Efficient Binding of Flexible and Redox-Active Coenzymes by

Letter pubs.acs.org/acscatalysis

Efficient Binding of Flexible and Redox-Active Coenzymes by Oxidoreductases C. Satheesan Babu*,† and Carmay Lim*,†,‡ †

Institute of Biomedical Sciences, Academia Sinica, Taipei 11529, Taiwan R.O.C. Department of Chemistry, National Tsing-Hua University, Hsinchu 300, Taiwan R.O.C.

‡

S Supporting Information *

ABSTRACT: Although how an enzyme binds its substrate and rate-limiting transition state is well-studied, how it binds a flexible redox-active coenzyme, which may adopt different native solution conformations when oxidized/reduced, remains unclear. Using the coenzyme nicotinamide adenine dinucleotide (NAD) as an example, we show that redox enzymes bind oxidized NAD+ and reduced NADH in similar conformations that are between folded NAD+ and extended NADH solution conformations. By preorganizing the coenzyme-binding site to bind such an “intermediate” conformation, a redox enzyme can efficiently bind oxidized and reduced NAD, whereas having to substantially reorganize the binding site to accommodate the NAD product might impede catalysis. KEYWORDS: oxidoreductases, dehydrogenases, coenzyme binding, QM/MM methods, umbrella sampling, enzyme proficiency

S

ince Pauling’s hypothesis that enzyme active sites are complementary in structure to “the molecular configuration that is intermediate between the reacting substances and the products” of the catalyzed reaction,1 much progress has been made in understanding how an enzyme binds this transitionstate configuration better than the substrate, thus lowering the activation barrier in solution.2−7 In contrast, how an enzyme binds its coenzyme, which is often large and flexible, has received little attention. A redox-active coenzyme, owing to its multiple rotatable bonds, could adopt multiple conformations in water, but which of these conformations are lowest in free energy (i.e., native) remains unclear. Whether oxidized and reduced coenzymes adopt similar or different native solution conformations also remains unclear. If they prefer different native solution conformations, how can they fit in the same coenzyme-binding site without compromising the enzyme’s catalytic ability? To shed light on this intriguing problem, we have focused on nicotinamide adenine dinucleotide (NAD), an essential coenzyme for oxidoreductase enzymes that reversibly catalyze hydride transfer from the substrate to the NAD+ nicotinamide ring, yielding dianionic NADH and product (Figure 1).8 This coenzyme was chosen because the NAD+:NADH concentration ratio is a key indicator of the well-being of cells, and NAD analogues are potential anticancer,9 neurological,10 and antibacterial11 drug candidates. Furthermore, NADH dissociation from certain enzymes such as lactate dehydrogenase is rate-limiting.12 Molecular dynamics simulations13,14 and X-ray structures show NAD+ conformations to be compact in water and more extended in proteins.15−17 While the hydride transfer mechanisms of oxidoreductases have been extensively studied,12,18−22 how these enzymes bind both NAD+ and NADH © XXXX American Chemical Society

Figure 1. Schematic diagram depicting a redox reaction involving NAD. A hydride from the C atom of the alcohol substrate is transferred to the C4 atom of the NAD+ nicotinamide ring (A) to form NADH and aldehyde product (B). The reaction coordinates, RC1B−C1D, RC6A−C2N, and χO5B−O5D dihedral angle employed are shown. Positively charged moiety in blue and negatively charged pyrophosphate in red.

has not been addressed. This raises the following questions: (1) Since NAD+ and NADH differ by only a hydride atom, do they adopt similar solution conformations? (2) If not, how do they bind to the same active site in an enzyme? Received: March 14, 2016 Revised: April 17, 2016

3469

DOI: 10.1021/acscatal.6b00743 ACS Catal. 2016, 6, 3469−3472

Letter

ACS Catalysis

The solution free energy profiles show that NAD+ and NADH prefer different conformations in water with NADH adopting more extended conformations than NAD+. Reduced NADH has free energy minima at much longer RC1B−C1D and RC6A−C2N distances (12.5 and 17 Å) than oxidized NAD+ (8.5 and 11.5 Å). This shows that a single hydride atom difference between NAD+ and NADH can influence the NAD solution conformation, which is compact when oxidized but extended when reduced. To elucidate why NAD+ and NADH prefer different solution conformations, their compact and extended conformers were constrained-optimized in implicit solvent and the resulting geometries were used to compute the compact → extended isomerization energy in the gas phase and in water (see Supporting Information). For NAD+, favorable electrostatic interactions between cationic nicotinamide and dianionic pyrophosphate stabilize compact structures.27 As the NADH nicotinamide is no longer positively charged, repulsion among the negatively charged pyrophosphate oxygen atoms favors extended structures. Solvation effects do not change the conformational preference of NAD+ or NADH, but attenuate the preference for NADH to adopt extended conformations. The solution free energy profiles also reveal an “intermediate” conformation between the native NAD+ and NADH solution conformations, located at the reaction coordinate where the free energy curves for NAD+ and NADH cross. When the reaction coordinate is RC1B−C1D, the free energy curves for NAD+ and NADH intersect at RC1B−C1D ∼ 10.5 Å, midway between the native NAD+ (8.5 Å) and NADH (12.5 Å) minima positions (Figure 2A). When it is RC6A−C2N, the curves cross at RC6A−C2N ∼ 14.75 Å, between the RC6A−C2N of NAD+ (∼11.5 Å) and NADH (∼17 Å) solution minima (Figure 2B). This “intermediate” conformation at the intersection of the free energy curves is equally destabilized relative to the NAD+ or NADH minimum whose free energy was set to zero. To determine if NAD+ and NADH also prefer different conformations when bound to enzymes as in aqueous solution, and if their enzyme-bound conformations are folded, semiextended, or extended, we computed the frequency distributions of RC1B−C1D, RC6A−C2N, and χO5B−O5D in oxidoreductase structures containing the ligand NAD (denoting NAD+) or NAI (denoting NADH)28 in the Protein Data Bank29 (PDB). The PDB codes, RC1B−C1D, RC6A−C2N, and χO5B−O5D values of NAD in 383 NAD+-bound and 85 NADH-bound oxidoreductase structures are listed in Tables S1 and S2, respectively. The total number of enzyme-bound NAD+ or NADH structures with a given range of RC1B−C1D or RC6A−C2N are shown as histograms in Figure 2. In contrast to aqueous solution, NAD+ and NADH adopt similar semiextended conformations when bound to oxidoreductases: This is shown by the similar RC1B−C1D/RC6A−C2N distributions in NAD+ and NADH-bound enzyme structures in Figure 2 with median RC1B−C1D = 10.50 ± 0.95 Å and RC6A−C2N = 14.75 ± 1.60 Å characteristic of the NAD “intermediate” solution conformation (see above). These distributions are independent of the enzyme-bound ligand/substrate type and are robust to the number/resolution of the structures used: They exhibit similar peaks and shapes to those derived from high (≤2 Å) resolution NAD+-bound enzyme structures (Figure S2). Thus, most enzymes bind NAD in an “intermediate” high-energy solution conformation rather than the native one.

To address these questions, we computed the free energies of NAD solution conformers along various reaction coordinates (Figure 1): (i) the RC1B−C1D distance between the ribose C1D and C1B, describing the ribose ring separation; (ii) the RC6A−C2N distance between the adenine C6A and nicotinamide C2N, defining the base separation; and (iii) the χO5B−O5D torsion angle (C1B−O5B−O5D−C1D), describing the orientation of the ribose rings relative to the pyrophosphate. The solution free energies along these reaction coordinates were computed using QM/MM simulations with umbrella sampling. The entire coenzyme was treated using semiempirical PM3,23 as this method could efficiently reproduce the highresolution (0.5-Å) X-ray structure of the NAD acid state in the Cambridge Structural Database24 (see Figure S1); water molecules and sodium counterions were treated using the TIP3P potential25 and the CHARMM force field,26 respectively. Details of the QM/MM and umbrella sampling simulations are given in Supporting Information. Figure 2A,B show the solution free energies of NAD+ (black curves) or NADH (blue curves) relative to the respective lowest free energy, which was set to zero, as a function of RC1B−C1D and RC6A−C2N, respectively.

Figure 2. Correlation between NAD+/NADH conformational free energies in water and the respective populations in enzyme-bound Xray structures. The free energies (left y-axis) as a function of (A) RC1B−C1D, and (B) RC6A−C2N, were computed using the QM/MM umbrella sampling. When RC6A−C2N of NADH exceeded 18 Å, the C− N was broken, hence the corresponding free energies could not be obtained. The histograms represent the total number of enzymebound NAD+ or NADH structures whose RC1B−C1D or RC6A−C2N fall within a given interval (right y-axis) with the NAD+ fraction shown in gray and NADH fraction in blue. 3470

DOI: 10.1021/acscatal.6b00743 ACS Catal. 2016, 6, 3469−3472

Letter

ACS Catalysis That most enzymes bind NAD in a conformation that is intermediate between the native NAD+ and NADH solution conformations is also evident by extracting the NAD/NADH populations from the respective unbiased QM/MM simulations in water and comparing them with the populations of enzymebound NAD+/NADH conformers. The populations of NAD+/ NADH conformers as a function of RC1B−C1D and χO5B−O5D in the X-ray structures and in the unbiased simulations are normalized and displayed as a contour plot with increasing populations reflected by increasing color intensity in Figure 3.

Figure 4. Binding scheme of NAD to a typical dehydrogenase in a catalytic cycle. (A) Native (compact, syn) solution conformation of NAD+. (B) An apo enzyme (alcohol dehydrogenase, PDB code 1A4U) with a Rossmann fold element (blue); the red spheres represent the two phosphates, while the green arrows indicate the binding directions of adenine and nicotinamide. (C) The corresponding holo enzyme (PDB entry 1B14) with NAD + bound in a semiextended conformation. (D) Native (extended, anti) solution conformation of NADH after its release from the enzyme’s active site into water.

Although rare, a few enzymes (highlighted in bold in Table S1) bind NAD+ in highly compact conformations with RC1B−C1D and RC6A‑C2N < 7 Å. How these enzymes bind NADH is unclear because the NADH-bound structures have not been solved. In summary, NAD+ and NADH adopt quite different structures in water, but both fit in the enzyme’s active site in a semiextended conformation that is intermediate between the folded NAD+ and extended NADH solution conformers. By preorganizing the coenzyme-binding site to fit an “intermediate” NAD solution conformation, redox enzymes evenly distribute the reorganization energy cost for binding the compact NAD+ and extended NADH solution conformers. In the future, it would be interesting to determine the NAD conformation in the rate-limiting transition state for the overall enzymatic reaction and see if it also lies between the native NAD+ or NADH solution conformers. If so, an enzyme would need to expend minimal reorganization energy to bind the NAD conformation in the rate-limiting transition state. It would also be interesting to examine how the different NAD conformers affect not only the height but also the width of the reaction barrier.

Figure 3. (A) Contour diagram of the RC1B−C1D vs χO5B−O5D frequencies in free NAD+ (red) and NADH (blue) in water and enzyme-bound NAD+/NADH (green). The color intensity is proportional to the occurrence frequency, which ranges from 0 to 1. Snapshots of simulation structures corresponding to solution free energy minima for (B) NADH and (C) NAD+.

The results in Figure 3 show that the enzyme-bound NAD+/ NADH conformations (green) fall between the unbound native NAD+ (red) and NADH (blue) solution conformations: In water, NAD+ favors compact conformations with RC1B−C1D ∼ 8.5 Å and χO5B−O5D ∼ 0° (denoted as syn), but NADH prefers extended conformations with longer RC1B−C1D (12.5 Å) and χO5B−O5D between 90° and 150° (denoted as anti). However, in most enzymes, both NAD+ and NADH adopt semiextended “syn” conformations with RC1B−C1D = 10.5 Å and χO5B−O5D ∼ 0°. Even in crystal structures containing ligands/coenzymes in addition to NAD, the RC1B−C1D and χO5B−O5D values of the enzyme-bound NADH/NAD+ correspond to those of the “high-energy” semiextended solution conformer. The above results provide important insight as to how an enzyme binds its coenzyme in oxidized and reduced states. In aqueous solution, NAD+ strongly favors compact structures with “syn” conformations of the ribose rings (Figure 4A). The common NAD-binding motif known as the Rossmann fold30 is poised to attract such a “syn” NAD+ conformation with the pyrophosphate drawn to the helical dipole, while the two bases flank either side of this fold (Figure 4B). The enzyme binds NAD+ in a configuration between the reactant and product NAD solution conformations either directly (a rare event) or by stretching the compact NAD+ solution structure (Figure 4C). In this semiextended conformation, NAD+ accepts a hydride from the substrate, yielding dianionic NADH in a similar conformation, which is then released into solution where it adopts a more extended conformation (Figure 4D).

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.6b00743. Computational methods and PDB survey (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Y.-H. Hua, C. Grauffel, and Y.-L. Lin for technical help, and J. Straub for helpful discussions. Supported by funds 3471

DOI: 10.1021/acscatal.6b00743 ACS Catal. 2016, 6, 3469−3472

Letter

ACS Catalysis from Academia Sinica and MOST (Grant No. 103-2113-M001-007), Taiwan.

■

REFERENCES

(1) Pauling, L. Nature 1948, 161, 707−709. (2) Kohen, A.; Klinman, J. P. Acc. Chem. Res. 1998, 31, 397−404. (3) Bruice, T. C.; Benkovic, S. J. Biochemistry 2000, 39, 6267−6274. (4) Garcia-Viloca, M.; Gao, J.; Karplus, M.; Truhlar, D. G. Science 2004, 303, 186−195. (5) Zhang, X. Y.; Houk, K. N. Acc. Chem. Res. 2005, 38, 379−385. (6) Warshel, A.; Sharma, P. K.; Kato, M.; Xiang, Y.; Liu, H. B.; Olsson, M. H. M. Chem. Rev. 2006, 106, 3210−3235. (7) Hammes, G. G.; Benkovic, S. J.; Hammes-Schiffer, S. Biochemistry 2011, 50, 10422−10430. (8) Pollak, N.; Dolle, C.; Ziegler, M. Biochem. J. 2007, 402, 205−218. (9) Peralta-Leal, A.; Rodríguez-Vargas, J. M.; Aguilar-Quesada, R.; Rodríguez, M. I.; Linares, J. L.; de Almodóvar, M. R.; Oliver, F. J. Free Radical Biol. Med. 2009, 47, 13−26. (10) Araki, T.; Sasaki, Y.; Milbrandt, J. Science 2004, 305, 1010− 1013. (11) Fieldhouse, R. J.; Merrill, A. R. Trends Biochem. Sci. 2008, 33, 546−556. (12) Cui, Q.; Elstner, M.; Karplus, M. J. Phys. Chem. B 2002, 106, 2721−2740. (13) Smith, P. E.; Tanner, J. J. J. Am. Chem. Soc. 1999, 121, 8637− 8644. (14) Smith, P. E.; Tanner, J. J. J. Mol. Recognit. 2000, 13, 27−34. (15) Moodie, S. L.; Thornton, J. M. Nucleic Acids Res. 1993, 21, 1369−1380. (16) Stockwell, G. R.; Thornton, J. M. J. Mol. Biol. 2006, 356, 928− 944. (17) Kuppuraj, G.; Sargsyan, K.; Hua, Y.-H.; Merrill, A. R.; Lim, C. J. Phys. Chem. B 2011, 115, 7932−7939. (18) Andrés, J.; Moliner, V.; Safont, V. S.; Domingo, L. R.; Picher, M. T.; Krechl, J. Bioorg. Chem. 1996, 24, 10−18. (19) Radkiewicz, J. L.; Brooks, C. L. J. Am. Chem. Soc. 2000, 122, 225−231. (20) Wong, K. F.; Selzer, T.; Benkovic, S. J.; Hammes-Schiffer, S. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 6807−6812. (21) Basner, J. E.; Schwartz, S. D. J. Am. Chem. Soc. 2005, 127, 13822−13831. (22) Nagel, Z. D.; Klinman, J. P. Chem. Rev. 2006, 106, 3095−3118. (23) Stewart, J. J. P. J. Mol. Model. 2004, 10, 155−164. (24) Allen, F. H. Acta Crystallogr., Sect. B: Struct. Sci. 2002, B58, 380− 388. (25) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. J. Chem. Phys. 1983, 79, 926−935. (26) MacKerell, J. A. D.; Bashford, D.; Bellott, M.; Dunbrack, R.; Evanseck, J. D.; Field, M. J.; Fischer, S.; Gao, J.; Guo, H.; Ha, S.; Joseph-McCarthy, D.; Kuchnir, L.; Kuczera, K.; Lau, F. T. K.; Mattos, C.; Michnick, S.; Ngo, T.; Nguyen, D. T.; Prodhom, B.; Reiher, W. E. I.; Roux, B.; Schlenkrich, M.; Smith, J. C.; Stote, R.; Straub, J.; Watanabe, M.; Wiorkiewicz-Kuczera, J.; Yin, D.; Karplus, M. J. Phys. Chem. B 1998, 102, 3586−3616. (27) Dudev, T.; Lim, C. J. Am. Chem. Soc. 2010, 132, 16533−16543. (28) Hua, Y. H.; Wu, C. Y.; Sargsyan, K.; Lim, C. Sci. Rep. 2014, 4, 6471. (29) Berman, H. M.; Battistuz, T.; Bhat, T. N.; Bluhm, W. F.; Bourne, P. E.; Burkhardt, K.; Feng, Z.; Gilliland, G. L.; Iype, L.; Jain, S.; Fagan, P.; Marvin, J.; Padilla, D.; Ravichandran, V.; Schneider, B.; Thanki, N.; Weissig, H.; Westbrook, J. D.; Zardecki, C. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2002, 58, 899−907. (30) Wu, C. Y.; Hwa, Y.-H.; Chen, Y. C.; Lim, C. J. Phys. Chem. B 2012, 116, 5644−5652.

3472

DOI: 10.1021/acscatal.6b00743 ACS Catal. 2016, 6, 3469−3472