Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Oct 1.
Published in final edited form as: J Inorg Biochem. 2012 Jun 1;115:204–210. doi: 10.1016/j.jinorgbio.2012.05.013

Ambidentate H-bonding of NO and O2 in Heme Proteins

Thomas G Spiro 1,*, Alexandra V Soldatova 1
PMCID: PMC3463650  NIHMSID: NIHMS395810  PMID: 22824153

Abstract

The affinity and reactivity of the gaseous molecules CO, NO and O2 (XO) in heme protein adducts are controlled by secondary interactions, especially by H-bonds donated from distal protein residues. Vibrational spectroscopy, supported by DFT modeling, has revealed that for NO and O2, but not for CO, a critical issue is whether the H-bond is donated to the outer or inner atom of the bound diatomic ligand. DFT modeling shows that bound NO and O2 are ambidentate, both atoms separately acting as H-bond acceptors. This is not the case for CO, whose π* orbital acts as a delocalized H-bond acceptor. Vibrational spectra of heme-XO adducts reveal a general pattern of backbonding variations, marked by families of negative correlations between frequencies associated with Fe-X and X-O bond stretches. For heme-CO adducts, H-bonding increases backbonding, the νFeX/νXO points moving up the backbonding correlation established with model compounds. For NO and O2 adducts, however, increased backbonding is only observed when the outer atom is the H-bond acceptor. H-bonding to the inner (X) atom instead produces a positive νFeX/νXO correlation. This effect can be reproduced by DFT modeling. Its mechanism is polarization of the sp2 orbital on the X atom, on the back side of the bent FeXO unit, drawing electrons from both the Fe-X and X-O bonds and weakening them together. Thus, the positioning of H-bond donors in the protein differentially affects bonding and reactivity in heme adducts of NO and O2.

Keywords: Heme, Nitric oxide, Dioxygen, H-bonding, Backbonding, DFT, Raman

1. Introduction

Heme proteins are nature’s receptors for the gaseous molecules CO, NO and O2 (XO), which play key roles in metabolism and in signaling. Different heme proteins respond very differently to the presence of these molecules, despite having a common prosthetic group, a Fe-porphyrin, generally protoporphyrin IX. The protein responses differ because of differences in axial ligation by a protein proximal residue, generally histidine or cysteine, and in a variety of secondary interactions, especially H-bonding, with residues in the heme pocket.

H-bonding to the bound XO molecules affects their stability and reactivity. For example, the CO/O2 affinity ratio is markedly reduced in myoglobin or hemoglobin, relative to protein-free heme, because the H-bond donated by a distal histidine residue is much stronger for bound O2 than for bound CO.[1,2] The reactivity of bound NO or O2 is also strongly affected by H-bonding. Thus, in cytochrome P450, which activates O2 for substrate oxygenation, the reaction path is critically dependent on H-bonding from a distal threonine residue.[3,4]

In this article, we focus on the issue of which atom, X or O, gets the H-bond, and what the consequences are. For NO and O2, but not for CO, the outer and inner atoms can both accept H-bonds, with different effects on the electronic structure of the FeXO adduct. These effects have been explored through vibrational spectroscopy and supported by DFT computations[56], (A.V. Soldatova, T.G. Spiro, manuscript in preparation).

2. Backbonding Pattern of Vibrational Frequencies

The vibrational spectra of heme XO adducts contain modes which have dominant contributions from Fe-X and X-O bond stretching coordinates, and are labeled νFeX and νXO. When their frequencies are plotted against one another for a series of 5-coordinate Fe(II) porphyrin XO adducts (Figure 1), the result is a set of negative linear correlations. This is the expected result for variations in backbonding. The filled Fe(II) dπ orbitals are well matched to the empty XO π* orbitals, and the resulting back-donation of electrons strengthens the Fe-X bond while weakening the X-O bonds (Figure 2), although to differing extents in the three FeXO adducts, due to their differing electronic and geometric structures (see below). The backbonding correlations were established with TPP-Y porphyrins (Figure 3) bearing variably electron-donating or -withdrawing substituents, which modulated back-donation through their inductive effect.[710] More electron-donating substituents increased back-donation, and shifted the νFeX/νXO point higher on the backbonding line. For different XO, the lines are shifted from one another because the Fe-X and X-O bonds have different intrinsic strengths for the different XO molecules. DFT computations, using a Fe(II)P-Y (porphine, P, bearing electron-donating and -withdrawing substituents, Y) (Figure 3) confirm that Fe-X and X-O bond distances and frequencies are negatively correlated for all the FeXO adducts[7, 8], (A.V. Soldatova, M. Ibrahim, T.G. Spiro, manuscript in preparation).

Figure 1.

Figure 1

Negative frequency correlations for vibrational modes associated with Fe-X and X-O bond stretching are seen for 5-coordinate Fe(II) porphyrin complexes of all three X-O molecules, X = C, N, O. See text for references to the data.

Figure 2.

Figure 2

Schematic diagram of geometric and electronic structure of Fe(II) porphyrin XO adducts, showing the orbitals available for π backbonding.

Figure 3.

Figure 3

Structural diagram for porphyrins used to establish Fe(II)XO backbonding correlations via substituent variation, experimentally via RR spectroscopy (left: FeTPP-Y), and computationally via DFT (right: FeP-Y)

The universality of the backbonding pattern is remarkable because the FeXO adducts have different electronic and geometric structures.[1114] In CO, both π* orbitals are empty, and FeCO is a linear adduct, with dπ - π* overlaps in both perpendicular directions (x and y, with z as the FeCO axis, Figure 2). However, NO and O2 have one and two π* electrons, and their adducts are forced to bend (~145° for FeNO, ~120° for FeO2) in order to accommodate a π-antibonding interaction. The bonding can be thought of in terms of sp2-hybridization at the X atom, with one or two electrons in the lobe pointing away from the direction of bending (Figure 2). (NO can also form Fe(III)NO adducts, which are isoelectronic with Fe(II)CO). However, the πy* orbital remains available for backbonding from the Fe dyz orbital in FeNO and FeO2. Moreover, the π* orbital energies decrease from CO to NO to O2. Thus, backdonation to πy* increases for FeNO and FeO2, somewhat compensating the loss of π bonding in the x direction.

In FeO2, backdonation results in nearly complete transfer of an electron, and the electronic structure is effectively Fe(III)(O2); computations yield an open shell singlet ground state, with antiferromagnetic coupling between the spins on Fe(III) and O2.[1317] A weak π-bond between singly occupied orbitals, Fe-dyz and O2y*, limits backdonation in Fe(III)(O2);[14] as a result, νXO is less responsive to changes in νFeX for FeO2 than for the other XO adducts, as can be seen in the much larger slope of the FeO2 backbonding line (Figure 1).

That FeNO and FeO2 vibrational frequencies follow the backbonding pattern is also remarkable in that ‘νFeX’ is not a simple Fe-X stretching vibration. Because these adducts are bent, the Fe-X stretching and Fe-X-O bending coordinates mix in the vibrational modes, something that is symmetry-forbidden for the linear FeCO adduct. The mode chosen as ‘νFeX’ is the one with the largest shift upon isotopic substitution at X.[8] While this mode has a major contribution from Fe-X stretching, Fe-X-O bending also contributes.[8,18] Despite this ambiguity, the backbonding lines are similar for all the adducts, albeit with different slopes, indicating that the coordinate mixing is constant, or else changes in proportion to the extent of backbonding.

In 6-coordinate FeXO adducts, the backbonding pattern is altered by the axial ligand, trans to the XO.[19] The axial ligand increases the electron density on the Fe, increasing backdonation and weakening the X-O bond. However, the expected Fe-X strengthening due to backbonding is offset by the σ competition between XO and the trans ligand for the Fe dz2 acceptor orbital, and the overall effect is Fe-X weakening. This effect is demonstrated by vibrational data for the Fe(II)(CO)TPP-Y adducts when 4-methylimidazole (4-MeImH – a surrogate for the histidine sidechain) is added as an axial ligand (Figure 4).[10,20] A new backbonding correlation is obtained, with νFeX shifted to lower values, as expected for a weaker Fe-X bond.

Figure 4.

Figure 4

Experimental correlations of νFeX and νXO for 5- and 6-coordinate CO and NO adducts of FeTPP-Y. Axial ligation shifts νFeX down for CO but up for NO, although the Fe-X bond is weakened in both cases. The anomalous νFeN upshift reflects an increased contribution of Fe-N-O bending to the ‘νFeN’ vibrational mode (see text).

A shifted backbonding line is also obtained from vibrational data on 6-coordinate (4-MeImH)Fe(II)(NO)TPP-Y adducts (Figure 4) but in this case, the νFeX values are higher, not lower, than for the 5-coordinate adducts, despite the weakening of the Fe-X bond which is observed and calculated [8]. This anomaly is a manifestation of altered coordinate mixing in the ‘νFeX’ mode, the issue mentioned above. In the 6-coordinate adducts, the contribution of Fe-X-O bending to this mode is increased [8], pushing the frequency higher. A similar effect is evident for 6- vs 5-coordinate Fe(II)O2 adducts [21], although the data are too limited to determine a 6-coordinate backbonding line [9].

Most of the vibrational data come from resonance Raman spectroscopy, because resonance with the porphyrin π-π* transition (the intense Soret band near 400 nm is generally utilized) enhances electronically coupled vibrations, including the Fe-X-O vibrations, without interference from background vibrations (solvent, substituents or protein.) These background vibrations obscure most regions of the infrared spectrum. However, νCO vibrations occur in an uncluttered region of the spectrum, and can be monitored via infrared spectroscopy.[2227]

3. H-bonding Effects

Heme protein vibrational spectroscopy, supported by DFT modeling, has led to key insights regarding distal H-bonding to bound XO molecules. For Fe(II)CO adducts, DFT modeling of distal H-bond donors indicates that the preferred orientation of the H-bond is towards the middle of the CO bond [28]. This is indeed the orientation of the distal imidazole in the CO adduct of myoglobin (Mb - Figure 5).[29] The CO π* orbitals, partially occupied by backbonding, act as delocalized acceptors for the H-bond. The effect of the H-bond is to polarize the Fe(II)CO unit, enhancing back-bonding. Accordingly, νFeC shifts up, while νCO shifts down. Indeed, for MbCO variants having distal residue replacements (Figure 6) that produce variable distal H-bonds, the νFeC/νCO points track the same backbonding correlation as do model compounds with variably electron donating substituents (Figure 7) [8,10]. The points shift higher on the backbonding line for stronger H-bonding, or other positively polar interactions, and lower for weaker H-bonding and for negatively polar interactions.

Figure 5.

Figure 5

Overlay of MbXO structures (gray, MbCO, pdb #: 1BZR; blue, MbNO, pdb #: 2FRJ; red, MbO2, pdb #: 1A6M), showing the sideways contact of the distal histidine with the bound XO.

Figure 6.

Figure 6

MbCO structure (pdb #:1A6G) showing the distal residues which have been mutated to investigate the effects of distal electric fields and H-bonds on MbXO adducts.

Figure 7.

Figure 7

Contrasting response of FeCO and FeNO adducts to distal H-bonds in heme proteins. Varying the H-bond strength shifts the νFeX/νXO point along the backbonding line established with model FeTPP-Y adducts with CO, and also with NO, provided that the H-bond is directed toward ONO. A positive correlation is seen if the H-bond is directed toward NNO (see text for references).

However, for both Fe(II)NO and Fe(II)O2, modeling of a distal H-bond donor demonstrates XO ambidenticity. A stable energy structure is obtained for H-bonding to either O or to X. Figure 8 shows the computed structures for (ImH)Fe(II)NO porphine having the H-bond through either N or O.[5] Similar structures are also computed for the O2 analog (A.V. Soldatova, T.G. Spiro, manuscript in preparation). The νFeX/νXO frequencies behave quite differently in these adduct, depending on whether H-bond is directed toward X or O, as computation reveals (vide infra).

Figure 8.

Figure 8

Ambidentate H-bonding by Fe(II)NO. Stable structures are obtained via DFT when a H-bond donor (HCl) is directed at either ONO or NNO, of NO bound to imidazole-ligated FeP [5].

3.1 Fe(II)NO

When νFeX/νXO data are plotted for heme protein Fe(II)NO adducts, the results are very different than for the Fe(II)CO adducts (Figure 7). The points can be seen to describe a fan. The top of the fan is the backbonding correlation obtained for the model compounds (4-MeImH)Fe(II)(NO)TPP-Y (Figure 4).[8] Some of the heme proteins fall on this backbonding line, but many do not.

DFT modeling provides the key to understanding the experimental fan. When H-bond donors of differing strength are allowed to interact with the bound NO, the effect on bonding depends on whether the interaction is with ONO or NNO (Figure 8).[5] H-bonding to ONO increases backbonding; the Fe-N distance decreases while the N-O distance increases and the associated frequencies shift in the opposite directions (Figure 9). A negative frequency correlation is obtained, much like the top of the experimental fan. However, H-bonds directed at NNO produce positive correlations between Fe-N and N-O distances and frequencies (Figure 9). For both sets of correlations, frequencies and distances obeyed Badger’s rule [30], despite the coordinate mixing in the ‘νFeN’ mode, because H-bonding has little effect on the mode composition.[5]

Figure 9.

Figure 9

DFT modeling [5] confirms that Fe-N and N-O bond distances and frequencies correlate negatively or positively when H-bond donors of varying strength are directed, respectively, at ONO or NNO of (ImH)FeP(NO).

The physical mechanism of the positive distance and frequency correlations is that H-bonding to NNO draws electrons from both the Fe-N and N-O bonds into the singly occupied non-bonding sp2 orbital on NNO (Figure 2); the Fe-N and N-O bonds therefore weaken together. The increased occupancy of the sp2 orbital should also decrease the Fe-N-O angle, and the computed angle does indeed decrease [5]. Although the decrease was only modest, 3°, between the weakest and strongest H-bonds (NH3 and NH4+), H-bonding to NNO was also found to soften the Fe-N-O bending potential significantly.

A similarly modest increase in the Fe-N-O angle was computed when the H-bond was directed at ONO. [5]. In this case, the H-bond draws electron density out of the sp2 orbital into the NO π* orbital, increasing both the Fe-N-O angle and the extent of backbonding, thus, accounting for the negative distance and frequency correlations.

The heme protein Fe(II)NO adducts that fall on the backbonding correlation, at the top of the fan in Figure 7, likely have ONO-directed H-bonds, if any. In the case of the heme cd1 nitrite reductase (NiR), whose point falls well up on the backbonding line, a strong H-bond from a distal histidine residue to ONO is evident in the crystal structure.[31]

The bottom of the fan is a line defined by the set of Mb variants having residue substitution distal to the heme (Figure 6).[9,32] The points along the line represent variants with different degrees of H-bonding to the bound NO. Wild-type Mb, having a distal histidine (H64), is in the middle of the line, but when it is replaced by hydrophobic residues (H64I, H64V), the point is close to the TPP-Y backbonding line. In between is H64Q, since the glutamine replacement forms a weaker H-bond than histidine. The point moves down the line from the wild-type position in the L29F variant, because the leucine at this position buttresses H64, and its replacement with the bulkier phenylalanine strengthens the H-bond. At the low end of the H-bonding line is the V68N variant; the valine in this position abuts the bound NO, and its replacement by asparagine introduces a strong H-bond.

For Fe(II)CO adducts, these same variably H-bonding Mb variants simply move the νFeX/νXO up and down the backbonding line defined by the TPP-Y models,[9] but for the Fe(II)NO they define an entirely different line, with a positive slope. The reason is that the distal histidine in Mb is oriented toward the side of the bound NO (Figure 5), with the NH group closer to NNO than to ONO.[3335] For the CO adduct, this orientation is of little consequence. The H-bond acceptor is the delocalized CO π* orbital, and H-bonding enhances backdonation, regardless of orientation. But for NO, the sideways interaction with NNO is critical.

There are four reported Mb(II)NO crystal structures, differing in the method of preparation, and in the source of the Mb.[3335] They all show distinctly shorter distances from the nearest N atom of the distal histidine to NNO than to ONO (Table 1). Interestingly, two of the reported Fe-N-O angles, 121° [35] and 112° [34] are much smaller than the other two, 147° [33] and 144°[35], and they are associated with significantly shorter (His64)N···NNO distances. This trend is as expected from the DFT computations. The extent of the extra Fe-N-O bending is surprisingly large, but may reflect the expected softening of the bending potential and the moderate resolution of the crystal structures (1.3 – 1.9 Å). The reason for the stronger H-bonds in the two highly bent structures is unclear, but perhaps reflects different distributions of solvent molecules or salt ions, stemming from different crystallization conditions.

Table 1.

Selected parameters for MbXO crystal structures.

Structure FeXO (deg) (His64)N···X (Å) (His64)N···O (Å) Reference
MbCOswa 171 3.59 3.21 [29]
MbNOhhb 147 2.98 3.18 [33]
MbNOhhc 144 3.05 3.34 [35]
MbNOhhd 121 2.72 3.33 [35]
MbNOswe 112 2.78 3.38 [34]
MbOOswf 122 3.08 2.97 [64]

pdb #:

a

1BZR (at 1.15 Å);

b

1NPF (at 1.90 Å);

c

2FRJ (at 1.30 Å);

d

2FRK (at 1.30 Å);

e

1HJT (at 1.70 Å);

f

1A6M (at 1.00 Å).

H-bonding to NNO can also have biochemical consequences, by modulating the reactivity of the bound NO. For example, it can facilitate reduction to HNO, a species of considerable biological interest.[3638] By pulling electron density into the sp2 orbital on NNO, and weakening the Fe-N and N-O bonds, the H-bond raises the Fe(II/I) reduction potential, and is itself poised to protonate the developing NO. HNO intermediates have been implicated in the enzymatic mechanisms of NO synthase [39], and of nitrite [40] and NO [41] reductases.

It is notable that reduction of Mb(II)NO is facile, and results in a stable HNO adduct,[42,43] although HNO itself is highly unstable. As expected, the Fe-N and N-O bonds are longer in the HNO adduct than in the NO adduct.[44,45] The reduction mechanism can be readily envisioned as complete transfer of the distal histidine NH proton to NNO, concerted with electron transfer from the reductant to the NO adduct (proton coupled electron transfer - Figure 10). Subsequent histidine re-protonation from the solvent, and reorientation of the HNO adduct would produce a new H-bond, now to the O atom of HNO, which NMR analysis [46] indicates to be a key aspect of Mb(HNO) stability.

Figure 10.

Figure 10

Suggested mechanism for the facile reduction of Mb(NO) to Mb(HNO) [5].

3.2 Fe(II)O2

The study of FeO2 vibrations poses special problems, because the adducts are prone to oxidation. The few 5-coordinate data, plotted in Figure 1 are from low-temperature and matrix isolation experiments, or from superstructured porphyrins that afford steric protection. Most data are from heme proteins, which sequester the FeO2 adducts, protecting them from rapid oxidation.

Another serious difficulty is that the O-O stretching mode, νOO, is normally not detectable in RR (resonance Raman) spectra, due to weak enhancement. And while the νOO mode is active in IR spectra, its frequency, ~1200 cm−1, places it in a spectral region that is crowded with porphyrin and protein vibrations, making detection and assignment difficult [47]. However, there are circumstances under which νOO does become detectable in RR spectra, presumably reflecting altered electronic coupling, although the mechanisms are currently uncertain.[48] RR enhancement of νOO is found when the axial ligand is thiolate; data are available for a number of cysteinyl heme proteins and model complexes.[4953]

RR enhancement of the νOO band can also be observed for some imidazole-ligated heme proteins that have distal H-bonds to the bound O2.[48,5458] A number of νFeO/νOO data pairs have been reported for such proteins, and are plotted in Figure 11. They describe a fan of points similar to that found for Fe(II)NO adducts (Figure 6). A clear connection to distal H-bond strength has been demonstrated for the points at the bottom of the fan, which belong to variants of Ctb [54,57], a truncated hemoglobin from the bacterium C. jejuni. The distal pocket of Ctb contains tryptophan, tyrosine and histidine residues, all within H-bond distance of bound heme ligands, and with one another. The wild-type protein occupies the lowest point of the νFeO/νOO fan, but when the H-bond donors are successively replaced by residues with non-polar sidechains, the Ctb points move steadily up a positive correlation, defining the bottom of the fan (Figure 11).

Figure 11.

Figure 11

RR data on O2 adducts for a number of imidazole-ligated heme proteins are suggestive of ambidentate distal H-bonding, showing behavior similar to NO adducts (Figure 7). For the Ctb truncated Hb [54,57], a positive correlation with the strength of distal H-bonds is revealed by site mutants, consistent with H-bonding to Op.

The similarity of the νFeX/νXO fans for Fe(II)NO and Fe(II)O2 suggests that, as in the former case, H-bonding to the terminal, or outer, atom, Ot, of Fe(II)O2, enhances backbonding, and leads to a negative νFeO/νOO correlation (top of the fan), while H-bonding to the proximal, or inner, atom, Op, weakens the Fe-O as well as the O-O bond, and produces a positive correlation (bottom of the fan). Unfortunately, structural data of high enough resolution to discern the dominant H-bond contacts are sparse. In Ctb, the positive correlation suggests Op as the dominant receptor. A crystal structure is available only for the cyanide adduct.[59] However, the distal tryptophan, whose mutation produces the largest νFeO/νOO shift, appears to be properly oriented to H-bond primarily with Op. Also consistent with the Op/Ot dichotomy is the position of hemoglobin, the highest point at the top of the fan [6062], whose crystal structure reveals short H-bonds to Ot [63]. Points falling within the fan (Figure 11) likely reflect situations in which H-bonds to both O atoms are of comparable importance.

DFT modeling designed to test the Op/Ot dichotomy (A.V. Soldatova, T.G. Spiro, manuscript in preparation), confirmed that H-bond donors of variable strength increase the Fe-O and O-O distances in concert when they are directed at Op, but decrease the Fe-O distance while increasing the O-O distance (backbonding) when they are directed at Ot (Figure 12). The mechanism is the same as for Fe(II)NO adducts, although the positive slope is larger (2.0 vs 1.5), and the negative slope is smaller (−0.31 vs −1.0), for Fe(II)O2. Vibrational mode computations for the Fe(II)O2 adducts are currently in progress (A.V. Soldatova, T.G. Spiro, manuscript in preparation).

Figure 12.

Figure 12

DFT modeling establishes that Fe-O and O-O bond distances correlate negatively or positively when H-bond donors of varying strength are directed, respectively, at Ot or Op of (ImH)FeP(O2).

Conclusions

Backbonding is the dominant motif in Fe(II)XO (X = C, N, O) porphyrin adducts, as revealed by families of νFeX/νXO negative correlations, the lines being modulated by the nature of X and by effects of trans ligands. The degree of backbonding is controlled by inductive effects of porphyrin substituents and by polar residues near the bound XO, including H-bond donors. For NO and O2, a straightforward backbonding effect is seen only if the H-bond acceptor is the outer atom of XO. H-bonding to the inner atom, X, weakens the Fe-X and X-O bonds in concert, by polarizing electrons in the sp2 orbital on the back side of X. The positioning of H-bond donors by the protein differentially controls bonding and reactivity in heme–NO and –O2 adducts.

Highlights.

  • DFT modeling together with available vibrational experimental data showed differential response of Fe(II)XO (XO = NO and O2) heme to the position of H-bond donors.

  • H-bonding to Fe(II)XO from heme pocket residues increases backbonding only if the H-bond acceptor is the outer atom of XO.

  • H-bonding to the inner atom, X, of the Fe(II)XO adduct weakens the Fe-X and X-O bonds in concert due to enhanced orbital mixing.

Acknowledgments

We dedicate this article to the memory of Hans Freeman, a guiding light of bioinorganic chemistry. This work was supported by NIH Grant GM 33567.

Abbreviations

IR

infrared

RR

resonance Raman

NMR

nuclear magnetic resonance

DFT

density functional theory

H-bond

hydrogen bond

Mb

myoglobin

Hb

hemoglobin

HbN,HbO

hemoglobin I & II from Mycobacterium tuberculosis

ChlHb

Chlamydomonas hemoglobin

ScHb

Synechocystis hemoglobin

HemDGC

heme-containing diguanylate cyclase

Ctb

hemoglobin III from Campylobacter jenjuni

NiR

nitrite reductase

AXCP

cytochrome c′ from Alcaligenes xylosoxidans

BjFixLH

Bradyrhozobium japonicum FixL heme PAS

MtDosH

Methanobacterium thermoautotrophicum Dos heme-PAS

TtTar4H

Thermoanaerobacter tengcongensis Tar4 protein heme domain

TPP,TMP, OEP

tetraphenyl-, tetramethyl- & octaethyl- porphyrins

Pc

phthalocyanine

ImH

MeImH, imidazole, 4-methylimidazole

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Olson JS, Phillips GN. J Biol Inorg Chem. 1997;2:544–552. [Google Scholar]
  • 2.Springer BA, Sligar SG, Olson JS, Phillips GN., Jr Chem Rev. 1994;94:699–714. [Google Scholar]
  • 3.Martinis SA, Atkins WM, Stayton PS, Sligar SG. J Am Chem Soc. 1989;111:9252–9253. [Google Scholar]
  • 4.Imai M, Shimada H, Watanabe Y, Matsushimahibiya Y, Makino R, Koga H, Horiuchi T, Ishimura Y. Proc Natl Acad Sci USA. 1989;86:7823–7827. doi: 10.1073/pnas.86.20.7823. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Xu C, Spiro TG. J Biol Inorg Chem. 2008;13:613–621. doi: 10.1007/s00775-008-0349-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Tangen E, Svadberg A, Ghosh A. Inorg Chem. 2005;44:7802–7805. doi: 10.1021/ic050486q. [DOI] [PubMed] [Google Scholar]
  • 7.Vogel KM, Kozlowski PM, Zgierski MZ, Spiro TG. J Am Chem Soc. 1999;121:9915–9921. [Google Scholar]
  • 8.Ibrahim M, Xu C, Spiro TG. J Am Chem Soc. 2006;128:16834–16845. doi: 10.1021/ja064859d. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Spiro TG, Ibrahim M, Wasbotten IH. In: The Smallest Biomolecules: Diatomics and Their Interactions with Heme Proteins. Ghosh A, editor. Elsevier; Amsterdam: 2008. pp. 95–123. [Google Scholar]
  • 10.Spiro TG, Wasbotten IH. J Inorg Biochem. 2005;99:34–44. doi: 10.1016/j.jinorgbio.2004.09.026. and references therein. [DOI] [PubMed] [Google Scholar]
  • 11.Spiro TG, Zgierski MZ, Kozlowski PM. Coord Chem Rev. 2001;219:923–936. [Google Scholar]
  • 12.Ghosh A. Acc Chem Res. 2005;38:943–954. doi: 10.1021/ar050121+. [DOI] [PubMed] [Google Scholar]
  • 13.Chen H, Ikeda-Saito M, Shaik S. J Am Chem Soc. 2008;130:14778–14790. doi: 10.1021/ja805434m. [DOI] [PubMed] [Google Scholar]
  • 14.Shaik S, Chen H. J Biol Inorg Chem. 2011;16:841–855. doi: 10.1007/s00775-011-0763-1. [DOI] [PubMed] [Google Scholar]
  • 15.Jensen KP, Ryde U. J Biol Chem. 2004;279:14561–14569. doi: 10.1074/jbc.M314007200. [DOI] [PubMed] [Google Scholar]
  • 16.Rovira C, Kunc K, Hutter J, Ballone P, Parrinello M. J Phys Chem A. 1997;101:8914–8925. [Google Scholar]
  • 17.Sigfridsson E, Ryde U. J Inorg Biochem. 2002;91:101–115. doi: 10.1016/s0162-0134(02)00426-9. [DOI] [PubMed] [Google Scholar]
  • 18.Lehnert N, Galinato MGI, Paulat F, Richter-Addo GB, Sturhahn W, Xu N, Zhao JY. Inorg Chem. 2010;49:4133–4148. doi: 10.1021/ic902181e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Oertling WA, Kean RT, Wever R, Babcock GT. Inorg Chem. 1990;29:2633–2645. [Google Scholar]
  • 20.Vogel KM, Kozlowski PM, Zgierski MZ, Spiro TG. Inorg Chim Acta. 2000;297:11–17. [Google Scholar]
  • 21.Liu Y, Sun H. J Comput Chem. 2011;32:1279–1285. doi: 10.1002/jcc.21709. [DOI] [PubMed] [Google Scholar]
  • 22.Barlow CH, Ohlsson PI, Paul KG. Biochemistry. 1976;15:2225–2229. doi: 10.1021/bi00655a031. [DOI] [PubMed] [Google Scholar]
  • 23.Yoshikawa S, Choc MG, Otoole MC, Caughey WS. J Biol Chem. 1977;252:5498–5508. [PubMed] [Google Scholar]
  • 24.Caughey WS, Shimada H, Choc MG, Tucker MP. Proc Natl Acad Sci USA. 1981;78:2903–2907. doi: 10.1073/pnas.78.5.2903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Paul KG, Smith ML, Ohlsson PI. Inorg Chim Acta. 1983;79:169–169. [Google Scholar]
  • 26.Li XY, Spiro TG. J Am Chem Soc. 1988;110:6024–6033. doi: 10.1021/ja00226a017. [DOI] [PubMed] [Google Scholar]
  • 27.Brown WE, Sutcliffe JW, Pulsinelli PD. Biochemistry. 1983;22:2914–2923. doi: 10.1021/bi00281a021. [DOI] [PubMed] [Google Scholar]
  • 28.Franzen S. J Am Chem Soc. 2002;124:13271–13281. doi: 10.1021/ja017708d. [DOI] [PubMed] [Google Scholar]
  • 29.Kachalova GS, Popov AN, Bartunik HD. Science. 1999;284:473–476. doi: 10.1126/science.284.5413.473. [DOI] [PubMed] [Google Scholar]
  • 30.Badger RM. J Chem Phys. 1935;3:710–714. [Google Scholar]
  • 31.Das TK, Wilson EK, Cutruzzola F, Brunori M, Rousseau DL. Biochemistry. 2001;40:10774–10781. doi: 10.1021/bi002327i. [DOI] [PubMed] [Google Scholar]
  • 32.Coyle CM, Vogel KM, Rush TS, Kozlowski PM, Williams R, Spiro TG, Dou Y, Ikeda-Saito M, Olson JS, Zgierski MZ. Biochemistry. 2003;42:4896–4903. doi: 10.1021/bi026395b. [DOI] [PubMed] [Google Scholar]
  • 33.Copeland DM, West AH, Richter-Addo GB. Proteins Struct Funct Gen. 2003;53:182–192. doi: 10.1002/prot.10495. [DOI] [PubMed] [Google Scholar]
  • 34.Brucker EA, Olson JS, Ikeda-Saito M, Phillips GN. Proteins: Struct, Funct, Gen. 1998;30:352–356. [PubMed] [Google Scholar]
  • 35.Copeland DM, Soares AS, West AH, Richter-Addo GB. J Inorg Biochem. 2006;100:1413–1425. doi: 10.1016/j.jinorgbio.2006.04.011. [DOI] [PubMed] [Google Scholar]
  • 36.Paolocci N, Katori T, Champion HC, St John ME, Miranda KM, Fukuto JM, Wink DA, Kass DA. Proc Natl Acad Sci USA. 2003;100:5537–5542. doi: 10.1073/pnas.0937302100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Miranda KM. Coord Chem Rev. 2005;249:433–455. [Google Scholar]
  • 38.Farmer PJ, Sulc F. J Inorg Biochem. 2005;99:166–184. doi: 10.1016/j.jinorgbio.2004.11.005. [DOI] [PubMed] [Google Scholar]
  • 39.Rusche KM, Spiering MM, Marletta MA. Biochemistry. 1998;37:15503–15512. doi: 10.1021/bi9813936. [DOI] [PubMed] [Google Scholar]
  • 40.Einsle O, Messerschmidt A, Huber R, Kroneck PMH, Neese F. J Am Chem Soc. 2002;124:11737–11745. doi: 10.1021/ja0206487. [DOI] [PubMed] [Google Scholar]
  • 41.Averill BA. Chem Rev. 1996;96:2951–2964. doi: 10.1021/cr950056p. [DOI] [PubMed] [Google Scholar]
  • 42.Lin R, Farmer PJ. J Am Chem Soc. 2000;122:2393–2394. [Google Scholar]
  • 43.Sulc F, Immoos CE, Pervitsky D, Farmer PJ. J Am Chem Soc. 2004;126:1096–1101. doi: 10.1021/ja0376184. [DOI] [PubMed] [Google Scholar]
  • 44.Immoos CE, Sulc F, Farmer PJ, Czarnecki K, Bocian DF, Levina A, Aitken JB, Armstrong RS, Lay PA. J Am Chem Soc. 2005;127:814–815. doi: 10.1021/ja0433727. [DOI] [PubMed] [Google Scholar]
  • 45.Zeng WQ, Silvernail NJ, Wharton DC, Georgiev GY, Leu BM, Scheidt WR, Zhao JY, Sturhahn W, Alp EE, Sage JT. J Am Chem Soc. 2005;127:11200–11201. doi: 10.1021/ja051052x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Sulc F, Fleischer E, Farmer PJ, Ma DJ, La Mar GN. J Biol Inorg Chem. 2003;8:348–352. doi: 10.1007/s00775-002-0422-7. [DOI] [PubMed] [Google Scholar]
  • 47.Tsuchida E, Komatsu T, Nakata T, Hasegawa E, Nishide H, Inoue H. J Chem Soc Dalton Trans. 1991:3285–3289. [Google Scholar]
  • 48.Das TK, Couture M, Ouellet Y, Guertin M, Rousseau DL. Proc Natl Acad Sci USA. 2001;98:479–484. doi: 10.1073/pnas.98.2.479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Chartier FJM, Blais SP, Couture M. J Biol Chem. 2006;281:9953–9962. doi: 10.1074/jbc.M513893200. [DOI] [PubMed] [Google Scholar]
  • 50.Hu SZ, Schneider AJ, Kincaid JR. J Am Chem Soc. 1991;113:4815–4822. [Google Scholar]
  • 51.Bangcharoenpaurpong O, Rizos AK, Champion PM, Jollie D, Sligar SG. J Biol Chem. 1986;261:8089–8092. [PubMed] [Google Scholar]
  • 52.Sjodin T, Christian JF, Macdonald ID, Davydov R, Unno M, Sligar SG, Hoffman BM, Champion PM. Biochemistry. 2001;40:6852–6859. doi: 10.1021/bi002510b. [DOI] [PubMed] [Google Scholar]
  • 53.Chottard G, Schappacher M, Ricard L, Weiss R. Inorg Chem. 1984;23:4557–4561. [Google Scholar]
  • 54.Lu C, Egawa T, Wainright LM, Poole RK, Yeh SR. J Biol Chem. 2007;282:13627–13636. doi: 10.1074/jbc.M609397200. [DOI] [PubMed] [Google Scholar]
  • 55.Lu C, Egawa T, Batabyal D, Mukai M, Yeh SR. In: The Smallest Biomolecules: Diatomics and Their Interactions with Heme Proteins. Ghosh A, editor. Elsevier; Amsterdam: 2008. pp. 235–266. [Google Scholar]
  • 56.Mukai M, Savard P-Y, Ouellet H, Guertin M, Yeh SR. Biochemistry. 2002;41:3897–3905. doi: 10.1021/bi0156409. [DOI] [PubMed] [Google Scholar]
  • 57.Manez PA, Lu C, Boechi L, Marti MA, Shepherd M, Wilson JL, Poole RK, Luque FJ, Yeh SR, Estrin DE. Biochemistry. 2011;50:3946–3956. doi: 10.1021/bi101137n. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Sawai H, Yoshioka S, Uchida T, Hyodo M, Hayakawa Y, Ishimori K, Aono S. Biochim Biophys Acta. 2010;1804:166–172. doi: 10.1016/j.bbapap.2009.09.028. [DOI] [PubMed] [Google Scholar]
  • 59.Nardini M, Pesce A, Labarre M, Richard C, Bolli A, Ascenzi P, Guertin M, Bolognesi M. J Biol Chem. 2006;281:37803–37812. doi: 10.1074/jbc.M607254200. [DOI] [PubMed] [Google Scholar]
  • 60.Hirota S, Ogura T, Appelman EH, Shinzawaitoh K, Yoshikawa S, Kitagawa T. J Am Chem Soc. 1994;116:10564–10570. [Google Scholar]
  • 61.Potter WT, Tucker MT, Houtchens RA, Caughey WS. Biochemistry. 1987;26:4699–4707. doi: 10.1021/bi00389a016. [DOI] [PubMed] [Google Scholar]
  • 62.Bruha A, Kincaid JR. J Am Chem Soc. 1988;110:6006–6014. doi: 10.1021/ja00226a014. [DOI] [PubMed] [Google Scholar]
  • 63.Park S-Y, Yokoyama T, Shibayama N, Shiro Y, Tame JR. J Mol Biol. 2006;360:690–701. doi: 10.1016/j.jmb.2006.05.036. [DOI] [PubMed] [Google Scholar]
  • 64.Vojtechovsky J, Chu K, Berendzen J, Sweet RM, Schlichting I. Biophys J. 1999;77:2153–2174. doi: 10.1016/S0006-3495(99)77056-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES