Molecular Recognition of Aromatic Rings by Flavin - American

Nov 15, 2013 - Department of Chemistry, University of Louisville, 1800 South Second Street, Louisville, Kentucky 40208, United States. ABSTRACT: Aroma...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCA

Molecular Recognition of Aromatic Rings by Flavin: Electrostatics and Dispersion Determine Ring Positioning above Isoalloxazine Lucas Koziol,† Neeraj Kumar,‡ Sergio E. Wong,† and Felice C. Lightstone*,† †

Physical and Life Sciences Division, Lawrence Livermore National Laboratory, 7000 East Avenue, Livermore, California 94550, United States ‡ Department of Chemistry, University of Louisville, 1800 South Second Street, Louisville, Kentucky 40208, United States ABSTRACT: Aromatic stacking interactions between isoalloxazine (ISA) of flavin and three prototypical aromatics (benzene, pyridine, chlorobenzene) were investigated using electronic structure calculations with Monte Carlo simulated annealing. The Effective Fragment Potential (EFP) method was used to locate the low-energy equilibrium configurations for the three dimer systems. These structures were further characterized through DFT (M06-2X) and MP2 calculations. One equilibrium configuration exists for ISA−benzene; characterizing the stacked dimer surface revealed a steep, single-welled potential that funnels benzene directly between rings II and III, positioning a substituent hydrogen adjacent to the redox-active N5. ISA−pyridine and ISA−chlorobenzene minimum-energy structures contain the aromatic ring in very similar position to that in ISA−benzene. However, the added rotational degree of freedom leads to two distinct binding motifs, having approximately antiparallel or parallel dipole moment alignment with ISA. The existence of the latter binding configuration was unexpected but is explained by the shape of the ISA electrostatic potential. Dispersion is the primary noncovalent interaction driving the positioning of aromatic rings above ISA, while electrostatics determine the orientation in dipole-containing substituted benzenes. The interplay of these interactions can be used to tune molecular recognition properties of synthetic redox cofactors, including positioning desired functional groups adjacent to the redox-active N5.

1. INTRODUCTION Flavin-containing enzymes (flavoenzymes) are ubiquitous in nature.1 They catalyze redox reactions in many important biochemical pathways including photosynthesis,2 DNA repair,3,4 electron transport,5,6 and bacterial cell-wall biosynthesis.7 The active site of all flavoenzymes contains flavin, a cofactor connected to apoenzyme primarily through noncovalent interactions. Flavins consist of isoalloxazine (ISA), a three-ringed, aromatic structure on which the redox chemistry occurs, and a ribose-containing side chain connected at the N10 position (Figure 1). These side chains differentiate riboflavin, flavin mononucleotide (FMN), and flavin adenine dinucleotide (FAD); the chemical structures of the different flavins are shown in Figure 1. ISAs are electron-deficient aromatic systems. Their redox cycle in flavins usually involves the nicotinamide cofactors NADH or NADPH.8 As electron pair donor, NAD(P)H, performs two-electron reduction via hydride transfer to ISAN5, generating a nonaromatic, reduced form of ISA that is highly reactive. Reduced ISA eventually transfers electrons to an enzyme substrate, regenerating the resting aromatic form shown in Figure 1. The mechanism for this reduction can vary; for example, the monooxygenase class of enzymes reduces O2 through an ISA-OOH adduct. Several flavoenzymes also © 2013 American Chemical Society

Figure 1. Chemical structures of the different flavins. Select atoms are labeled for discussion within the text; ISA rings are labeled I−III.

perform oxidation reactions (with NAD(P)+ as the most common electron pair acceptor), including amino acid oxidations and olefin-forming desaturations.8 Flavins also Received: July 19, 2013 Revised: November 13, 2013 Published: November 15, 2013 12946

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952

The Journal of Physical Chemistry A

Article

serve as step-down redox switches between obligate twoelectron donors (e.g., nicotinamide cofactors) and one-electron acceptors (e.g., iron−sulfur clusters and heme proteins). A number of experimental and theoretical studies have examined redox properties of flavin-containing enzymes and ISA models.9−12 The presence of nearby apoenzyme residues strongly affects the ISA redox potential, thus enabling the diversity of chemical reactions catalyzed by flavoenzymes in nature. Several common motifs for the modulation of redox potential have been found, including aromatic stacking interactions.13,14 A smaller amount of work has focused on examining the basic nature of noncovalent stacking interactions involving ISAs and other important molecules, despite their fundamental importance. For example, the importance of electrostatics and dispersion in π stacking as a function of substituent and relative orientation has been examined in detail by Sherrill,15−18 including in substituted benzene dimers as well as benzene−pyridine17 and benzene−indole.18 The design of synthetic enzymes targeting reactions not performed in nature will require a knowledge of the multiple noncovalent interactions that can occur between ISA and various molecular groups. The self-assembly of ISA-containing supramoleculer complexes utilizes their ability to form strong stacking interactions with aromatic rings.19,20 The key roles of flavoenzymes in several protein pathways makes their active sites important targets for therapeutic drugs that can address a wide variety of diseases. Identifying specific chemical motifs that maximize the interaction energy can aid in the design of new compounds with higher binding affinity and specificity to flavoenzyme targets.21,22 Molecular recognition and redox properties in flavins are closely interdependent, as highlighted by a recent review article.10 In this work, we focus on the latter properties by analyzing the quantum mechanical components of stacking interactions with prototypical aromatic rings. Aromatic stacking is one of two classes of noncovalent interactions known to govern redox and recognition properties in flavins (hydrogen bonding is the other one).10 We choose benzene, pyridine, and chlorobenzene as systems commonly incorporated into supramolecular scaffolds in bioengineering and drug design. We show that a single-welled potential energy surface funnels the benzene ring to a specific region directly above the CC bond separating rings II and III. This potential energy is composed largely of dispersion, with a smaller electrostatic contribution. The addition of a dipole moment onto the benzene ring leads to two distinct binding motifs, characterized by the dimer having either antiparallel or parallel dipole orientation relative to ISA. The concerted effects of dispersion and electrostatics controls the stable positioning of aromatic rings with respect to the ISA face. These interactions can be tuned to aid in the rational design of synthetic redox partners with electronaccepting groups positioned next to ISA-N5, to help direct the self-assembly of supramolecular complexes containing ISAs, or to modulate resulting redox properties.

Figure 2. Chemical structures, ESPs, and dipole moments (in Debye) for monomers studied in this work. Red (blue) ESP coloring corresponds to negative (positive) areas of molecular charge density. Left: ISA flavin model studied in this work. Right: Benzene (top), pyridine (center), and chlorobenzene (bottom).

generally too large to sample sufficiently with a small number of geometry optimizations, particularly when both monomers contain dipoles, and thus, in-plane rotations must be taken into account. EFP treats the monomer wave functions in terms of their electric multipoles and polarizabilities localized on atoms and bond midpoints. This allows analytical expressions to be derived to compute Eint. Importantly, Eint is expressed in terms of its individual components. There are four main interaction components in EFP; these are electrostatic (Coulomb), polarization (induction), dispersion, and exchange−repulsion. These terms and their definitions are extensively discussed in ref 24. In analogy with recent benchmarks of EFP, a chargetransfer term in the binding energy was not included.24,25 The charge-transfer interaction is expected to be negligible for aromatic stacking interactions due to the dominance of the dispersion interaction.24 After the initial cost of obtaining the monomer EFP potentials, the evaluation of Eint for a given dimer configuration is highly efficient (on the order of seconds even for large systems). EFP has been extensively benchmarked,26−28 including the S22 and S66 test sets,24 and its accuracy has been shown to be comparable to MP2 and dispersion-corrected DFT. In this work, all EFP monomer potentials (for ISA, benzene, pyridine, and chlorobenzene) were calculated using the 6-311(+,+)G(3df,2p) basis set. Distributed multipoles were generated using a numerical integration scheme. In order to account for charge-penetration effects, electrostatic screening was included at the level of charge−charge interactions, and polarization and dispersion were also screened. All EFP monomer and dimer calculations utilized M06-2X/aug-cc-pVTZ-optimized monomer geometries. Monte Carlo with simulated annealing (MC-SA) using EFP for Eint25,29 was performed to sample the potential energy surface and identify the global and (low-lying) local minima for ISA−benzene, ISA−pyridine, and ISA−chlorobenzene. MC-SA trajectories started from configurations randomized with respect to the xy (flavin) plane, with an initial interplanar distance of 3.0 Å. An MC step consisted of a three-dimensional translation and an in-plane rotation (keeping the monomers coplanar). The steps were scaled to ensure a translation distance of less than 0.2 Å and rotation angle less than 4°. The

2. COMPUTATIONAL DETAILS All calculations were performed on the ISA model shown in Figure 2. Noncovalent interaction energies Eint were computed for dimer configurations, Eint = Edimer − (EISA + EX), for X = benzene, pyridine, or chlorobenzene. A negative Eint corresponds to an attractive interaction. The Effective Fragment Potential (EFP) method23 for noncovalent interactions was used to sample the dimer configuration space. This space is 12947

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952

The Journal of Physical Chemistry A

Article

equilibrium configuration, shown in Figure 3. This global minimum has Eint = −13.29 kcal/mol. The individual Eint

initial temperature was set to 1000 K, and simulations were cooled to a final temperature below 100 K. One hundred MC steps were performed at each temperature, with a cooling factor of 0.9. A rectangular boundary extending at least 7 Å from ISA in each direction was employed to prevent the dimer from dissociating at high temperatures. For each system, a minimum of 60 independent MC-SA trajectories were performed. This corresponds to at least 132 000 total Eint evaluations for each of the three dimer systems. The final cooled structures from the MC-SA trajectories were subsequently used as starting guesses for EFP geometry optimizations to yield the final set of equilibrium structures. Here, the restriction of coplanarity was removed. For ISA− benzene, the coplanar potential energy surface was also mapped using grid points of 0.1 Å in the x and y coordinates. At each grid point, Eint was minimized along the intermonomer (z) coordinate using a Newton minimization scheme. Benzene rotation relative to ISA was not allowed and held fixed at the equilibrium angle found from MC-SA. Because this rotation is six-fold symmetric, it was not expected to have a significant effect on Eint. DFT and MP2 geometry optimizations were performed from final EFP equilibrium geometries in order to compare Eint for various methods. DFT calculations employed the M06-2X density functional30 with the aug-cc-pVTZ31 basis set for both geometry minimization and Eint. MP2 calculations utilized the aug-cc-pVDZ basis set for geometry minimization and the augcc-pVTZ basis for Eint. The use of a smaller basis for MP2 optimization was required due to the computational expense of obtaining MP2 first derivatives for systems of this size. Basis set superposition error was accounted for using the counterpoise correction method of Boys and Bernardi.32 DFT and MP2 geometry optimizations were performed using NWChem 6.1.33 All EFP calculations were performed using GAMESS.34 Dipole moments and electrostatic potentials (ESPs) for ISA, benzene, pyridine, and chlorobenzene were computed at the M06-2X/aug-cc-pVTZ-optimized geometries, using Hartree− Fock/6-311G(d,p) as implemented in Maestro.35

Figure 3. The calculated global minimum of the ISA−benzene stacked dimer.

components are listed in Table 1. The largest contribution is the dispersion energy, −17.28 kcal/mol, followed by exchange−repulsion, 8.05 kcal/mol. Computational studies have found that dispersion and exchange−repulsion generally both have very large magnitudes (>|Eint|) that cancel each other to a large extent.27 The large gain of dispersion over exchange− repulsion in this system (by more than −9 kcal/mol) is likely due to benzene’s position almost directly between rings II and III, allowing it to gain favorable dispersion interactions from two π systems. Similar equilibrium geometries have been reported, for example, in the indole−benzene system.18 Fused double rings may thus be an efficient way for positioning aromatic substrates through the dispersion force. The electrostatic contribution to Eint is −3.18 kcal/mol; the region of negative charge density above the benzene ring overlaps with the strongly positive region on the ISA between rings II and III (Figure 2). In order to examine electrostatic and dispersion stacking interactions over a larger region of the Eint potential energy surface, we mapped the parallel dimer surface along a fine (0.1 Å) evenly spaced grid in the x and y directions. At each grid point, the intermonomer distance (z) was optimized by Eint minimization, as described in the Computational Details. The resulting surface, as well as the individual energy components of this surface, is shown in Figure 4. Eint contains a single minimum, which is tightly centered around the CC bond between rings II and III (Figure 3), in agreement with the MCSA simulations. The electrostatic component of this surface shows one attractive region, which is shifted toward the center of ring III compared to Eint. This shift in the electrostatic interaction may originate from more favorable overlap that can be gained between two partially positive charged benzene hydrogens with the negative ESP above the ISA CO bonds. The dispersion component has the largest magnitude and is most negative in the same region as the Eint minimum. There also appears a strong, second dispersion minimum localized over the center of ring II. Thus, it is clear that the sum of electrostatics and dispersion leads to the narrow global Eint minimum. The exchange−repulsion surface is repulsive in all areas above the ISA atoms. The maximum in exchange repulsion is about 9 kcal/mol and occurs in approximately the same region as the dispersion minimum. Figure 4 shows that ISA utilizes both dispersion and electrostatic forces in tandem to recognize aromatic rings via a single-welled potential energy. Notably, this

3. RESULTS AND DISCUSSION 3.1. Monomer Geometries and ESPs. The ISA dipole and ESP are shown in Figure 2. ISA has a relatively large dipole moment of 9.6 D. The ESP contains two regions of negative charge density localized on the CO bonds of ring III and extending partially over N1 and C4a on ring II (Figure 2). A narrow region of positive charge density exists above N−H between the ring III CO bonds. Benzene, pyridine, and chlorobenzene ESPs are shown in Figure 2. Benzene has a region of negative charge density above the center of the ring, with hydrogens acquiring a partial positive charge. The pyridine negative charge density is strongly localized above N, reducing the negative area above the ring center compared to benzene. Chlorobenzene also contains a depleted negative charge density in the ring center relative to benzene. Recently, this depletion in substituted benzenes was shown to originate from additive, through-space effects of substituent lone pairs, rather than from an electron-withdrawing (through-bond) effect from the π ring.36 The chlorobenzene ESP contains a ring of negative charge density surrounding the Cl, with positive density above Cl in line with the C2 axis. Pyridine and chlorobenzene have similar dipole moments of 2.3 and 2.1 D, respectively. 3.2. ISA−Benzene Dimer. All MC-SA trajectories with subsequent EFP minimization converged to the same 12948

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952

The Journal of Physical Chemistry A

Article

Table 1. EFP Interaction Energies Eint and Components (kcal/mol) benzene pyridine (antipar) chlorobenzene (antipar) pyridine (par) chlorobenzene (par)

Eint

electrostatic

dispersion

XC repulsion

polarization

−13.29 −15.07 −17.01 −12.08 −15.24

−3.18 −4.70 −4.21 −2.14 −2.88

−17.28 −18.18 −21.03 −16.87 −20.33

8.05 8.58 9.01 7.58 8.64

−0.88 −0.76 −0.72 −0.66 −0.66

positioning of benzene aligns a hydrogen substituent adjacent to the redox-active ISA-N5 (Figure 3). 3.3. Monosubstituted Rings: ISA−Pyridine and ISA− Chlorobenzene. Pyridine and chlorobenzene both have nonnegligible dipole moments, leading to a larger dimer configuration space and the possibility to tune Eint through dipole−dipole interactions. Our MC-SA simulations found that both ISA−pyridine and ISA−chlorobenzene contain two distinct types of minimum-energy binding interactions. In both, the aromatic ring remains in a very similar position as ISA−benzene. One type has approximately antiparallel aligned dipole moments with respect to ISA, while the other type has approximately parallel dipoles, leading to a weaker electrostatic interaction, by 1.3 and 2.6 kcal/mol for chlorobenzene and pyridine, respectively. The antiparallel global minimum and a parallel configuration are shown for ISA−pyridine in Figure 5

Figure 5. The calculated global minimum (left) and a local minimum (right) of the ISA−pyridine stacked dimer.

Figure 6. The calculated global minimum (left) and a local minimum (right) of the ISA−chlorobenzene stacked dimer.

and ISA−chlorobenzene in Figure 6. We note that for both ISA−pyridine and ISA−benzene, each class sometimes contained a few different structures; these varied by small translations and rotations and had Eint within ±1.0 kcal/mol. We now consider the global minima (left structures in Figures 5 and 6). The Eint components are shown in Table 1. ISA−pyridine has Eint = −15.07 kcal/mol, a −1.8 kcal/mol difference (toward stronger binding) relative to ISA−benzene. The main part of this difference is due to the electrostatic contribution (−1.5 kcal/mol difference from ISA−benzene), compared to −0.9 and +0.5 kcal/mol from dispersion and exchange−repulsion, respectively. In ISA−chlorobenzene, the global minimum has Eint = −17.01 kcal/mol, a −3.7 kcal/mol difference from ISA−benzene and −1.9 kcal/mol difference from ISA−pyridine. Antiparallel ISA−chlorobenzene is thus the

Figure 4. A reduced-dimension parallel dimer potential energy surface of ISA−benzene. The color represents the corresponding contribution to the total binding energy, in kcal/mol. The x and y coordinates correspond to the location of the benzene molecule center-of-mass. The z coordinate (intermonomer distance) minimizes Eint. Top panel: Total binding energy. The position of the single energy minimum is indicated by the arrow. Middle three panels: Individual components of the Eint surface. Bottom panel: The location of ISA atoms in the xy plane. 12949

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952

The Journal of Physical Chemistry A

Article

most strongly bound of the three aromatic rings examined. The electrostatic contribution is −4.21 kcal/mol, which is stronger than that in ISA−benzene (by −1.0 kcal/mol) but weaker than that in ISA−pyridine (by +0.5 kcal/mol). This is possibly due to Cl positioning away from the ISA atoms (Figure 6), thus achieving less overlap with the positive region of the ISA ESP. The Cl positioning may be driven by the much stronger dispersion interaction, which tends to align the ring between rings II and III. ISA−chlorobenzene gains a significantly stronger dispersion interaction over either ISA−benzene (by −3.8 kcal/mol) or ISA−pyridine (by −2.9 kcal/mol). The presence of Cl on the benzene ring provides a significantly larger dispersion contribution to Eint than the electrostatic contribution due to dipole−dipole interactions. The strength of dispersion is closely tied to Cl’s higher polarizability (e.g., relative to pyridine N). At least in terms of flavin interactions, Cl can be considered more like a hydrophobic group than a polar one. We now compare the more weakly bound parallel-aligned structures of ISA−pyridine and ISA−chlorobenzene (right side, Figures 5 and 6). These configurations are less stable than the corresponding global minima by +3.0 kcal/mol in ISA− pyridine and +1.8 kcal/mol in ISA−chlorobenzene. The existence of these minima can be understood from analyzing the ISA ESP (Figure 2), which contains a narrow region of positive charge density above N3. This positive region has favorable overlap with negative charge density above pyridineN or chlorobenzene-Cl. Hence these minima are electrostatic in origin. The difference in the electrostatic energy from antiparallel to parallel is +2.6 kcal/mol for ISA-pyridine and +1.3 kcal/mol for ISA−chlorobenzene. The different EFP Eint components for all five structures are plotted in Figure 8. The polarization energy is relatively

Figure 8. Components of the EFP Eint for the five configurations.

respective potential energy surfaces, the five configurations detailed above were optimized using both the M06-2X density functional and MP2. Eint values for all five structures are listed in Table 2 and plotted in Figure 7. We note that the three Table 2. Calculated Eint from EFP, DFT, and MP2 (kcal/ mol) benzene pyridine (antipar) chlorobenzene (antipar) pyridine (par) chlorobenzene (par)

EFP

MP2

M06-2X

−13.29 −15.07 −17.01 −12.08 −15.24

−13.67 −15.58 −16.44 −12.86 −13.78

−8.77 −10.62 −10.86 −7.89 −8.74

methods predict the same E int ordering for all five configurations. The nonparallel error (NPE) is defined as NPE = max(ΔEint) − min(ΔEint) for the five configurations. NPE equals 2.31 kcal/mol between EFP and M06-2X, 1.84 kcal/mol between EFP and MP2, and 0.68 kcal/mol between MP2 and M06-2X. The nonplanarity involving EFP is largely due to EFP predicting a more strongly bound parallel ISA− chlorobenzene configuration than the other two methods. We also compare the interaction energies. EFP and MP2 show the best agreement, with a mean difference (MD) across the five structures of −0.07 kcal/mol and mean absolute difference (MAD) of 0.74 kcal/mol. M06-2X predicts a systematically less-negative Eint than either EFP or MP2 (MD = MAD = 5.16 kcal/mol with respect to EFP and 5.09 kcal/mol with respect to MP2). These curves are, too a large degree, parallel, as shown by their low NPE. MP2 is known to overbind dispersion, leading to systematic errors in dispersion-dominated systems,37 which may explain the systematic difference, at least between MP2 and DFT. EFP has comparative accuracy to MP2, as noted in other large π-stacked systems, for example, the styrene dimer.38 Overall, the agreement between these three methods in ordering the binding energy for the five configurations, including good agreement in absolute energies between EFP and MP2, establishes EFP as a reliable method for describing noncovalent interactions in ISA−aromatic dimers.

Figure 7. Eint at optimized configurations from EFP, M06-2X, and MP2.

constant and slightly attractive, varying between −0.88 to −0.66 kcal/mol. Exchange−repulsion increases with increasing dispersion; therefore, the two interactions balance themselves out to a large extent. Dispersion varies most going between different systems and depends less on the dipole orientation. In contrast, the electrostatic points in Figure 8 change most strongly going from antiparallel to parallel dipole alignment, while varying significantly less between pyridine and chlorobenzene. 3.4. Comparison of EFP Eint with DFT (M06-2X) and MP2. In order to compare different methods on their 12950

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952

The Journal of Physical Chemistry A

Article

(9) Breinlinger, E. C.; Keenan, C. J.; Rotello, V. M. Modulation of Flavin Recognition and Redox Properties through Donor Atom−π Interactions. J. Am. Chem. Soc. 1998, 120, 8606−8609. (10) Nandwana, V.; Samuel, I.; Cooke, G.; Rotello, V. M. Aromatic Stacking Interactions in Flavin Model Systems. Acc. Chem. Res. 2012, 46, 1000−1009. (11) Hermoso, J.; Mayoral, T.; Faro, M.; Gomez-Moreno, C.; SanzAparicio, J.; Medina, M. Mechanism of Coenzyme Recognition and Binding Revealed by Crystal Structure Analysis of Ferredoxin-NADP+ Reductase Complexed with NADP+. J. Mol. Biol. 2002, 319, 1133− 1142. (12) Umena, Y.; Yorita, K.; Matsuoka, T.; Kita, A.; Fukui, K.; Morimoto, Y. The Crystal Structure of L-Lactate Oxidase from Aerococcus viridans at 2.1 Anstrom Resolution Reveals the Mechanism of Strict Substrate Recognition. Biochem. Biophys. Res. Commun. 2006, 350, 249−256. (13) Breinlinger, E. C.; Rotello, V. M. Model Systems for Flavoenzyme Activity: Modulation of Flavin Redox Potentials through π-Stacking Interactions. J. Am. Chem. Soc. 1997, 119, 1165−1166. (14) Deans, R.; Niemz, A.; Breinlinger, E. C.; Rotello, V. M. Electrochemical Control of Recognition Processes: a ThreeComponent Molecular Switch. J. Am. Chem. Soc. 1997, 119, 10863− 10864. (15) Sinnokrot, M. O.; Valeev, E. F.; Sherrill, C. D. Estimates of the Ab Initio Limit for π−π Interactions: The Benzene Dimer. J. Am. Chem. Soc. 2002, 124, 10887−10893. (16) Ringer, A. L.; Sherrill, C. D. Substituent Effects in Sandwich Configurations of Multiply Substituted Benzene Dimers Are Not Solely Governed by Electrostatic Control. J. Am. Chem. Soc. 2009, 131, 4574−4575. (17) Hohenstein, E. G.; Sherrill, C. D. Effects of Heteroatoms on Aromatic π−π Interactions: Benzene−Pyridine and Pyridine Dimer. J. Phys. Chem. A 2009, 113, 878−886. (18) Geng, Y.; Takatani, T.; Hohenstein, E. G.; Sherrill, C. D. Accurately Characterizing the π−π Interaction Energies of Indole− Benzene Complexes. J. Phys. Chem. A 2010, 114, 3576−3582. (19) Caldwell, S. T.; Cooke, G.; Hewage, S. G.; Mabruk, S.; Rabani, G.; Rotello, V.; Smith, B. O.; Subramani, C.; Woisel, P. Model Systems for Flavoenzyme Activity: Intramolecular Self-Assembly of a Flavin Derivative via Hydrogen Bonding and Aromatic Interactions. Chem. Commun. 2008, 4126−4128. (20) McDonald, N. A.; Subramani, C.; Caldwell, S. T.; Zainalabdeen, N. Y.; Cooke, G.; Rotello, V. M. Simultaneous Hydrogen Bonding and π-Stacking Interactions between Flavin/Porphyrin Host−Guest Systems. Tetrahedron Lett. 2011, 52, 2107−2110. (21) Binda, C.; Newton-Vinson, P.; Hubalek, F.; Edmondson, D. E.; Mattevi, A. Structure of Human Monoamine Oxidase B, a Drug Target for the Treatment of Neurological Disorders. Nat. Struct. Biol. 2001, 9, 22−26. (22) Yang, Y.; Severin, A.; Chopra, R.; Krishnamurthy, G.; Singh, G.; Hu, W.; Keeney, D.; Svenson, K.; Petersen, P. J.; Labthavikul, P.; et al. 3,5-Dioxopyrazolidines, Novel Inhibitors of UDP-N-Acetylenolpyruvylglucosamine Reductase (MurB) with Activity against GramPositive Bacteria. Antimicrob. Agents Chemother. 2006, 50, 556−564. (23) Gordon, M. S.; Freitag, M. A.; Bandyopadhyay, P.; Jensen, J. H.; Kairys, V.; Stevens, W. J. The Effective Fragment Potential Method: A QM-Based MM Approach to Modeling Environmental Effects in Chemistry. J. Phys. Chem. A 2001, 105, 293−307. (24) Flick, J. C.; Kosenkov, D.; Hohenstein, E. G.; Sherrill, C. D.; Slipchenko, L. V. Accurate Prediction of Noncovalent Interaction Energies with the Effective Fragment Potential Method: Comparison of Energy Components to Symmetry-Adapted Perturbation Theory for the S22 Test Set. J. Chem. Theory Comput. 2012, 8, 2835−2843. (25) Smith, Q. A.; Gordon, M. S.; Slipchenko, L. V. Benzene− Pyridine Interactions Predicted by the Effective Fragment Potential Method. J. Phys. Chem. A 2011, 115, 4598−4609. (26) Smith, T.; Slipchenko, L. V.; Gordon, M. S. Modeling π−π Interactions with the Effective Fragment Potential Method: The

4. CONCLUSIONS MC-SA sampling was performed to locate low-lying equilibrium configurations on the ISA−benzene, ISA−pyridine, and ISA−chlorobenzene stacked dimer systems. For ISA−benzene, the energy surface was mapped to reveal the global shape of this interaction. The results present a unified picture of how flavins are able to recognize aromatic rings. Primarily, strong dispersion interactions funnel rings to a precise location above the ISA ring system. When these rings contain a dipole, electrostatic interactions discriminate between the various possible rotations. For the monosubstituted systems considered here, this leads to a meta substituent adjacent to the catalytic N5. Understanding the interplay between the interactions governing molecular recognition is important for developing general structure−function rules for these processes in flavins and other biological systems. In the case of ISAs, it is likely that interactions in synthetic cofactors can be tuned independently to position electron-donating functional groups close enough to ISA-N5 for redox chemistry to occur. These interactions also cast light onto the way that flavins and other enzyme active sites are able to recognize specific substrates among a range of diverse chemical groups in biochemical environments.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank the Defense Threat Reduction Agency for funding (CBM.THERB.02.11.LLNL.047). We also thank Livermore Computing for the computer time. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DEAC52-07NA27344.

■

REFERENCES

(1) Mansoorabadi, S. O.; Thibodeaux, C. J.; Liu, H. The Diverse Roles of Flavin Coenzymes: Nature’s Most Versatile Thespians. J. Org. Chem. 2007, 72, 6329−6342. (2) Gauden, M.; Yeremenko, S.; Laan, W.; van Stokkum, I. H. M.; Ihalainen, J. A.; van Grondelle, R.; Hellingwerf, K. J.; Kennis, J. T. M. Photocycle of the Flavin-Binding Photoreceptor AppA, a Bacterial Transcriptional Antirepressor of Photosynthesis Genes. Biochemistry 2005, 44, 3653−3662. (3) Weber, S. Light-Driven Enzymatic Catalysis of DNA Repair: A Review of Recent Biophysical Studies on Photolyase. Biochim. Biophys. Acta 2005, 1707, 1−23. (4) Epple, R.; Wallenborn, E.; Carell, T. Investigation of FlavinContaining DNA-Repair Model Compounds. J. Am. Chem. Soc. 1997, 119, 7440−7451. (5) Bruice, T. C. Mechanisms of Flavin Catalysis. Acc. Chem. Res. 1980, 13, 256−262. (6) Kurisu, G.; Kusunoki, M.; Katoh, E.; Yamazaki, T.; Teshima, K.; Onda, Y.; Kimata-Ariga, Y.; Hase, T. Structure of the Electron Transfer Complex Between Ferredoxin and Ferredoxin-NADP(+) Reductase. Nat. Struct. Biol. 2001, 8, 117−121. (7) Benson, T. E.; Filman, D. J.; Walsh, C. T.; Hogle, J. M. An Enzyme−Substrate Complex Involved in Bacterial Cell Wall Biosynthesis. Nat. Struct. Biol. 1995, 2, 644−653. (8) Walsh, C. Flavin Coenzymes: At the Crossroads of Biological Redox Chemistry. Acc. Chem. Res. 1980, 13, 148−155. 12951

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952

The Journal of Physical Chemistry A

Article

Benzene Dimer and Substituents. J. Phys. Chem. A 2008, 112, 5286− 5294. (27) Ghosh, D.; Kosenkov, D.; Vanovschi, V.; Williams, C. F.; Herbert, J. M.; Gordon, M. S.; Schmidt, M. W.; Slipchenko, L. V.; Krylov, A. I. Noncovalent Interactions in Extended Systems Described by the Effective Fragment Potential Method: Theory and Application to Nucleobase Oligomers. J. Phys. Chem. A 2010, 114, 12739−12754. (28) Slipchenko, L. V.; Gordon, M. S. Water−Benzene Interactions: An Effective Fragment Potential and Correlated Quantum Chemistry Study. J. Phys. Chem. A 2009, 113, 2092−2102. (29) Day, P. N.; Pachter, R.; Gordon, M. S.; Merrill, G. N. A Study of Water Clusters Using the Effective Fragment Potential and Monte Carlo Simulated Annealing. J. Chem. Phys. 2000, 112, 2063−2073. (30) Zhao, Y.; Truhlar, D. G. Density Functionals with Broad Applicability in Chemistry. Acc. Chem. Res. 2008, 41, 157−167. (31) Kendall, R. A.; Dunning, T. H., Jr.; Harrison, R. J. Electron Affinities of the First Row Atoms Revisited. Systematic Basis Sets and Wave Functions. J. Chem. Phys. 1992, 96, 6796−6806. (32) Boys, S. F.; Bernardi, F. The Calculation of Small Molecular Interactions by the Differences of Separate Total Energies: Some Procedures with Reduced Errors. Mol. Phys. 1970, 19, 553−566. (33) Valiev, M.; Bylaska, E. J.; Govina, N.; Kowalski, K.; Straatsma, T. P.; Van Dam, H. J. J.; Wang, D.; Nieplocha, J.; Apra, E.; Windus, T. L.; et al. NWChem: A Comprehensive and Scalable Open-Source Solution for Large Scale Molecular Simulations. Comput. Phys. Commun. 2010, 181, 1477−1489. (34) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S.; et al. General Atomic and Molecular Electronic Structure System. J. Comput. Chem. 1993, 14, 1347−1363. (35) Maestro, version 9.2; Schrödinger, LLC: New York, 2011. (36) Wheeler, S. E.; Houk, K. N. Substituent Effects in the Benzene Dimer Are Due to Direct Interactions of the Substituents with the Unsubstituted Benzene. J. Am. Chem. Soc. 2008, 130, 10854−10855. (37) Tkatchenko, A.; DiStasio, R. A., Jr.; Head-Gordon, M.; Scheffler, M. Dispersion-Corrected Moller−Plesset Second-Order Perturbation Theory. J. Chem. Phys. 2008, 131, 094106. (38) Adamovic, I.; Li, H.; Lamm, M. H.; Gordon, M. S. Modeling Styrene−Styrene Interactions. J. Phys. Chem. A 2006, 110, 519−525.

12952

dx.doi.org/10.1021/jp407193c | J. Phys. Chem. A 2013, 117, 12946−12952