Abstract
The H-NOX family of proteins have important functions in gaseous ligand signaling in organisms from bacteria to humans, including nitric oxide (NO) sensing in mammals, and provide a model system for probing ligand selectivity in hemoproteins. A unique vibrational feature that is ubiquitous throughout the Heme-Nitric oxide/OXygen binding (H-NOX) family is the presence of a high C-O stretching frequency. To investigate the cause of this spectroscopic characteristic, the Fe-CO and C-O stretching frequencies were probed in the H-NOX domain from Thermoanaerobacter tengcongensis (Tt H-NOX) using resonance Raman (RR) spectroscopy. Four classes of heme pocket mutants were generated to assess the changes in stretching frequency: (i) the distal H-bonding network, (ii) the proximal histidine ligand, (iii) modulation of the heme conformation via Ile-5 and Pro-115, and (iv) the conserved Tyr-Ser-Arg (YxSxR) motif. These mutations revealed important electrostatic interactions that dampen the back-donation of the FeII dπ electrons into the CO π* orbitals. The most significant change occurred upon disruption of the H-bonds between the strictly conserved YxSxR motif and the heme propionate groups, producing two dominant CO-bound heme conformations. One conformer was structurally similar to Tt H-NOX WT; whereas the other displayed a decrease in ν(C-O) of up to ~70 cm−1 relative to the WT protein, with minimal changes in ν(Fe-CO). Taken together, these results show that the electrostatic interactions in the Tt H-NOX binding pocket are primarily responsible for the high ν(C-O) by decreasing the Fe dπ → CO π* back-donation, and suggest that the dominant mechanism by which this family modulates the FeII-CO bond likely involves the YxSxR motif.
Keywords: Resonance Raman spectroscopy, H-NOX, Thermoanaerobacter tengcongensis, carbon monoxide, protein-ligand interactions, proximal ligand substitution, backbonding, YSR motif, heme proteins
Recently, a novel class of heme proteins, the Heme-Nitric oxide/OXygen binding (H-NOX) domains was identified through sequence similarity to the heme domain of mammalian soluble guanylate cyclase (sGC). sGC is the mammalian receptor for NO and is involved in various functions such as blood vessel dilation. In addition, studies have suggested that H-NOX domains may be gas sensors in a variety of prokaryotes. The H-NOX family of heme proteins has the distinctive property in that some members bind only NO and CO, whereas others can also form a stable complex with O2, thereby providing an avenue for probing ligand selectivity within heme proteins (1–3). A unique feature that is shared by all H-NOX domains, and may contribute to ligand binding properties, is an unusually high C-O stretching frequency observed in the heme-CO adducts. This mode, which typically occurs in other heme proteins at vibrational frequencies that are 20–50 cm−1 lower than that observed for the H-NOX family, is sensitive to the protein environment surrounding the bound ligand and can provide detailed insight on how the heme binding pocket possibly interacts with other diatomic gases, such as NO or O2. Despite the availability of NMR and X-ray crystal structures for several of these H-NOX domains (3–7), the factors that are responsible for the high ν(C-O) and the biochemical implications of this stretching mode remain unclear.
Resonance Raman (RR) spectroscopy is a powerful tool to probe heme protein-ligand interactions in solution, especially with heme-CO adducts because the nature of these interactions can be readily elucidated from the Fe-CO and C-O stretching frequencies since these modes are highly sensitive to the electrostatic field of the distal pocket and the donor properties of the proximal ligand. Extensive density functional theory (DFT) computations and vibrational studies on both heme proteins and synthetic model porphyrins have previously shown that the Fe-CO and C-O vibrational frequencies are negatively correlated due to backbonding effects; the FeII dπ electrons are back-donated to the CO π* orbitals, strengthening the Fe-CO bond while weakening the C-O bond (8–25). Thus, a key focus in investigating these H-NOX proteins via RR spectroscopy is to better understand the interactions between the exogenous diatomic ligands (e.g. O2, NO, and CO) and the H-NOX heme-binding pocket using the FeII CO adduct as a general model. A deeper appreciation of the complex protein-ligand interactions will not only reveal how the C-O stretching frequency is modulated in these heme domains, but may also help determine key components that are responsible for controlling ligand specificity in the H-NOX family.
In addition to the high ν(C-O) frequency in the H-NOX family, CO is also of interest as it weakly activates sGC upon binding to the heme prosthetic group (26). The 2–4-fold increase in catalytic activity upon CO binding is significantly lower than the 100–400-fold increase upon NO binding to sGC; however, the activity can be modulated by small molecules such as the benzylindazole derivative, YC-1 [3-(5'-hydroxymethyl-3'-furyl)-1-benzylindazole]. RR investigations have shown changes in the heme conformation and associated CO vibrational modes upon introduction of these small molecules, suggesting that they interact with the heme prosthetic group (27–34). Furthermore, previous work found that truncation of full length sGC to its heme domain, β1(1–194), resulted in a ~20 cm−1 ν(C-O) downshift to ~1968 cm−1 (35–38). These results suggest that the protein structure and conformational flexibility also influence the CO stretching frequency.
In this study, RR spectroscopy is used in conjunction with site-directed mutagenesis and isotopic substitution to evaluate how different features within the H-NOX domain from Thermoanaerobacter tengcongensis (Tt H-NOX) affect the C-O stretching frequency and elucidate the binding site interactions between the ligand-bound heme and the protein environment. The specific H-NOX properties investigated in this work include: (i) the presence of H-bonding distal residues in the binding pocket, (ii) the bond strength and steric constraints of the proximal ligand, (iii), heme deformation and (iv) H-bonding between the conserved Tyr-Ser-Arg (YxSxR) motif and the propionate substituents of the heme (Figure 1). Our RR spectra not only provide insight into these interactions in solution and their influence on the Fe-CO bond in Tt H-NOX, but additionally provide a plausible mechanism for signal transduction from the heme to associated signaling components in this broad family of diatomic sensors.
Figure 1.

Structure of the heme pocket of Tt H-NOX highlighting the locations of important residues (PDB 1U55). (A) Overview of the distal and proximal pockets, showing the hydrogen-bonding network, proximal His-102, and residues associated with heme deformation (Ile-5 and Pro-115). (B) The conserved YxSxR motif of the H-NOX family.
MATERIALS AND METHODS
Protein expression and purification
Expression and purification of the his-tagged (−H6) Tt H-NOX domain were performed as previously described (39). Site-directed mutagenesis was carried out using the QuikChange® protocol (Agilent), and verified by sequencing (UC Berkeley sequencing core).
The Tt H-NOX H102G mutant was expressed and purified with imidazole (Im) as the proximal ligand using the same methods described for the WT protein, with the following modifications. Imidazole (10 mM) was added to the expression cultures at the time of induction with isopropyl-β-D-1-thiogalactopyranoside (IPTG). In addition, 10 mM imidazole was added to buffer A (50 mM TEA, 20 mM NaCl, 5% glycerol, pH 7.5) during purification to ensure retention of the heme in the Tt H-NOX H102G mutant.
Sample preparation
Purified Tt H-NOX-H6 protein was brought into an anaerobic glovebag, and oxidized using 5–10 mM potassium ferricyanide to remove the bound O2. The ferricyanide was removed using a PD10 desalting column (GE Healthcare) equilibrated with buffer B (50 mM TEA, 50 mM NaCl, pH 7.5). Following oxidation and desalting, the protein was reduced with ~5–20 mM sodium dithionite that was removed using a PD10 desalting column upon complete reduction of the heme. To make the 12CO (Praxair) and 13CO (13C, 99%; 18O, < 2%; Cambridge Isotopes) complexes, gas was added to a sealed Reacti-Vial (Pierce) containing FeII-unligated protein. Final sample concentrations for the Raman experiments were typically 15 to 30 μM for the ferrous unligated proteins, and between 35 to 75 μM for the CO complexes. All UV/Vis absorption samples were prepared and measured as previously described (1, 40).
Other exogenous proximal ligands were substituted into the Tt H-NOX H102G mutant in addition to imidazole (Im); these included 1-methyl-imidazole (1-Me-Im), 2-methyl-imidazole (2-Me-Im), 4(5)-methyl-imidazole [4(5)-Me-Im], 4(5)-bromo-imidazole [4(5)-Br-Im], 4(5)-iodo-imidazole [4(5)-I-Im], pyridine (Pyr), and 3-fluoro-pyridine (3-F-Pyr). To exchange the proximal ligands, the purified Tt H-NOX H102G(Im) protein was heated for 1 h at 50 °C with ~5–10 mM potassium ferricyanide, then desalted with a PD10 column equilibrated with buffer B containing 10 mM of the appropriate proximal ligand. The desalted protein was then heated at 50 °C for an additional 30 min, and reduced with ~5–20 mM sodium dithionite that was removed using a PD10 desalting column upon complete reduction of the heme. This method was sufficient to exchange the imidazole with the alternate proximal ligands. Once the proximal ligand was substituted, generation of the ferrous unligated and CO-bound forms were performed as described above.
Resonance Raman spectroscopy
All spectra were collected using the 413.1 nm line from a Kr+ laser (Spectra-Physics model 2025) as previously described (41). The protein samples were anaerobically loaded into a microspinning Raman sample cell, described previously (42), in a glove bag to avoid the possibility of O2-binding. The power at the sample was 2 mW, except for the CO complexes, where the laser power was reduced to ~250 W to minimize CO photolysis. Typical data acquisition times were 60 to 90 minutes. Electronic absorption spectra were obtained before and after the Raman experiments to verify that no photoinduced degradation occurred. Raman spectra were corrected for wavelength dependence of the spectrometer efficiency with a white lamp, and the instrument was calibrated using the Raman frequencies from cyclohexane, CCl4, and toluene. The reported frequencies are accurate to ±1 cm−1, and the spectral bandpass was set to 8 cm−1. For each Raman spectrum, the raw data were baseline-corrected, and the buffer background signal was subtracted. Spectral analysis was performed using Igor Pro (WaveMetrics). Theoretical isotopic shifts were approximated on the basis of a simple harmonic oscillator model.
RESULTS AND DISCUSSION
To characterize the conformational and electronic changes of the heme chromophore in the Tt H-NOX mutants, electronic absorption and RR measurements on the ferrous unligated and CO-bound forms were carried out (Figures 2–6; Supporting Information, Tables S1–S4). The RR spectra of the Tt H-NOX CO complexes are organized by mutations made in the heme pocket that involve: (i) distal H-bonding residues, (ii) proximal ligand substitution, (iii) heme deformation, and (iv) the YxSxR motif. In addition, the heme skeletal modes, ν(Fe-His), and CO-associated modes are tabulated (Supporting Information, Tables S3 and S4) for comparison with other heme proteins (1, 27, 35–37, 43–46). Through careful analysis of the RR data presented here together with the available NMR and X-ray crystallographic results (3–7), a more comprehensive understanding of the heme structure and its interactions with the H-NOX protein environment has been developed.
Figure 2.
Resonance Raman spectra of the CO complexes of Tt H-NOX WT, W9F, Y140H, Y140F, P115A/Y140F, and Y140L. Panel (A) shows the low frequency 12CO (solid line), 13CO (dotted line), and magnified (2–4×) difference (13CO-12CO) spectra for clarity. Panel (B) shows the skeletal markers in the high frequency region; ν4 has been reduced in size to more clearly display the markers. Panel (C) displays a magnified view of the CO stretch region, with 13CO (dotted) and 12CO (solid) overlaid, and the difference spectra (13CO-12CO) included below. Spectral intensities were normalized to ν7 and ν4 for the low and high frequency regions, respectively.
Figure 6.
Resonance Raman spectra of the CO complexes of Tt H-NOX WT, Y131F, S133A, S133C, and R135Q. 13CO spectra (dotted line) are overlapped over the 12CO spectra to indicate the frequency shifts upon isotopic substitution, and the difference (13CO-12CO) spectra are shown below each protein for clarity. Insets show the isotopic shift trends for ν(C–O) in the 1850–2000 cm−1 region. Spectral intensities were normalized to ν7 and ν4 for the low and high frequency regions, respectively.
Distal electrostatic effects on CO-bound Tt H-NOX
As the distal pocket H-bonding residues have previously been shown to be important determinants of O2 binding and affinity in the H-NOX family, the effect of distal pocket mutants on the electronics of heme-ligand binding were investigated using the Fe(II)-CO adducts. Previous studies with other heme proteins reported modulations in the ν(Fe-CO)/ν(C-O) backbonding correlation upon altering the polar interactions between the CO ligand and distal residues (9, 14, 17–23, 44). A notable example is the Mb double mutant H64V/V68T, which produced vibrational frequencies of 479 and 1984 cm−1 for the Fe-CO and C-O stretching modes, respectively, in comparison to the native Mb protein [ν(Fe-CO) ~512 cm−1; ν(C-O) ~1944 cm−1] (20). The decreased backbonding in this double mutant was thought to result from negative polarity in the distal pocket due to lone electron pairs from the oxygen in Thr-68 being directed at the CO ligand, diminishing backbonding and increasing the C-O bond order. This negative polarity in the distal pocket forces electron density back to the Fe center and decreases the Fe-C bond order. Similar trends were previously demonstrated in synthesized superstructured iron twin-coronet porphyrin models with naphthalene hydroxyl groups overhanging the bound CO moiety (47).
The Tt H-NOX WT protein has a similar CO stretching frequency to that of the Mb H64V/V68T double mutant [Tt H-NOX WT, ν(C-O) ~1987 cm−1 vs Mb H64V/V68T, ν(C-O) ~1984 cm−1]; thus, the Tt H-NOX distal pocket may also contain a localized negative polarity with the oxygen from Tyr-140 directly above the CO ligand. Although no crystal structures of the CO-bound Tt H-NOX domain are available, previous structural studies from Pellicena et al. (3), Ma et al. (6), and Erbil et al. (7), strongly indicate that the CO moiety would bind in a mostly linear geometry in the heme pocket and could directly interact with the two lone pairs of the hydroxyl group of the Tyr-140 distal residue as indicated in the simplified schematic representation of the CO-bound Tt H-NOX (Figure 7).
Figure 7.

Interaction between Tyr-140 and CO in Tt H-NOX. The electron lone pairs from Tyr-140 point toward the CO moiety in Tt H-NOX WT. Upon mutation to a hydrophobic residue, the negative polarity introduced by Tyr-140 is abolished so back-donation from the FeII dπ orbital into the CO π* orbital increases, weakening the CO bond strength.
Several previously studied CO-bound myoglobin variants exhibited an increase in the ν(C-O) frequency upon replacement of the distal histidine residue with a nonpolar residue that cannot form H-bonding contacts with the bound CO (Supporting Information, Figure S4) (20). This results from the increased electron density around the CO moiety upon removal of the H-bond from the distal His-64 residue, decreasing the effects of back donation from the FeII dπ electrons into the CO π* orbitals and increasing the CO bond order. In contrast, the RR spectra reported here indicate a decrease in the C-O stretching frequency upon disruption of the distal H-bonding network in Tt H-NOX (Figure 2). In particular, both the Tt H-NOX Y140F and Y140L mutants display C-O stretching modes at 1966 cm−1 and 1930 cm−1, respectively, that are lower in frequency relative to the WT protein [ν(C-O) ~1987 cm−1]. This observation is the opposite of the trends previously observed in Mb (9, 14, 15, 17, 19, 20), and suggests that Tyr-140 does not form an H-bond with the CO moiety in the native Tt H-NOX protein. Furthermore, the previous work on the Mb H64V/V68T double mutant suggests the possibility that the lone electron pairs from the Tyr-140 oxygen point toward the CO moiety in Tt H-NOX. Therefore, the negative polarity should be abolished upon substitution of Tyr-140 with a hydrophobic residue, resulting in a decrease in the observed ν(C-O) (Figure 7).
In the Tt H-NOX WT protein, Tyr-140 is held in place by a H-bonding network that involves Trp-9 and Asn-74 (3). To investigate the influence of this extended H-bonding network on the C-O stretching frequency, attention was focused on the nearby Trp-9 residue in the distal pocket (Figure 1A). Mutation of Trp-9 to phenylalanine disrupts the H-bonding network, and a significant downshift of ~60 cm−1 to ~1930 cm−1 is observed in ν(C-O). It is plausible that by removing Trp-9, Tyr-140 is allowed to freely rotate into a H-bonding interaction with the bound CO ligand. The combination of replacing the original oxygen lone pair interaction with a H-bonding one would then produce the observed ~60 cm−1 ν(C-O) downshift. Similar to ν(C-O), the Fe-CO stretching mode also decreases to 485 cm−1 and broadens relative to the WT protein (Figure 2). The observed broadening of the Fe-CO stretching mode likely reflects the fact that the Tyr-140 residue has more freedom of motion upon disruption of the H-bonding network. The fact that the Fe-CO stretch also displayed a decrease in frequency despite such a large decrease in the C-O stretching frequency strongly suggests that there are other effects counterbalancing the increased back donation from the FeII dπ electrons into the CO π* orbitals. Although competition from the trans axial ligand for σ-overlap with the Fe dz2 orbital could counterbalance the expected increase in the Fe-CO stretching mode (20), this seems unlikely since there is no anticipated change to the donor strength of the proximal histidine ligand with the W9F mutation. One possibility is the presence of a modest Fe-CO displacement off the heme z-axis plane upon formation of distal H-bonding interactions, which would reduce the π bonding and result in deviations from the backbonding correlation (8, 21). From DFT studies, Kozlowski et al. found that both the C-O and Fe-CO stretching frequencies decrease as the oxygen atom is displaced from the heme z-axis due to orbital misalignment (8). Thus, it is possible that this modest displacement, in addition to the H-bonding interaction between the CO ligand and distal Tyr-140, may account for the decreased frequencies observed in both ν(Fe-CO) and ν(C-O) for the CO complex in Tt H-NOX W9F.
Another unique aspect of the RR spectra was the presence of multiple CO conformers upon mutation of Tyr-140 to histidine, evidenced by two observed ν(C-O) frequencies at 1928 cm−1 and 1952 cm−1 (Figure 2). This is further supported by the broadened Fe-CO stretching mode observed in Tt H-NOX Y140H with a peak at 501 cm−1 and a shoulder at ~515 cm−1. The presence of multiple conformers in the CO-bound Tt H-NOX Y140H mutant may be reflective of a loosely held histidine residue in the distal pocket, resulting in two different transient H-bonding interactions between the CO moiety and the protonated NHis. The reduced CO stretching frequencies that are observed would be consistent with an H-bond interaction between these two groups for both conformers. The only other distal pocket variant to display multiple CO conformations was the double mutant, Tt H-NOX P115A/Y140F, which displayed broad CO stretching modes at 1950 and 1964–1972 cm−1, in addition to broad Fe-CO stretching modes at 492 and 512 cm−1. We have previously suggested that the single P115A mutation sufficiently disrupts the hinging region of the H-NOX domain, resulting in a more `open' conformation that relaxes the heme (4, 7, 41). This increased opening of the distal and proximal regions of the H-NOX domain would certainly provide dynamic disorder within the heme pocket, making accessible protein-ligand contacts more transient. This effect, in combination with the structural rearrangements to accommodate the Y140F mutation, is plausibly responsible for the multiple conformations and peak broadening observed in the P115A/Y140F double mutant.
Although the idea of a localized negative polarity from the hydroxyl group of the Tyr-140 residue in the distal pocket provides a reasonable interpretation for the high ν(C-O) in Tt H-NOX and the consequent downshifts in frequency upon distal pocket mutation, it does not extend to the rest of the H-NOX family due to the absence of this residue in the H-NOX domains from facultative aerobes. In fact, structural studies on the H-NOX domains from Shewanella oneidensis and Nostoc sp indicate a predominantly nonpolar distal pocket lined with mostly hydrophobic residues (6, 7). Thus, other contributing factors (e.g. electrostatic or steric) from the H-NOX heme pocket must also play a role in the unusually high ν(C-O) frequency. These additional factors may provide insight as to how members of the H-NOX family with the same fold are able to exhibit different ligand selectivities.
Modulation via the proximal ligand
It is well known that the axial residue ligand trans to the CO influences the backbonding between the FeII and CO. This effect of the axial ligand is readily observed from the backbonding plot (Figure 8) showing the ν(Fe-CO)/ν(C-O) correlations among heme proteins with different proximal ligands (e.g. imidazole versus thiolate) (11, 12, 20, 21, 23, 48, 49). Vogel et al. also previously showed that the ligand proximal to the CO displaces the backbonding plots, altering the slopes and lowering the Fe-CO stretching frequency. Changes in the backbonding slope depend on the donor strength of the proximal ligand and occur due to the competition between the proximal ligand and CO for σ-orbital overlap with the FeII dz2 acceptor orbital (23, 24). With the increased electron density on the heme metal due to the proximal ligand, backdonation from FeII dπ to CO π* is increased and the C-O bond weakens.
Figure 8.

Backbonding correlation plot between ν(Fe-CO) and ν(C–O) for several H-NOX domains and other heme proteins. The CO-bound heme proteins and the corresponding symbols for the plot are as follows: H-NOX domains (∘); Tt H-NOX H102G with substituted proximal ligands (•); other histidyl-ligated heme proteins (×); P450's (■).
Tt H-NOX WT has a Fe-His stretching frequency of ~217 cm−1, which is indicative of a weak to moderate H-bond to the proximal ligand in comparison to no H-bond [ν(Fe-His) ~200 cm−1] and a strong H-bond [ν(Fe-His) ~240 cm−1] (49). Based on the similarity in vibrational frequency, it is possible that the Tt H-NOX WT protein contains a Mb-like imidazole H-bond to either a carbonyl group of the protein backbone, a nearby proximal residue, or water. No crystal structures of the FeII-unligated Tt H-NOX exist, so it is uncertain what is forming this moderate H-bond with His-102 in Tt H-NOX. However, based on the O2-bound Tt H-NOX crystal structure by Pellicena et al. (3), the H-bond is plausibly between the Nδ proton of His-102 and the backbone carbonyl of Met-98. The 2.8 Å distance between the Nδ of His-102 and backbone carbonyl oxygen from Met-98 is comparable to the ~2.9 Å distances between the corresponding H-bonding residues in Sw Mb, His-93, Ser-92 and Leu-89 (50).
In the H-NOX family, the vibrational frequency of the Fe-His stretching mode ranges from 208 to 224 cm−1 (Supporting information, Table S3). Despite these varying Fe-His bond strengths, the observed ν(Fe-CO) and ν(C-O) frequencies remain remarkably similar across the H-NOX family (Supporting information, Table S4). Thus, it does not appear that the strength of the Fe-His bond alone contributes to the unusually high C-O stretching frequency. However, it is important to note here that these observations do not necessarily rule out a functional role of the Fe-His bond within multi-component signaling mechanisms that involve the H-NOX domain. Previous studies have already shown that ν(Fe-His) increases from 204 cm−1 to 208 cm−1 as sGC is truncated down to the β(1–194) domain (36–38); thus, additional RR studies on the full-length proteins with their associated signaling components (e.g. histidine kinases, methyl-accepting chemotaxis domains, or diguanylate cyclases) should be completed before such conclusions are made.
To probe the role of the proximal ligand in modulating the ligand binding properties of the H-NOX family, the proximal ligand was systematically exchanged with a series of imidazole derivatives within the same protein scaffold via the Tt H-NOX H102G mutant. These imidazole derivatives produced a range of Fe-Nim stretching frequencies from 188 to 215 cm−1. Despite similarities in heme skeletal mode frequencies for the 5-coordinated fully reduced Tt H-NOX H102G variants, the RR spectra exhibited significant shifts in both the vibrational frequency and intensity for the CO complexes (Figures 3 and 4). Several CO-bound Tt H-NOX H102G variants [e.g. 2-Me-Im, 4(5)-Me-Im, 4(5)-Br-Im, and 4(5)-I-Im] displayed a dominant CO conformer with a ν(Fe-CO) at 526–531 cm−1 and a ν(C-O) at 1960–1964 cm−1 (Figure 3). However, when 1-Me-Im was incorporated as a proximal Fe ligand, an additional conformer appeared in the RR spectra with a ν(Fe-CO) at ~493–497 cm−1 and ν(C-O) at ~1984–1990 cm−1 (Figure 3). Since the vibrational frequencies for the Fe-CO and C-O stretching modes in Tt H-NOX H102G(Im) [ν(Fe-CO) ~492 cm−1, ν(C-O) ~1983 cm−1] are similar to those of the WT protein [ν(Fe-CO) ~490 cm−1, ν(C-O) ~1987 cm−1], it is clear that the substantial shifts observed for the other proximal ligands are directly related to the additional substituents. Interestingly, the Tt H-NOX H102G(Pyr) and H102G(3-F-Pyr) proteins exhibited increased ν(C-O) frequencies at 1997 cm−1 in comparison to the WT protein [ν(C-O) ~1987 cm−1] despite the significantly reduced ν(Fe-Npyr) frequencies of 203 and 197 cm−1 for pyridine and 3-F-pyridine, respectively (Figure 4; Supporting Information, Table S4).
Figure 3.
Resonance Raman spectra of the CO complexes of Tt H-NOX WT and the Tt H102G mutant with 1 mM of the following proximal ligands: imidazole (Im), 1-Me-Im, 2-Me-Im, 4(5)-Me-Im, 4(5)-Br-Im, and 4(5)-I-Im. 13CO spectra (dotted line) are overlapped over the 12CO spectra, and the difference (13CO-12CO) spectra are shown below each protein. Spectral intensities were normalized to ν7 and ν4 for the low and high frequency regions, respectively.
Figure 4.

Resonance Raman spectra of the CO complexes of Tt H-NOX WT and the Tt H102G mutant with 1 mM of the proximal ligands, pyridine (Pyr) and 3-F-Pyr. 13CO spectra (dotted line) are overlapped over the 12CO spectra, and the difference (13CO-12CO) spectra are shown below each protein. Spectral intensities were normalized to ν7 and ν4 for the low and high frequency regions, respectively.
There are several possible explanations for these spectral differences between the Tt H-NOX H102G proteins and WT (ν(Fe-CO) ~490 cm−1, ν(C-O) ~1987 cm−1), including: (i) the changes in the ligand pKa, (ii) the size, strain, and orientation of the ligand within the proximal pocket, and (iii) the ability to form H-bonds with nearby proximal residues or the protein backbone. From the inverse relationship between the Fe-CO and C-O stretching mode frequencies, it is evident that these modes are coupled. Depending on the change in the net coupling between these two modes, the vibrational frequencies for ν(Fe-CO) and ν(C-O) will shift accordingly. Thus, geometric distortions in either the Fe-Nim or Fe-CO bonds are likely to alter this vibrational coupling, modulating the Fe dπ → CO π* backdonation and influencing the magnitude of the ν(Fe-CO) and ν(C-O) shifts.
Furthermore, the heme conformation also appears to be affected by the introduction of the H102G mutation. NMR structures of the CO-bound WT H-NOX and homologous H103G mutant from Shewanella oneidensis (So H-NOX) were solved by Erbil et al. (7), and found notable ensemble differences between the two structures which suggested changes in the van der Waals contacts between the protein and the heme. The majority of these shifts occurred in the secondary structure region flanking the proximal pocket. Erbil et al. also showed that the heme deformation changed between the So H-NOX WT and H103G mutant structures, which is consistent with the RR spectra for the CO-bound Tt H-NOX WT and H102G variants. The significant changes observed in both frequency and RR intensity for the Tt H-NOX H102G mutants with bulkier proximal ligands [e.g. 2-Me-Im, 4(5)-Me-Im, 4(5)-Br-Im, and 4(5)-I-Im] are suggestive of a 5-coordinate CO complex, with a ν(Fe-CO) that is similar to previously reported 5-coordinate CO complexes (19, 20, 29, 31, 51). Notable differences in the RR spectra for these CO complexes with bulkier imidazole derivatives include increased Raman intensities in the 270–430 cm−1 range, a 9–11 cm−1 downshift in ν8 from 350 cm−1 in the WT protein, as well as changes in both the RR intensities and frequencies for several heme skeletal markers in the 1365–1635 cm−1 region (Figure 3, panels A and B). In the low frequency region, the increased Raman intensities of the porphyrin-substituent and pyrrole deformation modes (e.g. ν52, γ7, γ6) are consistent with a distortion of the heme as the Fe comes out of the porphyrin plane in an alternate 5-coordinate complex. The domed 5-coordinate porphyrin structure would also have longer Fe-Npyr distances, which is supported by the 9–11 cm−1 downshift in ν8 that is observed for the Tt H-NOX H102G mutants containing bulkier proximal ligands. In the high frequency region, the increased frequencies and intensities observed in the heme skeletal markers are also consistent with a smaller macrocycle cavity resulting from a domed, outof-plane Fe atom. These differences are further corroborated by electronic absorption spectra of the H102G mutants with bulkier imidazole derivatives, which showed Soret blue shifts of up to 12 nm from the typical ~423–424 nm position (Supporting Information, Table S2). Thus, it is possible that changing the proximal ligand from imidazole to derivatives containing substituents forces the heme into an alternate conformation that would elongate the Fe-Nim bond distance and form a transient 5-coordinate CO species, causing more pronounced changes in ν(Fe-CO) and ν(C-O) than previously observed with Mb H93G and with other H-NOX domains containing Fe-His bonds of varying strengths (1, 11, 12, 38, 48).
Influence of the Ile-5 and Pro-115 residues on heme conformation in CO adducts
To further examine the effects of heme deformation on the CO complex, mutations were generated at Ile-5 and Pro-115, which are known to influence the heme conformation of the O2-bound WT Tt H-NOX domain (3, 4, 41). Figure 5 shows the RR spectra for the CO complexes of Tt H-NOX WT, I5L, P115A, and the double mutant. In the low frequency region, the Fe-CO stretching mode is detected at 490 cm−1 for the I5L mutant, and at 488 cm−1 for both the P115A and I5L/P115A mutants, in comparison to 490 cm−1 for WT. In the high frequency region, a single C-O stretching mode is detected between 1981–1986 cm−1 for the three mutants [ν(C-O) ~ 1989 cm−1 for WT]. As is the case with the heme skeletal modes, these measured frequencies are also very similar to those of Tt H-NOX WT.
Figure 5.

Resonance Raman spectra of the CO complexes of Tt H-NOX WT, I5L, P115A, and I5L/P115A. 13CO spectra (dotted line) are overlaid upon the 12CO spectra to indicate the frequency shifts upon isotopic substitution, and the difference (13CO-12CO) spectra are shown below each protein for clarity. Insets show the isotopic shift trends for ν(C–O) in the 1850–2000 cm−1 region. Spectral intensities were normalized to ν7 and ν4 for the low and high frequency regions, respectively.
Our RR spectra agree with the DFT calculations of Xu et al. in that there is minimal effect of mutating the residues responsible for heme distortion on the Fe-CO and C-O stretching modes in Tt H-NOX (10). Although the Ile-5 and Pro-115 mutations have been shown to slightly alter the Fe-His strength and influence the heme conformation for the O2 complexes (41), both the Fe-CO and C-O frequencies remain remarkably similar to the WT protein at ~490 and ~1985 cm−1, respectively. In addition, the skeletal heme modes do not appreciably shift in frequency, nor do the heme deformation modes display changes in RR intensity. Thus, heme distortion does not appear to affect the Fe(II)-CO or C-O bonds, as the heme-CO adduct retains binding site interactions similar to those of the WT protein.
Importance of the conserved YxSxR motif
The YxSxR motif is strictly conserved throughout the H-NOX family (1–3, 7) and provides the anchor for the heme cofactor within the pocket. The crystal structure of WT Tt H-NOX shows that Tyr-131 and Ser-133 are within H-bonding distance to one of the heme propionates, while Arg-135 makes contacts with both propionate groups (Figure 1B). Recent DFT calculations by Xu et al. indicated that the neutralization of these negatively charged propionate groups on the heme could modulate the ν(Fe-CO)/ν(CO) backbonding correlations in a manner similar to electron-withdrawing substituents on porphines, which altered the ν(C-O) frequency by up to ~20 cm−1 (10). Thus, to investigate how the H-bonding of these residues with the propionate substituents influenced the CO complex, mutations were generated at Tyr-131, Ser-133, and Arg-135 of Tt H-NOX.
Figure 6 shows the RR spectra of the CO complexes for Tt H-NOX WT, Y131F, S133A, S133C, and R135Q. In all of these mutants, two isotope-sensitive bands at ~490 and ~569 cm−1 were detected in the low frequency region, which correspond to ν(Fe-CO) and δ(Fe-C-O), respectively. No significant frequency shifts were observed in these bands, although a slight spectral broadening of the Fe-CO stretching mode was evident for the Tt H-NOX Y131F, S133C, and R135Q mutants (Supporting Information, Figure S5). The overall similarity in vibrational frequency indicates that the Fe-CO bond energetics between the WT protein and the four mutants are comparable, whereas the slight broadening of the Fe-CO stretching mode would suggest a larger range of fluctuations in the softened distal pocket. In contrast, the RR spectra showed significant changes in the ν(C-O) frequency for these YxSxR mutants in Tt H-NOX [i.e. up to ~70 cm−1 shift for ν(C-O)], in addition to surprising dynamics within the heme pocket for the Y131F, S133C, and R135Q mutants, which have not been previously reported. These results provide direct evidence for the importance of the YxSxR motif in stabilizing the heme position and orientation in the H-NOX pocket when the protein is in solution (Figure 1B).
Mutation of Tyr-131 to phenylalanine in Tt H-NOX results in the presence of two dominant CO conformers (Figure 6). One form exhibits vibrational features similar to the WT protein, with ν(Fe-CO) and ν(C-O) at ~492 cm−1 and 1984 cm−1, respectively. The second conformer displays a significantly lower ν(C-O) at 1919 cm−1, yet no other isotope-sensitive band was detected in the lower frequency region for the Fe-CO stretching mode. This mutation abolishes one of the H-bonding contacts between the YxSxR motif and a propionate group. Similarly, the S133C and R135Q mutants display two ν(C-O) bands, but only one dominant ν(Fe-CO) band. Whereas S133C disrupts only one H-bonding contact to one propionate group, the R135Q mutation disrupts H-bonding interactions with both propionate groups. The presence of two CO stretching frequencies, one of which is very similar to the WT protein, may result from a fluctuating swivel motion of the heme prosthetic group upon destabilization of the coordinating H-bonding residues. It is plausible that the heme interchanges between a conformation that retains similar protein-ligand contacts as Tt H-NOX WT, and another conformation with a very strong H-bond to the CO ligand based on the significantly reduced ν(C-O) frequency of ~1919–1920 cm−1. The source of this strong H-bond could be a water molecule that enters the distal pocket upon mutation due to increased accessibility, although another possibility is that the fluctuating heme motion alters the interaction between Tyr-140 and the CO moiety into one that is H-bonding. The lack of a more appreciable shift in the Fe-CO stretching mode, despite such a large change in the C-O stretching frequency, indicates that there are likely two counterbalancing processes that influence the Fe-CO bond strength. DFT calculations have shown that the transition moment essentially lies along the vector formed by the Fe-C bond, so C-O stretching vibrations induce electrons to travel throughout the π system (8, 20, 52, 53). Thus, it is not unreasonable to consider a redistribution of the heme π-conjugation system upon disruption of the neutralizing H-bonds between the propionate groups and the YxSxR motif, which could affect the Fe-CO bond strength.
In contrast, only a single CO conformation is observed in the Tt H-NOX S133A mutant with similar Fe-CO and C-O stretching frequencies to those of the WT protein. Ser-133 is sandwiched between Tyr-131 and Arg-135 in the YxSxR motif, and is only within H-bonding distance of one of the propionate groups. It is possible that the propionate groups still retained electrostatic interactions with the other two residues; thus, the heme remains in its tightly held position within the H-NOX pocket as in the WT protein. This interpretation is further supported by the spectral similarities between the CO complexes of Tt H-NOX WT and S133A in both the vibrational frequencies of the skeletal modes and the relative intensities of the low frequency out-of-plane modes (i.e. γ15 ~706 cm−1). In contrast, the Tt H-NOX Y131F, S133C, and R135Q mutants all exhibit small shoulder bands [e.g. δ(Cβ-vinyl) and δ(Cβ-CH3)] or slightly decreased RR intensity with broad shoulder peaks (i.e. γ15 ~703–707 cm−1) in comparison to the WT protein (Supporting Information, Figures S5–S7); this is indicative of alternate heme conformations and/or interactions within the protein pocket.
Modulation of the CO bond in sGC and the H-NOX family
It is clear from the RR studies on Tt H-NOX and its mutants that a number of factors can influence and likely contribute to the ν(Fe-CO) and ν(C-O) frequencies in heme proteins (Figure 9). Within the H-NOX family, the YxSxR motif is conserved throughout this class of heme proteins, and clearly plays an important role in retaining the position of the heme molecule within the protein pocket, as is evident from the RR spectra presented here. Disruption of the tightly held H-bonds between these residues and the heme propionate groups results in altered protein-heme and protein-ligand interactions that can modulate the properties of the H-NOX domain. Not only would electron density from the H-bonds between the YxSxR motif and the propionate groups be redistributed to the heme π-conjugated macrocycle upon neutralizing the negative charges of the porphyrin substituents, but the increased electron density at the metal center would also increase the FeII dπ → CO π* donation. The subsequent heme swiveling motion within the binding site of these YxSxR mutants would also introduce the H-NOX domain to alternate interactions that are different from the WT protein, such as an H-bonding interaction between the CO ligand and distal Tyr-140 residue in Tt H-NOX, which can further modulate the Fe dπ → CO π* back-donation. It is well known that electron donating and withdrawing groups, whether from porphyrin substituents or nearby polar residues, influence the extent of backbonding in heme proteins (8, 9, 11, 12, 14–21, 23, 24, 48, 53, 54). As the backbonding increases, the Fe-CO stretching frequency rises and the C-O stretching frequency falls. Thus, the YxSxR motif is a critical feature throughout the H-NOX family, not only for the maintenance of both proximal and distal contacts with the heme-CO adduct, but for its electrostatic interactions with the porphyrin propionate substituent groups that also influence the ν(C-O) frequency.
Figure 9.

Schematic representation of the heme pocket of Tt H-NOX. The different components of the Tt H-NOX binding site that can influence the FeII dπ → CO π* backdonation are shown in this diagram: (i) the H-bonding network composed of Trp-9, Asn-74, and Tyr-140 affects the polarity of the distal pocket where CO is bound, (ii) modulation of the axial trans ligand strength can influence how the proximal ligand competes with CO for σ-orbital overlap with Fe, and (iii) H-bonding interactions between the conserved YxSxR motif and heme propionate groups can withdraw electron density from the heme.
Previous RR studies on sGC and the truncated β1 heme domain showed the presence of two CO conformers based on the detection of two Fe-CO stretching frequencies at ~478 and ~487–496 cm−1 (27–30, 33, 36–38). The majority of these results only reported one stretching frequency for ν(C-O); this discrepancy is likely due to the weak intensity inherent to the C-O stretching mode in RR spectra and the difficulty in obtaining large quantities of full length sGC. Still, an interesting aspect of these prior investigations was that the heme conformer population would shift in favor of one conformation [ν(Fe-CO) ~492 cm−1] over the other [ν(Fe-CO) ~474 cm−1] upon addition of the benzylindazole derivative, YC-1 [3-(5'-hydroxylmethyl-3'-furyl)-1-benzylindazole]. Although much is still unknown about how this small molecule causes the 6-coordinate CO-bound sGC to mimic the fully activated 5-coordinate NO-bound enzyme, it may involve the electrostatic interactions between the conserved YxSxR motif and the propionate side chains of the heme.
Our RR results with the YxSxR mutants of Tt H-NOX show remarkable vibrational shifts that are indicative of significant changes in the binding site interactions; thus, these residues may also play an important role in modulating the activation of CO-bound sGC by altering the heme conformation from inactive to active. In addition, our observation that electrostatic changes within the conserved YxSxR motif are coupled to changes in ligand binding and coordination at the Fe center could suggest a much broader role for this motif in the transmission of signal from the heme in the H-NOX domain to associated signaling components, such as the methyl-accepting chemotaxis proteins or histidine kinases. It is clear that there is interplay between ligand binding and the electrostatic interactions of the motif with the heme propionate groups. Since these residues are strictly conserved throughout the H-NOX family, it is plausible that the YxSxR motif may also be important for signal transduction involving H-NOX domains upon binding NO or O2.
Supplementary Material
ACKNOWLEDGMENTS
We thank Dr. Hans K. Carlson for his help with preparing some of the protein samples for RR measurements, members of the Mathies and Marletta laboratories for discussions, and Patrick Visperas for preliminary characterization of the YxSxR Tt H-NOX mutants. We also thank Dr. Mohammed Ibrahim for providing helpful feedback on the RR spectra.
Footnotes
SUPPORTING INFORMATION AVAILABLE Supporting material for this work include tabulated summaries of the electronic absorption characterization of the FeII-unligated (Table S1) and CO-bound (Table S2) Tt H-NOX, and observed vibrational frequencies for the 5-coordinate, FeII-unligated (Table S3) and CO-bound (Table S4) H-NOX proteins. In addition, charts showing correlation trends for ν(Fe-Nim), ν(Fe-CO) and ν(C-O) in CO-bound Tt H-NOX for distal pocket H-bonding mutants (Figure S1), H102G mutant with alternate proximal ligands (Figure S2), YxSxR mutants (Figure S3); and CO-bound myoglobin with distal mutants (Figure S4) are provided for clarity. Expanded views of different frequency regions of the resonance Raman spectra for the CO-bound YxSxR motif mutants is also provided (Figures S5–S7). This material is available free of charge via the Internet at http://pubs.acs.org.
Abbreviations: H-NOX, Heme-Nitric Oxide/Oxygen binding domain; RR, resonance Raman; Tt, Thermoanaerobacter tengcongensis; YxSxR, Tyrosine-Serine-Arginine motif; NO, nitric oxide; CO, carbon monoxide; TEA, triethanolamine; DFT, density functional theory.
This work was supported in part by NIH Grant GM070671 (M.A.M.) and the Mathies Royalty Fund.
REFERENCES
- 1.Karow DS, Pan D, Tran R, Pellicena P, Presley A, Mathies RA, Marletta MA. Spectroscopic characterization of the soluble guanylate cyclase-like heme domains from Vibrio cholerae and Thermoanaerobacter tengcongensis. Biochemistry. 2004;43:10203–10211. doi: 10.1021/bi049374l. [DOI] [PubMed] [Google Scholar]
- 2.Boon EM, Marletta MA. Ligand specificity of H-NOX domains: from sGC to bacterial NO sensors. J. Inorg. Biochem. 2005;99:892–902. doi: 10.1016/j.jinorgbio.2004.12.016. [DOI] [PubMed] [Google Scholar]
- 3.Pellicena P, Karow DS, Boon EM, Marletta MA, Kuriyan J. Crystal structure of an oxygen-binding heme domain related to soluble guanylate cyclases. Proc. Natl. Acad. Sci. U.S.A. 2004;101:12854–12859. doi: 10.1073/pnas.0405188101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Olea C, Boon EM, Pellicena P, Kuriyan J, Marletta MA. Probing the Function of Heme Distortion in the H-NOX Family. ACS Chem. Biol. 2008;3:703–710. doi: 10.1021/cb800185h. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Nioche P, Berka V, Vipond J, Minton N, Tsai AL, Raman CS. Femtomolar sensitivity of a NO sensor from Clostridium botulinum. Science. 2004;306:1550–1553. doi: 10.1126/science.1103596. [DOI] [PubMed] [Google Scholar]
- 6.Ma XL, Sayed N, Beuve A, van den Akker F. NO and CO differentially activate soluble guanylyl cyclase via a heme pivot-bend mechanism. EMBO J. 2007;26:578–588. doi: 10.1038/sj.emboj.7601521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Erbil WK, Price MS, Wemmer DE, Marletta MA. A structural basis for HNOX signaling in Shewanella oneidensis by trapping a histidine kinase inhibitory conformation. Proc. Natl. Acad. Sci. U.S.A. 2009;106:19753–19760. doi: 10.1073/pnas.0911645106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kozlowski PM, Vogel KM, Zgierski MZ, Spiro TG. Steric contributions to CO binding in heme proteins: a density functional analysis of FeCO vibrations and deformability. Journal of Porphyrins and Phthalocyanines. 2001;5:312–322. [Google Scholar]
- 9.Decatur SM, Boxer SG. A Test of the Role of Electrostatic Interactions in Determining the Co Stretch Frequency in Carbonmonoxymyoglobin. Biochemical and Biophysical Research Communications. 1995;212:159–164. doi: 10.1006/bbrc.1995.1950. [DOI] [PubMed] [Google Scholar]
- 10.Xu CL, Ibrahim M, Spiro TG. DFT analysis of axial and equatorial effects on Heme-CO vibrational modes: Applications to CooA and H-NOX heme sensor proteins. Biochemistry. 2008;47:2379–2387. doi: 10.1021/bi702254y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Decatur SM, Belcher KL, Rickert PK, Franzen S, Boxer SG. Hydrogen bonding modulates binding of exogenous ligands in a myoglobin proximal cavity mutant. Biochemistry. 1999;38:11086–11092. doi: 10.1021/bi9908888. [DOI] [PubMed] [Google Scholar]
- 12.Decatur SM, DePillis GD, Boxer SG. Modulation of protein function by exogenous ligands in protein cavities: CO binding to a myoglobin cavity mutant containing unnatural proximal ligands. Biochemistry. 1996;35:3925–3932. doi: 10.1021/bi952625t. [DOI] [PubMed] [Google Scholar]
- 13.Kerr EA, Mackin HC, Yu NT. Resonance Raman Studies of Carbon-Monoxide Binding to Iron Picket Fence Porphyrin with Unhindered and Hindered Axial Bases -an Inverse Relationship between Binding-Affinity and the Strength of Iron Carbon Bond. Biochemistry. 1983;22:4373–4379. doi: 10.1021/bi00288a005. [DOI] [PubMed] [Google Scholar]
- 14.Kushkuley B, Stavrov SS. Theoretical study of the distal-side steric and electrostatic effects on the vibrational characteristics of the FeCO unit of the carbonylheme proteins and their models. Biophys. J. 1996;70:1214–1229. doi: 10.1016/S0006-3495(96)79680-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Li TS, Quillin ML, Phillips GN, Olson JS. Structural Determinants of the Stretching Frequency of Co Bound to Myoglobin. Biochemistry. 1994;33:1433–1446. doi: 10.1021/bi00172a021. [DOI] [PubMed] [Google Scholar]
- 16.Li XY, Spiro TG. Is Bound Co Linear or Bent in Heme-Proteins - Evidence from Resonance Raman and Infrared Spectroscopic Data. J. Am. Chem. Soc. 1988;110:6024–6033. doi: 10.1021/ja00226a017. [DOI] [PubMed] [Google Scholar]
- 17.Phillips GN, Teodoro ML, Li TS, Smith B, Olson JS. Bound CO is a molecular probe of electrostatic potential in the distal pocket of myoglobin. J. Phys. Chem. B. 1999;103:8817–8829. [Google Scholar]
- 18.Ramsden J, Spiro TG. Resonance Raman Evidence That Distal Histidine Protonation Removes the Steric Hindrance to Upright Binding of Carbon-Monoxide by Myoglobin. Biochemistry. 1989;28:3125–3128. doi: 10.1021/bi00434a001. [DOI] [PubMed] [Google Scholar]
- 19.Ray GB, Li XY, Ibers JA, Sessler JL, Spiro TG. How Far Can Proteins Bend the Feco Unit - Distal Polar and Steric Effects in Heme-Proteins and Models. J. Am. Chem. Soc. 1994;116:162–176. [Google Scholar]
- 20.Spiro TG, Wasbotten IH. CO as a vibrational probe of heme protein active sites. J. Inorg. Biochem. 2005;99:34–44. doi: 10.1016/j.jinorgbio.2004.09.026. [DOI] [PubMed] [Google Scholar]
- 21.Spiro TG, Zgierski MZ, Kozlowski PM. Stereoelectronic factors in CO, NO and O-2 binding to heme from vibrational spectroscopy and DFT analysis. Coordination Chemistry Reviews. 2001;219:923–936. [Google Scholar]
- 22.Tsubaki M, Srivastava RB, Yu NT. Resonance Raman Investigation of Carbon-Monoxide Bonding in (Carbon Monoxy) Hemoglobin and (Carbon Monoxy) Myoglobin -Detection of Fe-Co Stretching and Fe-C-O Bending Vibrations and Influence of the Quaternary Structure Change. Biochemistry. 1982;21:1132–1140. doi: 10.1021/bi00535a004. [DOI] [PubMed] [Google Scholar]
- 23.Vogel KM, Kozlowski PM, Zgierski MZ, Spiro TG. Determinants of the FeXO (X = C, N, O) vibrational frequencies in heme adducts from experiment and density functional theory. J. Am. Chem. Soc. 1999;121:9915–9921. [Google Scholar]
- 24.Vogel KM, Kozlowski PM, Zgierski MZ, Spiro TG. Role of the axial ligand in heme-CO backbonding; DFT analysis of vibrational data. Inorganica Chimica Acta. 2000;297:11–17. [Google Scholar]
- 25.Yu NT, Kerr EA, Ward B, Chang CK. Resonance Raman Detection of Fe-Co Stretching and Fe-C-O Bending Vibrations in Sterically Hindered Carbonmonoxy Strapped Hemes - a Structural Probe of Fe-C-O Distortion. Biochemistry. 1983;22:4534–4540. doi: 10.1021/bi00288a028. [DOI] [PubMed] [Google Scholar]
- 26.Stone JR, Marletta MA. Soluble Guanylate-Cyclase from Bovine Lung -Activation with Nitric-Oxide and Carbon-Monoxide and Spectral Characterization of the Ferrous and Ferric States. Biochemistry. 1994;33:5636–5640. doi: 10.1021/bi00184a036. [DOI] [PubMed] [Google Scholar]
- 27.Denninger JW, Schelvis JPM, Brandish PE, Zhao Y, Babcock GT, Marletta MA. Interaction of soluble guanylate cyclase with YC-1: Kinetic and resonance Raman studies. Biochemistry. 2000;39:4191–4198. doi: 10.1021/bi992332q. [DOI] [PubMed] [Google Scholar]
- 28.Li ZQ, Pal B, Takenaka S, Tsuyama S, Kitagawa T. Resonance Raman evidence for the presence of two heme pocket conformations with varied activities in CO-bound bovine soluble guanylate cyclase and their conversion. Biochemistry. 2005;44:939–946. doi: 10.1021/bi0489208. [DOI] [PubMed] [Google Scholar]
- 29.Makino R, Obayashi E, Homma N, Shiro Y, Hori H. YC-1 facilitates release of the proximal his residue in the NO and CO complexes of soluble guanylate cyclase. J. Biol. Chem. 2003;278:11130–11137. doi: 10.1074/jbc.M209026200. [DOI] [PubMed] [Google Scholar]
- 30.Pal B, Li ZQ, Ohta T, Takenaka S, Tsuyama S, Kitagawa T. Resonance Raman study on synergistic activation of soluble guanylate cyclase by imidazole, YC-1 and GTP. J. Inorg. Biochem. 2004;98:824–832. doi: 10.1016/j.jinorgbio.2003.12.007. [DOI] [PubMed] [Google Scholar]
- 31.Ibrahim M, Derbyshire ER, Marletta MA, Spiro TG. Probing Soluble Guanylate Cyclase Activation by CO and YC-1 Using Resonance Raman Spectroscopy. Biochemistry. 2010;49:3815–3823. doi: 10.1021/bi902214j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Martin E, Czarnecki K, Jayaraman V, Murad F, Kincaid J. Resonance Raman and infrared spectroscopic studies of high-output forms of human soluble guanylyl cyclase. J. Am. Chem. Soc. 2005;127:4625–4631. doi: 10.1021/ja0440912. [DOI] [PubMed] [Google Scholar]
- 33.Pal B, Tanaka K, Takenaka S, Kitagawa T. Resonance Raman spectroscopic investigation of structural changes of CO-heme in soluble guanylate cyclase generated by effectors and substrate. J. Raman Spectrosc. 2010;41:1178–1184. [Google Scholar]
- 34.Pal B, Kitagawa T. Binding of YC-1/BAY 41–2272 to soluble guanylate cyclase: A new perspective to the mechanism of activation. Biochemical and Biophysical Research Communications. 2010;397:375–379. doi: 10.1016/j.bbrc.2010.05.122. [DOI] [PubMed] [Google Scholar]
- 35.Deinum G, Stone JR, Babcock GT, Marletta MA. Binding of nitric oxide and carbon monoxide to soluble guanylate cyclase as observed with resonance Raman spectroscopy. Biochemistry. 1996;35:1540–1547. doi: 10.1021/bi952440m. [DOI] [PubMed] [Google Scholar]
- 36.Schelvis JPM, Zhao Y, Marletta MA, Babcock GT. Resonance Raman characterization of the heme domain of soluble guanylate cyclase. Biochemistry. 1998;37:16289–16297. doi: 10.1021/bi981547h. [DOI] [PubMed] [Google Scholar]
- 37.Zhao Y, Schelvis JPM, Babcock GT, Marletta MA. Identification of histidine 105 in the beta 1 subunit of soluble guanylate cyclase as the heme proximal ligand. Biochemistry. 1998;37:4502–4509. doi: 10.1021/bi972686m. [DOI] [PubMed] [Google Scholar]
- 38.Karow DS, Pan DH, Davis JH, Behrends S, Mathies RA, Marletta MA. Characterization of functional heme domains from soluble guanylate cyclase. Biochemistry. 2005;44:16266–16274. doi: 10.1021/bi051601b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Weinert EE, Plate L, Whited CA, Olea C, Marletta MA. Determinants of Ligand Affinity and Heme Reactivity in H-NOX Domains. Angew Chem Int Edit. 2010;49:720–723. doi: 10.1002/anie.200904799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Boon EM, Davis JH, Tran R, Karow DS, Huang SH, Pan D, Miazgowicz MM, Mathies RA, Marletta MA. Nitric oxide binding to prokaryotic homologs of the soluble guanylate cyclase β1 H-NOX domain. J. Biol. Chem. 2006;281:21892–21902. doi: 10.1074/jbc.M600557200. [DOI] [PubMed] [Google Scholar]
- 41.Tran R, Boon EM, Marletta MA, Mathies RA. Resonance Raman Spectra of an O2-Binding H-NOX Domain Reveal Heme Relaxation upon Mutation. Biochemistry. 2009;48:8568–8577. doi: 10.1021/bi900563g. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kim M, Mathies RA, Hoff WD, Hellingwerf KJ. Resonance Raman Evidence That the Thioester-Linked 4-Hydroxycinnamyl Chromophore of Photoactive Yellow Protein Is Deprotonated. Biochemistry. 1995;34:12669–12672. doi: 10.1021/bi00039a024. [DOI] [PubMed] [Google Scholar]
- 43.Hu SZ, Smith KM, Spiro TG. Assignment of protoheme Resonance Raman spectrum by heme labeling in myoglobin. J. Am. Chem. Soc. 1996;118:12638–12646. [Google Scholar]
- 44.Tomita T, Gonzalez G, Chang AL, Ikeda-Saito M, Gilles-Gonzalez MA. A comparative resonance Raman analysis of heme-binding PAS domains: heme iron coordination structures of the BjFixL, AxPDEA1, EcDos, and MtDos proteins. Biochemistry. 2002;41:4819–4826. doi: 10.1021/bi0158831. [DOI] [PubMed] [Google Scholar]
- 45.Andrew CR, Green EL, Lawson DM, Eady RR. Resonance Raman studies of cytochrome c ' support the binding of NO and CO to opposite sides of the heme: Implications for ligand discrimination in heme-based sensors. Biochemistry. 2001;40:4115–4122. doi: 10.1021/bi0023652. [DOI] [PubMed] [Google Scholar]
- 46.Aono S, Kato T, Matsuki M, Nakajima H, Ohta T, Uchida T, Kitagawa T. Resonance Raman and ligand binding studies of the oxygen-sensing signal transducer protein HemAT from Bacillus subtilis. J. Biol. Chem. 2002;277:13528–13538. doi: 10.1074/jbc.M112256200. [DOI] [PubMed] [Google Scholar]
- 47.Tani F, Matsu-ura M, Ariyama K, Setoyama T, Shimada T, Kobayashi S, Hayashi T, Matsuo T, Hisaeda Y, Naruta Y. Iron twin-coronet porphyrins as models of myoglobin and hemoglobin: Amphibious electrostatic effects of overhanging hydroxyl groups for successful CO/O2 discrimination. Chem-Eur J. 2003;9:862–870. doi: 10.1002/chem.200390096. [DOI] [PubMed] [Google Scholar]
- 48.Franzen S, Boxer SG, Dyer RB, Woodruff WH. Resonance Raman studies of heme axial ligation in H93G myoglobin. J. Phys. Chem. B. 2000;104:10359–10367. [Google Scholar]
- 49.Kitagawa T. The Heme Protein Structure and the Iron-Histidine Stretching Mode. In: Spiro TG, editor. Biological Applications of Raman Spectroscopy: Resonance Raman Spectra of Heme and Metalloproteins. John Wiley & Sons; New York: 1988. pp. 97–131. [Google Scholar]
- 50.Vojtechovsky J, Chu K, Berendzen J, Sweet RM, Schlichting I. Crystal structures of myoglobin-ligand complexes at near-atomic resolution. Biophys. J. 1999;77:2153–2174. doi: 10.1016/S0006-3495(99)77056-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Yu NT, Kerr EA. Vibrational Modes of Coordinated CO, CN−, O2, and NO. In: Spiro TG, editor. Biological Applications of Raman Spectroscopy: Resonance Raman Spectra of Heme and Metalloproteins. John Wiley & Sons; New York: 1988. pp. 39–95. [Google Scholar]
- 52.Spiro TG, Kozlowski PM. Discordant results on FeCO deformability in heme proteins reconciled by density functional theory. J. Am. Chem. Soc. 1998;120:4524–4525. [Google Scholar]
- 53.Spiro TG, Kozlowski PM. Is the CO adduct of myoglobin bent, and does it matter? Accounts of Chemical Research. 2001;34:137–144. doi: 10.1021/ar000108j. [DOI] [PubMed] [Google Scholar]
- 54.Springer BA, Sligar SG, Olson JS, Phillips GN. Mechanisms of Ligand Recognition in Myoglobin. Chemical Reviews. 1994;94:699–714. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.



