Steric Factors Override Thermodynamic Driving Force in

13234

J. Phys. Chem. A 2010, 114, 13234–13243

Steric Factors Override Thermodynamic Driving Force in Regioselectivity of Proline Hydroxylation by Prolyl-4-hydroxylase Enzymes Baharan Karamzadeh,† Devesh Kumar,*,‡ G. Narahari Sastry,‡ and Sam P. de Visser*,† The Manchester Interdisciplinary Biocenter and the School of Chemical Engineering and Analytical Science, The UniVersity of Manchester, 131 Princess Street, Manchester M1 7DN, United Kingdom, and Molecular Modelling Group, Indian Institute of Chemical Technology, Hyderabad 500-607, India ReceiVed: September 20, 2010; ReVised Manuscript ReceiVed: NoVember 4, 2010

Prolyl-4-hydroxylase is an important nonheme iron-containing dioxygenase in humans involved in the regioselective hydroxylation of a proline residue in a peptide chain on the C4 position. In biosystems this process is important to create collagen cross-linking and cellular responses to hypoxia. We have performed a series of density functional theory (DFT) studies into the origin of the regioselectivity of proline hydroxylation by P4H enzymes using a minimal active site model (where substrate is unhindered in the binding site) and a larger active site model that incorporates steric hindrance of the substrate by several secondary sphere aromatic residues. Our studies show that thermodynamically the most favorable hydrogen atom abstraction position of proline is from the C5 position; hence, the small model gives a low reaction barrier and large exothermicity for this process. However, stereochemical repulsions of the substrate with aromatic residues of Tyr140 and Trp243 in the second coordination sphere prevent C5 hydroxylation and make C4 hydroxylation the dominant mechanism, despite a lesser driving force for the reaction. These studies explain the remarkable regioselectivity of proline hydroxylation by P4H enzymes and show that the regioselectivity is kinetically controlled but not thermodynamically. In addition, we calculated spectroscopic parameters and found good agreement with experimental data. Introduction

SCHEME 1: Proline Hydroxylation by P4H Enzymes

Mononuclear nonheme iron-containing enzymes are versatile enzymes in nature that catalyze important reaction processes in many biosystems, including the human body.1 These enzymes often share a common motif, whereby the metal is bound to the protein via interactions with two histidine and one carboxylic acid group of either an Asp or a Glu amino acid side chain to form a characteristic facial metal binding triad.2 A large group of mononuclear nonheme iron-containing enzymes utilize R-ketoglutarate (RKG) as a cofactor and catalyze aliphatic hydroxylation reactions as well as substrate halogenation, desaturation, and ring-closure reactions.1,3 In nature, these enzymes, for instance, are involved in the biosynthesis of vancomycin and fosfomycin in antibiotics,4 as well as in DNA and RNA base repair mechanisms in mammals.5 In addition, the R-ketoglutarate-dependent dioxygenases are also involved in cross-linking of collagen and responses to hypoxia.6 The most extensively studied R-ketoglutarate-dependent dioxygenase is taurine/R-ketoglutarate dioxygenase (TauD) found in Escherichia coli. TauD is involved in the hydroxylation of taurine and is highly soluble and relatively stable.7 Thanks to this, it was the first enzyme where a high-valent iron(IV)-oxo species was characterized with resonance Raman, Mo¨ssbauer, and X-ray absorption spectroscopy.8 As a consequence, TauD has become the template for mononuclear nonheme iron enzymes and has prompted many additional studies. The catalytic cycle of TauD enzymes is well-established9,10 and starts with RKG binding to an Fe center and substrate in its vicinity. * To whom correspondence should be addressed. E-mail: sam.devisser@ manchester.ac.uk (S.P.d.V.), [email protected] (D.K.). † The University of Manchester. ‡ Indian Institute of Chemical Technology.

Subsequently, dioxygen binds to the last binding site of the metal, which attacks RKG at the R-keto position to form a bicyclic ring structure. Subsequently, decarboxylation leads to succinate and a high-valent iron(IV)-oxo species, which has been spectroscopically characterized with various methods for TauD.8 The final steps in the catalytic cycle start with a hydrogen atom abstraction from the substrate by the iron(IV)-oxo oxidant followed by rebound of the hydroxyl group to form alcohol products. Another R-ketoglutarate-dependent dioxygenase with importance for human health is prolyl-4-hydroxylase (P4H), involved in the regioselective hydroxylation of proline groups on the C4 position (Scheme 1).11 This is an important reaction in biosystems due to cross-linking collagen helices, oxygen sensing, and cellular response to hypoxia through the hypoxia inducible factor (HIF).12,13 Thus, the cross-linking strand in collagen contains a series of repeated triads, where the first residue usually is proline, the second is often 4-hydroxyproline, while the last residue typically is a glycine.14 A combination of spectroscopic and kinetic studies on substrate hydroxylation by P4H using a small peptide (Pro-Ala-Pro-Lys)3 showed that the proline residues are readily hydroxylated by the enzyme.15 Studies using 18O2 showed that one oxygen atom of molecular oxygen is incorporated into

10.1021/jp1089855  2010 American Chemical Society Published on Web 11/29/2010

Regioselectivity of Proline Hydroxylation

Figure 1. Active site of P4H enzyme as taken from the 2JIG Protein Databank file. Amino acids labeled as in the crystal structure.

hydroxo-proline products.16 Further stopped-flow absorption and Mo¨ssbauer experiments provided evidence of an iron(IV)oxo intermediate with features similar to those detected for the iron(IV)-oxo oxidant of TauD above. Moreover, a large kinetic isotope effect was measured for replacement of the 2,3,3,4,4,5,5positions by deuterium atoms. Mechanistic studies using 5-oxaproline-containing peptides as a substrate found evidence of a Groves-type radical rebound mechanism.17 By contrast, studies of the reaction of P4H with fluorinated proline residues gave no reaction products, due to the strength of the C-F bond.15 On the other hand, with a sulfide group in this position, substrate sulfoxidation was observed. Figure 1 displays the active site structure of P4H enzymes as taken from the 2JIG Protein Databank file.18 The metal is bound to a 2His-1Asp binding motif through interactions with His143, Asp145, and His227 similar to other nonheme iron dioxygenases. Instead of RKG, the PDB contains pyridine-2,4dicarboxylate, which binds as a bidentate ligand to the metal similarly to RKG. Currently no substrate-bound crystal structures are available. The metal center is closely located to the protein surface, and a water channel is seen above the metal, and it has been proposed that the substrate will bind here.19 Directly above the metal are located a series of aromatic residues, such as Tyr140 and Trp243, that may restrict substrate approach to the binding site. At the moment, the origin of the substrate selectivity of P4H enzymes is unclear. Thus, the enzyme regioselectively hydroxylates a proline residue at the C4 position exclusively, and so far there is no evidence of products from hydroxylation of the C3 and C5 positions. To gain insight into the substrate hydroxylation mechanism of this important enzyme for human health and the factors that determine the regioselectivity of substrate hydroxylation, we have performed a density functional theory (DFT) study on models of the active site of P4H enzymes. In principle, the origin of this regioselectivity can arise from thermodynamic factors, i.e., the relative strength of the C4-H bond with respect to the C3-H and C5-H bonds or, alternatively, through electrostatic interactions such as substrate approach to the active center. In this study, we investigated both possibilities and show that stereochemical factors within the substrate binding pocket favor C4 hydroxylation over C3 and C5 hydroxylation.

J. Phys. Chem. A, Vol. 114, No. 50, 2010 13235 calculations with a triple-ζ quality LACV3P+ basis set on iron and 6-311+G* on H, C, O, N atoms, the basis set BS2. The effect of the environment was tested through single-point calculations in Jaguar using the self-consistent reaction field model with a dielectric constant of ε ) 5.7 and a probe radius of 2.72 Å; however, these single-point calculations had little effect on the overall energetics of the process; see Supporting Information, Table S6 and S7. In the past, the methods have been extensively benchmarked against experimental data and it was shown that free energies of activation are reproduced within 3 kcal mol-1,25 although this is a systematic error and generally barrier heights have standard deviations between 1 and 2 kcal mol-1.20,26 In addition, extensive calibration studies were done on related systems in which spectroscopic parameters as well as kinetic isotope effects were calculated.27 These studies generally reproduced experimental data well and confirm that the methods are reliable and suitable for these types of systems. Kinetic isotope effects (KIE) were calculated from the semiclassical Eyring equations (eq 1) using the free energy of activation (∆GH‡) of the reference structure and those where one or more hydrogen atoms of the substrate were replaced by deuterium atoms. In eq 1, R represents the gas constant and T the estimated temperature (298 K).

KIEEyring ) exp[(∆GD‡ - ∆GH‡)/RT]

(1)

Further corrections to the isotope effect due to tunneling were applied by multiplying KIEEyring with the tunneling ratio (QtH/ QtD) (eq 2). Two methods to estimate the tunneling ratio were used due to Wigner and Bell as given in eqs 3 - 6.28 In these equations h is Planck’s constant, kB is the Boltzmann constant, and ν the imaginary frequency in the transition state.

KIEtunneling ) KIEEyring × QtH /QtD Qt,Wigner ) 1 + ut2 /24

with

ut ) hν/kBT

Qt,Bell ) Q1 + Q2

Q1 )

∞

Q2 ) -

∑ (-1)n

n)1

(5)

1 2 sin ut 2

( ) with

(3)

(4)

ut

exp(Vt∆GH) Vt

(2)

Vt )

ut - 2nπ ut (6)

Methods The mechanism of proline hydroxylation by P4H enzymes was modeled using density functional theory using methods tested and benchmarked on nonheme iron enzyme models in our group.20 We use the hybrid density functional method B3LYP21 combined with a double-ζ quality basis set containing LACVP on iron and 6-31G on the rest of the atoms (H, C, O, N), the basis set BS1.22 Geometry optimizations were done with Jaguar 7.6,23 while analytical frequencies were calculated with Gaussian 03.24 Energetics were improved through single-point

In addition to replacing hydrogen atoms by deuterium atoms, we also investigated the effects of replacing 16O2 by 18O2 on the IR frequencies of the iron(IV)-oxo species. The geometry was optimized as described above and followed by a frequency calculation at the UB3LYP/BS1 level of theory in the Gaussian 03 suits of programs. A single-point calculation was performed on this structure with isotopic substitution of 16O2 by 18O2 and the obtained frequencies were analyzed. All frequencies were scaled by a factor of 0.9257.29

13236

J. Phys. Chem. A, Vol. 114, No. 50, 2010

Figure 2. Models of P4H studied in this work.

Results and Discussion Our models were based on the crystal structure coordinates (PDB file 2JIG) of P4H.18 We selected strand A of the dimer, since its peptide chain is longer. Subsequently, we added hydrogen atoms, replaced the metal by iron and pyridine-2,4dicarboxylate by RKG, and inserted an oxo group trans to His227. Two different models of the active site of P4H enzymes were set up; see Figure 2. Model A is a minimal model analogous to the one we used in our earlier studies of TauD33 and comprises of an iron(IV)-oxo group that is linked to two imidazole groups that mimic the His143 and His227 ligands, while the two carboxylic acid ligands of Asp145 and succinate (Succ) are abbreviated to acetate groups and proline as a saturated five-

Karamzadeh et al. membered ring without side chains. A more elaborate model B was built by inclusion of the peptide chain between His143 and Asp145, whereby amino acid 144 is modeled by Gly. Furthermore, His227 is mimicked as methylimidazole, and two distal aromatic amino acids (Tyr140 and Trp243) are included as p-cresol and indole groups, respectively. Finally, substrate proline is extended with a peptide bond on the N-terminus. To prevent the individual groups from reorienting from the original positions in the crystal structure, during the geometry optimization we constrained the position of the carbon atom of the methyl group of Tyr140 and Trp243 with respect to the metal center as well as the angle and dihedral of its neighboring carbon atom. However, no constraints were placed on the metal ligands or the substrate. The active species of P4H enzymes is a high-valent iron(IV)-oxo species, which is one of the few iron(IV)-oxo species of enzymatic systems that has been spectroscopically characterized with resonance Raman, EPR, and Mo¨ssbauer spectroscopic methods.15 In order to benchmark our methods and procedures, we started our work with a detailed analysis of the iron(IV)-oxo species (1) using models A and B as depicted in Figure 2. Here, for each considered species, the spin multiplicity is given as a superscript and model as a subscript before and after the species label, respectively. We ran a full geometry optimization on the lowest lying singlet, triplet, quintet, and septet spin states and subsequently calculated IR spectra as well as EPR and Mo¨ssbauer spectroscopic parameters. The high-lying occupied and low-lying virtual orbitals of the iron(IV)-oxo species of P4H (1) are shown in Figure 3. For simplicity, we only show the orbitals for the small model; those for the large model are essentially the same. The metal 3d type

Figure 3. High-lying occupied and virtual orbitals of the iron(IV)-oxo species of P4H.


J. Phys. Chem. A, Vol. 114, No. 50, 2010 13237

Figure 4. Optimized geometries of the iron(IV)-oxo species of P4H models A and B with bond lengths given in angstroms. Also given are relative energies of the quintet, septet, triplet, and singlet spin states. The inset gives a top view of the metal and its direct ligands for 51B.

orbitals split into three lower lying π* orbitals and two higher lying σ* orbitals. The lowest one is the π*xy orbital that is orthogonal to the Fe-O bond and interacts with the His143, Asp145, and succinate ligands. Slightly higher in energy are two π*FeO orbitals (labeled as π*xz and π*yz) for the antibonding interactions of the metal 3dxz,yz with corresponding 2p orbitals on the oxygen atom. Higher in energy are two σ* type orbitals for the antibonding interactions along the O-Fe-NHis227 axis (σ*z2) and one in the plane of the His143, Asp145, and succinate ligands (σ*x2-y2). This set of orbitals is occupied by four electrons to give the metal oxidation state Fe(IV). DFT calculations on the iron(IV)-oxo species of TauD predicted a quintet spin ground state with π*xy1 π*xz1 π*yz1 σ*x2-y21 occupation.30 In biomimetic nonheme iron(IV)-oxo species, on the other hand, usually a triplet spin ground state is found with π*xy2 π*xz1 π*yz1 occupation.31,32 DFT calculations showed that electron-donating axial ligands tend to reduce the energy gap between the π*xy and σ*x2-y2 orbitals and thereby lower the triplet-quintet energy gap in favor of a triplet spin ground state in biomimetic systems.33 Subsequently, we did a full geometry optimization of the lowest lying spin state structures of models A and B and the results are shown in Figure 4, which clearly demonstrates that both models predict the ground state of 1 to be a quintet spin state with orbital occupation π*xy1 π*xz1 π*yz1 σ*x2-y21. The triplet spin state has π*xy2 π*xz1 π*yz1 occupation, whereas the septet spin state has π*xy1 π*xz1 π*yz1 σ*x2-y21 σ*z21 lpO1 occupation. The latter orbital refers to a lone pair on the oxo group. Finally, the singlet spin state has π*xy2 π*xz1 π*yz1 occupation with the two unpaired electrons having opposing spin direction. Our assignment supports experimental EPR studies15 that indicated a quintet spin ground state. DFT and quantum mechanics/ molecular mechanics (QM/MM) studies on the active species of the related nonheme iron enzyme TauD predicted a similar spin state ordering to that obtained for P4H here with a quintet spin state as the molecular ground state.10,34 The quintet spin ground state is well-separated from the septet, triplet, and singlet spin states by 7.4, 13.6, and 25.2 kcal mol-1 for Model A. For the iron(IV)-oxo intermediate in the catalytic cycle of TauD, Siegbahn and co-workers10 found a quintet spin ground state that was 6.0 kcal mol-1 lower in energy than the

lowest septet spin state, in perfect agreement with our calculations here. This spin state ordering is similar to previous studies of nonheme iron(IV)-oxo complexes in enzymatic systems,30,34,35 where also large quintet-triplet and quintet-singlet energy gaps were obtained. In the nonheme iron enzyme isopenicillin N-synthase the iron(IV)-oxo intermediate is in a quintet spin ground state with the triplet and septet spin states 5.2 and 10.3 kcal mol-1 higher in energy.36 It appears therefore that the relative energy of the quintet and triplet spin states is very sensitive to the ligands bound to the metal, and minor changes lead to major shifts in the quintet-triplet energy gap. This is also apparent from studies on biomimetic nonheme iron(IV)-oxo complexes that tend to have a triplet spin ground state with the lowest lying quintet spin state several kcal mol-1 higher in energy.32,37 Usually in biomimetic nonheme iron(IV)-oxo complexes the septet spin state is high in energy and does not play a role of importance. Obviously, this is quite different from the situation in enzymatic systems such as TauD and P4H, where the septet spin state is significantly lower in energy, sometimes within 5 kcal mol-1 of the quintet spin ground state. Clearly, in P4H the triplet and singlet spin states are so high in energy in the reactants that they will not contribute significantly to the reaction mechanism; therefore, in the remainder of the paper, we will focus on the lowest lying quintet and septet spin states only. Extensive DFT studies on nonheme iron-containing enzymes such as TauD and R-ketoglutarate-dependent halogenase showed that the singlet and triplet spin state surfaces stay high in energy along the substrate hydroxylation mechanism.10,30,35,38 Interestingly, the optimized geometries do not seem dramatically affected by the size of the model. For instance, an iron-oxo bond length in the quintet spin state of 1.652 and 1.649 Å is found for 51A and 51B, respectively. In addition, the calculated metal-ligand bond distances between Fe and His143, Asp145, His227, and Succ give only small differences between the two models. It appears, therefore, that the optimized geometry is very little affected by increasing the size of the model through addition of mainly secondary sphere amino acids. The calculated iron(IV)-oxo distances are in excellent agreement with those calculated for complexes of heme and nonheme systems before.32,39 An Fe-O distance ranging from 1.62 to 1.65 Å was found for heme-based iron(IV)-oxo complexes

13238


dependent on the axial ligand. Our calculated Fe-O stretch vibration of 811 cm-1 for 51B is in line with previously determined values for iron(IV)-oxo complexes. Resonance Raman experiments of Hausinger et al.8b on TauD supplied with 16O2 and 18O2 provided a difference spectrum with oxygen sensitive peaks at 583, 821, and 859 cm-1 in the 16O2 spectrum. The latter vibration, however, coincided with a strong solvent band, so its identification was not unambiguous. To find the corresponding vibrations for the iron(IV)-oxo species of P4H, we analyzed the frequencies of 51B and its 18O2 substituted spectrum. Thus, in the catalytic cycle of P4H and TauD, molecular oxygen donates one oxygen atom to RKG to form succinate and an iron(IV)-oxo species. Therefore, we replaced the oxo group and one of the oxygen atoms of the carboxylic acid group of succinate that binds iron by 18O atoms to calculate the 18O2 spectrum. Vibrational frequency analysis of 51B[16O2] gives an FedO stretch vibration at νFedO ) 811 cm-1, while the stretch vibration along the Fe-O bond between the metal and the succinate group is located at νFe-OSucc ) 857 cm-1. Our calculations, therefore, support the experimental assignment of an FedO stretch vibration at 811 cm-1 and confirm that there indeed is an oxygen-sensitive band at 857 cm-1, which originates from the first oxygen atom transfer process in the catalytic cycle where succinate is formed. Two bending vibrations in the carboxylic acid group of succinate are identified at 557 cm-1 (out-of-plane) and 583 cm-1 (in-plane). Substitution of 16O2 by 18 O2 in 51B downshifts νFedO by 33 cm-1 and the νFe-OSucc by 18 cm-1. Much smaller frequency shifts are obtained for the two bending vibrations of -11 cm-1 for the in-plane vibration and -4 cm-1 for the out-of-plane vibration. However, these two bending vibrations give sharp differences in the calculated Raman intensities, whereby the higher vibration is more Raman active in the 16O2 spectrum, whereas the lower vibration is more active in the 18O2 spectrum. As a consequence of this, they implicate a single peak that shifts by about 30 cm-1 but in fact represent two peaks with different Raman intensities in the 16O2 and 18O2 spectra. In addition, we calculated Mo¨ssbauer spectroscopic parameters of 1 and focused on the electric field splitting (∆EQ), the asymmetry parameter of the nuclear quadrupole tensor (η), and the isomer shift (δ) (Supporting Information, Table S8). Although the error bars for these types of calculations are high, a modest agreement between the calculated values of ∆EQ for 5 1 with the experimental data from ref 15 is found. Hence, DFT calculations on the iron(IV)-oxo species of P4H support the spectroscopic assignments in the literature in favor of a quintet spin as the ground state. The benchmark calculations of spectroscopic parameters on models A and B give good agreement with experimental data and support the use of these models for our calculations. Substrate Hydroxylation by P4H Enzymes. The general mechanism of substrate hydroxylation by the iron(IV)-oxo species of P4H is shown in Scheme 2. Thus, the iron(IV)-oxo species (1) picks up a hydrogen atom from the substrate via a hydrogen atom abstraction barrier (TSHA) to form an iron(III)hydroxo complex with a nearby prolyl radical (2). A radical rebound barrier (TSreb) separates this radical intermediate from the alcohol product complex (3). This mechanism is similar to the rebound mechanism proposed for substrate hydroxylation reactions by the iron(IV)-oxo heme(+•) active species of cytochrome P450 enzymes and confirmed by biomimetic and computational studies.40 Theoretically, substrate hydroxylation by P4H enzymes should lead to products originating from hydroxylation at the

Karamzadeh et al. SCHEME 2: General Mechanism of Substrate Hydroxylation by P4H Enzymes

C3, C4, and C5 positions of proline; however, in the enzyme only the C4 position is activated. To find out what the most likely hydroxylation site of a proline molecule is, we did a series of DFT calculations to determine the bond dissociation energies (BDECH) of the various C-H bonds of proline; see Figure 5. The C-H bond strength or BDECH was calculated from eq 7:

Pro-H f Pro · + H · + BDECH

(7)

Experimental studies of hydrogen abstraction reactions by metal(IV)-oxo oxidants showed that the rate constants and hence free energies of activation correlated with the strength of the C-H bond that is broken in the process.41 Recent DFT studies in combination with valence bond modeling rationalized these hydrogen abstraction reactions and showed that the correlation follows from electron transfer processes from reactants to radical intermediates.26c In the case of testosterone hydroxylation by the iron(IV)-oxo species of cytochrome P450, the barrier heights of hydrogen atom abstraction for four different C-H positions of the substrate correlated to the strength of these C-H bonds.42 Figure 5 displays the calculated BDECH values of the C3-H, 4 C -H, and C5-H positions of proline. As follows, H-abstraction from the C5 position from proline should be the easiest because it has the weakest C-H bond strength (BDECH ) 85.4 kcal mol-1). Furthermore, the C4-H bond strength of proline apparently is the strongest of these C-H bonds and, therefore, the least likely position for hydroxylation with a BDECH ) 92.4

Figure 5. Calculated BDECH values of proline C-H bonds with energies in kcal mol-1.



Figure 6. Potential energy landscape of proline hydroxylation at the C5, C4, and C3 positions using model A. Relative energies are in kcal mol-1, obtained with basis set BS2, while ZPE corrections are calculated with basis set BS1. Optimized geometries contain bond lengths in angstroms and the imaginary frequency in wavenumbers.

kcal mol-1. Therefore, our BDECH calculations show that proline hydroxylation on the C4 position is thermodynamically the least expected hydroxylation site, despite the fact that this is the targeted site in the enzyme. Using a small Ala-Pro-Ala peptide, the calculated BDECH values at the C3, C4, and C5 positions of proline are 93.5, 92.9, and 86.7 kcal mol-1, respectively, which indicates that these BDECH values give little sensitivity regarding the size of the substrate. In order to find out why P4H enzymes hydroxylate proline on the C4 position while the driving force for the reaction is the smallest at this position, we ran a series of DFT calculations to establish environmental (stereochemical) effects on the reaction mechanism and barriers. Subsequently, we investigated the mechanism of proline hydroxylation at the C5, C4, and C3 positions using the minimal model (model A), and the results are shown in Figure 6. Proline hydroxylation at the C4 position was calculated on the lowest lying singlet, triplet, quintet, and septet spin states. The singlet and triplet reactants, however, are well higher in energy than 5 1A (Figure 4) by 25.2 and 13.6 kcal mol-1, respectively, and so are their corresponding hydrogen atom abstraction barriers leading to an iron-hydroxo intermediate (>28 kcal mol-1). Therefore, we will focus in the main text on the quintet and septet states only; details of the other spin states can be found in the Supporting Information. Proline hydroxylation at the C3 and C5 positions was only calculated on the quintet spin state surface. Although the reactant state has a quintet spin ground state, after hydrogen abstraction close-lying iron(III)-hydroxo complexes (2) in the quintet and septet spin states, 52A,C4 and 72A,C4, are found within 1 kcal mol-1. These two states have the same orbital occupation (π*xy1 π*xz1 π*yz1 σ*z21 σ*x2-y21 πSub1) with a singly occupied metal d-block coupled to an unpaired electron on the proline rest-group; this interaction is ferromagnetic in the septet spin state and antiferromagnetic in the quintet spin state. Even though the intermediate has close-lying quintet and

septet spin states, as a matter of fact, in the H-abstraction transition states, the quintet state is well-separated from the septet by more than 10 kcal mol-1 for the C4-hydroxylation mechanism. Consequently, P4H hydroxylation will proceed via single-state reactivity on a dominant quintet spin state surface only. The energy gap between 5TSHA and 7TSHA is in contrast to substrate hydroxylation by TauD models, where two-state reactivity patterns on close lying quintet and septet spin states were found.26b,30 Nevertheless, the H-abstraction step is ratedetermining as the radical rebound barrier is much smaller. In agreement with the thermodynamic analysis mentioned above, the relative energies of the H-abstraction transition states and radical intermediates follow the trend in BDECH with 5 TSHA,A,C5 < 5TSHA,A,C3 < 5TSHA,A,C4. Hydrogen abstraction of a H-atom from the C5 position (5TSHA,A,C5) is 10.5 kcal mol-1 lower in energy than that for H-atom abstraction from position C4 (5TSHA,A,C4), whereas the energy difference between the two radical intermediates (52A,C4 and 52A,C5) is 9.8 kcal mol-1. The difference in energy between the BDECH values for H-abstraction at the C5 and C4 positions (Figure 5) is calculated to be 7.0 kcal mol-1, which is comparable to the difference obtained between 52A,C5 and 52A,C4. Thermodynamically, the exothermicity of formation of 52A from 51A is the difference in energy between BDECH of the C-H bond of the substrate that is broken and the BDEOH of the FeO-H bond that is formed (eq 8).41a,b

∆H ) BDECH - BDEOH

(8)

For a small model of TauD that closely resembles model A here, a BDEOH of 95.7 kcal mol-1 was calculated.26b Combination of this BDEOH value with the BDECH values from Figure 5 gives predicted reaction exothermicities ∆H using eq 8 for formation of intermediate complexes for H-abstraction at the C4 and C5 position of -3.3 and -10.3 kcal mol-1. These values

13240


are a few kcal mol-1 less exothermic than those calculated for 5 2A,C4 and 52A,C5. The reason for this is the close proximity of the radical to the iron(III)-hydroxo complex. As a matter of fact, the hydrogen atom of the amide group of the radical forms a hydrogen bond with the carboxylic acid ligand of the metal and thereby lowers the energy of this radical intermediate. Nevertheless, the trends in relative energies of 52A,C4 and 52A,C5 in Figure 6, therefore, reflect the differences in BDECH of the C-H bonds of the substrate that are broken. Consequently, thermodynamically and kinetically one would expect P4H enzymes to hydroxylate the C5 position of proline rather than the C4 position, which was found in the enzyme. To find out whether stereochemical interactions change this regioselectivity preference, we calculated also a larger model of the P4H active site, vide infra. Recent studies of hydrogen abstraction trends by iron(IV)-oxo species of heme and nonheme oxidants showed that the barrier height from reactants to iron(III)-hydroxo intermediates correlates linearly with BDECH, whereas the reverse reaction barrier correlates with BDEOH instead.20,26 The calculated BDECH values in Figure 5, therefore, support the ordering of the hydrogen abstraction barriers calculated for the small P4H model complex with ordering 5TSHA,A,C5 < 5TSHA,A,C3 < 5TSHA,A,C4. The hydrogen abstraction barrier 5TSHA,A,C4 of 10.6 kcal mol-1 is slightly higher in energy than the barrier obtained before for toluene (7.6 kcal mol-1)26b hydroxylation by the iron(IV)-oxo species of TauD, where a C-H bond of BDECH ) 85.8 kcal mol-1 is broken. The C4-H bond strength in proline, actually, compares to the BDECH value for hydrogen abstraction from the secondary carbon atom of propane (BDECH ) 93.0 kcal mol-1), which would imply that the active oxidant of P4H enzymes is a powerful enough oxidant to hydroxylate very strong C-H bonds like in propane. Thermodynamically, one would expect a similar barrier height for hydrogen atom abstraction from the C3 and C4 positions. The TSHA barrier for abstraction of a hydrogen atom from the C3 position, however, is lowered with respect to that of the C4 position due to more stabilizing hydrogen-bonding interactions. The structure is stabilized by a short hydrogen bond of 2.231 Å between the amide proton of the substrate to one of the carboxylic acid oxygen atoms of the succinate residue, while the complementary hydrogen bond in the C4 mechanism is 2.650 Å long. This stronger O · · · H-N interaction in the C3 mechanism lowers the barrier height significantly with respect to that obtained for the C4 mechanism. Geometrically, 5TSHA,A,C4 is in a more upright position than 5 TSHA,A,C3 with an Fe-O-Cproline angle of 150.8° compared to 140.8°, respectively. Furthermore, 5TSHA,A,C4 is early with short C-H and long O-H distances (rCH ) 1.228 Å, rOH ) 1.320 Å) in comparison to those found for the 5TSHA,A,C3 (rCH ) 1.301 Å, rOH ) 1.207 Å). The magnitude of the imaginary frequency in the transition state for all H-abstraction barriers are large, which is indicative of narrow and steep barriers, which will result in a significant amount of tunneling in the H-abstraction process. The final H-abstraction transition state, 5TSHA,A,C5, is well lower than that expected based on the BDECH of the C5-H bond that is broken. Thus, a BDECH for H-abstraction from proline at the C5 position of 85.4 kcal mol-1 compares to the value obtained for aliphatic hydrogen abstraction from toluene, which gave a H-abstraction barrier of 7.6 kcal mol-1 using a similar nonheme iron-containing oxidant.26b The substantial lowering of the barrier height is due to the geometric distortion of the oxidant, whereby the imidazole group of His143 bends

Karamzadeh et al. upward and donates a hydrogen bond to the amide group of the substrate and hence is lowered in energy. This considerable distortion of the orientation of the histidine group may not be possible in the actual enzyme due to steric interactions. The small model (A), therefore, may not be an appropriate model for the studies of proline hydroxylation by P4H enzymes, and in particular, the C5 hydroxylation mechanism is disturbed due to artificial bonding patterns that cannot happen in the actual enzyme configuration. We, therefore, decided to study a larger enzyme model of P4H enzymes instead. Figure 7 shows the potential energy profile of proline hydroxylation using the large active site model of P4H (model B) as calculated on the quintet spin state surface. Similarly to the results shown above using the smaller model, the reaction is stepwise via a radical intermediate with a rate-determining hydrogen atom abstraction barrier. Substrate hydroxylation at the C4 position gives energies in line with those obtained with the small model above. The H-abstraction barrier (5TSHA,B,C4) is 12.6 kcal mol-1 higher in energy than reactants and is only slightly raised with respect to the small model system (10.6 kcal mol-1), where the substrate has no stereochemical interactions with second sphere amino acids. Hydrogen atom abstraction from the C3 position is raised by 7.3 kcal mol-1 in model B with respect to model A and is now slightly higher in energy to that obtained for the C4 position: 5TSHA,C,C3 ) 13.5 kcal mol-1. In addition to the mechanisms for hydrogen abstraction from proline at the C3 and C4 positions for model B, we also made efforts to find the C5-H abstraction mechanism. Although we were able to locate the radical intermediate (52B,C5), we did not manage to find its corresponding H-abstraction transition state (5TSHA,B,C5). Thus, a geometry scan for the hydrogen transfer step starting from 52B,C5 led to a high-energy pathway (>30 kcal mol-1), which implies that the steric hindrance of the two aromatic residues (Tyr140 and Trp243) included in the model block the hydrogen atom abstraction from the C5 position effectively. Although, the substrate approach leading to H-abstraction from the C5 position gave favorable hydrogen-bonding interactions and lowering of the barrier for the small model, as a matter of fact for the larger and more realistic model these hydrogenbonding interactions are not possible and instead the approach is hampered and the barrier considerably raised in energy. This is important for the enzyme due to the weak C5-H bond strength of proline and prevention of hydroxylation at that position. It appears, therefore, that the substrate approach to the active center is aligned in such a way as to favor the C4 hydroxylation mechanism and stereochemically disfavor hydroxylation on the other positions of proline. Furthermore, it may be anticipated that mutants whereby the Tyr140 and/or Trp243 amino acids are replaced by smaller size amino acids should give substantial hydroxylation at the C3 and C5 positions. Recent experimental studies on an R-ketoglutarate-dependent halogenase highlighted the intricate involvement of the active site amino acid residues by favoring a regioselective halogenation over hydroxylation of the substrate, which could be reversed by active site mutations.43 Geometrically, 5TSHA,B,C4 is quite different from 5TSHA,A,C4 (compare Figures 6 and 7) with a relatively long C-H distance of 1.326 Å (compared to 1.228 Å in 5TSHA,A,C4) and short O-H distance of 1.194 Å (compared to 1.320 Å in 5TSHA,A,C4). Thus, 5 TSHA,B,C4 is late with a geometry close to intermediate, whereas 5 TSHA,A,C4 has a more symmetrical C-H-O orientation. The hydrogen atom abstraction process leading to radical intermediates is slightly exothermic and has the largest exother-



Figure 7. Potential energy landscape of proline hydroxylation at the C4 and C3 positions using model B. Relative energies are in kcal mol-1, obtained with basis set BS2, while ZPE corrections are calculated with basis set BS1. Optimized geometries contain bond lengths in angstroms and the imaginary frequency in wavenumbers.

Figure 8. Kinetic isotope effects (KIE) as calculated with the Eyring, and Wigner and Bell models for the replacement of one or both of the hydrogen atoms at the C4 position of proline with deuterium atoms. Also given are the imaginary frequencies in the transition states. Ha is the transferring hydrogen atom, and replacement of Hb with a deuterium atom gives the secondary kinetic isotope effect.

micity for H-abstraction from the C5 position. This is in agreement with the discussion above that the energy difference between reactants and radical intermediates equals the BDECH - BDEOH difference of substrate and oxidant (eq 8). The reaction energy for formation of the radical intermediates in model A is significantly larger than that obtained for model B due to the formation of hydrogen-bonding interactions of the amide group of proline with hydrogen-bonding donors of the metal ligands. Due to stereochemical interactions of the aromatic residues in

model B, the substrate cannot approach the iron center closely and therefore these hydrogen-bonding interactions are not possible; hence, those structures are somewhat higher in energy. The reaction energies for formation of 52B,C4 and 52B,C5 are within 1 kcal mol-1 of the ∆H values calculated from eq 8 and hence follow thermodynamic principles. A comparison of Figures 6 and 7 reveals a further effect of the stereochemical interactions of the aromatic residues, namely, on the radical rebound barriers that increase from 1.5 kcal mol-1

13242


in model A (5TSreb,A,C4) to 8.2 kcal mol-1 in model B (5TSreb,B,C4). Similar trends are observed for proline hydroxylation at the C3 position. These enlarged radical rebound barriers will increase the lifetime of the radical intermediates significantly during which, for instance, rearrangement of the radical can take place, leading to side products.44 Despite the rise in rebound barrier, however, the H-atom abstraction barrier is still the ratedetermining step in the reaction mechanism. The reaction exothermicities found for the large models are similar to those found for the smaller models (compare Figures 6 and 7). These large values for reaction exothermicities follow thermodynamic principles and are comparable to values obtained before for substrate hydroxylation by iron(IV)-oxo complexes.45 In particular, for substrate dehydrogenation, which also starts with an initial hydrogen atom abstraction step, it was shown that the reaction energy can be written as a sum of bond dissociation energies of the bonds that are broken and formed during the reaction process.46 That way, reaction energies of 50 kcal mol-1 were rationalized. Subsequently, we calculated the kinetic isotope effect for replacement of one or two of the hydrogen atoms at the C4 position of substrate proline by deuterium atoms using model B, and the results are shown in Figure 8. These procedures use the semiclassical kinetic isotope effect due to Eyring and are based on the free energies of activation of the hydrogen and deuterium-substituted reactions (eq 1). Subsequently, two methods to estimate the difference in tunneling properties of a hydrogen versus a deuterium atom were used, namely, via the methods of Wigner and Bell.28 The Eyring kinetic isotope effect for replacement of the transferring hydrogen atom by deuterium is substantial (7.7) and increases to 9.9 when tunneling corrections due to Wigner are included and to 16.0 using the Bell model. Interestingly, the secondary isotope effect is larger than 1, although only slightly, which contrast the typical value of 0.9 obtained before.44 As a consequence, the KIE value for substitution of both hydrogen atoms by deuterium atoms increases to 11.2 and 18.4 with tunneling corrections included due to Wigner and Bell, respectively. These values point to a substantial kinetic isotope effect and considerable tunneling corrections, which are as expected for hydrogen atom transfer processes. Nevertheless, the values obtained here are somewhat smaller than those observed experimentally for the native substrate, where a KIE ) 60 was obtained.15 This may have to do with the difference in substrate used or due to interactions of the rest of the protein with the substrate-oxidant complex. Conclusions In this work, the regioselectivity of proline hydroxylation by P4H enzymes is investigated using density functional theory methods. It is shown that thermodynamically substrate hydroxylation at the C5 position should be favorable over that on the C4 position with significantly larger H-abstraction exothermicity. Stereochemical interactions, in particular, of two aromatic amino acid residues (Tyr140 and Trp243) hamper H-abstraction from the C3 and C5 positions and thereby lead to a regioselectivity preference of C4 hydroxylation. These stereochemical interactions affect the hydrogen abstraction as well as radical rebound barriers, but the former remains the rate-determining step in the reaction mechanism. Thus, these stereochemical interactions are essential for the enzyme activity and make sure that proline is regioselectively hydroxylated at the unfavorable C4 position. We validated the models by matching the computed spectroscopic values with experimental ones (IR spectra, Mo¨ssbauer and EPR spectroscopy). The iron(IV)-oxo model gives good

Karamzadeh et al. agreement between experimental and DFT calculated spectroscopic parameters. Acknowledgment. The research was supported by CPU time provided by the National Service of Computational Chemistry Software (NSCCS). D.K. holds a Ramanujan Fellowship from the Department of Science and Technology (DST), New Delhi (India), and acknowledges its financial support (Research Grants SR/S2/RJN-11/2008 and SR/S1/PC-58/2009). Supporting Information Available: Cartesian coordinates of all structures, tables with group spin densities, charges, and absolute energies of all structures calculated in this work, and figures with optimized geometries and geometry scans. This material is available free of charge via the Internet at http:// pubs.acs.org. References and Notes (1) (a) Solomon, E. I.; Brunold, T. C.; Davis, M. I.; Kemsley, J. N.; Lee, S.-K.; Lehnert, N.; Neese, F.; Skulan, A. J.; Yang, Y.-S.; Zhou, J. Chem. ReV. 2000, 100, 235–349. (b) Bugg, T. D. H. Curr. Opin. Chem. Biol. 2001, 5, 550–555. (c) Ryle, M. J.; Hausinger, R. P. Curr. Opin. Chem. Biol. 2002, 6, 193–201. (d) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que, L., Jr. Chem. ReV. 2004, 104, 939–986. (e) Abu-Omar, M. M.; Loaiza, A.; Hontzeas, N. Chem. ReV. 2005, 105, 2227–2252. (f) Bruijnincx, P. C. A.; van Koten, G.; Klein Gebbink, R. J. M. Chem. Soc. ReV. 2008, 37, 2716– 2744. (g) Bugg, T. D. H.; Ramaswamy, S. Curr. Opin. Chem. Biol. 2008, 12, 134–140. (2) Que, L., Jr. Nat. Struct. Biol. 2000, 7, 182–184. (3) (a) Krebs, C.; Fujimori, D. G.; Walsh, C. T.; Bollinger, J. M., Jr. Acc. Chem. Res. 2007, 40, 484–492. (b) Schofield, C. J.; Zhang, Z. Curr. Opin. Chem. Biol. 1999, 9, 722–731. (c) Bugg, T. D. H. Tetrahedron 2003, 59, 7075–7101. (4) (a) Choroba, O. W.; Williams, D. H.; Spencer, J. B. J. Am. Chem. Soc. 2000, 122, 5389–5390. (b) Higgins, L. J.; Yan, F.; Liu, P.; Liu, H.W.; Drennan, C. L. Nature 2005, 437, 838–844. (5) (a) Mishina, Y.; Duguid, E. M.; He, C. Chem. ReV. 2006, 106, 215–232. (b) O’Brien, P. J. Chem. ReV. 2006, 106, 720–752. (c) Simmons, J. M.; Mu¨ller, T. A.; Hausinger, R. P. Dalton Trans. 2008, 5132–5142. (d) Yi, C.; Yang, C. G.; He, C. Acc. Chem. Res. 2009, 42, 519–529. (6) Chowdhury, R.; Hardy, A.; Schofield, C. J. Chem. Soc. ReV. 2008, 37, 1308–1319. (7) Muthukumaran, R. B.; Grzyska, P. K.; Hausinger, R. P.; McCracken, J. Biochemistry 2007, 46, 5951–5959. (8) (a) Price, J. C.; Barr, E. W.; Tirupati, B.; Bollinger, J. M., Jr.; Krebs, C. Biochemistry 2003, 42, 7497–7508. (b) Proshlyakov, D. A.; Henshaw, T. F.; Monterosso, G. R.; Ryle, M. J.; Hausinger, R. P. J. Am. Chem. Soc. 2004, 126, 1022–1023. (c) Riggs-Gelasco, P. J.; Price, J. C.; Guyer, R. B.; Brehm, J. H.; Barr, E. W.; Bollinger, J. M., Jr.; Krebs, C. J. Am. Chem. Soc. 2004, 126, 8108–8109. (d) Grzyska, P. K.; Appelman, E. H.; Hausinger, R. P.; Proshlyakov, D. A. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 3982– 3987. (9) (a) Bollinger, J. M., Jr.; Price, J. C.; Hoffart, L. M.; Barr, E. W.; Krebs, C. Eur. J. Inorg. Chem. 2005, 4245–4254. (b) Clifton, I. J.; McDonough, M. A.; Ehrismann, D.; Kershaw, N. J.; Granatino, N.; Schofield, C. J. J. Inorg. Biochem. 2006, 100, 644–669. (10) Borowski, T.; Bassan, A.; Siegbahn, P. E. M. Chem.sEur. J. 2004, 10, 1031–1041. (11) (a) Winter, A. D.; Page, A. P. Mol. Cell. Biol. 2000, 4084–4093. (b) Kivirikko, K. I.; Myllyla, R.; Pihlajaniemi, T. FASEB J. 1989, 3, 1609– 1617. (12) (a) Bruick, R. K.; McKnight, S. L. Science 2001, 294, 1337–1340. (b) Berra, E.; Benizri, E.; Ginouve`s, A.; Volmat, V.; Roux, D.; Pouysse´gur, J. EMBO J. 2003, 22, 4082–4090. (c) Lee, K. A.; Lynd, J. D.; O’Reilly, S.; Kiupel, M.; McCormick, J. J.; LaPres, J. J. Mol. Cancer Res. 2008, 6, 829–842. (d) Seifert, A.; Katschinski, D. M.; Tonack, S.; Fischer, B.; Santos, A. N. Chem. Res. Toxicol. 2008, 21, 341–348. (13) West, C. M.; van der Wel, H.; Wang, Z. A. DeVelopment 2007, 134, 3349–3358. (14) Gorres, K. L.; Edupuganti, R.; Krow, G. R.; Raines, R. T. Biochemistry 2008, 47, 9447–9455. (15) Hoffart, L. M.; Barr, E. W.; Guyer, R. B.; Bollinger, J. M., Jr.; Krebs, C. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 14738–14743. (16) Fujimoto, D.; Tamiya, N. Biochem. J. 1962, 84, 333–335. (17) Wu, M.; Moon, H.-S.; Begley, T. P.; Myllyharju, J.; Kivirikko, K. I. J. Am. Chem. Soc. 1999, 121, 587–588. (18) Koski, M. K.; Hieta, R.; Bo¨llner, C.; Kivirikko, K. I.; Myllyharju, J.; Wierenga, R. K. J. Biol. Chem. 2007, 282, 37112–37123.

Regioselectivity of Proline Hydroxylation (19) (a) McDonough, M. A.; McNeill, L. A.; Tilliet, M.; Papamicae¨l, C. A.; Chen, Q.-Y.; Banerji, B.; Hewitson, K. S.; Schofield, C. J. J. Am. Chem. Soc. 2005, 127, 7680–7681. (b) Pappalardi, M. B.; Martin, J. D.; Jiang, Y.; Burns, M. C.; Zhao, H.; Ho, T.; Sweitzer, S.; Lor, L.; Schwartz, B.; Duffy, K.; Gontarek, R.; Tummino, P. J.; Copeland, R. A.; Luo, L. Biochemistry 2008, 47, 11165–11167. (c) Culpepper, M. A.; Scott, E. E.; Limburg, J. Biochemistry 2010, 49, 124–133. (20) (a) Aluri, S.; de Visser, S. P. J. Am. Chem. Soc. 2007, 129, 14846– 14847. (b) de Visser, S. P. J. Am. Chem. Soc. 2010, 132, 1087–1097. (21) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785–789. (22) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270–283. (b) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257– 2261. (23) Jaguar 7.6;Schro¨dinger, LLC.: New York, 2007. (24) Frisch, M. J. ; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.; Jr., Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, N.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, ReVision C.01; Gaussian, Inc.: Wallingford, CT, 2004. (25) (a) Kumar, D.; de Visser, S. P.; Shaik, S. Chem.sEur. J. 2005, 11, 2825–2835. (b) de Visser, S. P.; Oh, K.; Han, A.-R.; Nam, W. Inorg. Chem. 2007, 46, 4632–4641. (26) (a) de Visser, S. P.; Kumar, D.; Cohen, S.; Shacham, R.; Shaik, S. J. Am. Chem. Soc. 2004, 126, 8362–8363. (b) Latifi, R.; Bagherzadeh, M.; de Visser, S. P. Chem.sEur. J. 2009, 15, 6651–6662. (c) Shaik, S.; Kumar, D.; de Visser, S. P. J. Am. Chem. Soc. 2008, 130, 10128–10140. (27) (a) Kumar, D.; de Visser, S. P.; Shaik, S. J. Am. Chem. Soc. 2003, 125, 13024–13025. (b) de Visser, S. P. Chem.sEur. J. 2006, 12, 8168– 8177. (c) de Visser, S. P. Chem.sEur. J. 2008, 14, 4533–4541. (d) Porro, C. S.; Kumar, D.; de Visser, S. P. Phys. Chem. Chem. Phys. 2009, 11, 10219–10226. (e) Heyes, D. J.; Sakuma, M.; de Visser, S. P.; Scrutton, N. S. J. Biol. Chem. 2009, 284, 3762–3767. (28) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic Molecules; Robert, E. Krieger Publishing Co.: Malabar, FL, 1987. (29) de Visser, S. P. Coord. Chem. ReV. 2009, 253, 754–768. (30) (a) de Visser, S. P. J. Am. Chem. Soc. 2006, 128, 9813–9824. (b) de Visser, S. P. Chem. Commun. 2007, 171–173. (31) (a) Rohde, J.-U.; In, J.-H.; Lim, M. H.; Brennessel, W. W.; Bukowski, M. R.; Stubna, A.; Mu¨nck, E.; Nam, W.; Que, L., Jr. Science 2003, 299, 1037–1039. (b) Martinho, M.; Banse, F.; Bartoli, J.-F.; Mattioli, T. A.; Battioni, P.; Horner, O.; Bourcier, S.; Girerd, J.-J. Inorg. Chem. 2005, 44, 9592–9596. (c) Sastri, C. V.; Seo, M. S.; Park, M. J.; Kim, K. M.; Nam, W. Chem. Commun. 2005, 1405–1407. (d) De Oliveira, F. T.; Chanda, A.; Banerjee, D.; Shan, X.; Mondal, S.; Que, L., Jr.; Bominaar, E. L.; Mu¨nck, E.; Collins, T. J. Science 2007, 315, 835–838. (e) Jackson, T. A.; Rohde, J.-U.; Seo, M. S.; Sastri, C. V.; DeHont, R.; Stubna, A.; Ohta, T.; Kitagawa, T.; Mu¨nck, E.; Nam, W.; Que, L., Jr. J. Am. Chem. Soc. 2008, 130, 12394– 12407. (32) (a) Kumar, D.; Hirao, H.; Que, L., Jr.; Shaik, S. J. Am. Chem. Soc. 2005, 127, 8026–8027. (b) Hirao, H.; Kumar, D.; Que, L., Jr.; Shaik, S. J. Am. Chem. Soc. 2006, 128, 8590–8606. (33) Hirao, H.; Que, L., Jr.; Nam, W.; Shaik, S. Chem.sEur. J. 2008, 14, 1740–1756.

J. Phys. Chem. A, Vol. 114, No. 50, 2010 13243 (34) Godfrey, E.; Porro, C. S.; de Visser, S. P. J. Phys. Chem. A 2008, 112, 2464–2468. (35) (a) Sinnecker, S.; Svensen, N.; Barr, E. W.; Ye, S.; Bollinger, J. M., Jr.; Neese, F.; Krebs, C. J. Am. Chem. Soc. 2007, 129, 6168–6179. (b) Topol, I. A.; Nemukhin, A. V.; Salnikow, K.; Cachau, R. E.; Abashkin, Y. G.; Kasprzak, K. S.; Burt, S. K. J. Phys. Chem. A 2006, 110, 4223– 4228. (c) Nemukhin, A. V.; Topol, I. A.; Cachau, R. E.; Burt, S. K. Theor. Chem. Acc. 2006, 115, 348–353. (36) Lundberg, M.; Siegbahn, P. E. M.; Morokuma, K. Biochemistry 2008, 47, 1031–1042. (37) For examples of computational studies on biomimetic nonheme iron(IV)-oxo complexes, see the following:(a) Bautz, J.; Comba, P.; Lopez de Laorden, C.; Menzel, M.; Rajaraman, G. Angew. Chem., Int. Ed. 2007, 46, 8067–8070. (b) de Visser, S. P. J. Am. Chem. Soc. 2006, 128, 15809– 15818. (c) Decker, A.; Rohde, J.-U.; Klinker, E. J.; Wong, S. D.; Que, L., Jr.; Solomon, E. I. J. Am. Chem. Soc. 2007, 129, 15983–15996. (d) Bernasconi, L.; Baerends, E. J. Eur. J. Inorg. Chem. 2008, 1672–1681. (e) de Visser, S. P.; Nam, W. J. Phys. Chem. A 2008, 112, 12887–12895. (f) Comba, P.; Maurer, M.; Vadivelu, P. Inorg. Chem. 2009, 48, 10389–10396. (38) (a) Nemukhin, A. V.; Grigorenko, B. L.; Topol, I. A.; Burt, S. K. Int. J. Quantum Chem. 2006, 106, 2184–2190. (b) de Visser, S. P.; Latifi, R. J. Phys. Chem. B 2009, 113, 12–14. (c) Ye, S.; Price, J. C.; Barr, E. W.; Green, M. T.; Bollinger, J. M., Jr.; Krebs, C.; Neese, F. J. Am. Chem. Soc. 2010, 132, 4739–4751. (d) Borowski, T.; Noack, H.; Radon´, M.; Zych, K.; Siegbahn, P. E. M. J. Am. Chem. Soc. 2010, 132, 12887–12898. (39) (a) Green, M. T. J. Am. Chem. Soc. 1999, 121, 7939–7940. (b) Ogliaro, F.; de Visser, S. P.; Cohen, S.; Kaneti, J.; Shaik, S. ChemBioChem 2001, 2, 848–851. (c) Scho¨neboom, J. C.; Lin, H.; Reuter, N.; Thiel, W.; Cohen, S.; Ogliaro, F.; Shaik, S. J. Am. Chem. Soc. 2002, 124, 8142–8151. (d) de Visser, S. P.; Shaik, S.; Sharma, P. K.; Kumar, D.; Thiel, W. J. Am. Chem. Soc. 2003, 125, 15779–15788. (e) Behan, R. K.; Hoffart, L. M.; Stone, K. L.; Krebs, C.; Green, M. T. J. Am. Chem. Soc. 2006, 128, 11471– 11474. (40) (a) Groves, J. T. Models and Mechanisms of Cytochrome P450 Action. In Cytochrome P450: Structure, Mechanism and Biochemistry, 3rd ed.; Ortiz de Montellano, P. R., Ed.; Kluwer Academic/Plenum Publishers: New York, 2005; Chapter 1, pp 1-44. (b) Nam, W. Acc. Chem. Res. 2007, 40, 522–531. (c) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H. Chem. ReV. 1996, 96, 2841–2888. (41) (a) Bordwell, F. G.; Cheng, J.-P.; Ji, G.-Z.; Satish, A. V.; Zhang, X. J. Am. Chem. Soc. 1991, 113, 9790–9795. (b) Mayer, J. M. Acc. Chem. Res. 1998, 31, 441–450. (c) Kaizer, J.; Klinker, E. J.; Oh, N. Y.; Rohde, J.-U.; Song, W. J.; Stubna, A.; Kim, J.; Mu¨nck, E.; Nam, W.; Que, L., Jr. J. Am. Chem. Soc. 2004, 126, 472–473. (d) Yoon, J.; Wilson, S. A.; Jang, Y. K.; Seo, M. S.; Nehru, K.; Hedman, B.; Hodgson, K. O.; Bill, E.; Solomon, E. I.; Nam, W. Angew. Chem., Int. Ed. 2009, 48, 1257–1260. (e) Lansky, D. E.; Goldberg, D. P. Inorg. Chem. 2006, 45, 5119–5125. (f) Bell, S. R.; Groves, J. T. J. Am. Chem. Soc. 2009, 131, 9640–9641. (g) Prokop, K. A.; de Visser, S. P.; Goldberg, D. P. Angew. Chem., Int. Ed. 2010, 49, 5091–5095. (42) Zhang, Y.; Morisetti, P.; Kim, J.; Smith, L.; Lin, H. Theor. Chem. Acc. 2008, 121, 313–319. (43) Matthews, M. L.; Neumann, C. S.; Miles, L. A.; Grove, T. L.; Booker, S. J.; Krebs, C.; Walsh, C. T.; Bollinger, J. M., Jr. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 17723–17728. (44) (a) de Visser, S. P.; Ogliaro, F.; Shaik, S. Angew. Chem., Int. Ed. 2001, 40, 2871–2874. (b) Kumar, D.; de Visser, S. P.; Sharma, P. K.; Cohen, S.; Shaik, S. J. Am. Chem. Soc. 2004, 125, 1907–1920. (c) Kumar, D.; de Visser, S. P.; Shaik, S. J. Am. Chem. Soc. 2004, 126, 5072–5073. (45) (a) Ogliaro, F.; Harris, N.; Cohen, S.; Filatov, M.; de Visser, S. P.; Shaik, S. J. Am. Chem. Soc. 2000, 122, 8977–8989. (b) Kamachi, T.; Yoshizawa, K. J. Am. Chem. Soc. 2003, 125, 4652–4661. (c) de Visser, S. P.; Kumar, D.; Cohen, S.; Shacham, R.; Shaik, S. J. Am. Chem. Soc. 2004, 126, 8362–8363. (46) Kumar, D.; Tahsini, L.; de Visser, S. P.; Kang, H. Y.; Kim, S. J.; Nam, W. J. Phys. Chem. A 2009, 113, 11713–11722.

JP1089855

Steric Factors Override Thermodynamic Driving Force in

Recommend Documents