Density Functional Study on the Cytochrome-Mediated S‑Oxidation:Identification of Crucial Reactive Intermediate on the Metabolic Pathof ThiazolidinedionesNikhil Taxak,† Vaibhav A. Dixit,† and Prasad V. Bharatam*

Department of Medicinal Chemistry, National Institute of Pharmaceutical Education and Research (NIPER), S. A. S. Nagar (Mohali),160 062 Punjab, India

*S Supporting Information

ABSTRACT: S-Oxidation is an important cytochrome P450(CYP450)-catalyzed reaction, and the structural and energeticdetails of this process can only be studied by using quantumchemical methods. Thiazolidinedione (TZD) ring metabolisminvolving initial S-oxidation leads to the generation of reactivemetabolites (RMs) and subsequent toxicity forcing thewithdrawal of the glitazone class of drugs, thus, the study ofthe biochemical pathway of TZD ring metabolism is a subjectof interest. The S-oxidation of the TZD ring and the formationof the isocyanate intermediate (ISC) was implicated as apossible pathway; however, there are several questions stillunanswered in this biochemical pathway. The current studyfocuses on the CYP450-mediated S-oxidation, fate of thesulfoxide product (TZDSO), ring cleavage to ISC, and formation of nucleophilic adducts. The process of S-oxidation wasexplored by using Cpd I (iron(IV)-oxo porphyrin, to mimic CYP450) at TZVP/6-311+G(d) basis set. The barriers werecalculated after incorporating dispersion and solvent corrections. The metabolic conversion from TZDSO to ISC (studied atB3LYP/6-311++G(2df,3pd)//B3LYP/6-31+G(d)) required a novel protonated intermediate, TZDSOH+. The effect of higherbasis sets (6-311+G(d,p), aug-cc-pvqz) on this conversion was studied. TZDSOH+ was observed to be more reactive andthermodynamically accessible than ISC, indicating that TZDSOH+ is the actual reactive intermediate leading to toxicity of theTZD class of compounds.

■ INTRODUCTIONS-Oxidation is an important metabolic reaction for sulfur-containing drugs catalyzed by cytochrome P450 (CYP450).1

This reaction leads to a change in the metabolic state of thesubstrates, via the generation of S-oxidized metabolites. Thesemetabolites may be therapeutically active for some drugs,whereas, they may be reactive to cause toxicity in other cases.For example, sulindac sulfoxide is an anti-inflammatory drugthat undergoes S-oxidation to form sulindac sulfones, whichexhibit cancer chemopreventive activity.2 Similarly, benzimida-zole sulfides, namely, albendazole, fenbendazole, and triclaben-dazole, are oxidized to active sulfoxides, having antiparasiticactivity.3 Flosequinan and its sulfone metabolite possessvasodilatation, inotropic, and hemodynamic properties, andare useful in acute heart failure.4 On the other hand, the S-oxidation of thiazolidinedione (TZD) ring in the glitazone classof antidiabetic drugs leads to TZD ring-opening followed byseveral toxicological consequences.5−10

Quantum chemical studies have been reported by severalscientific groups to provide the molecular level details of the S-oxidation reaction using various oxidizing agents.11−21 Theatomic level details of the S-oxidation of dihydrogen sulfide(H2S) and dimethyl sulfide (DMS) were first explored by Bach

et al. using hydrogen peroxide (HOOH) and methylhydroperoxide as the model oxidants.11 Bach and co-workerslater reported the S-oxidation of DMS by a series of bicyclic andtricyclic model C4a-flavin hydroperoxides.12 A SN

2-like attack ofthe nucleophile on the distal oxygen of the hydroperoxide wasobserved in all the cases. They also reported the oxidation ofhydrocarbons, sulfides, and selenides using peroxynitrous acid(HOONO) as the oxidant and compared the results with otheroxidants, peroxyformic acid and dimethyldioxirane.13,14 Asimilar SN2-like mechanism was observed without anyinvolvement of metastable forms of peroxynitrous acid. Musaevet al. reported the S-oxidation reaction of DMS withperoxynitrite anion (ONOO−) and HOONO.15 The sulfox-idation study was extended to the actual biological system byKumar,16a Shaik,16b,18 Rydberg,17 and co-workers with the useof Cpd I (high-valent iron(IV)-oxo porphyrin complex) as themodel oxidant. Shaik and co-workers reported the sulfoxidationof DMS by Cpd I (using density functional theory), whichinvolved a direct oxygen transfer to the sulfur atom.18

Received: August 13, 2012Revised: September 29, 2012Published: October 1, 2012

Article

pubs.acs.org/JPCA

© 2012 American Chemical Society 10441 dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−10450

Moreover, de Visser and co-workers reported the quantummechanics/molecular mechanics (QM/MM) approach todescribe the S-oxidation mechanism in DMS by CYP450.19

They studied S-oxidation of DMS using Compound 0(iron(III)-hydroperoxo complex) and Cpd I and reportedthat Cpd 0 is a sluggish oxidant and Cpd I is the active oxidantinvolved in this process.19 Later, Shaik et al. reported the samemechanism on the sulfoxidation of para-substituted thioani-soles.20 Rydberg and co-workers discussed the processes ofsulfoxide, sulfur and nitrogen oxidation, and nitrogen deal-kylation by CYP450, utilizing theoretical methods.17 Taxak etal. have reported density functional theory based analysis of theprocess of S-oxidation of model substrate, 5-methyl-2,4-thiazolidinedione (MeTZD), by the use of various modeloxidants, such as HOOH, HOONO, and C4a-hydroperoxy-flavin.21 The mechanism involved the nucleophilic attack of Son the distal oxygen leading to the formation of sulfoxidemetabolite.21 Bharatam et al. reported the significance of sulfuroxidation in the racemization of the glitazone class of moleculesusing quantum chemical methods.22

The glitazone class of drugs include Rosiglitazone (RGZ),Pioglitazone (PGZ), and already withdrawn Troglitazone(TGZ). These show antidiabetic activity via peroxisomeproliferator activated receptors (PPARγ) activation and insulinsensitizing effect.23−25 Several reports have implicated thetoxicological consequences of these drugs, owing to theinvolvement of reactive metabolites (RMs) or reactiveintermediates originating from TZD ring metabolism.5−10

TGZ was withdrawn from the market owing to idiosyncratichepatotoxicity and associated mortality.26 In some countries,RGZ has been withdrawn owing to cardiovascular complica-tions,27 and PGZ has been withdrawn owing to the risk ofoccurrence of bladder cancer.28 Since all these drugs containingthe TZD ring have been withdrawn from active use in severalcountries, the compounds with the TZD ring are under scannerfor their toxicity and increased risk to benefit ratios. For thesame reason, FDA regulations have enforced a two-yearcarcinogenicity evaluation in rats and mice for all PPARagonists before conducting their clinical studies.29 Biotransfor-mation occurring on the TZD ring in this class of drugs hasbeen implicated for the toxicity and the subsequent withdrawalfrom therapeutic use. Since a general alert is in place againstTZD derivatives due to the toxicological implications arisingfrom TZD ring oxidation, it becomes necessary to establish themolecular level details of the biochemical reactions occurring

on the TZD ring, emphasizing the cytochrome P450(CYP450)-mediated oxidation reaction.Several experimental studies have been carried out to

establish the CYP450-mediated oxidation reaction takingplace on the TZD ring. The biotransformation of the TZDring involves the oxidation on the S atom of TZD, followed byring scission resulting in reactive intermediates (Figure1).5−10,30−32

Kassahun et al. suggested that the reactive metabolites ofTGZ are responsible for its hepatotoxicity; they proposed theformation of highly electrophilic isocyanate (ISC) intermediatevia CYP3A4-mediated oxidation of the S-atom of the TZD ringfollowed by ring cleavage.6 Prabhu et al. and He et al.highlighted the identification of glutathione conjugatesoriginating from the TZD ring.7,8 The S-oxidation-based ring-opened metabolites were also reported by Liu and co-workers,where they emphasized the formation of these metabolites viaoxidative cleavage of the TZD ring by S-oxidation.9 Reddy et al.studied the mechanism of metabolic scission of the TZD ring,highlighting the importance of the initial S-oxidation step.10

Baughman et al. reported the CYP3A4-mediated metabolicactivation of PGZ in human and rat liver microsomes andhepatocytes, and stressed various ring-opened reactive inter-mediates and conjugation with glutathione.30 The metabolitesgenerated by TZD ring-opening were recently reported byUchiyama et al. in PGZ and RGZ.31 Alvarez-Sanchez et al.compared the bioactivation potentials of TGZ, PGZ, and RGZfocusing on the common metabolic activation of the TZD ringleading to the formation of GSH adducts of ring-openedreactive intermediates.32

The molecular level details and energetics of the TZD ringmetabolic pathway have not been reported or interpreted byexperimental studies; only the indirect evidence of metabolicproducts are available. The details of the reaction path,including (i) structural details of the metabolites, intermediates,and transition states, (ii) energy requirements associated withthe reactions, (iii) reactivity of the metabolites andintermediates, etc., cannot be provided by experimentalmethods. In such cases, quantum chemical methods offersolutions providing necessary details, often correcting theexisting knowledge and providing new insights. Thus, in thisstudy, quantum chemical methods were employed to under-stand the whole biochemical pathway of TZD ring metabolismfor the glitazone series of drugs. The computational methodsprovide a clear understanding of several steps of the catalyticcycle of CYP450 and give vital insights into the mechanistic

Figure 1. CYP450-mediated metabolism of TZD consists of initial S-oxidation, formation of ISC intermediate, and its reaction with glutathione(GSH) and other nucleophiles leading to adducts, which cause cytotoxicity and cell death.

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010442

details behind the same. The TZD sulfoxide (TZDSO)formation, ring cleavage of TZDSO, generation and reactivityof various RMs or intermediates, and the mechanistic details forthe formation of ISC from TZDSO have been explored.The results obtained in this work are presented in three

sections: (i) CYP450-mediated S-oxidation of the TZD ring,(ii) identification of the novel intermediate, TZDSOH+ on thebiochemical pathway of TZD ring metabolism, and (iii)electrophilicity and reactivity of the TZDSOH+ and ISCintermediates. This study helped in identifying TZDSOH+ asthe most reactive novel species on this pathway; we proposethat the observed biochemical consequences of toxicityoriginate from TZDSOH+ rather than the previously suggestedISC intermediate.

■ COMPUTATIONAL DETAILSQuantum chemical calculations for all systems underconsideration have been carried out with the Gaussian03suite of programs.33 Geometry optimizations were performedby using DFT theory at the B3LYP/6-31+G(d) level.34a−c

Analytical frequencies have been estimated by carrying outfrequency calculations at the B3LYP/6-31+G(d) level tocharacterize the optimized structures as minima or transitionstates (one negative frequency) on the potential energy surface.Single point energies for the metabolic pathway after S-oxidation were determined at B3LYP/6-311++G(2df,3pd),using the B3LYP/6-31+G(d) optimized geometries. The zeropoint vibrational energy (ZPVE) values have been scaled by afactor of 0.9806.35 Partial atomic charges for geometries wereestimated by performing natural bond orbital (NBO) analysis

embedded in the G03 package.36 For the calculations involvingthe use of model Cpd I,16−18,37 the LanL2DZ basis set wasused on the Fe atom34d and the 6-31+G(d) basis set wasemployed on all the remaining atoms (denoted as BS1). Cpd Iconsisted of iron(IV)-oxo heme-porphine with SH− as the axialligand to model the human P450 and it is recognized as astandard model to study CYP-mediated oxidation reac-tions.16−18,37 Single point calculations were carried out usingthe TZVP triple-ζ basis set for iron38 and the 6-311+G(d) basisset34d,e for all remaining atoms to determine the final energies(denoted as BS2). The Integral Equation Formalism variant ofPolarizable Continuum Model (IEFPCM)39 was utilized tocarry out single point solvent calculations using a dielectricconstant (ε) of 5.7 corresponding to chlorobenzene thatmimics a nonpolar environment (denoted as BS3). Single pointenergy calculations were also carried out using the Grimme’s“D2” empirical dispersion corrections (denoted as BS4) in theGaussian09 suite of programs.17a,40 The multistate reactivity ofhigh-valent iron(IV)-oxo (Cpd I) was considered (bothdoublet and quartet spin state). Spin densities for Cpd Iwere determined by using Mulliken population analysis. Theenergy values and the geometric parameters discussed in thisarticle have been computed by using a large basis set includingdispersion corrections, solvent effects, and zero point vibra-tional corrections (BS4) whenever Cpd I is involved. Under allother conditions, the B3LYP/6-311++G(2df,3pd)//B3LYP/6-31+G(d) level data are inferred, unless otherwise specificallymentioned.

Figure 2. 3D structure of transition states of S-oxidation of MeTZD on the doublet (bold) and quartet spin surface (bold and italics withinparentheses) by Cpd I. The distance data are given in Å. Bond angles are in degrees. RC: reactant complex of MeTZD and Cpd I; TS: transitionstate for S-oxidation of MeTZD; and PC: product complex of TZDSO and the porphyrin ring.

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010443

■ RESULTS AND DISCUSSIONCYP450-Mediated S-Oxidation of the TZD Ring. The

detailed mechanistic study for the process of S-oxidation on themodel compound MeTZD by the use of model oxidant Cpd Iis discussed in this section. Figure 2 shows the 3D structures onthe reaction path, involving the reactant complex (RC),transition state (TS), and product complex (PC) geometrieson both the doublet and quartet spin states of Cpd I, for the S-oxidation of MeTZD.The energy profile indicates that the reactant complex of the

quartet spin state is marginally more stable than thecorresponding reactant complex on the doublet spin state by1.22 kcal/mol (Figure 3). The S−O distances in the transition

state geometries for the doublet and quartet spin surfaces are2.16 Å and 1.97 Å, respectively, indicating that the transitionstate is reached early on the doublet spin surface. The Fe−Odistance gets elongated in the transition states from 1.62 Å to1.73 Å and 1.81 Å for the doublet and quartet spin states,respectively (Figure 2). The energy barrier for the S-oxidationhas been estimated to be 11.59 kcal/mol on a path involvingthe doublet spin state of Cpd I, after solvent corrections (BS3).The inclusion of dispersion corrections (BS4) significantlylowered the barrier to 7.26 kcal/mol (Figure 3). This value ismuch lower than that of a path involving the quartet spin stateof Cpd I (14.30 kcal/mol). These values were obtained afterincorporating dispersion corrections and implicit solventconditions (BS4). It was observed that more accurate barrierswere observed with higher basis sets (BS2), solvent corrections(BS3), and dispersion corrections (BS4) as compared to thegas phase (BS1). The energy estimated for S-oxidation in DMSusing Cpd I on doublet and quartet spin surfaces have beenreported to be 7.1 and 9.1 kcal/mol.18 Hence, the energybarrier is observed to be higher for the process of S-oxidation inMeTZD (11.59 kcal/mol, BS3) as compared to DMS (7.1kcal/mol). Similarly, the S-oxidation barrier in thioanisoles hasbeen reported to be 7.4 kcal/mol.20 The higher barrier in S-oxidation of MeTZD can be attributed to the lowernucleophilicity of the S atom due to the electron delocalizationin the lone pair of the sulfur atom in MeTZD. The process of

S-oxidation of MeTZD by Cpd I involved the interactionbetween p-lone pair on the sulfur atom and the π*/σ*z

2 orbitalof iron-oxo (FeO) of Cpd I similar to the observationreported by Shaik et al. on thioanisoles.20 The Fe−O−S anglesfor the doublet and quartet spin states were found to be 125.5°and 140.1°, respectively, in the transition state geometries(Figure 2). S-Oxidation through a quartet spin state involvesthe interaction of pS → σ*z

2FeO orbitals, whereas on the

doublet spin surface, a pS → π*FeO orbital interaction isfavored. The spin densities and NBO charge analyses alsosupport the favorable S-oxidation of MeTZD on the doubletspin state over the quartet spin state; the spin density on the Satom ofMeTZD is near zero (−0.003) for doublet TS, whereasthe same in quartet spin state is −0.197 (see the SupportingInformation, Table S2). The NBO charge on the S atom ofMeTZD was found to be quite comparable for both the TSgeometries (doublet: 0.627; quartet: 0.736) indicating asignificant “Electron Transfer” character in both species;however, the mechanism observed involves a direct oxygentransfer (DOT) from Cpd I to MeTZD (see the SupportingInformation, Table S3). The results are in accordance with thethioether sulfoxidation reported by Shaik et al.20 In addition, aC−H---O hydrogen-bonding interaction between the hydrogenof the methyl group at carbon C4 was observed in thetransition states, for both doublet and quartet spin states. Theproduct complexes were found to be stable by 3.97 and 6.77kcal/mol for the sulfoxide formation in doublet and quartetspin states, respectively (Figure 3).In the product complexes, the Fe−O bond is almost broken

(2.42 and 2.76 Å for doublet and quartet spin states,respectively) and the S−O bond (1.52 Å) is completelyformed; however, there is a sufficient electrostatic interactionbetween Fe and O. Overall, the above computational analysison the S-oxidation of MeTZD suggests that a direct oxygentransfer mechanism from the doublet spin state of Cpd I is thefavorable path. These results are in accordance with theconclusions reported in the literature on the S-oxidation ofDMS.16−18,20

The energy barriers for the process of S-oxidation ofMeTZDby various organic and inorganic catalysts have already beenreported, HOOH (33.84 kcal/mol), bicyclic C4a-flavin hydro-peroxide (26.79 kcal/mol), and tricyclic C4a-flavin hydro-peroxide (24.55 kcal/mol).31 Thus, it can be concluded thatCpd I is the most efficient oxidant catalyzing the process of S-oxidation as compared to other catalysts.

Identification of the Novel Intermediate, TZDSOH+ onthe Biochemical Pathway of TZD Ring Metabolism. S-Oxidation of the TZD ring and the formation of nucleophilicadducts from the corresponding ISC intermediate have beenreported in the metabolism studies on the glitazone class ofcompounds by many research groups.5−10,30−32,41−46 However,a mechanistic pathway with the molecular level details for therearrangement of TZDSO to ISC have not been envisionedearlier. TZDSO and ISC are regioisomers with an equalnumber of atoms but with very different three-dimensionalstructures. With respect to connectivity, the position of onlyone proton is different between the two structures; hence, it isevident that some type of proton transfer must be leading tothe formation of ISC from TZDSO. A literature search revealedthat no study addressing this fact has been undertaken so far. Aclear understanding of this isomerization is essential withrespect to the fate of the metabolites and resulting toxicityprofile. This could lead to the exploration of some specific

Figure 3. Potential energy surface for the path of S-oxidation ofMeTZD by Cpd I on the doublet (◆, blue) and the quartet (■, red)spin surface. RC: reactant complex of MeTZD and Cpd I; TS:transition state for S-oxidation of MeTZD; and PC: product complexof TZDSO and the porphyrin ring. The energies are in kcal/molrelative to the reactant complexes on both spin states of Cpd I. Valuesin bold are at the basis set BS4 after dispersion and solvent corrections.Values in normal font without parentheses are at basis set BS3 aftersolvent and frequency corrections. Values in parentheses are at basisset BS2 without solvent corrections. Values in square bracket are atbasis set BS1 (gas phase).

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010444

details that might exist on the pathway of TZD ringmetabolism.A comparison of structures of TZD and TZDSO showed

that the S−C2 bond is elongated in TZDSO (from 1.79 Å to1.91 Å). This is the same bond that is completely brokenleading to the formation of ISC. This indicates that S−C2 bondweakening occurs upon the oxidation of the TZD ring. Thisweakening can be attributed to the simultaneous impact ofn(CO) → σ*(S−C2) and n(SO) → σ*(S−C2) negative hyper-conjugative interactions. The energy change associated withsuch second order interactions, E(2), can be estimated by usingNBO analysis.36 In TZD, the n(CO) → σ*(S−C2) interaction isvery weak (E(2): 2.62 kcal/mol), while in TZDSO, the E(2)

values associated with n(CO) → σ*(S−C2) and n(SO) →σ*(S−C2) second order interactions are 39.21 and 18.24 kcal/mol, respectively, indicating a strong cumulative influence ofnegative hyperconjugative interactions. This explains theweaker S−C2 bond in the TZDSO in comparison to TZD.The S−C2 bond breaking is further facilitated by the loss ofelectron delocalization in the S−C2−N3−C4 framework ofTZD due to the oxidation at the S atom.The species TZDSO and ISC are almost isoenergetic,

showing that the ring-opening is not a thermodynamicallydriven process. This was confirmed by analyzing their energiesat different basis sets and in the presence of implicit solventconditions (see the Supporting Information, Table S4). Protontransfer from nitrogen (N3) to oxygen (O1) in TZDSO occursvia hydronium ion catalysis as shown in Figure 4. Any directand uncatalyzed rearrangement of TZDSO to ISC is less likelyconsidering the fact that the distance between the N3 hydrogenand the O1 oxygen is more than 4.20 Å and both groups face inopposite directions (Figure 4) and this process involves 1,4-Hshift, which is not thermodynamically favorable.It is possible that the proton transfer from the NH center

to the SO center can take place through a network of watermolecules. We studied this mechanism employing one, two,and three water molecules. With two water molecules, theenergy barrier required was 11.87 kcal/mol. Increasing thenumber of water molecules to three decreased the barrier to ∼8kcal/mol. Further increase in water molecules did not causemuch change. However, the structural features of the TZDSO+ 3H2O molecules were not clear. Considering the fact that inbiochemical conditions, many molecules of water and floatingprotons are available, the formation of a single chain of three or

more water molecules cannot be clearly envisaged. Hence, thisoption was not considered further for this study.Thus, a most likely mechanism for the conversion of TZDSO

to ISC was explored, based on H3O+ (an acceptable model of

proton in aqueous medium) catalysis.47a The transition state forthe protonation of TZDSO by H3O

+ was envisaged; however,every attempt gave the geometry closer to the product, andthus, it can be suggested that TZDSO gets directly protonatedby H3O

+ (see the Supporting Information, Figure S1). Theoxygen atom in TZDSO has a charge of −0.922 (NBO) andthe proton of H3O

+ has a +0.628 charge. This indicates that thehighly electronegative oxygen atom of TZDSO has a propensityto immediately accept the proton present in the aqueous mediain the biological conditions. Similar observation for the directprotonation by H3O

+ has been reported for acetonitrile, leadingto the formation of protonated acetonitrile.47b It was envisagedthat a proton transfer from H3O

+ to O1 oxygen in TZDSOgenerated a protonated intermediate, along with a watermolecule. This protonated intermediate, TZDSOH+ is thenovel intermediate identified on the metabolic pathway of theTZD class of drugs. This intermediate then leads to thegeneration of the ISC intermediate, via a proton abstractionfrom the NH group by the water molecule to give back thecatalyst (H3O

+). The energy released due to the proton transferfrom hydronium ion to TZDSO was estimated to be 37.65kcal/mol (a highly exothermic process). On the other hand, theproton transfer from TZDSOH+ to water (to form ISC) wasestimated to be equally endothermic by 37.72 kcal/mol.Therefore, such a bimolecular process is not associated withany energy gain on thermodynamic grounds.Figure 4 shows that the S1−C2 bond is nearly broken (1.97

Å) in TZDSOH+, which thereafter completely breaks in theISC (3.12 Å) intermediate. Since both intermediates are closelyrelated to each other, it becomes necessary to characterize theirelectrophilicity and reactivity properties. This would probablyunravel the dilemma concerning the actual intermediate, onwhich the nucleophilic addition reactions occur, leading to theadduct formation. The various electrophilic parameters andreactivity for TZDSOH+ and ISC intermediates are comparedand results are discussed in the next section. Thus, a detailedstudy of the reactivity of both these intermediates would openthe black box that existed between TZDSO and identificationof RMs, formed from the nucleophilic addition to TZDSOH+

and ISC intermediate.

Figure 4. CYP450-mediated S-oxidation of TZD ring and H3O+ catalyzed TZDSO isomerization to TZDSOH+ (novel intermediate) and ISC

intermediates on the pathway. The increase in S−C2 bond length after S-oxidation is observed (shown in 3D structures along the reaction path) insubsequent steps indicating weakening of the S−C2 bond, and ring-opening.

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010445

Electrophilicity and Reactivity of TZDSO, TZDSOH+,and ISC. The relationship between electrophilicity/nucleophil-icity and reactivity has been studied by various groups.45−49

Applications of this concept in toxicology have also beenrealized over the years.48−51 Global and local electrophilicityindices can be calculated by using standard equationsmentioned in these references (see the Supporting Information,S1). Dixit et al. have utilized these parameters to understandthe difference in the electrophilicity and toxicity of o-quinonemethide reactive metabolites generated from TGZ.52 Bothglobal (ω) and local electrophilicity (ωc

+) indices point to thefact that TZDSO is a highly electrophilic species (Tables 1 and2). Especially, the C-2 center of the TZDSO ring is highly

electron deficient, because two electron-withdrawing atoms,namely, carbonyl oxygen and oxidized sulfur, are attached to it.The protonation of TZDSO to TZDSOH+ leads to an

increase in the electrophilicity, as shown by a high local C2electrophilicity value of 1.064 (Table 1) and a very high globalelectrophilicity index of 9.951 (Table 2). The ωc

+ (0.177) andω (2.091) values of the ISC intermediate are relatively less ascompared to those in TZDSOH+. These data prompted us toconsider TZDSOH+ as the most reactive and electrophilicintermediate in relation to TZDSO as well as the ISCintermediate for nucleophilic addition reactions leading to theformation of adducts in vivo. Thus, there is a marked differencein the electrophilic character of C2 in TZDSOH+ and ISCintermediates, and it becomes necessary to understand theirdifferential reactivity toward various nucleophiles, as discussedin the next section.Reactivity of TZDSOH+ and ISC toward Different

Nucleophiles. Isocyanates are well-known reactive specieswith a variety of applications in synthetic chemistry and with awell-understood carcinogenic potential.53−57 Theoretical stud-ies on water-catalyzed hydrolysis of hydrogen isocyanatesuggested a concerted nature of the process.58 Similar studiesby Raspoet et al. suggested a multimolecular mechanism formethanolysis of hydrogen isocyanate.58 According to thesestudies, addition reactions of water and methanol to isocyanates

were found to be favorable.58,59 Both these studies focused onthe oxygen-based nucleophiles trying to explain the observedrates of hydrolysis. Considering the fact that sulfur- andnitrogen-based nucleophiles are equally important for the fateof any electrophile generated in vivo, we have explored thereaction mechanism for the formation of sulfur-, oxygen-, andnitrogen-based adducts of TZDSOH+ and ISC derived after theS-oxidation of the TZD ring.The reaction mechanism was determined for the addition of

nucleophiles water (HOH), methanethiol (MeSH), methanol(MeOH), and methylamine (MeNH2) to the carbonyl group ofTZDSOH+ and ISC. In TZDSOH+, the mechanism involves atwo-step process initiated by the nucleophilic attack at the C2center followed by proton transfer from the nucleophile to thecarbonyl oxygen (Figure 5). On the other hand, in ISC, themechanism involves a simultaneous processa nucleophilicattack at the C2 center accompanied by the transfer of X−Hhydrogen to the nitrogen atom of ISC (Figure 5). The reactionmechanism was found to be highly dependent on the identityof the nucleophile. The activation barriers observed for thenucleophilic addition to the carbonyl carbon in TZDSOH+

were found to be lower as compared to those for the additionto the ISC intermediate, confirming the highly electrophilicnature of the protonated species.Figure 6 shows the structures of transition state geometries

for the nucleophilic addition of water, methanol, methanethiol,and methylamine to TZDSOH+ along with the activationbarriers (in kcal/mol). The geometries of transition states fornucleophilic addition of water, methanol, and methanethiol tothe ISC intermediate are shown in Figure 7. A comparisonbetween four nucleophiles reveals that methylamine is thestrongest nucleophile with activation barriers of 13.48 and22.18 kcal/mol toward TZDSOH+ and ISC intermediates,respectively. It can also be observed that the activation barriersfor the nucleophilic addition to TZDSOH+ are 8−13 kcal/mollower than that with the ISC intermediate. This shows a higherreactivity for TZDSOH+ as compared to the ISC intermediate,thus indicating its crucial role in leading to toxicity in the TZDclass of compounds.Figure 8 compares the potential energy surfaces for the

nucleophilic addition reactions to TZDSOH+ and ISC. Theproducts obtained are also thermodynamically more stable forthe amino adducts of TZDSOH+, in comparison to that of ISC.This comparative study establishes that the nucleophilicreactions involving TZDSOH+ are spontaneous, whereasreaction processes involving ISC require higher energy barriers.Thus, the formation of the ISC intermediate on thebiochemical pathway involving S-oxidation of TZD is lesslikely, in contrast to the earlier proposed metabolicpaths.5−10,30−32

The whole biochemical pathway elucidated on the oxidationpath starting from TZD to the formation of TZDSOH+

nucleophilic adducts can be summed up as shown in Figure 9.

Table 1. Condensed Atomic NBO Charges (q) for N + 1 and N Electron Systems, Local Fukui Functions ( fc+) and Local

Electrophilicities (ωc+) Calculated at the B3LYP/6-311++G(2df,3pd)//B3LYP/6-31+G(d) Level

molecule qN+1(C) qN(C) qN−1(S) qN(S) fc+(C) fc

−(S) ωc+ NS

−

TZD 0.515 0.548 0.782 0.265 0.033 −0.517 0.056 −0.306TZDSO 0.371 0.532 1.291 1.178 0.161 −0.113 0.341 −0.053TZDSOH+ 0.511 0.618 0.107 1.064ISC 0.798 0.883 0.085 0.177DMS 0.896 0.174 −0.722 −0.902

Table 2. Energy of the Highest Occupied Molecular Orbital(EHOMO, au), Lowest Unoccupied Molecular Orbital (ELUMO,au), Electronegativity (χ, eV), Hardness (η, eV),Electrophilicity Index (ω, eV), and Nucelophilicty Index (N,eV), Using the B3LYP/6-311++G(2df,3pd)//B3LYP/6-31+G(d) Method in the Gas Phase

molecule EHOMO ELUMO χ η ω N

TZD −0.281 −0.054 4.559 6.177 1.685 0.593TZDSO −0.268 −0.077 4.697 5.203 2.120 0.471TZDSOH+ −0.473 −0.279 10.23 5.259 9.951 0.100ISC −0.264 −0.076 4.626 5.116 2.091 0.479DMS −0.223 −0.004 3.088 5.959 0.800 1.250

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010446

The above-mentioned analysis establishes that the proto-nated S-oxide (TZDSOH+) is the crucial reactive intermediateleading to ISC formation. TZDSOH+ formed reacts quicklywith the nucleophiles in vitro and in vivo (as compared toISC), owing to its high electrophilicity. Hence, it is possiblethat the observed mass spectral products (GSH adducts),

during the experimental metabolism studies by variousgroups,5−10,30−32 originate from TZDSOH+ without under-going biotransformation to the ISC intermediate. Owing to thesimilarity in mass for the GSH adducts for both intermediates,the above observation has remained elusive until this study.Thus, the involvement of the ISC intermediate on the

Figure 5. Schematic diagram for the nucleophilic attack of water, methanol, methanethiol, and methylamine on TZDSOH+ and ISC intermediates.

Figure 6. Transition state structures for the nucleophilic addition of water (HOH), methanol (MeOH), methanethiol (MeSH), and methylamine(MeNH2) to TZDSOH+ reactive intermediate. Values in parentheses indicate the activation barriers for nucleophilic addition (in kcal/mol). Colorcode: red, oxygen; blue, nitrogen; off-white, hydrogen; yellow, sulfur; gray, carbon.

Figure 7. Transition state structures for the nucleophilic addition of water (HOH), methanol (MeOH), methanethiol (MeSH), and methylamine(MeNH2) to ISC reactive intermediate. Values in parentheses indicate the activation barriers for nucleophilic addition (in kcal/mol). Color code:red, oxygen; blue, nitrogen; off-white, hydrogen; yellow, sulfur; gray, carbon.

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010447

metabolic path of oxidation of TZD seems to be less important.This argument is supported by the fact that deprotonation fromthe product complexes of nucleophile adducts of TZDSOH+ toISC requires an energy of 30−40 kcal/mol following anendothermic process. Owing to the higher reactivity ofTZDSOH+ (low activation barriers) and more stability of thenucleophilic adducts of TZDSOH+ as compared to ISCadducts, TZDSOH+ becomes the most important and criticalreactive intermediate leading to cytotoxicity. The nucleophilicadducts (GSH adducts) of TZDSOH+ can be identifiedexperimentally by mass spectral studies such as collision-induced dissociation (CID) followed by ESI(+)-MS/MScharacterization.60−62 Therefore, this study identified a criticalnovel reactive intermediate on the TZD ring-opening metabolicreaction. This study also points out that the reactiveintermediate TZDSOH+ reacts easily with amines in relationto other nucleophiles. Thus, the toxicity originates due to thereaction of TZDSOH+ with the biological amines in thebiological conditions, even in the presence of glutathione. Thisimplies that attention should be paid to the reactions between

TZDSOH+ and biological amines to understand thecomplications arising from the administration of the glitazoneclass of drugs.

■ CONCLUSIONS

Quantum chemical analysis was carried out to elucidate thebiochemical pathway of TZD ring metabolism and to obtainclues regarding the actual mechanism and the structures ofreactive intermediates behind the toxicity associated with theTZD class of drugs. Density functional theory (DFT)calculations have been performed on the CYP450-mediatedS-oxidation of TZD by using Cpd I as the model oxidant tomimic CYP450. The biochemical mechanism of S-oxidationinvolves a direct oxygen transfer from the oxidant, Cpd I withan energy barrier of 7.26 kcal/mol on the doublet spin surfaceof Cpd I (including dispersion effect) and implicit solvent (ε =5.7) conditions. The S-oxidation process was favorable on thedoublet spin state owing to a favorable pS → π*FeOinteraction, as compared to higher energy pS → σ*z

2FeO

interaction on the quartet spin surface.The fate of the sulfoxide product (TZDSO) and its

biotransformation to ISC intermediate via H3O+ catalysis was

explored. A novel intermediate, protonated TZDSO(TZDSOH+), connecting TZDSO and ISC intermediates wasidentified during this conversion. The thermodynamicallyneutral proton exchange process between TZDSO and H3O

+

favors the generation of TZDSOH+ from TZDSO. To gainfurther insights into the characteristics of these intermediates, acomparative analysis of their electrophilicity and reactivityproperties was carried out. The charge and electrophilicityanalyses indicated that TZDSOH+ is highly electrophilic ascompared to the ISC intermediate. The marked difference inthe electrophilic character of these intermediates was confirmedby their differential reactivity toward various nucleophiles (S, N,and O based). The activation barriers observed for thenucleophilic addition to TZDSOH+ intermediate were foundto be lower as compared to the addition to the ISCintermediate, thus indicating a higher reactivity for theTZDSOH+ intermediate. The reactivity toward differentnucleophiles was found to be highly dependent on the nature

Figure 8. Potential energy surface for the nucleophilic addition of water (◆, blue), methanol (■, red), methanethiol (▲, green), and methylamine(×) to TZDSOH+ and ISC intermediates. The black line shows the path of conversion of TZDSO to TZDSOH+ and ISC via TZDSOH+. All thecalculated values are in kcal/mol. TZDSO: TZD sulfoxide; TZDSOH+: protonated TZDSO; ISC: isocyanate; TS″: transition states for nucleophilicaddition to ISC; ISCa: ISC adducts; RC: reactant complex of nucleophiles and TZDSOH+; and TS′: transition states for nucleophilic addition toTZDSOH+.

Figure 9. The reaction pathway for the formation of TZDSOH+

adducts, starting from the S-oxidation of TZD.

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010448

of the attacking nucleophile; amino group containingnucleophiles are shown to be highly reactive (lowest energybarriers) toward the TZDSOH+ intermediate. This studyexplains that nitrogen- (more preferably), sulfur-, or oxygen-based nucleophilic centers in the protein and peptides reactwith TZDSOH+ preferentially than ISC to ultimately formhighly stable nucleophilic adducts of TZDSOH+.Thus, the whole biochemical pathway of TZD ring

metabolism starting from the initial CYP450-mediated S-oxidation of the TZD ring to the eventual formation ofnucleophilic adducts of RMs was elucidated. TZDSOH+ wasidentified as the crucial and the novel reactive intermediateinvolved in the toxicological implications of the TZD class ofdrugs. All the mass spectral data can be explained withreference to TZDSOH+, without invoking the ISC inter-mediate. Hence, TZDSOH+ is the important and crucialreactive intermediate existing on the TZD ring metabolicpathway.

■ ASSOCIATED CONTENT*S Supporting InformationAbsolute energies of geometries involved in the S-oxidation ofTZD by Cpd I at different basis sets (Table S1), table of spindensity analysis and NBO charge distribution (Tables S2 andS3), relative energies of TZDSOH+ and ISC at different basissets (Table S4), absolute energies for the optimized geometriesinvolved in the nucleophilic addition to TZDSOH+ and ISC(Table S5), optimized geometry for the direct protonation ofTZDSO by H3O

+ (Figure S1), formulas for calculatingelectrophilicity and nucleophilicity parameters (S1), completeauthor details for ref 43 (S2), and Cartesian coordinates for allthe optimized geometries (S3) discussed in the text. Thismaterial is available free of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Phone: +91 172 2292018. Fax: +91 172 2214692. E-mail:pvbharatam@niper.ac.in.Author Contributions†These authors contributed equally to the work.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis research was financially supported by Department ofScience and Technology (DST), New Delhi. N.T. is thankful toDST for the INSPIRE Fellowship.

■ REFERENCES(1) (a) Nassar, A. F.; Hollenberg, P. F.; Scatina, J. Drug MetabolismHandbook: Concepts and Applications; John Wiley & Sons: Hoboken,NJ, 2009. (b) Meunier, B.; de Visser, S. P.; Shaik, S. Chem. Rev. 2004,104, 3947−3980.(2) (a) Thompson, H. J.; Jiang, C.; Lu, J.; Mehta, R. G.; Piazza, G. A.;Paranka, N. S.; Pamukcu, R.; Ahnen, D. J. Cancer Res. 1997, 57, 267−271. (b) Reddy, B. S.; Kawamori, T.; Lubet, R. A.; Steele, V. E.; Kelloff,G. J.; Rao, C. V. Cancer Res. 1999, 59, 3387−3391.(3) Soria-Arteche, O.; Castillo, R.; Hernandez-Campos, A.; de laPena, M. H.; Navarrete-Vazquez, G.; Medina-Franco, J. L.; Gomez-Flores, K. J. Mex. Chem. Soc. 2005, 49, 353−358.(4) (a) Falotico, R.; Haertlein, B. J.; Lakas-Weiss, C. S.; Salata, J. J.;Tobia, A. J. J. Cardiovasc. Pharmacol. 1989, 14, 412−418. (b) Weishaar,

R. E.; Kirker, M. L.; Wallace, A. M.; Ferraris, V. A.; Britton, L. W.; Sim,M. F. Eur. J. Pharmacol. 1993, 236, 363−366.(5) Bolton, G. C.; Keogh, J. P.; East, P. D.; Hollis, F. J.; Shore, A. D.Xenobiotica 1996, 26, 627−636.(6) Kassahun, K.; Pearson, P. G.; Tang, W.; McIntosh, I.; Leung, K.;Elmore, C.; Dean, D.; Wang, R.; Doss, G.; Baillie, T. A. Chem. Res.Toxicol. 2001, 14, 62−70.(7) Prabhu, S.; Fackett, A.; Lloyd, S.; McClellan, H. A.; Terrell, C.M.; Silber, P. M.; Li, A. P. Chem.-Biol. Interact. 2002, 142, 83−97.(8) He, K.; Talaat, R. E.; Pool, W. F.; Reily, M. D.; Reed, J. E.;Bridges, A. J.; Woolf, T. F. Drug Metab. Dispos. 2004, 32, 639−646.(9) Liu, D. Q.; Karanam, B. V.; Doss, G. A.; Sidler, R. R.; Vincent, S.H.; Hop, C. E. C. A. Drug Metab. Dispos. 2004, 32, 1023−1031.(10) Reddy, V. B. G.; Karanam, B. V.; Gruber, W. L.; Wallace, M. A.;Vincent, S. H.; Franklin, R. B.; Baillie, T. A. Chem. Res. Toxicol. 2005,18, 880−888.(11) Bach, R. D.; Ming-Der, S.; Schlegel, H. B. J. Am. Chem. Soc.1994, 116, 5379−5391.(12) Canepa, C.; Bach, R. D.; Dmitrenko, O. J. Org. Chem. 2002, 67,8653−8661.(13) Bach, R. D.; Glukhovtsev, M. N.; Canepa, C. J. Am. Chem. Soc.1998, 120, 775−783.(14) Bach, R. D.; Dmitrenko, O.; Estevez, C. M. J. Am. Chem. Soc.2004, 127, 3140−3155.(15) (a) Musaev, D. G.; Geletii, Y. V.; Hill, C. L. J. Phys. Chem. A2003, 107, 5862−5873. (b) Geletii, Y. V.; Musaev, D. G.; Khavrutskii,L.; Hill, C. L. J. Phys. Chem. A 2004, 108, 289−294.(16) (a) Kumar, D.; Sastry, G. N.; de Visser, S. P. Chem.Eur. J.2011, 17, 6196−6205. (b) Shaik, S.; Kumar, D.; de Visser, S. P.; Altun,A.; Thiel, W. Chem. Rev. 2005, 105, 2279−2328.(17) (a) Rydberg, P. J. Chem. Theory Comput. 2012, 8, 2706−2712.(b) Rydberg, P.; Ryde, U.; Olsen, L. J. Chem. Theory Comput. 2008, 4,1369−1377.(18) (a) Sharma, P. K.; de Visser, S. P.; Shaik, S. J. Am. Chem. Soc.2003, 125, 8698−8699. (b) Li, C.; Zhang, L.; Zhang, C.; Hirao, H.;Wu, W.; Shaik, S. Angew. Chem., Int. Ed. 2007, 46, 8168−8170.(19) Porro, C. S.; Sutcliffe, M. J.; de Visser, S. P. J. Phys. Chem. A2009, 113, 11635−11642.(20) Shaik, S.; Wang, Y.; Chen, H.; Song, J.; Meir, R. Faraday Discuss.2010, 145, 49−70.(21) Taxak, N.; Parmar, V.; Patel, D. S.; Kotasthane, A.; Bharatam, P.V. J. Phys. Chem. A 2011, 115, 891−898.(22) Bharatam, P. V.; Khanna, S. J. Phys. Chem. A 2004, 108, 3784−3788.(23) Saltiel, A. R.; Olefsky, J. M. Diabetes 1996, 45, 1661−1669.(24) Smith, B. W.; Adams, L. A. Nat. Rev. Endocrinol. 2011, 7, 456−465.(25) (a) Lehrke, M.; Lazar, M. A. Cell 2005, 123, 993−999.(b) Mudaliar, S.; Henry, R. R. Annu. Rev. Med. 2001, 52, 239−257.(26) Gale, E. A. M. Lancet 2001, 357, 1870−1875.(27) (a) Palee, S.; Chattipakorn, S.; Phrommintikul, A.;Chattipakorn, N. World J. Cardiol. 2011, 26, 144−152. (b) Woodcock,J.; Sharfstein, J. M.; Hamburg, M. N. Engl. J. Med. 2010, 363, 1489−1491.(28) (a) Withdrawal from the Market of Pioglitazone-ContainingProducts: Actos® and Competact® in France, July 2011. Available:http://www.takeda.com/press/article_42712.html [date visited: 31/05/2012]. (b) The Antidiabetic Actos Withdrawn from the Market,June 2011. Available: http://www.lefigaro.fr/flash-actu/2011/06/09/97001-20110609FILWWW00505-info-le-figaro- antidiabetique-actos-retire-du-marche.php [date visted: 31/05/2012]. (c) FDA DrugSafety Communication: Update to Ongoing Safety Review of Actos(pioglitazone) and Increased Risk of Bladder Cancer, June 2011.Available: http://www.fda.gov/Drugs/DrugSafety/ucm259150.htm[date visited: 31/05/2012].(29) Guidance for Industry Diabetes Mellitus: Developing Drugs andTherapeutic Biologics for Treatment and Prevention, February 2008.A v a i l a b l e : h t t p : / / www . f d a . g o v / d ow n l o a d s /D r u g s /

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010449

GuidanceComplianceRegulatoryInformation/Guidances/ucm071624.pdf [date visited: 31/05/2012].(30) Baughman, T. M.; Graham, R. A.; Wells-Knecht, K.; Silver, I. S.;Tyler, L. O.; Wells-Knecht, M.; Zhao, Z. Drug Metab. Dispos. 2005, 33,733−738.(31) Uchiyama, M.; Fischer, T.; Mueller, J.; Oguchi, M.; Yamamura,N.; Koda, H.; Iwabuchi, H.; Izumi, T. Drug Metab. Dispos. 2010, 38,946−956.(32) Alvarez-Sanchez, R.; Montavon, F.; Hartung, T.; Pahler, A.Chem. Res. Toxicol. 2006, 19, 1106−1116.(33) Frisch, M. J.; Rega, N.; Petersson, G. A.; Trucks, G. W.;Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,J.; Ishida, M.; Burant, J. C.; Nakajima, T.; Honda, Y.; Kitao, O.;Schlegel, H. B.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H.P.; Cross, J. B.; Millam, J. M.; Bakken, V.; Adamo, C.; Jaramillo, J.;Gomperts, R.; Scuseria, G. E.; Stratmann, R. E.; Yazyev, O.; Austin, A.J.; Cammi, R.; Pomelli, C.; Iyengar, S. S.; Ochterski, J. W.; Ayala, P. Y.;Morokuma, K.; Voth, G. A.; Salvador, P.; Robb, M. A.; Dannenberg, J.J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Tomasi, J.; Strain,M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;Foresman, J. B.; Cheeseman, J. R.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;Barone, V.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;Liashenko, A.; Piskorz, P.; Komaromi, I.; Montgomery, J. A., Jr.;Martin, R. L.; Fox, D. J.; Mennucci, B.; Keith, T.; Al-Laham, M. A.;Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;Johnson, B.; Chen, W.; Vreven, T.; Wong, M. W.; Cossi, M.; Gonzalez,C.; Pople, J. A.; Kudin, K. N.; Scalmani, G. Gaussian 03, Revision C.02;Gaussian, Inc: Wallingford, CT, 2004.(34) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098−3100. (b) Lee, C.T.; Yang, W. T.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789. (c) Becke,A. D. J. Chem. Phys. 1993, 98, 5648−5652. (d) Hay, P. J.; Wadt, W. R.J. Chem. Phys. 1985, 82, 270−283. (e) Friesner, R. A.; Murphy, R. B.;Beachy, M. D.; Ringnalda, M. N.; Pollard, W. T.; Dunietz, B. D.; Cao,Y. J. Phys. Chem. A 1999, 103, 1913−1928.(35) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502−16513.(36) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,83, 735−746.(37) (a) de Visser, S. P.; Ogliaro, F; Sharma, P. K.; Shaik, S. J. Am.Chem. Soc. 2002, 124, 11809−11826. (b) Shaik, S.; Hirao, H.; Kumar,D. Acc. Chem. Res. 2007, 40, 532−542. (c) Shaik, S.; Cohen, S.; Wang,Y.; Chen, H.; Kumar, D.; Thiel, W. Chem. Rev. 2010, 110, 949−1017.(d) Rittle, J.; Green, M. T. Science 2010, 330, 933−937. (e) Taxak, N.;Desai, P. V.; Patel, B.; Mohutsky, M.; Klimkowski, V. J.; Gombar, V.;Bharatam, P. V. J. Comput. Chem. 2012, 33, 1740−1747.(38) Schaefer, A.; Huber, C.; Ahlrichs, R. J. J. Chem. Phys. 1994, 100,5829−5835.(39) (a) Cances, M. T.; Mennucci, B.; Tomasi, J. A. J. Chem. Phys.1997, 107, 3032−3037. (b) Mennucci, B.; Tomasi, J. J. Chem. Phys.1997, 106, 5151−5158. (c) Mennucci, B.; Cances, E.; Tomasi, J. J.Phys. Chem. B 1997, 101, 10506−10517. (d) Tomasi, J.; Mennucci, B.;Cances, E. J. Mol. Struct.: THEOCHEM 1999, 464, 211−226.(40) (a) Lonsdale, R.; Harvey, J. N.; Mulholland, A. J. J. Chem. TheoryComput. 2012, X, X ; DOI DOI: 10.1021/ct300329h. (b) Grimme,S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104−154119. (c) Grimme, S. J. Comput. Chem. 2006, 27, 1787−1799.(41) Madsen, K. G.; Gronberg, G.; Skonberg, C.; Jurva, U.; Hansen,S. H.; Olsen, J. Chem. Res. Toxicol. 2008, 21, 2035−2041.(42) Smith, M. T. Chem. Res. Toxicol. 2003, 16, 679−687.(43) Lebovitz, H. E. Diabetes Metabol. Res. Rev. 2002, 18, S23−S29.(44) Kawai, K.; Kawasaki-Tokui, Y.; Odaka, T.; Tsuruta, F.; Kazui,M.; Iwabuchi, H.; Nakamura, T.; Kinoshita, T.; Ikeda, T.; Yoshioka, T.;et al. Arzneim. Forsch. 1997, 47, 356−368.(45) Krieter, P. A.; Colletti, A. E.; Doss, G. A.; Miller, R. R. DrugMetab. Dispos. 1994, 22, 625−630.(46) (a) Toufar, H.; Nulens, K.; Janssens, G. O. A.; Mortier, W. J.;Schoonheydt, R. A.; De Proft, F.; Geerlings, P. J. Phys. Chem. B 1996,100, 15383−15387. (b) De Proft, F.; Geerlings, P. Chem. Rev. 2001,101, 1451−1464. (c) Geerlings, P.; De Proft, F.; Langenaeker, W.Chem. Rev. 2003, 103, 1793−1873. (d) Fievez, T.; Sablon, N.; De

Proft, F.; Ayers, P. W.; Geerlings, P. J. Chem. Theory Comput. 2008, 4,1065−1072. (e) Roos, G.; Geerlings, P.; Messens, J. J. Phys. Chem. B2009, 113, 13465−13475. (f) Muya, J. T.; De Proft, F.; Geerlings, P.;Nguyen, M. T.; Ceulemans, A. J. Phys. Chem. A 2011, 115, 9069−9080.(47) (a) Brown, T. N.; Mora-Diez, N. J. Phys. Chem. B 2006, 110,9270−9279. (b) Barbosa, L. A. M. M.; van Santen, R. A. J. Mol. Struct.:THEOCHEM 2000, 497, 173−188.(48) Chattaraj, P. K.; Sarkar, U.; Roy, D. R. Chem. Rev. 2006, 106,2065−2091.(49) Chattaraj, P. K.; Chakraborty, A.; Giri, S. J. Phys. Chem. A 2009,113, 10068−10074.(50) Fukui, K. Science 1982, 218, 747−754.(51) Fukui, K.; Yonezawa, T.; Shingu, H. J. Chem. Phys. 1952, 20,722−725.(52) Dixit, V. A.; Bharatam, P. V. Chem. Res. Toxicol. 2011, 24, 1113−1122.(53) Pratihar, S.; Roy, S. J. Org. Chem. 2010, 75, 4957−4963.(54) Wondrousch, D.; Bohme, A.; Thaens, D.; Ost, N.; Schuurmann,G. J. Phys. Chem. Lett. 2010, 1, 1605−1610.(55) Shellhamer, D. F.; Davenport, K. J.; Hassler, D. M.; Hickle, K.R.; Thorpe, J. J.; Vandenbroek, D. J.; Heasley, V. L.; Boatz, J. A.;Reingold, A. L.; Moore, C. E. J. Org. Chem. 2010, 75, 7913−7916.(56) Benigni, R.; Bossa, C. Chem. Rev. 2011, 111, 2507−2536.(57) (a) Saunders, J. H.; Slocombe, R. J. Chem. Rev. 1948, 43, 203−218. (b) Braunstein, P.; Nobel, D. Chem. Rev. 1989, 89, 1927−1945.(c) Raders, S. M.; Verkade, J. G. J. Org. Chem. 2010, 75, 5308−5311.(d) Loscutoff, P. W.; Wong, K. T.; Bent, S. F. J. Phys. Chem. C 2010,114, 14193−14201.(58) Arroyo, S. T.; Garcia, A. H.; Martin, J. A. S. J. Phys. Chem. A2009, 113, 1858−1863.(59) Raspoet, G.; Nguyen, M. T. J. Org. Chem. 1998, 63, 6878−6885.(60) Zagorevskii, D.; Song, M.; Breneman, C. J. Am. Soc. MassSpectrom. 2003, 14, 881−892.(61) Zhang, D.; Wu, L; Koch, K. J.; Cooks, R. G. Eur. Mass Spectrom.1999, 5, 353−361.(62) Eberlin, M. N. Eur. Mass Spectrom. 2007, 13, 19−28.

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp308023g | J. Phys. Chem. A 2012, 116, 10441−1045010450