Insights into the Catalytic Mechanism of Unsaturated Glucuronyl

Jan 10, 2017 - New Class of Glycoside Hydrolase Mechanism-Based Covalent Inhibitors: Glycosylation Transition State Conformations. Journal of the Amer...
0 downloads 5 Views 1MB Size
Subscriber access provided by University of Colorado Boulder

Article

Insights into Catalytic Mechanism of Unsaturated Glucuronyl Hydrolase of Bacillus sp. GL1 Jing Xiong, and Dingguo Xu J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.6b10501 • Publication Date (Web): 10 Jan 2017 Downloaded from http://pubs.acs.org on January 12, 2017

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 36

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

Insights into Catalytic Mechanism of Unsaturated Glucuronyl Hydrolase of Bacillus sp. GL1 Jing Xiong1 and Dingguo Xu1,2* 1

MOE Key Laboratory of Green Chemistry, College of Chemistry, Sichuan

University, Chengdu, Sichuan 610064, People’s Republic of China 2

Geonome Research Center for Biomaterials, Sichuan University, Chengdu, Sichuan

610064, People’s Republic of China

* To whom correspondence should be addressed: [email protected] (D.X). Tel: 86-28-85406156.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

Abstract Together with polysaccharide lyases, the unsaturated glucuronyl hydrolase of Bacillus sp. GL1 are responsible for the metabolism of glycosaminoglycans, which exhibits important role in various crucial physiological events. More importantly, the degradation mechanism of glycosaminoglycans often causes extracellular bacterial infection, and is thought to be one of virulence factors. We have previously studied the first degradation step catalyzed by polysaccharide lyases. In this work, we then focused the degradation of the unsaturated chondroitin disaccharide, products of chondroitin lyases. A combined quantum mechanical and molecular mechanical method was employed in all simulations. First of all, molecular dynamics simulations were performed to obtain a stable initial enzyme-substrate complex structure. Almost all interactions between substrate and enzyme were found to be related with Dglucuronic acid unit, while no recognition specificity can be found for N-acetyl-Dgalactosamine unit. Experimentally, two different pathways have been proposed based on X-ray structures and kinetic isotopic effects. In our simulation, the pathway involving formation of an epoxide intermediate should be favorable rather than the direct hydration around the vinyl ether group around carbons 4 and 5. A meta-stable oxocarbenium ion-like intermediate can be found in our simulation.

2

ACS Paragon Plus Environment

Page 2 of 36

Page 3 of 36

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

1. Introduction Glycosaminoglycans (e.g. chondroitin, hyaluronan, and heparin) are negatively charged polysaccharides, which is composed by a repeating disaccharide unit consisting of an uronic acid and an amino sugar.1, 2 They are widely distributed in the

extracellular

matrix

of

mammalian

tissues.

These

polysaccharides

play a critical role in some important physiological events, such as cell signaling, growth and differentiation. Their degradation mechanism usually represents a way to remove a barrier during extracellular bacterial infection, and thus is thought be a virulence factor.3 Generally, metabolic pathway of glycosaminoglycans (GAGs) involves two kinds of mechanisms.2 In the first stage, polysaccharides lyases (PLs) can recognize the uronic acid residue of polysaccharide, and generate oligosaccharide with unsaturated uronic acid units at the non-reducing terminus.4,

5, 6

Subsequently, the

second class of enzymes, unsaturated glucuronyl hydrolases (UGL), can act on the resulted oligosaccharides, and release the monomer units at the non-reducing terminus. The degradation of polysaccharide catalyzed by lyases via the so-called β-elimination mechanism4 has already been studied to some details by both experiments7, 8, 9 and theoretical simulations10, 11, 12, in which the product clearly contains an unsaturated uronic acid. However, the degradation mechanism catalyzed by UGL still remains debate. Considering important physiological functions of GAGs, it is interesting to note that actions of both PLs and UGL on the GAGs are through different mechanisms from endogenous mammalian enzymes in GAGs degradation.2 This could make the bacterial degradation of GAGs become an attractive target for design of some small molecule inhibitors, which could restrain bacteria growth. Design of such inhibitors relies on a thorough understanding of the mechanism underlying the recognition and degradation of GAGs. In this work, we will focus on the UGL, which is a member of the glycoside hydrolase family 88 (GH88).13 Crystal structure of UGL complexed with unsaturated chondroitin disaccharide, which is a product of chondroitin lyase,9 was determined with D88N mutant in 2006.14 The chondroitin disaccharide molecule consists of D3

ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

glucuronic acid (∆GlcA) and N-acetyl-D-galactosamine (GalNAc) units, which are connected by β-1, 3 linkage. A direct hydration mechanism, namely pathway I, has been proposed accordingly, as shown in Scheme 1A. The reaction is initiated by the protonation of the double bond between C4 and C5 in the substrate unsaturated moiety. The obtained oxo-carbenium ion like species is then attacked by an active site water molecule at the C5 with a hydrolysis of the double bond. The generated hemiketal intermediate undergoes a series of rearrangements, and finally cleaves the glycosidic bond. D88 was supposed not to directly take part in the catalytic reaction, but just serve as one of stabilization factors to whole substrate molecule. Later on, this mechanism proposal was revisited by a different group.15, 16, 17 A step-wise reaction pathway was proposed in Scheme 1B, namely pathway II. Such newly proposed mechanism is based on some experimental evidences. First of all, the exceptionally slow turnover rate of the 2-deoxy-2-fluoro substrate analogue is not consistent with the direct hydration mechanism. In addition, an attempt of designing inhibitors based on the direct hydrolysis mechanism finally failed.16 Furthermore, the kinetic isotope effects (KIE) analyses suggest a stepwise catalytic pathway with a metastable epoxide intermediate.17 Once carbon 4 is protonated, the D88 can act as the general base catalyst to activate the hydroxyl group at C2 for nucleophilic attack. Subsequently, this oxygen atom can attack at C1 to generate an epoxide intermediate with the pyranose ring opening simultaneously. This forms a ketone at C5 to stabilize the occurring positive charge. Although the presence of the epoxide intermediate seems to be more reasonable in interpreting the new experimental findings, direct observance of epoxide is unreal due to its metastable characteristics. Therefore, detailed theoretical simulations are highly required to obtain insights into the mechanism from microscopic angle. Clearly, the precise role of D88 in the hydrolysis of unsaturated uronic acid catalyzed by UGL is crucial to fully understand the catalytic mechanism. In this work, we will try to address this interesting mechanistic issue using the combined quantum mechanical and molecular mechanical method (QM/MM).18 Two possible mechanisms will be evaluated, respectively. It is believed that detailed simulations of 4

ACS Paragon Plus Environment

Page 4 of 36

Page 5 of 36

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

the mechanism can provide some useful help in further development of enzyme inhibitors. 2. Computational Details 2.1 QM/MM System Setup It has been well accepted that the QM/MM method is one of the most useful tools to combine the computational cost and accuracy in dealing with those large and complicated systems, such as enzymes.18, 19, 20 Basically, in QM/MM treatment, the whole system is divided into two parts, QM and MM regions. The QM region, which consists of those atoms participating in the reaction, is treated by means of high level quantum mechanical method. The MM region, which represents the surrounding environment of the QM region, is treated using a classical force field. In this work, the self-consistent charge-density functional tight binding (SCC-DFTB) method21, 22 was employed for the electronic structure calculations, which has been implemented into the CHARMM package23. It has been widely tested for various enzyme systems24, 25, 26, 27

, including some glycosidases.28, 29 The CHARMM all atom force field was used

to describe those atoms in the MM region.30 The initial structure in our simulation was extracted from protein data bank (PDB code: 2AHG),14 which is the X-ray structure of D88N complexed with its natural substrate of chondroitin disaccharide. First of all, we manually transformed N88 back to D88 to recover its wild type form. The atom definition and putative interactions between substrate and enzyme are given in Scheme 2. The nomenclature for enzyme was adopted using the CHARMM convention. Hydrogen atoms were added to the heavy atoms using the HBUILD module. The protonation status of those polar amino residues was determined by carefully examining their interactions with their nearby residues. In particular, the D149 is assigned to be protonated since it was thought to be the general acid catalyst to initiate the reaction via donating its proton to carbon 4. The whole system was solvated in a pre-equilibrated TIP3P31 water sphere of 25 Å radius centering at the C5 (∆GlcA) atom. This process was repeated several 5

ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 36

times with rotated water sphere to ensure uniform solvation. The solvent was relaxed by 30 ps molecular dynamics (MD) with all protein and substrate atoms fixed. Stochastic boundary condition32 was applied to reduce the computational cost. All atoms of 25 Å away from the origin will be deleted. The link atom approach was applied to describe the covalent interfaces between QM and MM regions. 2.2 QM/MM dynamics A total of 3 ns QM/MM molecular dynamics (MD) simulations were carried out to examine the stability and substrate binding of the system. The temperature was slowly heated to 300 K and system equilibrium dynamics were carried out within 300 ps. The subsequent 2.7 ns MD trajectory was saved for data analysis. The integration time step is set to be 1 fs. SHAKE algorithm33 was applied to account for all hydrogen atoms covalently connected with heavy atoms. A group-based switching scheme was used for nonbond interactions.34 The definition of a suitable QM region is critical to understand the reactive mechanism. In this work, as shown in Scheme 2, the QM region consists of the side chain groups of D88, D149, H193, Q211, a water molecule, and the entire substrate of chondroitin disaccharide, resulting in a total of 88 atoms. The total charge of the QM region is calculated to be -2. The objective in this work is to identify which mechanism is more reasonable. Two experimental proposed mechanisms were then systematically explored. Minimum energy pathways (MEPs) for both the mechanisms along the putative reaction coordinates were calculated using the adiabatic mapping approach, respectively. The MEPs are further refined using the conjugate peak refinement (CPR) approach.35 It has been suggested that CPR corrected energy activation barrier height can be sufficiently close to experimental value, especially the proton transfer reactions.36 The reaction coordinates for CPR simulation can be defined as:  = ∑   −  − 1 | /√3 ,  |

(1)

where N is the order of the points along the CPR trajectory, x(i) is the coordinate of point i, and n is the number of atoms. Similar with our previous work,37 λ was 6

ACS Paragon Plus Environment

Page 7 of 36

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

normalized throughout this work if not otherwise stated, so that 0 ≤ λ ≤ 1. The first pathway (Pathway I) follows the suggestion of Itoh et al,14 in which all of bond breaking and formation only occur around C4-C5 bond. It involves migration of the proton on D149 to C4 atom and quenching by an active site water molecule at C5 atom. Such process can generate an intermediate, namely hemi-ketal. This two-dimensional pathway is then characterized by the reaction coordinates:  =  ⋯ − ⋯ for the protonation of C4, and  = 

⋯ ! for

water quenching at C5.

The second pathway (pathway II) was suggested recently by Jongkees et al.,17 in which the contribution of D88 during the reaction should not be ignored. This mechanism was also triggered by the protonation of the double bond of C4-C5 firstly. But the participation of D88 will facilitate the formation of an epoxide ring and cleavage of C1-O5 bond. The breaking of epoxide ring could further generate an oxocarbenium ion like species. This will be followed by the quenching of water molecule at the C5 atom. Therefore, for the first step the reaction coordinates are then described as " =  ⋯ − ⋯ and  = ⋯ for the formation of carbon-oxygen bond between C1 and O2. In the subsequent step, the reaction coordinates are defined as  = ⋯ for epoxy ring opening, and # =  ⋯$ − !⋯$ for nucleophilic attacking by water molecule to complete the reaction, respectively. The two pathways are illustrated in Scheme 1A and 1B, respectively. The resulting structures based on above MEP calculations were then adopted as the initial structures for the potentials of mean force (PMFs) computations. To obtain free energy profiles, umbrella sampling method was applied to enhance sampling around the peak region by adding a biasing potential to the force field. Corresponding harmonic constraints were set to be around 50-300 kcal/(mol·Å2). A total of 100 ps constrained MD simulations were performed for each window with the first 60 ps heating to 300 K and equilibrium, and the rest of the 40 ps trajectory was saved for data analysis. Such selection could lead to more than 28 ns for pathway I 7

ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

and 65 ns for pathway II. The MD calculations were performed with a 1 fs time step and SHAKE algorithm33 for covalent bonds involving non-transferring H atoms. Finally, the weighted histogram analysis method (WHAM)38, 39 was used to obtain final PMFs for both pathways. 3. Results and Discussion 3.1 Validation of SCC-DFTB method The validity for the SCC-DFTB method has been assessed by several previous studies.40, 41 Particularly, it has been shown that the SCC-DFTB optimized geometries can be compared with those obtained at the B3LYP/6-31++G(d,p) level of theory.42 For the substrate molecule investigated in this work, we have fully optimized the structure of ∆GlcA-GalNAc using both SCC-DFTB and B3LYP/6-31++G(d,p) level of theory. Optimized structures are displayed in Figure 1 with some selected geometric parameters labeled. Clearly, the geometries obtained by both methods are quite close with typical bond distance difference smaller than 0.1 Å and angle difference less than 4 degree. Such close structure optimized by both methods can further ensure the viability of SCC-DFTB in the simulation of the unsaturated polysaccharide. The density functional theory (DFT) calculations were performed using Gaussian09 suite of program.43 3.2. Michaelis Complex To discuss the detailed catalytic process that occurs in the enzyme active site, a reasonable enzyme-substrate (ES) complex structure stays in the core position. As we know, it is not that easy to get the direct complex structure with both native enzyme and the substrate molecule via an experimental way, as well as the UGL. The QM/MM MD simulation, on the other hand, represents an alternative way to predict a reasonable ES structure. First of all, one of key issues related with the mechanistic studies is to identify the protonation state of D149. The protonated D149 requires a high pKa environment, and there is another possible catalyst candidate, H193, which also stays close to the active site water molecule. Hence, to avoid artificial effect, we then constructed a model in which D149 is assigned to be deprotonated, while H193 is doubly 8

ACS Paragon Plus Environment

Page 8 of 36

Page 9 of 36

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

protonated. One water molecule stays as a bridge to connect D149 and H193. Quite interestingly, such model is not stable at all. Even in the geometrical minimization stage, the proton on H193 is found to be easily abstracted by the active site water molecule, since one proton on water molecule migrates to D149 simultaneously. Such process finally generates a protonated D149. To further confirm the protonation status of D149 and H193, we calculated the pKa values for these two residues using PROPKA tool.44, 45 Quite interestingly, pKa(D149) = 6.75, while pKa(H193) = 1.75, which can ensure that a protonated D149 and neutral H193 without doubt. To this end, the D149 is simulated to be protonated in following mechanistic studies. The optimized structure is subject to a 3 ns QM/MM MD simulation. The whole system is maintained very well throughout the simulation, evidenced by the root-mean-square deviation (rmsd) of 0.75 ± 0.03 Å for the backbone atoms compared to the crystal structure as depicted in Figure 2. One of snapshots randomly extracted from the MD trajectory is displayed in Figure 3, and an overlay representation with the X-ray structure (2AHG) is also included for comparison. The simulated overall structure shows a good overlap with the X-ray structure. Some selected statistical averaged geometrical properties were listed in Table 1, in which data from X-ray structure is also included for comparison. One of key structural features for ES complex structure of UGL is the location of D149 and its recognition status, since it has been suggested to be the general acid/base in both mechanisms. We can see that D149 stays close to vinyl ether group of ∆GlcA with ⋯ = 3.17 ± 0.29 Å. No stable hydrogen bond formed between D149 and the active water molecule can be observed during the MD simulation. At the same time, this active site water molecule is just occasionally hydrogen bonded with H193 and Q211, respectively. Such geometric features might have two folds meaning. The first, such kind of initial structure will be not so easy to undergo the subsequent direct hydration pathway at the C4-C5 bond. The second, it can be also in agreement with experimental kinetic data show that mutants of H193A and Q211A do not have significant effects to overall catalytic activity. On the other hand, this water stays nearly perpendicular to the plane formed around C5 atom. This could ensure that 9

ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 36

once the water is activated, the hydroxyl group can easily attack the C5 atom for subsequent reaction. As suggested by experimental works, the hydration of the unsaturated chondroitin disaccharide substrate does not directly occur at the glycosidic linkage. Therefore, it would be of particular interest to examine recognition features of ∆GlcA unit. This unit is deeply embedded inside into a polar pocket formed by some active site residues like D88, Q211, R221, and W225. Significant hydrogen bond network formed between ∆GlcA and the enzyme can be found, evidenced by the distances of 2.76

±

0.12

Å

for

NE(R221)···O62(∆GlcA),

2.67

±

0.09

Å

for

NH2(R221)···O61(∆GlcA), 2.87 ± 0.16 Å for NE1(W225)···O62(∆GlcA), and 2.83 ± 0.12 Å for NE2(Q211)···O61(∆GlcA), respectively. Indeed, we can also observe similar hydrogen bond network in the X-ray structure, as shown in Figure 2 and Table 1. At the same time, the ∆GlcA unit is also recognized by D88. Throughout the simulation, the hydrogen bonds formed between D88 and O2/O3 hydroxyl groups are maintained very well, although occasional position switching between OD1 and OD2 of D88 can be observed. Details can be found in the supporting material (Figure S1). Clearly, such geometric characteristics make it be possible that D88 serves as a general base to abstract the proton of the C2 hydroxyl group. Besides, such hydrogen bonds could also prevent O3 hydroxyl group forms the hydrogen bond with D149, which can make proton migration from D149 to C4 atom more difficult. In addition, the hydrophobic stacking interactions between hydrophobic residues and sugar rings could provide further stabilization as shown in our simulation, e.g., the indole group of W42 is almost parallel to the ∆GlcA unit of the substrate. Surprisingly, the unit of GalNAc seems to be totally exposed to solvent environment, which indicates no recognition specificity can be related with this unit. For the system like UGL we studied in this work, two possible mechanisms have been proposed, direct hydration (pathway I) and formation of an epoxide intermediate (pathway II). The experimental work suggested that two mechanisms share a same initial ES complex structure. Indeed, our simulated initial ES complex can fully match structural requirement for both mechanisms. For example, appropriate 10

ACS Paragon Plus Environment

Page 11 of 36

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

location of D149 and the active site water molecule can ensure the triggering of hydrolysis of the vinyl ether group, while hydrogen bonds formed between D88 and O2/O3 hydroxyl groups of ∆GlcA also makes the formation of the epoxide intermediate become possible. Certainly, further catalytic reaction simulations are needed to understand the hydrolysis of unsaturated chondroitin disaccharide substrate. 3.2. Mechanistic Simulations In this work, we have systematically explored two mechanisms, namely direct hydration and epoxide intermediate formation, along their putative reaction coordinates, respectively. Calculated reaction pathways by SCC-DFTB/MM minimizations are given in Supporting Information (SI), with the transition states optimized by CPR method. 3.2.1. Pathway I (direct hydration) For the first pathway proposed by Itoh et al.,14 the mechanism has the name of direct hydration, since all bond formation and breaking processes just occur around C4-C5 double bond of ∆GlcA unit. In particular, the whole reaction involves the protonation of the C4-C5 double bond, accompanying with the quenching of positively charged C5 by a water molecule, leading to generation of a metastable hemi-ketal intermediate. In our simulation, it would be natural to calculate a two-dimensional (2-D) free energy surface for this process involving all significant bonds developments. Calculated PMF is displayed in Figure 4. A nearly concerted reaction pathway could be observed, which features a relatively high barrier of 35.0 kcal/mol. Snapshots for ES, transition state (TS) and enzyme-intermediate complexes (EI, hemi-ketal) are plotted in Figure 5, respectively. Selected key geometric parameters are listed in Table 2. In particular, the geometry of TS is adopted using CPR approach as we described above. Throughout the reaction, the hydrogen bond formed between the substrate and D88 is maintained very well. In addition, the C5-carboxylate group of ∆GlcA unit is also hydrogen bonded with Q211, W225 and R221, respectively. In TS, with the proton (HD2) on D149 migration to C4 (∆GlcA) atom, the distance of HD2-C4 was 11

ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

shortened to 1.18 Å from 3.09 Å in ES. At this stage, the D149 clearly serves as the general acid. At the same time, the active site water molecule becomes much closer to C5 atom, with 

⋯ !

= 2.11 Å vs. 3.50 Å in ES. Notably, a strong hydrogen bond

was then formed between D149 and the active site water molecule, judged by the distance between OD1 of D149 and H2 (water) atom calculated to be 1.54 Å. The D149 acts as the general base at this stage. Additionally, since C4 will accept another proton, the bond distance of C4-C5 is elongated from 1.35 to 1.46 Å accordingly, which indicates that the lost of a double bond. Interestingly, with formation of hemiketal intermediate (EI), we can easily observe that the hydration of C4-C5 is completed with C4 protonated and C5 atom attacked by the hydroxyl group. In hemiketal intermediate, the bond distance of C4-C5 becomes 1.52 Å, a typical distance for C-C single bond. As suggested by Itoh et al,14 an oxocarbeniun ion-like species should exist during the hydration of the substrate. However, they did not assign this species to be transition state or intermediate. Therefore another interesting aspect deserving to understand is the charge distribution on C5/O5 atom along the reaction coordinates. Based on our NBO46 analyses using B3LYP/6-31G(d,p)/MM//SCC-DFTB/MM single point calculations, the charge on C5 changes from 0.21 (ES) to 0.53 (TS), while the charge on O5 changes from -0.56 to -0.50 as well. All of these evidences strongly suggest the existence of an oxo-carbenium ion like transition state during the hydration. Apart from this, the nature of C4 atom before the reaction is predominately sp2 hybridization, which suggests a double bond: '(,= 0.7091234.5+(+ + ) *(+ 0.7051234.85 () . With proton on D149 transferring to C4 atom, this atom becomes sp3 hybridization at TS: '(,= 0.7161234.):(+ + 0.698123$.8