Which Oxidation State Initiates Dehalogenation in the B12-Dependent

Nov 4, 2015 - Subtle changes in the active site architecture untangled overlapping substrate ranges and mechanistic differences of two reductive ...
0 downloads 0 Views 3MB Size
Download PDF
Research Article pubs.acs.org/acscatalysis

Which Oxidation State Initiates Dehalogenation in the B12Dependent Enzyme NpRdhA: CoII, CoI, or Co0? Rong-Zhen Liao,*,† Shi-Lu Chen,‡ and Per E. M. Siegbahn§ †

Key Laboratory of Material Chemistry for Energy Conversion and Storage, Ministry of Education, School of Chemistry and Chemical Engineering, Huazhong University of Science and Technology, Wuhan 430074, People’s Republic of China ‡ School of Chemistry, Beijing Institute of Technology, Beijing 100081, People’s Republic of China § Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, SE-10691 Stockholm, Sweden S Supporting Information *

ABSTRACT: The quantum chemical cluster approach was used to elucidate the reaction mechanism of debromination catalyzed by the B12-dependent reductive dehalogenase NpRdhA. Various pathways, involving different oxidation states of the cobalt ion and different protonation states of the model, have been analyzed in order to find the most favorable one. We find that the reductive C−Br cleavage takes place exclusively at the CoI state via a heterolytic pathway in the singlet state. Importantly, the C−H bond formation and the C−Br bond cleavage proceeds via a concerted transition state, as opposed to the stepwise pathway suggested before. C−Br cleavage at the CoII state has a very high barrier, and the reduction of CoI to Co0 is associated with a very negative potential; thus, reductive dehalogenation at CoII and Co0 can be safely ruled out. Examination of substrates with different halogen substitutions (F, Cl, Br, I) shows that the dehalogenation reactivity follows the order C−I > C−Br > C−Cl > C−F, and the barrier for defluorination is so high that NpRdhA cannot catalyze that reaction. KEYWORDS: density functional, reductive dehalogenation, reaction mechanism, enzyme catalysis, cobalamin

1. INTRODUCTION

substitutions showed higher activity, while the activity of substrates with fluoro substitution has not been reported. On the basis of the crystal structure and the substrate docking studies, two different mechanisms (Scheme 1) have been proposed for the carbon−bromide cleavage, being either heterolytic or homolytic.4 One-electron transfer from the adjacent [4Fe-4S] cluster to the CoII complex 1 first takes place to generate the CoI intermediate 2. In the heterolytic pathway, proton transfer from Tyr426 to the brominated carbon atom proceeds to form the tetrahedral ketone intermediate 3. Then, the C−Br bond breaks heterolytically, coupled with twoelectron transfer from the CoI center, generating the CoIII− bromide intermediate 4. Finally, a second electron transfer from the [4Fe-4S] cluster occurs to form the CoII−bromide product 7. In this mechanism, the C−Br bond cleavage takes place at the CoI stage. In the alternative homolytic pathway, proton transfer and electron transfer to the CoI intermediate (via 5) produce the CoI substrate radical intermediate 6 (formally Co0). This is followed by a homolytic C−Br bond cleavage to form the product complex 7. These two mechanisms thus differ in the nature of the C−Br bond cleavage and the oxidation state of the cobalt that initiates the C−Br bond cleavage.

Organohalides are generally known as prevalent contaminants, mainly produced by agricultural and industrial activities.1 The use of microorganisms for the degradation of organohalides has been advanced as an attractive environmental remediation strategy.2 In this context, organohalide respiration catalyzed by B12-dependent reductive dehalogenases has gathered special attention, and great efforts have been devoted in the last few decades to understanding this process and its reaction mechanism.3 A breakthrough in this field came very recently with the report of the crystal structures of two reductive dehalogenases: namely, NpRdhA from Nitratireductor pacificus ph-3B4 and PceA from Sulfurospirillum multivorans.5 The crystal structure of NpRdhA has been solved by Leys et al. at 2.3 Å resolution, and it reveals a “base off” cobalamin cofactor in the active site and two nearby iron−sulfur cubane clusters.4 In the active site, the cobalt ion (CoII) is fivecoordinated with a chloride in the axial position (Figure 1). The chloride was replaced by the substrate during the reaction. On the basis of docking studies, a number of second-shell residues have been suggested to be involved in substrate binding, including Ser422, Tyr426, Lys488, and Arg552. Substrate screening studies indicated that this enzyme strictly accepts ortho-halogenated phenolic substrates.4 In addition, 2,6-dihalogenated substrates with chloro, bromo, and iodo © XXXX American Chemical Society

Received: July 16, 2015 Revised: October 15, 2015

7350

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis

pathways. Using density functional theory (DFT) with the hybrid functional B3LYP*-D3 (B3LYP12 functional with 15% HF13 exchange and D3 dispersion14 from B3LYP), we have investigated three different debromination mechanisms in NpRdhA, involving three different oxidation states: namely, CoII, CoI, and Co0. On the basis of the best mechanism from our calculations, the dehalogenation of substrates with fluoro, chloro, and iodo substitutions has also been studied to understand the trends of reactivity. This kind of approach has proven very successful in the study of various classes of enzymes, including several related Co-dependent enzymes.15−21 It should be pointed out that the alternative QM/MM approach22−27 has also been used to investigate B12-dependent enzymes.28−33 However, comparison of the quantum chemical cluster and the alternative QM/MM approaches with acetylene hydratase as a representative example showed that both methods give very similar mechanistic conclusions already with a QM region of less than 200 atoms.34,35

2. COMPUTATIONAL DETAILS We used the wild-type crystal structure (PDB code 4RAS)4 to design a quite large model of the active site (Figure 2). The model consists of the cobalt ion (CoII) along with its corrin ligand, which has been truncated to include only the corrin ring, extended by one or two carbons. In addition, the side chains of five important second-shell residues, Phe291, Ser422, Tyr426, Lys488, and Arg522, and the peptide chain of Cys561−Gly562 as well as two water molecules were also included in the model. The 2,6-dibromophenolate substrate was added in its anionic

Figure 1. X-ray structure of the active site of NpRdhA from Nitratireductor pacificus ph-3B complexed with chloride (coordinates taken from PDB entry 4RAS).4 The Co and Cl ions are labeled in red.

To resolve the mechanism, the quantum chemical cluster approach6−11 was used to investigate the various possible

Scheme 1. Proposed Reductive Debromination Mechanisms for NpRdhA4

7351

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis

the sensitivity of pKas, redox potentials, and barriers to the choice of functional. To estimate the polarization effects from the protein environment on the active site model, single-point calculations were carried out at the same level of theory as the geometry optimizations using the SMD42 solvation model method. The dielectric constant (ε) was chosen to be 4, which is the standard value used for the modeling of the enzyme surroundings.11 For comparison, an ε value of 8 has also been tested to check the sensitivity of the results (especially pKa and one-electron redox potential) to the choice of dielectric constant. Analytic frequency calculations were performed at the same level of theory as the geometry optimizations to obtain zero-point energies (ZPE) and to establish the nature of the various stationary points. Unless otherwise specified, the B3LYP*-D3 energies, including solvation, ZPE, and dispersion corrections from B3LYP, are reported. The energies reported can be approximately considered as the free energies in solution. A similar procedure has been used by Noodleman and co-workers in the calculation of pKas and redox potentials. To calculate the redox potentials of various reduction steps, the experimental absolute redox potential of the standard hydrogen electrode (SHE) is used (4.281 V, 98.7 kcal/mol) as a reference.43 For the calculation of pKas, the experimental value of −264.0 kcal/mol for the solvation free energy of a proton44 is used. This kind of approach has been successfully applied to the study of water oxidation catalysis.45−48 For more details on calculating redox potentials and pKas, see the Supporting Information. The uncertainties of density functionals for the calculations of absolute redox potentials and pKas have been shown to be around 0.2 V and 3 units, respectively.49,50 In addition, the error of the cluster approach used on relative energies in enzymatic reactions has been shown to be around 3 kcal/mol.8 An increase in the size of the QM region may improve the energetics by a few kcal/mol;35 however, this is beyond the scope of the present study. In any case, this increase will not alter the main conclusion for the present enzyme, as the barrier difference between different mechanisms is more than 20 kcal/mol (vide infra).

Figure 2. Optimized structure of the active site model of NpRdhA (total charge of 1+). Atoms marked in red (including certain hydrogen atoms) were fixed at their X-ray structure positions during the geometry optimization.

form, as it forms a number of hydrogen bonds with the nearby residues, including one positively charged residue Arg552. Lys488 was chosen to be unprotonated, as its pKa was calculated to be 6.0 (the protonated form has also been considered; vide infra). The total charge of the model is 1+, and the total number of atoms is 211. The quantum chemical calculations were accomplished with the B3LYP functional12 as implemented in the Gaussian 09 program.36 For geometry optimization, the 6-31G(d,p) basis sets were used for the C, N, O, and H elements and the LANL08 pseudopotential for Co and Br (with corresponding f and d polarization).37 On the basis of these optimized geometries, single-point calculations were performed at the B3LYP* level13 using larger basis sets, in which all elements, except Co and Br, were described by 6-311+G(2d,2p). It has been shown that B3LYP* performs better in describing relative energies in transition-metal complexes.13 D3 dispersion corrections (with the original D3 damping function) proposed by Grimme14 were also added at single points. Single-point calculations using BP86-D3,38,39 B3LYP-D3, M06L-D3,40 and M06-D341 functionals have also been carried out to examine

3. RESULTS AND DISCUSSION In this section, we first consider the dehalogenation of 2,6dibromophenolate initiated by CoII, CoI, and Co0 (sections 3.1−3.3). Next, the results calculated with different density functionals, including B3LYP*-D3, B3LYP-D3, BP86-D3, M06L-D3, and M06-D3, are presented (section 3.4). Finally, we present a comparison of the reactivity of different

Figure 3. Optimized structures of reactant, transition state, and intermediate (doublet) for CoII-catalyzed debromination with an unprotonated Lys488. Distances are given in Å. Spin densities on selected atoms are given in red italics. Energies are given in kcal/mol relative to A-React. The imaginary frequency for A-TS1 is indicated. For the full model, see Figure 2. 7352

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis halogenated substrates, with fluoro, chloro, and iodo substitution (section 3.5). 3.1. Debromination at CoII. We first considered C−Br bond cleavage directly from CoII without acceptance of an electron from the iron−sulfur cluster. The reactant complex (equivalent to structure 1 in Scheme 1) is labeled as A-React (Figure 3), in which Lys488 is unprotonated. The pKa of Lys488 was calculated to be 6.0 when ε = 4 is used, and it becomes 11.6 when ε = 8 is used. The calculation of pKas is quite sensitive to the choice of the dielectric constant, and the reason for this is that the total charges of the two species with different protonation states are different and consequently the solvation free energies are quite sensitive. An unprotonated Lys488 is not common but has been seen in pyridoxaldependent enzymes.51−53 Therefore, we cannot make a safe conclusion about the protonation state of Lys488, and both protonation states were therefore considered. A-React is a doublet, and the spin is mainly localized on the metal (spin density of 1.07). The quartet lies at +24.2 kcal/mol relative to the doublet and is too high to be accessible. From AReact, the C−Br bond cleavage and the protonation of C1 by Tyr426 proceed concertedly via A-TS1 (Figure 3) with a barrier of 39.0 kcal/mol (Figure 4), and the reaction is

thus keeps its oxidation state as +2 during the reaction. In AInt1, the spin densities on Tyr426 and the 1-bromophenolate product are 0.98 and −0.62, respectively. The protonation of Lys488 at A-React (structure labeled as A-React′, pKa = 6.0; see Figure 5) was calculated to be endergonic by 1.4 kcal/mol at pH 7.0. The protonated Lys488 can deliver a proton to the substrate using Tyr426 as a bridge to facilitate the C−Br bond cleavage. Consequently, Tyr426 cannot donate an electron to the substrate during the reaction as for the case when Lys488 is unprotonated. The calculations (see Figure 5) suggest that the debromination reaction along this pathway takes place via a homolytic C−Br bond cleavage, generating a substrate radical and a Br radical, which reacts with CoII to form a CoIII−bromide intermediate. The barrier for the debromination was calculated to be 38.1 kcal/mol relative to AReact′, and the reaction is endergonic by 25.9 kcal/mol. These results suggest that the dehalogenation reaction cannot take place directly at the CoII state, since the barriers are prohibitive. This agrees with the experimental conclusion that the reduction of CoII is needed before the C−Br bond cleavage occurs.4 3.2. Debromination at CoI. From A-React, a protoncoupled electron transfer, with a proton donated to Lys488 from the solution, leads to a reduction to form a CoI complex (labeled as B-React in Figure 6). B-React (equivalent to structure 2 in Scheme 1) is a singlet, and the triplet lies at +12.3 kcal/mol. The reduction potential was calculated to be −0.55 V, and the formation of B-React using nicotinamide adenine dinucleotide phosphate (NADPH, reduction potential of −0.32 V)54 as the electron donor is thus endergonic by 5.2 kcal/mol. The pKa of B-React (for the deprotonated form, see Figure S3 in the Supporting Information) was calculated to be 17.4, and the pKa of its protonated form (labeled as B-React′, Figure 7) is −6.3, where the phenol substrate is in its neutral form. These results suggest that B-React is the dominant CoI species at pH 7.0. From B-React, the debromination takes place in the singlet state via the concerted transition state B-TS1 (Figure 6), associated with a barrier of only 7.3 kcal/mol relative to BReact, and the total barrier becomes 12.5 kcal/mol when the energy penalty for the formation of B-React is added. This leads to the formation of the CoIII−Br intermediate B-Int1 (equivalent to structure 4 in Scheme 1), which is found to lie at −14.0 kcal/mol relative to B-React. No stable tetrahedral intermediate as proposed in Scheme 1 (structure 3) can be located. B-TS1 has an imaginary frequency of 1201.3i cm−1, with the vibrational mode corresponding to mainly proton transfer from Tyr426 to the substrate C1. The overlaid structures of B-React, B-TS1, and B-Int1 (see Figure S2 in the Supporting Information) shows a high level of structural consistency in which the major structural change takes place in the reacting region. At B-TS1, the critical C1−Br1 and Co−Br1 distances are 2.62 and 2.49 Å, respectively. In this transition state, the proton is being transferred from Tyr426 to the substrate, with O1−H1 and C1−H1 distances of 1.19 and 1.43 Å, respectively. This mechanism can be considered as a C−Br heterolytic pathway, in which the CoI ion acts as a two-electron donor to bind the Br+ ion generated during the C−Br bond cleavage. Debromination in the triplet state has also been considered. For the reactant complex (B-Reacttriplet; Figure S4 in the Supporting Information), a population analysis showed that the electronic structure can be interpreted as a low-spin CoII ion (spin density of 1.10) interacting with a corrin radical ion (spin

Figure 4. Calculated Gibbs free energy diagram (in kcal/mol) for NpRdhA-catalyzed debromination of 2,6-dibromophenolate.

endergonic by 28.9 kcal/mol. A-TS1 has an imaginary frequency of 117.2i cm −1 , with the vibrational mode corresponding to mainly C−Br bond cleavage. At A-TS1, the C1−Br1 bond is elongated to 2.17 Å, and the Br1 atom is 4.70 Å away from the cobalt ion, indicating only a very small catalytic effect from the cobalamin. In addition, H1 is transferred as a proton, rather than a hydrogen atom, as little spin density can be seen on H1 (−0.01) at A-TS1. In this pathway, a proton-coupled electron transfer from Tyr426 to the substrate enables a heterolytic C−Br bond cleavage, which creates a Br− species bound to CoII (A-Int1), and the cobalt ion 7353

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis

Figure 5. Optimized structures of reactant, transition state, and intermediate (doublet) for CoII-catalyzed debromination with a protonated Lys488. Distances are given in Å. Spin densities on selected atoms are given in red italics. Energies are given in kcal/mol relative to A-React′. The imaginary frequency for A-TS1′ is indicated.

Figure 6. Optimized structures of reactant, transition state, and intermediate (singlet) for CoI-catalyzed debromination with an ionic substrate. Distances are given in Å. Energies are given in kcal/mol relative to B-React. The imaginary frequency for B-TS1 is indicated.

Figure 7. Optimized structures of reactant, transition state, and intermediate (singlet) for CoI-catalyzed debromination with a neutral substrate. Distances are given in Å. Energies are given in kcal/mol relative to B-React′. The imaginary frequency for B-TS1′ is indicated.

can protonate the substrate after C−Br bond cleavage, which provides additional driving force for the dehalogenation. Protonation of the phenolate substrate in B-React to form BReact′ (Figure 7) is endergonic by 18.1 kcal/mol at pH 7.0. Such a large energy penalty can be understood from the fact that two positively charged residues, namely Lys488 and Arg552, are already hydrogen-bonded to the substrate phenolate oxygen, which therefore becomes very difficult to become protonated. B-React′ is a singlet, and the triplet state is 16.0 kcal/mol higher. Importantly, no electron transfer from Co to the substrate takes place, which is different from the case for the suggested intermediate 5 in the homolytic cleavage pathway in Scheme 1.4 From B-React′, the debromination is very facile, with a barrier of only 1.5 kcal/mol, and the reaction is exergonic by 22.9 kcal/mol. The total barrier, however, is 24.8 kcal/mol relative to A-React, while it is only 12.5 kcal/mol for the heterolytic pathway with a deprotonated substrate.

density of 0.86) in a ferromagnetic fashion. Different from that in the singlet state, the C−Br bond cleavage proceeds via a single-electron-transfer pathway (B-TS1triplet; Figure S3 in the Supporting Information) with a barrier of 18.7 kcal/mol, which is 11.4 kcal/mol higher than that in the singlet state (7.3 kcal/ mol). During the reaction, one electron is transferred from corrin to the C−Br σ* orbital, which facilitates the C−Br bond cleavage. Consequently, a CoII−Br phenolate radical complex is generated, and the spin densities on Co and the phenolate are 1.01 and 0.98, respectively. Different from the case for the singlet transition state, the proton transfer from Tyr426 to the substrate is very late and the imaginary frequency of B-TS1triplet mainly corresponds to the C−Br bond cleavage and the Co−Br bond formation. This is somewhat similar to the singleelectron-transfer mechanism as proposed by Adrian and coworkers,55 in which one electron is transferred from CoI to the substrate. In the present case, a nearby acid (Lys488-Tyr426) 7354

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis 3.3. Debromination at Co0. Further reduction of B-React to generate the formally Co0 complex C-React (pKa = 7.6, Figure 8; for the deprotonated form see Figure S5 in the

Figures S6−S10 in the Supporting Information. BP86-D3 gives a positive reduction potential (0.133 V) for the CoII/CoI couple, suggesting an exergonic process for the formation of B-React. All other functionals give reduction potentials lower than that for the reference NADPH (−0.32 V), implying an endergonic process for the first reduction step. B3LYP*-D3, M06L-D3, and M06-D3 give quite similar reduction potentials, all around −0.5 V, while B3LYP-D3 gives the lowest reduction potential (−0.739 V). For the CoI/Co0 reduction, all functionals give a reduction potential of lower than −1.4 V, which matches the data derived from other synthetic Co complexes.56−59 The very low reduction potential further excludes the possibility of the involvement of Co0 during the reaction, or the homolytic pathway shown in Scheme 1. For the barrier, M06-D3 gives the highest value (24.7 kcal/ mol), while BP86-D3 gives the lowest (only 2.0 kcal/mol). B3LYP*-D3, B3LYP-D3, and M06L-D3 give somewhat more reasonable barriers, in the range of 12−19 kcal/mol. By comparison of the barrier from M06L-D3 and M06-D3, it can be seen that the barrier decreases upon lowering the percentage of the exact Hartree−Fock exchange, and similar results can be seen from the B3LYP-D3 and B3LYP*-D3 methods. Further experimental kinetic studies would be helpful to validate the accuracy of these three functionals. It should be pointed out that the reduction potential is somewhat sensitive to the choice of the dielectric constant, which then affects the total barrier. When ε = 8 is used (for an energy diagram see Figure S11 in the Supporting Information), A-React′ becomes the resting state, and the first reduction is a one-electron process with a potential of −0.883 V. B-TS1 is 6.3 kcal/mol higher than B-React, and the total barrier becomes 19.3 kcal/mol. This is about 7 kcal/mol higher than that with ε = 4, mainly due to the difference in the first reduction potential. However, it does not alter the main mechanistic conclusion, in which the heterolytic pathway is preferred. We have also tested geometry optimization with dispersion corrections, in this case using the B3LYP-D3 functional. The Gibbs free energy diagram obtained is shown in Figure S12 in the Supporting Information. For a geometry optimization with and without dispersion corrections, the differences in redox potentials are about 0.1 V. The total barrier is 13.0 kcal/mol for optimization with dispersion, while it is 12.5 kcal/mol for optimization without dispersion. These results suggest that the inclusion of dispersion in geometry optimizations is not very important, as it gave similar results and the same conclusion for the present system. 3.5. Dehalogenation of Substrates with F, Cl, and I Substitution. Various halogenated aromatic substrates with different substitutions have been tested to explore the substrate dependence.4 Dechlorination, debromination, and deiodination were found to have similar activities, while no defluorination

Figure 8. Optimized structures of reactant and product (doublet) for Co0-catalyzed debromination with a neutral substrate. Distances are given in Å. Spin densities on selected atoms are given in red italics. Energies are given in kcal/mol relative to C-React.

Supporting Information) was found to be a PCET process, leading to a spontaneous C−Br bond cleavage. No tetrahedral intermediate such as structure 6 in Scheme 1 can be located. The calculations showed that the electron goes to the substrate rather than to the metal and the proton goes to the substrate phenolate oxygen. The spin density on C1 is 0.96, while it is 0 on the Co, suggesting an oxidation state of +1. The redox potential was calculated to be −1.67 V, which implies that the formation of C-React is endergonic by as much as 31.2 kcal/ mol using NADPH as the electron donor. This is 1.12 V lower than that for the CoII/CoI couple, and similar differences have also been seen in a number of synthetic Co-based water reduction catalysts.56−59 The energy penalty (36.4 kcal/mol, Figure 4) is too high to be a viable option, and the proposed heterolytic pathway4 involving a CoI substrate radical (equivalent to Co0) shown in Scheme 1 can thus be safely ruled out. The PCET reduction of B-Int1 was calculated to have a potential of 0.95 V, generating the CoII−Br product complex CProd (pKa = 10.0; for the deprotonated form see Figure S5 in the Supporting Information), in which Lys488 gets protonated. The whole reaction is exergonic by 38.0 kcal/mol. 3.4. Comparison of Different Density Functionals. To investigate the sensitivity of the results to the choice of density functionals, single-point calculations using a number of other popular functionals have been performed, including B3LYP-D3, BP86-D3, M06L-D3, and M06-D3. The redox potentials and total barriers calculated using these functionals are displayed in Table 1, and the corresponding energy diagrams are shown in

Table 1. Comparison of Reduction Potentials (in V) and Barriers (in kcal/mol) Calculated Using Different Functionals

reduction potential B-React/A-React C-React/B-React C-Prod/B-Int1 barrier B-TS1 relative to B-React B-TS1 relative to A-React

B3LYP*-D3

B3LYP-D3

BP86-D3

M06L-D3

M06-D3

−0.546 −1.671 0.947

−0.739 −1.444 1.126

0.133 −1.907 0.656

−0.459 −1.755 0.865

−0.450 −1.952 1.234

7.3 12.5

8.7 18.4

2.0 2.0

10.5 13.7

21.7 24.7

7355

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis

bond cleavage. The reduction potentials for these three substrates are around −1.7 V. However, the fluoro substrate is a very poor electron acceptor, and during the reduction the electron goes to the metal, generating a true Co0 complex (spin density of 0.78 on Co). Consequently, the reduction potential is as low as −2.23 V. F, Cl, and Br substitutions have rather small effects on the first reduction potential (−0.68 to −0.65 V), while I substitution make the first reduction less negative, being −0.34 V. For the fluoro substrate, the following C−F bond cleavage transition state B-TS1F involves a late proton transfer (C1−H1 distance of 2.01 Å) in comparison with the other substrates. In addition, the Co−F1 and C−F1 distances are 1.99 and 2.19 Å, respectively. These two distances are much shorter than those for the chloro, bromo, and iodo substrates (Table 2). B-TS1F is 29.0 kcal/mol relative to B-ReactF. Consequently, the total barrier becomes 36.6 kcal/mol when the energetic penalty for the formation of B-ReactF is added. The very high barrier indicates that this enzyme cannot catalyze the defluorination reaction. Indeed, the C−F bond activation is well-known to be much more difficult than C−Cl, C−Br, and C−I activations.60 To achieve the hydrodefluorination reaction, a very common strategy is to use a metal hydride to initiate the reductive defhuorination, as shown by many synthetic organometallic catalysts.61−63 The total barrier for the dechlorination reaction is 16.6 kcal/mol, 4.1 kcal/mol higher than that for debromination. The deiodination reaction is very facile, with a barrier of only 8.2 kcal/mol. The reactivity trend is consistent with the experimental results,4 although the relative activity seems to be somewhat overestimated.

activity has been reported. Here, we compare the reactivity of substrates with different substitutions (X = F, Cl, Br, I). The energy diagram for these dehalogenation reactions is shown in Figure 9, and the key distances in the reactant, transition state, and intermediate of the CoI-initiated C−X bond cleavage are shown in Table 2.

4. CONCLUSION We have in this work used the quantum chemical cluster approach to investigate the reaction mechanism of the B12dependent reductive dehalogenase NpRdhA. On the basis of our calculations, we have proposed a CoI-initiated concerted dehalogenation mechanism (Scheme 2) for the reductive dehalogenase NpRdhA. With 2,6-dibromophenolate as a representative substrate, three possible oxidation states, namely, CoII, CoI, and Co0, have been considered to identify which oxidation state leads to the reductive debromination. It is found that the reaction starts from the CoII state, which undergoes a PCET reduction to form a CoI complex. In the following step, the C−Br bond cleavage proceeds concomitantly with a proton transfer from Lys488 to the substrate carbon (C1) facilitated by Tyr426, without the formation of a tetrahedral intermediate as suggested before. In the singlet state, the reaction proceeds via a heterolytic C−Br cleavage pathway, while the triplet state follows a single-electron-transfer mechanism, in which one electron transfers from corrin to the C−Br σ* orbital. The singlet state is energetically more favorable, and the total barrier was calculated to be 12.5 kcal/mol. The CoI ion functions as a two-electron donor to accept the Br+ generated from the C−Br heterolytic cleavage, leading to the formation of a CoIII−Br 2bromophenolate intermediate. Finally, another PCET reduction takes place to produce the CoII−Br product. The debromination at CoII has a very high barrier, and the formation of Co0 is thermodynamically quite unfeasible due to the very negative reduction potential for the CoI/Co0 couple. Substrates with different halogen substitutions (F, Cl, Br, I) were also examined, and the calculations reproduce quite well the experimentally observed trends in reactivity, although the

Figure 9. Calculated Gibbs free energy diagram (in kcal/mol) for NpRdhA-catalyzed dehalogenation of fluoro, chloro, bromo, and iodo substrates.

Table 2. Comparison of Geometric Parameters (Distances in Å) in CoI-Catalyzed Dehalogenation B-React O1−H1 H1−C1 C1−X1 X1−Co B-TS1 O1−H1 H1−C1 C1−X1 X1−Co B-Int1 O1−H1 H1−C1 C1−X1 X1−Co

F

Cl

Br

I

0.97 2.77 1.38 4.52

0.97 3.07 1.78 3.95

0.97 2.56 1.97 3.15

0.97 2.43 2.21 2.97

0.99 2.01 2.19 1.99

1.15 1.48 2.64 2.34

1.19 1.43 2.62 2.49

1.22 1.40 2.66 2.66

2.75 1.09 4.12 1.77

2.74 1.09 3.83 2.21

2.76 1.09 3.79 2.35

2.75 1.09 3.73 2.53

The reactions of the four halogenated substrates follow the same pattern, and similar stationary points have been optimized, except that no C−F bond cleavage takes place during the formation of C-React by the reduction of B-React, while C−X (X = Cl, Br, I) bond cleavage is seen for the other three substrates. The reason is that the substrates with Cl, Br, and I substitutions are better electron acceptors than CoI, and the electron prefers to go to the substrate, leading to C−X 7356

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis

(4) Payne, K. A.; Quezada, C. P.; Fisher, K.; Dunstan, M. S.; Collins, F. A.; Sjuts, H.; Levy, C.; Hay, S.; Rigby, S. E. J.; Leys, D. Nature 2014, 517, 513−516. (5) Bommer, M.; Kunze, C.; Fesseler, J.; Schubert, T.; Diekert, G.; Dobbek, H. Science 2014, 346, 455−458. (6) Himo, F.; Siegbahn, P. E. M. Chem. Rev. 2003, 103, 2421−2456. (7) Noodleman, L.; Lovell, T.; Han, W.-G.; Li, J.; Himo, F. Chem. Rev. 2004, 104, 459−508. (8) Siegbahn, P. E. M.; Borowski, T. Acc. Chem. Res. 2006, 39, 729− 738. (9) Ramos, M. J.; Fernandes, P. A. Acc. Chem. Res. 2008, 41, 689− 698. (10) Siegbahn, P. E. M.; Himo, F. JBIC, J. Biol. Inorg. Chem. 2009, 14, 643−651. (11) Blomberg, M. R. A.; Borowski, T.; Himo, F.; Liao, R.-Z.; Siegbahn, P. E. M. Chem. Rev. 2014, 114, 3601−3658. (12) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (13) Reiher, M.; Salomon, O.; Hess, B. A. Theor. Chem. Acc. 2001, 107, 48−55. (14) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. (15) Jensen, K. P.; Ryde, U. J. Am. Chem. Soc. 2003, 125, 13970− 13971. (16) Alfonso-Prieto, M.; Biarnés, X.; Kumar, M.; Rovira, C.; Kozlowski, P. M. J. Phys. Chem. B 2010, 114, 12965−12971. (17) Kozlowski, P. M.; Kamachi, T.; Toraya, T.; Yoshizawa, K. Angew. Chem., Int. Ed. 2007, 46, 980. (18) Durbeej, B.; Sandala, G. M.; Bucher, D.; Smith, D. M.; Radom, L. Chem. - Eur. J. 2009, 15, 8578. (19) Siegbahn, P. E. M.; Blomberg, M. R. A.; Chen, S.-L. J. Chem. Theory Comput. 2010, 6, 2040−2044. (20) Chen, S.-L.; Blomberg, M. R. A.; Siegbahn, P. E. M. J. Phys. Chem. B 2011, 115, 4066−4077. (21) Hirao, H. J. Phys. Chem. A 2011, 115, 9308−9313. (22) Warshel, A. Annu. Rev. Biophys. Biomol. Struct. 2003, 32, 425− 443. (23) Friesner, R. A.; Guallar, V. Annu. Rev. Phys. Chem. 2005, 56, 389−427. (24) Lin, H.; Truhlar, D. G. Theor. Chem. Acc. 2007, 117, 185−199. (25) Senn, H. M.; Thiel, W. Top. Curr. Chem. 2007, 268, 173−290. (26) Senn, H. M.; Thiel, W. Angew. Chem., Int. Ed. 2009, 48, 1198− 1229. (27) Hu, H.; Yang, W. Annu. Rev. Phys. Chem. 2008, 59, 573−601. (28) Jensen, K. P.; Ryde, U. Coord. Chem. Rev. 2009, 253, 769−778. (29) Jensen, K. P.; Ryde, U. J. Am. Chem. Soc. 2005, 127, 9117−9128. (30) Sharma, P. K.; Chu, Z. T.; Olsson, M. H. M.; Warshel, A. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 9661−9666. (31) Li, X.; Chung, L. W.; Paneth, P.; Morokuma, K. J. Am. Chem. Soc. 2009, 131, 5115−5125. (32) Bucher, D.; Sandala, G. M.; Durbeej, B.; Radom, L.; Smith, D. M. J. Am. Chem. Soc. 2012, 134, 1591−1599. (33) Pang, J.; Li, X.; Morokuma, K.; Scrutton, N. S.; Sutcliffe, M. J. J. Am. Chem. Soc. 2012, 134, 2367−2377. (34) Liao, R.-Z.; Thiel, W. J. Chem. Theory Comput. 2012, 8, 3793− 3803. (35) Liao, R.-Z.; Thiel, W. J. Comput. Chem. 2013, 34, 2389−2397. (36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, M. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;

Scheme 2. Suggested Reductive Debromination Mechanisms for NpRdhA on the Basis of Calculations

relative reactivity is somewhat overestimated. The defluorination of fluoro substrate has a very high barrier (36.6 kcal/mol), suggesting that this enzyme cannot catalyze the defluorination of 2,6-difluorophenolate. Further experimental studies would be helpful to verify our results.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.5b01502. Structures, energy diagrams, and Cartesian coordinates (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail for R.-Z.L.: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by startup funding from the Huazhong University of Science and Technology, the Swedish Research Council, the Knut and Alice Wallenberg Foundation, the National Natural Science Foundation of China (21503083, 21373027), the 111 Project (B07012), and Beijing Nova Program (Z151100000315055). Computer time was generously provided by the Swedish National Infrastructure for Computing.

■

REFERENCES

(1) Henschler, D. Angew. Chem., Int. Ed. Engl. 1994, 33, 1920−1935. (2) Fetzner, S. Appl. Microbiol. Biotechnol. 1998, 50, 633−657. (3) Hug, L. A.; Maphosa, F.; Leys, D.; Löffler, F. E.; Smidt, H.; Edwards, E. A.; Adrian, L. Philos. Trans. R. Soc., B 2013, 368, 20120322. 7357

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358

Research Article

ACS Catalysis Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision D.01; Gaussian, Inc., Wallingford, CT, 2009. (37) Roy, L. E.; Hay, P. J.; Martin, R. L. J. Chem. Theory Comput. 2008, 4, 1029−1031. (38) Becke, A. D. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098− 3100. (39) Perdew, J. Phys. Rev. B: Condens. Matter Mater. Phys. 1986, 33, 8822−8824. (40) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101. (41) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241. (42) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2009, 113, 6378−6396. (43) Isse, A. A.; Gennaro, A. J. Phys. Chem. B 2010, 114, 7894−7899. (44) Camaioni, D. M.; Schwerdtfeger, C. A. J. Phys. Chem. A 2005, 109, 10795−10797. (45) Liao, R.-Z.; Li, X.-C.; Siegbahn, P. E. M. Eur. J. Inorg. Chem. 2014, 2014, 728−741. (46) Liao, R.-Z.; Siegbahn, P. E. M. ACS Catal. 2014, 4, 3937−3949. (47) Liao, R.-Z.; Kärkäs, M. D.; Lee, B.-L.; Åkermark, B.; Siegbahn, P. E. M. Inorg. Chem. 2015, 54, 342−351. (48) Su, X.-J.; Gao, M.; Jiao, L.; Liao, R.-Z.; Siegbahn, P. E. M.; Cheng, J.-P.; Zhang, M.-T. Angew. Chem., Int. Ed. 2015, 54, 4909− 4914. (49) Marenich, A. V.; Ho, J.; Coote, M.; Cramer, C. J.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2014, 16, 15068−15106. (50) Marenich, A. V.; Majumdar, A.; Lenz, M.; Cramer, C. J.; Truhlar, D. G. Angew. Chem., Int. Ed. 2012, 51, 12810−12814. (51) Eliot, A. C.; Kirsch, J. F. Annu. Rev. Biochem. 2004, 73, 383−415. (52) Liao, R.-Z.; Ding, W.-J.; Yu, J.-G.; Fang, W.-H.; Liu, R.-Z. J. Comput. Chem. 2008, 29, 1919−1929. (53) Oliveira, E. F.; Cerqueira, N. M. F. S. A.; Fernandes, P. A.; Ramos, M. J. J. Am. Chem. Soc. 2011, 133, 15496−15505. (54) Rodkey, F. L.; Donovan, J. A., Jr. J. Biol. Chem. 1959, 234, 677− 680. (55) Cooper, M.; Wagner, A.; Wondrousch, D.; Sonntag, F.; Sonnabend, A.; Brehm, M.; Schuurmann, G.; Adrian, L. Environ. Sci. Technol. 2015, 49, 6018−6028. (56) Lee, C. H.; Dogutan, D. K.; Nocera, D. G. J. Am. Chem. Soc. 2011, 133, 8775−8777. (57) Wiedner, E. S.; Appel, A. M.; Dubois, D. L.; Bullock, R. M. Inorg. Chem. 2013, 52, 14391−14403. (58) Nippe, M.; Khnayzer, R. S.; Panetier, J. A.; Zee, D. Z.; Olaiya, B. S.; Head-Gordon, M.; Chang, C. J.; Castellano, F. N.; Long, J. R. Chem. Sci. 2013, 4, 3934−3945. (59) Solis, B. H.; Maher, A. G.; Honda, T.; Powers, D. C.; Nocera, D. G.; Hammes-Schiffer, S. ACS Catal. 2014, 4, 4516−4526. (60) O’Hagan, D. Chem. Soc. Rev. 2008, 37, 308−319. (61) Whittlesey, M. K.; Peris, E. ACS Catal. 2014, 4, 3152−3159. (62) Amii, H.; Uneyama, K. Chem. Rev. 2009, 109, 2119−2183. (63) Kuehnel, M. F.; Lentz, D.; Braun, T. Angew. Chem., Int. Ed. 2013, 52, 3328−3348.

7358

DOI: 10.1021/acscatal.5b01502 ACS Catal. 2015, 5, 7350−7358