D M., Jr Arch Biochem Biophys. 2001;394:76–86. [PubMed]
6. Coelho A V, Matias P, Fülõp V, Thompson A, Gonzalez A, Carrondo M A. J Biol Inorg Chem. 1997;2:680–689.
7. Yeh A P, Hu Y, Jenney F E, Adams M W W, Rees D C. Biochemistry. 2000;39:2499–2508. [PubMed]
8. Clay M D, Jenney F E, Jr, Hagedoorn P L, George G N, Adams M W W, Johnson M K. J Am Chem Soc. 2002;124:788–805. [PubMed]
9. Clay M D, Jenney F E, Jr, Noh H J, Hagedoorn P L, Adams M W W, Johnson M K. Biochemistry. 2002;41:9833–9841. [PubMed]
10. Tavares P, Ravi N, Moura J J G, LeGall J, Huang Y-H, Crouse B R, Johnson M K, Huynh B H, Moura I. J Biol Chem. 1994;269:10504–10510. [PubMed]
11. Moura I, Tavares P, Moura J J G, Ravi N, Huynh B H, Liu M -Y, LeGall J. J Biol Chem. 1990;265:21596–21602. [PubMed]
12. Coulter E D, Emerson J P, Kurtz D M, Jr, Cabelli D E. J Am Chem Soc. 2000;122:11555–11556.
13. Nivière V, Lombard M, Fontecave M, Houée-Levin C. FEBS Lett. 2001;497:171–173. [PubMed]
14. Emerson J P, Coulter E D, Cabelli D E, Phillips R S, Kurtz D M., Jr Biochemistry. 2002;41:4348–4357. [PubMed]
15. Lombard M, Houée-Levin C, Touati D, Fontecave M, Nivière V. Biochemistry. 2001;40:5032–5040. [PubMed]
16. Abreu I A, Saraiva L M, Soares C M, Teixeira M, Cabelli D E. J Biol Chem. 2001;276:38995–39001. [PubMed]
17. Mathé C, Mattioli T A, Horner O, Lombard M, Latour J-M, Fontecave M, Nivière V. J Am Chem Soc. 2002;124:4966–4967. [PubMed]
18. Scheidt W R, Ellison M K. Acc Chem Res. 1999;32:350–359.
19. Hu S, Kincaid J R. J Am Chem Soc. 1991;113:2843–2850.
20. Hu S, Kincaid J R. J Am Chem Soc. 1991;113:9760–9766.
21. Tomita T, Hirota S, Ogura T, Olson J S, Kitagawa T. J Phys Chem B. 1999;103:7044–7054.
22. Chen V J, Orville A M, Harpel M R, Frolik C A, Surerus K K, Münck E, Lipscomb J D. J Biol Chem. 1989;264:21677–21681. [PubMed]
23. Nelson M J. J Biol Chem. 1987;262:12137–12142. [PubMed]
24. Salerno J C, Siedow J N. Biochim Biophys Acta. 1979;579:246–251. [PubMed]
25. Twilfer H, Bernhardt F-H, Gersonde K. Eur J Biochem. 1985;147:171–176. [PubMed]
26. Arciero D M, Lipscomb J D. J Biol Chem. 1986;261:2170–2178. [PubMed]
27. Arciero D M, Orville A M, Lipscomb J D. J Biol Chem. 1985;260:14035–14044. [PubMed]
28. Zhang Y, Pavlosky M A, Brown C A, Westre T E, Hedman B, Hodgson K O, Solomon E I. J Am Chem Soc. 1992;114:9189–9191.
29. Nocek J M, Kurtz D M, Jr, Sage J T, Xia Y-M, Debrunner P, Shiemke A K, Sanders-Loehr J, Loehr T M. Biochemistry. 1988;27:1014–1024. [PubMed]
30. Neese F, Solomon E I. Inorg Chem. 1999;38:1847–1865. [PubMed]
31. Brown C A, Pavlosky M A, Westre T E, Zhang Y, Hedman B, Hodgson K O, Solomon E I. J Am Chem Soc. 1995;117:715–732.
32. Hauser C, Glaser T, Bill E, Weyhermüller T, Wieghardt K. J Am Chem Soc. 2000;122:4352–4365.
33. Enemark J H, Feltham R D. Coord Chem Rev. 1974;13:339–406.
34. Farrar J A, Grinter R, Pountney D L, Thomson A J. J Chem Soc Dalton Trans. 1993;18:2703–2709.
35. Wanat A, Schneppensieper T, Stochel G, van Eldik R, Bill E, Wieghardt K. Inorg Chem. 2002;41:1–10. [PubMed]
36. Nakamoto K. Infrared and Raman Spectra of Inorganic and Coordination Compounds. New York: Wiley; 1997.
37. Westre T E, Di Cicco A, Filipponi A, Natoli C R, Hedman B, Solomon E I, Hodgson K O. J Am Chem Soc. 1994;116:6757–6768.
38. Miki E, Mizumacji K, Ishimoro T, Okuno H. Bull Chem Soc Jpn. 1973;46:3779–3784.
39. Quinby-Hunt M, Feltham R D. Inorg Chem. 1978;17:2515–2520.
40. Schelvis J P M, Seibold S A, Cerda J F, Garavito R M, Babcock G T. J Phys Chem B. 2000;104:10844–10850.
41. Reynolds M F, Parks R B, Burstyn J N, Shelver D, Thorsteinsson M V, Kerby R L, Roberts G P, Vogel K M, Spiro T G. Biochemistry. 2000;39:388–396. [PubMed]
42. Li M, Bonnet D, Bill E, Neese F, Weyhermüller T, Blum N, Sellmann D, Wieghardt K. Inorg Chem. 2002;41:3444–3456. [PubMed]
43. Lehnert N, Ho R Y N, Que L, Solomon E I. J Am Chem Soc. 2001;123:12802–12816. [PubMed]
44. Lehnert N, Ho R Y N, Que L, Solomon E I. J Am Chem Soc. 2001;123:8271–8290. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...