Isomerization of Δ5-Androstene-3, 17-dione into Δ4-Androstene-3, 17

Article pubs.acs.org/JPCA

Isomerization of Δ5‑Androstene-3,17-dione into Δ4‑Androstene-3, 17-dione Catalyzed by Human Glutathione Transferase A3-3: A Computational Study Identifies a Dual Role for Glutathione Daniel F. A. R. Dourado,†,∥ Pedro Alexandrino Fernandes,† Bengt Mannervik,‡,§ and Maria Joaõ Ramos*,† †

REQUIMTE/Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre, s/n, 4169-007 Porto, Portugal ‡ Department of Chemistry-BMC, Uppsala University, Box 576, SE-75123 Uppsala, Sweden § Department of Neurochemistry, Stockholm University, SE-10691 Stockholm, Sweden S Supporting Information *

ABSTRACT: Glutathione transferases (GSTs) are important enzymes in the metabolism of electrophilic xenobiotic and endobiotic toxic compounds. In addition, human GST A3-3 also catalyzes the double bond isomerization of Δ5-androstene-3,17-dione (Δ5-AD) and Δ5-pregnene3,20-dione (Δ5-PD), which are the immediate precursors of testosterone and progesterone. In fact, GST A3-3 is the most efficient human enzyme known to exist in the catalysis of these reactions. In this work, we have used density functional theory (DFT) calculations to propose a refined mechanism for the isomerization of Δ5-AD catalyzed by GST A3-3. In this mechanism the glutathione (GSH) thiol and Tyr9 catalyze the proton transfer from the Δ5-AD C4 atom to the Δ5-AD C6 atom, with a rate limiting activation energy of 15.8 kcal·mol−1. GSH has a dual function, because it is also responsible for stabilizing the negative charge that is formed in the O3 atom of the enolate intermediate. The catalytic role of Tyr9 depends on significant conformational rearrangements of its side chain. Neither of these contributions to catalysis has been observed before. Residues Phe10, Leu111, Ala 208, and Ala 216 complete the list of the important catalytic residues. The mechanism detailed here is based on the GST A3-3:GSH:Δ4-AD crystal structure and is consistent with all available experimental data.

■

catalyze such isomerization reactions (Table 1),16,17 but to date GST A3-3 is the most efficient human enzyme known to carry out that role.15 It has been shown that GST A3-3 is highly expressed only in tissues that synthesize steroid hormones, such as gonads, placenta, and the adrenal gland. On account of these findings it was suggested15 that GST A3-3 could be an important enzyme in steroid hormone biosynthesis. Since then, the highly efficient steroid isomerase activity of GST A3-3 has been the subject of several scientific studies.18,19 It was discovered that, despite the very high sequence identity between GST A3-3, GST A1-1, and GST A2-2 (>80%), the steroid isomerase activity of GST A3-3 is 10-fold higher than that of GST A1-1 and ∼5000fold higher than that of GST A2-220 (Table 1). The G-sites (the conserved pocket where GSH binds) of the three isoenzymes are strictly identical and therefore not responsible for the differences.

INTRODUCTION Glutathione transferases (GSTs) are enzymes mainly involved in the metabolism of electrophilic xenobiotic and endobiotic compounds. They catalyze the nucleophilic addition of glutathione (GSH) to a wide range of electrophilic substrates, forming less toxic and more soluble derivatives that subsequently can be expelled from the cell by ATP-dependent transmembrane pumps, such as MRP family proteins. Furthermore, some GSTs isoenzymes possess other highly specific catalytic activities.1,2 For instance, GST enzymes of class sigma have a prostaglandin D2 synthase activity;3,4 GSTs of the theta class have a specific sulfatase activity;5,6 GSTs of the omega class have a dehydroascorbate reductase activity and a GSH-dependent thioltransferase activity7 and are associated with the biotransformation pathway of arsenic;8 GSTs of the zeta class are involved in the tyrosine degradation pathway9,10 and are responsible for the catalytic biotransformation of dichloroacetate.11−14 In this work we are concerned with the study of GST A3-3.15 This enzyme is able to catalyze the isomerization of Δ5-androstene-3,17-dione (Δ5-AD) into Δ4-androstene-3,17dione (Δ4-AD) and also Δ5-pregnene-3,20-dione (Δ5-PD) into Δ4-pregnene-3,20-dione (Δ4-PD). The compounds Δ4-AD and Δ4-PD are the immediate precursors of testosterone and progesterone, respectively. Some GSTs were already known to © 2014 American Chemical Society

Special Issue: Energetics and Dynamics of Molecules, Solids, and Surfaces - QUITEL 2012 Received: November 2, 2013 Revised: April 10, 2014 Published: April 16, 2014 5790

dx.doi.org/10.1021/jp410810q | J. Phys. Chem. A 2014, 118, 5790−5800

The Journal of Physical Chemistry A

Article

Table 1. Kinetic Parameters of the Δ5-AD Isomerization Reaction for GST A1-1, GST A2-2, GST A3-3, and Four Mutant Enzymesa enzymes

kcat, s −1

KmAD, μM

kcat/KmAD, μM−1 s−1

GST A3-3 GST A1-1 GST A2-2 GST A2-2/S10F/I12G/ F111L/M208A/S216A GST A3-3/F10S/L111F/ A216S GST A3-3/Y9F GST A1-1/Y9F

228 ± 9 29.2 ± 0.8 0.26 ± 0.01 195 ± 11

45 ± 4 58 ± 4 250 ± 20 55

5.0 ± 04 0.50 ± 0.04 0.00104 ± 0.00009 3.5 ± 0.45

2.3 ± 0.2

60 ± 12

0.038 ± 0.004

0.66 ± 0.01 0.61 ± 0.02

9.3 ± 0.6 25 ± 3

0.07 ± 0.004 0.024 ± 0.003

The first corresponds to a GST A2-2 mutant that mimics the GSTA33 H-site. The second corresponds to a triple point GST A3-3 mutant, which was transmuted with corresponding GST A2-2 H-site residues.17,20,21 The last two mutants are GST A3-3 Y9F and GST A1-1 Y9F. a

Table 2. Differences in the Amino Acid Residues among the GST A1-1, GST A2-2, and GST A3-3 H-Site Pockets residue enzyme

10

12

111

208

216

GST A3-3 GST A1-1 GST A2-2

Phe Phe Ser

Gly Ala Ile

Leu Val Phe

Ala Met Met

Ala Ala Ser

However, the H-sites of GST A3-3 and GST A1-1 (the pocket where a large spectrum of electrophilic compounds are able to bind to) differ only in residues 12, 111, and 208, whereas the H-sites of GST A3-3 and GST A2-2 differ in residues 10, 12, 111, 208, and 216 (Table 2). When these five amino acids are mutated in GST A2-2, to mimic the GST A3-3 H-site, the steroid isomerase activity increases by 3500-fold, approaching closely the activity of GST A3-3 (Table 1).20 An extensive study with GST A3-3 mutants elucidated some fundamental aspects.21 As mentioned above, five residues distinguish the GST A3-3 H-site from the one of GST A2-2. Residues 10, 111, and 216, in a cumulative fashion, are the ones that have the strongest influence in catalysis. When they are mutated in GST A3-3 to the corresponding GST A2-2 residues (Ser10, Phe111, and Ser 216), the rate of catalysis decreases 131fold. On the other hand, the same study suggests that Tyr9, beyond its classical role of lowering the pKa of the GSH thiol, also participates in the steroid isomerase catalysis. For the GST A3-3-Y9F mutant, the steroid isomerase catalytic activity (kcat) decreases 354-fold (Table 1). In fact, the mutants GST A3-3-Y9F and GST A1-1-Y9F have identical values of kcat (Table 1), showing that for these mutants the differences in the H-site have no significant influence on the steroid isomerase turnover number. Furthermore, GSH itself also has an important role in the steroid isomerase catalysis. When GSH is absent, the value of kcat decreases 30-fold and the value of KMAD increases 7-fold, showing that GSH plays a role in catalysis and that it increases the affinity of Δ5-AD to the H-site. With the above findings, it was first suggested20,21 that the GSH thiolate should act as an acid/base catalyst deprotonating the Δ5AD C4 atom and protonating the Δ5-AD C6 atom. Additionally, the Tyr9 hydroxyl group would stabilize the charge formed at the O3 atom of the enolate intermediate (please see atoms numbering in Figure 1). We shall identify this hypothesis as mechanism A.

Figure 1. Reaction mechanisms, A and B, proposed for the Δ5-AD isomerization reaction catalyzed by GST A3-3. The two mechanisms differ only in the second step: (a) mechanism A;23 (b) mechanism B.22

Subsequent to these findings, the crystal structure of GST A3-3 complexed with GSH was solved22 and it was used to build a computational model of the GST A3-3:GSH:Δ4-AD ternary complex. The structure suggested22 that the thiolate of GSH deprotonates the Δ5-AD carbon atom C4, whereas the Tyr9 hydroxyl protonates the Δ5-AD carbon atom C6. The charge that builds in the O3 atom of the enolate intermediate is stabilized by a water molecule. In the end, GSH would promptly transfer a proton to the ionized Tyr9, restoring the initial state of the enzyme. Throughout the article we will invoke this hypothesis as mechanism B. Figure 1 summarizes all the events. Recently, the Mannervik group that proposed mechanism A solved the crystal structure of GST A3-3 complexed with both GSH and the final steroid isomerase product, Δ4-AD.23 This structure shows that Tyr9 is too far away from the O3 atom and so the previous suggested that the Tyr9 catalytic role was to be reconsidered. It was proposed23 that Tyr9 promotes, even if in an indirect way, the proton transfer from GSH to the C6 carbon atom by making a strong hydrogen bond with the thiol group. The suggested role of GSH20,21 was not altered. Throughout the 5791



Article

present article whenever we mention mechanism A we are invoking this restructured hypothesis. Figure 1 summarizes all the events. Despite all these investigations, the catalytic mechanism of GST A3-3 Δ5-AD isomerization remains not completely understood. Bottoni and co-workers have studied this mechanism computationally18,19 and provided many atomistic insights, but their reported energy barriers (29.9 and 22.4 kcal·mol−1, respectively) are quite different from the experimentally reported one (14.5 kcal·mol−1), and they could not observe the dual function GSH activity. Recently, we studied the GSH activation mechanism in the isoenzymes GST A1-1, GST P1-1, and GST M1-1.24−27 This catalytic step (GSH thiol deprotonation) precedes most reactions catalyzed by GSTs, including the steroid isomerization catalysis. In the present study, we have resorted to density functional theory (DFT) calculations, to clarify the Δ5-AD isomerization mechanism of GST A3-3. We have obtained fundamental insights in the catalytic role played by glutathione. This mechanism is based on the GST A3-3:GSH:Δ4-AD crystal

Figure 2. Stereoview of the G-site (green carbons) and H-site (blue carbons) pockets model (276 atoms).

Figure 3. Free energy barriers of the Δ5-AD isomerization reaction catalyzed by GST A3-3. The full mechanism is composed of three catalytic steps and by two Tyr9 conformational rearrangements (CR). The free energy barrier for the second step of mechanism A proposal is also represented.

Figure 4. Superimposition of the transition state structures for the second step of mechanism A (violet) and mechanism B (green) shown in two different orientations. 5792



Article

Figure 5. First step of the Δ5-AD isomerization reaction catalyzed by GST A3-3. The reaction coordinate (δ) corresponds to the distance between the GSH sulfur and the C4 proton. Relevant distances (Å) and charges (au) are shown.

the GSH already activated (deprotonated). The α carbons of all the amino acids were fixed during the whole study to avoid eventual local unfolding. The GSH glutamyl α-amino group was modeled deprotonated. Hydrogen atoms were used as link atoms at the truncated bonds. Linear scans for the proton transfers and scans for the necessary Tyr9 structural rearrangements were performed. All the geometry optimizations were conducted with the ONIOM hybrid method,28−30 as implemented in Gaussian 09.31 The high layer (117 atoms) was described with density functional theory (DFT) using the B3LYP functional32,33 and the 6-31G(d)32,33 basis set as they both have proven to give good results in the past for this type of system.22,24 It included GSH, Δ5-AD, and the atoms from residues Tyr9, Arg45, Gln67, and Thr68, which directly interact with GSH and may have important roles in the proton transfer (Figure 2). The low layer, treated

structure23 and is consistent with mechanism B as well as with all the available experimental data.

■

METHODOLOGY We used the GST A3-3:GSH:Δ4-AD crystal structure23 to transform Δ4-AD into Δ5-AD by changing the hybridization of carbon atom C4 from sp3 to sp2 and the hybridization of carbon atom C6 from sp2 to sp3. Subsequently, the G-site and H-site pockets were removed from the enzyme generating an enzyme model with 276 atoms (Figure 2). This model includes GSH, Δ5-AD, and all the residues critical to catalysis (Phe10, Gly12, Leu111, Ala208, Ala216, Arg15, Arg45, Gln67, Thr68, and Glu104). The XYZ coordinates of the model can be seen in the Supporting Information. It should be noted that, to study the isomerization reaction, we started with 5793



Article

Figure 6. First conformational rearrangement of Tyr9 in the Δ5-AD isomerization reaction catalyzed by GST A3-3: (A) reactants, before the rearrangement; (B) products, after rearrangement. Relevant distances (Å) are shown.

with the semiempirical method PM3MM,34,35 includes the rest of the model atoms. Initial guesses for the structures of the stationary points (proton transfers: reagent, transition state and product; Tyr9 structural rearrangements: reagent and product) were taken from the scans. The stationary points were later optimized and their nature was confirmed by frequency calculations. As a consequence of the constraints applied, several imaginary frequencies were a result of these calculations for the transition state. However, these imaginary frequencies were several orders of magnitude smaller than the one associated with the reaction coordinate in the transition state (the only true imaginary frequency). Entropic corrections based on such calculations offer normally a relatively safe estimate to free energy differences between stationary states. After obtaining the stationary points, we performed single point calculations to obtain the energy of the high layer with the basis set 6-311++G(2d,2p). A continuum model was used as an approximation to the effect of the protein beyond the atomistic region. We chose the ONIOM-PCM continuum model,36,37 as implemented in G09, with a dielectric constant of 4. Zero point corrections and thermal and entropic effects were also added to obtain the final free energies values (T = 310.15 K, P = 1 bar). All the partial charges were obtained from a Mulliken population analysis.38 This methodology has been used recently in the catalytic study of several enzymes with excellent results.24,26,27,39−45 To further demonstrate the flexibility of the Tyr9 side chain, molecular dynamic (MD) simulations starting with the reagent structures of both Tyr9 conformational rearrangements were performed. Both systems were solvated with ∼16 700 single point charge waters (SPC).46 Then they were submitted to 100 steps of steepest descent energy minimization to remove bad contacts between the solvent and the protein. Subsequently, they were equilibrated with a 200 ps MD simulation by maintaining the protein atoms restrained by weak harmonic constraints to allow for the structural relaxation of the water molecules. Later, production MD simulations of 20 ns were performed. A time step of 0.002 ps was applied, the trajectories being saved at each 20 ps.

Periodic boundary conditions were used in all simulations. The temperature and pressure were maintained constant by the use of the v-rescale thermostat47 and Parrinello−Rahman barostat48 (parameters: τT = 0.1 ps, Tref = 300 K, Pref =1 bar). The particle mesh Ewald (PME)49 method was applied to compute electrostatic interactions with a cutoff of 1.0 nm. A twin range cutoff with neighbor list cutoff 1.0 and cutoff of 1.0 was used for the van der Waals interactions. All the bonds involving hydrogen atoms were constrained by the LINCS constraint algorithm.50 All the simulations and subsequent analyses were carried out using the Gromacs software package conjugated with the Amber99 force field.51−54

■

RESULTS AND DISCUSSION Analysis of the Two Mechanistic Proposals Previously Presented in the Literature. First Step. Mechanism A. In the first step of both proposals, GS− abstracts a proton from the Δ5-AD C4 carbon atom (Figure 1). In mechanism A23 it was not suggested how the charge formed at the O3 atom of the enolate intermediate could be stabilized. In mechanism B, however, it was suggested22 that a water molecule could perform that role through H-bonding to oxygen atom O3. We have carried out the calculations according to this mechanism and our results suggest that, as the proton moves from atom C4 to the GS−, the negative charge that builds upon C4 tends to delocalize through the region between atoms O3 and C6 (please see atoms numbering in Figure 1). This charge delocalization leads to an increase in the negative charge of atom O3. This charge is compensated by an ionic hydrogen bond interaction, taking place between atom O3 and the GSH-Gly main chain amide. Later in the manuscript we will go into this further, while explaining our own mechanism proposal. We have obtained 12.7 kcal·mol−1 for the activation free energy (ΔG‡) of this step and 10.7 kcal·mol−1 for the reaction free energy (ΔGr) (Figure 3). Mechanism B. We have tested mechanism B’s22 suggestion of including a water molecule in the active site, which would 5794



Article

Figure 7. Second step of the Δ5-AD isomerization reaction catalyzed by GST A3-3. The reaction coordinate (ζ) corresponds to the distance between the Tyr9 proton and carbon C6. Relevant distances (Å) and charges (au) are shown.

molecules are neither necessary nor responsible for the stabilization of the intermediate enolate. Second Step. As Figure 1 demonstrates, the second step is completely different in each proposal. Mechanism A. In the second step of mechanism A, GSH protonates atom C6, yielding product Δ4-AD. We tested this hypothesis. Our studies demonstrate that the free energy barrier associated with this proton transfer is 7.3 kcal·mol−1 and the reaction free energy (ΔGr) is −10.5 kcal·mol−1 (Figure 3). The associated stationary point structures can be found in the Supporting Information. Mechanism B. According to the mechanism B proposal,22 the Tyr9 hydroxyl proton can easily be transferred to the C6 atom, forming isomer Δ4-AD. Subsequently, GSH would spontaneously transfer a proton to the ionized Tyr9 hydroxyl group, restoring the initial state of the enzyme. Our results indicate that the Tyr9 residue side chain needs to undergo a conformational rearrangement before the proton transfer takes place, so that it can interact directly with carbon C6. This rearrangement is not spontaneous.

stabilize the charge formed in atom O3 at the transition state and products of this step. In the first place, we submitted the ternary complex GST A3-3:GSH:Δ5-AD to a 20 ns molecular dynamics simulation in explicit solvent. Subsequently, we calculated the radial distribution functions for the two oxygen atoms of Δ5-AD relatively to the water oxygen atoms (OW). Around each oxygen atom we observed a defined first water coordination sphere that corresponds to two water molecules. On the basis of it we built a model identical to that shown in Figure 2, which includes two water molecules, hydrogen-bonded to each oxygen atom of Δ5-AD. The corresponding figure and the XYZ coordinates of this model can be seen in the Supporting Information. With this new model we performed the same proton transfer scan from the C4 atom to the GS− one. We observed no relevant difference from the results obtained for the model with no water molecules. The same ionic hydrogen bond interaction between the negative charge at the O3 atom and GSH-Gly main chain amide still formed. On the basis of these results, we conclude that water 5795



Article

Figure 8. Second conformational rearrangement of Tyr9 in the Δ5-AD isomerization reaction catalyzed by GST A3-3: (A) reactants, before the rearrangement; (B) products, after the rearrangement. Relevant distances (Å) are shown.

It has associated a free energy barrier of 2.8 kcal·mol−1 (Figure 3), and when overcome, the Tyr9 hydroxyl proton can then be transferred to carbon C6. The activation free energy of this step is as small as 1.9 kcal·mol−1, and the reaction is very exergonic (ΔGr = −14.1 kcal mol−1) (Figure 3). However, Tyr9 is not spontaneously protonated by GSH at this point, contrary to what was predicted in mechanism B. When explaining our own mechanism proposal, we will show in detail how carbon C6 is protonated by Tyr9 and what is necessary to restore the initial state of the enzyme. Comparison of the Second Step of Both Mechanisms. The free energy barrier obtained for the second step of mechanism A is considerably higher than that obtained for mechanism B (Figure 3). Why does this happen? In Figure 4 it is possible to compare the transition state structures from both mechanisms. In mechanism A the direct interaction between GSH and atom C6 involves a significant rearrangement of the intermediate enolate within the H-site and a distortion in the GSH arrangement. This GSH distortion leads to a structural redisposition of all the side-chains of the G-site residues that are directly interacting with GSH, most likely with a significant energy cost involved. In the case of mechanism B, we just observe a Tyr9 conformational rearrangement, prior to the actual reaction, with an energy barrier of 2.8 kcal·mol−1 (Figure 3). It could be hypothesized that the difference in energy between having GSH or the Tyr9 side chain as acid catalysts could be due to differences in the microscopic pKa of the residues in the active site (in this case with Tyr having a lower pKa). However, this hypothesis seems less plausible as it is experimentally known that, within the active center, GSH has a pKa as low as 6.6 pH units,55,56 and the Tyr9 side chain has a pKa of ∼9 pH units,56 which contrarily would represent an extra cost of ∼3 kcal·mol−1 for using Tyr9 as the proton donor, instead of using GSH. Moreover, the transfer of a proton from GSH to Tyr9 has a very low energy barrier ruling out that a difference in the pKa values could be the basis of the difference found.

Our Mechanistic Proposal. Our results demonstrate that neither mechanism A nor mechanism B can explain the steroid isomerization catalysis in its full extension. While analyzing these two mechanistic proposals, we were defining our own mechanism as follows. In the first step GS− acts as base and deprotonates carbon C4 (Figure 5) as happens in both literature proposals (Figure 1). Our calculated activation free energy (ΔG‡) for this step is 12.7 kcal·mol−1 and the reaction free energy (ΔGr) is 10.7 kcal·mol−1 (Figure 3), as stated before. However, our results demonstrate that the negative charge that builds up in the O3 oxygen atom of the enolate intermediate is stabilized by a strong ionic hydrogen bond with the GSH-Gly main chain amide (Figure 5). The difference in charge of oxygen O3 between the reactants and products is −0.28 au. The charge of carbon C6 does not change significantly from reactants (−0.38 au) to products (−0.36 au). This charge analysis indicates that, presumably, the enolate resonance structure (instead of the keto structure with an ionic carbon C6; Figure 10) is the major contributor to the products structure. The dual function observed for GSH is also supported by experimental kinetics results. When GSH is5 absent, the value of kcat decreases 30-fold while the value of KMΔ ‑AD increases from 45 to 310 μM.21 For the second step, mechanism A has a higher free energy barrier than mechanism B. Therefore, the suggestion that Tyr9 may protonate carbon C6 should be considered. However, mechanism B does not predict the necessary conformational rearrangements of Tyr9, which are essential for the reaction to take place, and does not restore the initial state of the enzyme. In our proposal we take these aspects into account. The initial conformational rearrangement of Tyr9 side chain allows a direct interaction of this residue with carbon C6 (Figure 6). This rearrangement has a free energy barrier of 2.8 kcal·mol−1 and is facilitated by the prior protonation of the GSH thiolate, thus weakening the hydrogen bond that the latter was making with the Tyr9 hydroxyl. An MD simulation performed with this active center structure demonstrates that the distance between GSH thiol 5796



Article

Figure 9. Third step of the Δ5-AD isomerization reaction catalyzed by GST A3-3. The reaction coordinate (θ) corresponds to the distance between the GSH proton and the Tyr9 side chain oxygen. Relevant distances (Å) and charges (au) are shown.

This reaction has an activation energy (ΔG‡) of 1.9 kcal·mol−1, and a reaction free energy (ΔGr) of −14.1 kcal·mol−1 (Figure 3). Later, in a third chemical step, GSH protonates Tyr9 restoring the initial state of the enzyme. However, at this stage, GSH and Tyr9 are no longer hydrogen-bonded. The Tyr9 side chain needs to change its structural arrangement to interact directly with GSH. Such conformational rearrangement has a free energy barrier of 3.2 kcal·mol−1 (Figure 8). We performed an MD simulation starting with this active center structure. The distance between GSH thiol and Tyr9-O− groups fluctuates between 3 and 5 Å (with a short peak of 6.5 Å), further proving the Tyr9 side chain flexibility shown in Figure 8 (Figure S4, Supporting Information). Residue Tyr9 is now able to receive the proton from GSH (Figure 9). We performed a linear scan for the proton transfer from GSH to Tyr9. The TS frequency calculation demonstrated the existence of an energy barrier (note that as stated before, as we froze some atomic positions, the frequency calculations

and Tyr9 hydroxyl groups fluctuates between 3 and 5 Å (with a short peak of 6.5 Å), which is in accordance with the distances shown in Figure 6, further proving the high flexibility of Tyr 9 side chain in this step (Figure S3, Supporting Information). When residue Tyr9 is finally interacting with carbon C6, its side chain polarizes the C6 atom resulting in a considerable increase in carbon C6 negative charge. The difference in charge before and after the conformational rearrangement corresponds to approximately −0.26 au, displacing the keto/enolate hybrid structure more toward the keto form. Subsequently, residue Tyr9, in a second chemical step, protonates carbon C6 forming the product of the isomerization reaction, Δ4-AD (Figure 7). As the proton is being transferred from residue Tyr9 to carbon C6, the hydrogen bond between the O3 oxygen and GSH-Gly becomes weaker and longer (from 1.87 Å at the reactants to 2.14 Å at the products). The difference in the O3 oxygen charge between the reactants and products corresponds to 0.19 au. 5797



Article

Figure 10. This work’s mechanistic proposal for the Δ5-AD isomerization reaction catalyzed by GST A3-3. The mechanism is composed of three catalytic steps and two Tyr9 conformational rearrangements.

cycle is expected to be close to zero. The difference in the internal energy of the two isomers is small (ΔGΔ4‑AD−Δ5‑AD = −2.4 kcal·mol−1) and the difference in the interaction energy is expected to be negligible, due to the electrostatic similarity between the initial reactants and the final products of the cycle. The obtained value, −1.0 kcal·mol−1, makes full sense in light of these considerations. Figure 10 recapitulates all the events of our mechanism proposal.

yielded more than one imaginary value, but only one had a significant value; all the others were extremely small, negligible really, in comparison). However, after adding the continuum solvent energy, the zero point corrections, the thermal and the entropic effects to the stationary points energies, the transition state became 0.4 kcal·mol−1 below that of the reactants (Figure 3). We can conclude that a potential energy barrier exists, but probably not a free energy barrier at physiological temperature. This step should be spontaneous after the second conformational rearrangement of Tyr9. The rate determining states of the whole cycle are the transition state of the second step and the reactants of the first step, leading to an energetic span of 15.8 kcal·mol−1. This means that the apparent barrier for the whole turnover is 15.8 kcal·mol−1, a result quite close to the experimental value obtained for this reaction, 14.5 kcal·mol−1.15 The net reaction free energy of the whole

■

CONCLUSION

We used the crystal structure of GST A3-3 complexed with GSH and Δ4-AD, the product of the isomerase reaction, to model the reactant Δ5-AD bound to the H-site in a catalytically productive conformation. 5798



Article

(5) Hiratsuka, A.; et al. A New Class Of Rat Glutathione S-transferase Yrs-Yrs Inactivating Reactive Sulfate Esters As Metabolites Of Carcinogenic Arylmethanols. J. Biol. Chem. 1990, 265, 11973−81. (6) Hussey, A. J.; Hayes, J. D. Characterization of a Human ClassTheta Glutathione-S-Transferase with Activity Towards 1-Menaphthyl Sulfate. Biochem. J. 1992, 286, 929−935. (7) Board, P. G.; et al. Identification, Characterization, And Crystal Structure Of The Omega Class Glutathione Transferases. J. Biol. Chem. 2000, 275, 24798−806. (8) Zakharyan, R. A.; Sampayo-Reyes, A.; Healy, S. M.; Tsaprailis, G.; Board, P. G.; Liebler, D. C.; Aposhian, H. V. Human Monomethylarsonic Acid (MMA(V)) Reductase Is A Member Of The Glutathione-Stransferase Superfamily. Chem. Res. Toxicol. 2001, 14, 1051−7. (9) Fernandez-Canon, J. M.; Penalva, M. A. Characterization Of A Fungal Maleylacetoacetate Isomerase Gene And Identification Of Its Human Homologue. J. Biol. Chem. 1998, 273, 329−37. (10) Blackburn, A. C.; Woollatt, E.; Sutherland, G. R.; Board, P. G. Characterization and Chromosome Location Of The Gene GSTZ1 Encoding The Human Zeta Class Glutathione Transferase And Maleylacetoacetate Isomerase. Cytogenet. Cell Genet. 1998, 83, 109−14. (11) James, M. O.; Cornett, R.; Yan, Z.; Henderson, G. N.; Stacpoole, P. W. Glutathione-dependent Conversion To Glyoxylate, A Major Pathway Of Dichloroacetate Biotransformation In Hepatic Cytosol From Humans And Rats, Is Reduced In Dichloroacetate-Treated Rats. Drug Metab. Dispos. 1997, 25, 1223−1227. (12) Tong, Z.; Board, P. G.; Anders, M. W. Glutathione Transferase Zeta Catalyses The Oxygenation Of The Carcinogen Dichloroacetic Acid To Glyoxylic Acid. Biochem. J. 1998, 331 (Pt 2), 371−4. (13) Anderson, W. B.; Board, P. G.; Gargano, B.; Anders, M. W. Inactivation Of Glutathione Transferase Zeta By Dichloroacetic Acid And Other Fluorine-Lacking Alpha-Haloalkanoic Acids. Chem. Res. Toxicol. 1999, 12, 1144−9. (14) Tzeng, H. F.; Blackburn, A. C.; Board, P. G.; Anders, M. W. Polymorphism- And Species-Dependent Inactivation Of Glutathione Transferase Zeta By Dichloroacetate. Chem. Res. Toxicol. 2000, 13, 231− 6. (15) Johansson, A.-S.; Mannervik, B. Human Glutathione Transferase A3-3, A Highly Efficient Catalyst Of Double-Bond Isomerization In The Biosynthetic Pathway Of Steroid Hormones. J. Biol. Chem. 2001, 276, 33061−5. (16) Benson, A. M.; Talalay, P.; Keen, J. H.; Jakoby, W. B. Relationship Between The Soluble Glutathione-Dependent Delta 5−3-Ketosteroid Isomerase And The Glutathione S-Transferases Of The Liver. Proc. Natl. Acad. Sci. U. S. A. 1977, 74, 158−62. (17) Pettersson, P. L.; Mannervik, B. The Role Of Glutathione In The Isomerization Of Delta(5)-Androstene-3,17-Dione Catalyzed By Human Glutathione Transferase A1-1. J. Biol. Chem. 2001, 276, 11698−11704. (18) Matteo, C.; Stenta, Marco; Altoè, Piero; Bottoni, Andrea; Garavelli, Marco; Spinelli, Domenico Computational QM/MM Study of the Reaction Mechanism of Human Glutathione S-Transferase A3-3. AIP Conf. Proc. 2007, 963, 696−698. (19) Matteo Calvaresi, M. S.; Garavelli, Marco; Altoé, Piero; Bottoni, Andrea Computational Evidence for the Catalytic Mechanism of Human Glutathione S-Transferase A3-3: A QM/MM Investigation. ACS Catal. 2012, 2, 280−286. (20) Pettersson, P. L.; Johansson, A. S.; Mannervik, B. Transmutation of Human Glutathione Transferase A2-2 With Peroxidase Activity Into An Efficient Steroid Isomerase. J. Biol. Chem. 2002, 277, 30019−22. (21) Johansson, A. S.; Mannervik, B. Active-Site Residues Governing High Steroid Isomerase Activity In Human Glutathione Transferase A33. J. Biol. Chem. 2002, 277, 16648−54. (22) Gu, Y.; Guo, J.; Pal, A.; Pan, S. S.; Zimniak, P.; Singh, S. V.; Ji, X. Crystal Structure Of Human Glutathione S-Transferase A3-3 And Mechanistic Implications For Its High Steroid Isomerase Activity. Biochemistry 2004, 43, 15673−9. (23) Tars, K.; Olin, B.; Mannervik, B. Structural Basis For Featuring Of Steroid Isomerase Activity In Alpha Class Glutathione Transferases. J. Mol. Biol. 2010, 397, 332−40.

GSH and Tyr9 were shown to be able to catalyze the proton transfer from Δ5-AD C4 atom to the Δ5-AD C6 atom. The obtained overall barrier, 15.8 kcal·mol−1, seems to be in accordance with the experimental kinetics (kcat 228 ± 9 s−1; ∼14.5 kcal·mol−1 15). GSH has a dual function because it also stabilizes the negative charge that builds up in the O3 oxygen of the enolate intermediate by establishing a strong ionic hydrogen bond with the GSH-Gly main chain amide. This GSH dual function is supported by experimental kinetics.5 When GSH is absent, the kcat decreases 30-fold while the KMΔ ‑AD increases from 45 to 310 μM.21 The mechanism is proposed to occur through three chemical steps plus two conformational rearrangements, in a total of five elementary steps. These results provide for the first time a clear atomic level portrait of the isomerization reaction, clarifying the role of the enzyme and of the GSH cofactor in the transformation that takes place during isomerization.

■

ASSOCIATED CONTENT

* Supporting Information S

Second step of mechanism A for the Δ5-AD isomerization reaction. Stereoview of the G-site and H-site pockets model. Change in the distance between GSH thiol and Tyr9 hydroxyl and hydroxylate groups. Cartesian coordinates. This material is available free of charge via the Internet at http://pubs.acs.org.”

■

AUTHOR INFORMATION

Corresponding Author

*M. J. Ramos: e-mail, [email protected]; telephone, +351 220 402 506. Present Address

∥ Department of Cell and Molecular Biology, Computational and Systems Biology, Uppsala Biomedicinska Centrum BMC, Box 596 751 24, Uppsala, Sweden.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank the FCT (Fundaçaõ para a Ciência e Tecnologia) for financial support linked to Project POCI/QUI/61563/2004 as well as for a Postdoctoral grant SFRH/BPD/74441/2010 for DFARD. B.M. is supported by the Swedish Research Council

■

ABBREVIATIONS GSTs, glutathione transferases; GSH, glutathione; Δ5-AD, Δ5androstene-3,17-dione; Δ5-PD, Δ5-pregnene-3,20-dione; Δ4AD, Δ4-androstene-3,17-dione; DFT, density functional theory

■

REFERENCES

(1) Josephy, P. D., and Mannervik, B. Molecular Toxicology; Oxford University Press: New York, 2006; p 608. (2) Dourado, D. F. A. R.; Fernandes, P. A.; Ramos, M. J. Mammalian Cytosolic Glutathione Transferases. Curr. Protein Pept. Sci. 2008, 9, 325−37. (3) Meyer, D. J.; Thomas, M. Characterization of Rat Spleen Prostaglandin-H D-Isomerase as a Sigmaclass Gsh Transferase. Biochem. J. 1995, 311, 739−742. (4) Jowsey, I. R.; Thomson, A. M.; Flanagan, J. U.; Murdock, P. R.; Moore, G. B.; Meyer, D. J.; Murphy, G. J.; Smith, S. A.; Hayes, J. D. Mammalian Class Sigma glutathione S-transferases: Catalytic Properties And Tissue-Specific Expression Of Human And Rat GSH-dependent Prostaglandin D2 Synthases. Biochem. J. 2001, 359, 507−16. 5799



Article

(46) Berendsen, H., Postma, J.P.M.; vanGunsteren, W.F. Intermolecular Forces; D. Reidel Publishing Company: Dordrecht, The Netherlands, 1981; pp 331−342. (47) Bussi, G.; Donadio, D.; Parrinello, M. Canonical Sampling Through Velocity Rescaling. J. Chem. Phys. 2007, 126. (48) Parrinello, M.; Rahman, A. Polymorphic Transitions in SingleCrystals - a New Molecular-Dynamics Method. J. Appl. Phys. 1981, 52, 7182−7190. (49) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem. Phys. 1995, 103, 8577−8593. (50) Hess, B.; Bekker, H.; Berendsen, H.; Fraaije, J. LINCS: A Linear Constraint Solver for Molecular Simulations. J. Comput. Chem. 1997, 18, 1463−1472. (51) Erik Lindahl, B. H.; van der Spoel, David. GROMACS 3.0: A Package For Molecular Simulation And Trajectory Analysis. J. Mol. Model. 2001, 7, 306−317. (52) Wang, J. M.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Development And Testing Of A General Amber Force Field. J. Comput. Chem. 2004, 25, 1157−1174. (53) Sorin, E. J.; Pande, V. S. Exploring The Helix-Coil Transition Via All-Atom Equilibrium Ensemble Simulations. Biophys. J. 2005, 88, 2472−93. (54) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M., Jr.; 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. (55) Graminski, G. F.; Kubo, Y.; Armstrong, R. N. Spectroscopic And Kinetic Evidence For The Thiolate Anion Of Glutathione At The Active Site Of Glutathione S-transferase. Biochemistry 1989, 28, 3562−8. (56) Björnestedt, R.; Stenberg, G.; Widersten, M.; Board, P. G.; Sinning, I.; Jones, T. A.; Mannervik, B. Functional Significance Of Arginine 15 In The Active Site Of Human Class Alpha Glutathione Transferase A1-1. J. Mol. Biol. 1995, 247, 765−73.

(24) Dourado, D. F. A. R.; Fernandes, P. A.; Mannervik, B.; Ramos, M. J. Glutathione Transferase: New Model for Glutathione Activation. Chem.Eur. J. 2008, 14, 9591−9598. (25) Dourado, D. F. A. R.; Fernandes, P. A.; Ramos, M. J. Glutathione Transferase A1-1: Catalytic Role Of Water. Theor. Chem. Acc. 2009, 124, 71−83. (26) Dourado, D. F.; Fernandes, P. A.; Mannervik, B.; Ramos, M. J. Glutathione Transferase A1-1: Catalytic Importance Of Arginine 15. J. Phys. Chem. B 2010, 114, 1690−7. (27) Dourado, D. F.; Fernandes, P. A.; Ramos, M. J. Glutathione Transferase Classes Alpha, Pi, And Mu: GSH Activation Mechanism. J. Phys. Chem. B 2010, 114, 12972−80. (28) Bakowies, D.; Thiel, W. Hybrid Models For Combined Quantum Mechanical And Molecular Mechanical Approaches. J. Phys. Chem. 1996, 100, 10580−10594. (29) Dapprich, S.; Komaromi, I.; Byun, K. S.; Morokuma, K.; Frisch, M. J. A New ONIOM Implementation In Gaussian98. Part I. The Calculation Of Energies, Gradients, Vibrational Frequencies And Electric Field Derivatives. J. Mol. Struct. (THEOCHEM) 1999, 462, 1−21. (30) Cosman, P. C.; Rogers, J. K.; Sherwood, P. G.; Zeger, K. Combined Forward Error Control And Packetized Zerotree Wavelet Encoding For Transmission Of Images Over Varying Channels. IEEE Trans. Image Process 2000, 9, 982−993. (31) Frisch, M. J.; et al. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009. (32) Lee, C. T.; Yang, W. T.; Parr, R. G. Development of the ColleSalvetti Correlation-Energy Formula into a Functional of the ElectronDensity. Phys. Rev. B 1988, 37, 785−789. (33) Becke, A. D. Density-Functional Thermochemistry 0.3. The Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652. (34) Stewart, J. J. P. Optimization of Parameters for Semiempirical Methods 0.1. Method. J. Comput. Chem. 1989, 10, 209−220. (35) Stewart, J. J. P. Optimization of Parameters for Semiempirical Methods 0.2. Applications. J. Comput. Chem. 1989, 10, 221−264. (36) Vreven, T.; Mennucci, B.; da Silva, C. O.; Morokuma, K.; Tomasi, J. The ONIOM-PCM Method: Combining The Hybrid Molecular Orbital Method And The Polarizable Continuum Model For Solvation. Application To The Geometry And Properties Of A Merocyanine In Solution. J. Chem. Phys. 2001, 115, 62−72. (37) Mo, S. J.; Vreven, T.; Mennucci, B.; Morokuma, K.; Tomasi, J. Theoretical Study Of The S(N)2 Reaction Of Cl-(H2O)+CH3Cl Using Our Own N-Layered Integrated Molecular Orbital And Molecular Mechanics Polarizable Continuum Model Method (ONIOM-PCM). Theor. Chem. Acc. 2004, 111, 154−161. (38) Mulliken, R. S. Electronic Population Analysis on Lcao-Mo Molecular Wave Functions 0.1. J. Chem. Phys. 1955, 23, 1833−1840. (39) Cerqueira, N. M. F. S. A.; Fernandes, P. A.; Eriksson, L. A.; Ramos, M. J. Dehydration Of Ribonucleotides Catalyzed By Ribonucleotide Reductase: The Role Of The Enzyme. Biophys. J. 2006, 90, 2109−2119. (40) Cerqueira, N. M. F. S. A.; Fernandes, P. A.; Eriksson, L. A.; Ramos, M. J. Ribonucleotide Activation By Enzyme Ribonucleotide Reductase: Understanding The Role Of The Enzyme. J. Comput. Chem. 2004, 25, 2031−2037. (41) Carvalho, A. T.; Fernandes, P. A.; Ramos, M. J. Theoretical Study Of The Unusual Protonation Properties Of The Active Site Cysteines In Thioredoxin. J. Phys. Chem. B 2006, 110, 5758−61. (42) Carvalho, A. T.; Fernandes, P. A.; Ramos, M. J. Determination of the DeltapKa between The Active Site Cysteines Of Thioredoxin And DsbA. J. Comput. Chem. 2006, 27, 966−75. (43) Carvalho, A. T. P.; Fernandes, P. A.; Ramos, M. J. The Catalytic Mechanism of RNA Polymerase II. J. Chem. Theory Comput. 2011, 7, 1177−1188. (44) Ramos, M. J.; Fernandes, P. A. Computational Enzymatic Catalysis. Acc. Chem. Res. 2008, 41, 689−698. (45) Oliveira, E. F.; Cerqueira, N. M.; Fernandes, P. A.; Ramos, M. J. Mechanism of Formation Of The Internal Aldimine In Pyridoxal 5′Phosphate-Dependent Enzymes. J. Am. Chem. Soc. 2011, 133, 15496− 505. 5800


Isomerization of Δ5-Androstene-3, 17-dione into Δ4-Androstene-3, 17

Recommend Documents