A comparative study of O2, CO and CN binding to metalloporphyrins
Francisco Torrens
Institut Universitari de Ciència Molecular, Universitat de València, Dr. Moliner 50, E‑46100 Burjassot (València), Spain. Fax: 34 9 6354 3156; Tel: 34 9 6354 3182; E‑mail: Francisco.Torrens@uv.es; http://www.uv.es/~icmol
Parametrization of a molecular-mechanics program to include terms specific for five‑ and six-coordinate transition metal complexes results in computer-simulated structures of heme complexes. The principal new feature peculiar to five and six coordination is a term that measures the effect of electron-pair repulsion modified by the ligand electronegativity and takes into account the different structural possibilities. The model system takes into account the structural differences of the fixing centre in the haemoglobin subunits. The customary proximal histidine is added. The prosthetic group heme IX is wholly considered in our model. The calculations show clearly that certain conformations are much more favourable that others for fixing O2. From the O2 binding in haemoglobin, myoglobin and simple Fe porphyrin models it is concluded that the bent O2 ligand is best viewed as bound superoxide O2‑. Rotation of axial ligands are practically free. A small modification of the model in both crystal and protein matrix affects the orientation of the ligands in experimental systems.
Keywords: electron-pair repulsion, polarizing molecular mechanics, iron–porphyrin complex, oxygen fixation, CO/O2 discrimination.
Introduction
The heme prosthetic group is in the active centre of a number of relevant proteins as the oxygen (O2) transport proteins haemoglobin (Hb) and myoglobin (Mb) [1,2], as well as enzymes involved in catabolism as peroxidases [3], catalases, oxidases [4] and cytochromes [5]. The replacement of Fe by Mg in heme leads to chlorophyll [6], and the replacement of Fe by other transition metals coupled with modifications in the aromatic ring leads to species as vitamin B12 [7] and cofactor F‑430 [8,9].
Rohmer’s group
characterized the electronic state of Fe(P) (
porphyrin) complexes [10–12], and predicted an
electronic structure, which was later proved by experiment [13]. Another
theoretical studies are the real position of the CO group in Fe(P)(Im)(CO)
(Im = imidazole) [14,15], the position of the CN group in Fe(mdi)2(py)(CN)
(mdi = malondialdiminate) complexes [16], the role of distal and proximal
histidines (His) on the binding of O2 in Hb [17–20], and structural
aspects of the binding of O2 and other ligands to heme [21–27]. The
amount of information obtained from the calculations is seriously limited by
the size of the heme group, which has allowed only recently the appearance of
theoretical studies on reactivity [28–33]. The coordination of O2
to Fe(P)(Im) 5‑coordinate species leads to 6‑coordinate species with
octahedral geometry, i.e., the
biomimetic forms of Mb–O2 and Hb–O2. X-ray data were
reported only on two complexes: Fe(TpivPP)[1‑(Me)Im](O2)
[34] and Fe(TpivPP)[2‑(Me)Im](O2) [35]. Both are quite
similar, sharing the same porphyrin TpivPP =
meso‑tetrakis(a,a,a,a‑o‑pivalamidophenyl)porphyrin.
Maseras’ group optimized the geometry of Fe(TpivPP)[1‑(Me)Im](O2)
with hybrid quantum mechanics/molecular mechanics (QM/MM) IMOMM(DFT-B3LYP:MM3)
[36] and pure QM DFT-B3LYP [37]. Ghosh and Bocian optimized the geometry of
Fe(P)(Im)(CO) with density functional theory (DFT) [19]. Salzmann
et al. optimized under constraint
the geometry of a Fe(TpivPP)[1-(Me)Im](CO) model with DFT-B3LYP
[38]. Han et al. calculated a
heme model–CO system employing the ab
initio pseudopotential method with local density approximation
(LDA) exchange correlation [39].
The reversibe binding of O2 and carbon monoxide (CO) played a central role in studies of heme-protein structure and function. The affinity of Hb for CO is 200 times that for O2, which is a risk for CO poisoning. Numerous encumbered FeII porphyrin models were synthesized in an effort to elucidate the structural details of small ligand binding. The steric bulk of certain axial ligands bonded to synthetic FeII porphyrins provided model compounds of reduced O2 and CO affinity, and models of the tense (T) state of hemoproteins. There is only one single-crystal X‑ray structure determination on such a complex [40]. There was much discussion on the mechanistic basis of the variation of affinity values in heme proteins and model compounds. This focused on the nature of the axial ligand, distal steric effects, distal polar effects, and enforced doming and ruffling of the porphyrin skeleton. Johansson et al. showed by QM calculations on a haem a model that, upon reduction, the spin pairing at Fe is accompanied by effective delocalization of electrons, from the Fe towards the periphery of the porphyrin ring including its substituents [41,42]. In previous works, both non‑interacting (NID) and interacting (ID) induced-dipole polarization models were implemented in the program molecular mechanics (MM2) [43]. The polarizing force field (MMID2) [44] was applied to For‑Gly‑NH2 [45]. In the present report, terms specific for five‑ and six-coordination have been included in MMID2. The new program is called MMIDX. The next section presents the computational method. Following that the results are discussed. Finally, the conclusions are summarized.
Computational method
The molecular polarizability, aabmol, is defined as the linear response to an external electric field,
(1)
where maind is the induced molecular dipole moment. Considering a set of N interacting atomic polarizabilities, the atomic mind has a contribution also from the other atoms,
(2)
where Tpq,bg(2) is the interaction tensor as modified by Thole [46]
(3)
where vpq = rpq/spq if rpq < spq; otherwise vpq = 1, with s = 1.662(apaq)1/6. Molecular polarizability can be written as
(4)
where B is the relay matrix defined as (in a supermatrix notation)
(5)
1,3‑Interactions between atoms bonded to a common atom are not specifically included in Equation (6) in the MM2+polarization approach because they are effectively already included in the bond-length and bond-angle strainless parameters.
(6)
For 5‑coordinate structures, there is a need to consider the effect of 1,3‑interactions because an energy term is needed for ligand–ligand repulsion to account for (1) the stability difference of various possible geometries and (2) structural effects due to the variation of ligand electronegativity. The geometries of the methylfluorophosphoranes (CH3)nPF5‑n (n = 0®3) were qualitatively correlated with Gillespie’s VSEPR theory, which was adopted as a model for the present approach. Bond electron-pair repulsion (EPR) terms were introduced via the non‑bonded term Enb of Equation (6) modified to express EPR for atoms bonded to 5‑ or 6‑coordinated atoms. The unmodified term is
(7)
where P = [rVDW(A)+rVDW(B)]/rAB, rVDW is the van der Waals radius of the specific atom, and rAB is the non‑bonded distance between A and B; e = (eAeB)1/2, where eA and eB are parameters specific to atoms A and B, and are related to the hardness of the atoms. Hill evaluated the constants in Equation (7), which give the energy Enb(AB) in units of kilocalories per mole [47]. The modification of Equation (7) to express EPR terms is [48]
(8)
The addition of a scaling factor D, to obtain a suitable balance between this term and the other terms in Equation (6), and the replacement of P with P* (where P* = [rVDW(A)+rVDW(B)]/rAB* and rAB* is the distance between atoms A and B calculated from modified bond lengths, dCA* and dCB*, between the central atom C and either atom A or B) provide the necessary adjustments to quantitatively reproduce the (CH3)nPF5‑n structures [49]. The variation in ligand electronegativity is introduced by a distance factor RA in the relation dCA* = dCARA. The magnitude of R is inversely related to the electronegativity difference between atoms C and A. The R factors are the means of including EPR between atoms A and B. If the electronegativity difference DXCA is large, the bonding electron pair can be considered to move away from atom C, thus decreasing EPR between C–A and C–B bonds. When DXCA > DXCA’, EPR term E(1,3)AB < E(1,3)A’B even when the actual bond lengths are equal. A set of distance factors R may be obtained from the bond ionic character I,
(9)
and using the relation
(10)
where rA and rC are covalent radii of atoms A and C.
Calculation results and discussion
The structure of the model heme(‑His)‑O2 is shown in Figure 1. Heme is the prosthetic group of oxy‑Hb. The molecular mechanics calculations use the MM2/MMX+polarization force fields. Van der Waals parameters for the Fe atom have been taken from the UFF force field [50]. Torsional contributions involving dihedral angles with the metal atom and the bending terms involving Fe in central position have been set to zero.
Molecular mechanics dipole moments m are collected in Table 1. In general m increases with the oxidation state of Fe. In particular, the FeIII heme(‑His)‑CN m results the greatest due to the highly polar Fed+–C–Nd‑ complex. The FeIII heme(‑His)‑O2 m is relatively large due to the extremely polar Fed+–O–Od‑. The inclusion of polarization in MM2 corrects, in general, m in the correct direction when compared with MMX+ID. In particular for heme(‑His) the MM2+polarization m remains almost constant. The binding of His in heme increases m by a factor of 5. Moreover the subsequent binding of CN doubles m and the binding of O2 or CO increases m 7–16%.
The 4‑coordinate Fe(P) system
The heme group is a natural starting point for both experimental and theoretical studies. Crystal structures were reported for a number of heme derivatives with different substituents in the ring. In particular Fe(TPP) (TPP = meso‑tetraphenylporphyrin) has been chosen. Its electronic state is well known experimentally to correspond to a low spin triplet (S = 1). Selected structural parameters are collected in Table 2. The agreement in bond angles between both MM2/MMX+polarization geometries and the X‑ray structure is good, with discrepancies smaller than 5°. Differences in bond distances are larger in a number of cases as for the C–Cbridge, C–C’ and C’–C’’ distances. These have values of 1.395Å, 1.439Å and 1.365Å in X‑ray, and ca. 1.336Å, 1.340Å and 1.332Å in MM2/MMX+polarization, respectively. Agreement between experiment and MM2/MMX+polarization is good, ca. 0.06Å. All these atoms are in part purely described with MM2. The optimized MM2/MMX+polarization values are close to the optimal bond distance for these types of atoms in the applied force field, which is 1.337Å. Another discrepancy in the geometries appears in the Fe–N distance. This is more puzzling, because the calculated distance 1.877–1.881Å is smaller than the experimental value of 1.966Å.
The 5‑coordinate Fe(P)(Im) system
Coordination of an Im ligand to the heme group leads to a 5‑coordinate species with a square pyramidal geometry. These compounds are good biomimetic models of Mb and Hb, with Im replacing the proximal His of the biological systems. The need to avoid dimerization and formation of 6‑coordinate species with two axial ligands poses serious restrictions on the nature of the porphyrins able to give this kind of complexes. For this study, the species Fe(Piv2C8)[1‑(Me)Im] {Piv2C8 = a,a,5,15‑[2,2’‑(octanediamido)diphenyl]‑a,a,10,20‑bis(o‑pivalamidophenyl)porphyrin} has been chosen. This species has the advantage of having 1‑methylimidazole as axial ligand, in constrast with the more common 2‑methylimidazole, which is more sterically demanding. Neither for Fe(Piv2C8)[1‑(Me)Im] nor for other 5‑coordinate derivatives of heme the electronic state is experimentally known. Electronic spectroscopy, magnetic susceptibility and Mössbauer measurements identify it as high spin (S = 2). Selected parameters are resumed in Table 3. The Fe–Nporph distances are longer than those in the 4‑coordinate system by ca. 0.02Å (MMX+ID). This trend is in qualitative agreement with the reference values. The result is fully consistent with the shift from low spin to high spin in the metal. Overall agreement in the geometric parameters is correct. Moreover one has to take with suspicion the X‑ray parameters of Im, which would make the N=C double bond NIm–Ce in Im longer than the N–C single bond NIm–Cd. Furthermore the MM calculations are, in general, in agreement with the QM/MM reference, which provides the expected result.
The sharper discrepancy concerns the Nporph–Fe–NIm–Ce dihedral angle. This angle measures the rotation around the Fe–NIm single bond, and rules the placement of the Im plane with respect to the porphyrin ring. Its sign is arbitrary because the x and y directions are equivalent in absence of axial ligand. In this work, a positive sign has been chosen for consistence with data on the 6‑coordinate complexes presented below. An angle of 90° (like in the pure QM reference) means that the Im plane is eclipsing one of the Fe–Nporph bonds, while an angle of 135° (ca. the 135.2° in MMX) indicates a staggered orientation of Im with respect to the Fe–Nporph bonds. Therefore both pure QM and MMX values are just opposite with the experimental value, 126.0°, lying in between although closer to MMX results. The structural minima lead to structures where the Im ligand is located about the bisector of an angle Nporph–Fe–Nporph. The MMX+polarization results lie, in general, in the range 90–133° of the references. The importance of the large discrepancy between different values is, however, arguable because there is also a large dispersion in different experimental 5‑coordinate derivatives of heme, as well as in experimental reports of both Mb and Hb. The corresponding interpretation is that the rotation around this single bond has a very low barrier.
The 6‑coordinate Fe(P)(Im)(ligand) systems
Coordination of O2 to the 5‑coordinate heme–His species leads to 6‑coordinate species with octahedral geometry. These are the biomimetic forms of Mb–O2 and Hb–O2. X‑ray data are reported only on two complexes: Fe(TpivPP)[1‑(Me)Im](O2) and Fe(TpivPP)[2‑(Me)Im](O2). Both are quite similar, sharing the same porphyrin TpivPP, which is meso‑tetrakis(a,a,a,a‑o‑pivalamidophenyl)porphyrin. Fe(TpivPP)[1‑(Me)Im](O2), containing the less sterically demanding 1‑methylimidazole ligand, has been chosen for comparison. The state of this system is a low spin open-shell singlet (S = 1) resulting in a FeIII–O2‑ charge distribution. Selected parameters are reported in Table 4. The parameters concerning the coordination of O2, which are probably the most critical for the biochemical activity of Hb are well reproduced. The computed values for the Fe–O distance, 1.8–1.9Å, are close to the experimental value of 1.746Å. The calculated values for the O–O distance, 1.21–1.26Å, are far from the experimental report of 1.163Å. However the experimental value, even shorter than the 1.21Å for free O2 is suspect because of the disorder on the placement of the ending O atom within the crystal, as admitted by the authors of the X‑ray experiment [34]. The O–O interatomic distance increases from 1.21Å in free O2 to 1.261Å (MMX+NID), suggesting that electronic charge is transferred from FeP to O2, in qualitative agreement with the experimental observation that the Fe–O2 bond can be formally described as FeIII–O2‑ [9].
The most significant feature of the present structure is the bent Fe–O–O bond. The Fe–O–O bond angles are in all cases indicative of a bent h1 coordination mode, where only one O atom is directly attached to the metal. These calculations are in agreement with the experiment and calculation references. Atomic net charges have been calculated with our program POLAR [51]. The results, qFe = 3.078, qObegin = –0.412 and qOend = –0.531e are in agreement with Weiss’ model for the Fe–O2 bond in which the bond might be ionic between a Fe3+ and a superoxide ion, net charge being transferred from Fe to O2 (Fe3+O2–) [52]. Experimental support for Weiss’ model was advanced by Misra and Fridovich [53]. The geometry agrees with Pauling’s prediction of a bent FeO2 bond, and the O–O distance is close to that of 1.27Å, predicted by him [54]. It is slightly shorter than that of 1.34Å in O2– close to the electron spin resonance results, which show that no more than 2/3 of the density of one electron is transferred from the metal to the antibonding p* orbitals of O2.
The sharper discrepancy concerns the Nporph–Fe–O–O dihedral angle. This angle measures the rotation around the Fe–O single bond, and rules the placement of the Fe–O–O plane with respect to the porphyrin ring. An angle of 0° (ca. the –3.3° in MM2) means that the Fe–O–O plane is eclipsing one of the Fe–Nporph bonds, while an angle of –45° (ca. the –44.6° in pure QM) indicates a staggered orientation of the O2 with respect to the Fe–Nporph bonds. The structural minima lead to structures where the O2 ligand is located about the bisector of an angle Nporph–Fe–Nporph. The MMX/MMX+NID results lie near the reference results. The structural minimum correspond to a trans isomer, where the ending O is placed above the opposite quadrant where the Nd is located.
Structural parameters of Fe(P)(Im)(CO) are shown in Table 5. The geometric parameters concerning the coordination of CO, which are probably the most critical for the biochemical activity of Hb, are well reproduced. The computed values for the Fe–C distance, 1.97–2.01Å, are close to the experimental value of 1.793Å. The calculated values for the C–O distance, 1.11–1.14Å, are close to the experimental report of 1.095Å. The C–O distance is similar in the free molecule, 1.171Å, calculated with AM1 [55] and in Fe(P)(Im)(CO) 1.143Å (MMX). The corresponding interpretation is that electronic charge is not transferred from FeP to CO. This is in qualitative agreement with the experimental observation that the Fe–CO bond can be formally described as FeII–CO [9].
The most significant feature of the present structure is the linear Fe–C–O bond. While such a linear bond is to be expected based on the extensive literature on transition metal CO complexes, the result is nonetheless highly significant since bent Fe–C–O bonds with bond angles of 135–145º appear to be the rule in various CO complexes of hemoproteins [56–58]. The global energy minimum is linear for Fe–C–O (MM2/MMX+polarization). These calculations are in agreement with the experiment and calculation references. The Im ring is rotated so that the N atom is directed toward the Fe atom, and the rotation angle Nporph–Fe–NIm–Ce reaches ca. 176° (MM2+polarization) in conformity with the LDA reference (174.2°).
Selected bond lengths and angles are summarized in Table 6. The geometric parameters concerning the coordination of CN are well reproduced. The computed values for the Fe–C distance, 1.9–2.0Å, are in agreement with the experimental value of 1.930Å. The calculated values for the C–N distance, 1.16–1.20Å, are close from the experimental report of 1.150Å. The C–N distance increases from 1.147Å in the free molecule (AM1) to 1.202Å in Fe(P)(Im)(CN) (MMX+NID). The corresponding interpretation is that electronic charge is transferred from FeP to CN. This is in qualitative agreement with the experimental observation that the Fe–CN bond can be formally described as FeIII–(CN)‑ [9]. The most significant feature of the present structure is the linear Fe–C–N bond. The global energy minimum is linear for Fe–C–N (MM2/MMX+polarization). These calculations are in conformity with the experiment and calculation references.
Conclusions
From the preceding results the following conclusions can be drawn.
1. For the heme‑IX adducts, NID or ID polarization energy represents 73% of the total energy MM2+polarization. EPR corresponds to 68% of the total energy MMX+polarization.
2. Three different Fe-binding models are proposed for O2, CO and CN: bent superoxide FeIII–O2‑, linear FeII–CO and linear FeIII–CN‑. The nature of O2 binding in Hb, Mb and simple Fe–porphyrin models is becoming clear. When O2 is bound as a bent rather than a triangular ligand it is best described as bound superoxide. This bent geometry may be critical to biological functioning because it allows the discrimination between O2 and CO.
3. The distal His protects the heme FeII acting as a proton trap. The distal His has a pKa of ca. 5.5; at neutral pH it is protonated only at Nd, which faces the solvent. Any proton entering the heme pocket of Hb would be bound by Ne, and simultaneously Nd would release its proton to the solvent. When the His side chain swings out of the hemo pocket, the protons would interchange, restoring the previous state. No other amino acid side chain could function in this way.
4. The fact that in a close analogue of the CO complexes of the hemoproteins the Fe–C–O linkage is linear strongly suggests that another interpretation of the results of the protein studies is in order. The allegedly bent Fe–C–O linkage in these proteins was derived from Fourier maps on which the C and O atoms remained unresolved. These maps were interpreted on the assumption that the Fe–C vector was perpendicular to the porphyrin plane. It is much more reasonable to expect that owing to the fixed nature of the globin pocket bending will occur at the Fe atom, leading to a linear Fe–C–O bond, which is not perpendicular to the porphyrin plane.
5. Unlike Fe–CO, the Fe–O2 complex is highly polar and bound O2 is selectively stabilized by H‑bonding to distal His in Mb and Hb. The geometry optimizations performed on the experimental structures present a good agreement with the X‑ray results, in spite of that only certain residues of the fixing centre have been taken into account. In Hb–O2 distal His H‑bonds with the O directly linked with Fe. The H‑bond in the fixing centres of Hb differs from that customarily observed in the biomimetic systems and Mb, in which distal His H‑bonds the ending O. The electronic density of the beginning O is lower than that of the ending O, suggesting that the H‑bond between O and the distal His is weaker than in Mb.
Acknowledgement
The author acknowledges financial support from the Spanish MCT (Plan Nacional I+D+I, Project No. BQU2001‑2935‑C02‑01).
References
1. M. F. Perutz, Proc. R. Soc. London, B, 1980, 208, 135-162.
2. M. F. Perutz, G. Fermi, B. Luisi, B. Shaanan and R. C. Liddington, Acc. Chem. Res., 1987, 20, 309-321.
3. K. G. Welinder, Curr. Opin. Struct. Biol., 1992, 2, 388-393.
4. B. G. Malmström, Chem. Rev., 1990, 90, 1247-1260.
5. M. Sono, M. P. Roach, E. D. Coulter and J. H. Dawson, Chem. Rev., 1996, 96, 2841-2888.
6. J. Barber and B. Andersson, Nature (London), 1994, 370, 31-34.
7. C. L. Drennan, S. Huang, J. T. Drummond, R. G. Matthews and M. L. Ludwig, Science, 1994, 266, 1669-1674.
8. M. A. Halcrow and G. Christou, Chem. Rev., 1994, 94, 2421-2481.
9. M. Momenteau and C. A. Reed, Chem. Rev., 1994, 94, 659-698.
10. A. Dedieu, M.‑M. Rohmer and A. Veillard, Adv. Quantum Chem., 1982, 16, 43-95.
11. M.‑M. Rohmer, A. Dedieu and A. Veillard, Chem. Phys., 1983, 77, 449-462.
12. M.‑M. Rohmer, Chem. Phys. Lett., 1985, 116, 44-49.
13. N. Li, Z. Su, P. Coppens and J. Landrum, J. Am. Chem. Soc., 1990, 112, 7294-7298.
14. T. G. Spiro and P. M. Kozlowski, J. Biol. Inorg. Chem., 1997, 2, 516-520.
15. T. G. Spiro and P. M. Kozlowski, J. Am. Chem. Soc., 1998, 120, 4524-4525.
16. T. Vangberg, D. F. Bocian and A. Ghosh, J. Biol. Inorg. Chem., 1997, 2, 526-530.
17. P. Jewsbury, S. Yamamoto, T. Minato, M. Saito and T. Kitagawa, J. Am. Chem. Soc., 1994, 116, 11586-11587.
18. P. Jewsbury, S. Yamamoto, T. Minato, M. Saito and T. Kitagawa, J. Phys. Chem., 1995, 99, 12677-12685.
19. A. Ghosh and D. F. Bocian, J. Phys. Chem., 1996, 100, 6363-6367.
20. E. Sigfridsson and U. Ryde, J. Biol. Inorg. Chem., 1999, 4, 99-110.
21. C. Rovira, P. Ballone and M. Parrinello, Chem. Phys. Lett., 1997, 271, 247-250.
22. C. Rovira, K. Kunc, J. Hutter, P. Ballone and M. Parrinello, J. Phys. Chem. A, 1997, 101, 8914-8925.
23. C. Rovira, K. Kunc, J. Hutter, P. Ballone and M. Parrinello, Int. J. Quantum Chem., 1998, 69, 31-35.
24. C. Rovira and M. Parrinello, Chem. Eur. J., 1999, 5, 250-262.
25. C. Rovira, P. Carloni and M. Parrinello, J. Phys. Chem. B, 1999, 103, 7031-7035.
26. R. Salzmann, M. T. McMahon, N. Godbout, L. K. Sanders, M. Wojdelski and E. Oldfield, J. Am. Chem. Soc., 1999, 121, 3818-3828.
27. N. Godbout, L. K. Sanders, R. Salzmann, R. H. Havlin, M. Wojdelski and E. Oldfield, J. Am. Chem. Soc., 1999, 121, 3829-3844.
28. G. Loew and M. Dupuis, J. Am. Chem. Soc., 1996, 118, 10584-10587.
29. D. L. Harris and G. H. Loew, J. Am. Chem. Soc., 1996, 118, 10588-10594.
30. D. Harris, G. Loew and L. Waskell, J. Am. Chem. Soc., 1998, 120, 4308-4318.
31. D. E. Woon and G. H. Loew, J. Phys. Chem. A, 1998, 102, 10380-10384.
32. O. Zakharieva, M. Grodzicki, A. X. Trautwein, C. Veeger and I. M. C. M. Rietgens, J. Biol. Inorg. Chem., 1996, 1, 192-204.
33. M. T. Green, J. Am. Chem. Soc., 1998, 120, 10772-10773.
34. G. B. Jameson, G. A. Rodley, W. T. Robinson, R. R. Gagne, C. A. Reed and J. P. Collman, Inorg. Chem., 1978, 17, 850-857.
35. G. B. Jameson, F. S. Molinaro, J. A. Ibers, J. P. Collman, J. I. Brauman, E. Rose and K. S. Suslick, J. Am. Chem. Soc., 1980, 102, 3224-3237.
36. F. Maseras, New J. Chem., 1998, 22, 327-332.
37. J.‑D. Maréchal, G. Barea, F. Maseras, A. Lledós, L. Mouawad and D. Pérahia, J. Comput. Chem., 2000, 21, 282-294.
38. R. Salzmann, C. J. Ziegler, N. Godbout, M. T. McMahon, K. S. Suslick and E. Oldfield, J. Am. Chem. Soc., 1998, 120, 11323-11334.
39. S. Han, K. Cho and J. Ihm, Phys. Rev. E, 1999, 59, 2218-2221.
40. K. Kim, J. Fettinger, J. L. Sessler, M. Cyr, J. Hugdahl, J. P. Collman and J. A. Ibers, J. Am. Chem. Soc., 1989, 111, 403-405.
41. M. P. Johansson, M. R. A. Blomberg, D. Sundholm and M. Wikström, Biochim. Biophys. Acta, 2002, 1553, 183-187.
42. M. P. Johansson, D. Sundholm, G. Gerfen and M. Wikström, J. Am. Chem. Soc., 2002, 124, 11771-11780.
43. N. L. Allinger, J. Am. Chem. Soc., 1977, 99, 8127-8134.
44. F. Torrens, M. Ruiz-López, C. Cativiela, J. I. García and J. A. Mayoral, Tetrahedron, 1992, 48, 5209-5218.
45. F. Torrens, Mol. Simul., 2000, 24, 391-410.
46. B. T. Thole, Chem. Phys., 1981, 59, 341-350.
47. T. L. Hill, J. Chem. Phys., 1948, 16, 399-404.
48. J. A. Deiters, J. C. Gallucci, T. E. Clark and R. R. Holmes, J. Am. Chem. Soc., 1977, 99, 5461-5471.
49. H. Yow and L. S. Bartell, J. Mol. Struct., 1973, 15, 209-215.
50. A. K. Rappé, C. J. Casewit, K. S. Colwell, W. A. Goddard III and W. M. Skiff, J. Am. Chem. Soc., 1992, 114, 10024-10035.
51. F. Torrens, J. Sánchez-Marín and I. Nebot-Gil, J. Mol. Struct. (Theochem), 1999, 463, 27-39.
52. J. J. Weiss, Nature (London), 1964, 202, 83-84.
53. H. P. Misra and I. Fridovich, J. Biol. Chem., 1972, 247, 6960-6962.
54. L. Pauling, Proc. Natl. Acad. Sci. USA, 1977, 74, 2612.
55. M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. J. P. Stewart, J. Am. Chem. Soc., 1985, 107, 3902-3909.
56. R. Huber, O. Epp and H. Formanek, J. Mol. Biol., 1970, 52, 349-354.
57. E. A. Padian and W. E. Love, J. Biol. Chem., 1974, 249, 4067-4078.
58. J. C. Norvell, A. C. Nunes and B. P. Schoenborn, Science, 1975, 190, 568-570.
59. M. Momenteau, W. R. Scheidt, C. W. Eigenbrot and C. A. Reed, J. Am. Chem. Soc., 1988, 110, 1207-1215.
60. W. R. Scheidt and K. Hatano, Acta Crystallogr., Sect. C, 1991, 47, 2201-2203.
Figure caption
Figure 1. Structure of heme(‑His)‑O2. Heme is the prosthetic group of oxyhaemoglobin.
