Single-Amino Acid Modifications Reveal Additional Controls on the

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Single Amino Acid Modifications Reveal Additional Controls on the Proton Pathway of [FeFe]-Hydrogenase Adam James Cornish, Bojana Ginovska, Adam Thelen, Júlio Cosme Santos Da Silva, Thereza A. Soares, Simone Raugei, Michel Dupuis, Wendy J Shaw, and Eric L. Hegg Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.5b01044 • Publication Date (Web): 17 May 2016 Downloaded from http://pubs.acs.org on May 19, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Single Amino Acid Modifications Reveal Additional Controls on the Proton Pathway of [FeFe]-Hydrogenase Adam J. Cornish1,2‡, Bojana Ginovska3, Adam Thelen1, Julio C. S. da Silva4, Thereza A. Soares4, Simone Raugei3, Michel Dupuis3‡‡, Wendy J. Shaw3*, Eric L. Hegg1,2*

1

Department of Biochemistry & Molecular Biology, Michigan State University, East Lansing,

MI 48824, U.S.A. 2

Great Lakes Bioenergy Research Center, Michigan State University, East Lansing, MI 48824,

U.S.A. 3

Physical Sciences Division, Pacific Northwest National Laboratory, Richland, WA 99352

4

Department of Fundamental Chemistry, Federal University of Pernambuco, Cidade

Universitária 50740-560, Recife, PE, Brazil

*

Correspondence to:

Eric L. Hegg, 510A Biochemistry, Michigan State University, East Lansing, MI 48824-1319. Phone: (517) 353-7120; Fax: 517-353-9334; E-mail: [email protected] Wendy Shaw, Pacific Northwest National Laboratory, PO Box 999, Richland, WA 99352, Phone: 509-375-5922, E-mail: [email protected]

‡

Current address: Department of Physiology, Johns Hopkins University

‡‡

Current address: University at Buffalo, Department of Chemical and Biological Engineering

ACS Paragon Plus Environment

1

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

ABBREVIATIONS DFT, density functional theory; QM/MM, quantum mechanics/molecular mechanics; MV, methyl viologen; MD, Molecular Dynamics; SEM, standard error of the mean


2

Page 3 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

ABSTRACT The proton pathway of [FeFe]-hydrogenase is essential for enzymatic H2 production and oxidation and is composed of four residues and a water molecule. A computational analysis of this pathway in the [FeFe]-hydrogenase from Clostridium pasteurianum revealed that the solvent-exposed residue of the pathway (Glu282) forms hydrogen bonds to two residues outside of the pathway (Arg286 and Ser320), implicating that these residues could function in regulating proton transfer. In the present study, we show that substituting Arg286 with leucine eliminates hydrogen bonding with Glu282 and results in a ~3-fold enhancement in H2 production activity when methyl viologen is used as an electron donor, suggesting that Arg286 may help control the rate of proton delivery. In contrast, substitution of Ser320 with alanine reduces the rate approximately 5-fold, implying that it either acts as a member of the pathway or influences Glu282 to enable proton transfer. Interestingly, QM/MM and molecular dynamics calculations indicate that Ser320 does not play a structural role nor does it indirectly influence the barrier for proton movement at the entrance of the channel. Rather, it may act as an additional proton acceptor for the pathway or serve in a regulatory role. While further studies are needed to elucidate the role of Ser320, collectively these data provide insights into the complex proton transport process.


3

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 38

INTRODUCTION Hydrogenases catalyze the reversible reduction of protons to molecular hydrogen and are critical to anaerobic metabolism in a variety of microorganisms. [FeFe]-hydrogenases, which utilize a unique diiron active site,1-2 have garnered particular interest due to their high catalytic rates of H2 production and their associated promise in fuel cell and electrolyzer applications.3-4 The catalytic production of H2 exhibited by [FeFe]-hydrogenases is dependent on, among other factors, the efficient transport of protons between the enzyme surface and the active site.5 Therefore, obtaining a comprehensive mechanistic understanding of intraprotein proton transfer is essential in engineering [FeFe]-hydrogenases to optimize H2 production, as well as in designing efficient biomimetic H2-forming catalysts. Proton transfer in proteins is accomplished via hydrogen-bonding interactions utilizing water wires or residue side chains.5 Examination of the Clostridium pasteurianum [FeFe]hydrogenase crystal structure identified three water-based pathways that could participate in proton transport,2, 6-7 although both mutagenesis studies8 and theoretical calculations9 indicated that two of these pathways are unlikely to be major pathways. The third pathway has been specifically proposed to deliver the second proton during H2 evolution, opening only upon formation of the metal hydride species,7, 10 although direct biochemical studies are needed to test this hypothesis. The positions of residues within the proposed pathways are shown as color coded spheres in Figure 1 of Ginovska-Pangovska et al.,9 or in Sode et al.10 Further analysis of the crystal structure identified a fourth potential proton transfer pathway consisting of two glutamates (Glu279 and Glu282), a serine (Ser319), a cysteine (Cys299), and a water molecule (H2O-612) (C. pasteurianum HydA numbering; Fig. 1).1 Together, these residues connect the active site to the enzyme surface and are strictly conserved in all sequenced [FeFe]-


4

Page 5 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

hydrogenases.8 Subsequent site-directed mutagenesis studies confirmed the importance of these residues in hydrogenase activity and provided strong evidence for their role in proton transfer,8 conclusions further supported by multiple molecular dynamics studies.6, 9, 11 Recently Ginovska-Pangovska et al.9 computationally evaluated the hydrogen bonding interactions in the proton transport pathway of [FeFe]-hydrogenases, testing the hypothesis that a well-defined hydrogen bonding network is necessary for facile proton transfer.9 Interestingly, this study suggested that Glu282, which was proposed to serve as the interface between the proton transfer network and the solvent,8-9, 11 was only transiently hydrogen bonded with the rest of the pathway. Instead, Glu282 was found almost exclusively in a salt bridge with Arg286 along the enzyme surface. This computational investigation also indicated that this interaction with Arg286 is disrupted when Glu282 is protonated. Interestingly, the loss of the salt bridge does not result in an enhancement of the hydrogen bonding between Glu282 and Ser319 (an integral proton transport residue). Instead, protonation of Glu282 is found to form a hydrogen bond with Ser320 with ~60% occupancy, a residue not previously proposed to impact the proton transfer pathway. These computational studies suggested that both Arg286 and Ser320 might influence proton transport, and elucidating the roles of these residues could provide insight into the structural framework required for optimized proton transfer. In this manuscript, the potential roles of the surface-exposed Arg286 and Ser320 residues in the H2 production activity of the C. pasteurianum [FeFe]-hydrogenase were explored using both biochemical and computational methods. Substitution of Arg286 with leucine (R286L) caused a significant increase in hydrogenase activity when methyl viologen (MV) was used as the electron donor, while substituting Ser320 for alanine (S320A) dramatically decreased activity. Computational analyses for hydrogen bonding, geometry, water interactions, and the


5

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 38

barrier to proton movement were utilized to further probe the role of these residues in the proton transfer process and their relative importance to catalysis.

MATERIALS AND METHODS Overexpression Plasmid Construction—A modified pAC-BAD plasmid containing the C. pasteurianum hydA1 gene was generated as described previously.8 Specific amino acid codons were altered by site-directed mutagenesis PCR (pfu turbo, Stratagene) using the primers listed in Table S1.8 Plasmids containing the mutated genes were isolated, sequenced to confirm the mutation, and transformed into electrocompentent Shewanella oneidensis MR-1 ∆hydA∆hyaB cells.8 Transformed cells were selected for resistance to 50 µg/mL kanamycin sulfate.

Cell Growth, Gene Overexpression, and Purification—Cells transformed with pAC-BAD containing either a wild-type or a mutated hydA1 gene were grown in 1 L of LB media, followed by induction and purification as described by Cornish et al.8 All purification steps were performed in the absence of O2 using an anaerobic chamber (Coy Laboratory Products) or airtight centrifuge bottles (Nalgene). The proteins were purified using anaerobic Ni-NTA resin (Qiagen) in 5 mL columns and eluted in 1 mL fractions using 100 mM imidazole. A qualitative assessment of hydrogenase activity was performed by separately adding 20 µL of each elution fraction to 20 µL of a 10 mM benzyl viologen dichloride solution in a H2-rich environment and monitoring the change in benzyl viologen from colorless to purple as it is reduced during H2 oxidation. The second elution fraction consistently turned a deeper purple color than the other fractions and was used for subsequent assays. The protein concentration of each elution fraction


6

Page 7 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

was determined by a BCA Protein Assay Kit (Pierce), comparing the measured absorbance to a BSA standard curve.

Hydrogenase Activity Assays—H2 evolution assays were modified from Cornish et al.,8 utilizing 100 mM potassium phosphate buffer (pH 6.8) for all reactions. The reaction was prepared in a 10 mL serum vial (Wheaton) containing 1.9 mL phosphate buffer and 100 mM Na2S2O4. Assays with an abiotic electron donor included 10 mM methyl viologen dichloride (MV) in the reaction mixture, while assays with a physiological electron donor were performed using 24 µM C. pasteurianum ferredoxin (accession number AAA83524.1). Ar was used to purge the headspace to remove N2 and residual O2 before 0.1 mL of protein sample (~0.1 mg/mL protein) was injected into the vial to initiate the reaction. The vial was shaken at ~100 rpm at 25 ˚C throughout the assay. Every 3 min for 15 min, 50 µL of headspace gas was removed from the reaction vial and injected into a TRACE GC Ultra gas chromatograph (Thermo Scientific) equipped with a capillary molecular sieve column.12 The resulting peak area at 1.4 min was compared to a standard curve to calculate the µmol of H2 produced. Hydrogen evolution activity was measured as the change in µmol of H2 per mg of protein over time. Chemical rescue with sodium azide and Zn2+-inhibition assays were performed as previously described by Cornish et al.8 Briefly, the enzymes were incubated with 0, 62.5, 125, 250, or 500 mM sodium azide in standard H2 production activity assay conditions (100 mM sodium dithionite and 10 mM MV) for chemical rescue experiments. Zn2+ inhibition was assessed by incubating the enzymes with 0 or 500 µM ZnCl2, followed by standard H2 production activity measurements.


7

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 38

Structural Evaluation Using Classical Molecular Dynamics (MD) Classical molecular dynamics simulations were performed for the native [FeFe]hydrogenase and two amino acid-substituted variants (R286L and S320A) with the enzyme in a protonation state configured for H2 evolution, where Glu279 and Glu282 residues are protonated (Fig. 1),9 and using the C. pasteurianum crystal structure (Protein Databank Bank code: 3C8Y)1 as the starting structure. The active site was modeled in the oxidation state Fe(I)Fe(I), with an overall charge of -4. The details of the classical model of the H-cluster are in the Supporting Information. The enzyme was solvated in a cubic water box with 25,453 water molecules and 14 Na+ counter ions that neutralize the charge. The force field parameters for the standard residues were taken from the CHARMM force field,13 and the parameters for the non-standard residues from the work of Chang and coworkers,14 adjusted as described previously9 and in the Supporting Information. With the exception of Glu282 and Glu279, all Asp and Glu residues were not protonated and all His residues were represented with a proton on δ-N, based on pKa values obtained from customary estimates via the Poisson-Boltzmann equation. This choice was found to reproduce a number of experimental properties from vibrational spectroscopy and the crystal structure, including vibrational frequencies, bond lengths, bond angles, and dihedrals, providing confidence in the appropriateness of the model employed for these studies.11, 14 All simulations were performed using the NAMD simulation package.15 The following simulation protocol was used: (1) the initial geometry of the systems were optimized using a conjugate gradient approach; (2) the optimized structures were gradually heated by carrying out 100–250 ps-long equilibrations at increasingly higher temperatures from 0 K to 250 K in increments of 50 K, followed by a 10 ns equilibration at 300 K; (3) trajectories were collected for 80 to 100 ns. All simulations were run at constant pressure (1 atm), constant temperature (298 K),16 and with a


8

Page 9 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

time step of 2 fs. As demonstrated in our previous publication,9 this simulation protocol reproduces the beta factors of the protein residues.

Analysis of the MD Trajectories Hydrogen bonding was analyzed using the VMD program.17 A distance cutoff of 3.5 Å between the hydrogen bond donor and the hydrogen bond acceptor was used to define an occupied hydrogen bond, along with an angle cutoff ∠(D-H…A) of ≥120 degrees, where D is the hydrogen bond donor and A is the hydrogen bond acceptor. Backbone structural studies were performed based on the pairwise distances between the Cα atoms between Glu279, Glu282, Arg286, Ser319, and Ser320. Orientation of the side chain residues was determined by the distance between the last carbon on arginine, leucine, alanine, and glutamic acid, or the oxygen on serine. The analysis of water was performed by counting the number of water molecules hydrogen bonding with the side chains of the residues.

Proton Transport Using Hybrid QM/MM Simulations Hybrid quantum mechanics/molecular mechanics (QM/MM) Born-Oppenheimer, umbrella sampling18 molecular dynamics simulations were employed to study two proton transfer processes. The first process investigated was the proton transfer between Glu282 and Glu279 via Ser319. The second proton transfer process studied was between Glu282 and a hydronium ion at the mouth of the channel via Ser320. In these simulations, the side chains of Glu282, either Ser319 or Ser320, and one water molecule were treated at the QM level, while the rest of the enzyme and the solvent were described classically, using the same force field parameters that were used for the classical molecular dynamics simulations described above. For


9

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 38

the Glu residues, the QM region was truncated between the Cα and Cβ carbon atoms, and for Ser residues between the Cα and O atoms. The bonds were capped in the QM region with H atoms. Simulations were performed within the density functional theory (DFT) framework using the PBE exchange and correlation functional,19-20 augmented with Grimme’s correction for the dispersion energy (PBE+D2),21 using a hybrid Gaussian and plane waves framework22-23 and norm-conserving pseudopotentials24 to describe the interaction between valence and core electrons. For these calculations, we used a triple-zeta basis set, TZVP (double-zeta, DZVP, for Fe), optimized for reproducing intermolecular gas-phase and condensed phase energy of molecular systems,25 along with an auxiliary plane wave basis set with an energy cut-off of 280 Ry. The QM/MM simulations were performed using the CP2K package22 in the canonical ensemble16 using an integration time step of the equation of motion of 0.25 fs. Umbrella sampling is an enhanced sampling technique for free energy calculations where sequences of MD simulations are performed by restraining the reaction coordinate along a chosen reaction pathway. For the proton transfer event from Glu282 to Glu279 via Ser319, the reaction coordinate in our simulation was described in terms of a linear combination (r = r1 - r2) of the O-H bond on Ser319 (r1) and the O-H distance of the O on Ser319 and H on Glu282 (r2). The reactant state is located at r = 0.5 Å, and the product state is located at r = -0.5 Å. A total of nine umbrella sampling runs were performed, equally spaced in the r dimension (∆r = 0.2 Å) spanning a range of -0.9 Å to 0.7 Å. For the proton transfer event from hydronium ion to Glu282 via Ser320, the reaction coordinate was described in terms of a linear combination (r = r3 + r4), where r3 is the distance between the O atom of Ser320 and the proton on Glu282, and r4 is the distance between the H atom of Ser320 and the O atom of the water molecule. The value of the reaction coordinate is 2.0 Å when the proton resides on the hydronium ion, and it is 3.5 Å when


10

Page 11 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

it resides on Glu282. The sampling was performed in nine equally spaced sampling windows spanning a range of 2.0 to 3.5 Å. For both simulations, a force constant of 0.1 Hartrees/Bohr2 was used as a harmonic restraint. Each trajectory was equilibrated for 2.5 ps and statistics were collected from runs spanning 4.5 to 5.0 ps at a time step of 0.25 fs. The trajectories were analyzed with the Weighted Histogram Analysis Method,26-27 extracting the Potential of Mean Force and calculating the free energy for proton transfer. The starting configurations for the QM/MM umbrella sampling simulations were prepared as follows: 1) An equilibrated structure of the enzyme and the solvent was selected from the equilibrated classical MD trajectory; the structure was selected with hydrogen bonds present between Glu282 and Ser319 and between Ser319 and Glu279, and between Glu282 and Ser320 and between Ser320 and the H2O molecule. 2) The structure was minimized using the simulated annealing approach. The classical region was equilibrated at a constant volume and a constant temperature (300 K) for 2 ns, while the QM region was kept constrained to maintain the hydrogen bonding interactions for the proton transfer study. The velocity rescaling approach was used to reach the correct temperature of the system.16 3) Once the MM region was equilibrated, the constraints were removed from the QM region and the full system was allowed to equilibrate at 300 K for 2.0 ps at a constant volume and constant temperature using a Nose-Hoover thermostat.28-29 This equilibrated system was used as the initial configuration for the umbrella sampling simulations.

Assessment of the DFT Accuracy for Proton Transport The uncertainty of the results due to the DFT level of theory used in the QM/MM calculations was evaluated by carrying out post-Hartree-Fock MP2 calculations for a model system


11

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 38

containing the three residues in the QM region. The positions of the Cα were fixed and transition states were optimized in the gas phase. The MP2 calculations were performed in the gas phase and employed an aug-cc-pVQZ basis set30-31 for all atoms except the protic H, for which we used 6-311G**.32 Harmonic corrections to free energies were adopted. As expected, DFT underestimates the barrier for the proton transfer. Nevertheless, the calculated MP2 free energy barrier to reaction is 8.9 kcal/mol and is within reasonable agreement with a value of 4.4 kcal/mol obtained using the setup at the DFT/PBE level of theory. The Gaussian09 program was employed for this benchmark.33

RESULTS Multiple Sequence Alignment—Glu282 is directly involved in proton transfer based on previous studies and is strictly conserved in all sequenced [FeFe]-hydrogenases.8 In addition to interacting with Ser319 as part of the proton pathway, the crystal structure1 reveals that the side chain of Glu282 is also within hydrogen bonding distance of the Arg286 side chain and is near to Ser320 (Fig. 2). Furthermore, MD calculations suggest that deprotonated Glu282 forms a favorable salt bridge with Arg286, while protonated Glu282 forms a H-bond interaction with Ser320 at 60% occupancy.9 Importantly, a multiple sequence alignment of >900 hydrogenase amino acid sequences indicates that Arg286 is also highly conserved (Fig. S1) and is very rarely substituted with a lysine, which maintains a positive charge at this position. Residue 320 contained a hydroxyl group in 100% of the sequences analyzed, with serine observed in 53% of the sequences and threonine observed in 47% of the 956 sequences analyzed. The observed conservation of Arg286 and the hydroxyl group at position 320, as well as their ability to interact


12

Page 13 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

with the surface-exposed Glu282 in the proton transport pathway, suggests that both Arg286 and Ser320 have the potential to modulate proton transfer.

Amino Acid-Substituted Enzyme Activity Assays—The positively-charged Arg286 is exposed to solvent and is hypothesized to form a salt-bridge with the nearby Glu282. In an effort to eliminate the putative salt-bridge and alter proton transfer, Arg286 was substituted with leucine to generate the R286L variant. Using MV as an abiotic electron donor, both native and variant enzymes were tested for hydrogenase activity. The R286L variant demonstrated a ~3-fold increase in H2 evolution activity compared to the native enzyme (Fig. 3; p ≤ 0.01), suggesting that the salt bridge between Arg286 and Glu282 limits the rate of H2 production when MV is used as an abiotic electron donor. The apparent increase in H2 evolution activity upon substituting Arg286 with leucine poses an interesting question: why has an activity-enhancing amino acid substitution not been favored by evolution? It is important to note that the use of an abiotic electron mediator allows for rapid electron transfer, presumably resulting in the H2 production rate being largely dominated by changes in the rate of proton transfer.1 In a biological system using a physiological donor, however, electron transfer is often rate-determining, removing evolutionary pressure for the R286L substitution. To test this hypothesis, H2 production assays were performed using C. pasteurianum ferredoxin as an electron mediator, and the activity of the native enzyme with ferredoxin was ~3-fold less than with MV (0.34 vs. 1.0 µmol/mg/s, respectively). As predicted, the R286L variant did not exhibit increased activity relative to the native enzyme when

1

In this instance, the rate of proton transfer could include any step in the catalytic cycle, including both individual steps within the proton pathway or proton transfer events at the active site.


13

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 38

ferredoxin was used as the electron donor (Fig. 3), indicating that the leucine substitution did not have a stimulatory effect on H2 production under natural conditions when the delivery of electrons was the rate-limiting step. H2 evolution activity was also determined for the S320A variant using either MV or ferredoxin as the electron donor. In both cases, H2 production activity in the S320A variant was dramatically decreased relative to the native enzyme (Fig 3; p ≤ 0.01). There was a nonsignificant difference between the specific activities of the S320A variant with MV and ferredoxin (0.13 ± 0.06 µmol/mg/s vs. 0.07 ± 0.015 µmol/mg/s, respectively). These results suggest that this substitution reduces activity regardless of the electron transfer agent (i.e. proton transfer becomes rate-limiting) and that Ser320 may play an important role in [FeFe]hydrogenase activity.

Sodium Azide Rescue Assays—Sodium azide is a common chemical utilized to rescue activity in proton transfer deficient variant enzymes.34-35 To evaluate the direct impact of substituting Arg286 and Ser320 on proton transfer, sodium azide was included in H2 production assays for the native and variant enzymes (Fig. 4). The activities of neither the native [FeFe]-hydrogenase nor the R286L variant were stimulated by treatment with 500 mM sodium azide relative to untreated enzyme. In contrast, a two-fold increase in activity was observed for the S320A variant under these same conditions, suggesting that this residue may play a critical role in proton transport.

Zinc Inhibition Assays—Divalent metal cations, such as Zn2+, have been previously demonstrated to bind to and occlude solvent-exposed proton-binding sites in several enzymes.36-37 Previous


14

Page 15 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

work demonstrated that Glu282 is essential for Zn2+-based inhibition in the C. pasteurianum [FeFe]-hydrogenase,8 but the impact of other residues on the interaction between Glu282 and Zn2+ had not been assessed. H2 production activity was measured in the presence of 500 µM ZnCl2, and both the native enzyme and the S320A variant exhibited similar inhibition in the presence of Zn2+, 34% and 28%, respectively (Fig. 5). These results indicate that the S320A variant did not affect the exposure of Glu282 to the solvent. Conversely, the R286L variant was inhibited by 53% in the presence of ZnCl2. This suggests that eliminating the salt bridge between Arg286 and Glu282 further exposed Glu282 to solvent and may explain the enhanced H2 evolution activity observed in the R286L variant.

Structural Modulation in the Variant Proteins—Analysis of the classical MD simulations indicated that the structure of the protein backbone was unchanged in the R286L and S320A variants, determined by monitoring the pairwise distances separating the Cα atoms between Glu282, Arg286, Ser319, and Ser320 (Fig. S2). The space between the side chains, however, decreased as a function of both R286L and S320A substitutions, measured pairwise as a distance between the last carbon on arginine, leucine, alanine and glutamic acid, or the oxygen on serine (Fig. 6). Specifically, the average space between the side chains of Glu282-Ser320 and Glu282Ser319 was decreased by ~30% in both variants, with reduced fluctuations (expressed as standard deviations about the average) in both cases. In contrast, the distance between the Glu282-Arg286 side chains only decreased for the S320A variant (30%), with reduced fluctuations, while no change was observed in the R286L variant. As the side chains of arginine and leucine are of different lengths, it was initially surprising that the distance between Glu282 and the residue at position 286 was unchanged when Arg286 was substituted with leucine. This


15

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 38

disparity is due to a change in the orientation of the Glu282 side chain in the R286L variant, which results in a stronger hydrogen bond between Ser319 and Glu282 and positions Glu282 closer to the 286 side chain in the R286L variant (Fig. S3). In the native enzyme, the Ser319Glu282 hydrogen bond is less persistent, allowing the Glu282 side chain more flexibility such that it moves further from Arg286. Interestingly, in the R286L variant (which shows increased activity with MV), both the orientation of Glu282 and the Ser319-Glu282 hydrogen bond are similar to the S320A variant (which shows reduced activity relative to the native enzyme). Thus, overall, there does not appear to be a strong correlation between hydrogenase activity and the calculated distances between Glu282 and the side chains for Arg286, Ser319, and Ser320. It is possible that differences in water accessibility around Glu282 in the variants relative to the native enzyme could result in the observed changes in activity. The number of water molecules that participate in hydrogen bonding with Glu282 were calculated from the MD simulations of the two mutants and were found to be similar to those for the native enzyme (Fig. S4). This finding suggests that neither the decrease in activity in S320A nor the increase of activity in R286L is caused by variation in water content near Glu282. A dramatic increase in hydrogen bonding occupancy was computationally determined between Glu282 and Ser319 in both variants, while a concomitant general decrease in occupancy was noted between Glu279 and Ser319 (Fig. 7). These alterations in hydrogen bond occupancy, however, do not correlate with changes in activity. Additionally, no significant difference was observed in the hydrogen bond occupancy between Glu282 and Arg286 in the S320A variant (~13%) relative to the native enzyme (9%) (Fig. 7). This result is consistent with the similar accessibility of Zn2+ to Glu282 in both the native protein and the S320A variant observed in the zinc inhibition assays. This


16

Page 17 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

observation is in contrast to what was observed with the R286L variant, where the hydrogen bonding was completely lost and accessibility of Zn2+ to Glu282 increased. Barrier of proton transfer at the entrance of the pathway —To test the hypothesis that Ser320 modulates proton transfer by altering the electronic properties of the other known residues in the pathway, QM/MM simulations were performed to evaluate proton transfer between Glu282 and Glu279 via Ser319 in both the native enzyme and the S320A variant. QM/MM simulations suggest that proton transfer is fast, synchronous, and indistinguishable for both proteins, with estimated activation free energies of 2.5 ± 0.5 for the native enzyme and 2.6 ± 0.5 kcal/mol for the S320A variant, respectively (Fig. S5). The lack of distinction in calculated energy barriers indicates that the decreased activity in S320A was not due to changes in the orientation of nearby residues. Additionally, these low values suggest that, for both native and variant proteins, proton movement is very facile between Glu282 and Glu279 and that this barrier is not limiting proton transfer. To explore the possibility that Ser320 is directly involved in the proton pathway, QM/MM simulations were utilized to calculate the free energy barriers to transfer a proton from a H3O+ to Glu282, either directly or via Ser320. The simulations show that proton transfer directly from H3O+ to Glu282 is a spontaneous reaction, observed on the picosecond timescale in the QM/MM simulation without any constraints. This observation implies barrierless transfer, and it was deemed unnecessary to employ free energy calculations for this transfer. The free energy calculations for the H+ transfer from H3O+ to Glu282 via Ser320 also indicate that this is a barrierless process, demonstrating there is no significant difference in energy, whether the proton is transferred through the serine or is transferred directly from water to Glu282. It is important to note that the H3O+ was solvated by classical water molecules, which limited our


17

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 38

ability to quantitate the energetics of the proton transfer process. The initial position of the H3O+ would also have an effect on the calculation of the absolute values for the energetics. The effects would be similar for both processes allowing for the relative comparison of the energetics of the two proton processes (i.e. proton transfer mediated with Ser320 or without). In both cases, the results indicated that the process was downhill and barrierless for H2 production by approximately 6 kcal/mol (Fig. 8).

DISCUSSION We have performed a series of experiments and computational simulations to examine the role of two residues, Arg286 and Ser320, that were previously unexamined for their impact on proton transfer in the C. pastuerianum [FeFe]-hydrogenase. Both residues are solventexposed and proximal to Glu282, a residue proposed to accept protons from the surface during H2 production. Both Arg286 and Ser320 are highly conserved in [FeFe]-hydrogenases, implying that they are important for structure and/or function. The role of Arg286 was explored by measuring H2 evolution activity in the native enzyme and in an R286L variant. When an abiotic electron donor system (sodium dithionite and MV) was utilized, the R286L variant had ~3-fold higher activity relative to the native enzyme. Conversely, when a sodium dithionite/ferredoxin electron donor system was utilized to mimic more closely biological conditions in which interprotein electron transfer was the rate-limiting step, the two enzymes had statistically indistinguishable H2 evolution activities. We propose that the increased activity of the R286L variant when using dithionite/MV is due to more rapid proton transfer,1 presumably caused by eliminating the salt-bridge between Glu282 of the proton transport pathway and Arg286. It must be noted, however, that the amino acid substitutions


18

Page 19 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

could have had unequal effects on the Km of the enzyme towards MV and ferredoxin, providing an alternate, but unlikely, explanation for these results. Enhanced proton transport is consistent with the increased Zn2+-based inhibition in R286L, suggesting that the Glu282 side-chain is more solvent exposed and available to bind Zn2+ in this variant. Additionally, previous reports indicate that replacing Arg286 with leucine creates a more negative electrostatic potential in this region of the protein, potentially enhancing proton entry at the enzyme surface.9 Based on these data, we hypothesize that Arg286 serves a regulatory or gating role for proton transport, although the physiological function is not immediately apparent. It is interesting to note that the enzyme has not evolved for maximal proton transfer rates, implying that biomimetic synthetic catalysts have the potential to exceed the performance of natural enzymes. When protonated, Glu282 is predicted to interact not with Arg286, but rather with solvent-exposed Ser320, forming a H-bond with ~60% occupancy according to molecular dynamics simulations.9 Intriguingly, substituting Ser320 with alanine resulted in a dramatic and unexpected decrease in activity (Fig. 3). This observation can be explained by at least two different scenarios: 1) The side chain of Ser320 is important for modulating either the structure or the electrostatic environment of the proton pathway or 2) Ser320 is itself a member of the proton pathway. To explore the role of Ser320, we conducted a series of computational analyses using both force field-only and hybrid QM/MM molecular dynamics simulations. No distinguishable differences in the backbone structure of residues Glu282, Ser319, Ser320, or Arg286 were observed between the native enzyme and the protein variants. There were, however, modest changes in side chain orientation among the variants. In particular, the distance between Arg286 and Glu282 decreased in the S320A variant (Fig. 6). However, the shortened distance does not


19

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 38

appear to correlate to the change in activity, based on the experimental Zn2+ inhibition assays, which demonstrated no change in the solvent exposure of Glu282 in the S320A variant, and no change in the computationally determined hydrogen bonding occupancy between Arg286 and Glu282. To ascertain if the Ser → Ala substitution modulated the pKa values of nearby pathway residues, the free energy for proton transfer from Glu282, through Ser319, to Glu279 was calculated, and the reaction was found to be nearly thermoneutral for both the native enzyme and the S320A variant, with low barriers to reaction (~2.5 kcal/mol), further suggesting a limited impact of the S320A variant on the electronic environment. Another possible explanation for the observed results is that Ser320 serves as an additional substituent in the proton transfer pathway, potentially acting as the initial proton acceptor from solvent. The S320A variant exhibits a 5- to 6-fold decrease in activity relative to the native enzyme and a 2-fold stimulation of activity in the presence of sodium azide, both of which are strikingly similar to that observed for the E282D variant.8 QM/MM calculations are also consistent with a role of Ser320 in proton transfer, based on the equi-energetic proton transfer pathways either including or excluding Ser320. Although classical proton transfer systems utilize solvent-exposed side chain moieties that are acidic or neutral at the proton pathway entrance (e.g. bacteriorhodopsin,5, 38 cytochrome c oxidase,39 and carbonic anhydrase40), Ser320 (or the conserved threonine in this position; Fig. S1) could operate purely by a Grotthuss mechanism,41 serving as a proton shuttle and never needing to be fully protonated or deprotonated. This may be the mechanism for the threonine in halorhodopsin.5, 42-44 The rationale for Ser320 participating in the proton pathway is not initially intuitive. Specifically, Glu282 appears to have good access to the solvent based on the inhibition studies with ZnCl2.8 However, Ser320 has greater solvent exposure compared to Glu282 and also lacks


20

Page 21 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

the positive charge from Arg286, both of which could enable Ser320 to have enhanced accessibility to hydronium ions. Therefore, the data reported here may suggest that Ser320 is an additional member of the linear proton pathway. In this potential extended pathway, both the Glu282 and Ser320 residues could serve as an entrance for proton transfer. Indeed, substitution of Glu282 with a non-protic residue caused a 95% decrease in activity,8 while the S320A variant activity was reduced by ~80%, consistent with a linear pathway of residues starting with Ser320 (Fig. 2). Conversely, the collected data do not allow us to rule out the possibility that Ser320 could also perform some other undetermined function. Further studies are necessary to clarify the exact role of this amino acid residue. It is difficult to establish a unique role for each residue in the proton transport pathway, but the unexpected importance of Ser320, and the role of Arg286 in regulating proton transfer, points to a mechanism more complex than was originally envisioned.8 While additional experiments are necessary to understand fully the mechanism for proton transport, one of the most interesting observations is that the C. pasteurianum [FeFe]-hydrogenase is not optimized for H2 production in an artificial system. Thus, further studies on this class of enzymes can reveal ways to enhance catalytic performance, and this information could be applied to synthetic molecular catalysts that mimic the enzyme.

CONCLUSIONS In conclusion, we have investigated the roles of two solvent-exposed residues, Arg286 and Ser320, which form strong interactions with Glu282, the presumed proton acceptor in [FeFe]-hydrogenase. The S320A variant demonstrates decreased activity and can be rescued by azide, indicating that this residue might be directly involved in proton transfer, a role that is


21

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 38

supported by QM/MM calculations. Conversely, Arg286 appears to help regulate the rate of proton transfer via salt-bridge formation with Glu282, as indicated by increased rates of reaction in the R286L variant. The observation that the Arg286-Glu282 hydrogen bonding interaction represses the rate of proton transfer highlights the fact that these enzymes have been optimized for a complex series of reactions in a living system and that there are still opportunities for improvements in the reaction of interest (i.e. H2 production). This is a critical observation in attempting to use or mimic [FeFe]-hydrogenases for practical or industrial applications.

FUNDING This work was supported by the DOE Great Lakes Bioenergy Research Center (DOE BER Office of Science, DE-FC02-07ER64494). In addition, support from the DOE Office of Science Early Career Research Program through the Office of Basic Energy Sciences (WJS, BG), the Center for Molecular Electrocatalysis, an Energy Frontier Research Center funded by the US DOE, Office of Science, Office of Basic Energy Sciences (SR), and the Division of Chemical Sciences, Geosciences, and Bio-Sciences of the US DOE, Office of Science, Office of Basic Energy Sciences (SR) is gratefully acknowledged. Pacific Northwest National Laboratory is operated by Battelle for the U.S. Department of Energy. JCSS acknowledges a post-doctoral fellowship from CNPq. TAS is a Productivity Fellow in Research from CNPq.

ACKNOWLEDGEMENTS Computational resources were provided at W. R. Wiley Environmental Molecular Science Laboratory (EMSL), a national scientific user facility sponsored by the Department of Energy’s Office of Biological and Environmental Research located at Pacific Northwest National


22

Page 23 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Laboratory, and a portion of the research was performed using PNNL Institutional Computing at Pacific Northwest National Laboratory.

SUPPORTING INFORMATION The following data can be found in the supplementary information: Force field parameters and simulation details utilized for MD; multiple sequence alignment of selected [FeFe]-hydrogenase protein sequences (Fig. S1); computational analysis of the distance between Cα atoms (Fig. S2); the distances between side chains (Fig. S3); the average number of water molecules around Glu282 in water density maps (Fig. S4); and QM/MM simulations of proton transfer from Glu282 to Glu279 in native and S320A variant (Fig. S5).

REFERENCES 1.

2.

3. 4. 5. 6.

7.

8.

Peters, J. W., Lanzilotta, W. N., Lemon, B. J., and Seefeldt, L. C. (1998) X-ray crystal structure of the Fe-only hydrogenase (Cpl) from Clostridium pasteurianum to 1.8 angstrom resolution, Science 282, 1853-1858. Nicolet, Y., Piras, C., Legrand, P., Hatchikian, C. E., and Fontecilla-Camps, J. C. (1999) Desulfovibrio desulfuricans iron hydrogenase: the structure shows unusual coordination to an active site Fe binuclear center, Structure 7, 13-23. (2010) Fuel Cell Science: Theory, Fundamentals, and Biocatalysis, John Wiley & Sons, Inc., Hoboken, New Jersey. Bhatt, S., Gupta, B., Sethi, V. K., and Pandey, M. (2012) Polymer exchange membrane (PEM) fuel cell: a review, Int J Curr Eng Tech 2, 219-226. Wraight, C. A. (2006) Chance and design--proton transfer in water, channels and bioenergetic proteins, Biochim Biophys Acta 1757, 886-912. Hong, G., Cornish, A. J., Hegg, E. L., and Pachter, R. (2011) On understanding proton transfer to the biocatalytic [Fe-Fe]H sub-cluster in [Fe-Fe] H2ases: QM/MM MD simulations, Biochim Biophys Acta 1807, 510-517. McCullagh, M., and Voth, G. A. (2013) Unraveling the role of the protein environment for [FeFe]-hydrogenase: a new application of coarse-graining, J Phys Chem B 117, 40624071. Cornish, A. J., Gartner, K., Yang, H., Peters, J. W., and Hegg, E. L. (2011) Mechanism of proton transfer in [FeFe]-hydrogenase from Clostridium pasteurianum, J Biol Chem 286, 38341-38347.


23

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

9.

10. 11. 12.

13.

14.

15.

16. 17. 18. 19. 20. 21. 22.

23.

24. 25. 26.

Page 24 of 38

Ginovska-Pangovska, B., Ho, M.-H., Linehan, J. C., Cheng, Y., Dupuis, M., Raugei, S., and Shaw, W. J. (2014) Molecular dynamics study of the proposed proton transport pathways in [FeFe]-hydrogenase, Biochim Biophys Acta 1837, 131-138. Sode, O., and Voth, G. A. (2014) Electron transfer activation of a second water channel for proton transport in [FeFe]-hydrogenase, J Chem Phys 141, 22D527. Long, H., King, P. W., and Chang, C. H. (2014) Proton transport in Clostridium pasterianum [FeFe]-hydrogenase I: a computational study, J Phys Chem B 118, 890-900. Gärtner, K., Lechno-Yossef, S., Cornish, A. J., Wolk, C. P., and Hegg, E. L. (2012) Expression of Shewanella oneidensis MR-1 [FeFe]-hydrogenase genes in Anabaena sp. strain PCC 7120, Appl Environ Microbiol 78, 8579-8586. MacKerell, A. D., Bashford, D., Bellott, M., Dunbrack, R. L., 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., Roux, B., Schlenkrich, M., Smith, J. C., Stote, R., Straub, J., Watanabe, M., Wiorkiewicz-Kuczera, J., Yin, D., and Karplus, M. (1998) All-atom empirical potential for molecular modeling and dynamics studies of proteins, J Phys Chem B 102, 35863616. Chang, C. H., and Kim, K. (2009) Density functional theory calculation of bonding and charge parameters for molecular dynamics studies on [FeFe] hydrogenases, J Chem Theory Comput 5, 1137-1145. Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C., Skeel, R. D., Kale, L., and Schulten, K. (2005) Scalable molecular dynamics with NAMD, J Comput Chem 26, 1781-1802. Allen, M. P., and Tildesley, D. J. (1989) Comp Sim Liq, Clarendon Press, Oxford. Humphrey, W., Dalke, A., and Schulten, K. (1996) VMD: visual molecular dynamics, J Mol Graph 14, 33-38, 27-38. Torrie, G. M., and Valleau, J. P. (1977) Non-physical sampling distributions in MonteCarlo free-energy estimation - umbrella sampling, J Comput Phys 23, 187-199. Perdew, J. P., Burke, K., and Ernzerhof, M. (1996) Generalized gradient approximation made simple, Phys Rev Lett 77, 3865-3868. Perdew, J. P., Burke, K., and Ernzerhof, M. (1997) Generalized gradient approximation made simple (vol 77, pg 3865, 1996), Phys Rev Lett 78, 1396-1396. Grimme, S. (2006) Semiempirical GGA-type density functional constructed with a longrange dispersion correction, J Comput Chem 27, 1787-1799. VandeVondele, J., Krack, M., Mohamed, F., Parrinello, M., Chassaing, T., and Hutter, J. (2005) QUICKSTEP: Fast and accurate density functional calculations using a mixed Gaussian and plane waves approach, Comput Phys Comm 167, 103-128. Lippert, G., Hutter, J., and Parrinello, M. (1999) The Gaussian and augmented-planewave density functional method for ab initio molecular dynamics simulations, Theor Chem Acc 103, 124-140. Goedecker, S., Teter, M., and Hutter, J. (1996) Separable dual-space Gaussian pseudopotentials, Phys Rev B Condens Matter 54, 1703-1710. VandeVondele, J., and Hutter, J. (2007) Gaussian basis sets for accurate calculations on molecular systems in gas and condensed phases, J Chem Phys 127, 114105. Grossfield, A. (2013) WHAM: an implementation of the weighted histogram analysis method, version 2.0.9, http://membrane.urmc.rochester.edu/content/wham.


24

Page 25 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

27.

28. 29. 30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Kumar, S., Bouzida, D., Swendsen, R. H., Kollman, P. A., and Rosenberg, J. M. (1992) The weighted histogram analysis method for free-energy calculations on biomolecules. 1. The method, J Comput Chem 13, 1011-1021. Hoover, W. G. (1985) Canonical dynamics - equilibrium phase-space distributions, Phys Rev A 31, 1695-1697. Nose, S. (1984) A unified formulation of the constant temperature molecular-dynamics methods, J Chem Phys 81, 511-519. Petersson, G. A., and Allaham, M. A. (1991) A complete basis set model chemistry. 2. Open-Shell systems and the total energies of the 1st-row atoms, J Chem Phys 94, 60816090. Petersson, G. A., Bennett, A., Tensfeldt, T. G., Allaham, M. A., Shirley, W. A., and Mantzaris, J. (1988) A complete basis set model chemistry. 1. The total energies of closed-shell atoms and hydrides of the 1st-row elements, J Chem Phys 89, 2193-2218. Krishnan, R., Binkley, J. S., Seeger, R., and Pople, J. A. (1980) Self-consistent molecular-orbital methods. 20. Basis set for correlated wave-functions, J Chem Phys 72, 650-654. Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman, J. R., Scalmani, G., Barone, V., Mennucci, B., Petersson, G. A., Nakatsuji, H., Caricato, M., Li, X., Hratchian, H. P., Izmaylov, A. F., Bloino, J., Zheng, G., Sonnenberg, J. L., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Montgomery Jr, J. A., Peralta, J. E., Ogliaro, F., Bearpark, M. J., Heyd, J., Brothers, E. N., Kudin, K. N., Staroverov, V. N., Kobayashi, R., Normand, J., Raghavachari, K., Rendell, A. P., Burant, J. C., Iyengar, S. S., Tomasi, J., Cossi, M., Rega, N., Millam, N. J., Klene, M., Knox, J. E., Cross, J. B., Bakken, V., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R. E., Yazyev, O., Austin, A. J., Cammi, R., Pomelli, C., Ochterski, J. W., Martin, R. L., Morokuma, K., Zakrzewski, V. G., Voth, G. A., Salvador, P., Dannenberg, J. J., Dapprich, S., Daniels, A. D., Farkas, Ö., Foresman, J. B., Ortiz, J. V., Cioslowski, J., and Fox, D. J. (2009) Gaussian 09, Gaussian, Inc., Wallingford, CT, USA. Takahashi, E., and Wraight, C. A. (2006) Small weak acids reactivate proton transfer in reaction centers from Rhodobacter sphaeroides mutated at AspL210 and AspM17, J Biol Chem 281, 4413-4422. Chang, Y. H., Chuang, L. Y., and Hwang, C. C. (2007) Mechanism of proton transfer in the 3α-hydroxysteroid dehydrogenase/carbonyl reductase from Comamonas testosteroni, J Biol Chem 282, 34306-34314. Whitehead, S. J., Iwaki, M., Cotton, N. P., Rich, P. R., and Jackson, J. B. (2009) Inhibition of proton-transfer steps in transhydrogenase by transition metal ions, Biochim Biophys Acta 1787, 1276-1288. Faxen, K., Salomonsson, L., Adelroth, P., and Brzezinski, P. (2006) Inhibition of proton pumping by zinc ions during specific reaction steps in cytochrome c oxidase, Biochim Biophys Acta 1757, 388-394. Gunner, M. R., Mao, J., Song, Y., and Kim, J. (2006) Factors influencing the energetics of electron and proton transfers in proteins. What can be learned from calculations, Biochim Biophys Acta 1757, 942-968. Brzezinski, P., and Adelroth, P. (2006) Design principles of proton-pumping haemcopper oxidases, Curr Opin Struct Biol 16, 465-472.


25

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

40. 41. 42.

43.

44. 45.

Page 26 of 38

Silverman, D. N., and McKenna, R. (2007) Solvent-mediated proton transfer in catalysis by carbonic anhydrase, Acc Chem Res 40, 669-675. Cukierman, S. (2006) Et tu, Grotthuss! and other unfinished stories, Biochim Biophys Acta 1757, 876-885. Sasaki, J., Brown, L. S., Chon, Y. S., Kandori, H., Maeda, A., Needleman, R., and Lanyi, J. K. (1995) Conversion of bacteriorhodopsin into a chloride ion pump, Science 269, 7375. Tittor, J., Haupts, U., Haupts, C., Oesterhelt, D., Becker, A., and Bamberg, E. (1997) Chloride and proton transport in bacteriorhodopsin mutant D85T: different modes of ion translocation in a retinal protein, J Mol Biol 271, 405-416. Kolbe, M., Besir, H., Essen, L. O., and Oesterhelt, D. (2000) Structure of the light-driven chloride pump halorhodopsin at 1.8 A resolution, Science 288, 1390-1396. Hess, B. (2002) Determining the shear viscosity of model liquids from molecular dynamics simulations, J Chem Phys 116, 209-217.


26

Page 27 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure Legends

Figure 1. Proton pathway of [FeFe]-hydrogenase with hydrogen bonds positioned for H2 production used in the MD simulations. The arrows show the expected movement of the protons.

Figure 2. Proton pathway and active site in [FeFe]-hydrogenase. The enlarged area shows the proposed entrance to the proton pathway in [FeFe]-hydrogenases as well as the relative orientations of Ser320, Arg286, and Glu282 to each other and to the surface of the protein.

Figure 3. Relative activity of native and amino-acid substituted variants of [FeFe]-hydrogenase. Hydrogen production was initiated by incubating protein samples with 100 mM sodium dithionite and either 10 mM methyl viologen (abiotic electron donor) or 16 µM C. pasteurianum ferredoxin and measuring the accumulation of headspace H2 by GC. Values are reported as kobsrel (ratio of variant enzyme activity to native enzyme activity). The activity of the native enzyme with MV and ferredoxin as the electron donor was 1.0 ± 0.2 and 0.34 ± 0.08 µmol/mg/s, respectively; the activity of the R286L variant with MV and ferredoxin as the electron donor was 3.2 ± 0.6 and 0.40 ± 0.01 µmol/mg/s, respectively; the activity of the S320A variant with MV and ferredoxin as the electron donor was 0.13 ± 0.06 and 0.07 ± 0.015 µmol/mg/s, respectively. Error bars represent the standard error of the mean (SEM; n ≥ 3). An asterisk (*) indicates that the value is statistically different from the analogous value in the native enzyme (p ≤ 0.01).

Figure 4. Stimulation of enzymatic activity by sodium azide. Native and variant enzymes were incubated with 0, 62.5, 125, 250, or 500 mM sodium azide under standard assay conditions using


27

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 38

100 mM sodium dithionite and 10 mM MV. H2 production activities are normalized to 1 using 0 mM sodium azide and reported as fold stimulation. Error bars represent the SEM (n ≥ 3).

Figure 5. Inhibition of hydrogenase activity by ZnCl2. Native and variant enzymes were incubated with 0 or 500 µM ZnCl2 as described in the Materials and Methods section using 100 mM sodium dithionite and 10 mM MV. H2 production activities were normalized to 0 µM ZnCl2 and expressed as a percent. Error bars represent the SEM (n ≥ 3).

Figure 6. Calculated pairwise distances between side chain residues. The distances between side chain residues of Glu282 and Ser319 and Ser320 reveal small but significant changes (~30%) in the space between the residues as a function of substitution, while the distance between Arg286 and Glu282 varies as a function of mutant. While the distances are smaller for each variant, there is no clear correlation to this reduction in size and the resulting rates. The fluctuations of the distances are expressed as standard deviations from the average values and are denoted by brackets. The error on the averages obtained from block averaging45 is between 0.04 Å to 0.34 Å. Distances were

measured between the following carbon or oxygen atoms: arginine (CZ), leucine (CG), alanine (CB), glutamic acid (CD) or serine (OG).

Figure 7. Calculated hydrogen bond occupancy rates for Glu282-Ser319 and Glu282-Arg286 for the native protein and each of the variants. Significant differences are observed for both mutations for Glu282-Ser319, but not for Glu282-Arg286. The errors were calculated using the block average approach.45


28

Page 29 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 8. Free energy calculated using QM/MM calculations for proton transfer from H3O+ to Glu282. The reaction coordinate r = r3 + r4 is shown on the left. For r = 2 Å the proton is on the H3O+ ion, and for r = 3.5 Å the proton is on the Glu282 side-chain. The lack of an observable minimum when the proton resides on the H3O+ ion indicates that the reaction is spontaneous. The error bars reported were calculated using Monte Carlo Bootstrap Error Analysis implemented in WHAM.26-27


29

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 38

Figure 1


30

Page 31 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 2


31

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 38

Figure 3


32

Page 33 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 4


33

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 38

Figure 5


34

Page 35 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 6


35

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 38

Figure 7


36

Page 37 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 8


37

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 38

For Table of Contents Use Only


38

Single-Amino Acid Modifications Reveal Additional Controls on the

Recommend Documents