ONIOM (DFT:MM) Study of the Catalytic Mechanism of myo-Inositol

Dec 26, 2012 - Our calculations show that the three-metal mechanism is energetically unfavorable; the initial inline attack of a hydroxide ion on the ...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF ALABAMA BIRMINGHAM

Article

ONIOM(DFT:MM) Study of the Catalytic Mechanism of myo-Inositol Monophosphatase: Essential Role of Water in Enzyme Catalysis in the Two-Metal Mechanism Xiaoqing Wang, and Hajime Hirao J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/jp312483n • Publication Date (Web): 26 Dec 2012 Downloaded from http://pubs.acs.org on January 6, 2013

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 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

The Journal of Physical Chemistry

ONIOM(DFT:MM) Study of the Catalytic Mechanism of myoInositol Monophosphatase: Essential Role of Water in Enzyme Catalysis in the Two-Metal Mechanism Xiaoqing Wang and Hajime Hirao*,† † Division of Chemistry and Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, 21 Nanyang Link, Singapore 637371

* Corresponding author. E-mail: [email protected]

Abstract myo-Inositol monophosphatase (IMPase), a putative target of lithium therapy for bipolar disorder, is an enzyme that catalyzes the hydrolysis of myo-inositol-1-phosphate (Ins(1)P) into myo-inositol (MI) and inorganic phosphate. It is known that either two or three Mg2+ ions are used as cofactors in IMPase catalysis; however, the detailed catalytic mechanism and the specific number of Mg2+ ions required have long remained obscure. To obtain a clearer view of the IMPase reaction, we undertook extensive ONIOM hybrid quantum mechanics and molecular mechanics (QM/MM) calculations, to evaluate the reaction with either three or two Mg2+ ions. Our calculations show that the three-metal mechanism is energetically unfavorable; the initial inline attack of a hydroxide ion on the Ins(1)P substrate markedly destabilized the system without producing any stable transition state or intermediate. By contrast, for the two-metal mechanism, a favorable pathway was obtained from QM/MM calculations. In our proposed two-metal mechanism, the phosphoryl oxygen of the substrate acts as an acid–base catalyst, activating a water molecule in the first step, and the resultant hydroxide ion attacks the substrate in an inline fashion. A second water molecule, bound to a Mg2+ ion, was found to play an essential role in the final proton-transfer step that leads to the formation of an MI product; this is achieved by lowering the energy barrier by 2.5 kcal/mol compared with the barrier for the mechanism that does not use this water molecule. Our results should advance our understanding of the IMPase mechanism, and this could have profound implications for the treatment of disease in the central nervous system.

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

1. Introduction myo-Inositol (MI) is the most abundant of the nine stereoisomers of cyclohexanehexol,1-4 and relatively high concentrations are present in the central nervous system (CNS).5 MI is a precursor of phosphatidylinositol (PtdIns) as well as its mono-, di-, and triphosphates (PtdInsP or phosphoinositides) that play pivotal roles in PtdIns signaling and mediate a range of other essential cellular processes.6-11 MI also forms the structural backbone of inositol phosphates (InsP) and their pyrophosphate derivatives; the prominent roles of InsP and derivatives in signal transduction and as cell controllers have been well documented.5,12-18 For example, the extensively studied inositol 1,4,5-triphosphate (Ins(1,4,5)P3) induces the release of intracellular calcium.19-23 Given that MI is involved in a number of signaling processes and that changes in its level may alter the concentration of many other inositol derivatives, it can be predicted that without strict maintenance of the MI level, many intracellular processes will be altered, which might then trigger CNS pathology. The kidney is the major organ in which MI is catabolized5 and the first committed step is catalyzed by myo-inositol oxygenase (MIOX), a dinuclear iron enzyme that cleaves one of the C–C bonds of MI.24-35 The mechanism of the MIOX reaction has recently been investigated computationally by one of the present authors.30-31 Regarding the replenishment of MI, while there is about 1 g of dietary intake of MI per day, it is also produced biosynthetically in several organs including the kidney and brain.36 In the brain, de novo synthesis, recycling, and transporter uptake are the major sources of cellular MI,5,37 of which the former two involve MI biosynthesis. In both of the two biosynthetic pathways, myo-inositol monophosphatase (IMPase) plays a key role in the catalysis of the reaction step in which myo-inositol-1phosphate (Ins(1)P) is converted into MI and inorganic phosphate (Scheme 1).38-44 In de novo biosynthesis, the first step is the conversion of glucose 6-phosphate (G-6-P) into Ins(1)P by myo-inositol-1-phosphate synthase (mIPS),45 and Ins(1)P is then converted into MI and

2

ACS Paragon Plus Environment

Page 2 of 29

Page 3 of 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

The Journal of Physical Chemistry

inorganic phosphate by IMPase (Scheme 1). On the other hand, the production of MI by recycling occurs within the PtdIns signaling pathway wherein MI is converted to PtdIns and various PtdInsP and InsP species, and the Ins(1)P is finally converted back to MI by IMPase. As MI has poor permeability across the blood–brain barrier,46 maintenance of its concentration in the brain is largely dependent on these local biosynthesic pathways. Understanding the molecular mechanism of IMPase activity is particularly important in relation to the so-called “inositol depletion hypothesis” that posits that lithium therapy in the treatment of bipolar disorder47-50 is associated with depletion of the MI level,51-52 and one of the most likely protein targets for the lithium inhibition is IMPase.53-61 The inositol depletion hypothesis was formulated based on the experimental finding that injection of lithium chloride resulted in a decrease in the MI level in the cerebral cortex of rats as well as an increase in Ins(1)P.62-64 Depletion of the MI level was also observed in the right frontal lobe of human patients after lithium administration.65 In fact, it was experimentally verified that Li+ ions inhibit IMPase.66-67 Further, it was shown that Li+ ions act as uncompetitive inhibitors of the binding of Ins(1)P to IMPase, most probably by binding to a postcatalytic complex.68-71 However, despite its having been the mainstay in the treatment of bipolar disorder, lithium therapy is known to induce adverse side effects.43,72 For the development of less toxic drugs, an improved understanding of the mechanism of MI biosynthesis is essential.

Scheme 1. The IMPase-catalyzed reaction

IMPase requires magnesium ions (Mg2+) as a cofactor38-39,68-69,73-74 and the maximum catalytic activity is seen at a Mg2+ ion concentration of 3–4 mM,68 while higher

3

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

concentrations result in an uncompetitive inhibition of the enzyme.68 Several mechanisms have been proposed to explain the IMPase reaction, such as the ping-pong mechanism that proceeds through a phosphoenzyme intermediate,75-76 the non-inline displacement mechanism that is accompanied by pseudorotation,77-78 and the inline or SN2 displacement that results in inversion of the configuration around phosphorous.41,79-81 The current consensus is that the reaction proceeds via inline displacement, as this mechanism is consistent with the experimental configuration analysis using oxygen isotopes.82-83 This inline mechanism involves the initial activation of a water molecule by a nearby Mg2+ ion at the metal-binding site 1 (M1 in Figure 1) to produce a hydroxide ion. The hydroxide then attacks the phosphorous in an SN2 fashion to release the hydrolysis products, MI, and inorganic phosphate.

Figure 1. Putative active-site structure of IMPase in complex with Ins(1)P and three Mg2+ ions. The structure was prepared from atomic coordinates derived by X-ray crystallography of bovine IMPase (PDB code 2BJI). Another PDB structure of human IMPase (PDB code 1IMB) was superimposed on the first and the coordinates for Ins(1)P were extracted. Subsequently, energy minimization was performed on Ins(1)P, the associated water molecules,

4

ACS Paragon Plus Environment

Page 4 of 29

Page 5 of 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

The Journal of Physical Chemistry

and the hydrogen atoms of the protein.

X-ray crystallography has been used extensively to determine the atomic coordinates of IMPase in the last two decades.79-81,84-89 Such studies revealed that there are three potential metal-binding sites (M1–M3 in Figure 1) within the active site; however, the number of sites actually occupied by Mg2+ ions during catalysis is still under debate. Nevertheless, the accumulated knowledge on this subject suggests that Mg2+ ions bind weakly at M3 compared with M1 and M2. Currently, a major question concerning the reaction mechanism is whether Mg2+ ions occupy two sites (M1 and M2) or three sites (M1, M2, and M3). If it is assumed that the three binding sites are fully occupied, then the structure shown in Figure 1 should accurately represent the initial ternary complex. Bone et al. determined crystal structures of human IMPase in complex with a single Gd3+ ion and Ins(1)P (PDB codes 1IMA and 1IMB),80 and in these complexes Gd3+ ions occupied only the M1 site. Bone et al. also determined the structures of human IMPase in complex with Mn2+ ions (without substrate, PDB code 1IMC) and in complex with Mn2+ ions and inorganic phosphate (PDB code 1IMD).81 Interestingly, while there were three Mn2+ ions in the 1IMC structure, the Mn2+ ion at M3 in the 1IMD structure was displaced upon addition of phosphate, leading to the proposition that IMPase uses only two Mg2+ ions for catalysis. In another study, Ganzhorn et al. used Ca2+ ions to delineate how metal ions bind to IMPase, and they found three Ca2+ ions in their X-ray crystal structure of human IMPase in complex with D-Ins(1)P (PDB code 1AWB).84 By contrast, only two Ca2+ ions (at M1 and M2) were observed in the X-ray structure of archaeal IMPase solved by Johnson et al. (PDB code 1G0H).85 Since the structures of the human and archaeal IMPase enzymes are very similar, the discrepancy in the number of Ca2+ ions might be due to the different concentrations of Ca2+ ions present during protein crystallization (200 mM in 1AWB vs. 10 mM in 1G0H). Arai et al. determined the

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

crystal structure of human IMPase 2 (IMPA2) in complex with Ca2+ ions and phosphate (PDB code 2CZI).87 IMPA2 is similar to the above-described human IMPase enzyme (IMPA1), sharing 54% sequence identity, and in the 2CZI structure, two Ca2+ ions were observed at M1 and M2. Gill et al. solved a crystal structure of bovine IMPase in complex with Mg2+ ions, but without a substrate, and observed three Mg2+ ions in the active site.86 Very recently, Singh et al. solved crystal structures of human and mouse IMPase in complex with inorganic phosphate, also finding three Mg2+ ions. Despite these many X-ray studies, no unified view on the binding of Mg2+ ions during catalysis has emerged. This is mainly because the determination of the structure of a reactive ternary complex comprising IMPase, Ins(1)P, and Mg2+ ions is very difficult. Therefore, in all the published structures, IMPase was crystallized either without substrate, with substrate but with alternative metal ions, or with inorganic phosphate in place of Ins(1)P. Besides, the number of metal ions found in the active site of a crystal structure appears to depend on the concentration of metal ions present in the crystallization process. Hence, although there is no doubt that the X-ray studies have provided invaluable insights into the structural details of IMPase and its active site, it remains to be clarified specifically how many Mg2+ ions are required for catalysis. As it is evident that the catalytic mechanism of IMPase is very important in understanding the role of MI in CNS pathology, there has been an extensive research effort into this topic; however, very few computational studies of IMPase have been reported. Pollack et al. performed molecular mechanical (MM) energy minimization and short (11 ps) molecular dynamics (MD) simulation on several models of IMPase. Based on computationally predicted structural features, they concluded that the two-metal mechanism is more plausible than the three-metal mechanism.41 Wilkie et al. also ran MM energy minimization and MD simulation on IMPase, and concluded that the reaction follows a noninline displacement mechanism that is accompanied by pseudorotation.78 A more recent MM

6

ACS Paragon Plus Environment

Page 6 of 29

Page 7 of 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

The Journal of Physical Chemistry

study by the same group amended this conclusion and also suggested that a water molecule bound to the Mg2+ ion at M2 serves as a proton donor.90 Fujita et al. performed docking simulation on human IMPA2 in complex with Mg2+ ions and Ins(1)P, and proposed that the dephosphorylation reaction proceeds with three Mg2+ ions.91 Dudev and Lim computationally studied the Mg2+ → Li+ exchange processes in simplified IMPase models, employing density functional theory (DFT) calculations and solvation free-energy calculations based on Poisson’s equation.92 Very recently, Lu et al. applied ONIOM(DFT:MM) to IMPase in the absence of the Ins(1)P substrate,93 showing that with three Mg2+ ions, the bridging water molecule between M1 and M3 spontaneously donated a proton to the nearby Thr95 that is hydrogen bonded to Asp47, thereby generating a nucleophilic hydroxide ion, whereas such spontaneous deprotonation of water was not observed when Mg2+ was absent at M3. These results led them to conclude that IMPase adopts the three-metal mechanism for catalysis. Despite these computational analyses, there is no DFT or DFT/MM study that directly probes the reaction pathway of IMPase, so there remains no definitive answer to the number of Mg2+ ions involved. In view of the difficulty of obtaining this data from X-ray crystallography alone, computational studies should help fill this knowledge gap. In this report we examine the catalytic mechanism of IMPase with a state-of-the-art quantum mechanics and molecular mechanics (QM/MM) hybrid method defined within the two-layer ONIOM(DFT:MM) framework. In this approach, DFT and MM are used together to describe the reaction mechanism within an enzyme that consists of thousands of atoms.

2. Methods The X-ray structure of dimeric bovine IMPase (PDB code 2BJI) was taken from the Protein Data Bank (PDB) website,86 and the protonated states of histidine residues were determined by PROPKA94-95 analysis and visual inspection to be as follows: HIE65, HID100,

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

HIP150, HID188, and HIP217. The notations HID and HIE mean that the histidine takes the neutral form with Nδ1 and Nε2 atoms protonated, respectively, whereas in HIP both of these nitrogen atoms are protonated. As the substrate was missing in the 2BJI structure, atomic coordinates of the L-Ins(1)P substrate were taken from the crystal structure of human IMPase (PDB code 1IMB) after superimposing the two PDB structures.80 IMPase has been shown to hydrolyze both D- and Lenantiomers of Ins(1)P.39,80 The structures of human and bovine IMPase are quite similar to each other, with the root-mean-square deviation (RMSD) of backbone atoms being only ~0.6 Å. The hydrogen atoms of the Ins(1)P substrate were added using the AmberTools Reduce program,96 and the total charge of this substrate was –2. In the active sites of the 2BJI structure (chains A and B), there were eleven crystallographic water molecules that had steric overlap with the inserted substrate (Figures S1 and S2). These water molecules were moved out of the active site to the solvent layer that was later added for MM-based energy minimization calculation (see Figure S2 in the Supporting Information). Some of the MM parameters of Ins(1)P were determined by the combined use of Gaussian 0997 and AmberTools.96 The geometry of Ins(1)P was optimized at the B3LYP/631G* level of theory with Gaussian 09. We initially optimized the geometry of the deprotonated dianionic form of Ins(1)P. However, this gas-phase optimization resulted in intramolecular proton transfer from a hydroxyl group on the cyclohexane ring to a phosphate oxygen atom. The atomic charges obtained for this state were considered to be unreasonable and we therefore optimized a monoprotonated, monoanionic form of Ins(1)P. Without changing the optimized Ins(1)P geometry, we removed a proton from the phosphate, and then derived the RESP atomic charges at the HF/6-31G* level. The specific charges used are summarized in the Supporting Information (Table S1). Atom types from the Amber ff99SB force field were assigned to the substrate (Scheme 2).98 Additional angle parameters were

8

ACS Paragon Plus Environment

Page 8 of 29

Page 9 of 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

The Journal of Physical Chemistry

required for atom type HC, and literature values were adopted (Table S2).99

Scheme 2. Assigned atom types for the Ins(1)P substrate.

After building the model, the protein containing three Mg2+ ions per active site was immersed in a TIP3P100 solvent box of 85 × 115 × 85 Å, and the total system was neutralized by adding eight Na+ ions (counterions) in the solvent layer, resulting in 71739 total atoms in the system. The system was then subjected to stepwise energy minimization using the NAMD 2.8 package101 with the Amber99 force field. In the first energy minimization, the coordinates of amino acid residues, crystallographic water molecules, and the three Mg2+ ions were held fixed, while the solvent water molecules, the substrate, and all hydrogen atoms were allowed to move. The system was subjected to energy minimization for 10000 steps using the steepestdescent algorithm. In the second step, the coordinates of heavy atoms in the protein were held fixed while other constraints were removed in order to run energy minimization for an additional 10000 steps. Reaction mechanisms were investigated applying the two-layer ONIOM(DFT:MM)

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

QM/MM approach,102-103 and the final structure of the above NAMD energy minimization (Geom-0) was used as a starting geometry for this purpose. To reduce the computational cost, only chain A was used for the subsequent QM/MM calculations and the solvent layer was removed, and both the three-metal model (Model-I) and two-metal model (Model-II) were constructed. In Model-I, M1–M3 were fully occupied by Mg2+ ions, whereas in Model-II, the Mg2+ at M3 was deleted. In this paper the Mg2+ ions bound to these sites will be referred to as Mg1, Mg2, and Mg3 respectively, and the high-level QM layer for the ONIOM calculations is shown in Scheme 3. In Model-I, the QM layer comprised the following moieties: Ins(1)P; three Mg2+ ions; the side chains of Glu70, Asp90, Asp220, and Asp45; the backbone carbonyl group of Ile92, entire Asp93 and Gly94; Thr95 except the backbone carbonyl group; water 33 (wat33), wat34, wat66, wat82, wat86, and wat184. In Model-II, the QM layer was essentially the same as for Model-I except that Mg3 was absent. The total numbers of atoms in Model-I and Model-II were 5009 and 5008, respectively, and the numbers of atoms in the QM layers of Model-I and Model-II were 114 and 113, respectively. The atoms within 12 Å of the phosphorous atom of the phosphate in Geom-0 were relaxed in all geometry optimization calculations, while all other atoms were frozen. The ONIOM(B3LYP:AMBER) electronicembedding (EE) scheme was used, whereby the influence of MM charges can be included in the one-electron Hamiltonian of the QM calculation, and the polarization of the QM wave function is taken into account. In the other ONIOM approach, i.e., the mechanical embedding (ME) scheme, the highly charged active site is essentially treated as being in the gas phase, which will necessarily overestimate the Lewis acidity of the Mg2+ ions and the stability of the neutral form of a carboxyl group. We performed frequency calculations on the optimized geometries at the same level with outer atoms frozen, to obtain approximate zero-point energies (ZPEs). Despite the high computational demand of the ONIOM geometry optimizations, the basis set (6-31G* or B1)

10

ACS Paragon Plus Environment

Page 10 of 29

Page 11 of 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

The Journal of Physical Chemistry

used was not large enough to gain reliable information about energetics; therefore, a larger basis set (6-311+G(d,p) or B2) was used for energy evaluation to more reliably compare the stabilities of different states. Furthermore a recent benchmark study done by Ribeiro et al.104 demonstrated that the MPWB1K functional105 performs better than B3LYP in the evaluation of barrier heights for the hydrolysis of phosphodiester bonds. Therefore, we reevaluated the energies at the ONIOM(MPWB1K/B2:AMBER)-EE level by performing single-point calculations and it is these energies that we have used for the discussion, although ONIOM(B3LYP/B2:AMBER)-EE calculations were also performed for comparison. A few Amber parameters had to be appended to run ONIOM calculations. The appended parameters are summarized in Table S3. All ONIOM calculations were carried out using the Gaussian 09 package. Chimera and VMD software tools were used for visualization of proteins.106-107

Scheme 3. QM model used for the QM/MM optimization of Model-I. The wavy lines indicate a QM/MM boundary.

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

The high computational demand of the above-mentioned ONIOM calculations did not allow us to examine the dependence of the computational results on the initial protein geometry, which would be done by calculating reaction energy profiles for multiple initial geometries produced from MD simulations. In this study we used an X-ray geometry that was subjected only to the above-described partial MM energy minimization to get Geom-0, prior to ONIOM calculations. Nevertheless, we performed MD simulations on solvated Model-I and Model-II in dimeric form for 30 ns with a step time of 1 fs. The changes in RMSD values of backbone atoms with respect to the original crystal structure during the MD simulation are shown in the Supporting Information (Figure S5). Although the active site of the enzyme is situated close to the solvent layer, a relatively small change in the number of water molecules in the active site, particularly around w86 and the phosphate moiety of Ins(1)P, was observed during the MD simulations (Figure S6).

3. Results and Discussion 3.1. Structure of initial ternary complex. The ONIOM-optimized geometries of the ternary reactant complex (RC) of Model-I and Model-II are shown in Figure 2. In the structure of Model-I (Figure 2a), the nucleophilic water (w86, see also Scheme 3) is seen to bridge Mg1 and Mg3. The water was also hydrogen bonded to the hydroxyl group of Thr95. Lu et al. have recently reported that in the absence of Ins(1)P, w86 spontaneously underwent deprotonation, donating its proton to the Thr95/Asp47 dyad.93 However, this spontaneous proton transfer was not observed in our calculation with Ins(1)P present, suggesting that the influence of Ins(1)P on the Lewis acidity of the Mg2+ ions is significant. Thus, the coordination of the phosphonyl oxygen atoms of Ins(1)P to Mg1 and Mg3 results in electron donation to these Mg2+ ions, thereby attenuating their Lewis acidity, such that a spontaneous proton transfer from w86 to Thr95 is rendered more difficult. The RC

12

ACS Paragon Plus Environment

Page 12 of 29

Page 13 of 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

The Journal of Physical Chemistry

geometry of Model-II (Figure 2b) was similar to that of Model-I, except that in Model-II there is no Mg3, and thus w86 interacted with Mg1, Thr95, and a phosphate oxygen of Ins(1)P. Our ONIOM calculations demonstrated that spontaneous proton transfer from w86 to Thr95/Asp47 does not occur in either Model-I or Model-II, and so it was not possible at this stage to establish the number of Mg2+ ions required for the catalysis. To answer this question, it was clear that reaction energy analyses were required.

Figure 2. The optimized geometries of RC: (a) Model-I; (b) Model-II.

3.2. Three-metal mechanism. Although spontaneous proton transfer from w86 to Thr95 was not observed in the RC geometry of Model-I (Figure 2a), hydroxide formation and nucleophilic attack on the phosphorous of Ins(1)P might be expected, and so we assessed this possibility with two sets of relaxed scan calculations. In the first set of scan calculations, we gradually decreased the distance between the hydrogen of wat86 and the hydroxyl oxygen of Thr86, and at each scan step, we performed geometry optimization with this H–O distance frozen. Figure 3a plots the 13

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

relative energy profiles (with respect to the energy of Model-I RC) obtained from these scans. It can be seen that the energy initially increases with decreasing H–O distance, and the barrier for the proton transfer step is about 7 kcal/mol. Thus an intermediate, which is less stable than RC, is formed after proton transfer to Thr95. Starting from this local minimum structure at r(H–O) ~1.05 Å, we performed a second set of relaxed scan calculations for the next step, i.e., SN2 reaction between the hydroxide ion and Ins(1)P. For these calculations, we gradually decreased the distance between the hydroxide oxygen and the Ins(1)P phosphorous. As shown in Figure 3b, this relaxed scan calculation resulted in significant energy growth without providing any local energy minimum state. In fact at r(P–O) = 1.7 Å, the energy was ~45 kcal/mol higher than that of RC itself. Accordingly, our ONIOM calculations did not support the three-metal mechanism, which can be explained by assuming a diminished nucleophilicity of the hydroxide in Model-I. Thus, even though the presence of Mg2+ ions may lower the pKa of w86, the resultant hydroxide ion will also be stabilized by the two Mg2+ ions, thereby attenuating its nucleophilicity.

14

ACS Paragon Plus Environment

Page 14 of 29

Page 15 of 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

The Journal of Physical Chemistry

Figure 3. ONIOM(B3LYP/B1:AMBER)-EE energies obtained from relaxed scan calculations for (a) proton transfer from w86 to Thr95, and (b) hydroxide attack on the phosphorous of Ins(1)P.

3.3. Two-metal mechanism. As shown in Scheme 4, in the two-metal mechanism starting from RC in Figure 2b, there are three possible bases to which wat86 could potentially donate a proton to produce hydroxide. Paths A, B, and C (Scheme 4) differ in the initial hydroxide formation step, in that they require proton transfer from wat86 to the side chain oxygen atom of Glu70, a phosphoryl oxygen atom, and the side chain oxygen atom of Thr95, respectively.

15

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

Scheme 4. Three possible pathways to hydroxide formation.

All pathways subsequently follow an SN2-like backside attack of a nucleophilic hydroxide ion on the Ins(1)P phosphorous. Unlike Model-I, ONIOM calculations on the reactions of Model-II produced a number of intermediates and transition states. The proton transfer in Scheme 4 did not lead to any minimum energy structure, but instead it was found that the proton transfer and the SN2 reaction are synchronous processes. Figure 4 summarizes the calculated energy profiles for the three pathways obtained with the MPWB1K and B3LYP functionals. Although it is often postulated that Thr95 acts as a base in the hydroxide formation of IMPase, our ONIOM calculation suggests that Path C can be ruled out, as its first energy barrier is higher than 45 kcal mol–1. Therefore, in the following descriptions we will deal only with Paths A and B. During the SN2 reactions the phosphate ester bond undergoes inversion, generating an MI oxyanion and a phosphate monoanion on Path B, or a protonated Glu70 and a phosphate dianion on Path A. Figure 4 shows that the energy barrier for Path A (A_TS1) is 4 kcal/mol higher than for Path B (B_TS1), suggesting that Path B is more favorable in the early stage of the reaction. This indicates that the phosphoryl oxygen of 16

ACS Paragon Plus Environment

Page 16 of 29

Page 17 of 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

The Journal of Physical Chemistry

Ins(1)P will act as a base to generate a nucleophilic hydroxide ion from wat86. The resultant intermediate on Path A (A_Int1) was slightly more stable than that on Path B (B_Int) by 1.1 kcal/mol. However, calculations showed that B_Int can easily be converted to A_Int1 via proton transfer from phosphate to Glu70. Thus, the initial SN2 reaction proceeds via B_TS1, leading to the formation of A_Int1.

Figure 4. ONIOM-calculated relative energy profiles. The two values for each state were obtained at the ONIOM(MPWB1K/B2:AMBER)//ONIOM(B3LYP/B1:AMBER)-EE and ONIOM(B3LYP/B2:AMBER)//ONIOM(B3LYP/B1:AMBER)-EE levels, respectively. The blue, black, and red lines indicate the profiles for Paths A, B, and C, respectively. The respective energies, which include the ZPE effects, are given in kcal mol–1.

Mg2+-dependent hydrolytic reactions of phosphate-containing molecules have been extensively studied using computational approaches for many other proteins, including cAMP-dependent protein kinase (PKA),108-115 myosin,116-117 F1-ATPase,118-119 Ras-GAP,120-122 and HIV-1 integrase.123 For the PKA-mediated reaction, several such studies aimed to identify the acid–base catalyst that deprotonates and activates the hydroxyl group of a nucleophilic

17

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

serine in the substrate. Hart et al.108 and Hutter and Helms109 performed semiempirical QM/MM calculations and suggested that a phosphoryl oxygen, and not Asp166, acts as the base. By contrast, Valiev et al. used B3LYP DFT to examine the process in which a proton transfer from a nucleophilic serine toward nearby Asp166 occurs,112 and the barrier for the nucleophilic attack was calculated to be only 11 kcal/mol. Díaz and Field used B3LYP calculations to compare two mechanisms in which the serine donates a proton to either Asp166 or a phosphoryl oxygen atom of adenosine triphosphate (ATP), and found that the former is much more favorable.113 Cheng et al. performed B3LYP(6-31+G*)/MM calculations on PKA, and demonstrated that the Asp166-base mechanism has a barrier height of 10.0–14.3 kcal/mol. Thus, for PKA, there seems to be an emerging consensus toward using Asp166 as the base. This mechanism is analogous to Path A in our IMPase Model-II, while the mechanism that utilized phosphoryl oxygen to activate serine corresponds to Path B (Scheme 4). Interestingly, Path B was more favorable than Path A in the case of IMPase Model-II, which might be because the phosphoryl oxygen is bound to Mg2+ in PKA, so reducing its basicity.113 In IMPase Model-II, the phosphoryl oxygen that acts as a base does not interact with Mg2+, which appears to enable more efficient proton transfer. It should be noted that for the IMPase reaction, either Glu70 or Thr95 is proposed to act as the base accepting the proton from w86.41,80 Interestingly, in the IMPase Model-II reaction (Figure 4), the Glu70-base mechanism (Path A) and the Thr95-base mechanism (Path C) were less favorable than Path B, in which the phosphoryl oxygen activates w86. To generate the final MI product, the second step involves the transfer of a proton to the MI oxyanion, and there are two possibilities for this transfer. One is a direct proton transfer from phosphate in B_Int to inositol oxyanion, and this is characterized as B_Int → B_TS2 → B_Prod in Figure 4. The other pathway involves water-mediated proton transfer to yield an MI product complex (i.e., A_Int1 → A_TS2 → A_Int2 → A_TS3 → A_Prod in Figure 4), in

18

ACS Paragon Plus Environment

Page 18 of 29

Page 19 of 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

The Journal of Physical Chemistry

which w184 on Mg2 (see Figure 2b) is positioned between the proton on Glu70 in A_Int1 and the anionic oxygen of MI, and so facilitates proton transfer between these two moieties. Interestingly, our ONIOM calculations show that wat184 acts essentially as a catalyst, lowering the energy barrier (26.0 kcal/mol) for proton transfer by 2.5 kcal mol–1 (A_TS2 in Figure 4), compared with that for the mechanism in the absence of the water molecule (B_TS2 in Figure 4). In the latter mechanism, Path B, proton transfer to the inositol oxyanion occurs directly from the phosphate moiety, without the aid of wat184, and this process has to pass a higher barrier (28.5 kcal mol–1) to complete the reaction. Indeed, B_Prod could be converted readily to the more stable A_Prod by minor rearrangements of the hydrogen bond network (Figures 4 and S2). Overall, the reaction energy profile (Figure 4) suggests that the reaction starts with proton transfer from wat86 to a phosphonyl oxygen atom of Ins(1)P, followed by proton transfer to a carboxylate group of Glu70 and then to the MI oxyanion via a water molecule (wat184). This important role of wat184 in the catalysis was previously proposed by Miller et al.90 and Gill et al.86 The most favorable pathway in Figure 4 involves a barrier of just 26 kcal/mol, which does not seem high enough to exclude the two-metal mechanism. Moreover, we have demonstrated above that no favorable pathway could be described for the three-metal mechanism. Therefore, we conclude that the two-metal mechanism is the most likely pathway for IMPase-mediated hydrolysis. The ONIOM-optimized geometries of key species appearing during this preferred pathway (i.e., B_RC, B_TS1, B_Int, A_Int1, A_TS2, A_Int2, A_TS3, and A_Prod) are shown in Figure 5. The reaction starts with proton transfer from wat86 to phosphate (Path B), generating B_RC (Figure 5a). In this state, the distance between the proton of wat86 and the phosphate oxygen atom is 1.74 Å, whereas the oxygen of wat86 is distant from the phosphorus atom (3.09 Å). Subsequently, the hydroxide ion that originates from wat86 attacks the phosphorus atom for inline displacement. At the transition state for this step (B_TS1), the

19

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

distance between the hydroxide oxygen and the phosphorus decreases to 2.12 Å (Figure 5b). Concurrently, the P–O distance for the oxygen atom on the other side increases from 1.67 Å in B_RC to 1.76 Å in B_TS1, indicating that an SN2 reaction is indeed taking place. During these processes, the Mg1 and phosphoryl oxygen (Figure 1) activate the water molecule and stabilize the resultant nucleophilic hydroxide ion. Furthermore, as the reaction proceeds via Path B in the initial stages, the phosphoryl oxygen of Ins(1)P can also be viewed as activating wat84, by assisting in hydroxide ion formation.

Figure 5. The optimized geometries of key species in the IMPase-catalyzed reaction. Bond distances are given in Å.

Upon completion of the SN2 displacement, an intermediate B_Int is formed (Figure 5c) in which the distance between the oxygen of the hydroxide ion and the phosphorus atom (1.67 Å) is much less than in B_RC (3.09 Å). At the same time, the distance between the phosphorus atom of phosphate and the oxygen of the MI oxyanion is much greater (3.04 Å)

20

ACS Paragon Plus Environment

Page 20 of 29

Page 21 of 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

The Journal of Physical Chemistry

than in the B_RC state (1.67 Å). At this stage, the Mg1 still interacts with the oxygen of inorganic phosphate and the side chain of Glu70. It appears that the Mg2 stabilizes the increased negative charge on the MI oxyanion. In the next step, proton transfer occurs from phosphoryl oxygen to carboxylate oxygen of Glu70. The proton–oxygen distance decreases from 1.63 Å in B_Int to 1.02 Å in A_Int. During this process, some minor states are involved that are associated with slight geometric changes of wat184. The optimized geometries of all these states are shown in Figure S2. In the A_Int state, the distance between the proton from wat86 (say proton A) and the oxygen of wat184 is 2.90 Å, while the distance between the proton of wat184 (say proton B) and the oxygen of the MI oxyanion is 1.96 Å (Figure 5d). The final step involves proton transfer to the MI oxyanion through wat184. The first transition state (A_TS2) is associated with rotation of the side chain of Glu70, to a position where the distance between proton A and the oxygen of wat184 is 2.04 Å (Figure 3e). Subsequently, the reaction passes through an intermediate (A_Int2) and another transition state (A_TS3) via the geometries shown in Figures 5f and 5g. This results in the transfer of proton B to the MI oxyanion and a simultaneous transfer of proton A from Glu70 to wat184. Following this wat184-mediated proton transfer, a product complex B_Prod is generated, which would easily be converted to more stable A_Prod (Figure 3f). During the protontransfer step, the Mg2+ ion at Mg2 stabilizes the MI oxyanion as well as wat184. Figure 6 displays how the distance between Mg2 and the oxygen of w184 changes with simulation time. Consistent with the result presented by Lu et al.,93 our 30 ns MD simulation showed that w184 is bound to Mg2 during the simulation of Model-I (Figure 6a). The separation between O(w184) and Mg2 was 2.06 ± 0.08 Å (Chain A) and 2.05 ± 0.08 Å (Chain B). Furthermore, the simulation of Model-II (Figure 6b) also showed that w184 keeps its coordination to Mg2 during the simulation, with the distance between the w184 oxygen and Mg2 being 2.02 ± 0.06

21

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

Å for chain A and 2.02 ± 0.07 Å for chain B. This reinforces the hypothesis that w184 is critically involved in the hydrolysis.

Figure 6. Changes in the distance between Mg2 and the oxygen of w184 with simulation time: (a) Model-I; (b) Model-II.

4. Conclusion We have applied ONIOM(B3LYP:AMBER) calculations to the long-standing issue regarding the mechanism of IMPase catalysis. In this study, we examined the three-metal and two-metal reaction mechanisms, in which three and two Mg2+ ions are present in the metalbinding sites of IMPase, respectively. Our ONIOM study disfavored the three-metal mechanism, for which no reasonably stable transition state or intermediate was obtained for the SN2 reaction between a hydroxide ion and Ins(1)P. By contrast, the two-metal mechanism was found to have pathways leading to the hydrolyzed products. Within the two-metal mechanism, several mechanisms differ in how initial proton transfer occurs to produce reactive hydroxide. Of all the possible mechanisms, it was shown that the most favorable

22

ACS Paragon Plus Environment

Page 22 of 29

Page 23 of 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

The Journal of Physical Chemistry

initial proton transfer occurs from wat86 to phosphoryl oxygen, which triggers an SN2 attack of the resultant hydroxide on the phosphorous of Ins(1)P. In the final step, a proton is relayed through wat184 to the MI oxyanion to give a final MI product. Hence, based on the computational results, we propose that IMPase adopts the two-metal mechanism in which wat184 plays an essential catalytic role in donating proton to the MI oxyanion. This is the first study that investigated the reaction mechanism of IMPase-catalyzed hydrolysis using a QM/MM method.

Supporting Information Complete ref 97, procedure for MD simulation, appended MM parameters, raw energy data, comparison of PDB structures, water molecules in the active site, results of MD simulation, ONIOM-optimized geometries of other minor states, and XYZ coordinates of QM atoms in optimized geometries. This material is available free of charge via the Internet at http://pubs.acs.org.

Acknowledgment H.H. thanks the financial support from a Nanyang Assistant Professorship and the Takeda Science Foundation. The computer resources at the High Performance Computing Centre (HPCC) at Nanyang Technological University are gratefully acknowledged.

23

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

References (1) Scherer, J. Liebigs Ann. Chem. 1850, 73, 322-328. (2) Bouveault, L. Bull. Soc. Chim. Fr. 1894, 11, 44-147. (3) Young, L. Biochem. J 1934, 28, 1435-1443. (4) Folch, J.; Woolley, D. W. J. Biol. Chem. 1942, 142, 963-964. (5) Fisher, S. K.; Novak, J. E.; Agranoff, B. W. J. Neurochem. 2002, 82, 736-754. (6) Fisher, S. K.; Agranoff, B. W. J. Neurochem. 1987, 48, 999-1017. (7) Fisher, S. K.; Heacock, A. M.; Agranoff, B. W. J. Neurochem. 1992, 58, 18-38. (8) Potter, B. V. L.; Lampe, D. Angew. Chem. Int. Ed. 1995, 34, 1933-1972. (9) Payrastre, B.; Missy, K.; Giuriato, S.; Bodin, S.; Plantavid, M.; Gratacap, M. P. Cell. Signal. 2001, 13, 377-387. (10) Toker, A. Cell. Mol. Life Sci. 2002, 59, 761-779. (11) Michell, R. H. Nat. Rev. Mol. Cell Biol. 2008, 9, 151-161. (12) York, J. D. Biochim. Biophys. Acta 2006, 1761, 552-559. (13) Potter, B. V. L. Nat. Prod. Rep. 1990, 7, 1-24. (14) Irvine, R. F.; Schell, M. J. Nat. Rev. Mol. Cell Biol. 2001, 2, 327-338. (15) Bennett, M.; Onnebo, S. M. N.; Azevedo, C.; Saiardi, A. Cell. Mol. Life Sci. 2006, 63, 552-564. (16) Burton, A.; Hu, X.; Saiardi, A. J. Cell. Physiol. 2009, 220, 8-15. (17) Shears, S. B. Mol. Pharmacol. 2009, 76, 236-252. (18) Majerus, P. W. Annu. Rev. Biochem. 1992, 61, 225-250. (19) Berridge, M. J. Annu. Rev. Biochem. 1987, 56, 159-193. (20) Berridge, M. J.; Irvine, R. F. Nature 1989, 341, 197-205. (21) Berridge, M. J. Nature 1993, 361, 315-325. (22) Berridge, M. J.; Lipp, P.; Bootman, M. D. Nat. Rev. Mol. Cell Biol. 2000, 1, 11-21. (23) Berridge, M. J. Biochim. Biophys. Acta 2009, 1793, 933-940. (24) Xing, G.; Barr, E. W.; Diao, Y.; Hoffart, L. M.; Prabhu, K. S.; Arner, R. J.; Reddy, C. C.; Krebs, C.; Bollinger Jr, J. M. Biochemistry 2006, 45, 5402-5412. (25) Xing, G.; Diao, Y.; Hoffart, L. M.; Barr, E. W.; Prabhu, K. S.; Arner, R. J.; Reddy, C. C.; Krebs, C.; Bollinger Jr, J. M. Proc. Nat. Acad. Sci. U.S.A. 2006, 103, 6130-6135. (26) Xing, G.; Hoffart, L. M.; Diao, Y.; Prabhu, K. S.; Arner, R. J.; Reddy, C. C.; Krebs, C.; Bollinger Jr, J. M. Biochemistry 2006, 45, 5393-5401. (27) Bollinger, J. M., Jr.; Diao, Y.; Matthews, M. L.; Xing, G.; Krebs, C. Dalton Trans. 2009, 905-914. (28) Brown, P. M.; Caradoc-Davies, T. T.; Dickson, J. M. J.; Cooper, G. J. S.; Loomes, K. M.; Baker, E. N. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 15032-15037. (29) Thorsell, A.-G.; Persson, C.; Voevodskaya, N.; Busam, R. D.; Hammarström, M.; Gräslund, S.; Gräslund, A.; Hallberg, B. M. J. Biol. Chem. 2008, 283, 15209-15216. (30) Hirao, H.; Morokuma, K. J. Am. Chem. Soc. 2009, 131, 17206-17214. (31) Hirao, H. J. Phys. Chem. B 2011, 115, 11278-11285. (32) Arner, R. J.; Prabhu, K. S.; Thompson, J. T.; Hildenbrandt, G. R.; Liken, A. D.; Reddy, C. C. Biochem. J 2001, 360, 313-320. (33) Arner, R. J.; Prabhu, K. S.; Reddy, C. C. Biochem. Biophys. Res. Commun. 2004, 324, 1386-1392. (34) Prabhu, K. S.; Arner, R. J.; Vunta, H.; Reddy, C. C. J. Biol. Chem. 2005, 280, 1989519901. (35) Arner, R. J.; Prabhu, K. S.; Krishnan, V.; Johnson, M. C.; Reddy, C. C. Biochem. Biophys. Res. Commun. 2006, 339, 816-820. (36) Holub, B. J. Annu. Rev. Nutr. 1986, 6, 563-597.

24

ACS Paragon Plus Environment

Page 24 of 29

Page 25 of 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

The Journal of Physical Chemistry

(37) Deranieh, R. M.; Greenberg, M. L. Biochem. Soc. Trans. 2009, 37, 1099-1103. (38) Chen, I. W.; Charalampous, F. C. J. Biol. Chem. 1966, 241, 2194-2199. (39) Eisenberg. F J. Biol. Chem. 1967, 242, 1375-1382. (40) Pollack, S. J.; Knowles, M. R.; Atack, J. R.; Broughton, H. B.; Ragan, C. I.; Osborne, S. A.; McAllister, G. Eur. J. Biochem. 1993, 217, 281-287. (41) Pollack, S. J.; Atack, J. R.; Knowles, M. R.; McAllister, G.; Ragan, C. I.; Baker, R.; Fletcher, S. R.; Iversen, L. L.; Broughton, H. B. Proc. Nat. Acad. Sci. U.S.A. 1994, 91, 57665770. (42) Atack, J. R.; Broughton, H. B.; Pollack, S. J. Trends Neurosci. 1995, 18, 343-349. (43) Atack, J. R.; Broughton, H. B.; Pollack, S. J. FEBS Lett. 1995, 361, 1-7. (44) Atack, J. R. Brain Res. Rev. 1996, 22, 183-190. (45) Majumder, A. L.; Johnson, M. D.; Henry, S. A. Biochim. Biophys. Acta 1997, 1348, 245-256. (46) Spector, R.; Lorenzo, A. V. Am. J. Physiol. 1975, 228, 1510-1518. (47) Cade, J. F. J. Med. J. Aust. 1949, 2, 349-352. (48) Malhi, G. S.; Tanious, M.; Das, P.; Berk, M. Aust. N. Z. J. Psychiatry 2012, 46, 192211. (49) Machado-Vieira, R.; Manji, H. K.; Zarate, C. A. Bipolar Disord. 2009, 11, 92-109. (50) Belmaker, R. H. N. Engl. J. Med. 2004, 351, 476-486. (51) Berridge, M. J.; Downes, C. P.; Hanley, M. R. Biochem. J 1982, 206, 587-595. (52) Berridge, M. J.; Downes, C. P.; Hanley, M. R. Cell 1989, 59, 411-419. (53) Shaldubina, A.; Agam, C.; Belmaker, R. H. Prog. Neurossychopharmacol. Biol. Psychiatry 2001, 25, 855-866. (54) Harwood, A. J. Mol. Psychiatr. 2005, 10, 117-126. (55) Miller, D. J.; Allemann, R. K. Mini Rev. Med. Chem. 2007, 7, 107-113. (56) Quiroz, J. A.; Gould, T. D.; Manji, H. K. Mol. Interv. 2004, 4, 259-272. (57) Beaulieu, J. M.; Caron, M. G. Mol. Interv. 2008, 8, 230-241. (58) King, J. S.; Teo, R.; Ryves, J.; Reddy, J. V.; Peters, O.; Orabi, B.; Hoeller, O.; Williams, R. S. B.; Harwood, A. J. Dis. Model. Mech. 2009, 2, 306-312. (59) Quiroz, J. A.; Machado-Vieira, R.; Zarate, C. A.; Manji, H. K. Neuropsychobiology 2010, 62, 50-60. (60) Gani, D.; Downes, C. P.; Batty, I.; Bramham, J. Biochim. Biophys. Acta 1993, 1177, 253-269. (61) Phiel, C. J.; Klein, P. S. Annu. Rev. Pharmacool. Toxicol. 2001, 41, 789-813. (62) Allison, J. H.; Stewart, M. A. Nat. New Biol. 1971, 233, 267-268. (63) Allison, J. H.; Blisner, M. E.; Holland, W. H. Biochem. Biophys. Res. Commun. 1976, 71, 664-670. (64) Sherman, W. R.; Leavitt, A. L.; Honchar, M. P.; Hallcher, L. M.; Phillips, B. E. J. Neurochem. 1981, 36, 1947-1951. (65) Moore, G. J.; Bebchuk, J. M.; Parrish, J. K.; Faulk, M. W.; Arfken, C. L.; StrahlBevacqua, J.; Manji, H. K. Am. J. Psychiatry 1999, 156, 1902-1908. (66) Naccarato, W. F.; Ray, R. E.; Wells, W. W. Arch. Biochem. Biophys. 1974, 164, 194201. (67) Haimovich, A.; Eliav, U.; Goldbourt, A. J. Am. Chem. Soc. 2012, 134, 5647-5651. (68) Hallcher, L. M.; Sherman, W. R. J. Biol. Chem. 1980, 255, 896-901. (69) Gee, N. S.; Ragan, C. I.; Watling, K. J.; Aspley, S.; Jackson, R. G.; Reid, G. G.; Gani, D.; Shute, J. K. Biochem. J 1988, 249, 883-889. (70) Leech, A. P.; Baker, G. R.; Shute, J. K.; Cohen, M. A.; Gani, D. Eur. J. Biochem. 1993, 212, 693-704. (71) Greasley, P. J.; Hunt, L. G.; Gore, M. G. Eur. J. Biochem. 1994, 222, 453-460.

25

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

(72) Reisberg, B.; Gershon, S. Arch. Gen. Psychiatry 1979, 36, 879-887. (73) McAllister, G.; Whiting, P.; Hammond, E. A.; Knowles, M. R.; Atack, J. R.; Bailey, F. J.; Maigetter, R.; Ragan, C. I. Biochem. J 1992, 284, 749-754. (74) Ganzhorn, A. J.; Chanal, M. C. Biochemistry 1990, 29, 6065-6071. (75) Shute, J. K.; Baker, R.; Billington, D. C.; Gani, D. J. Chem. Soc. Chem. Commun. 1988, 626-628. (76) Baker, G. R.; Gani, D. Bioorg. Med. Chem. Lett. 1991, 1, 193-196. (77) Cole, A. G.; Gani, D. J. Chem. Soc. Chem. Commun. 1994, 1139-1141. (78) Wilkie, J.; Cole, A. G.; Gani, D. J. Chem. Soc. Perkin Trans. 1995, 2709-2727. (79) Bone, R.; Springer, J. P.; Atack, J. R. Proc. Nat. Acad. Sci. U.S.A. 1992, 89, 1003110035. (80) Bone, R.; Frank, L.; Springer, J. P.; Pollack, S. J.; Osborne, S.; Atack, J. R.; Knowles, M. R.; McAllister, G.; Ragan, C. I.; Broughton, H. B.; Baker, R.; Fletcher, S. R. Biochemistry 1994, 33, 9460-9467. (81) Bone, R.; Frank, L.; Springer, J. P.; Atack, J. R. Biochemistry 1994, 33, 9468-9476. (82) Fauroux, C. M. J.; Lee, M.; Cullis, P. M.; Douglas, K. T.; Freeman, S.; Gore, M. G. J. Am. Chem. Soc. 1999, 121, 8385-8386. (83) Fauroux, C. M. J.; Lee, M.; Cullis, P. M.; Douglas, K. T.; Gore, M. G.; Freeman, S. J. Med. Chem. 2002, 45, 1363-1373. (84) Ganzhorn, A. J.; Rondeau, J. M. Protein Eng. 1997, 10, 61-61. (85) Johnson, K. A.; Chen, L. J.; Yang, H. Y.; Roberts, M. F.; Stec, B. Biochemistry 2001, 40, 618-630. (86) Gill, R.; Mohammed, F.; Badyal, R.; Coates, L.; Erskine, P.; Thompson, D.; Cooper, J.; Gore, M.; Wood, S. Acta Crystallogr., Sect D: Biol. Crystallogr. 2005, 61, 545-555. (87) Arai, R.; Ito, K.; Ohnishi, T.; Ohba, H.; Akasaka, R.; Bessho, Y.; Hanawa-Suetsugu, K.; Yoshikawa, T.; Shirouzu, M.; Yokoyama, S. Proteins 2007, 67, 732-742. (88) Bhattacharyya, S.; Dutta, D.; Saha, B.; Ghosh, A. K.; Das, A. K. Biochimie 2012, 94, 879-890. (89) Singh, N.; Halliday, A. C.; Knight, M.; Lack, N. A.; Lowe, E.; Churchill, G. C. Acta Cryst. F 2012, 68, 1149-1152. (90) Miller, D. J.; Beaton, M. W.; Wilkie, J.; Gani, D. ChemBioChem 2000, 1, 262-271. (91) Fujita, S.; Ohnishi, T.; Okuda, S.; Kobayashi, R.; Fukuno, S.; Furuta, D.; Kikuchi, T.; Yoshikawa, T.; Fujita, N. J. Mol. Model. 2011, 17, 2559-2567. (92) Dudev, T.; Lim, C. J. Am. Chem. Soc. 2011, 133, 9506-9515. (93) Lu, S. Y.; Huang, W. K.; Li, X. B.; Huang, Z. M.; Liu, X. Y.; Chen, Y. Y.; Shi, T.; Zhang, J. J. Chem. Inf. Model. 2012, 52, 2398-2409. (94) Li, H.; Robertson, A. D.; Jensen, J. H. Proteins 2005, 61, 704-721. (95) Bas, D. C.; Rogers, D. M.; Jensen, J. H. Proteins 2008, 73, 765-783. (96) Case, D.; Darden, T.; Cheatham Iii, T.; Simmerling, C.; Wang, J.; Duke, R.; Luo, R.; Crowley, M.; Walker, R. C.; Zhang, W.; al., e. University of California, San Francisco 2008, 32. (97) Frisch, M.; Trucks, G.; Schlegel, H.; Scuseria, G.; Robb, M.; Cheeseman, J.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G.; Inc: 2009. (98) Hornak, V.; Abel, R.; Okur, A.; Strockbine, B.; Roitberg, A.; Simmerling, C. Proteins 2006, 65, 712-725. (99) Rosen, S. A. J.; Gaffney, P. R. J.; Gould, I. R. Phys. Chem. Chem. Phys. 2011, 13, 1070-1081. (100) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. J. Chem. Phys. 1983, 79, 926-935. (101) Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.; Villa, E.; Chipot, C.;

26

ACS Paragon Plus Environment

Page 26 of 29

Page 27 of 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

The Journal of Physical Chemistry

Skeel, R. D.; Kale, L.; Schulten, K. J. Comput. Chem. 2005, 26, 1781-1802. (102) Vreven, T.; Byun, K. S.; Komaromi, I.; Dapprich, S.; Montgomery, J. A.; Morokuma, K.; Frisch, M. J. J. Chem. Theor. Comput. 2006, 2, 815-826. (103) Chung, L. W.; Hirao, H.; Li, X.; Morokuma, K. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2011, 2, 327-350. (104) Ribeiro, A. J. M.; Ramos, M. J.; Fernandes, P. A. J. Chem. Theor. Comput. 2010, 6, 2281-2292. (105) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2004, 108, 6908-6918. (106) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004, 25, 1605-1612. (107) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graph. 1996, 14, 33-38. (108) Hart, J. C.; Sheppard, D. W.; Hillier, I. H.; Burton, N. A. Chem. Commun. 1999, 7980. (109) Hutter, M. C.; Helms, V. Protein Sci. 1999, 8, 2728-2733. (110) Hirano, T.; Oi, T.; Nagao, H.; Morokuma, K. Inorg. Chem. 2003, 42, 6575-6583. (111) Cavalli, A.; De Vivo, M.; Recanatini, M. Chem. Commun. 2003, 1308-1309. (112) Valiev, M.; Kawai, R.; Adams, J. A.; Weare, J. H. J. Am. Chem. Soc. 2003, 125, 99269927. (113) Díaz, N.; Field, M. J. J. Am. Chem. Soc. 2004, 126, 529-542. (114) Henkelman, G.; LaBute, M. X.; Tung, C. S.; Fenimore, P. W.; McMahon, B. H. Proc. Nat. Acad. Sci. U.S.A. 2005, 102, 15347-15351. (115) Cheng, Y. H.; Zhang, Y. K.; McCammon, J. A. J. Am. Chem. Soc. 2005, 127, 15531562. (116) Okimoto, N.; Yamanaka, K.; Ueno, J.; Hata, M.; Hoshino, T.; Tsuda, M. Biophys. J. 2001, 81, 2786-2794. (117) Li, G. H.; Cui, Q. J. Phys. Chem. B 2004, 108, 3342-3357. (118) Dittrich, M.; Hayashi, S.; Schulten, K. Biophys. J. 2003, 85, 2253-2266. (119) Hayashi, S.; Ueno, H.; Shaikh, A. R.; Umemura, M.; Kamiya, M.; Ito, Y.; Ikeguchi, M.; Komoriya, Y.; Iino, R.; Noji, H. J. Am. Chem. Soc. 2012, 134, 8447-8454. (120) Shurki, A.; Warshel, A. Proteins 2004, 55, 1-10. (121) Cavalli, A.; Carloni, P. J. Am. Chem. Soc. 2002, 124, 3763-3768. (122) Grigorenko, B. L.; Nemukhin, A. V.; Topol, I. A.; Cachau, R. E.; Burt, S. K. Proteins 2005, 60, 495-503. (123) Ribeiro, A. J. M.; Ramos, M. J.; Fernandes, P. A. J. Am. Chem. Soc. 2012, 134, 1343613447.

27

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

TOC

28

ACS Paragon Plus Environment

Page 28 of 29

Page 29 of 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

The Journal of Physical Chemistry

ACS Paragon Plus Environment