Dehydrochlorination of Hexachlorocyclohexanes Catalyzed by the

Subscriber access provided by The Chinese University of Hong Kong

Article

Dehydrochlorination of Hexachlorocyclohexanes Catalyzed by the LinA Dehydrohalogenase. A QM/MM Study Rabindra Nath Manna, Kirill Zinovjev, Iñaki Tuñón, and Agnieszka Dybala-Defratyka J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.5b07538 • Publication Date (Web): 11 Nov 2015 Downloaded from http://pubs.acs.org on November 19, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Dehydrochlorination of Hexachlorocyclohexanes Catalyzed by the LinA Dehydrohalogenase. A QM/MM Study

Rabindra Nath Manna,1 Kirill Zinovjev,2 Iñaki Tuñón2 and Agnieszka Dybala-Defratyka1*

1

Institute of Applied Radiation Chemistry, Faculty of Chemistry, Lodz University of Technology, Zeromskiego 116, Lodz 90-924, Poland 2

Departament de Química Física, Universitat de València, 46100 Burjassot, Spain

* To whom the correspondence should be addressed: Institute of Applied Radiation Chemistry, Lodz University of Technology Zeromskiego 116, 90-924 Lodz, Poland Phone: +48 42 631 3198 Fax: +48 42 636 5008 E-mail: [email protected]

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 32

Abstract The elucidation of the catalytic role of LinA dehydrohalogenase in the degradation processes of hexachlorocyclohexane (HCH) isomers is extremely important to further studies on the bioremediation of HCH polluted areas. Herein, QM/MM free energy simulations are employed to provide the details of the dehydrochlorination reaction of two HCH isomers (γ and β). In particular, the role of the protonation state of one of the catalytic residues - His73 is explored. Based on our calculations two distinct minimum free energy pathways (concerted and stepwise) were found for γ-HCH and β-HCH. The choice of the reaction channel for the dehydrochlorination reactions of γ- and β-HCH was shown to depend on the initial mutual orientations of the reacting species in the active site and the protonation form of His73. The sequential pathway comprises the transfer of the proton (Hδ1) between His73 and Asp25 and subsequent the H1/Cl2 pair elimination from the substrate molecule. Within a concerted mechanism, the dehydrochlorination reaction of γ-/β-HCH is initiated with neutral His73 and the Hδ1 proton is transferred upon final product formation. We found that the concerted pathway for β-HCH results in significantly higher free energy of activation than the stepwise route and therefore can be disregarded as not a feasible mechanism. On the other hand the reaction that occurs with much lower energetic barrier requires a stronger base (i.e., anionic His73) to abstract the proton (H1) from the substrate molecule. The presence of such transient form of His results in higher energy than the respective Michaelis complex and was observed only in the stepwise pathway for both isomers. Furthermore, we have concluded that both pathways (concerted and stepwise) are feasible for the dehydrochlorination reaction of γHCH. The activation free energies obtained from the M05-2X/6-31+G(d,p) corrected path coordinate PMF profiles for the dehydrochlorination reactions of the γ-/β-HCH are in good agreement with the experimental values.

2


Page 3 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Keywords LinA, HCH, dehydrochlorination, QM/MM, free energy simulations, string method, path collective variable, PMF

1.

Introduction Hexachlorocyclohexane (HCH) isomers are highly persistent chlorinated pollutants.1

Based on the toxicological study, HCH isomers cause damage of central nervous system, renal, liver, kidney and reproductive system.2 Although the usage of HCH isomers have been banned in most of the countries under the Stockholm convention,3 they are still being manufactured in some countries as a seed dressing in agriculture and as a human medicinal for control of lice and scabies.4 Dehalogenases from the Lin family of enzymes have been shown to metabolize HCHs to less toxic compounds via two distinctive mechanisms. LinA dehydrochlorinates the HCH substrates via elimination mechanism5 (Scheme 1) whereas LinB catalyzes the hydrolytic dehalogenation via SN2 pathway.6,7

Scheme 1. Conversion of the γ-HCH and β-HCH isomers to 1,3R,4S,5S,6R- and

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 32

1,3R,4S,5R,6S-pentachlorocyclohexene (PCCH) catalyzed by LinA in which active site His73 and Asp25 act as a catalytic dyad.

LinA belongs to a class of enzymes known as the dehydrohalogenases.8 It readily degrades α-HCH, γ-HCH, and δ-HCH isomers9 and has been shown to have very low but still detectable activity towards β-HCH.10 γ-HCH known as lindane,11 has three axial chlorine atoms and this isomer is claimed to be responsible for insecticidal activity.12 Among these four HCH isomers, β-HCH has a much higher melting point and higher bio-concentration factor in human.13 All chlorine atoms of β-HCH are in equatorial positions in the cyclohexane ring which results in larger chemical stability of this isomer. LinA and LinB dehalogenases have numerous potential bioremediation applications, including roles in industrial biocatalysis.14,15,16,17 The dehydrochlorination reactions of HCHs may involve several chemical steps (Scheme 2 and Scheme S1). However, key metabolites of the dehydrochlorination reactions of HCH isomers have to be properly described in order to assess the fate of HCH isomers in the environment.

4


Page 5 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scheme 2. Postulated reaction mechanisms for γ-HCH catalyzed by LinA. Displayed atom numbering and nomenclature are used throughout this study.

Although some efforts towards the elucidation of the reaction mechanisms have been already made,10,10 also by some of us,18 the sequence in which the transfers of the H1 and Hδ1 protons occur remains unresolved. Using reduced models of the LinA active site, Brittain et al.,10 showed that the most favorable scenario is when the Hδ1 proton of His73 is first transferred to Asp25 and an anionic His73 is formed. Then, this residue abstracts the axial (H1) proton from the HCH substrate and the chlorine substituent is eliminated. Furthermore, they predicted that γ-HCH and β-HCH react via E2 and E1CB elimination mechanisms, respectively. On the other hand, the results coming from our previous study18 clearly indicated that the dehydrochlorination reaction of HCH followed a concerted mechanism in which these two protons transfers, the C2-Cl2 bond cleavage and the C1-C2 double bond formation occur 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 32

within a single reaction step. In Scheme 2, this mechanism would correspond to a diagonal leading from the bottom left-side corner to the upper right-side corner (see also SI, Scheme S1 for β-HCH). Furthermore, we showed18 that although γ-HCH and β-HCH share the same net mechanism, they differ in the transition state structures (concerted anti-E2 for γ-HCH and

syn-E2 for β-HCH), and the latter one is not readily degraded by LinA. The concerted pathway, resulted in an energetic barrier higher by about 7 kcal·mol-1 for β-HCH than for γHCH. Comparison of some key distances in the transition state structures located in these two earlier studies10,18 shows that the Oδ1-Hδ1 distance in β-HCH is similar whereas, the C1-H1 bond in β-HCH was significantly more elongated (1.95 Å) in the structure obtained by us than by Brittain et al., (1.49 Å) (SI, Scheme S1). With respect to γ-HCH the prediction of the protonation state of the Nδ1 atom differs significantly (Nδ1-Hδ1 = 1.75 Å and 1.16 Å at the HF/6-31+G(d,p)10 at the M05-2X/6-31+G(d,p) levels of theory,18 respectively). Part of these discrepancies may be a consequence of using different theory levels that led to different mechanistic scenarios. A similar observation has been very recently provided by Mlynsky et al.,19 for the mechanism of hairpin ribozyme. Another possible reason for the different findings obtained in these two studies may arise from the lack of considerations of a full enzymatic environment in the models employed.10,18 Additionally, both previous studies10,18 are the only reports in literature showing the possibility of β-HCH metabolism by LinA. Other recent works20 did not find any activity of LinA towards the β isomer. Atomiclevel studies using a realistic protein environment are, therefore, essential to improve the understanding of degradation processes of HCH isomers. We have employed a quantum mechanical (QM)/molecular mechanical (MM) methodology, which allows to include the protein environment in the models of enzymatic reactions. However, the main problem of studying complex chemical processes (the LinA catalyzed dehydrochlorination reactions not being an exception) is the large number of order

6


Page 7 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


parameters involved. The conventional free energy calculation techniques, such as umbrella sampling21 and metadynamics22 are limited by the so-called “curse of dimensionality”: they become prohibitively expensive when the dimensionality of the free energy surface to be explored becomes larger than two. To address this issue we have employed the string method -path collective variable approach.23,24,25,26,27 This combination of techniques was shown to successfully treat complex reactions catalyzed by enzymes, such as isochorismate pyruvate lyase,25,27,28 guanidinoacetate methyltransferase,26 and N6-adenine methyltransferase.29 Shortly, the main idea of such an approach is as follows: instead of calculating the entire free energy surface, the effort is made in the determination of the minimum free energy path (MFEP) on this surface, because the path followed by most of the reactive trajectories must be close to it. Then, a one-dimensional reaction coordinate is defined along this MFEP and the monodimensional free energy profile is determined. This framework reduces the multidimensional process to one single dimension and hence the process becomes tractable by any conventional free energy sampling technique.

2. Computational details 2.1. Enzyme models preparation Initial coordinates of the LinA-γ-HCH and LinA-β-HCH complexes were taken from our previous docking study.18 In order to prepare these complexes for further simulations the Amber 12 package30 was used. The AM1-BCC charges31 of γ-HCH and β-HCH were calculated using the Antechamber module. The ff99SB32 and GAFF force fields were used to model LinA, γ-HCH, and β-HCH isomers. 21 sodium ions were added to neutralize the systems. The resulting complexes were solvated with a 10 Å radius buffer zone of water molecules around the enzyme complex in each direction. The final model systems of LinA-γHCH and LinA-β-HCH consisted of 38107 and 38110 atoms, respectively, including 10338 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 32

and 10339 water molecules described by TIP3P potential.33 The minimization procedure used was as previously described.34 The system was heated from 100 to 300 K during 20 ps using the NVT ensemble with a time step of 1 fs. Then, 1.2 ns of classical molecular dynamics were performed using the NPT ensemble. The final configurations of the enzyme-substrate (ES) complexes (LinA-γ/β-HCH) resulted from the classical molecular dynamics (MD) were used as initial structures for the QM/MM MD simulations. The QM part consisted of the γ-HCH or β-HCH molecule, respectively, the imidazole ring of His73 and the side chain of Asp25. The QM subsystem was treated with the PM3 Hamiltonian.35 Two link atoms were placed between the Cβ and Cγ atoms of His73 and Asp25 to saturate the valence of the QM subsystem. The rest of the protein, water molecules and 21 Na+ ions were part of the MM region (where MM part was defined by amberff99SB force fields). Each ES complex was subjected to 3 ns of the QM/MM MD simulation using the NVT ensemble at 300 K.

2.2. QM/MM free energy simulations At first, we have performed QM/MM simulations using the on-the-fly string method24,25, 26

to obtain minimum free energy paths (MFEPs).27,29 A total of nine collective variables was

defined to find these MFEPs. Seven collective variables (CVs) out of these nine were related to key distances: C2-Cl2, C1-H1, Nε2-H1, Nδ1-H1, Oδ1-H1, C1-Nε2, and Nδ1-Oδ1, respectively (Scheme 2 and Scheme S1). The other two CVs were associated with the hybridization change for the C1 and C2 atoms from sp3 to sp2 during the dehydrochlorination reaction (Scheme 2). The hybridization coordinate is defined as the distance between the atom and the plane formed by its three substituents in the sp2 state. This multidimensional space forms a sufficient basis to cover all the different mechanistic proposals for the systems under study. Different initial guesses were used in this method to explore all possible mechanisms for the 8


Page 9 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


dehydrochlorination reactions of γ-/β-HCH. The force constants were 6000 for the distance and 15000 kJ·mol-1·Å-2 for the hybridization CVs, respectively. A total of 60 string nodes was used to trace the MFEPs for each reaction. Each string node was subjected to 50 ps of Langevin dynamics at 300 K with 1 fs time step. Replica Exchange36 between the string nodes was used to improve the sampling with exchange attempts performed every 100 fs. Once the MFEPs were obtained, a metric-corrected path collective variable26,27 was defined to obtain the potential of mean force (PMF) using the umbrella sampling technique.21 A set of 100 points was interpolated from 60 string nodes to define the reference path and ensure the smooth behavior of the coordinate. For each converged string, 100 umbrella sampling windows were used with force constant ranging from 400 to 3000 kJ·mol-1·Å-2. Each umbrella sampling window was equilibrated for 2 ps followed by 20 ps of production MD simulation with the time step of 0.5 fs. In the PMF calculations, replica exchanges were attempted with interval of 50 fs. The weighted histogram analysis method (WHAM)37 was then used to obtain the PMF. The error bars for the PMFs were calculated using Bayesian bootstrap technique.38 Each of 100 trajectories were split into 10 fragments, resulting in 1000 histograms. 200 bootstrap samples were generated for each case. The PMFs obtained at the PM3/MM level (where, MM=OPLS-AA39 and TIP3P33 potential for enzyme and water molecules, respectively) were corrected using interpolation correction technique.40,41 The corrected free energy was defined as a function of the path collective variable (s) using following equation, E = E + E/ + E + Spl ∆E s

where, Spl denotes a one dimensional cubic splines interpolation for the correction term.

∆E s is a difference between the single point high level (HL) energy (such as HL = M052X/6-31+G(d,p))42,43 of the QM system and the energy resulted from the low level (LL) of

9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 32

theory (LL=PM3) along the path coordinate (s). In these QM/MM simulations the Lennard Jones parameters for all atoms were taken from the OPLS-AA potential except for the chlorine atoms which were adopted from Kaminski et al.44 All calculations using the string method, path collective variables PMF and interpolation corrections were carried out with fDynamo libraries45 and its interface with Gaussian 09.46

3. Results and Discussion 3.1

MD simulations of enzyme-substrate (ES) complexes

3 ns of QM/MM MD simulations of the LinA-γ-HCH and LinA-β-HCH complexes were performed to obtain equilibrated structures for modeling LinA reactivity. The monitored root-mean-square-deviations (RMSD) of protein backbone atoms and active site residues relative to the respective starting structures resulted from docking (see SI, Figure S1, Table S1) indicated that both complexes reached equilibrium after 1 ns of simulations and remained stable. The LinA-β-HCH complex exhibited a slightly higher RMSD compared to the LinA-γHCH. The average active site structures of the two LinA-substrate complex resulting from the PM3/MM MD simulations are shown in Figure 1. Both isomers remained in the chair conformation resulted from docking after the simulation time. The active site of LinA consists of the side chains of Lys20, Leu21, Val24, Asp25, Trp42, Leu64, Phe68, Cys71, His73, Val94, Leu96, Ile109, Phe113, and Arg129 residues (Figure 1). Although only minor positional distortions of Phe68, Val94, Leu96 and Ile109 residues were spotted in the active site of the ES complexes some interesting differences were observed between the LinA-γHCH and LinA-β-HCH complexes.

10


Page 11 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Superimposed structures of active site residues obtained for the LinA-γ/β-HCH complexes. Structures shown in green and red color represent the average structures of LinAγ-HCH, and LinA-β-HCH complexes, respectively calculated from the last ns of 3 ns PM3/MM MD simulation.

During the 3 ns of QM/MM dynamics, the average Nε2(His73)···H1 distance in the LinA-γ-HCH complex is larger than in the LinA-β-HCH complex (2.59 vs 1.94 Å, respectively) (Table S2). This difference can be explained by the fact that γ-HCH has two biaxial pairs (H1/Cl2 and H5/Cl4) and Nε2 can abstract either the H1 or H5 proton from γ-HCH. During our simulation Nε2(His73) alternatively interacts with H1 and H5 (Figure S2). The H1/Cl2 pair elimination leads to 1,3(R),4(S),5(S),6(R)-PCCH whereas the elimination of H5/Cl4 pair yields 1,3(S),4(R),5(R),6(S)-PCCH.9,10,18 It has been reported that LinA enantiotopologically distinguishes between these two biaxial H/Cl pairs giving the 1,3(R),4(S),5(S),6(R)-PCCH enantiomer exclusively.5,9,10,18 The results obtained from our previous truncated models in which the elimination of two different H1/Cl2 and H5/Cl4 pairs of

11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 32

γ-HCH was investigated turn out to be very similar.18 Therefore, in this work we focus on the elimination of the H1/Cl2 pair. Another interesting observation is related to the interactions between the leaving chloride anion (Cl2) (in γ-/β-HCHs) and the neighboring residues being postulated to stabilize it right after its elimination from the substrate molecule. Due to the different HCH conformation the hydrogen bond distances between Cl2 and Arg129 and Trp42 in β-HCH are significantly larger than those found in γ-HCH (Table S2).

3.2

Dehydrochlorination reactions of γ-/β-HCH catalyzed by LinA

All possible reaction pathways for the dehydrochlorination reactions of γ-/β-HCH have been explored using different initial guesses in the string-method (Figure S3). At the PM3/MM level two distinct reaction channels, named path-A and path-B, were found for γ-/βHCH. Each string converged within the first 20 ps of QM/MM dynamics (Figure S5). Evolution of seven distance CVs and two hybridization CVs along the string nodes is shown in Figure S6. Within path-A for γ-/β-HCH isomers (Figure 2 and Figure S4) the H1 proton abstraction, the chloride elimination, and the Hδ1 transfer take place in a single step. However, at the TSa structure the Hδ1 proton is still attached to the His73 moiety and thus proton abstraction from the substrate is carried out by a neutral histidine. Path-B for both substrates (γ- and β-HCH) corresponds to a stepwise reaction mechanism (ESTSb1 EITSb2EP) in which the first step comprises the Hδ1 proton transfer to Asp25, and the second one involves the elimination of the H1/Cl2 pair with concomitant C1 and C2 carbon atoms hybridization change (Figure 2). The residue in charge of proton abstraction from the substrate is then an anionic histidine. The free energy of activation along path-A from the PM3/MM PMF profile after the M05-2X/6-31+G(d,p) corrections is 13.2 kcal·mol-1 for the dehydrochlorination reaction of γHCH, which is slightly lower than the value derived from the experimental10 rate constant10 12


Page 13 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


16.0 kcal·mol-1 (Table 1). The average C2-Cl2 bond lengths at the ES, TSa and EP structures of γ-HCH are 1.79, 1.95 and 3.60 Å, respectively. The average C1-H1 and Nε2-H1 interatomic distances at TSa for γ-HCH are 1.54 and 1.19 Å, which are consistent with our previous results (1.51 Å and 1.22 Å, respectively).18 Moreover, the average Nδ1-Hδ1 and Oδ1-Hδ1 distances at TSa are 1.05 and 1.69 Å, respectively, which indicates that the Hδ1 proton is still attached to His73. Based on the PMF profile, path-A for γ-HCH follows a concerted (ESTSaEP) anti-E2 mechanism via TSa structure in which the Nε2 and Nδ1 atoms of His73 are protonated (Figure 3). For the dehydrochlorination of β-HCH, the free energy of activation along the path-A obtained from the M05-2X corrected PMF is 25.0 kcal·mol-1 which is significantly higher than the experimental values of 21.5 kcal·mol-1. At the TSa structure of β-HCH (Figure 4), the Hδ1 atom is bound to the Nδ1 atom (Nδ1-Hδ1 = 1.05 Å) of His73 whereas, the Oδ1-Hδ1 and Nε2H1 distances are 1.68 and 1.14 Å, respectively. The average C1-H1 and C2-Cl2 bond distances at TSa of β-HCH are 1.69 and 2.13Å which are larger than the aforementioned distances at TSa of γ-HCH.

Path-A and Path-B of γ-HCH

Path-A and Path-B of β-HCH

Figure 2. M05-2X/6-31+G(d,p) corrected PMFs obtained as a function of the path collective 13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 32

variable for the dehydrochlorination reactions of γ-/β-HCH isomers. TSa represents transition state along the path-A of γ-/β-HCH dehydrochlorination, TSb1 and TSb2 represent the Hδ1 proton transfer transition state and the H1/Cl2 pair elimination transition state along the path-B for both γ-/β-HCH isomers. ES, EI, and EP denote enzyme-substrate, enzyme-intermediate, and enzyme-product, respectively.

Table 1. Relative free energies (kcal·mol-1) of all stationary points along the reaction pathways obtained for the dehydrochlorination reactions of γ- and β-HCH at the different levels of theory. γ-HCH Path-A PM3/MM PMF M05-2X/6-31+G(d,p) corrected PM3/MM PMF Path-B PM3/MM PMF M05-2X/6-31+G(d,p) corrected PM3/MM PMF β-HCH Path-A PM3/MM PMF M05-2X/6-31+G(d,p) corrected PM3/MM PMF Path-B PM3/MM PMF M05-2X/6-31+G(d,p) corrected PM3/MM PMF

ES 0.0 0.0 ES TSb1 0.0 3.7 0.0 5.1

TSa 10.8 13.2 EI TSb2 1.0 15.2 3.3 17.2

EP -22.1 -33.4 EP -21.8 -32.2

ES 0.0 0.0 ES TSb1 0.0 5.6 0.0 6.5

TSa EP 17.3 -26.2 25.0 -14.2 EI TSb2 EP 1.5 17.3 -21.7 2.3 18.8 -25.3

On the other hand, during the first step along path-B the average Nδ1-Hδ1 and Oδ1-Hδ1 distances at the TSb1 structure of the γ-HCH reaction are 1.25 Å and 1.33 Å, respectively. The same distances in the case of β-HCH are 1.30 and 1.32 Å, respectively (Figures 3 and 4). The free energy of activation for the Hδ1 proton transfer step in the case of β-HCH is 1.4 kcal·mol1

higher than for γ-HCH: 5.6 and 6.5 kcal·mol-1, respectively (Table 1, M05-2X/6-31+G(d,p)

corrected PMFs). In the intermediate (EI) structures corresponding to both substrates, the average Oδ1-Hδ1 and Nδ1-Hδ1 distances are ~1.0 Å and ~1.7 Å, respectively, which indicates that the Nδ1 atom of imidazole ring of His73 becomes fully deprotonated. Moreover, the C214


Page 15 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Cl2 as well as the C1-H1 bond lengths at TSb1 and EI in the reactions for both substrates remain unchanged with respect to the initial values (Table 2). Only the C1-Nε2 distance becomes significantly shortened at the EI structure which might indicate the system preparation for the next step, dehydrochlorination. The EI structure is not particularly stabilized in the LinA active site and it does not constitute a ground state species of the enzyme. In the case of both isomers its energy is higher than that for the ES complex (Table 1). In the Michaelis, ground state complex His73 is found in its neutral form. Therefore, an unprotonated His73 should be only considered as a transient species which is formed during the reaction to assist proton abstraction from the substrate. This higher energy form of His73 seems to play a crucial role in the reaction with β-HCH. However, in the case of γ isomer the pathway assuming the formation of such transient species results in higher activation energy. Earlier QM calculations by Brittain et al. also showed the possibility of such an anionic intermediate form of His73.

Table 2. Average key distances (Å) in all stationary points located for the dehydrochlorination reactions of the γ-/β-HCH isomers obtained from the PM3/MM PMFs.

Path-A C1-H1 C2-Cl2 Nε2-H1 Nδ1-Hδ1 Oδ1-Hδ1 C1-Nε2 Nδ1-Oδ1 Cl2···Hη21(Arg129) Cl2···Hε1(Trp42) Path-B C1-H1 C2-Cl2 Nε2-H1 Nδ1-Hδ1 Oδ1-Hδ1 C1-Nε2

γ-HCH ES 1.13 ± 0.03 1.79 ± 0.04 2.13 ± 0.09 1.04 ± 0.04 1.68 ± 0.06 3.02 ± 0.10 2.65 ± 0.08 6.80 ± 0.84 3.78 ± 0.69 ES TSb1 1.14 ± 0.04 1.13 ± 0.04 1.79 ± 0.04 1.79 ± 0.04 2.48 ± 0.17 2.14 ± 0.13 1.02 ± 0.03 1.25 ± 0.11 1.80 ± 0.07 1.33 ± 0.14 3.24 ± 0.08 3.07 ± 0.07

TSa 1.54 ± 0.10 1.95 ± 0.04 1.19± 0.05 1.05 ± 0.03 1.69 ± 0.08 2.65 ± 0.06 2.63 ± 0.08 4.66 ± 0.48 3.52 ± 0.69 EI TSb2 1.16 ± 0.04 1.51 ± 0.10 1.79 ± 0.04 1.97 ± 0.03 1.80 ± 0.08 1.20 ± 0.14 1.71 ± 0.05 1.72 ± 0.06 0.99 ± 0.03 0.98 ± 0.03 2.88 ± 0.05 2.74 ± 0.09

EP 2.64 ± 0.31 3.60 ± 0.10 1.00 ± 0.03 1.74 ± 0.07 0.98 ± 0.03 3.39 ± 0.19 2.69 ± 0.07 2.32 ± 0.20 2.80 ± 0.31 EP 2.63 ± 0.19 3.72 ± 0.09 1.00 ± 0.03 1.74 ± 0.07 0.98 ± 0.03 3.26 ± 0.14 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nδ1-Oδ1 Cl2···Hη21(Arg129) Cl2···Hε1(Trp42)

2.92 ± 0.06 6.35 ± 0.36 3.78 ± 0.55

2.55 ± 0.09 2.66 ± 0.05 6.36 ± 0.46 6.10 ± 0.48 3.70 ± 0.75 3.45 ± 0.30 β-HCH Path-A ES TSa C1-H1 1.15 ± 0.04 1.69 ± 0.20 C2-Cl2 1.79 ± 0.04 2.13 ± 0.04 Nε2-H1 1.90 ± 0.07 1.14± 0.26 Nδ1-Hδ1 1.04 ± 0.03 1.05 ± 0.03 Oδ1-Hδ1 1.71 ± 0.07 1.68 ± 0.07 C1-Nε2 3.02 ± 0.07 2.64 ± 0.07 Nδ1-Oδ1 2.68 ± 0.06 2.82 ± 0.09 7.33 ± 0.37 6.46 ± 0.30 Cl2···Hη21(Arg129) Cl2···Hε1(Trp42) 5.01 ± 0.29 5.20 ± 0.32 Path-B ES TSb1 EI C1-H1 1.14 ± 0.03 1.13 ± 0.03 1.16 ± 0.03 C2-Cl2 1.79 ± 0.04 1.79 ± 0.03 1.80 ± 0.04 Nε2-H1 2.33 ± 0.15 2.20 ± 0.18 1.77 ± 0.06 Nδ1-Hδ1 1.02 ± 0.03 1.30 ± 0.14 1.73 ± 0.06 Oδ1-Hδ1 1.90 ± 0.10 1.32 ± 0.09 0.99 ± 0.03 C1-Nε2 3.21 ± 0.11 3.11 ± 0.14 2.90 ± 0.05 Nδ1-Oδ1 2.81 ± 0.02 2.53 ± 0.13 2.66 ± 0.06 Cl2···Hη21(Arg129) 7.91 ± 0.30 7.87 ± 0.37 7.58 ± 0.50 Cl2···Hε1(Trp42) 5.56 ± 0.44 4.92 ± 0.34 4.93 ± 0.26

Page 16 of 32

2.67 ± 0.06 5.82 ± 0.53 3.44 ± 0.58

2.66 ± 0.07 2.22 ± 0.13 3.44 ± 0.61

EP 3.24 ± 0.14 3.46 ± 0.37 1.01 ± 0.04 1.74 ± 0.07 0.99 ± 0.03 4.06 ± 0.05 2.68 ± 0.07 3.90 ± 1.12 4.46 ± 0.51 TSb2 EP 1.78 ± 0.12 3.28 ± 0.19 2.05 ± 0.04 3.98 ± 0.44 1.10 ± 0.20 1.02 ± 0.04 1.74 ± 0.07 1.75 ± 0.07 0.98 ± 0.03 0.99 ± 0.03 2.81 ± 0.08 4.00 ± 0.15 2.69 ± 0.07 2.68 ± 0.07 6.88 ± 0.33 3.03 ± 1.21 4.84 ± 0.44 4.42 ± 0.37

TSa

16


Page 17 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TSb1

TSb2

Figure 3. Transition state structures along with average key distances (Å) for the dehydrochlorination reaction of γ-HCH (top: path-a and bottom: path-B).

In the second step of path-B, the anionic His73 abstracts the axial H1 proton from the substrate, the C2-Cl2 bond is broken and both C1 and C2 carbon atoms change their hybridization from sp3 to sp2 forming the product, a PCCH molecule. Representative snapshots of all transition state structures with the average key distances for the second step of path-B are shown in Figure 3 (γ-HCH) and Figure 4 (β-HCH). The free energy of activation for the elimination of the H1/Cl2 pair from the enzyme intermediate (EI) for γ- and β-HCH are 13.9 and 16.5 kcal·mol-1, respectively. Hence, the H1/Cl2 pair elimination step determines the rate of the overall dehydrochlorination reaction for both γ- and β-HCH isomers along path-B. The M05-2X corrected free energy barriers obtained in this study for the overall reactions along path-B are 17.2 and 18.8 kcal·mol-1 for γ- and β-HCH, respectively. The free energy barriers found along path-A for γ-HCH and β-HCH are 13.2 and 25.0 kcal·mol-1, respectively. Thus, the best predicted activation free energies for the dehydrochlorination reactions of γ-HCH and β-HCH are 13.2 and 18.8 kcal·mol-1, respectively. These values seem

17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 32

to be in a reasonable agreement with experimentally derived10 values for the dehydrochlorination reactions of γ-HCH (16.0 kcal·mol-1) and β-HCH (21.5 kcal·mol-1) although we are aware that the activation free energies derived from the experimental turnover numbers and the first order rate constants for the LinA reaction with the γ and β isomers provided so far in the literature should be treated with caution as preliminary numbers constituting upper limits for the chemical step catalyzed by LinA. However, interestingly, the difference between the activation free energies obtained for both isomers (5.6 kcal·mol-1) found in our simulations is in agreement with the difference derived from the experimental rate constants (5.5 kcal·mol-1)10, indicating that QM/MM simulations provide an excellent strategy to reproduce and rationalize the enzymatic preference for the γ isomer. Our simulations clearly indicate the preference of LinA for this isomer and that the activity towards the β isomer must be significantly lower. It should be also noted that the difference found between the activation free energies for path-A and path-B in the case of γ-HCH is modest, 4 kcal·mol-1, and then the preference for a particular mechanism could be shifted changing the strength of the abstracting base. Finally, the significantly higher free energy of activation found for path-A of the β-HCH dehydrochlorination reaction makes this mechanism the least feasible and therefore we have excluded it from further analysis. Once both reaction paths have been analyzed we can compare the differences between them and between substrates. The average C2-Cl2 (1.97 Å) and C1-H1 (1.51 Å) bond lengths at TSb2 for γ-HCH are similar to those at TSa (Table 2). However, the hybridization state change of the C1 and C2 atoms at TSb2 for the γ-HCH reaction is slightly more advanced than at TSa (Table S3). The two pathways described for the γ isomer share also some features, in particular related to the magnitude of the H1-C1-C2-Cl2 dihedral. In both cases the initial value is about -160° and deviates from this value only at TSa and TSb2 responsible for the H1/Cl2 pair elimination (Table S9). Since the adjacent trans-di-axial orientation of the H1/Cl2 pair is

18


Page 19 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


not present in β-HCH and all chlorine atoms of β-HCH are placed in the equatorial position of the cyclohexane ring a different behavior is expected in the dehydrochlorination of this isomer.

TSa

TSb1

TSb2

Figure 4. Transition state structures with the average key distances (Å) for the dehydrochlorination reaction of β-HCH (top: path-a and bottom: path-B).

As discussed in the Section 3.1 and shown in Table S2 the binding pose of the γ-HCH

19



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 32

produces a more favorable environment to stabilize one of the products - the leaving chloride anion. Thus, elimination reaction of the γ-HCH isomer can be initiated with a weaker base. However, in the case of β-HCH, the reaction seems to take place only if a stronger base (the anionic His73) abstracts the proton (H1). This observation is in contrast with our previous findings generated using the truncated model of the LinA active site, in which neither Trp42 nor Arg129 was present, and the leaving chloride anion was found to interact "artificially" with neighboring His73 polar hydrogen atoms. Additionally we observed two C-Cl bonds elongation during the dehydrochlorination reaction of β-HCH. These interactions resulted in a concerted pathway and the largest product stabilization among the studied HCH isomers.18 Therefore, the positioning of the substrate in the active site in the model containing a realistic enzymatic environment would explain the differences in the reaction mechanisms. The C1-H1 bond (1.78 Å) at TSb2 for β-HCH is significantly longer than at TSb2 for γ-HCH (1.51 Å). Furthermore, the C2-Cl2 bond at TSb2 of β-HCH is 2.05 Å whereas at TSb2 for γ-HCH is 1.97 Å. Although both isomers remain in a staggered chair conformation in all analyzed transition states (Figures S7 and S8) we found some interesting differences between them while analyzing the values of the H1-C1-C2-Cl2 dihedral and the C1-C2 bond length along the reaction coordinate (Table S9). First of all, in the case of γ-HCH we observe antiperiplanarity in the H1-C1-C2-Cl2 framework throughout the entire stepwise pathway. However, some deviation from a perfect anti staggered conformation was observed (the dihedral of -160°). The H1 proton transfer, which is quite advanced, is concertedly followed by the C1-C2 bond rotation, double bond formation and finally chloride anion elimination. On the other hand in the reaction with β-HCH we located a gauche transition state (Table S9). It is also noteworthy that the ES complex for the β isomer is in a gauche conformation too. In the case of both isomers the C1-C2 bond rotation and double bond formation take place soon before crossing the second transition state of path-B. The contraction of this bond relatively to the EI complex

20


Page 21 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


is only by 0.01 and 0.07 Å for γ and β-HCH, respectively. Since the difference between the activation energies obtained for path-B for two isomers is only 1.6 kcal⋅mol-1 one could expect that any hyperconjugation effect will be very similar and nearly negligible. Steric repulsions should not play any important role even so we observe some deviations from perfect staggered conformations (gauche and trans). One could rather envision that in the case of γHCH some stabilization of the transition state may originate from the 1,3-diaxial interactions between chlorine substituents and hydrogen atoms. These interactions do not appear in the βisomer. Nevertheless, the dehydrochlorination reaction of two isomers proceeds through E2 mechanism although none of them is synchronous in respect of the H1/Cl2 elimination step. Further differences between the isomers are also reflected in the distances between the leaving chloride ions and stabilizing moieties of Arg129 and Trp42 (Table 2). At the TSb2 structures for both isomers, the Nδ1 atom of His73 is not protonated. However, the anionic character of His73 is less pronounced in the case of β-HCH where the transfer of the H1 proton is more advanced at TSb2 compared with the same process for γ-HCH (Table 2). Among the reaction mechanisms studied in this work γ-HCH results in the final products (EP) that are the most stable (Table 1). The distances between the leaving chloride ion and the neighboring Arg129 and Trp42 at the EP complex formed for this isomer are the shortest (Table 2). Larger distances observed at the EP complex for the reaction with β-HCH are a consequence of the axial orientation of chlorine atoms in this isomer. Detailed atomic charges analysis performed using the ChelpG scheme47 for all stationary points located for all three pathways have also revealed interesting patterns (SI, Tables S4-S6). In particular distinguishing features between two isomers spotted for path-B are worth discussing. When we summed up the atomic charges on respective substrate, His73, and Asp25 separately, we observed that β-HCH is negatively charged in ES as compared with γ-HCH. This negative charge seems to be compensated by His73 carrying positive charge in 21



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 32

total, whereas in the case of γ-HCH there is an opposite trend (Tables S5-S6). Subsequently, upon the reaction the β isomer becomes gradually more negatively charged, up to -0.83 a.u. at EP. Naturally most of this charge comes from the chloride anion leaving group (-1.04 a.u.). γ-HCH, in contrast, exhibits almost no change during the first step of the reaction, small decrease upon crossing the second transition state and a sudden drop upon reaching the product state. Furthermore, in the case of β-HCH drastic changes of the sum of atomic charges on His73 and Asp25 were monitored, as expected, during the first step of the overall reaction. They resulted in the charge change of -0.50 and 0.43 a.u. on His and Asp, respectively. In the reaction with γ-HCH these values were -0.15 and 0.17 a.u., respectively. The EI complex formed as a result of the Hδ1 proton abstraction from the base in the case of γHCH exhibits some polarization with the HCH part being positively charged and the rest of it carrying negative charge. In the reaction with β-HCH all three components are negatively charged. Analysis of the HOMO/LUMO orbitals at TSb2 also reveals interesting differences between the two isomers (see Figure S9). In both cases the HOMO is located on the anionic His73 and the LUMO on the substrate. However, for γ-HCH the LUMO is essentially located on the leaving chloride atom, facilitating the development of a negative charge on this atom and the C-Cl bond breaking. In β-HCH the LUMO is delocalized on the entire substrate structure. In order to find the source of these differences we performed additional population analysis but without the presence of the environment and for the geometries of the ES complexes resulted from the QM/MM calculations (SI, Tables S7-S8). The results obtained within this analysis indicate that when γ-HCH is bound to the active site of LinA the effect of the environment is rather small and observed only for few atoms (H1, Nε2, Nδ1). Interestingly, the same analysis performed for β-HCH revealed much larger influence of the environment. All atomic charges on carbon atoms of the β isomer changed upon the presence of the enzyme 22


Page 23 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(Table S8).

4.

Conclusions In our present study we have explored the role of His73 protonation during the

dehydrochlorination reactions of γ-/β-HCH isomers in LinA, using QM/MM calculations along with the string method and path collective variable. Two distinct reaction pathways (concerted and stepwise) were found for γ-HCH and β-HCH. In the stepwise mechanism (here denoted as path-B) His73 deprotonation takes place before the elimination of the H1/Cl2 pair, which was found to be the rate-limiting step. The concerted asynchronous mechanism (denoted as path-A) proceeds with chlorine elimination and then His73 deprotonation. The analysis of the ES complexes obtained for two paths in the case of γ-HCH revealed a more compact arrangements of the reacting species in the active site in the case of path-A. The H1 proton donor-acceptor distance is shorter than in path-B, which in consequence leads to a mechanism in which the H1 abstraction happens first, while the proton transfer from His73 to Asp25 takes place once the transition state has been crossed. The estimated activation free energies are in good agreement with the range of the experimental values and, in particular, the differences observed between the both isomers match the experimental trend. This opens the possibility of future enzyme redesign to improve the activity with respect to a particular isomer, using the results of present simulations. Our results show that although two reaction channels could be possible for each of the two isomers, path-A seems to be preferred by γHCH and path-B by β-HCH. According to our simulations the orientation of the substrate in the active site may control whether singly protonated or un-protonated form of His73 will be required for the elimination reaction to proceed. The necessity of an unprotonated form of His73 in the active site of LinA for the metabolism of the β isomer may indicate that,

23



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 32

although the probability of this process is low, it should not be excluded from considerations. The initiation of the H1/Cl2 elimination by the neutral form of His73 resulted in the most energetically expensive process among all reactions studied in this work. Nevertheless, the findings for β-HCH presented in this paper may open up a possibility for the design of LinA variants where this process could take place with larger rate constants. In the case of γ-HCH, the binding of the substrate leads to a favorable environment for the stabilization of the resulting chloride anion and where the elimination process can take place with a weaker base (neutral His73) abstracting the proton. Furthermore, our results confirm that the degradation of β-HCH by LinA is possible although at lower rate, in excellent agreement with recent experimental findings.10 Additionally, both pathways found for γ-HCH exhibit some similar features with respect to the H/Cl elimination event which could at least constitute an initial set of conditions that need to be fulfilled for a reaction to occur at lower activation energy. It must be noticed that the concerted pathway described in this study is similar to our previous findings using the truncated model of LinA active site while the stepwise reaction mechanisms have many resemblances with that previously reported by Brittain et al. None of the groups, however, considered the possibility of two reaction mechanisms for the γ-HCH isomer. For this the presence of the full enzymatic environment was evidently necessary.

Supporting Information Proposed reaction steps for β-HCH, RMSD plots of ES complexes, key inter-atomic distances (Å) of LinA, LinA-γ/β-HCH complexes and convergence plots obtained from the string method are provided. The evolution of distances, hybridization coordinates of C1 and C2 atoms along the string and ChelpG charges of the QM atoms are also listed. This material is available free of charge via the Internet at http://pubs.acs.org.

24


Page 25 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Author information Corresponding Author E-mail: [email protected] Notes The authors declare no competing financial interest.

Acknowledgements RNM acknowledges the support from the EU MCA-ITN grant "CSI-Environment" (FP7264329, 2011-2015) and the fellowship for foreign graduate students from the Lodz University of Technology (2014-2015). Access to “Tirant” supercomputing facilities at the Universitat de València, is gratefully acknowledged. KZ and IT acknowledge support from the Ministerio de Economía y Competitividad, Spain (project CTQ2012-36253-C03-03). KZ acknowledges a FPU fellowship from Ministerio de Educación, Spain.

Abbreviations HCH: Hexachlorocyclohexane, QM/MM: Quantum Mechanics/Molecular Mechanics, PMF: Potential of Mean Force, MD: Molecular Dynamics, E2: Bimolecular elimination, ES: Enzyme-substrate, TS: Transition state, EI: Enzyme-intermediate, EP: Enzyme-product, CVs: collective variables.

References

25



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 32

(1) Lal, R.; Pandey, G.; Sharma, P.; Kumari, K.; Malhotra, S.; Pandey, R.; Raina, V.; Kohler, H. –P. E.; Holliger, C.; Jackson, C.; et al. Biochemistry of Microbial Degradation of Hexachlorocyclohexane and Prospects for Bioremediation. Microbiol. Mol. Biol. Rev. 2010,

74, 58–80. (2) Willett, K. L.; Ulrich, E. N.; Hites, R. A. Differential Toxicity and Environmental Fates of Hexachlorocyclohexane Isomers. Environ. Sci. Technol. 1998, 32, 2197–2207. (3) Lallas. P. L. The Stockholm Convention on Persistent Organic Pollutants. Am. J. Int. Law. 2001, 95, 692–708. (4) Weber, R.; Gaus, C.; Tysklind, M.; Johnston, P.; Forter, M.; Hollert, H.; Heinisch, H.; Holoubek, I.; Lloyd-Smith, M.; Masunaga, S.; et al. Dioxin- and POP-Contaminated SitesContemporary and Future Relevance and Challenges. Environ. Sci. Pollut. Res. 2008, 15, 363–393. (5) Okai, M.; Kubota, K.; Fukuda, M.; Nagata, Y.; Nagata, K.; Tanokura, M. Crystal Structure of γ-Hexachlorocyclohexane Dehydrochlorinase LinA from Sphingobium japonicum UT26. J. Mol. Biol. 2010, 403, 260–269. (6) Koudelakova, T.; Chovancova, E.; Brezovsky, J.; Monincova, M.; Fortova, A.; Jarkovsky, J.; Damborsky, J. Substrate Specificity of Haloalkane Dehalogenases. Biochem. J. 2011, 435, 345–354. (7) Heeb, N. V.; Zindel, D.; Geueke, B.; Kohler, H. –P. E.; Lienemann, P. Biotransformation of Hexabromocyclododecanes (HBCDs) with LinB-An-HCH-Converting Bacterial Enzyme.

Environ. Sci. Technol. 2012, 46, 6566–6574. (8) Lal, R.; Dadhwal, M.; Kumari, K.; Sharma, P.; Singh, A.; Kumari, H.; Jit, S.; Gupta, S. K.; Nigam, A.; Lal, D.; et al. Pseudomonas sp. to Sphingobium Indicum: a Journey of 26


Page 27 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Microbial Degradation and Bioremediation of Hexachlorocyclohexane. Indian J. Microbiol. 2008, 48, 3–18. (9) Tantírek, L.; Hynková, K.; Nagata, Y.; Murzin, A.; Ansorgova, A.; Sklenář, V.; Damborský, J. Reaction Mechanism and Stereochemistry of γ-Hexachlorocyclohexane Dehydrochlorinase LinA. J. Biol. Chem. 2001, 276, 7734–7740. (10) Brittain, D. R. B.; Pandey, R.; Kumari, K.; Sharma, P.; Pandey, G.; Lal, R.; Coote, M. L.; Oakeshott, J. G.; Jackson, C. J. Competing SN2 and E2 Reaction Pathways for Hexachlorocyclohexane Degradation in the Gas Phase, Solution and Enzymes. Chem.

Commun. 2011, 47, 976–978.

(11) Volchek. K.; Thouin, G.; Kuang, W.; Li, K.; Tezel, F. H.; Brown, C. E. The Release of Lindane from Contaminated Building Materials. Environ Sci Pollut Res. 2014, 21, 11844– 11855. (12)

Buser,

H.;

Muller,

M.

D.

Isomer

and

Enantioselective

Degradation

of

Hexachlorocyclohexme Isomers in Sewage Skdye under Anaerobic Conditions. Environ. Sci.

Technol. 1995, 29, 664–672. (13) Geyer, H. J.; Scheunert, I.; Korte, F. Correlation between the Bioconcentration Potential of Organic Environmental Chemicals in Humans and their n-octanol/water Partition Coefficients. Chemosphere 1987, 16, 239–252. (14) Swanson. P. Dehalogenases Applied to Industrial-Scale Biocatalysis. Curr. Opin.

Biotechnol. 1999, 10, 365–369. (15) Prokop, Z.; Sato, Y.; Brezovsky, J.; Mozga, T.; Chaloupkova, R.; Koudelakova, T.; Jerabek, P.; Stepankova, C. E.; Natsume, R.; Van Leeuwen, J. G. V.; et al. Enantioselectivity of Haloalkane Dehalogenases and its Modulation by Surface Loop Engineering. Angew.

Chem. Int. Ed. 2010, 49, 6111–6115. 27



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 32

(16) Stucki, G.; Thuer, M. Experiences of a Large-Scale Application of 1,2-Dichloroethane Degrading Microorganisms for Groundwater Treatment. Environ. Sci. Technol. 1995, 29, 2339–2345. (17) Prokop, Z.; Oplustil, F.; Defrank, J.; Damborsky, J. Enzymes Fight Chemical Weapons.

Biotechnol. J. 2006, 1, 1370–1380. (18) Manna, R. N.; Dybala-Defratyka, A. Insights into the Elimination Mechanisms Employed for the Degradation of Different Hexachlorocyclohexane Isomers using Kinetic Isotope Effects and Docking Studies. J. Phys. Org. Chem. 2013, 26, 797–804.

(19) Mlýnský, V.; Banáš, P.; Šponer, J.; Van der Kamp, M. W.; Mulholland, A. J.; Otyepka. M. Comparison of ab Initio, DFT, and Semiempirical QM/MM Approaches for Description of Catalytic Mechanism of Hairpin Ribozyme. J. Chem. Theory Comput. 2014, 10, 1608–1622.

(20) Geueke, B.; Garg, N.; Ghosh, S.; Fleischmann, T.; Holliger, C.; Lal, R.; Kohler, H. –P. E. Metabolomics of Hexachlorocyclohexane (HCH) Transformation: ratio of LinA to LinB Determines Metabolic Fate of HCH Isomers. Environ. Microbiol. 2013, 15, 1040–1049. (21) Torrie, G. M.; Valleau, J. P. Nonphysical Sampling Distributions in Monte Carlo FreeEnergy Estimation: Umbrella Sampling. J. Comput. Phys. 1977, 23, 187−199. (22) Barducci, A.; Bonomi, M.; Parrinello, M. Metadynamics. Wiley Interdiscip. Rev.

Comput. Mol. Sci. 2011, 1, 826−843. (23) Branduardi, D.; Gervasio, F. L.; Parrinello, M. From A to B in Free Energy Space. J.

Chem. Phys. 2007, 126, 054103. (24) Maragliano, L.; Vanden-Eijnden, E. On-the-fly String Method for Minimum Free Energy Paths Calculation. Chem. Phys. Lett. 2007, 446, 182–190.

28


Page 29 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(25) Zinovjev, K.; Martí, S.; Tuñón, I. A Collective Coordinate to Obtain Free Energy Profiles for Complex Reactions in Condensed Phases. J. Chem. Theory Comput. 2012, 8, 1795−1801. (26) Zinovjev, K.; Ruiz-Pernía, J. J.; Tuñón, I. Toward an Automatic Determination of Enzymatic Reaction Mechanisms and Their Activation Free Energies. J. Chem. Theory

Comput. 2013, 9, 3740−3749. (27) Zinovjev, K.; Tuñón, I. Exploring Chemical Reactivity of Complex Systems with Path-Based Coordinates: Role of the Distance Metric. J. Comput. Chem. 2014, 35, 1672−1681. (28) Sanchez-Martinez, M.; Field, M.; Crehuet, R. Enzymatic Minimum Free Energy Path Calculations Using Swarms of Trajectories. J. Phys. Chem. B 2015, 119, 1103−1113. (29) Aranda, J.; Zinovjev, K.; Roca, M; Tuñón, I. Dynamics and Reactivity in Thermus Aquaticus N6-Adenine Methyltransferase. J. Am. Chem. Soc. 2014, 136, 16227–16239. (30) Case, D. A.; Darden, T. A.; Cheatham, III. T. E.; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M.; et al. AMBER 12, 2012, University of California, San Francisco. (31) Jakalian, A.; Bush, B. L.; Jack, B. D.; Bayly, C. I. Fast, Efficient Generation of HighQuality Atomic Charges. AM1-BCC Model: I. Method. J. Comp. Chem. 2000, 21, 132–146.

(32) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. A Second Generation Force Field for the Simulation of Proteins, Nucleic Acids, and Organic Molecules. J. Am. Chem. Soc. 1995, 117, 5179–5197.

29



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 32

(33) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J.; Impey, R.; Klein, M. Comparison of Simple Potential Functions for Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926–935. (34) Manna, R. N.; Dybala-Defratyka, A. A Computational Study of the Dechlorination of βHexachlorocyclohexane (β-HCH) Catalyzed by the Haloalkane Dehalogenase LinB. Arch.

Biochem. Biophys. 2014, 562, 43–50. (35) Stewart, J. J. P. Optimization of Parameters for Semiempirical Methods I. Method. J.

Comp. Chem. 1989, 10, 209–220. (36 ) Sugita, Y.; Kitao, A.; Okamoto, Y. Multidimensional Replica-Exchange Method for Free-Energy Calculations. J. Chem. Phys. 2000, 113, 6042−6051. (37) Kumar, S.; Bouzida, D.; Swendsen, R. H.; Kollman, P. A.; Rosenberg, J. M. The Weighted Histogram Analysis Method for Free-Energy Calculations on Biomolecules. I. The Method. J. Comput. Chem. 1992, 13, 1011−1021. (38) Hub, J. S.; de Groot, B. L.; van der Spoel, D. g_wham—A Free Weighted Histogram Analysis Implementation Including Robust Error and Autocorrelation Estimates, J. Chem.

Theory Comput. 2010, 6, 3713–3720. (39) Jorgensen, W. L.; Tirado-Rives, J. The OPLS Potential Functions for Proteins. Energy Minimizations for Crystals of Cyclic Peptides and Crambin. J. Am. Chem. Soc. 1988, 110, 1657–1666. (40) Ruiz-Pernía, J. J.; Silla, E.; Tuñón, I.; Martí, S.; Moliner, V. Hybrid QM/MM Potentials of Mean Force with Interpolated Corrections. J. Phys. Chem. B 2004, 108, 8427−8433. (41) Ruiz-Pernía, J. J.; Silla, E.; Tuñón, I.; Martí, S. Hybrid Quantum Mechanics/Molecular Mechanics Simulations with two-dimensional Interpolated Corrections: Application to Enzymatic Processes. J. Phys. Chem. B 2006, 110, 17663−17670.

30


Page 31 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(42) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. Design of Density Functionals by Combining the Method

of

Constraint

Satisfaction

with

Parametrization

for

Thermochemistry,

Thermochemical Kinetics, and Noncovalent Interactions. J. Chem. Theory Comput. 2006, 2, 364–382. (43) Frisch, M. J.; Pople, J. A.; Binkley, J. S. Self‐Consistent Molecular Orbital Methods 25. Supplementary Functions for Gaussian Basis Sets. J. Chem. Phys. 1984, 80, 3265–3269. (44) Kaminski, G. A.; Jorgensen, W. L. A Quantum Mechanical and Molecular Mechanical Method Based on CM1A Charges: Applications to Solvent Effects on Organic Equilibria and Reactions. J. Phys. Chem. B 1988, 102, 1787–1796. (45) Field, M. J.; Albe, M.; Bret, C.; Proust-De Martin, F.; Thomas, A. The Dynamo Library for Molecular Simulations Using Hybrid Quantum/Mechanical and Molecular Mechanical Potentials. J. Comput. Chem. 2000, 21, 1088–1100. (46) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, Gaussian, Inc.: Wallingford, CT, 2009. (47) Breneman, C. M.; Wiberg, K. B. Determining Atom-Centered Monopoles from Molecular Electrostatic Potentials. The Need for High Sampling Density in Formamide Conformational Analysis. J. Comput. Chem. 1990, 11, 361-373.

31



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 32

Table of Contents graphics

32


Dehydrochlorination of Hexachlorocyclohexanes Catalyzed by the

Recommend Documents