Vibrational analyses of the reaction of oxymyoglobin with NO using a photolabile caged NO donor at cryogenic temperatures

Abstract

The NO dioxygenation reaction catalyzed by heme-containing globin proteins is a crucial aerobic detoxification pathway. Accordingly, the second order reaction of NO with oxymyoglobin and oxyhemoglobin has been the focus of a large number of kinetic and spectroscopic studies. Stopped-flow and rapid-freeze-quench (RFQ) measurements have provided evidence for the formation of a Fe(III)-nitrato complex with millisecond lifetime prior to release of the nitrate product, but the temporal resolution of these techniques is insufficient for the characterization of precursor species. Most mechanistic models assume the formation of an initial Fe(III)-peroxynitrite species prior to homolytic cleavage of the O-O bond and recombination of the resulting NO₂ and Fe(IV)=O species. Here we report vibrational spectroscopy measurements for the reaction of oxymyoglobin with a photolabile caged NO donor at cryogenic temperatures. We show that this approach offers efficient formation and trapping of the Fe(III)-nitrato, enzyme-product, complex at 180 K. Resonance Raman spectra of the Fe(III)-nitrato complex trapped via RFQ in the liquid phase and photolabile NO release at cryogenic temperatures are indistinguishable, demonstrating the complementarity of these approaches. Caged NO is released by irradiation <180 K but diffusion into the heme pocket is fully inhibited. Our data provide no evidence for Fe(III)-peroxynitrite of Fe(IV)=O species, supporting low activation energies for the NO to nitrate conversion at the oxymyoglobin reaction site. Photorelease of NO at cryogenic temperatures allows monitoring of the reaction by transmittance FTIR which provides valuable quantitative information and promising prospects for the detection of protein sidechain reorganization events in NO-reacting metalloenzymes.

Graphical Abstract

INTRODUCTION

Nitric oxide (NO) is a stable radical in anaerobic aqueous solutions, but it readily reacts with other radical species and transition metals leading to its broad toxicity and its critical role in the mammalian immune response [1–3]. In macrophages, NO can accumulate to micromolar concentrations and reacts with superoxide (O₂⁻), produced by NADPH oxidase, in a diffusion-controlled fashion to generate peroxynitrite (ONOO⁻):

$${\text{NO}}^{•}+{{\text{O}}_2}^{•-}\to {\text{ONOO}}^{-}$$ (equation 1)

Peroxynitrite is highly reactive and versatile with a near neutral pK_a of 6.8 and cis and trans isomers of comparable energies [4,5]. While it can decay via isomerization to nitrate (NO₃⁻) in aqueous solution, it also reacts with a wide range of biomolecules, including metalloenzymes, to produce additional reactive nitrogen species (RNS) and deleterious molecular defects. Pathogenic microorganisms depend on metal-containing superoxide dismutases and low molecular weight thiol antioxidants to scavenge reactive oxygen species (ROS), and on NO-sensing proteins to control and regulate the gene expression of NO-detoxifying enzymes and repair proteins. Flavohemoglobins and other heme-containing proteins such as truncated hemoglobins are the primary NO-detoxifying enzymes in aerobic conditions. These enzymes catalyze NO dioxygenase activity that uses O₂ to efficiently convert NO to NO₃⁻ [6,7]:

$$\text{heme-Fe}\left(\text{II}\right)+{\text{O}}_2\to \text{heme-Fe}\left(\text{III}\right){{\text{O}}_2}^{•-}+{\text{NO}}^{•}\to \text{heme-Fe}\left(\text{III}\right)+{{\text{NO}}_3}^{-}$$ (equation 2)

In this reaction, a second-order radical combination of NO with the heme iron(III)-superoxo complex forms a putative iron(III)-peroxynitrite intermediate. This intermediate is proposed to subsequently isomerize to NO₃⁻ via homolytic O-O bond cleavage and radical recombination of the resulting NO₂ and ferryl heme Fe(IV)=O species (Scheme 1) [7–9].

Scheme 1.

Proposed mechanisms for the NO dioxygenase reaction in hemoproteins

Monitoring this NO reaction with oxymyoglobin (oxyMb) using resonance Raman spectroscopy along with a rapid-freeze-quench approach (RFQ-RR), we showed that a millisecond intermediate, previously characterized by UV-vis stopped-flow and RFQ-EPR as a high-spin ferric species [10–12], corresponds to a transient Fe(III)-nitrato complex prior to NO₃⁻ release [13]. Accordingly, recent theoretical modeling of the NO deoxygenation reaction in Mycobacterium tuberculosis truncated hemoglobin predicts very low activation energy barriers throughout this reaction with predicted lifetimes for putative Fe(III)-peroxynitrite and Fe(IV)=O intermediates in the picosecond timescale [14,15].

Recently, Tosha, Shiro and coworkers used N,N’-bis-(carboxymethyl)-N,N’-dinitroso-p-phenylenediamine (Scheme 2), a photolabile caged NO initially designed by Fujimori and coworkers [16], to access intermediate species of the enzymatic reaction in various metalloenyzmes [17–19].

Scheme 2.

Structure and photoreaction of N,N’-bis-(carboxymethyl)-N,N’-dinitroso-p-phenylenediamine

In one of these studies, the NO donor is irradiated at cryogenic temperatures to produce a trapped non-heme iron-nitrosyl intermediate in the cytochrome c-dependent NO reductase of Pseudomonas aeruginosa (P.a. cNOR) [19]. Accordingly, we decided to test this approach to probe the NO deoxygenation reaction in oxyMb. In contrast to time-resolved spectroscopy and RFQ techniques, the approach used here cannot generate kinetic information, but it can allow the accumulation of intermediate species if their decay routes are inhibited at lower temperatures than their formation routes. Using low-temperature UV-vis, RR, and FTIR spectroscopies, we show that photorelease and migration of NO at cryogenic temperature forms the same Fe(III)-nitrato complex previously characterized via a RFQ approach [13]. A significant advantage of the cryogenic NO photorelease over the RFQ approach is its amenability to FTIR spectroscopy. In the experiments described herein, the cryogenic light-induced FTIR difference spectra simultaneously identify the presence of the dominant Fe(III)-nitrato complex and of minor ferrous- and ferric-nitrosyl contaminants. The FTIR data also detect the production of nitrous oxide (N₂O) in experiments where the NO donor is used alone, providing indirect evidence for the sub-stoichiometric release of HNO from the photolabile NO donor.

MATERIAL AND METHODS

The N,N’-bis-(carboxymethyl)-N,N’-dinitroso-p-phenylenediamine NO donor was synthesized following published protocols [16,20]. p-phenylenediamine, ethyl bromoacetate, sodium acetate and sodium nitrite, were used as received from Sigma-Aldrich. Sodium nitrite (¹⁵N) was used as received from Cambridge Isotope Laboratories.

Lyophilized horse heart myoglobin was purchased from Sigma. The protein was brought into a glovebox with a controlled atmosphere of less than 1 ppm O₂ (Omni-Lab System; Vacuum Atmospheres Co.) and dissolved in 100 mM phosphate buffer pH 7.5 to a concentration of ~ 3 mM before reduction by addition of 10 mM sodium dithionite before washing the excess reductant and dithionite oxidation product with a PD10 column or desalting spin column (Zebra, Pierce).

Complete heme reduction and dithionite removal was confirmed by UV-vis spectroscopy with a Cary 50 spectrometer and the deoxyMb concentration was determined and adjusted to the desired concentration by dilution or further concentrated with microcon filtrating device (10 kDa cutoff, Millipore). DeoxyMb solutions were then transferred to sealed serum bottles and exposed to ¹⁶O₂ (Airgas) or ¹⁸O₂ (99% ¹⁸O; ICON Stable Isotopes) or simply exposed to air outside the glovebox. Formation of oxyMb was confirmed by UV-vis spectroscopy before addition of 2 equiv NO donor.

Low-temperature UV-vis and FTIR photolysis experiments were conducted on thin films prepared with myoglobin solutions with 1-to-2 equiv NO donor sandwiched between two CaF₂ windows and a 15-μm Mylar spacer. FTIR cells were mounted to a closed-cycle cryostat system (Displex, Advanced Research Systems) equipped with CaF₂ windows before placing the cryostat inside the sample compartment of a Cary-50 UV-vis or Bruker Vertex 80 FTIR spectrometer and kept in the dark until reaching the desired cryogenic temperature. The temperature of the sample was monitored and controlled with a Cryo-Con 32 unit (Cryogenic Control Systems). The FTIR spectrometer was equipped with a liquid N₂-cooled MCT detector and with constant purging of the instrument with compressed air depleted of water vapor and CO₂ (purge gas generator, Puregas LLC). Multiple sets of 2000-scan accumulations at 4-cm⁻¹ resolution were collected and compared before and after sample illuminations. Photolysis was performed by illumination of the sample directly in the FTIR sample chamber using the 355-nm laser emission of a frequency-triple YAG laser (Surelite SLII-10, Continuum) at 10 Hz repetition rate. The 600-mW laser beam was defocused to insure illuminate of the entirety of the IR film inside the cryostat. The photolysis process was efficient and required ~ 1-minute illumination to maximize the UV-vis and FTIR differential signals at 180 K.

Low temperature RR spectra (~110 K) were recorded in a backscattering geometry on frozen samples in EPR tubes maintained at 110 K in a liquid nitrogen Dewar and cold-finger sample holder. All RR spectra were collected using a backscattering geometry on a custom McPherson 2061/207 spectrograph (0.67 m with variable gratings) equipped with a Princeton Instruments liquid N₂-cooled CCD detector (LN-1100PB) with a 407 nm excitation from a krypton laser (Innova 302, Coherent). A long-pass edge filter (RazorEdge, Semrock) was used to attenuate the Rayleigh scattering. Frequencies were calibrated relative to aspirin and are accurate to ±1 cm⁻¹. The lack of photosensitivity of all samples was confirmed by comparing rapid acquisitions within a range of laser powers.

RESULTS AND DISCUSSION

Low-temperature UV-vis analysis of the reaction of deoxyMb and oxyMb with the caged NO donor

As a first step to determining the photolabile efficiency of our 355-nm laser excitation with the photolabile caged NO and diffusion efficiency of the photogenerated NO at cryogenic temperatures, we prepared a deoxyMb sample with 2 equiv of the NO donor and monitored UV-vis spectral changes after illumination at 77, 140, and 180 K. Traces obtained before and after illumination at 77 and 140 K are indistinguishable, indicating that the frozen solvation cage of deoxyMb prevents the photo-released NO to reach the heme Fe(II) ion center. In contrast, illumination for 1 minute at 180 K with the 355-nm laser source results in a complete blue shift of the Soret maximum from 438 to 423 nm with appearance of distinctive Q bands at 547 and 581 nm indicative of the formation myoglobin’s nitrosyl complex [21] (Figure 1, top traces). Using the same illumination conditions (same average power and duration) at 77 K before raising the sample temperature to 180 K produces only a partial conversion of deoxyMb to the nitrosyl complex, suggesting a less efficient photorelease of NO at 77 K, possibly as recombination of photolyzed NO with the caged radical (see Scheme 2) becomes more significant at 77 K.

Figure 1:

UV-vis spectra monitoring the cryogenic reaction of deoxyMb (top traces) and oxyMb (bottom traces) with the NO donor at 180 K. Black and red traces correspond to spectra obtained before and after the 355-nm illumination.

Reproducing these UV-vis experiments with oxyMb and 2 equiv of the caged NO reveals similar temperature dependence. Specifically, the Soret maximum of oxyMb at 77 K is observed at 419 nm and is unchanged after UV-illumination at 77 or 140 K. However, either warming these samples to 180 K, or with direct illumination of samples at 180 K, the Soret maximum change to 413 nm with appearance of a CT band around 635 nm indicative of the formation of high-spin ferric heme species (Figure 1, bottom traces). Further increase of the samples’ temperature to 210 and 220 K do not generate any significant UV-vis changes.

Resonance Raman analysis of the reaction of oxyMb with the caged NO donor

Low-temperature RR spectra of oxyMb show fully symmetric porphyrin modes v₄, v₃, v₂ and v₁₀ at 1378, 1508, 1586 and 1644 cm⁻¹, respectively, that are typical of a 6-coordinate low-spin ferric heme species [22] (Figure 2A). While these RR modes are unchanged after UV irradiation at 110 K, annealing these irradiated samples at 190 K for 10 mins before returning to 110 K downshifts the v₄, v₃ and v₂ modes to 1373, 1482 and 1566 cm⁻¹, respectively, indicating conversion to a 6-coordinate high-spin ferric-heme species [23].

Figure 2:

High- (A) and mid-frequency (B) regions of the RR spectra of oxyMb with unlabeled and ¹⁵N-labeled caged NO after illumination with the 355-nm emission from the YAG laser at 110 K. The blue trace corresponds to the RR spectrum of a sample maintained at 110 K while the black and red traces are spectra of illuminated oxyMb samples with unlabeled and ¹⁵N-labeled NO photo-donor incubated for 10 minutes at 190 K before data collection at 110 K. The green trace corresponds to the difference between the black and red traces.

Experiments performed with ¹⁵N-labeled caged NO identifies a single isotope sensitive RR band at 1282 cm⁻¹ that downshifts to 1260 cm⁻¹ with ¹⁵N (Figure 2B). These RR frequencies are exact match for the unlabeled and ¹⁵N-labeled v_s(NO₂) vibration of the iron(III)-nitrato complex previously trapped on the millisecond time using a rapid-freeze-quench approach on samples at 5 °C [13].

Low-temperature FTIR photolysis analysis of the reaction of oxyMb with caged NO

Low-temperature FTIR experiments were performed with 4 mM oxyMb and 1 equiv caged NO, collecting series of 2000-scan spectra before and after 355-nm illumination at 180 K. Light minus illuminated difference spectra produce numerous differential signals in the 1100 to 1950 cm⁻¹ frequency window (Figure 3A). Control experiments with the caged NO alone identify differential signals associated with the photoconversion of the starting material to the quinone-imine byproduct after release of NO (Figure 3B). In these control experiments, the most prominent signals are negative bands corresponding to IR absorption features of the NO donor, indicating that the quinone-imine product is a much weaker IR absorbing species than the starting material. Labeling of the nitrosyl groups with ¹⁵N-atoms affects most severely two bands at 1416 and 1463 cm⁻¹ with vibrational contributions from the NNO groups. As expected, the N-O stretching absorption for the photolyzed NO is too weak and broadened for detection in these FTIR difference spectra, but a distinctive positive band at 2229 cm⁻¹ that downshifts to 2160 cm⁻¹ with ¹⁵N-labeled caged NO is clearly observed in the high-frequency region (Figure 3B). These frequencies are 5 cm⁻¹ higher than literature values for gas phase frequencies of the v₃ mode of unlabeled and doubly labeled ¹⁵N¹⁵NO at 2224 and 2155 cm⁻¹, respectively [24]. These IR absorptions are observed whether the samples are prepared in presence or absence of O₂ and are indirect reporters of the photoproduction of nitroxyl (³NO⁻ and ¹HNO). Evaluating the concentration of N₂O produced from the intensity of the v₃ absorption bands and standard solutions of N₂O suggest a yield of 0.10 ± 0.05 N₂O per caged NO. This HNO yield at cryogenic temperatures is comparable to values reported earlier for the photochemistry of aqueous NONOate solutions at ambient temperatures [25,26]. As the HNO concentrations remain low relative to NO, HNO dimerization is expected to be outpaced by the reaction of HNO with NO to produce nitrite and N₂O, which imply that we can assign the ~10% production of N₂O to a ~10% photoproduction of HNO per NO donor.

Figure 3:

180-K “illuminated” minus “dark” FTIR photolysis difference spectra of oxyMb in presence of unlabeled (black) and ¹⁵N-labeled (red) caged NO donor (A), and of the caged NO alone (B).

Compared to the control FTIR difference spectra obtained with the caged NO alone, the data obtained with oxyMb show a positive band at 1283 cm⁻¹ that downshifts to 1260 cm⁻¹ with the ¹⁵N-labeled NO donor (Figure 3A). These frequencies match those observed by RR spectroscopy for the v_s(NO₂) of the nitratoMb complex. The assignment of these FTIR signals to the nitrato v_s(NO₂) mode is further supported by the splitting of this signal observed when oxyMb is prepared with ¹⁸O₂ gas (see Figure S1). Isotope sensitive differential signals are also seen around 1500 cm⁻¹ in a frequency range where a nitrato v_as(NO₂) mode would be expected [27]. Accordingly, a strong positive band at 1507 cm⁻¹ is likely to correspond to the nitrato the v_as(NO₂) mode as it is only seen when both oxyMb and the unlabeled caged NO are present. Unfortunately, the expected v_as(¹⁵NO₂) counterpart to this unlabeled v_as(NO₂) mode is not clearly defined, even in the isotope-edited double difference spectrum as isotope sensitive modes from the caged NO also contribute differential signals in that same frequency range (Figure 4). Accidental degeneracy of nitrato group vibrations with porphyrin skeletal modes occurring between 1380 and 1460 cm⁻¹ may also contribute to the complexity seen in this spectral range of the isotope-edited difference spectrum. FTIR signals from free nitrate in aqueous solution are broad [28] and are not detected in our experiments with oxyMb and the caged NO (see Figure S2).

Figure 4:

Isotope-edited double difference spectra of oxyMb with the caged NO donor (top black trace) and the caged NO alone (bottom blue trace). The traces are obtained by subtracting the “illuminated-¹⁵N” minus “dark-¹⁵N” difference spectrum from the “illuminated-¹⁴N” minus “dark-¹⁴N” spectrum.

Sharp differential signals in the 1600 to 1700 cm⁻¹ region are only seen when both oxyMb and the NO donor are present (Figure 3A). However, these signals are unaffected by the ¹⁵N-labeleing of the NO donor and cancel out in the isotope-edited double difference spectrum (Figure 4). These signals must be associated with minor frequency shifts of protein vibrations upon conversion of oxyMb to nitratoMb.

The isotope-edited double difference spectrum of oxyMb also show a very weak positive signal at 1614 and 1924 cm⁻¹ that downshift 28 and 38 cm⁻¹ with ¹⁵N-labeled caged NO. These frequencies are characteristic of the v(N-O) mode of the ferrous- and ferric-nitrosyl complexes of Mb [29,30]. The concentration of these nitrosyl species can be estimated to ~ 0.1 mM based on the intensity of the v(N-O) bands, which presumably reflects ~ 2% metMb and deoxyMb contamination in our starting oxyMb solutions. The photoproduction of nitroxyl from the NO donor may also contribute to the formation of these ferric and ferrous-nitrosyl Mb contaminants. Indeed, the v₃ signals from N₂O seen with the NO donor alone and attributed to the formation of HNO are not seen in the spectra obtained with oxyMb (Figure 4).

Interestingly, the N₂O signals observed in control experiments with the caged NO alone are absent of the FTIR different spectra obtained with oxyMb samples, implying that the sub-stoichiometric HNO produced by the NO donor is consumed by myoglobin before it can produce N₂O. The product of oxyMb reaction with nitroxyl would be expected to lead to the formation of nitrito-metMb species, but anticipated concentration of this side product is too low to obtain direct evidence for this species in our FTIR spectra. Moreover, we cannot rule out metMb and deoxyMb contaminants in the oxyMb samples as possible source for the HNO consumption; both are known to react with HNO and produce ferrous-nitrosyl and ferrous-nitroxyl species [31,32].

CONCLUSION

Using a photolabile caged NO donor at cryogenic temperatures, we have trapped a ferric-nitrato enzyme-product complex, for the NO dioxygenation reaction catalyzed by oxyMb. Below 180 K, the diffusion of the photolyzed NO into the heme distal pocket of Mb is inhibited, but above that temperature the radical reaction between NO and the superoxo group of oxyMb proceeds all the way to ferric-nitrate complex. The low-temperature UV-vis, RR, and FTIR data provide no evidence for the trapping of any of the putative early intermediate species such as ferric-peroxynitrite and Fe(IV)-oxo species. This failure to trap early intermediate species is consistent with theoretical modeling of this reaction and the prediction of near barrierless isomerization of Fe(III)-peroxynitrito to Fe(III)-nitrato species, but it clearly does not rule out the viability of these species as transient states in the NOD reaction of myoglobin. The 180-K thermal barrier for the initiation of the NO reaction in deoxyMb can be assumed to reflect collective motions of the protein matrix required for NO to gain access to near-attack position by the Fe(III)-superoxo group. Slightly higher thermal barriers (200 to 220 K) have been observed for non-geminate rebinding of CO to deoxyMb after photodissociation [33–35]. At ambient temperature, O₂ access to the iron distal pocket is known to be gated by the distal histidine [36]; molecular dynamic simulations have proposed additional gated pathways across the surface of myoglobin [37,38], but experimental support for these alternative pathways remain limited. Tosha, Shiro and coworkers reported an equivalent 180-K thermal barrier for the reaction of NO with the detergent solubilized P.a. cNOR [19], but the gaping contrast in structures of these two proteins leave no doubts on the coincidental nature of these matching temperatures. Unlike Mb, P.a. cNOR is equipped with a large hydrophobic channel connecting the outer-hydrophobic surface of its twelve transmembrane α-helix bundle with the diiron active site anchored at its core [39].

Previous RFQ experiments with oxyMb and NO were performed in pH 9.0 buffer to allow for easy distinction between the high-spin nitrato complex and the low-spin hydroxy-metMb product [13]. It was unclear at the time if the alkaline conditions increase the lifetime of the nitrato complex and whether the same nitrato complex would form at neutral pH. The perfect match in the v_s(NO₂) frequencies seen in the RFQ samples and photogenerated complex indicates that the environment of the nitrato group is unaffected by the pH difference and that if pH 9.0 increase the lifetime of the nitrato complex it must be through changes at the egress route used by the nitrate group.

Our work here has shown how photoreleasing NO in frozen samples allows characterization of trapped reaction steps by low-temperature transmittance FTIR. In contrast to RR spectroscopy, transmittance FTIR provides direct access to quantitative information as well as evidence for protein sidechain conformational changes. Specifically, in the study discussed here, low-temperature FTIR spectra have revealed the presence of trace ferric- and ferrous-nitrosyl species, undetected in the RR data, and which we assigned to minor deoxyMb and metMb contaminants in the oxyMb starting material. It is conceivable that in some multistep reactions, this FTIR approach could afford information on conversion rates between intermediate species. Sharp derivative signals seen in the FTIR difference spectra associated with the conversion of oxyMb to the Fe(III)-nitrato complex that were unaffected by the labeling of the NO donor and absent from control experiments with the NO donor alone were indicators of perturbations at the porphyrin macrocycle and protein sidechains. Follow-up experiments with isotopic labeling of the protein and porphyrin substitutions would allow accurate assignments for these differential signals and lead to detailed reaction mechanism schemes.

Control FTIR experiments with the NO donor provide indirect evidence for the release of ~10 % nitroxyl upon UV irradiation. This small fraction of NO⁻/HNO production does not affect the promise of this photocryogenic approach to mechanistic studies of NO catalysis, but it will be important to keep this side-reaction in mind in future studies if multiple intermediate species are detected after photolysis.

Synopsis.

The graphic shows the structure of oxymyoglobin, the caged NO compound, and the result of uncaged at 180 K.

Highlights.

• Nitric oxide readily diffuses into the active site of oxymyoglobin at 180 K.

• The reaction of NO with oxymyoglobin at 180 K traps a Fe(III)-nitrato complex.

• Transmittance FTIR data quantifies trace ferric- and ferrous-nitrosyl contaminants.