Gas-Phase Mechanisms of Degradation of Hazardous

Jul 17, 2008 - Yoon Jeong Jang , Kibong Kim , Olga G. Tsay , David A. Atwood , and David G. Churchill. Chemical Reviews 2015 115 (24), PR1-PR76...
1 downloads 0 Views 1MB Size
Download PDF
9982

J. Phys. Chem. B 2008, 112, 9982–9991

Gas-Phase Mechanisms of Degradation of Hazardous Organophosphorus Compounds: Do They Follow a Common Pattern of Alkaline Hydrolysis Reaction As in Phosphotriesterase? Edyta Dyguda-Kazimierowicz,† W. Andrzej Sokalski,*,† and Jerzy Leszczynski‡ Department of Chemistry, Wrocław UniVersity of Technology, Wybrzez˙e Wyspian´skiego 27, 50-370 Wrocław, Poland, and Jackson State UniVersity, Jackson, Mississippi, 39217 ReceiVed: January 15, 2008; ReVised Manuscript ReceiVed: April 2, 2008 嘷 w This paper contains enhanced objects available on the Internet at http://pubs.acs.org/JPCB.

A comprehensive ab initio analysis of the gas-phase mechanisms of alkaline hydrolysis for a number of phosphotriesterase substrates;O,O-diisopropyl phosphorofluoridate (DFP), O-isopropyl methyl phosphonofluoridate, O,O-diethyl p-nitrophenyl phosphate (paraoxon), O,O-diethyl p-nitrophenyl thiophosphate (parathion), N-acetyl phosphoramidothioate (acephate), O,O-diethyl S-2-ethylthioethyl phosphorothioate (demetonS) and O-ethyl N,N-dimethyl phosphoramidocyanidate;has been presented herein. The results indicate that, although an associative mechanism of alkaline hydrolysis is followed by all these compounds, P-F and P-CN bonds are cleaved according to the multistep addition-elimination scheme, whereas the breakage of P-O and P-S bonds appears to be consistent with the one-step direct-displacement mechanism. Of the two alternative reaction pathways present in all those cases (except of acephate), the most probable one involves the proton from a nucleophilic hydroxide experiencing an additional stabilization by the phosphoryl oxygen atom. Introduction High toxicity of organophosphorus compounds, directed not only toward crop protection but also rising threat of being used against human population,1,2 resulted in an urgent need to develop safe and efficient decontamination methods. Consequently, there is a growing interest in the fundamental mechanisms of organophosphate hydrolysis as reflected by a number of contributions devoted to the computational investigation of phoshotriesters degradation.3–5 The non-harmful transformation of such hazardous chemicals could also be attained by enzymatic biodegradation6 and the factors determining the selection of leading enzyme candidates include both the efficiency of biocatalysis and the ability to react with a broad spectrum of substrates. From this perspective, bacterial phosphotriesterase (PTE)7 constitutes a prime target, as it has been shown to cleave a variety of phosphorus-ester bonds8,9 and to approach the limits of diffusion in the case of its best known substrate.10 Moreover, much experimental evidence (e.g., directed evolution of mutant enzymes) has shown that PTE substrate specificity as well as enantioselectivity can be modified, leading to engineered enzymes capable of enhanced decontamination of toxic agents.11 Such approaches, however, require a great amount of time-consuming laboratory work as well as dealing with hazardous materials that one could possibly avoid by application of theoretical methods allowing the relationship between mutations and catalytic activity to be predicted. Noticeably, despite many experimental (as referenced below) and theoretical efforts,12–21 the actual catalytic mechanism of PTE has remained unexplained. Since a detailed understanding of the driving force behind catalysis constitutes a crucial prerequisite in computer-aided catalyst design, the overall objective of this study has been to gain insight into the catalytic * Corresponding author. Tel./fax: +48 71 320 2457; [email protected]. † Wrocław University of Technology. ‡ Jackson State University.

properties of PTE that could further be utilized in the rational, knowledge-based control of enzyme catalytic activity. Considering that the mechanism of an enzyme-catalyzed reaction resembles the analogous gas phase process rather than the one taking place in a solution,22–25 the preliminary research presented herein has focused on the comprehensive study of possible gas phase mechanisms of the hydrolysis of PTE substrates. PTE exhibits wide substrate specificity encompassing the hydrolysis of phosphorus-oxygen (e.g., O,O-diethyl p-nitrophenyl phosphate (paraoxon), O,O-diethyl p-nitrophenyl thiophosphate (parathion)), phosphorus-fluorine (e.g., O,Odiisopropyl phosphorofluoridate (DFP), O-isopropyl methyl phosphonofluoridate (sarin, SA)), phosphorus-sulfur (e.g., O,Sdimethyl N-acetyl phosphoramidothioate (acephate), O,O-diethyl S-2-ethylthioethyl phosphorothioate (demeton-S)), and phosphoruscyanide (e.g., O-ethyl N,N-dimethyl phosphoramidocyanidate (tabun, TB)) bonds.9,26 Phosphomonoesters as well as phosphodiesters are not cleaved by PTE at a reasonable rate.27 Interestingly, a natural PTE substrate has not been revealed, whereas limits of diffusion are reached in the hydrolysis of paraoxon, the best substrate identified to date.10 Taking into account the chemical diversity of PTE substrates, it is essential to determine some common features of their hydrolysis that allow all of them to be accommodated in the PTE active site and subsequently subjected to catalysis. Experimental data for PTE chemical and kinetic mechanism suggest that phosphotriester hydrolysis occurs upon inversion of configuration at the phosphorus atom, consistent with an inline displacement mechanism.28 The pH-dependence of the activity suggests an involvement of one ionizable residue with a pKa value of 6.1.29 On the basis of the available X-ray structures and theoretical studies, it was proposed that this ionizable residue is the hydroxyl group that bridges the two zinc ions within the PTE active site.12 Alternatively, the most recent crystallographic analysis30 might indicate that the hy-

10.1021/jp800386f CCC: $40.75  2008 American Chemical Society Published on Web 07/17/2008

Gas-Phase Hydrolysis of PTE Substrates

J. Phys. Chem. B, Vol. 112, No. 32, 2008 9983

TABLE 1: The Kinetic Parameters for the Hydrolysis of Selected PTE Substrates DFP SA a

TB acephate

25.8 7.49 k2 1 E 9.1 10.1 kcatb 465 56 76.7 2.8

paraoxon 2.7 × 13.5 3170

10-2

SCHEME 1

parathion demeton-S 3.6 × 10-4 16.6 630 1.3

a Second-order rate constant, k2 (M-1 s-1), and activation energy, E ( kcal · mol-1), refer to the reaction with an aqueous base.34 b Turnover number, kcat (s-1) applies to the PTE-catalyzed process.9,26 Since the particular experiments were not carried out under identical conditions, caution should be used when comparing these numbers.

droxide in question is another solvent molecule terminally coordinated to the more buried zinc ion. Nonetheless, hydrolysis seems to be initiated by the hydroxide nucleophilic attack on the phosphorus atom followed by the phosphoester bond cleavage and the expulsion of a leaving group in the anionic form. In addition, the reaction might be facilitated by the zinc ion-substrate interaction resulting in the polarization of a phosphoryl oxygen bond and the increase in phosphorus atom electrophilicity.31 Comparison of kinetic isotope effects for alkaline and enzymatic hydrolysis of a P-O bond (paraoxon)32 suggests that the PTE-catalyzed process is slightly less associative than the one in solution. Nevertheless, it can be interpreted as consistent with an SN2-like concerted associative mechanism in both enzymatic and nonenzymatic phosphotriesters hydrolysis. The latter finding, along with the apparent involvement of a hydroxide as a nucleophile, was used to rationalize our approach that utilizes gas phase results for alkaline hydrolysis as a reasonable starting point for further study of the reaction occurring in the PTE active site. While the hydrolysis of anionic phosphomonoesters is likely to proceed via a dissociative, unimolecular mechanism that assumes the presence of a metaphosphate intermediate,32 phosphotriester hydrolysis seems to involve some kind of an associative pathway. As to the latter, two limiting scenarios are possible,33,34 i.e., a fully associative mechanism comprising an addition-elimination reaction associated with the presence of a pentacoordinate phosphorane intermediate or a direct-displacement mechanism accompanied by a single SN2-like transition state. Thus, the addition-elimination pathway is a two-step process in which the attack of a nuclephile results in the formation of an intermediate that further decomposes into products. On the other hand, the direct-displacement mechanism involves only one step in which the expulsion of a leaving group occurs at the same time the substituting nucleophile is entering. Current theoretical studies have suggested that, depending on the actual trisubstituted phosphoric acid derivative, both these pathways are possible. In particular, basic hydrolysis of several phosphofluoridate compounds (i.e., DFP, SA) was shown to proceed according to an addition-elimination pathway involving trigonal bipyramidal intermediates,3,5 whereas the hydrolysis of paraoxon is probably a single-step process consistent with a direct-displacement pathway.3 Since no theoretical analysis has yet been performed for a hydrolysis of a number of PTE substrates and, as it will be argued in what follows, alternative reaction pathways have remained to be explored, this contribution aims at systematic ab initio investigation of the possible gas phase mechanisms of alkaline hydrolysis of a variety of phosphoester bonds in compounds known as PTE substrates (Table 1 and Scheme 1). Additionally, the influence of solvent on the relative stability of structures occurring along particular reaction coordinates is examined.

For the analysis of P-F bond breakdown, two toxic organophosphate compounds, i.e., DFP and SA, were selected. Since an additional goal of the present analysis comprised establishing and validating the most cost-effective model chemistry, O,Odimethyl phoshorofluoridate was chosen as a model compound for a preliminary study of reaction pathways as well as the sensitivity of both geometry and energetics to the level of theory applied in calculations. The cleavage of the P-O bond was analyzed based on paraoxon and the related phosphorothioate, parathion. Similarly to DFP and SA, the toxicity of these organophosphate pesticides is attributed to the acetylcholinesterase inhibition that impairs nerve functions. As mentioned above, paraoxon is the best PTE substrate identified to date;its hydrolysis for the enzyme proceeds with kcat and kcat/Km of 104 s-1 and 108 M-1 s-1, respectively.35 Exchange of the phosphoryl oxygen atom by sulfur results in a slight decrease of the hydrolysis rate for both alkaline and enzymatic catalysis.9,34 Finally, alkaline hydrolysis of P-S and P-CN bonds was studied for the following trisubstituted phosphoryl derivatives: acephate, demeton-S, and TB. Importance of these particular compounds comes from the fact that acephate and demeton-S are commonly used as insecticides, while TB constitutes another toxic substance. It is noteworthy that, to our best knowledge, no computational study has yet been carried out for these compounds. In the case of acephate, the presence of electronrich domains (i.e., phosphoryl and carbonyl moieties) stabilizes the P-N bond, resulting in the cleavage of the P-S bond under alkaline conditions.36 A similar mechanism of decomposition applies to demeton-S, and, upon PTE action, both acephate and demeton-S are also hydrolyzed via the breakdown of a thiolester bond. Under neutral and basic conditions, TB is hydrolyzed to O-ethyl N,N-dimethylamido phosphoric acid and cyanide.2 Since the same products were found for the PTE-catalyzed process,26 a reaction coordinate involving P-CN bond cleavage was the subject of this investigation. Computational Details The gas phase reaction profiles were initially studied at the Hartree-Fock (HF) level with the 6-31+G(d) basis set. The nature of stationary points was verified by vibrational analysis. For all the first-order saddle points, intrinsic reaction coordinate (IRC) calculations37 were performed, revealing the geometries of the local minima associated with a given transition state. Since HF theory neglects the electron correlation effects, it is also unable to accurately predict the barrier heights. Theoretical studies of closely related compounds have shown that results of second-order Møller-Plesset (MP2) calculation of energy with the use of HF-optimized geometries are comparable to the analogous outcome obtained using structures optimized at the MP2 level of theory.3 Such an approach was further tested for

9984 J. Phys. Chem. B, Vol. 112, No. 32, 2008

Dyguda-Kazimierowicz et al.

Figure 1. HF/6-31+G(d) geometries of the stationary points along a reaction coordinate for the alkaline hydrolysis of O,O-dimethyl phosphorofluoridate (A path). Selected interatomic distances (Å) are indicated. The analogous distances associated with structures from the B path are given in parentheses. An animated graphic of the entire reaction pathway (based on IRC simulation, see Figure 2) is available in.mpg format.

a model compound, O,O-dimethyl phosphorofluoridate, and our calculation also indicated the relative insensitivity of HF geometries to the electron correlation effects (please refer to the Supporting Information). Thus, except for O,O-dimethyl phoshorofluoridate, where the size of a system allowed for the MP2/6-311++G(d,p) calculations of both geometry and energy, the energies reported herein refer to MP2/6-311++G(d,p) model chemistry applied to HF-optimized structures (denoted as the MP2/6-311++G(d,p)//HF/6-31+G(d) level of theory). Additionally, the model of demeton-S was simplified by truncation of the last methyl group belonging to the S-2-ethylthioethyl moiety (see Scheme 1 for the full structure). Thermodynamic properties (enthalpies and Gibbs free energies) were determined from vibrational frequencies computed at the fully optimized structures of stationary points along a reaction coordinate. Unless stated otherwise, energy values reported herein include zero-point vibrational energy (ZPE). To account for the influence of aqueous solvation, the polarizable continuum model (PCM)38 was applied in single-point MP2/ 6-311++G(d,p) calculations employing gas-phase HF/631+G(d) geometries. All calculations were performed using the Gaussian 03 program.39 Results and Discussion Multistep Mechanism: Hydrolysis of a Model Compound. Two possible reaction pathways starting from a common structure of a prereactive complex of subtrates (INT1) were determined at the HF/6-31+G(d) level and further refined at the MP2/6-311++G(d,p) level of theory. The difference between these two reaction coordinates consists in the orientation of an attacking hydroxide ion; in a lower energy pathway, its hydrogen atom points in the same direction as the phosphoryl oxygen (path designated as “A”), whereas the other one involves hydroxide positioned in an opposite way (path “B”). In what follows, the “a” and “b” suffixes will be used to distinguish between the structures associated with A and B paths (e.g., TSa and TSb). Structures related to A path are presented in Figure 1. For comparison of geometries for both paths, please refer to the Supporting Information. Since the four transition-state

Figure 2. Energy profiles for the alkaline hydrolysis of O,O-dimethyl phosphorofluoridate. Potential energy changes with respect to the separately optimized reactants structure, INT1 (black dotted line; ZPE not included), were generated by four independent HF/6-31+G(d) IRC calculations starting with TS1, TSr1, TSr2, and TS2 geometries. Reaction coordinate “0” corresponds to the INT3 structure; its negative and positive values indicate the directions toward reactants (INT1) and products (INT5a or INT5b), respectively. Variation in the interatomic distances is given for a phosphorus-hydroxide oxygen as well as phosphorus-fluorine atom separation. (a) Reaction pathway denoted as A (hydroxide proton pointing in a direction of a phosphoryl oxygen atom). (b) Reaction pathway denoted as B (with a hydroxide proton positioned oppositely).

geometries were located along each of those pathways, HF/631+G(d) IRC simulation was employed to verify the connectivity of particular reaction pathway components. The results for A and B paths are plotted in Figure 2. Additionally, MP2/6311++G(d,p)//HF/6-31+G(d) level energetics of both paths is compared in Figure 3. The molecular scenario of the alkaline hydrolysis of O,Odimethyl phoshorofluoridate is consistent with the additionelimination mechanism, as described in the Introduction. The consecutive steps of this process can be summarized as follows. First, a stable complex of reactants (INT1) is formed that subsequently proceeds through the two possible transition states for nucleophile attack (TS1a or TS1b) to the two pentacoordinate intermediates (INT2a and INT2b, respectively). TS1a with the hydroxide pointing “upward” (i.e., in the same direction as a phosphoryl oxygen atom) is associated with a lower energy barrier (∆E ) 3.6 compared to 5.4 kcal · mol-1; see Table 2). Since the intermediate decomposition requires both methyl groups of O,O-dimethyl phoshorofluoridate to be oriented toward a fluorine atom, the next steps involve two subsequent

Gas-Phase Hydrolysis of PTE Substrates

J. Phys. Chem. B, Vol. 112, No. 32, 2008 9985

Figure 3. Potential energy surface along a reaction coordinate for the alkaline hydrolysis of O,O-dimethyl phosphorofluoridate. MP2/6311++G(d,p) relative energies (normalized with respect to the reactants geometry, INT1; ZPE included) were evaluated for structures fully optimized at the HF/6-31+G(d) level of theory. Energy barriers for the chemical steps encompassing the formation of the first intermediate and decomposition of the final one are indicated.

TABLE 2: MP2/6-311++G(d,p)//HF/6-31+G(d) Relative Energies (kcal · mol-1) for the Hydrolysis of O,O-Dimethyl Phoshorofluoridate INT1 INT1 INT2 INT2 INT3 INT3 INT4 INT4 INT1

f f f f f f f f f

TS1 INT1 TSr1 INT2 TSr2 INT3 TS2 INT5 INT5

∆Ea

∆Gb

3.6 (5.4)c -11.9 (-6.9) 2.5 (0.7) -1.5 (-6.4) 3.1 (4.1) 0.1 (0.7) 4.8 (4.4) -1.5 (-3.3) -14.7 (-16.0)

4.8 (7.2) -10.0 (-4.9) 3.0 (1.1) -1.5 (-6.7) 3.6 (4.7) 0.7 (0.7) 4.4 (4.4) -4.3 (-4.7) -15.2 (-15.5)

Figure 4. HF/6-31+G(d) geometries of the stationary points along a reaction coordinate for the alkaline hydrolysis of DFP (A path). Selected interatomic distances (Å) are indicated. The analogous distances associated with structures from the B path are given in parentheses.

a ∆E corresponds to the sum of electronic and zero-point energies. b ∆G stands for the Gibbs free energy. c The results for the B path are given in parentheses.

rotations proceeding through the two transition states (TSr1a, TSr2a or TSr1b, TSr2b) separated by a common intermediate, INT3, with the attached hydroxyl group situated as in the A path (the alternative hydroxyl orientation was found to be unstable). Energy barriers within the A path are similar for both rotations (∆E ) 2.5 and 3.1 kcal · mol-1), while, in case of the B path, the first rotation almost lacks any barrier, and the second one has a slightly higher activation barrier of 4.1 kcal · mol-1. Trigonal bipyramidal intermediates with both methyl groups oriented to stabilize the leaving fluoride (INT4a and INT4b) seem to be of similar stability: activation barriers for their decomposition are ∆E ) 4.8 and 4.4 kcal · mol-1 for the A and B paths, respectively. Finally, the transition states associated with the loss of fluoride (TS2a, TS2b) lead to the expected products, i.e., phosphate and fluoride. The overall energetic effects of both O,O-dimethyl phoshorofluoridate hydrolysis pathways (i.e., INT1 f INT5a or INT1 f INT5b) encompass ∆E ) -14.7 (A path) and -16.0 kcal · mol-1 (B path). Among the four transition state structures located along each of the reaction coordinates, those associated with the formation and decomposition of the pentacoordinate intermediates (i.e., TS1 and TS2) seem to influence the overall rate of reaction. Judging from the comparison of Gibbs free energy changes (∆G; see Table 2) associated with A and B paths, the A path involving additional stabilization of hydroxyl proton by the phosphoryl oxygen atom is energetically more favorable. In both cases, intermediate formation is probably the rate-limiting step with the highest Gibbs free energy barrier.

Figure 5. HF/6-31+G(d) geometries of the stationary points along a reaction coordinate for the alkaline hydrolysis of SA (A path). Selected interatomic distances (Å) are indicated. The analogous distances associated with structures from the B path are given in parentheses.

While our findings regarding the alkaline hydrolysis of O,Odimethyl phoshorofluoridate are in agreement with the results obtained by Ornstein et al.,3 the reaction coordinate proposed therein corresponds to the B path from our study, that is the lower energy pathway was not determined. Multistep Mechanism: Hydrolysis of DFP and SA. Alkaline DFP and O-isopropyl methyl phosphonofluoridate hydrolysis was found to be analogous to that of O,O-dimethyl phoshorofluoridate (see Figures 4 and 5 for A path structures; the comparison of both path geometries is provided in the Supporting Information). Since the intermediate formation (following the hydroxide attack) as well as its decomposition (i.e., fluoride departure) were assumed to play the dominant role in an overall reaction kinetics, the rotation steps were not taken into account. In the case of SA hydrolysis, the second part of a reaction coordinate (starting with the intermediate that immediately precedes the loss of fluoride) proceeds only via the A path. The MP2/6-311++G(d,p) level energetics of

9986 J. Phys. Chem. B, Vol. 112, No. 32, 2008

Dyguda-Kazimierowicz et al.

TABLE 3: Relative Energies (kcal · mol-1) for the Multistep Hydrolysis of DFP, SA, TB, and Acephate as Determined at the MP2/6-311++G(d,p)//HF/6-31+G(d) Level of Theory ∆Ea DFP

sarin

tabun

acephate

INT1 INT1 INT2 INT3 INT3 INT1 INT1 INT1 INT2 INT3 INT3 INT1 INT1 INT1 INT2 INT3 INT3 INT1 INT1 INT1 INT2 INT2 INT1

f f f f f f f f f f f f f f f f f f f f f f f

TS1 INT2 INT3 TS2 INT4 INT4 TS1 INT2 INT3 TS2 INT4 INT4 TS1 INT2 INT3 TS2 INT4 INT4 TS1d INT2 TS2 INT3 INT3

(5.0)c

3.3 -9.5 (-5.2) -1.9 (-5.6) 2.3 (1.9) -2.7 (-4.3) -14.2 (-15.0) 3.2 (6.3) -9.1 (-4.4) -2.4 (-7.1) 3.0 -2.2 -13.6 5.7 (8.6) -10.8 -0.5 0.4 -17.3 -28.5 38.5 25.1 1.8 -3.1 22.0

∆Gb 5.3 (7.6) -7.5 (-3.6) -1.8 (-4.6) 2.7 (1.9) -5.2 (-6.5) -14.5 (-14.7) 4.4 (7.6) -7.4 (-3.3) -2.7 (-6.7) 3.0 -3.3 -13.3 6.2 (9.4) -9.5 -1.0 1.0 -21.6 -32.1 41.1 27.4 1.9 -7.8 19.6

a ∆E corresponds to the sum of electronic and zero-point energies. b ∆G stands for the Gibbs free energy. c The results for the B path are given in parentheses. d Results for the first step of acephate hydrolysis correspond to the reaction between a water molecule and a deprotonated acephate.

particular steps is compared in Table 3. A graphical representation of the corresponding potential energy surface is provided in Figure 6. For both A and B paths of DFP and SA hydrolysis, an intermediate formation step seems to be associated with a higher energy barrier compared to the fluoride expulsion (see Table 4 for the direct comparison of activation energy barriers). While comparison of A and B path energy barriers for the hydroxide attack reveals an almost two-fold difference in favor of the A path, the fluoride departure step for both paths of DFP hydrolysis is associated with a similar energy barrier. This finding along with a more favorable orientation of hydroxide occurring in A path structures (e.g., INT2a and INT3a geometries are lower in energy than their B path counterparts by 4.3 and 0.7 kcal · mol-1, respectively; see the energy diagram in Figure 6) support the conclusion that alkaline hydrolysis of DFP proceeds mainly according to the A path. Since the loss of fluoride in the case of SA occurs only via the A path, which also favors the initial formation of a pentavalent intermediate, the overall sarin hydrolysis appears to involve only the A path structures, i.e., those with a hydroxide stabilized by an additional interaction with a phosphoryl oxygen atom. Similarly to the above-presented hydrolysis of O,O-dimethyl phoshorofluoridate, the results discussed herein are consistent with the mechanism of DFP and SA hydrolysis proposed by Ornstein et al.3 Nonetheless, close inspection of the structures and energetics of the corresponding potential energy surfaces supports the notion that only a higher energy reaction pathway was proposed therein. While the results of Cramer et al. regarding alkaline SA hydrolysis5 encompass the reaction coordinate characterized by a more favorable hydroxide orientation, no attempts have been made to investigate an alternative reaction pathway. Regarding the available experimental data, the highest Gibbs free energy barrier along the SA hydrolysis

pathway, i.e., 4.4 kcal · mol-1 (see Table 4) does not seem to compare favorably with an experimental value of 9.1 kcal · mol-1 (Table 1). Again, this theoretical analysis has been carried out for gas-phase structures, while experimental data apply to an aqueous solution. Multistep Mechanism: Hydrolysis of TB. Alkaline hydrolysis of the P-CN bond in O-ethyl N,N-dimethyl phosphoramidocyanidate molecule resembles that of P-F bond cleavage in DFP and sarin. The structure summary is provided in Figure 7. As discussed previously, starting with the same structure of reactants complex (INT1), two transition state structures are possible (either TS1a or TS1b) depending on the way a nucleophile approaches a phosphorus atom. The proton pointing “upward” (A path) is probably better stabilized by the interaction with a phosphoryl oxygen atom, which additionally lowers the activation energy barrier (∆E ) 5.7 kcal · mol-1 as compared to 8.7 kcal · mol-1 for the B path; see Table 3). However, all the remaining structures along a reaction coordinate correspond to the A path, and efforts to locate the corresponding geometries with a hydroxide proton pointing “downward” led to the cyanide departure (i.e., INT4 structure). Nonetheless, complete A path for TB hydrolysis comprises the TS1a structure followed by the formation of a pentavalent intermadiate (INT2) that undergoes a slight reorganization of N,N-dimethyl substituent (another pentacoordinate intermediate denoted as INT3) and can finally be decomposed via the second transition state, TS2. The latter process is almost barrierless and, similarly to the P-F bond hydrolysis, the rate-controlling step of the P-CN bond breakage consists in the formation of the first pentacoordinate intermediate. Experimental activation energy for the hydrolysis of TB by an aqueous base is higher by 1.0 kcal · mol-1 compared to that of SA (Table 1). Noticeably, reasonable resemblance also occurs for Gibbs free energy barriers associated with a rate-limiting step of SA and TB (Table 4). Multistep Mechanism: Hydrolysis of Acephate. The mechanism of acephate hydrolysis via a thiolester bond cleavage differs remarkably from the prior results regarding P-F and P-CN bond hydrolysis. In all those previous cases, two alternative orientations of an attacking nucleophile were observed, resulting in the two parallel reaction pathways. Herein, only the hydroxide position that allows its proton to interact with a phosphoryl oxygen is predicted (Figure 8). Furthermore, despite the determination of a transition state geometry corresponding to direct hydroxyl ion attack on the phosphorus atom, we were unable to optimize the structure of hydrogen-bonded reactants (INT1) that would comprise the neutral amide moiety (as in the case of TS1) and the separate hydroxide; all the attempts resulted in a water molecule stabilized by two hydrogen bonds to a negatively charged nitrogen and an oxygen atom from the carbonyl group. Consequently, the energy barrier for TS1 formation is rather large (∆E ) 38.5 kcal · mol-1; see Table 3), and it probably does not encompass a single step, i.e., some intermediate structure(s) or an alternative pathway might be involved. Similarly to the case of P-F bond cleavage, TS1 leads to a pentacoordinate intermediate (INT2) that is further decomposed through methylthiolate expulsion. The height of an energy barrier associated with the transition state for intermediate decomposition, TS2 (∆E ) 1.8 kcal · mol-1) is of minor importance compared to the barrier for its formation (Table 4). The overall reaction is endothermic, and this finding might also suggest the existence of some alternative pathway, especially for the intermediate formation step. It is then interesting to consider the molecular events occurring along a reaction coordinate leading from an acephate-water molecule complex

Gas-Phase Hydrolysis of PTE Substrates

J. Phys. Chem. B, Vol. 112, No. 32, 2008 9987

Figure 6. Potential energy surface along a reaction coordinate for the hydrolysis of (a) DFP, (b) SA, (c) TB, and (d) acephate. MP2/6-311++G(d,p) relative energies (normalized with respect to the reactants geometry, INT1; ZPE included) were evaluated for structures fully optimized at the HF/6-31+G(d) level of theory. Energy barriers for the chemical steps encompassing the formation of the first intermediate and decomposition of the final one are indicated. Results for the first step of acephate hydrolysis correspond to the reaction between a water molecule and a deprotonated acephate.

TABLE 4: Multistep Hydrolysis: The Activation Energy Barriersa as Determined at the MP2/6-311++G(d,p)//HF/ 6-31+G(d) Level of Theory DFP A B

b

∆Ea ∆Gae ∆EaPCMf ∆Ea ∆Ga ∆EaPCM

(5.0)c

3.3 5.3 (7.6) 3.4 (0.6) 2.3 (1.9) 2.7 (1.9) 0.5 (1.9)

sarin

tabun

acephate

3.2 (6.3) 4.4 (7.6) 3.1 (3.1) 3.0 3.0 -2.3

5.7 (8.6) 6.2 (9.4) 4.6 (4.9) 0.4 1.0 0.1

38.5d 41.1 33.4 1.8 1.9 1.3

a In units of kcal · mol-1; part “A” ) intermediate formation; part “B” ) intermediate decomposition. b ∆Ea corresponds to the sum of electronic and zero-point energies. c The results for the B path are given in parentheses. d Results for the first step of acephate hydrolysis correspond to the reaction between a water molecule and a deprotonated acephate. e ∆Ga stands for the Gibbs free energy in solution. f ∆EaPCM stands for the electronic energy in solution.

to the TS1 encompassing separate hydroxide and a neutral amide moiety. An energy profile for the first step of acephate hydrolysis (i.e., intermediate formation) was obtained from the IRC simulation starting with the TS1 structure, as plotted in Figure 9. Progress along the reaction coordinate in the direction of the reactants is associated with a gradual increase in the phosphorushydroxide oxygen atom distance. Simultaneously, the amide proton is being attracted by the latter, and the N-H distance also increases while decreasing the O-H separation. That results in the formation of a water molecule and deprotonation of an amide moiety. Although no stable geometry was found along this part of the reaction coordinate, the value of an energy barrier

Figure 7. HF/6-31+G(d) geometries of the stationary points along a reaction coordinate for the alkaline hydrolysis of TB (A path). Selected interatomic distances (Å) are indicated. The analogous distances associated with structures from the B path are given in parentheses.

for the formation of TS1 starting with the structure comprising neutral amide nitrogen was estimated to be 6.9 kcal · mol-1 (at the HF/6-31+G(d) level of theory, no ZPE included). The experimental 2.2-fold acceleration of PTE-catalyzed acephate hydrolysis as compared to demeton-S degradation (Table 1) is associated with the 0.5 kcal · mol-1 decrease in a free energy barrier. Therefore, the corresponding ∆EHF (without ZPE

9988 J. Phys. Chem. B, Vol. 112, No. 32, 2008

Figure 8. HF/6-31+G(d) geometries of the stationary points along a reaction coordinate for the hydrolysis of acephate. Selected interatomic distances (Å) are indicated. The analogous distances associated with structures from the B path are given in parentheses. An animated graphic of the first part of the reaction pathway (based on IRC simulation, see Figure 9) is available in.mpg format.

Figure 9. Energy profile for the first step of acephate hydrolysis. Potential energy changes with respect to the separately optimized reactants structure, INT1 (black dotted line; ZPE not included) were generated by HF/6-31+G(d) IRC calculation starting with the TS1 structure (reaction coordinate “0”). Variation in the interatomic distances is given for a phosphorus-hydroxide oxygen, water molecule hydrogen-hydroxideoxygenaswellaswatermoleculehydrogen-nitrogen atoms separation. The value of 6.9 kcal · mol-1 reflects the difference between the TS1 energy and the energy associated with a reaction coordinate, where the curves describing O-H and N-H distances intersect each other.

correction) for demeton-S hydrolysis, i.e., 6.9 kcal · mol-1, is in qualitative agreement with experiment. However, further analysis of the influence of acephate protonation state on the mechanism of hydrolysis is required. Single-Step Mechanism: Hydrolysis of Paraoxon and Parathion. In contrast to the complex addition-elimination reaction pathways already discussed, base-catalyzed hydrolysis of paraoxon and parathion proceeds according to the directdisplacement mechanism. Experimental kinetic isotope effects data suggest the presence of a transition state structure that is best described by an associative SN2-like mechanism for the hydrolysis occurring in a concerted, asynchronous way with an inversion of configuration at the phosphorus center.32 What should also be mentioned is that analogous conclusions were

Dyguda-Kazimierowicz et al. put forward for the enzymatic paraoxon hydrolysis.32 Moreover, the enzyme does not seem to contribute to the departing group activation by protonation of the latter.31 Thus, similarly to the reaction proceeding in an aqueous solution, collapse of the transition state results in the expulsion of an anionic pnitrophenolate. Considering the influence of a sulfur atom substitution for the phosphoryl oxygen (as observed in case of parathion relative to paraoxon), the slower rate of chemical hydrolysis of tiophosphates is consistent with the reduced electronegativity of sulfur and the resulting decrease in an electrophilic character of the phosphorus center.34 Probably, the extent of a phenolic oxygen bond breakage in thiophosphates is less developed in comparison with the corresponding phosphates.31 To the best of our knowledge, base-catalyzed paraoxon hydrolysis has been investigated theoretically only by Ornstein et al.3 Surprisingly, although our findings encompass the two possible reaction pathways for gas phase hydrolysis of paraoxon, neither of them is associated with such a low energy barrier (i.e., 0.8 kcal · mol-1) as the one proposed by Ornstein et al.3 The structures along the A reaction pathway for basecatalyzed hydrolysis of both paraoxon and parathion are given in Figure 10 (for comparison of structural data for A and B paths, please refer to the Supporting Information). Table 5 provides activation energy barriers and reaction energies. The corresponding mechanism involves a single transition state that is further decomposed by a p-nitrophenolate departure. No pentacoordinate intermediates were located along the reaction coordinate. The results for paraoxon and parathion are analogous and differ mainly in terms of energetics (see Table 6 for a summary). Starting with a stable complex of reactants (INT1), the two alternative transition state structures may develop that differ in the position of an attacking hydroxyl group (TSa and TSb). Either of them undergoes decomposition, leading to the products (INT2a and INT2b, respectively). The difference in the height of activation energy barriers for the A and B paths is not significant: ∆E ) 7.4 compared to 8.7 kcal · mol-1 in the case of paraoxon and ∆E ) 8.6 and 9.7 kcal · mol-1 for parathion. Nevertheless, the lower energy pathway is again the one with a hydroxide proton being stabilized by the interaction with a phosphoryl oxygen (sulfur). Since the attacking hydroxide approaches closer to the phosphorus center in the case of the B path (see structural details indicated in Figure 10), the higher energy barrier associated with the latter may result from an additional strain. Simultanously, the extent of a phosphoester bond breakage is larger. In agreement with experimental evidence,31 the extent of phenolic oxygen bond cleavage in parathion is less developed in comparison with paraoxon (Figure 10). Noticeably, the results discussed above are consistent with experimental data suggesting associative bimolecular mechanism of alkaline hydrolysis that involves a single SN2-like transition state structure.32 Relative activation energy barriers for parathion and paraoxon decomposition are in a remarkable agreement with the chemical kinetics data. Assuming the 5.6-fold rate acceleration when comparing parathion and paraoxon hydrolysis40 (the experimental data from Table 1 are not used herein, as they were not gathered under common temperature), it corresponds to a decrease in activation free energy of about 1.1 kcal · mol-1 (at 40 °C). Moreover, a comparison of turnover numbers for a PTE-catalyzed reaction results in similar values. Thus, the difference in the activation free energy barrier for parathion and paraoxon (i.e., ∆∆G ) 1.1 kcal · mol-1; see Table 6) resulting from our model is in excellent agreement with the experimental results related to the enzymatic reaction.

Gas-Phase Hydrolysis of PTE Substrates

J. Phys. Chem. B, Vol. 112, No. 32, 2008 9989

Figure 10. HF/6-31+G(d) geometries of the stationary points along a reaction coordinate for the alkaline hydrolysis of (a) paraoxon, (b) parathion, and (c) demeton-S. Selected interatomic distances (Å) are indicated. The analogous distances associated with structures from the B path are given in parentheses. An animated graphic of the entire reaction pathway for demeton-S (based on IRC simulation, see Figure S2, Supporting Information) is available in.mpg format.

TABLE 5: Relative Energies (kcal · mol-1) for the Single-Step Hydrolysis of Paraoxon, Parathion, and Demeton-S as Determined at the MP2/6-311++G(d,p)//HF/ 6-31+G(d) Level of Theory paraoxon parathion demeton-S

INT1 INT1 INT1 INT1 INT1 INT1

f f f f f f

TS INT2 TS INT2 TS INT2

∆Ea

∆Gb

7.4 (8.7)c -33.1 (-34.1) 8.6 (9.7) -32.0 (-32.5) 4.8 (6.4) -28.7 (-29.9)

7.7 (9.3) -38.5 (-39.8) 8.8 (10.3) -37.6 (-37.7) 6.8 (8.7) -32.7 (-34.4)

a ∆E corresponds to the sum of electronic and zero-point energies. b ∆G stands for the Gibbs free energy. c The results for the B path are given in parentheses.

TABLE 6: Single-Step Hydrolysis: The Activation Energy Barriers (in units of kcal · mol-1) as Determined at the MP2/6-311++G(d,p)//HF/6-31+G(d) Level of Theory ∆Eaa ∆Gac ∆EaPCMd

paraoxon

parathion

demeton-S

7.4 (8.7)b 7.7 (9.3) 6.2 (4.9)

8.6 (9.7) 8.8 (10.3) 7.1 (5.3)

4.8 (6.4) 6.8 (8.7) 9.1 (7.2)

a

∆Ea corresponds to the sum of electronic and zero-point energies. b The results for the B path are given in parentheses. c ∆Ga stands for the Gibbs free energy in solution. d ∆EaPCM stands for electronic energy in solution.

Single-Step Mechanism: Hydrolysis of Demeton-S. Basic hydrolysis of demeton-S follows a route analogous to the directdisplacement mechanism as determined for paraoxon and parathion (see Figure 10 and Table 5). In particular, there are two possible reaction pathways (denoted as A and B paths; B path geometries are provided in the Supporting Information) differing in the hydroxide orientation. Each of them involves a single SN2-like transition state leading to thiolate formation. As it was shown for paraoxon and parathion, the attacking hydroxide approaches closer to the phosphorus center in the case of the B path (P-O(H) distance of 2.543 compared to 2.690 Å; see Figure 10) but probably because of the lack of favorable

interaction between a hydroxide proton and a phoshoryl oxygen, the transition state representing the B path is associated with a higher energy barrier (∆E ) 6.4 and 4.8 kcal · mol-1 for B and A paths, respectively; see Table 6). Since thiolate is a better leaving group than is the alkoxide anion,34 the activation energy barrier associated with demeton-S hydrolysis is lower compared to the corresponding values for P-O bond hydrolysis of paraoxon and parathion. The difference in the way of P-S bond cleavage in the case of demeton-S and acephate likely results from the remarkably dissimilar substituents at a phosphorus atom, i.e., an ethoxyl group being replaced by an amide moiety. Presumably, phosphorothioates containing two alkoxy groups in addition to the thiolate substituent are hydrolyzed according to a single-step direct-displacement mechanism. Finally, a qualitative agreement with the chemical kinetics data can be seen in the case of demeton-S and DFP turnover numbers for PTE-catalyzed hydrolysis. A 360-fold rate acceleration (Table 1) is reflected by the decrease in the activation free energy of about 3.6 kcal · mol-1. The corresponding difference in the free energy barrier for demeton-S and DFP (i.e., ∆∆G ) 1.5 kcal · mol-1) resulting from our model is qualitatively consistent with the above experimental data. Alkaline Hydrolysis versus PTE-Catalyzed Process. Because of the lack of experimental data for gas-phase alkaline hydrolysis of the compounds studied in this contribution, a concluding validation of the theoretical results presented herein cannot be performed. To relate our results to available experimental quantities corresponding to analogous processes in aqueous solution (see Table 1), the self-consistent reaction field model38 of bulk water has been applied. Solvation-induced effects on the potential energy surface for particular reactions are given in the Supporting Information. Additionally, the activation energy barriers in solution are compared in Tables 4 and 6. In general, the height of energy barriers is decreased upon solvation, with the most pronounced differences occurring for the B paths. As a consequence, the reaction pathways involving a hydroxide proton positioned in an opposite way relative to the phosphoryl oxygen atom are associated with lower

9990 J. Phys. Chem. B, Vol. 112, No. 32, 2008 energy barriers than those employing an alternative hydroxide arrangement. Apparently, solvent is better suited for stabilization of the attacking hydroxyl group than is the phosphoryl oxygen alone. Since analogous solvent effects can be observed regarding the energy barriers for intermediate decomposition, it is presumably the intermediate formation step that controls the overall reaction rate, similar to the gas-phase results. Additionally, on account of the strong polarity of the compounds studied herein, the reactant complex may no longer be associated with a minimum on the potential energy surface, and the reaction barrier for the intermediate formation (or the overall reaction barrier in the case of single-step mechanisms) might correspond to the amount of energy required for separated reactants to reach the transition state geometry.3 This could possibly explain an apparent discrepancy between experimental data and the results discussed herein. Another likely reason for the observed inconsistency is the error introduced by the use of the implicit solvation method itself.41 As it has been shown for SA and soman,42 PCMs do not seem to be capable of reproducing the experimental data. It is possibly due to the greater resemblance between the gas phase and the enzymatic transition state22–25 that an exact agreement has been obtained for paraoxon and parathion gas-phase results and experimental turnover numbers for the PTE-catalyzed process.40 Since including the solvent effects by means of quantum mechanical continuum model does not seem to yield a reliable outcome, a comparison to experimental data obtained for processes occurring in an aqueous solution is also inconclusive. As it can be seen from Table 1, the ordering of compounds in terms of their susceptibility to hydrolysis undergoes a major change upon PTE action. The most pronounced example encompasses paraoxon, which, despite being relatively insensitive to hydrolysis, is the best PTE substrate. Nonetheless, two recent theoretical analyses of PTE-catalyzed hydrolysis of paraoxon 20,21 have confirmed the mechanism proposed on the basis of the experimental kinetic isotope effects.32 In particular, no protonation of a leaving p-nitrophenolate is required, and the bridging hydroxide is a sufficiently strong nucleophile to initiate the reaction. It is worth emphasizing, however, that the reaction coordinate observed for an enzyme-catalyzed process might differ from that in a gas phase or an aqueous solution. In the case of PTE-catalyzed cleavage of paraoxon, the singlestep process proposed herein is likely to be modified, as additional intermediates have been observed along a reaction pathway.21 Still, the general characteristics of these two mechanisms have remained unaltered, and therefore the knowledge of the most probable events following the attachment of a hydroxide is of great importance for the successful modeling of a reaction taking place in an enzyme active site. Conclusions The following conclusions can be drawn upon our results of quantum chemical modeling of the gas-phase alkaline hydrolysis of a variety of organophosphorus compounds known as PTE substrates: • While all base-catalyzed hydrolysis reactions studied herein appear to follow an associative mechanism, the cleavage of P-O and P-S bonds (except of acephate molecule) occurs according to a one-step direct-displacement mechanism involving the presence of a single SN2-like transition state, whereas the hydrolysis of P-F and P-CN bonds is consistent with an addition-elimination scenario employing several trigonal bipyramidal intermediates. • Except for acephate, two alternative reaction pathways are possible for each of these mechanisms that differ in the position

Dyguda-Kazimierowicz et al. of the attacking hydroxide relative to the phosphoryl oxygen atom. Apparently, the most energetically favorable reaction coordinate involves the hydroxide proton being stabilized by a phosphoryl oxygen. It remains an open question which mechanism will be exploited inside the PTE active site. • In the case of a multistep addition-elimination mechanism, relatively significant energy barriers are associated with the nucleophilic attack of a hydroxide (i.e., formation of the first intermediate) and the departure of a leaving group (i.e., decomposition of the final intermediate). Judging from the results of O,O-dimethyl phoshorofluoridate hydrolysis, the energy barriers for conformational transitions are of minor importance relative to the chemical steps encompassing the formation or breakage of a chemical bond. • The rate-limiting step of multistep mechanisms appears to be associated with an intermediate formation. • Since all the reaction pathways reported herein constitute essentially the variants of an associative mechanism of hydrolysis, they could presumably be accommodated by a common active site of PTE. Therefore, our current results will be useful as a starting point for a QM/MM simulation of a PTE-catalyzed process. Acknowledgment. This work was supported by Jackson State University subcontract #W912HZ-04-2-0002. We also thank Wrocław University of Technology for support. Professor Frank M. Raushel is gratefully acknowledged for reading and commenting on this manuscript. Calculations were performed at the Wrocław Supercomputer and Networking Center (WCSS), Poznan´ Supercomputer and Networking Center (PCSS), and the Interdisciplinary Center for Modeling (ICM) in Warsaw. Supporting Information Available: Comparison of structures and energies for the model compound hydrolysis as determined at different levels of theory, IRC energy profile for demeton-S hydrolysis, illustrations of stationary point geometries along A and B paths of all the compounds studied herein, energetics of particular reaction pathways in solution, as well as animation of the B path trajectory for the model compound hydrolysis. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Toy, A. D. F.; Walsh, E. N. Phosphorus Chemistry in EVeryday LiVing; American Chemical Society Publications: Washington, DC, 1987; pp 285-294. (2) Munro, N. B.; Talmage, S. S.; Griffin, G. D.; Waters, L. C.; Watson, A. P.; King, J. F.; Hauschild, V. EnViron. Health Perspect. 1999, 107, 933– 974. (3) Zheng, F.; Zhan, C. G.; Ornstein, R. L. J. Chem. Soc., Perkin. Trans. 2001, 2, 2355–2363. (4) Xiong, Y.; Zhan, C. G. J. Org. Chem. 2004, 69, 8451–8458. (5) Seckute, J.; Menke, J. L.; Emnett, R. J.; Patterson, E. V.; Cramer, C. J. J. Org. Chem. 2005, 70, 8649–8660. (6) Raushel, F. M. Curr. Opin. Microbiol. 2002, 5, 288–295. (7) Dumas, D. P.; Caldwell, S. R.; Wild, J. R.; Raushel, F. M. J. Biol. Chem. 1989, 264, 19659–19665. (8) Dumas, D. P.; Durst, H. D.; Landis, W. G.; Raushel, F. M.; Wild, J. R. Arch. Biochem. Biophys. 1990, 277, 155–159. (9) Lai, K.; Stolovich, N. J.; Wild, J. R. Arch. Biochem. Biophys. 1995, 318, 59–64. (10) Caldwell, S. R.; Newcomb, J. R.; Schlecht, K. A.; Raushel, F. M. Biochemistry 1991, 30, 7438–7444. (11) Hill, C. M.; Li, W. S.; Thoden, J. B.; Holden, H. M.; Raushel, F. M. J. Am. Chem. Soc. 2003, 125, 8990–8991. (12) Zhan, C. G.; de Souza, O. N.; Rittenhouse, R.; Ornstein, R. L. J. Am. Chem. Soc. 1999, 121, 7279–7282. (13) Koca, J.; Zhan, C. G.; Rittenhouse, R. C.; Ornstein, R. L. J. Am. Chem. Soc. 2001, 123, 817–826.

Gas-Phase Hydrolysis of PTE Substrates (14) Koca, J.; Zhan, C. G.; Rittenhouse, R. C.; Ornstein, R. L. J. Comput. Chem. 2003, 24, 368–378. (15) Pang, Y.-P. Proteins: Struct., Funct., Genet. 2001, 45, 183–189. (16) Kafafi, S. A.; Krauss, M. Int. J. Quantum Chem. 1999, 75, 289– 299. (17) Krauss, M. J. Chem. Inf. Model. 2001, 41, 8–17. (18) Krauss, M.; Olsen, L.; Antony, J.; Hemmingsen, L. J. Phys. Chem. B 2002, 106, 9446–9453. (19) Soares, T. A.; Osman, M. A.; Straatsma, T. P. J. Chem. Theory Comput. 2007, 3, 1569–1579. (20) Wong, K.-Y.; Gao, J. Biochemistry 2007, 46, 13352–13369. (21) Chen, S.-L.; Fang, W.-H.; Himo, F. J. Phys. Chem. B 2007, 111, 1253–1255. (22) Dewar, M. J. S.; Storch, D. M. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 2225–2229. (23) Lightstone, F. C.; Zheng, Y.-J.; Maulitz, A. H.; Bruice, T. C. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 8417–8420. (24) Rucker, V. C.; Byers, L. D. J. Am. Chem. Soc. 2000, 122, 8365– 8369. (25) Devi-Kesavan, L. S.; Gao, J. J. Am. Chem. Soc. 2003, 125, 1532– 1540. (26) Raveh, L.; Segall, Y.; Leader, H.; Rothschild, N.; Levanon, D.; Henis, Y.; Ashani, Y. Biochem. Pharmacol. 1992, 44, 397–400. (27) Shim, H.; Hong, S.-B.; Raushel, F. M. J. Biol. Chem. 1998, 273, 17445–17450. (28) Lewis, V. E.; Donarski, W. J.; Wild, J. R.; Raushel, F. M. Biochemistry 1988, 27, 1591–1597. (29) Dumas, D. P.; Raushel, F. M. J. Biol. Chem. 1990, 265, 21498– 21503. (30) Jackson, C. J.; Foo, J.-L.; Kim, H.-K.; Carr, P. D.; Liu, J.-W.; Salem, G.; Ollis, D. L. J. Mol. Biol. 2008, 375, 1189–1196. (31) Hong, S.-B.; Raushel, F. M. Biochemistry 1996, 35, 10904–10912. (32) Caldwell, R.; Raushel, F. M.; Weiss, P. M.; Cleland, W. W. Biochemistry 1991, 30, 7444–7450.

J. Phys. Chem. B, Vol. 112, No. 32, 2008 9991 (33) Knowles, J. R. Annu. ReV. Biochem. 1980, 49, 877–919. (34) Cox, J. R.; Ramsay, O. B. Chem. ReV. 1964, 64, 317–352. (35) Omburo, G. A.; Kuo, J. M.; Mullins, L. S.; Raushel, F. M. J. Biol. Chem. 1992, 267, 13278–13283. (36) Singh, A. K.; White, T.; Spassova, D.; Jiang, Y. Comp. Biochem. Physiol. 1998, 119C, 107–117. (37) Gonzales, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154–2161. (38) Cance´s, E.; Mennucci, B.; Tomasi, J. J. Phys. Chem. 1996, 100, 16502–16513. (39) 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, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; 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.; 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.02; Gaussian, Inc.: Wallingford, CT, 2004. (40) Faust, S. D.; Gomaa, H. M. EnViron. Lett. 1972, 3, 171–201. (41) Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.; Frisch, M. J. J. Phys. Chem. 1996, 100, 16098–16104. (42) Kaczmarek, A.; Gorb, L.; Sadlej, A. J.; Leszczynski, J. Struct. Chem. 2004, 15, 517–525.

JP800386F