MM Studies of Dph5 – A Promiscuous Methyltransferase in the

Jun 21, 2018 - Eukaryotic diphthine synthase, Dph5, is a promiscuous methyltransferase that catalyzes an extraordinary N,O-tetramethylation of ...
0 downloads 0 Views 2MB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Computational Biochemistry

QM/MM Studies of Dph5 – A Promiscuous Methyltransferase in the Eukaryotic Biosynthetic Pathway of Diphthamide Johanna Hörberg, Patricia Saenz-Méndez, and Leif A Eriksson J. Chem. Inf. Model., Just Accepted Manuscript • DOI: 10.1021/acs.jcim.8b00217 • Publication Date (Web): 21 Jun 2018 Downloaded from http://pubs.acs.org on June 22, 2018

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 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 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.

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 20 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

Journal of Chemical Information and Modeling

QM/MM Studies of Dph5 – A Promiscuous Methyltransferase in the Eukaryotic Biosynthetic Pathway of Diphthamide Johanna Hörberga, Patricia Saenz-Mendezb and Leif A. Erikssona* a Department of Chemistry and Molecular Biology, University of Gothenburg, 405 30 Göteborg, Sweden. b Computational Chemistry and Biology Group, Facultad de Química, Universidad de la República, 11800 Montevideo, Uruguay. *Corresponding autor: [email protected]

Graphical Abstract

Abstract Eukaryotic Diphthine Synthase, Dph5, is a promiscuous methyltransferase that catalyzes an extraordinary N,O-tetramethylation of 2-(3-carboxy-3-aminopropyl)-L-histidine (ACP) to yield diphthine methyl ester (DTM). These are intermediates in the biosynthesis of the posttranslationally modified histidine residue diphthamide (DTA), a unique and essential residue part of the eukaryotic elongation factor 2 (eEF2). Herein, the promiscuity of Saccharomyces cerevisiae Dph5 has been studied with in silico approaches, including homology modeling to provide the structure of Dph5, protein-protein docking and molecular dynamics to construct the Dph5-eEF2 complex, and quantum mechanics/molecular mechanics (QM/MM) calculations to outline a plausible mechanism. The calculations show that the methylation of ACP follows a typical SN2 mechanism, initiating with a complete methylation (trimethylation) at the Nposition, followed by the single O-methylation. For each of the three N-methylation reactions, our calculations support a stepwise mechanism, which first involve proton transfer through a bridging water to a conserved aspartate residue D165, followed by a methyl transfer. Once fully methylated, the trimethyl amino group forms a weak electrostatic interaction with D165, which allows the carboxylate group of diphthine to attain the right orientation for the final methylation step to be accomplished.

ACS Paragon Plus Environment

1

Journal of Chemical Information and Modeling 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 20

Introduction Diphthamide (DTA) is a post-translationally modified histidine residue of archaeal and eukaryotic elongation factor 2 (aEF2 and eEF2), where it is believed to increase the fidelity of translation by preventing -1 frame shift mutations.1-2 DTA also constitutes the target for ADPribosylation by bacterial toxins, including the Diphtheria toxin (Corynebacterium diphtheriae), Exotoxin A (Pseudomonas aeruginosa) and Cholix toxin (Vibrio cholerae). The ribosylation irreversibly inhibits EF2 from operating in protein translation, and thus automatically induces cell arrest.2-3 In recent years, the eukaryotic pathway for modification of histidine to diphthamide has been shown by Lin et al. to involve an additional step compared to that of archaea (Figure 1A).4 Seven enzymes, Dph1-7,1-2, 4-7 participate in the eukaryotic biosynthesis of DTA, where the first four enzymes catalyze the first step, namely the transfer of a 3-amino-3-carboxypropyl (ACP) tail from S-adenosylmethionine (SAM) to the CE1 position of the histidine residue (H699 in yeast and H715 in human).2 In archaea, the next step involves diphthine (DTI) formation through trimethylation of the amino group of ACP by the methyltransferase aDph5 (Figure 1A(i)). However, eukaryotic Dph5 (Figure 1A(ii)) exhibits a remarkable N,O-promiscuity to yield the previously overlooked intermediate, methylated diphthine (DTM),2,

4

which is

subsequently hydrolyzed to DTI by the esterase Dph7; the first WD40 protein with a known enzymatic activity. DTI, which constitutes the substrate of Dph6, is finally amidated to DTA.5 The evolutionary reason behind the extra O-methylation/demethylation step has been suggested to provide additional regulation,2, 4 which is consistent with the work of Uthman et al.7 In addition to co-immune precipitation results, demonstrating that the fraction of eEF2 bound to Dph5 was elevated in mutant cells lacking either Dph7 or Dph1, Uthman et al. observed that overexpression of Dph5 resulted in a reduction of cell growth in yeast strains lacking any of the genes coding for Dph1-Dph4 and Dph7, but had negligible effect on Dph6 deficient strains and no effect on the wild type.7 Hence, besides catalyzing a promiscuous N,Omethylation, Dph5 also seems to act as a key regulator of the pathway. It is very likely that such a regulatory mechanism has evolved to prevent erroneous translation by improperly modified eEF2 lacking the amidated DTA, and that the functional role of Dph7 has emerged to disrupt the Dph5-eEF2 interactions, thereby allowing for completion of dipthamide modification. This may indeed be a key process for evolution of higher developed eukaryotes, as DTA has been observed to be critical for the survival of those organisms.1-2 The promiscuous behavior of eukaryotic Dph5 and its role as a potential regulator in the biosynthetic pathway of DTA motivated us to further understand its mechanism. ACS Paragon Plus Environment

2

Page 3 of 20 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

Journal of Chemical Information and Modeling

Methyltransferases, which constitute a major class of proteins in the cell, predominately use SAM as a source for transfer of methyl groups to their substrates through either an SN2 mechanism or (less commonly) through a radical mechanism if the reaction involves methylation of unreactive C-H bonds.8 Since the catalyzed tetramethylation of the ACP intermediate involves methylation of nucleophilic functional groups, the reactions are expected to follow typical SN2 mechanisms. Available crystal structures of archaeal Dph5 have provided insight into the binding site of SAM and EF2 (Figure S1A-B).9 Structural analyzes have shown that the protein belongs to the homodimeric class III family of methyltransferases,9-10 where each monomer adopts a kidney type of shape with two similar α/β-domains linked by a hinge region. However, in contrast to the other members of this class, Dph5 only binds one molecule of SAM instead of two. Upon binding of SAM to its pocket, which is located at the hinge region, the two-fold symmetry of the Dph5 dimer is broken. This property has been suggested to be of great importance for recognition of larger proteins such as EF2 as substrates.9 In this study, in silico approaches are used to unveil the structural and functional features of Saccharomyces cerevisiae Dph5. Starting with homology modeling, followed by proteinprotein docking and finishing with QM/MM studies, we outline a plausible mechanism for the tetramethylation of ACP (Figure 1B). This involves complete methylation of the amino group through a stepwise mechanism that includes proton transfer through a bridging water to a conserved aspartate residue followed by SN2 methyl transfer. Once fully methylated, the amino group forms a salt-bridge with the conserved aspartate residue, which allows the carboxylate group to become methylated.

Results System preparation The required Dph5-eEF2 complex for QM/MM studies was constructed as follows. Since there were no available crystal structures of Saccharomyces cerevisae Dph5 (Dph5_SaCe), a homology model was generated in YASARA.11 The advantage of YASARA is that it utilizes multiple templates to construct a high quality hybrid model. The provided hybrid model of Dph5_SaCe (Figure S2A-B) was based on Entamoeba histolytica Dph5 and four mutated constructs of Pyrococcus horikoshii Dph5 (Dph5_PyHo). Quality assessment (Figure S2C-D and Table S2), both by YASARA and three additional validation tools (Veryfy3D,12 Errat13 and

ACS Paragon Plus Environment

3

Journal of Chemical Information and Modeling 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 20

RAMPAGE14) confirmed that the structure was well-modeled. In addition, the hybrid model included the demethylated form of SAM, S-adenosyl-L-homocystein (SAH), thus already defining the active site. Through multiple sequence alignment (Figure S3) and analysis of the interactions of SAH with Dph5_SaCe (Figure S2B), it was evident that most of the residues that corresponded to the key residues of the Dph5_PyHo-SAH complex were involved in interactions, which implied that the ligand was well positioned during the homology modeling process, and that the active site was accurately modeled. SAH was modified to SAM, and the stability of the Dph5_SaCe-SAM complex was explored through MD simulations in Gromacs15; as indicated by RMSD calculations, the complex converged after 20 ns and remained highly stable throughout the whole 100 ns run (Figure S2E). The active site remained stable, and the complex exhibited a large number of hydrogen bond interactions throughout the whole simulation (Figure S2F). Through clustering and analysis of the number of hydrogen bond interactions between SAM and key residues of Dph5, a snapshot was selected for protein-protein docking (Figure S2G-H). The selected crystal structure of Saccharomyces cerevisiae eEF2 (PDB ID: 1N0V)16 also needed to be prepared as it contained a missing loop of 17 residues (K50-G67; Figure S4). Hence, loop modeling in YASARA was performed before the generation of the Dph5-eEF2 complex. Following loop modeling, H699 was modified to ACP.

Protein-Protein Docking The Dph5-eEF2 complex was constructed through protein-protein docking in PatchDock,17-18 in which a distance constraint between the amino group of ACP and the methyl group of SAM was set to 3-6 Å to allow them to be in close proximity without forming too much steric clash between the docking partners. This generated three decoys of which the most reasonable structure of the complex, illustrated in Figure 1B, corresponded to the highest scoring decoy. Seven conserved residues of archaeal Dph5, located near the surface, had been predicted by Kishishita et al.9 to potentially participate in interactions with EF2. The corresponding analogues for Dph5_SaCe (R59, E63, A89, T91, Y126, S133 and D165, highlighted in Figure S5) were identified through multiple sequence analysis (Figure S3). By analyzing the obtained Dph5-eEF2 complex (Figure 1B), we observed that the amino group of ACP interacts with the peptide backbone of A89 and the carboxylate group interacts with the backbone and side chain of T91 of Dph5. Other parts of eEF2 were also in close proximity allowing the interaction with residues R59, E63 and Y126. Noteworthy, the provided docking pose was highly similar to the

ACS Paragon Plus Environment

4

Page 5 of 20 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

Journal of Chemical Information and Modeling

manually constructed complex of Dph5_PyHo and Saccharomyces cerevisiae eEF2 by Kishishita et al.9 This further supports the reliability of the predicted pose by PatchDock. Regarding methylation at the nitrogen position of the substrates, some additional thoughts are required as these are predominately protonated at physiological pH and thus need to lose a proton before methylation can occur. Some methyltransferase enzymes bind the substrates in their neutral form,19-20 whereas others abstract the proton from the nitrogen atom through a negatively charged residue.21-23 ACP, which will be methylated three times at the nitrogen position, simply cannot bind in its deprotonated form; it needs to lose the protons within the active site of Dph5. In this sense, D165 is of particular interest and could potentially be involved in the proton abstraction; however, the distance to the closest proton of ACP (5.67 Å) is too far to form a hydrogen bond interaction (Figure 1C). It also seems rather unlikely that the complex will undergo considerable conformational changes to decrease the distance between D165 and ACP

and

at

the same time maintain

a reasonable distance

and

SN2

angle

(NACP•••MeSAM•••SSAM) between SAM and ACP. We thus hypothesized that the proton abstraction would involve a bridging water between D165 and ACP; a mechanism that has been observed for other methyltransferases.22-23

Molecular Dynamic Simulations To account for effects of protein interactions upon complex formation, the Dph5-eEF2(ACP) complex was subjected to MD refinement. This process required an initial restrained simulation (220 ns) with position constraints on various Cα atoms to relax large fluctuations of flexible parts and avoid dissociation of the complex, and to assure parts important for the reaction were maintained, until a 30 ns unrestrained simulation run was successful. RMSD calculations of the constraint run and production run (Figure S6A-B) indicated a stabilized system. For the production run, the SN2-angle and the distance between the amino group of ACP and the methyl group of SAM were analyzed over the trajectory (Figure S6C-D). As seen, these values fluctuate during the simulations, which highlight the dynamic feature of the complex and the spacious nature of the active site. For collection of a proper snapshot for QM/MM studies, the SN2-angle, the distance between ACP and SAM, and the presence of a bridging water were considered (a bridging water could be seen for snapshots with an SN2-angle > 140° and an ACPSAM distance < 5 Å). The selected snapshot (Figure 1D) showed a proper distance (3.73 Å) and angle (177°) for a nucleophilic attack. In addition, the interactions with A89 and T91 had been retained.

ACS Paragon Plus Environment

5

Journal of Chemical Information and Modeling 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 20

Each methylation step will be followed by large dynamical changes, where SAH is exchanged for another methylated SAM molecule and D165 is deprotonated. To account for these dynamical features during the reaction event, ACP of the equilibrated Dph5-eEF2 complex was modified to the expected intermediates of the tetramethylation (monomethylated ACP (ACM), dimethylated ACP (ACD) and DTI, respectively) and additional MD simulations were performed to collect snapshots required for investigation of the different stages of the reaction mechanism (Figure S7A-D, S8A-D and S9A-D, respectively). ACM and ACD exhibited similar fluctuations in SN2-angle and distance toward SAM (Figure S7B-C and S8B-C). In contrast, DTI exhibited a more rigid structure during the simulation, with an angle closer to 180° and a distance toward SAM below 3 Å (Figure S9B-C). The collected snapshots showed that a bridging water had been retained for both ACM and ACD (Figure S7D and S8D). For DTI (Figure S9D), the trimethyl amino group had instead approached D165 to form weak ionic interactions.

QM/MM Calculations Monomethylation of ACP For the QM/MM studies, a two layered ONIOM scheme24-25 provided by Gaussian 0926 was utilized. The selected QM part included the methionine part of SAM, the side chain of D165, the bridging water, the ACP tail and most atoms of the residues G88-T91. The latter turned out to be important in order to reduce the nucleophilicity of the carboxylate group of ACP, as this would otherwise abstract a proton from the amino group already at the stage of geometry optimization. The QM region was treated with a standard DFT protocol (B3LYP/6-31G(d,p)), whereas the MM region was described with UFF to avoid parameterization. The optimized Michaelis complex (R in Figure 1E) showed a slight decrease in the distance (3.58 Å) and angle (155°) between ACP and SAM. Stronger hydrogen bond interactions between ACP, D165 and the bridging water could also be observed. Following optimization, we investigated if the mechanism was either concerted or stepwise. Assuming a concerted mechanism this was found to result in the protonation of the wrong functional group, in that an intramolecular proton transfer to the carboxylate group of ACP occurred. This was followed by unreasonable structural rearrangements of the carboxyl group to transfer the proton to D165 before the methyl group had been properly transferred to the amine group of ACP. As a result, a true TS could not be localized, and it was concluded that the reaction was best described as occurring in a stepwise manner. The stepwise mechanism (Figure 2A-B and Figure S11A(a)-A(b)) first

ACS Paragon Plus Environment

6

Page 7 of 20 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

Journal of Chemical Information and Modeling

involves a proton transfer from ACP to D165 through the bridging water. This provides an intermediate (Figure S11A(a)) where strong hydrogen bond interactions between ACP, the bridging water and D165 exists, and the distance between ACP and the methyl group of SAM has started to decrease. The neutralized amino group then rotates and loses its contact with the backbone of A89 to approach the second TS (Figure 2B), in which the methyl group is transferred through an SN2 mechanism with an almost perfectly linear configuration (174°). During TS2, the methyl group is almost completely planar, and located slightly closer to the sulphur atom than to the nitrogen (2.21 Å compared to 2.29 Å) . SAM is thus converted to SAH and inversion at the carbon center of the methyl group yields monomethylated ACP (ACM). At the product state (Figure S11A(b)), ACM also regains its interaction with the backbone of A89. The barriers for the proton transfer and methyl transfer were predicted to be approximately 5.0 and 10 kcal/mol respectively, and the overall reaction is exothermic, -38 kcal/mol (Table 1) To understand the nature of the optimized geometries (R, TS1, I, TS2 and P) from the QM/MM calculations, DFT calculations on only the QM region were performed (Figure S13A(a)-A(b)); this provided imaginary frequencies for the proton transfer (1085i cm-1) and methyl transfer (406i cm-1), respectively, but none for the reactants, intermediates and products, which assured that true minima and transition states had been found and that the QM/MM data was reliable. The obtained frequency for the methyl transfer is similar to those reported in studies of SN2-methyl transfer reactions catalyzed by other known methyltransferases.20, 23 The estimated barriers for the DFT calculations (5.0 kcal/mol for TS1 and 18.8 kcal/mol for TS2) show that the surrounding protein environment is important for stabilization of the methyl transfer (black lines in Figure 3A-B). The DFT calculations also show a reduction in the exothermicity of the reaction (-23 kcal/mol compared to -38 kcal/mol at the QM/MM level).

Dimethylation of ACP For the second step, that is the methylation of ACM to yield ACD (Figure 2C-D and Figure S11B(a)-B(c)), the mechanism involving proton transfer followed by methyl transfer was repeated, and the obtained energy profile (red lines in Figure 3A-B) was almost identical to that of the methylation of ACP. The barrier for the proton transfer was slightly larger (6.5 kcal/mol compared to 5.0 kcal/mol; Table 1), whereas for the methyl transfer the barrier was instead lower (7.7 kcal/mol compared to 10 kcal/mol; Table 1). In addition, the SN2 angle deviated a bit from an optimal linear configuration (164° instead of 180°). Frequency calculations on the methyl transfer (Figure S13B(a)-B(b)) confirmed the identified stationary point to be an SN2 TS.

ACS Paragon Plus Environment

7

Journal of Chemical Information and Modeling 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 20

Tri- and Tetramethylation of ACP For the trimethylation and tetramethylation of ACP, some difficulties were encountered. For ACD, the proton transfer to D165 occurred already at the stage of optimization (Figure S12A(a)). By analyzing the selected snapshot from the MD simulation (Figure S8D), we could observe that the bridging water was highly activated by D165 with a OSOL•••HSOL•••OD165 angle of 179°. This might explain the negligible barrier for proton transfer.27-28 In the subsequent step, the methyl group was transferred to ACD through an SN2 angle close to a linear configuration (168°) to complete the methylation at the N-position (Table 2, Figure 4A and Figure S13C(a)). In the TS, the methyl group was closer to the amino group than to the sulphur (2.21 Å compared to 2.41 Å), and the estimated barrier (30.4 kcal/mol) was significantly higher compared to the first two methylations (Table 1, Figure 3). Such a high barrier is not in agreement with those observed for other studied methyl transfers.19-23 We hypothesized that the amino group will be fully methylated first, since trimethylation is the only process that occurs in archaea. However, to investigate the possibility of other pathways, scanning of the methyl group towards the carboxylate group was performed. In the optimized Michaelis complex (Figure S12A(a)), the methyl group of SAM was closer to the carboxylate group than to the dimethylated amine group (3.89 Å compared to 4.58 Å) but the angle was lower (95° compared to 134°). Scanning followed by TS optimization yielded a barrier of approximately 11 kcal/mol (Table 2, Figure 3), which indicates that this alternative pathway cannot be rejected; both the barrier and reaction energy implies that this pathway is energetically more favorable and thus more likely to occur. The lower barrier can be understood by analyzing the TS (Figure 4B and Figure S12C(b)); the TS in the O-methylation pathway is stabilized through a hydrogen bond network, where both the bridging water and T91 stabilize the orientation of the carboxylate group to facilitate methylation. This hydrogen bond network also stabilizes the provided product, DDM (Figure S12A(c)), which explains the increase in exothermicity (Table 2 and Figure 3). Given that the tetramethylation of ACP is an enzyme catalyzed reaction that involves large dynamical events, which cannot entirely be accounted for with the QM/MM approach, and the fact that proton transfer occurred already at the stage of optimization for ACD, another reasonable pathway could be that the deprotonation of the amino group occurs while SAH is exchanged for another SAM, and once the methyl transfer is about to proceed, the proton has been transferred out from the active site. To explore this possibility, the bridging water and the proton were removed from the QM part, and through optimization, both the distance (3.56 Å)

ACS Paragon Plus Environment

8

Page 9 of 20 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

Journal of Chemical Information and Modeling

and angle (145°) between SAM and ACD were improved (Figure S12B(a)). Scanning of the methyl transfer, followed by TS optimization, provided a TS (Figure 4C and Figure S13C(c)), with an estimated barrier of approximately 7 kcal/mol (Table 2, Figure 3), which suggests that improper description of dynamic features could be responsible for the questionably high barrier initially estimated for the N-methylation of ACD (Figure 4A). To predict which one is the most probable pathway for the complete methylation of ACP, the two possible pathways for the tetramethylation (methylation of amine group in DDM or methylation of carboxylate group in DTI) were evaluated. Additional MD simulations were first performed to identify a suitable snapshot for the QM/MM study of the methylation of DDM (Figure S10A-D). In the first situation, the N-methylation of DDM (Figure 5A, Figure S12C(a)-C(b) and Figure S13D(a)), no hydrogen bond interactions were present to stabilize the TS, which thus provided an unreasonably high barrier of approximately 28 kcal/mol (Table 2, Figure 3A). In, addition, the exothermicity (-5.4 kcal/mol) of the reaction was considerably lower than for the other studied methylations (Figure 3A). For the other potential pathway of the tetramethylation, that is the O-methylation of DTI (Figure 5B, Figure S12D(a)-D(b) and Figure S13D(b)), the substrate lost its contact with the backbone of residues A89-T91 as the trimethyl amino group approached D165, which in turn positioned the carboxylate group at a proper distance (2.91 Å) and angle (159°) toward SAM (Figure S12D(a)) for the reaction to take place. The methylation of DTI proceeded through a SN2 mechanism (171°) to yield DTM (Figure 5B and Figure S12D(b)), with a barrier of approximately 20 kcal/mol (Table 2). The reaction energy for this pathway is thus significantly lower than for the methylation of DDM. In addition, by comparing the energy profile plots (Figure 3A-B), we observe that in comparison to the DFT calculations, the QM/MM calculations show lower barriers and higher exothermicity for all studied methylations except for the N-methylation of DDM, where the reverse trend is observed. This indicates that the surrounding protein environment has a destabilizing effect on that reaction. We thus conclude that the most probable reaction sequence for the methylation of ACP involves complete methylation at the N-position followed by methylation at the O-position.

Conclusions This study set out to unveil the promiscuity of eukaryotic Dph5 trough in silico approaches. Saccharomyces cerevisiae was selected as the target organism, and the structure of Dph5 was obtained through homology modeling. Protein-protein docking provided the Dph5-eEF2 ACS Paragon Plus Environment

9

Journal of Chemical Information and Modeling 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 20

complex, and through MD simulations, snapshots for ACP and the other methylated intermediates were obtained, which were used for QM/MM studies of the explicit reactions. Through the use of QM/MM calculations we were able to outline a plausible mechanism, where the estimated barrier for each methylation was comparable to other known methyltransferases.19-23 The calculations revealed that the tetramethylation proceeded through a SN2 mechanism, with complete methylation at the N-position to provide DTI occurring first, followed by methylation at the O-position to provide the previously overlooked intermediate DTM. The trimethylation of the amino group of ACP followed stepwise mechanisms, where first a proton transfer proceeded through a bridging water to D165, followed by the SN2-methyl transfer. The trimethyl amino group of DTI then approached D165 to form weak ionic interactions, which positioned the carboxylate group at a good distance and angle toward SAM for the final methylation. The mechanism for the proton transfer was readily determined for the first two methylation steps; however, for deprotonation of dimethylated ACP (ACD), the barrier was negligible as the proton transfer occurred already during the optimization step. The subsequent barrier for methyl transfer to ACD was unreasonably high, which could be rationalized in terms of lack of dynamics in the calculations, an inherent limitation with the QM/MM approach. For future studies, we recommend to focus more on the dynamics of the reaction and the SAH/SAM exchange process, as this could provide more information about the proton transfer, to determine more definitively at which stage it occurs and when and how the proton is transferred out of the active site thereby re-activating D165. This could also provide more details about other aspects of the dynamic process where the amine group disrupts its interactions with the bridging water to approach SAM in order to become methylated. This in turn could provide even more accurate barriers. Based on our calculations it seems likely that, in the case of ACD, the proton has already been transported out from the active site before the third N-methylation reaction occurs. The current study provides a good starting point for experimental studies, where the catalytic role of D165 could be unveiled through mutational studies. Dph5 was shown through multiple alignments (Figure S3) to be a highly conserved enzyme, hence the exhibited promiscuity cannot be explained only on the basis of its structure; it is thus possible that differences in aEF2 and eEF2 in combination with different cellular conditions have a more prominent influence on the distinct behavior of archaeal and eukaryotic Dph5.

ACS Paragon Plus Environment

10

Page 11 of 20 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

Journal of Chemical Information and Modeling

Acknowledgement We gratefully acknowledge funding provided by the Swedish Research Council (VR) through grant number 2014-3914 and the Faculty of Science at University of Gothenburg, and generous grants of computing time provided by the Swedish National Infrastructure for Computing (SNIC).

Author contributions All authors designed the initial study. JH performed all calculations and wrote first version of manuscript. All authors revised manuscript and contributed with analysis of data. All authors approved the final version of the manuscript.

Conflict of Interest The authors declare no conflicting interests.

Supporting Information File 1: Method section. Tables for settings for the homology modeling and quality assessment of the homology model of Dph5. Supplementary figures for the crystal structure of Pyrococcus horikoshii Dph5; preparation of the homology model of Saccharomyces cerevisae Dph5; multiple sequence alignment; preparation of the crystal structure of Saccharomyces cerevisae eEF2; the active site of Dph5; analysis of the MD simulations of the Dph5eEF2(ACP,ACM,ACD,DTI,DDM) complexes; additional optimized geometries in the QM/MM calculations; all optimized transition states with DFT calculations for only the QMpart scanning diagrams, and parameters for running MD simulations with the studied diphthamide precursors. File 2: Cartesian coordinates for all ONIOM-optimized geometries (only the QM-part is shown), and Cartesian coordinates for all QM-optimized geometries.

ACS Paragon Plus Environment

11

Journal of Chemical Information and Modeling 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 20

Table 1: Geometric Parameters and Barriers for the Proton Transfer (TS) and Methyl Transfer (TS), Together with Reactions Energies for the Monomethylation and Dimethylation of ACP

Table 2: Geometric Parameters and Barriers for the Methyl Transfer (TS), Together with Reactions Energies for the Trimethylation and Tetramethylation of ACP

ACS Paragon Plus Environment

12

Page 13 of 20 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

Journal of Chemical Information and Modeling

Figure Legends Figure 1: (A) Biosynthetic pathway of diphthamide in (i) archaea and (ii) eukaryotes. Modified from ref.4. (B) Proposed mechanism for the tetramethylation of ACP. (C) The Dph5-eEF2 complex, obtained through protein-protein docking in PatchDock, and a zoom-in of the active site showing the interactions between ACP and Dph5. The distance and angle between ACP and SAM is 4.18 Å and 163° respectively, and the distance between D165 and the closest proton of ACP is 5.67 Å. ACP exhibits hydrogen bond interactions with A89 and T91. The figure was generated in MOE2016.08.29 (D) Selected snapshot from MD refinement of the Dph5-eEF2(ACP) complex to be used in QM/MM studies. Blue lines illustrate hydrogen bond interactions. The figure was generated in Chimera.30 (E) The QM part of the QM/MM optimized Michaelis complex (ONIOM(B3LYP/6-31G(d,p):UFF)) with key distances and angles highlighted. The figure was generated in GaussView.31 𝛉𝛉: SN2-angle, 𝟇𝟇 (OSOL•••HSOL•••OD165) and 𝞿𝞿 (NACP•••HACP•••OSOL). Figure 2: Optimized transition states (stepwise mechanism) for (A) deprotonation of ACP (B) methylation of ACP (C) deprotonation of ACM (D) methylation of ACM. Key distances and angles are highlighted. Figure 3: Relative energies (∆E) for all the studied methylations, obtained through single point energy calculations for (A) the QM/MM system and (B) the QM system. The sequence R-TS1I corresponds to proton transfer, and the sequence I-TS2-P corresponds to methyl transfer. Figure 4: Optimized TS for the three studied pathways for the trimethylation of ACP (A) Nmethylation of ACD to provide DTI, (B) O-methylation of ACD to provide DDM, and (C) Nmethylation of ACD with deprotonated D165 and no bridging water to give DTI. Figure 5: Optimized TS for the two potential pathways of the tetramethylation of ACP (A) Nmethylation of DDM (B) O-methylation of DTI.

ACS Paragon Plus Environment

13

Journal of Chemical Information and Modeling

Page 14 of 20

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

ACS Paragon Plus Environment

14

Page 15 of 20 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

Journal of Chemical Information and Modeling

Figure 1

Figure 2

ACS Paragon Plus Environment

15

Journal of Chemical Information and Modeling 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 20

Figure 3

ACS Paragon Plus Environment

16

Page 17 of 20 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

Journal of Chemical Information and Modeling

Figure 4

ACS Paragon Plus Environment

17

Journal of Chemical Information and Modeling 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 20

Figure 5

ACS Paragon Plus Environment

18

Page 19 of 20 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

Journal of Chemical Information and Modeling

References 1. Su, X.; Lin, Z.; Lin, H., The Biosynthesis and Biological Function of Diphthamide. Crit. Rev. Biochem. Mol. Biol. 2013, 48, 515-521. 2. Schaffrath, R.; Abdel-Fattah, W.; Klassen, R.; Stark, M. J. R., The Diphthamide Modification Pathway from Saccharomyces Cerevisiae – Revisited. Mol. Microbiol. 2014, 94, 1213-1226. 3. Jørgensen, R.; Wang, Y.; Visschedyk, D.; Merrill, A. R., The Nature and Character of the Transition State for the ADP‐Ribosyltransferase Reaction. EMBO Rep. 2008, 9, 802-809. 4. Lin, Z.; Su, X.; Chen, W.; Ci, B.; Zhang, S.; Lin, H., Dph7 Catalyzes a Previously Unknown Demethylation Step in Diphthamide Biosynthesis. J. Am. Chem. Soc. 2014, 136, 6179-6182. 5. Su, X.; Chen, W.; Lee, W.; Jiang, H.; Zhang, S.; Lin, H., YBR246W Is Required for the Third Step of Diphthamide Biosynthesis. J. Am. Chem. Soc. 2012, 134, 773-776. 6. Dong, M.; Su, X.; Dzikovski, B.; Dando, E. E.; Zhu, X.; Du, J.; Freed, J. H.; Lin, H., Dph3 Is an Electron Donor for Dph1-Dph2 in the First Step of Eukaryotic Diphthamide Biosynthesis. J. Am. Chem. Soc. 2014, 136, 1754-1757. 7. Uthman, S.; Bär, C.; Scheidt, V.; Liu, S.; ten Have, S.; Giorgini, F.; Stark, M. J. R.; Schaffrath, R., The Amidation Step of Diphthamide Biosynthesis in Yeast Requires DPH6, a Gene Identified through Mining the DPH1-DPH5 Interaction Network. PLoS Genet. 2013, 9, e1003334. 8. Zhang, Q.; van der Donk, W. A.; Liu, W., Radical-Mediated Enzymatic Methylation: A Tale of Two SAMS. Acc. Chem. Res. 2012, 45, 555-564. 9. Kishishita, S.; Shimizu, K.; Murayama, K.; Terada, T.; Shirouzu, M.; Yokoyama, S.; Kunishima, N., Structures of two Archaeal Diphthine Synthases: Insights into the PostTranslational Modification of Elongation Factor 2. Acta Cryst. D 2008, 64, 397-406. 10. Schubert, H. L.; Blumenthal, R. M.; Cheng, X., Many Paths to Methyltransfer: a Chronicle of Convergence. Trends Biochem. Sci. 2003, 28, 329-335. 11. Krieger, E.; Vriend, G., YASARA View—Molecular Graphics for All Devices— from Smartphones to Workstations. Bioinformatics 2014, 30, 2981-2982. 12. Bowie, J. U.; Luthy, R.; Eisenberg, D., A Method to Identify Protein Sequences that Fold into a Known Three-Dimensional Structure. Science 1991, 253, 164-170. 13. Colovos, C.; Yeates, T. O., Verification of Protein Structures: Patterns of Nonbonded Atomic Interactions. Protein Sci. 1993, 2, 1511-1519. 14. Lovell, S. C.; Davis, I. W.; Arendall, W. B.; de Bakker, P. I. W.; Word, J. M.; Prisant, M. G.; Richardson, J. S.; Richardson, D. C., Structure Validation by Cα Geometry: ϕ,ψ and Cβ Deviation. Proteins: Struct., Funct., Genet. 2003, 50, 437-450. 15. Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E., GROMACS 4:  Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J. Chem. Theory Comput. 2008, 4, 435-447. 16. Jorgensen, R.; Ortiz, P. A.; Carr-Schmid, A.; Nissen, P.; Kinzy, T. G.; Andersen, G. R., Two Crystal Structures Demonstrate Large Conformational Changes in the Eukaryotic Ribosomal Translocase. Nat. Struct. Mol. Biol. 2003, 10, 379-385. 17. Duhovny, D.; Nussinov, R.; Wolfson, H. J., Efficient Unbound Docking of Rigid Molecules. Lect. Notes Comput. Sci. 2002, 2452, 185-200. 18. Schneidman-Duhovny, D.; Inbar, Y.; Nussinov, R.; Wolfson, H. J., PatchDock and SymmDock: Servers for Rigid and Symmetric Docking. Nucleic Acids Res. 2005, 33, W363-W367.

ACS Paragon Plus Environment

19

Journal of Chemical Information and Modeling 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 20

19. Georgieva, P.; Himo, F., Quantum Chemical Modeling of Enzymatic Reactions: The Case of Histone Lysine Methyltransferase. J. Comput. Chem. 2010, 31, 1707-1714. 20. Velichkova, P.; Himo, F., Theoretical Study of the Methyl Transfer in Guanidinoacetate Methyltransferase. J. Phys. Chem. B 2006, 110, 16-19. 21. Zhang, R.; Li, X.; Liang, Z.; Zhu, K.; Lu, J.; Kong, X.; Ouyang, S.; Li, L.; Zheng, Y. G.; Luo, C., Theoretical Insights into Catalytic Mechanism of Protein Arginine Methyltransferase 1. PLoS One 2013, 8, e72424. 22. Hou, Q. Q.; Wang, J. H.; Gao, J.; Liu, Y. J.; Liu, C. B., QM/MM Studies on the Catalytic Mechanism of Phenylethanolamine N-Methyltransferase. Biochim. Biophys. Acta 2012, 1824, 533-541. 23. Saen-oon, S.; Lee, S. G.; Jez, J. M.; Guallar, V., An Alternative Mechanism for the Methylation of Phosphoethanolamine Catalyzed by Plasmodium Falciparum Phosphoethanolamine Methyltransferase. J. Biol. Chem. 2014, 289, 33815-33825. 24. Vreven, T.; Byun, K. S.; Komáromi, I.; Dapprich, S.; Montgomery, J. A.; Morokuma, K.; Frisch, M. J., Combining Quantum Mechanics Methods with Molecular Mechanics Methods in ONIOM. J. Chem. Theory Comput. 2006, 2, 815-826. 25. Shaik, S.; Cohen, S.; Wang, Y.; Chen, H.; Kumar, D.; Thiel, W., P450 Enzymes: Their Structure, Reactivity, and Selectivity—Modeled by QM/MM Calculations. Chem. Rev. 2010, 110, 949-1017. 26. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, B.; 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, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; 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.; Fox, D. J. Gaussian 09;, Gaussian, Inc.: Wallingford, USA, 2009. 27. Duan, X.; Scheiner, S., Energetics, proton transfer rates, and kinetic isotope effects in bent hydrogen bonds. J. Am. Chem. Soc. 1992, 114, 5849-5856. 28. Rangelov, M. A.; Vayssilov, G. N.; Yomtova, V. M.; Petkov, D. D., The synOriented 2-OH Provides a Favorable Proton Transfer Geometry in 1,2-Diol Monoester Aminolysis:  Implications for the Ribosome Mechanism. J. Am. Chem. Soc. 2006, 128, 49644965. 29. Molecular Operating Environment (MOE), 2016.08; Chemical Computing Group Inc: 1010 Sherbrooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7, 2017. 30. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E., UCSF Chimera—A Visualization System for Exploratory Research and Analysis. J. Comput. Chem. 2004, 25, 1605-1612. 31. Dennington, R.; Keith, T. A.; Millam, J. M. GaussView, 5.0; Semichem Inc: Shawnee Mission KS, 2009.

ACS Paragon Plus Environment

20