A Major Structural Change of the Homocitrate Ligand of Probable

Jan 5, 2018 - Because model systems exist where N2 is activated by complexes containing a five-coordinated MoIII,(25) attempts were now made to stretc...
0 downloads 6 Views 701KB Size
Article pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

A Major Structural Change of the Homocitrate Ligand of Probable Importance for the Nitrogenase Mechanism Per E. M. Siegbahn* Arrhenius Laboratory, Department of Organic Chemistry, Stockholm University, SE-106 91 Stockholm, Sweden S Supporting Information *

ABSTRACT: Mo-containing nitrogenase is the main enzyme that is able to take N2 from the air and form ammonia. The active-site cofactor of the enzyme, termed FeMoco, is unique in nature. It has seven Fe and one Mo atoms connected by S bridges, with a C atom in the center of the cofactor. Another unusual feature is that it has a large homocitrate ligand known to be of importance for catalysis. In the present computational study, the role of the homocitrate ligand is investigated. It is found that a large structural change, which makes MoIII fivecoordinated, is energetically favorable in the more reduced states. This is of probable importance for the nitrogenase mechanism.

I. INTRODUCTION Nitrogenase is of huge importance in nature because it is the leading enzyme for transforming N2 in the air to ammonia, which can then be used by all living species to make compounds containing N. The active site for the formation of ammonia is, not surprisingly, very unusual.1 The metallocofactor at the active site, termed FeMoco, contains seven Fe atoms and one Mo atom bound together by S bridges. This is not the only unusual part of the structure. After decades of investigations, it was found that there is a C in the center of the complex. To obtain this finding required more and more refined crystal structures down to a resolution of 1.0 Å.2 Even at the rather high resolution of 1.6 Å,3 the C atom could not be seen, and at a resolution of 1.16 Å, it was still unclear whether the central atom was a C, a N, or an O atom.4 Spectroscopy and theory also played an important role in the final assignment.5 The above very unusual parts of the cofactor structure are not enough; there is also a large homocitrate ligand bound to Mo. By quantum mechanics/molecular mechanics (QM/MM) modeling, this ligand was recently suggested by Cao et al. to have a negative charge of 3−.6 This finding is the same as the one found in the present study. The homocitrate has three carboxylate and one hydroxyl groups and binds to Mo with one carboxylate O atom and with the hydroxyl O atom. An interesting aspect of the structure is also that there are no positive amino acids hydrogen-bonded directly to it to compensate for the large negative charge. Because homocitrate is so unusual as a metal ligand in nature, it has since long been of interest to find out what its role is. In the beginning, before the crystal structures, obviously mainly experimental studies were used for the investigations. Around 1990, Madden et al. found in a study of several homocitrate © XXXX American Chemical Society

substitutes that erythro-1-fluorohomocitrate was the only one capable of keeping some reasonable activity.7 Substitutes like citrate, (R)-citramalate, cis-aconitate, and 1-fluorohomocitrate had diminished or undetectable activities. The significant role of homocitrate became even more confusing when it was shown by spectroscopy that Mo is most probably not the site for N2 binding.8,9 It should be noted in this context that N2 reduction for citrate is still 7% and that for isocitrate 3% of the activity for homocitrate, and this only means an increase by about 2 kcal/mol on the rate-limiting barrier. It is still clear that the precise form of the homocitrate ligand has an effect on this barrier. An attempt to answer the question of the role of homocitrate has been made in theoretical model studies by Dance.10 He suggested that the homocitrate is an essential part of a proton relay chain, leading to the N2 site. He argued that this could answer some of the detailed differences between the different substitutes by modifications of the hydrogen-bonding network. There are two possibilities for this hypothesis to be true. The first one is that there is a very large effect on the proton-transfer barrier. The second possibility, with only a small effect on the barrier, requires that proton transfer should be rate-limiting for N2 reduction. Both of these possibilities appear quite unlikely. Rate-limiting proton transfer is very unusual in enzymes with a slow turnover like the one in nitrogenase, and the only case where this occurs, is when there is a special purpose for having a slow transfer, like in the gating of proton pumping in cytochrome oxidase.11,12 Normally, proton transfer is achieved on the microsecond scale.13 Because these energetic questions were not investigated in the previous studies, the role of homocitrate is still not clear. Received: October 2, 2017

A

DOI: 10.1021/acs.inorgchem.7b02493 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Fe1 and ending with Fe7, this spin can be described as + - - - + + + . It can be noted that the optimal spin for the quartet ground state in the X-ray structure is different with + - + - + + -.6,18,19 The reason for this difference in the spin coupling is that MoIII has low spin (s = −1.4) in the ground state but high spin (s = −2.5) in most of the reduced states, for example, in E4. The larger spin on the Mo atom leads to a strong antiferromagnetic coupling to the three nearest Fe atoms. Broken symmetry was used in the calculations. The spin populations on the Fe atoms are between 3.5 and 3.7 but are not very useful for assigning the individual oxidation states. A few comments can be added concerning the choice of the method and model. The size of the model has been tested by going from smaller to larger models, and the energies of interest are stable. Compared to the energies of interest in the previous study, there are actually very small differences in going from the smaller model of 170 atoms to the present one of 270 atoms. For the structural change of the homocitrate, some of the results are obviously due to the added atoms outside the previous model. The charge state has been tested against known experimental facts (see above). Within the model, there are no assumptions. The structure with the lowest energy has always been chosen. There are no significant effects by varying the basis set or by changing the functional by varying the amount of exact exchange.

II. METHODS AND MODELS The methods used here are the same as the ones used in the previous study on nitrogenase14 and also essentially the same as those used on many other enzymes, summarized in a recent review.12 The geometries were optimized using the density functional theory (DFT) functional b3lyp,15 with the lacvp* basis set. This basis set was also used for obtaining dielectric effects with a dielectric constant equal to 4.0 at the b3lyp level. At the optimized geometries, single-point calculations were made using a large basis set, with cc-pvtz(-f) for all nonmetal atoms. For the metals, a lacvp3+ basis set was used. Dispersion effects were added using the empirical D2 formula of Grimme.16 Zero-point effects were not calculated because they are quite time-consuming for a model with 270 atoms. Furthermore, they should be small for the present energy differences. All of the calculations have been performed using the Jaguar program.17 The model of FeMoco is based on the high-resolution structure2 and is shown in Figure 1. This model is larger than the one used before

III. RESULTS Before the results for the structural changes of the homocitrate ligand are described, a few remarks have to be made. After extension of the model to 270 atoms from the 170 atoms used previously,14 most features of the cofactor are the same as before. However, with the present finding for the homocitrate ligand described below, it was investigated whether additional reductions of the cofactor could be possible. It was then found that two additional electrons and protons were energetically possible to add to the cofactor compared to the previous model of the E4 state. To avoid confusion, the present active state is still termed E4, in line with the experimental assignments. In the new E4 state, there are three protonated S ligands: S3A, S2B, and S5A (or homocitrate). With three protons on C in an activation sequence and with two hydrides, the (formal) oxidation state of the cofactor in E4 now becomes Mo3+7Fe2+, where MoIII has high spin. When the hydrides leave as H2, the cluster is reduced further to Mo3+5Fe2+2Fe+ when N2 binds. The presence of Fe1+ in the state that binds N2 is not surprising. In the reduction of protons in the FeFe-hydrogenase, which is a much less demanding process, the key to the accepted mechanism is that one of the Fe atoms becomes Fe1+,20 but it should be noted that the ligand field is stronger in the hydrogenase case. Still, the present calculations are in full agreement with this simplified argument, which is, furthermore, in line with the fact that the P cluster is an all-FeII cluster but does not activate N2. The important part of the N2 activation mechanism remains the same, with a reductive elimination of two hydrides directly followed by N2 binding, in full agreement with the suggestion made from electron paramagnetic resonance (EPR) experiments.8,9 An important difference is that N2 activation now occurs for a doublet state, in agreement with experiments. In the previous computational study,14 a singlet was found to be the active state. The central C atom is in the present study still protonated to CH3 as before. In the experimentally suggested mechanism for N 2 activation in nitrogenase, 8,9 C is unprotonated and only two protons (on S) and two hydrides are added to the ground-state cofactor to reach the active E4 state. This means that the oxidation state of E4 is the same as that for the ground state, now generally accepted as Mo3+4Fe3+3Fe2+,19 even though the lowest redox potential in nature of −1.6 V is used to reach E4.14,21−24 With three

Figure 1. Model of the ground state used for the present calculations showing which residues were included. The numbering of the Fe atoms and three of the S atoms is also shown. and has included also a region outside His195 and residues and water hydrogen bonding to the homocitrate. In most cases, the backbone atoms are fixed from the X-ray structure; exceptions are the arginines, for which smaller models were used and only one H atom was frozen. The model consists of 270 atoms. Asp386 and Glu427 become protonated. The charge of the entire model is 2−, leading to a charge of 4− for the cofactor itself without its hydrogen-bonding neighbors. The decision on this charge state was reached by a comparison to other charge states, where only this charge state leads to a reasonable activation of N2 and a protonation state of the cofactor in agreement with the X-ray structure.14 The oxidation state of the cofactor then becomes Mo3+4Fe3+3Fe2+, which is in full agreement with recent assignments based on spectroscopy.19 The process leading from the ground state in Figure 1 to the E4 state, studied here, is described in detail in the previous paper.14 In short, there is an activation process in which the central C atom is protonated three times, leading to a corresponding reduction of the metal cofactor. In addition, three S atoms of the cofactor become protonated (in the previous study, it was only two) and two hydrides are being bound to the cofactor in the E4 state. In each reduction, one electron and one proton are added, so the charge of the model remains the same (2−). The binding of two hydrides is in line with a detailed spectroscopic analysis.8,9 The oxidation state of the E4 state of the cofactor then becomes Mo3+7Fe2+ with a doublet spin. The best spin coupling found for the reduced states was described previously in ref 14. It has alternating spins, where the leftmost Fe atom has up-spin and the Mo atom has down-spin. The three Fe atoms closest to the Mo atom have up-spin, while the three Fe atoms closest to the leftmost Fe atom have down-spin. With the numbering from the X-ray structures (see Figure 1) starting with B

DOI: 10.1021/acs.inorgchem.7b02493 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 2. Starting structure (A) for the homocitrate region, essentially taken from the X-ray structure.

diagram in Figure 3, if the energy is set to zero for the starting structure A with a proton on S5A. A major part of this energy

protonations of the C atom in an activation sequence and three protonations of the S atoms, the cofactor becomes much more reduced because all of the electrons in these steps (formally) will go to the metals, which should make it significantly more reactive toward N2. All details of the structure and mechanism for the extended model will be described in a forthcoming paper. The present paper will instead focus on the homocitrate ligand. Pathway for the Structural Change of the Homocitrate Ligand. The underlying reason for the investigation of a possible structural change of the homocitrate was that no positive free energy of binding (in the present convention, binding energies are considered positive when the binding is exergonic) was found for N2 on the FeMoco complex in the E4 state, in spite of a very large number of attempts. Because model systems exist where N2 is activated by complexes containing a five-coordinated MoIII,25 attempts were now made to stretch each one of the two Mo−O bonds of homocitrate to possibly open a site for N2 binding. The starting structure for the homocitrate region of the E4 state is shown in Figure 2 and is labeled A. It is essentially taken from the X-ray structure. A model with 270 atoms was used, as described in section II, where most of the surrounding hydrogen bonding to the homocitrate was kept. To computationally model this structure, hydrogen bonding has to be saturated within the model, which is a common procedure in cluster modeling.12 This means that there are minor differences in hydrogen bonding compared to the experimental structure. In the optimal form of this model, there is only one proton on the homocitrate, which is positioned at OA, forming a hydrogen bond to OD. This is very similar to the structure reached in a recent QM/MM study by Ryde et al.6 The structure of the homocitrate region is shown in Figure 2 and is labeled A. To stretch the Mo−O bonds in structure A is energetically quite costly unless an additional proton is placed on OD of the homocitrate pointing away from OA. The proton is taken from S5A. A consequence of this proton transfer is that hydrogen bonding around the homocitrate is changed. Several hydrogenbonding structures were tried until the best one was found. In that structure, a water (number 3) was moved to allow a direct hydrogen bonding between Lys426 and OF, which significantly stabilizes the energy. When the water has been moved, there are obviously also other changes of the hydrogen bonding. The energy is now still rather high at +11.3 kcal/mol, point B in the

Figure 3. Approximate path for the structural change of the homocitrate region described in the text. The energies are given in kcal/mol The distance R (Å) is the Mo−OA distance, while the distance S is the Mo−OB distance.

increase comes from the solvation effects of +6.8 kcal/mol, but there is also a contribution from the dispersion of +1.3 kcal/ mol. The bond distance between Mo and OA has now increased to 2.43 Å from 2.19 Å in A. With an extra proton on the homocitrate, it is much easier to stretch the Mo−OA bond. After stretching the Mo−OA bond from 2.43 to 2.80 Å, the energy goes up to +16.2 kcal/mol for point C, only 4.9 kcal/ mol higher than that for point B. After point C, a very flat region of the energy surface is reached. Stretching the bond further leads to an energy of +15.7 kcal/mol at point D, which is actually a local minimum with the MoA bond equal to 3.56 Å. This rather large structural change is therefore still within reach at the time scale of seconds for each step of the nitrogenase mechanism. When OA is pushed further away from Mo, the energy goes down to +7.3 kcal/mol at point E for a distance of C

DOI: 10.1021/acs.inorgchem.7b02493 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

parts of the catalytic cycle. The N2 protonation steps after the initial binding of N2 were therefore investigated all the way to the formation of two NH3. However, no clear advantage was found anywhere, and this idea was also left. Binding N2H to the Fe Atoms. A difference from the previous study is that there are now three protons bound to the ligands of the cofactor in E4. This difference is due to the two additional reduction steps, found to be energetically possible. In the old E4, only S3A was protonated. S2B only became protonated in E5. One of the additional protons in the new E4 is now on S2B and the other one on the homocitrate, where the latter is the interesting difference. The question is whether the homocitrate proton can be used for the binding of N2. Attempts were therefore made to move this proton to N2. The removal of a proton means that the structure of the homocitrate in Figure 4 becomes unstable and therefore rotates back to the original position in Figure 2. This turns out to finally solve the problem of binding N2. Compared to the E4 structure with a rotated homocitrate, the removal of the two hydrides as H2 and binding N2 as N2H are essentially thermoneutral, just as observed experimentally in the E4 state. The removal of H2 from E4, forming an excited E2 state (termed E2*), is endergonic by 2.5 kcal/mol, including an energy gain of entropy for the free H2 of 10.3 kcal/mol. The binding of N2 to E2* is after this removal exergonic by 2.5 kcal/ mol, including an entropy loss of 9.9 kcal/mol from the binding of a free N2. In the previous study, it was found that there is an entropy gain for the cofactor itself of 5.7 kcal/mol for the more open structure with a bound N2H compared to the more compact structure of the E4 and E2* states. This entropy gain, taken from the old study, is included in the 2.5 kcal/mol for the total free energy of binding. The energy gain of using the additional proton on the homocitrate to form a bound N2H is therefore enough to compensate for the cost of the backrotation of the homocitrate. Without the structural change of the homocitrate, this nearly perfect agreement with the experiments is lost. It should be added that the accuracy of the present modeling is not high enough to completely trust these exact values, but the results should give the rough energetics for these events. Starting State for the Rotation. The structural change of the homocitrate for the other states in the catalytic mechanism was also investigated. Only the energy difference between structure A in Figure 2 and structure H in Figure 4 was investigated. It can first be concluded that, for the X-ray structure, previously termed A0, the structural change cannot occur because there is no additional proton at that stage, which could be moved to the homocitrate. As described in the previous paper,14 the mechanism is suggested to start with a set of four A states, before the actual catalytic cycle starts. For these states, S3A and C are protonated. CH3 is formed and goes out of the center of the cofactor to form a terminal ligand. These protonation positions are all preferred compared to protonation of the homocitrate, and no structural change of the homocitrate will therefore occur. The E0 state is the same as the A4 state. For the E1 state, there is a bound hydride and a proton on S3A. Both are fairly strongly bound, so moving any one of them to the homocitrate is not favorable, even after the structural change of the homocitrate. For E2−E4, there is an additional proton on S2B, which is much easier to move to the homocitrate, and these states therefore prefer a protonated homocitrate in the rotated form. As described above, the

4.59 Å. At this stage, it was also decided to increase the Mo−OB distance. At a Mo−OB distance of 2.90 Å, there is a slight structural change in which OC moves closer and eventually binds to Mo. The energy has now increased again to +13.1 kcal/mol at point F for a Mo−OB distance fixed at 2.90 Å. When OB is moved still further out and the constraint is relaxed, the energy goes down to +0.3 kcal/mol at point G for a Mo−OB distance of 3.74 Å. Only a very small barrier (less than 2 kcal/mol) was found in this structural change. A final change of hydrogen bonding around the homocitrate leads to an energy at −7.6 kcal/mol at point H. The final structure is shown in Figure 4. In conclusion, a structural change of the

Figure 4. Final point (H) after the structural change of the homocitrate region.

homocitrate to create an open site on Mo is therefore quite favorable energetically in the E4 state. The Mo part of the cofactor is now very similar to the synthesized Mo complex with a five-coordinated MoIII. Binding N2 to Mo. The first investigation made after the above finding of the structural change was obviously to try to bind N2 to the open site of MoIII. It should first be noted for the structure in Figure 4 that the open site created on Mo is not where OA was situated but where OB was. This led to another very surprising finding. It turns out that the CH2 link to the backbone of His442 is blocking this OB site, leading to a strongly bent Mo−N2 bond and a poor free energy of binding. Many attempts were then made to get an open site at the original OA position on Mo, but this led to significantly higher energies. The conclusion is therefore that the open site of MoIII on the cofactor does not appear to help the activation of N2 at all. Instead, N2 binds better to the Fe atoms of the FeMoco cluster as before, but the free energy of binding is still negative for the E4 state. Binding N2 to the Fe Atoms. Since the structural change of the homocitrate described above is quite favorable energetically, the investigation was continued to find out whether there is another advantage with this change. The first idea tested was that the energy gained by the structural change might appear concertedly with the binding of N2 and make the step downhill. This scenario turned out not to be true. Because the structural change is favorable already in E2 and E3, it cannot occur concertedly with the binding of N2 in the E4 state. Another obvious possibility might be that N2 binds better to the Fe atoms in E4 after the structural change. However, N2 is actually quite unbound in the best local minimum between the Fe atoms, much like in the E4 state of the previous study, where an additional electron had to be added to bind N2. Other Possible Advantages of the Rotation. Even though the structural change of the homocitrate did not improve the binding of N2, there might be advantages in other D

DOI: 10.1021/acs.inorgchem.7b02493 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry structure changes back from structure H to structure A when the homocitrate proton is placed at N2 at the end of the E4 step. Importance of Lys426. As described above, a key difference between structures A and H is that there is a direct hydrogen bond between the positive Lys426 and one of the negative carboxylates of the homocitrate in structure H. The importance of this interaction was studied by replacing the NH3+ of the lysine by a H atom for structures A and H. The structures were reoptimized, but no different local minima than those for structures A and H were studied. It was then found that the energy preference of structure H compared to structure A disappears entirely, decreasing by about 8 kcal/mol from the value of −7.6 kcal/mol in Figure 3 to 0 kcal/mol. Kinetic Considerations. The thermodynamic energetics given above are not the only factors that lead to the unexpectedly complicated mechanism of nitrogenase. There are probably also kinetic reasons. The most sensitive situation in the entire mechanism is when N2 becomes activated in E4, which is most probably also the rate-limiting step. One difficulty not discussed yet is to avoid binding another proton as a hydride in the E2* state. Access to the additional proton in E4 must therefore be sufficiently slow that N2 has become bound before the proton arrives. The proton must go to N2 rather than into the cavity as a hydride. Access of the homocitrate proton is indeed quite slow because it depends on the back-rotation of the homocitrate. If the additional proton would be bound to one of the bridging S atoms, it would probably rapidly go into the cavity before N2 arrives. Finally, I am grateful to one of the reviewers, who pointed out two papers by Durrant et al.26,27 of relevance for the present study. These papers are more than 10 years old, and I was not aware of them. In these papers, a structural change of the homocitrate was discussed and modeled. By an MM simulation, they found that it was possible to open up an empty site on Mo, much like in the present study. The importance of Lys426 was emphasized, just like here. This was followed up by experimental mutation studies, where Lys426 was mutated to other amino acids. Significant effects on the N2 activation rate were found. For subsequent DFT calculations, very small models by the present standards were obviously used. The conclusion from those DFT studies was that N2 was activated at this empty Mo site, which is different from the present conclusions.

to N2H using the proton on the homocitrate, also in a nearly isoenergetic step. Experimentally, these steps in E4 can be reversed by varying the partial pressure of N2, indicating a nearly thermoneutral process.8,9 There are a few features of the X-ray structure that indicate that a structural change might be possible. First, the charge of 3− for the homocitrate is quite unusual for a metal ligand in biology. Furthermore, in spite of the high negative charge, the homocitrate is not anchored by directly binding positive groups. Instead, there is a lysine, Lys426, positioned a bit away from the homocitrate. After the structural change, one of the carboxylates of the homocitrate forms a strong hydrogen bond to Lys426, which is the main reason for the favorable structural change. Another indication of the correctness of the modeling results is that this gives a significant role for the homocitrate in the rate-limiting step when N2 becomes activated. It has been experimentally shown that the form of this ligand is important for catalysis.7 For example, changing homocitrate to citrate has a significant effect on the rate. An explanation could be that citrate is shorter than homocitrate and will therefore not as easily reach Lys426. The present paper has focused on the structural change of the homocitrate. In forthcoming papers, the entire mechanism of nitrogenase, including this change, will be described in detail.

IV. SUMMARY By the present model calculations, it has been shown that a structural change of the homocitrate is energetically preferred for the higher E states of the nitrogenase mechanism. The structural change starts with protonation of the homocitrate, followed by rotation. The preference for the rotated structure is −7.6 kcal/mol for the E4 state, which is much more than the uncertainty in the calculations; for further information, see section II. The structural change opens up an empty coordination site on Mo, which unexpectedly does not lead to a binding of N2 at this position. Instead, the advantage with the structural change is that a proton is saved on the homocitrate, which can be used to protonate N2 at the end of the E4 step. This leads to a nearly perfect agreement with the experimental findings based on EPR.8,9 Because the agreement disappears without the structural change, this is the strongest evidence that it actually occurs. There is first a nearly isoenergetic loss of two hydrides as H2 by a reductive elimination step. After this, N2 binds and becomes protonated

ACKNOWLEDGMENTS This work was generously supported by the Knut and Alice Wallenberg Foundation and by grants from the Swedish Research Council. Computer time was provided by the Swedish National Infrastructure for Computing.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.7b02493. Coordinates for the structures mentioned in the text (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Per E. M. Siegbahn: 0000-0001-7787-1881 Notes

The author declares no competing financial interest.

■ ■

REFERENCES

(1) Kim, J.; Rees, D. C. Structural Models for the Metal Centers in the Nitrogenase Molybdenum-iron Protein. Science 1992, 257, 1677− 1682. (2) Spatzal, T.; Aksoyoglu, M.; Zhang, L.; Andrade, S. L. A.; Schleicher, E.; Weber, S.; Rees, D. C.; Einsle, O. Evidence for Interstitial Carbon in Nitrogenase FeMo Cofactor. Science 2011, 334, 940. (3) Mayer, S. M.; Lawson, D. M.; Gormal, C. A.; Roe, S. M.; Smith, B. E. New insights into structure-function relationships in nitrogenase: A 1.6 angstrom resolution X-ray crystallographic study of Klebsiella pneumoniae MoFe-protein. J. Mol. Biol. 1999, 292, 871−891. (4) Einsle, O.; Tezcan, F. A.; Andrade, S. L. A.; Schmid, B.; Yoshida, M.; Howard, J. B.; Rees, D. C. Nitrogenase MoFe-protein at 1.16 E

DOI: 10.1021/acs.inorgchem.7b02493 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Free Mg2+ on ATP Linked Enzymes and the Calculation of Gibbs Free Energy of ATP Hydrolysis. J. Phys. Chem. B 2010, 114, 16137−16146. (25) Yandulov, D. V.; Schrock, R. R. Catalytic reduction of dinitrogen to ammonia at a single molybdenum center. Science 2003, 301, 76−78. (26) Durrant, M. C. An Atomic-Level Mechanism for Molybdenum Nitrogenase. Part 1. Reduction of Dinitrogen. Biochemistry 2002, 41, 13934−13945. (27) Durrant, M. C.; Francis, A.; Lowe, D. J.; Newton, W. E.; Fisher, K. Evidence for a dynamic role for homocitrate during nitrogen fixation: the effect of substitution at the α-Lys426 position in MoFeprotein of Azobacter vinelandi. Biochem. J. 2006, 397, 261−270.

angstrom resolution: A central ligand in the FeMo-cofactor. Science 2002, 297, 1696−1700. (5) Lancaster, K. M.; Roemelt, M.; Ettenhuber, P.; Hu, Y.; Ribbe, M. W.; Neese, F.; Bergmann, U.; DeBeer, S. X-ray Emission Spectroscopy Evidences a Central Carbon in the Nitrogenase Iron-Molybdenum Cofactor. Science 2011, 334, 974−977. (6) Cao, L.; Caldararu, O.; Ryde, U. Protonation States of Homocitrate and Nearby Residues in Nitrogenase Studied by Computational Methods and Quantum Refinement. J. Phys. Chem. B 2017, 121, 8242−8262. (7) Imperial, J.; Hoover, T. R.; Madden, M. S.; Ludden, P. W.; Shah, V. K. Substrate Reduction Properties of Dinitrogenase Activated Invitro are Dependent upon the Presence of Homocitrate or its Analogs during Iron Molybdenum Cofactor Synthesis. Biochemistry 1989, 28, 7796−7799. (8) Hoffman, B. M.; Lukoyanov, D.; Dean, D. R.; Seefeldt, L. C. Nitrogenase: A Draft Mechanism. Acc. Chem. Res. 2013, 46, 587−595. (9) Hoffman, B. M.; Lukoyanov, D.; Yang, Z. Y.; Dean, D. R.; Seefeldt, L. C. Mechanism of Nitrogen Fixation by Nitrogenase: The Next Stage. Chem. Rev. 2014, 114, 4041−4062. (10) Dance, I. The pathway for serial proton supply to the active site of nitrogenase: enhanced density functional modeling of the Grotthuss mechanism. Dalton Trans. 2015, 44, 18167−18186. (11) Siegbahn, P. E. M.; Blomberg, M. R. A. Proton Pumping Mechanism in Cytochrome c Oxidase. J. Phys. Chem. A 2008, 112, 12772−12780. (12) Blomberg, M. R. A.; Borowski, T.; Himo, F.; Liao, R.-Z.; Siegbahn, P. E. M. Quantum Chemical Studies of Mechanisms for Metalloenzymes. Chem. Rev. 2014, 114, 3601−3658. (13) Siegbahn, P. E. M.; Blomberg, M. R. A. Quantum Chemical Studies of Proton-Coupled Electron Transfer in Metalloenzymes. Chem. Rev. 2010, 110, 7040−7061. (14) Siegbahn, P. E. M. Model Calculations Suggest that the Central Carbon in the FeMo-Cofactor of Nitrogenase Becomes Protonated in the Process of Nitrogen Fixation. J. Am. Chem. Soc. 2016, 138, 10485− 10495. (15) Becke, A. D. Density functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652. (16) Grimme, S. Semiempirical hybrid density functional with perturbative second-order correlation. J. Chem. Phys. 2006, 124, 034108. Schwabe, T.; Grimme, S. Phys. Chem. Chem. Phys. 2007, 9, 3397−3406. (17) Jaguar, version 8.9; Schrodinger, Inc.: New York, 2015. Bochevarov, A. D.; Harder, E.; Hughes, T. F.; Greenwood, J. R.; Braden, D. A.; Philipp, D. M.; Rinaldo, D.; Halls, M. D.; Zhang, J.; Friesner, R. A. Jaguar: A high-performance quantum chemistry software program with strengths in life and materials sciences. Int. J. Quantum Chem. 2013, 113, 2110−2142. (18) Lukoyanov, D.; Pelmenschikov, V.; Maeser, N.; Laryukhin, M.; Yang, T. C.; Noodleman, L.; Dean, D. R.; Case, D. A.; Seefeldt, L. C.; Hoffman, B. M. Testing if the Interstitial Atom, X, of the Nitrogenase Molybdenum−̌ Iron Cofactor Is N or C: ENDOR, ESEEM, and DFT Studies of the S = 3/2 Resting State in Multiple Environments. Inorg. Chem. 2007, 46, 11437−11449. (19) Bjornsson, R.; Neese, F.; DeBeer, S. Revisiting the Mössbauer Isomer Shifts of the FeMoco cluster of Nitrogenase and the Cofactor Charge. Inorg. Chem. 2017, 56, 1470−1477. (20) Siegbahn, P. E. M.; Tye, J. W.; Hall, M. B. Computational Studies of [NiFe] and [FeFe] Hydrogenases. Chem. Rev. 2007, 107, 4414−4435. (21) CRC Handbook of Chemistry and Physics, 85th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2003. (22) Burgess, B. K.; Lowe, D. J. Mechanism of Molybdenum Nitrogenase. Chem. Rev. 1996, 96, 2983−3011. (23) Rees, D. C. Great Metalloenzymes in Enzymology. Annu. Rev. Biochem. 2002, 71, 221. (24) Veech, R. L.; Lawson, J. W. R.; Cornell, N. W.; Krebs, H. A. Cytosolic Phosphorylation Potential. J. Biol. Chem. 1979, 254, 6538− 6547. Bergman, C.; Kashiwaya, Y.; Veech, R. L. The effect of pH and F

DOI: 10.1021/acs.inorgchem.7b02493 Inorg. Chem. XXXX, XXX, XXX−XXX