The bacterial enzyme MenD, or 2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexene-1-carboxylate (SEPHCHC) synthase, catalyzes an essential Stetter reaction in menaquinone (vitamin K2) biosynthesis via thiamine diphosphate (ThDP)-bound tetrahedral post-decarboxylation intermediates. The detailed mechanism of this intermediate chemistry, however, is still poorly understood, but of significant interest given that menaquinone is an essential electron transporter in many pathogenic bacteria. Here, we used site-directed mutagenesis, enzyme kinetic assays, and protein crystallography to reveal an active–inactive intermediate equilibrium in MenD catalysis and its modulation by two conserved active site arginine residues. We observed that these conserved residues play a key role in shifting the equilibrium to the active intermediate by orienting the C2-succinyl group of the intermediates through strong ionic hydrogen bonding. We found that when this interaction is moderately weakened by amino acid substitutions, the resulting proteins are catalytically competent with the C2-succinyl group taking either the active or the inactive orientation in the post-decarboxylation intermediate. When this hydrogen-bonding interaction was strongly weakened, the succinyl group was re-oriented by 180° relative to the native intermediate, resulting in the reversal of the stereochemistry at the reaction center that disabled catalysis. Interestingly, this inactive intermediate was formed with a distinct kinetic behavior, likely as a result of a non-native mode of enzyme–substrate interaction. The mechanistic insights gained from these findings improve our understanding of the new ThDP-dependent catalysis. More importantly, the non-native-binding site of the inactive MenD intermediate uncovered here provides a new target for the development of antibiotics.

Introduction

MenD, or 2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexene-1-carboxylate (SEPHCHC) synthase, is involved in the biosynthesis of menaquinone or vitamin K2 [14]. It is a thiamine diphosphate (ThDP)-dependent enzyme catalyzing the conversion of α-ketoglutarate and isochorismate into SEPHCHC in a multistep reaction (Figure 1) [5,6], which is irreversible and indispensable in the biosynthetic pathway [79]. Since menaquinone is an essential electron transporter in many pathogenic bacteria, such as Staphylococcus aureus and Haemophilus influenzae, MenD is an attractive target like other enzymes in the same pathway for the development of new antibiotics [1014].

Catalytic mechanism of the SEPHCHC synthase (MenD) in menaquinone biosynthesis.

Figure 1.
Catalytic mechanism of the SEPHCHC synthase (MenD) in menaquinone biosynthesis.

(A) Overall reaction catalyzed by MenD. (B) Proposed structure of the tetrahedral intermediate (E-Int) in the MenD-catalyzed reaction. The N4′-C distance and the hydrogen-bonding interaction of Arg395 and Arg413 with the tetrahedral intermediate are from the crystal structure 5EJ8 deposited in the Protein Data Bank [15]. (C) Simplified kinetic mechanism in MenD catalysis. kapp1 and k2 represent a first-order or pseudo first-order process determined previously in the absence of the isochorismate substrate using stopped-flow CD spectroscopy [16]. E-ThDP-2KG is the pre-decarboxylation intermediate, while both E-Int* and E-Int are post-decarboxylation intermediates. 2KG: α-ketoglutarate; SEPHCHC: (1R, 2S, 5S, 6S)-2-succinyl-5-enolpyruvyl-6-hydroxyl-3-cyclohexene-1-carboxylate; ThDP: thiamine diphosphate.

Figure 1.
Catalytic mechanism of the SEPHCHC synthase (MenD) in menaquinone biosynthesis.

(A) Overall reaction catalyzed by MenD. (B) Proposed structure of the tetrahedral intermediate (E-Int) in the MenD-catalyzed reaction. The N4′-C distance and the hydrogen-bonding interaction of Arg395 and Arg413 with the tetrahedral intermediate are from the crystal structure 5EJ8 deposited in the Protein Data Bank [15]. (C) Simplified kinetic mechanism in MenD catalysis. kapp1 and k2 represent a first-order or pseudo first-order process determined previously in the absence of the isochorismate substrate using stopped-flow CD spectroscopy [16]. E-ThDP-2KG is the pre-decarboxylation intermediate, while both E-Int* and E-Int are post-decarboxylation intermediates. 2KG: α-ketoglutarate; SEPHCHC: (1R, 2S, 5S, 6S)-2-succinyl-5-enolpyruvyl-6-hydroxyl-3-cyclohexene-1-carboxylate; ThDP: thiamine diphosphate.

MenD is structurally similar to other ThDP-dependent enzymes [17,18], but has been shown to adopt new intermediate chemistry. It was found to form a tetrahedral Breslow intermediate (E-Int) in catalysis [15,19], rather than the enamine intermediate found in many other ThDP-dependent enzymes [2023]. This tetrahedral intermediate is proposed to exist in a tautomerization equilibrium between a carbanion and a fully protonated species (Figure 1B) and is structurally distinct from other non-enamine Breslow intermediates [2428]. More recently, single-turnover kinetic studies have suggested a new tetrahedral post-decarboxylation intermediate (E-Int*) [16], which is formed in a pseudo first-order reaction with its rate constant proportional to the concentration of the α-ketoglutarate substrate and converted at a moderate rate (∼4.8 s−1) to the more stable E-Int likely through rotation of the C2–C bond (Figure 1C). In Escherichia coli MenD, formation of this new intermediate is rate-limited by binding of the α-ketoglutarate substrate in a pseudo first-order process. Owing to the low physiological concentration of the electrophilic substrate, isochorismate, the tetrahedral intermediates are kinetically stabilized in the catalytic process with the product release as the rate-limiting step, in contrast with coupling of enamine generation with binding of the electrophilic acceptor in some other ThDP-dependent enzymes [29]. Despite these achievements, mechanism of the new intermediate chemistry is not fully understood, particularly in stabilization of the intermediates and modulation of their reactivity.

In the E. coli MenD active site, Arg395 and Arg413 were found to form multiple ionic hydrogen bonds with the terminal succinyl carboxylate of the intermediate (Figure 1B) [15]. This strong interaction is believed to orient the C2-succinyl group to result in the unique stereochemistry of the C reaction center, which is distinct from that of the ThDP-bound tetrahedral adducts in many other enzymes [23,3032]. The importance of these residues is supported by more than 100-fold decrease in catalytic efficiency when E. coli MenD Arg395 or its equivalent residue in Bacillus subtilis MenD is mutated [13,18,33]. These previous results strongly suggest that the two conserved residues play an important role in stabilizing and modulating the tetrahedral intermediates.

In this study, we mutated Arg395 and Arg413 of E. coli MenD into alanine or lysine to further understand their roles in the catalysis. The site-directed mutants were evaluated for their ability to form the post-decarboxylation intermediate and the resulting complexes were structurally determined by X-ray crystallography. Meanwhile, the mutant-catalyzed reactions were analyzed by real-time monitoring of the chemical state of the ThDP cofactor with circular dichroism (CD) spectroscopy. Results from these structural and kinetic studies have provided fresh mechanistic insights into the catalytic roles of the two conserved residues. More importantly, these studies have revealed an alternative enzyme–substrate interaction mode, which produces an inactive post-decarboxylation intermediate with a totally different C2-succinyl orientation and provides a new inhibitory site for the development of new antibiotics.

Experimental

MenD and its mutants

The wild-type E. coli MenD and its mutants were expressed in pET28a without any tag. The plasmid containing the menD sequence from the previous study [15,16] was used for wild-type enzyme expression and as a template for introducing the point mutations Q118A, R395K, R395A, R413K, and R413A using the QuikChange Mutagenesis Kit (Stratagene, La Jolla, CA). For production of the R395K/R413K mutant protein, the R413K-expressing plasmid was used as the template for introducing the second mutation, R395K, into the menD gene. Oligodeoxynucleotide primers used in the site-directed mutagenesis are listed in Table 3. The mutated menD gene was sequenced in full length by Beijing Genomics Institute (BGI, Shenzhen, China) to ensure that only the desired mutation was introduced.

For expression of the wild-type enzyme or a mutant protein, BL21 (DE3) cells containing the corresponding expression plasmid were grown in Luria broth (LB) 50 µg/ml kanamycin at 37°C to OD600 = 0.8 and then induced for protein overexpression at 18°C for 20 h by adding 0.1 mM IPTG. The protein was precipitated from the crude extract with 40% saturated ammonium sulfate and then dissolved in 25 mM Tris–HCl buffer (pH 8.0) containing 10% glycerol. The protein solution was desalted with a HiPrep 26/10 Desalting column (GE Healthcare) before loaded onto a HiPrep DEAE FF 16/10 column (GE Healthcare) for fractionation. The undesired proteins were eluted from the DEAE column by 93% buffer A (25 mM Tris–HCl, pH 8.0, and 10% glycerol) and 7% buffer B (25 mM Tris–HCl, pH 8.0, 1 M NaCl, and 10% glycerol). Subsequently, the gradient was set from 7% to 20% buffer B and the protein was eluted from 10% to 18% buffer B. The fractions were collected and concentrated to 1 ml before being loaded onto a HiPrep Sephacryl S-200 column for further purification. The protein was purified to homogeneity with a purity >95% by SDS–PAGE analysis and concentrated to ∼130 mg/ml in 25 mM Tris–HCl, pH 8.0, 10% glycerol for storage at −20°C.

Steady-state and single-turnover kinetics

The activity of MenD or its mutants was determined by monitoring the consumption of isochorismate with its absorbance at 278 nm (ε = 8300 M−1 cm−1) using a previously reported method [1]. In the activity assay, ThDP and α-ketoglutaric acid were purchased from Sigma–Aldrich and isochorismate was prepared from chorismic acid using EntC [34,35] and purified by high-performance reverse-phase liquid chromatography using an Xterra semi-preparative C18 column and a Waters 600E system with model 2487 dual λ absorbance detector, as previously reported [5,15,16]. Similar to the previous study [5], the single-substrate steady-state kinetics of MenD and its mutants were determined by keeping one substrate at a saturated concentration while varying the concentration of the other substrate. The assay reaction was typically initiated by adding the enzyme at 90 nM to a mixture of the substrates, 50 µM ThDP and 5 mM MgCl2 in 50 mM phosphate buffer at pH 7.0. For R395A, the isochorismate concentration was set at 20 µM in the determination of the steady-state kinetic constants for α-ketoglutarate because the saturated concentration of isochorismate was more than 50 µM and its UV absorbance at 278 nm was too high to measure accurately. For other mutant enzymes, the saturating concentration of α-ketoglutarate and isochorismate was 600 and 3.0 µM, respectively. For R395K/R413K, the mutant protein concentration was increased to 500 nM due to its low activity. The kinetic constants were determined from the initial rates according to the Michaelis–Menten equation,

 
formula

where ν is the initial reaction rate (μM/min); νmax is the maximum reaction rate (μM/min) and is equal to kcat multiplied by the total concentration of the enzyme; [S] is the substrate concentration (μM); and KM is the Michaelis–Menten constant (μM).

Single-turnover kinetic experiments were carried out with a Chirascan circular dichroism spectrometer (Applied Photophysics, U.K.) and its stopped-flow accessory, using the procedure developed for the wild-type enzyme [16]. Typically, 5.4 mg/ml protein (wild-type MenD or its mutants) solution containing 400 μM ThDP and 5 mM MgCl2 in 50 mM phosphate buffer (pH 7.0) in one syringe was incubated at room temperature (21°C) for 20 min and then mixed with 2 mM α-ketoglutarate (2KG) in the same buffer in another syringe. The structural change of the ThDP cofactor was monitored at 302 nm and the time curves were averaged from 80 repeated experiments and fitted with two exponential terms corresponding to two first-order chemical reactions (kapp1 and k2). When the second-phase reaction was too slow to measure accurately, the first-order rate constant (k2) was determined by fitting the time curve of the CD signal collected at 334 nm, as reported recently [16].

Crystallization, soaking, and data collection

Similar to crystallization of the wild-type enzyme, MenD mutants were screened for conditions for co-crystallization with ThDP at 289 K using commercially available screening kits. Initial crystallization conditions were found by the sitting-drop diffusion method to be similar to that for the wild-type enzyme. After optimization and additive screening, conditions were found by the hanging-drop diffusion method to successfully grow large single crystals that diffracted to a high resolution of <2.1 Å on an in-house Oxford Xcalibur X-ray diffractometer. The R395K–ThDP complex was crystallized by mixing 1 µl of protein solution containing 10 mg/ml R395K, 1.0 mM ThDP, 5 mM MnSO4, and 10% glycerol in 25 mM Tris–HCl buffer (pH 8.0) with 1 µl of reservoir solution containing 0.2 M ammonium acetate, 0.06 M magnesium formate, 3% (w/v) polyethylene glycol 3350, and 12% (w/v) polyethylene glycol 10 000 in 0.1 M Tris–HCl (pH 7.5). The R395A–ThDP complex was crystallized by mixing 1 µl solution of 10 mg/ml R395A, 1.0 mM ThDP, 5 mM MnSO4, and 10% glycerol in 25 mM Tris–HCl buffer (pH 8.0) and 1 µl of reservoir solution containing 30% (v/v) Jeffamine M-600 (pH 7.0), 11.2% (w/v) polyethylene glycol 3350 and 0.16 M magnesium formate in 0.1 M HEPES (pH 7.5). The R413A–ThDP crystals were obtained by mixing 1 µl of R413A solution in the same buffer and at the same concentration as other mutant proteins with 1 µl of reservoir solution containing 0.16 M magnesium formate, 1% Tacsimate (pH 7.0), 14% polyethylene glycol 3350, and 2% polyethylene glycol MME 5000 in 0.02 M HEPES (pH 7.0).

The single crystals of the mutant enzymes complexed with ThDP were then soaked in the mother liquor supplemented with 10% glycerol and 16 mM α-ketoglutarate (4 µl of reservoir solution mixed with 1 µl of 100 mM α-ketoglutarate and 1.25 µl of 50% glycerol) for a time period ranging from 5 to 35 min. For each mutant protein, the substrate soaking was stopped at multiple time points by directly transferring the soaked single crystals into liquid nitrogen. The soaked crystals were tested again on the in-house Oxford Xcalibur X-ray diffractometer and found to suffer a resolution decrease by at least 0.5 Å. Soaked single crystals with a resolution higher than 3.0 Å were then used for the collection of diffraction data at the beamline BL17U of the Shanghai Synchrotron Radiation Facility (SSRF) with an ADSC Quantum 315R charge-coupled device detector [36] or BL19U1 with a PILATUS3 6M detector at the National Facility for Protein Science Shanghai (NFPS). Diffraction images were either processed with HKL2000 [37] or indexed, integrated with iMosfilm [38], and scaled with Aimless [39]. Statistics for data collection are summarized in Table 4.

Structural determination and refinement

Structures of the soaked crystals were solved by molecular replacement with Phaser [40] using the wild-type E. coli MenD structure (PDB entry: 5EJ8) [15] as the search model. The generated model was further manually built by COOT [41] based on the extra electron density and refined by PHENIX-Refine [42]. Non-crystal symmetry and TLS anisotropic refinement were incorporated in the middle refinement cycles. Subsequently, the co-ordinates and geometry constraints of the ligand molecules were generated using PRODRUG [43] and optimized by eLBOW [44]. To create the two rotamers of the C2-succinyl group of the ThDP-bound intermediate in chain C and chain G in R395A, the ligand file of the native tetrahedral intermediate in the wild-type enzyme was extracted from the PDB file 5EJ8 [15] and manually merged with that of the inactive tetrahedral intermediate in R413A to create restraints of the new ligand by eLBOW [44], setting the occupancy of each original rotamer to 0.5. The new ligand was then incorporated into the R395A active sites and subjected to refinement of the structure and occupancy by PHENIX-Refine [42]. The overall quality of the finalized structures was assessed by PROCHECK [45] and MolProbity [46]. The data statistics for refinement of the structures are summarized in Table 4.

Structural analysis, sequence alignment, and docking simulation

PyMOL version 1.3 [47] was used to perform structural analysis of the protein crystal structures and to generate all graphics. The protein interfaces were analyzed and the quaternary structure was determined using PISA [48], while the electrostatic potential surface was calculated using PDB2PQR plus APBS [49]. Reviewed MenD orthologs from UniProt [50] were clustered with a 50% sequence identity and were aligned and analyzed with Clustal Omega [51]. The alignment was presented with the E. coli MenD crystal structure using ESPript 3.0 [52].

Docking of α-ketoglutarate into the active sites of the wild-type MenD (PDB ID: 5EJ8) and its R413A mutant (PDB ID: 5EJM) was performed with Autodock (4.2.6) and the graphical user interface AutoDockTools (1.5.6.) [53]. A functional dimer formed by chain A and chain B was chosen as the receptor for the substrate and the ThDP adduct TD6 was replaced by ThDP in the modeling. The ligand file was created by the Online SMILES Translator and Structure File Generator and optimized by eLBOW in the PHENIX software suite [46]. The grid map was calculated by AutoGrid with a grid box centered on the C2 atom of the ThDP cofactor in chain A and set to a size of 100 × 100 × 100 with a point spacing of 0.2 Å. Genetic algorithm was used in the docking process with its run number set to 25 and all other parameters were set to the default values.

Results

MenD mutants and steady-state kinetics

Arg395 and Arg413 were mutated individually to either alanine or lysine in E. coli MenD. These mutants were readily overexpressed and purified to homogeneity with a high stability like the wild-type enzyme. They were also found to exhibit a far-ultraviolet CD spectrum closely resembling that of the wild-type enzyme, demonstrating that the point mutations have negligible influence on the protein structure. Single-substrate steady-state kinetic constants of these mutants were determined by varying the concentration of one substrate and keeping the other substrate at a saturating concentration.

As shown in Table 1, mutations at Arg413 generally cause a larger activity decrease than at Arg395, while the alanine mutation affects enzymic activity more than the lysine mutation at the same amino acid residue. Only the R413A mutant protein is completely inactive, while other proteins containing a single mutation are decreased in catalytic efficiency (kcat/KM) mainly due to the increased KM with a moderate decrease in kcat. KM for α-ketoglutarate is more significantly increased by R413K (39-fold) than R395K (22.8-fold), whereas KM for isochorismate is much less increased by the former (∼4.2-fold) than the latter (26.3-fold). These Michaelis constants show that both Arg395 and Arg413 contribute significantly to binding of the two substrates with the former more important in recognizing isochorismate and the latter more important in recognizing α-ketoglutarate. The role of Arg395 as a major contributor to isochorismate binding is further supported by the R395A-caused 197-fold KM increase for isochorismate and 9.9-fold KM increase for α-ketoglutarate. The steady-state kinetic constants of proteins containing the single mutations are consistent with those determined for the same R395A, R395K, and R413K mutants of E. coli MenD in a previous study [33]. When both R395K and R413K mutations were introduced into the enzyme, KM was increased 59-fold, roughly equal to the combined effect of the single mutations. However, kcat of the double-mutant protein was decreased more than 100-fold for both α-ketoglutarate and isochorismate, demonstrating that both Arg395 and Arg413 also play an important role in the catalytic reaction likely through positioning the substrates for reaction and stabilizing the transition states and reaction intermediates. Considering the KM effects and the fact that R413K causes a significantly larger decrease in catalytic efficiency (270-fold) for α-ketoglutarate than R395K (45-fold), Arg413 is likely the main residue regulating the binding, positioning of the ketoacid substrate, and its reaction with the ThDP cofactor to form the tetrahedral intermediate E-Int.

Table 1
Steady-state kinetic parameters of MenD and its mutants
Protein Substrate kcat (min−1kcat (wt)/kcat KM (μM−1KM/KM (wt) kcat/KM (M−1 min−1Fold decrease in kcat/KM 
WT α-Ketoglutarate 17.2 ± 1.5 – 2.9 ± 1.3 – 5.9 × 106 – 
Isochorismate 20.2 ± 1.6 – 0.08 ± 0.03 – 2.4 × 107 – 
R395K α-Ketoglutarate 8.62 ± 0.84 1.99 67 ± 18 23 1.3 × 105 45 
Isochorismate 8.07 ± 0.39 2.49 2.2 ± 0.4 26 3.7 × 105 66 
R395A1 α-Ketoglutarate 1.07 ± 0.06 16.1 29 ± 8 9.9 3.7 × 104 1.6 × 102 
Isochorismate 3.91 ± 0.08 5.16 16 ± 1 197 2.4 × 105 1.0 × 103 
R413K α-Ketoglutarate 2.57 ± 0.31 6.68 115 ± 37 39 2.2 × 104 2.7 × 102 
Isochorismate 2.30 ± 0.20 8.77 0.35 ± 0.14 4.2 6.6 × 106 37 
R395K/R413K α-Ketoglutarate 0.16 ± 0.02 108 172 ± 46 59 9.3 × 102 6.3 × 103 
Isochorismate 0.15 ± 0.02 135 0.82 ± 0.25 10 1.8 × 104 1.3 × 103 
R413A No activity 
Q118A No activity 
Protein Substrate kcat (min−1kcat (wt)/kcat KM (μM−1KM/KM (wt) kcat/KM (M−1 min−1Fold decrease in kcat/KM 
WT α-Ketoglutarate 17.2 ± 1.5 – 2.9 ± 1.3 – 5.9 × 106 – 
Isochorismate 20.2 ± 1.6 – 0.08 ± 0.03 – 2.4 × 107 – 
R395K α-Ketoglutarate 8.62 ± 0.84 1.99 67 ± 18 23 1.3 × 105 45 
Isochorismate 8.07 ± 0.39 2.49 2.2 ± 0.4 26 3.7 × 105 66 
R395A1 α-Ketoglutarate 1.07 ± 0.06 16.1 29 ± 8 9.9 3.7 × 104 1.6 × 102 
Isochorismate 3.91 ± 0.08 5.16 16 ± 1 197 2.4 × 105 1.0 × 103 
R413K α-Ketoglutarate 2.57 ± 0.31 6.68 115 ± 37 39 2.2 × 104 2.7 × 102 
Isochorismate 2.30 ± 0.20 8.77 0.35 ± 0.14 4.2 6.6 × 106 37 
R395K/R413K α-Ketoglutarate 0.16 ± 0.02 108 172 ± 46 59 9.3 × 102 6.3 × 103 
Isochorismate 0.15 ± 0.02 135 0.82 ± 0.25 10 1.8 × 104 1.3 × 103 
R413A No activity 
Q118A No activity 
1

Kinetic parameters for α-ketoglutarate were determined at an unsaturated concentration of isochorismate at 20 μM.

Notably, the steady-state kinetic profile of the R395A mutant is similar to that of the corresponding site-directed mutant at Arg409 (R409A) of B. subtilis MenD [18]. However, the complete activity loss of the R413A mutant of E. coli MenD is significantly different from a 100-fold decrease in catalytic efficiency reported for the alanine mutant of the corresponding arginine residue (Arg428) in B. subtilis MenD [18]. This difference is likely due to the 275-fold higher affinity of the E. coli enzyme for α-ketoglutarate (KM = 0.08 µM, Table 1) than the B. subtilis ortholog (KM = 22 µM) [18]. When Arg413 that plays a crucial role in the interaction of α-ketoglutarate with the active site is replaced by alanine in E. coli MenD, this higher affinity is completely lost to greatly exacerbate the activity decrease compared with the equivalent mutation in the B. subtilis enzyme, leading to complete loss of the catalytic activity.

CD spectra of the intermediates

Similar to the wild-type enzyme [15,16], the site-directed mutants were subjected to test for their ability in forming a ThDP-bound intermediate and analysis of the resulting intermediate by CD spectroscopy in the near-UV and UV regions. As shown in Figure 2, a ThDP-bound intermediate is indeed formed in R395A, R395K, and R413K, which is similar to the native intermediate (E-Int, Figure 1B) in the shape of its CD spectrum. However, the intensity of the CD signals for the three mutants is significantly lower compared with that of the wild-type enzyme. These observations suggest that a similar ThDP-bound post-decarboxylation intermediate is formed in the three mutants as in the wild-type enzyme, but with a lower yield.

CD spectrum of the ThDP-bound intermediate formed in wild-type MenD and its mutants.

Figure 2.
CD spectrum of the ThDP-bound intermediate formed in wild-type MenD and its mutants.

The intermediate was formed from reaction containing 20 mg/ml protein, 500 µM ThDP, 5 mM MgCl2, and 2 mM α-ketoglutarate in 50 mM phosphate buffer at pH 7.0 and its spectrum was taken after 15 min incubation at room temperature.

Figure 2.
CD spectrum of the ThDP-bound intermediate formed in wild-type MenD and its mutants.

The intermediate was formed from reaction containing 20 mg/ml protein, 500 µM ThDP, 5 mM MgCl2, and 2 mM α-ketoglutarate in 50 mM phosphate buffer at pH 7.0 and its spectrum was taken after 15 min incubation at room temperature.

Interestingly, despite the complete loss of catalytic activity of R413A mutant in the steady-state kinetic determination, a ThDP-bound intermediate was also detected for the mutant with an intense CD signal in the same region, of which the positive 302 nm signal is similar in shape with a lower intensity but the broad negative peak from 310 to 365 nm is significantly different from that for all other mutants and the wild-type enzyme. Combined with the steady-state kinetic properties of the mutants (Table 1), these results strongly suggest that all single-point mutants are able to recognize and react with the α-ketoglutarate substrate to form a ThDP-bound intermediate, which is inactive for R413A but active with a structure similar to the native tetrahedral intermediate E-Int for R395A, R395K, and R413K.

Overall structure of the mutants

The mutant proteins were readily crystallized in the presence of ThDP under conditions that were optimal for crystallization of the wild-type enzyme [15]. The obtained single crystals were then soaked in saturated α-ketoglutarate solution and successfully used for structural determination. Like the MenD structure [15], all the mutant protein structures belong to the P1 space group and their unit cells contain two tetramers in a dimer-of-dimer assembly in which the functional unit is a dimer. In each unit cell, the eight polypeptide chains are essentially identical with an RMSD of <0.25 Å and continuous high-quality electron densities throughout the primary sequence. In addition, a ThDP-bound post-decarboxylation intermediate with additional electron densities at C2 is found in the active site of most protein subunits of the two tetramers. The ThDP component of the intermediate is the same and all identical with that of the tetrahedral intermediate in the wild-type MenD with the same active site interactions. However, differences are found in the C2-succinyl group of the ThDP-bound intermediate and the active site amino acid residues interacting with this group.

Two C2-succinyl orientations in mutants

The ThDP-bound post-decarboxylation intermediate is highly similar in each subunit in the crystal structures of both R395K and R413K, but its C2-succinyl group is lower in electron density in R413K with an occupancy of 0.50–0.80 for the ethylene group and the terminal carboxylate. Nonetheless, the C2-succinyl group is essentially superimposable on that of the native tetrahedral intermediate in the wild-type enzyme (Figure 3A,D) with some subtle changes. Take R395K for example, the unchanged Arg413 still makes hydrogen-bonding contact with the C2-succinyl terminal carboxylate with slightly altered bond lengths, but Lys395 makes no contact with the terminal carboxylate with its ε-NH2 located 7.2 Å away (Figure 3B). However, the hydrogen bond between this terminal carboxylate and the backbone amide of Ser391 is apparently strengthened with its bond length shortened from 2.8 Å in the wild-type enzyme to 2.5 Å in R395K. These subtle changes in the stabilizing interactions cause a slight positional shift of the C2-succinyl terminal carboxylate towards the Ser391 backbone, leading to change in the ethylene moiety from anti-conformation in the wild-type enzyme to gauche-conformation in R395K (Figure 3E). Despite these slight differences, the overall structural similarity of the R395K intermediate to the native intermediate is consistent with the mild mutational effect on the enzymatic activities (Table 1).

Two C2-succinyl orientations in the tetrahedral ThDP-bound intermediates in the MenD mutants.

Figure 3.
Two C2-succinyl orientations in the tetrahedral ThDP-bound intermediates in the MenD mutants.

(A) Stereo diagram of the intermediate in the wild-type MenD. (B) Stereo diagram of the intermediate in R395K. (C) Stereo diagram of the intermediate in R413A. (D) Overlap of the intermediates in different proteins. (E) Two views of the ethylene group in the succinyl groups in different proteins. Relevant active site residues are presented in sticks with C, O, N, P colored green, red, blue and brown, respectively, while the intermediates are also presented in sticks with a different color for the carbon atoms in each protein. FMT in (A) is a formate ion (in green sticks) from the crystallization buffer. Blue mesh lines represent the mFo − DFc electron density map contoured at 2.0σ for the C2-appendage of the intermediates only and golden dashed lines denote hydrogen bonds with a distance of ≤3.5 Å. The mutated residues are labeled in purple and water molecules are showed in red dots, while Mn2+ ions are presented in purple spheres and the residues labeled with a primed number are from a different subunit in the functional dimer.

Figure 3.
Two C2-succinyl orientations in the tetrahedral ThDP-bound intermediates in the MenD mutants.

(A) Stereo diagram of the intermediate in the wild-type MenD. (B) Stereo diagram of the intermediate in R395K. (C) Stereo diagram of the intermediate in R413A. (D) Overlap of the intermediates in different proteins. (E) Two views of the ethylene group in the succinyl groups in different proteins. Relevant active site residues are presented in sticks with C, O, N, P colored green, red, blue and brown, respectively, while the intermediates are also presented in sticks with a different color for the carbon atoms in each protein. FMT in (A) is a formate ion (in green sticks) from the crystallization buffer. Blue mesh lines represent the mFo − DFc electron density map contoured at 2.0σ for the C2-appendage of the intermediates only and golden dashed lines denote hydrogen bonds with a distance of ≤3.5 Å. The mutated residues are labeled in purple and water molecules are showed in red dots, while Mn2+ ions are presented in purple spheres and the residues labeled with a primed number are from a different subunit in the functional dimer.

In contrast, the ThDP-bound intermediate in R413A is significantly different from the native tetrahedral intermediate E-Int in the wild-type enzyme. Its C center takes an opposite stereochemistry (R-configuration) compared with that in the wild-type enzyme (S-configuration), leading to a 180° change in the orientation of the C2-succinyl group (Figure 3C,D). Consequently, the C2-succinyl group of the R413A intermediate is the mirror image of the same group in the wild-type enzyme about a plane cutting through the C2–C bond, except that the ethylene moiety of the succinyl group takes an eclipsed-conformation with the two non-hydrogen groups forming an angle of ∼120° in the former and an anti-conformation with the two non-hydrogen groups forming an angle of 180° in the latter (Figure 3E). In R413A, the terminal carboxylate of the C2-succinyl group is stabilized by multiple hydrogen bonds with the side chain of Gln118′, the backbone amide of Ser32′, and a water molecule bound by the Arg33′ and Arg107′ side chains (Figure 3C). Despite these differences, the C-N4′ of the intermediate in R413A is 3.2 Å and similar to that found in the native tetrahedral intermediate. Since the short C-N4′ distance is unable to accommodate two hydrogen atoms [15], the R413A intermediate should also take both the anionic aminopyrimidinium (APH+) form and the neutral iminopyrimidine (IP) form in tautomerization equilibrium and is thus nucleophilic like the native intermediate (Figure 1B). However, this R413A intermediate loses all its activity (Table 1) probably due to the non-native stereochemistry at the C center. Nevertheless, the structural isomerism of the ThDP-bound tetrahedral intermediates provides a rationale for their broad negative CD 310–365 nm signal with a distinctive shape (Figure 2). In addition, the retained nucleophilicity of the intermediate in R413A offers an opportunity to access carboligation products with different stereochemistry in chemoenzymatic application of the ThDP-dependent enzyme [54].

Very surprisingly, four different ligands of ThDP or its derivatives are found in the unit cell of the 2KG-soaked R395A crystal structure. Among the eight chains in the unit cell, chain F is the only subunit that contains a free ThDP cofactor without any electron density extended beyond the C2 atom, whereas all other subunits contained the post-decarboxylation product with a C2-succinyl group in one of the two conformations or a mix of both conformations (Figure 4). In chains A and B, the ThDP-bound intermediate is conformationally similar to the inactive, off-pathway intermediate in the R413A mutant (Figure 4A), while the intermediate is closely similar in conformation to the active tetrahedral intermediate in the wild-type enzyme or the R395K mutant in chains D, E, and H (Figure 4C). In chains C and G, the intermediate takes both active and inactive conformations with the C2-succinyl group taking both orientations with an occupancy of 0.53 and 0.47, respectively (Figure 4B). In all the chains, the quality of the electron density is significantly poorer for the succinyl group than that in the native tetrahedral intermediate (Figure 3A).

Three different ThDP-bound intermediates in the R395A active site.

Figure 4.
Three different ThDP-bound intermediates in the R395A active site.

(A) Stereo diagram of the inactive intermediate in chain B. (B) Stereo diagram of the intermediate with the C2-succinyl group in both the active and inactive orientations in chain C. (C) Stereo diagram of the active intermediate in chain H. For simplicity, only the active site residues interacting with the succinyl group are presented in green sticks and the ThDP-bound intermediates are presented in sticks with gray carbon atoms. Labels with a primed number denote residues from the different chain in the functional dimer and the mutated residue is labeled in purple. The mFo − DFc omit map of the C2-succinyl group is simultaneously contoured at 2.0σ in blue meshes and at 1.0σ in pink meshes. The active site water molecules are indicated by red dots and hydrogen bonds with a distance ≤3.5 Å are denoted by golden dashed lines.

Figure 4.
Three different ThDP-bound intermediates in the R395A active site.

(A) Stereo diagram of the inactive intermediate in chain B. (B) Stereo diagram of the intermediate with the C2-succinyl group in both the active and inactive orientations in chain C. (C) Stereo diagram of the active intermediate in chain H. For simplicity, only the active site residues interacting with the succinyl group are presented in green sticks and the ThDP-bound intermediates are presented in sticks with gray carbon atoms. Labels with a primed number denote residues from the different chain in the functional dimer and the mutated residue is labeled in purple. The mFo − DFc omit map of the C2-succinyl group is simultaneously contoured at 2.0σ in blue meshes and at 1.0σ in pink meshes. The active site water molecules are indicated by red dots and hydrogen bonds with a distance ≤3.5 Å are denoted by golden dashed lines.

Single-turnover kinetics of the mutants

As reported recently [16], the tetrahedral intermediate E-Int* is formed from α-ketoglutarate in a pseudo first-order process (kapp1 = k · [2KG]) and is then converted to E-Int in another first-order process (k2) (Figure 1). In the same single-turnover experiments monitored with CD signal at 302 nm characteristic of the IP form of the cofactor [55,56], the mutants gave rise to kinetic traces that were fitted well with two similar exponential phases. For R413A, the spectral change in the second phase was negligible at 302 nm and the CD change at 334 nm was used to increase the accuracy in the determination of the rate constant. As shown in Table 2, kapp1 is slightly reduced by 2–3-fold and k2 is more significantly decreased by 1.7–12-fold for R395K, R395A, and R413K. R395A is most similar to the wild-type enzyme among the three mutant proteins and shows the same linear kapp1 dependence (Figure 5), but k2 independence on α-ketoglutarate (data not shown), indicating that the rate-limiting step in forming the intermediate E-Int* is unchanged and is still the binding of the ketoacid substrate [16]. For R395K and R413K, kapp1 loses its linear dependence on α-ketoglutarate in comparison with the wild-type enzyme, suggesting a shift of the rate-limiting step to another step, for example, the nucleophilic addition of the activated cofactor to the bound α-ketoglutarate or the decarboxylation step. For these mutants, kapp1 decreases in response to weakening of the α-ketoglutarate binding, strongly suggesting that the ketoacid substrate interacts with the same active site containing the conserved arginine residues as in the wild-type enzyme.

Dependence of the pseudo first-order rate constant kapp1 on 2KG in MenD and its mutants.

Figure 5.
Dependence of the pseudo first-order rate constant kapp1 on 2KG in MenD and its mutants.

Each dataset is fitted with a linear trend line in the plot. Reaction conditions: the stopped-flow experiments were carried out in 50 mM phosphate buffer, pH 7.0, containing 200 µM ThDP and 2.5 mM MgCl2; [MenD] = 44 µM when [2KG] was varied from 50 to 1000 µM. Data for MenD were from a recent investigation [16] and included in the plot for comparison.

Figure 5.
Dependence of the pseudo first-order rate constant kapp1 on 2KG in MenD and its mutants.

Each dataset is fitted with a linear trend line in the plot. Reaction conditions: the stopped-flow experiments were carried out in 50 mM phosphate buffer, pH 7.0, containing 200 µM ThDP and 2.5 mM MgCl2; [MenD] = 44 µM when [2KG] was varied from 50 to 1000 µM. Data for MenD were from a recent investigation [16] and included in the plot for comparison.

Table 2
Single-turnover kinetic constants of MenD and its mutants

The reaction was monitored with CD change at 302 nm after rapid equal-volume mixing of a solution containing 5.4 mg/ml MenD or its mutants, 400 µM ThDP, and 5 mM MgCl2 in 50 mM phosphate buffer (pH 7.0) with an α-ketoglutarate (2.0 mM) solution in the same buffer at room temperature.

Protein kapp1 (s−1k2 (s−1
MenD 90 ± 11 5.2 ± 0.6 
R395K 46 ± 7 3.0 ± 0.1 
R395A 29 ± 3 0.43 ± 0.02 
R413K 67 ± 10 0.47 ± 0.04 
R413A 98 ± 3 2(3.6 ± 0.1) × 10−3 
Protein kapp1 (s−1k2 (s−1
MenD 90 ± 11 5.2 ± 0.6 
R395K 46 ± 7 3.0 ± 0.1 
R395A 29 ± 3 0.43 ± 0.02 
R413K 67 ± 10 0.47 ± 0.04 
R413A 98 ± 3 2(3.6 ± 0.1) × 10−3 
1

k2 was determined from the change of the CD signal at 334 nm.

Table 3
Synthetic oligodeoxynucleotide primers for MenD cloning and site-directed mutagenesis
Protein  Primer sequence (the mutated codons are underlined) 
MenD Forward CCCGCCATGGCAATGTCAGTAAGCGCATTTAA 
Reverse CCGCCTCGAGTCATAAATGGCTTACACTCG 
Q118A Forward GACTGCGGCGCGAAT GCG GCAATTCGCCAGCCG 
Reverse CGGCTGGCGAATTGC CGC ATTCGCGCCGCAGTC 
R395K Forward AACAGCCTGGTGGTA AAA CTGATTGATGCGCTT 
Reverse AAGCGCATCAATCAG TTT TACCACCAGGCTGTT 
R395A Forward AACAGCCTGGTGGTA GCT CTGATTGATGCGCTT 
Reverse AAGCGCATCAATCAG AGC TACCACCAGGCTGTT 
R413K Forward CCGGTGTACAGCAAC AAA GGGGCCAGCGGTATC 
Reverse GATACCGCTGGCCCC TTT GTTGCTGTACACCGG 
R413A Forward CCGGTGTACAGCAAC GCT GGGGCCAGCGGTATC 
Reverse GATACCGCTGGCCCC AGC GTTGCTGTACACCGG 
Protein  Primer sequence (the mutated codons are underlined) 
MenD Forward CCCGCCATGGCAATGTCAGTAAGCGCATTTAA 
Reverse CCGCCTCGAGTCATAAATGGCTTACACTCG 
Q118A Forward GACTGCGGCGCGAAT GCG GCAATTCGCCAGCCG 
Reverse CGGCTGGCGAATTGC CGC ATTCGCGCCGCAGTC 
R395K Forward AACAGCCTGGTGGTA AAA CTGATTGATGCGCTT 
Reverse AAGCGCATCAATCAG TTT TACCACCAGGCTGTT 
R395A Forward AACAGCCTGGTGGTA GCT CTGATTGATGCGCTT 
Reverse AAGCGCATCAATCAG AGC TACCACCAGGCTGTT 
R413K Forward CCGGTGTACAGCAAC AAA GGGGCCAGCGGTATC 
Reverse GATACCGCTGGCCCC TTT GTTGCTGTACACCGG 
R413A Forward CCGGTGTACAGCAAC GCT GGGGCCAGCGGTATC 
Reverse GATACCGCTGGCCCC AGC GTTGCTGTACACCGG 
Table 4
Data collection and refinement statistics
 R395K R395A R413A R413K 
PDB code 5Z2R 5Z2U 5EJM 5Z2P 
Data collection 
Wavelength (Å) 0.979 0.979 0.979 0.979 
Space group PPPP
Unit cell 
 a, b, c (Å) 90.63, 90.75, 172.37 90.67, 90.70, 169.07 90.66, 90.67, 169.37 90.26, 90.27, 167.07 
 α, β, γ (°) 82.91, 75.71, 64.20 83.26, 75.96, 64.17 83.29, 76.07, 64.32 83.43, 76.14, 63.49 
Reflections (unique) 353 375 (187 417) 346 161 (177 286) 939 018 (482 851) 355 213 (159 069) 
Multiplicity3 1.9 (1.8) 2.0 (2.0) 1.9 (1.9) 2.2 (2.0) 
Completeness (%)3 88.0 (77.1) 90.5 (92.4) 95.9 (93.6) 78.4 (77) 
I/σI3 9.2 (5.5) 9.1 (6.4) 6.3 (2.1) 7.0 (4.9) 
Rmerge3 0.068 (0.108) 0.096 (0.083) 0.064 (0.290) 0.081 (0.126) 
CC1/23 0.983 (0.963) 0.981 (0.965) – 0.984 (0.928) 
Structural refinement 
Resolution range (Å) 37.34–2.30 (2.38–2.30) 37.27–2.35 (2.43–2.35) 31.48–1.72 (1.7 5–1.72) 20.64–2.30 (2.38–2.30) 
Rfree/Rwork (%) 19.99/16.98 20.24/16.39 19.14/16.47 20.39/16.71 
No. of non-hydrogen atoms 37 691 37 825 40 065 36 992 
Macromolecules 34 137 34 202 35 063 34 123 
Solvent 3190 3282 4650 2516 
Ligands/ions 364 341 352 353 
Average B-factor (Å222.56 16.59 22.20 16.71 
Macromolecules 21.94 16.00 20.83 16.32 
Ligands 26.28 15.25 22.34 18.37 
Solvent 28.83 22.94 32.54 21.77 
RMSD for ideal value in 
 Bond length (Å) 0.009 0.008 0.007 0.012 
 Bond angles (o1.30 1.00 1.07 1.47 
 Ramachandran statistics (%)4 97.6/2.4/0.0 97/2.5/0 98/1.8/0.18 98/2.2/0 
 R395K R395A R413A R413K 
PDB code 5Z2R 5Z2U 5EJM 5Z2P 
Data collection 
Wavelength (Å) 0.979 0.979 0.979 0.979 
Space group PPPP
Unit cell 
 a, b, c (Å) 90.63, 90.75, 172.37 90.67, 90.70, 169.07 90.66, 90.67, 169.37 90.26, 90.27, 167.07 
 α, β, γ (°) 82.91, 75.71, 64.20 83.26, 75.96, 64.17 83.29, 76.07, 64.32 83.43, 76.14, 63.49 
Reflections (unique) 353 375 (187 417) 346 161 (177 286) 939 018 (482 851) 355 213 (159 069) 
Multiplicity3 1.9 (1.8) 2.0 (2.0) 1.9 (1.9) 2.2 (2.0) 
Completeness (%)3 88.0 (77.1) 90.5 (92.4) 95.9 (93.6) 78.4 (77) 
I/σI3 9.2 (5.5) 9.1 (6.4) 6.3 (2.1) 7.0 (4.9) 
Rmerge3 0.068 (0.108) 0.096 (0.083) 0.064 (0.290) 0.081 (0.126) 
CC1/23 0.983 (0.963) 0.981 (0.965) – 0.984 (0.928) 
Structural refinement 
Resolution range (Å) 37.34–2.30 (2.38–2.30) 37.27–2.35 (2.43–2.35) 31.48–1.72 (1.7 5–1.72) 20.64–2.30 (2.38–2.30) 
Rfree/Rwork (%) 19.99/16.98 20.24/16.39 19.14/16.47 20.39/16.71 
No. of non-hydrogen atoms 37 691 37 825 40 065 36 992 
Macromolecules 34 137 34 202 35 063 34 123 
Solvent 3190 3282 4650 2516 
Ligands/ions 364 341 352 353 
Average B-factor (Å222.56 16.59 22.20 16.71 
Macromolecules 21.94 16.00 20.83 16.32 
Ligands 26.28 15.25 22.34 18.37 
Solvent 28.83 22.94 32.54 21.77 
RMSD for ideal value in 
 Bond length (Å) 0.009 0.008 0.007 0.012 
 Bond angles (o1.30 1.00 1.07 1.47 
 Ramachandran statistics (%)4 97.6/2.4/0.0 97/2.5/0 98/1.8/0.18 98/2.2/0 
1

Statistics for the highest-resolution shell are shown in parentheses.

2

Ramachandran statistics indicate the fraction of residues in the most favored, allowed, and disallowed regions of the Ramachandran diagram.

In comparison, formation of the inactive intermediate in R413A is essentially a single first-order process in the single-turnover experiment because k2 is decreased more than 1000-fold relative to the wild-type enzyme. More importantly, its kapp1 is increased rather than decreased and is independent of α-ketoglutarate. This very different single-turnover kinetic profile strongly suggests that R413A binds the ketoacid substrate likely without involving the conserved arginine residues, raising the possibility that R413A interacts with the ketoacid substrate via a different mechanism.

Discussion

MenD is a unique enzyme that catalyzes the ThDP-dependent SEPHCHC synthesis via tetrahedral intermediates [15,16], rather than an enamine intermediate found in the catalysis of many other ThDP-dependent enzymes [2023,57]. As noted earlier, the tetrahedral intermediates are distinct in that their C2-succinyl group forms strong hydrogen-bonding interactions at the terminal carboxylate with two conserved arginine residues, Arg395 and Arg413 in E. coli MenD [15], whereas the C2-appendages in the enamine intermediates interact weakly with the active site residues in other ThDP-dependent enzymes [23,3032]. This difference strongly suggests that the two conserved arginine residues are critical to the new ThDP-dependent catalysis without knowing their exact roles. In this study, we used both structural and biochemical methods to shed lights on their catalytic roles and to provide new insights into the new ThDP-dependent catalytic mechanism.

From the crystal structures of the wild-type enzyme and the mutant proteins, it is obvious that there are two stable orientations/conformations for the C2-succinyl group of the post-decarboxylation intermediates, of which one is active with the terminal carboxylate interacting with Arg395 and Arg413 in the wild-type enzyme and the other is inactive as found in the R413A mutant with the terminal carboxylate interacting with Ser33′, Arg33′, Thr78′, Arg107′, and Gln118′ (Figure 3C). Interestingly, both sets of the stabilizing amino acid residues are highly conserved among MenD orthologs (Figure 6A), forming a native site and a non-native site for binding the succinyl group in two opposite orientations (Figure 6B). The presence of both conformations in the crystal structure of the active mutant R395A strongly supports that they interconvert and equilibrate with each other in the active site, not just in R395A but also in other mutants and the wild-type enzyme. Notably, the conserved residues at the non-native-binding site are also important for the SEPHCHC synthase activity. The enzyme was inactive when Gln118 was mutated to alanine (Table 1), whereas its catalytic efficiency (kcat/KM) was decreased 3–100-fold when the arginine residues corresponding to Arg33 and Arg107 were individually mutated to alanine in B. subtilis MenD [18].

The MenD active site.

Figure 6.
The MenD active site.

(A) Conservation of the residues at the two C2-succinyl-binding pockets. The alignment contains 71 MenD orthologs filtered at 50% identity from the UniProt, of which representative sequences from seven largest clusters containing >50 orthologs are shown only with the fragments contributing to binding of the succinyl group. The residues with 100% and 80–100% identity in the alignment are shown in red boxes and yellow boxes with a blue line, respectively, while the residues in direct contact with the C2-succinyl group are denoted with orange triangles. The secondary structures in the alignment are from the wild-type MenD structure (PDB ID: 5EJ8). (B) Stereo diagram of the MenD active site represented in translucent electrostatic potential surface. The wild-type MenD active site (PDB ID: 5EJ8) is shown with the native ThDP-bound tetrahedral intermediate (gold sticks), of which the terminal succinyl carboxylate is circled in white dashed line to denote the native succinyl-binding pocket. A formate (in blue sticks) in the structure occupies the same position as the terminal carboxylate of the ThDP-bound tetrahedral intermediate with an enantiomeric C in R413A and is also circled in white dashed line to denote the alternate succinyl-binding pocket. Electrostatic potentials less than −20 kT, neutral, and greater than 20 kT are displayed in red, white, and blue, respectively. Residues indicated by orange triangles in (A) are represented in green sticks, of which those labeled with primed numbers are from a different subunit. (C) Stereo diagram of α-ketoglutarate docked to the active site of R413A. The substrate is represented in salmon sticks and the protein structure (PDB ID: 5EJM) is represented as translucent electrostatic potential surface as in (B).

Figure 6.
The MenD active site.

(A) Conservation of the residues at the two C2-succinyl-binding pockets. The alignment contains 71 MenD orthologs filtered at 50% identity from the UniProt, of which representative sequences from seven largest clusters containing >50 orthologs are shown only with the fragments contributing to binding of the succinyl group. The residues with 100% and 80–100% identity in the alignment are shown in red boxes and yellow boxes with a blue line, respectively, while the residues in direct contact with the C2-succinyl group are denoted with orange triangles. The secondary structures in the alignment are from the wild-type MenD structure (PDB ID: 5EJ8). (B) Stereo diagram of the MenD active site represented in translucent electrostatic potential surface. The wild-type MenD active site (PDB ID: 5EJ8) is shown with the native ThDP-bound tetrahedral intermediate (gold sticks), of which the terminal succinyl carboxylate is circled in white dashed line to denote the native succinyl-binding pocket. A formate (in blue sticks) in the structure occupies the same position as the terminal carboxylate of the ThDP-bound tetrahedral intermediate with an enantiomeric C in R413A and is also circled in white dashed line to denote the alternate succinyl-binding pocket. Electrostatic potentials less than −20 kT, neutral, and greater than 20 kT are displayed in red, white, and blue, respectively. Residues indicated by orange triangles in (A) are represented in green sticks, of which those labeled with primed numbers are from a different subunit. (C) Stereo diagram of α-ketoglutarate docked to the active site of R413A. The substrate is represented in salmon sticks and the protein structure (PDB ID: 5EJM) is represented as translucent electrostatic potential surface as in (B).

The active–inactive intermediate equilibrium is apparently crucial to the enzymic catalysis and is controlled by the strength of the stabilizing interactions at the native relative to the non-native succinyl-binding site. In the wild-type enzyme, it is dominated by the active intermediate because of the strong hydrogen-bonding interaction with the conserved arginine residues, which should be significantly stronger than the interactions at the alternate site. In R395K and R413K, this equilibrium is minimally affected as evidenced by the little-changed crystal structures (Figure 3B) and the mild activity decreases (Table 1) because the lysine residue is able to form similar hydrogen bonds like the arginine residues. In contrast, the C2-succinyl interaction with the native-binding site is significantly attenuated in R395A to a level comparable to that at the non-native-binding site and to allow both intermediates of the equilibrium to be observed, whereas the native-binding site in R413A is weakened to a level at which it is unable to compete with the non-native-binding site for the succinyl group, resulting in dominance of the equilibrium by the inactive intermediate (Figure 3C). Thus, the active–inactive intermediate equilibrium is consistent with all the structural data and is an intrinsic feature of the MenD catalysis.

Although both Arg395 and Arg413 significantly contribute to shifting the intermediate equilibrium to the active form, Arg413 is more important for the active intermediate. This is consistent with the observation of the inactive intermediate in R413A (Figure 3C) and both active and inactive intermediates in R395A (Figure 4). It is also supported by the observed poorer quality of electron density and lower occupancy of the C2-sucinyl group of the intermediate in R413K compared with R395K. In addition, the differential role of the two arginine residues is also consistent with the CD spectra of the intermediates formed in the mutants in comparison with the wild-type enzyme (Figure 2). Interestingly, the post-decarboxylation intermediate is not converted to the enamine intermediate found in the catalysis of many other ThDP-dependent enzymes, even when its interactions with the conserved arginine residues are significantly weakened at the native succinyl-binding site. This absence of the enamine intermediate in the mutants suggests that enamine is not even a minute form of the intermediates in the MenD catalytic process. This specific avoidance of the enamine intermediate is likely due to its vulnerability to oxidation [58], which may lead to damage of the intermediates in the absence of the isochorismate substrate [16].

Besides interacting with the intermediate E-Int, the conserved arginine residues are also involved in binding and orienting both substrates, as well as modulating the activities and structures of intermediates in other chemical steps (Figure 1C). From the Michaelis constants in Table 1, it is obvious that Arg395 affects the binding of isochorismate more than Arg413, while the latter affects the binding of 2-ketoglutarate more than the former. In addition, the slight decrease in kcat suggests that these residues also affect the reactivity of the intermediates, probably through their effects on the intermediate structure. The role of these residues in interacting with α-ketoglutarate is corroborated and better demonstrated by the single-turnover kinetic results (Table 2). For R395K, R395A, and R413K, the mode of their interaction with α-ketoglutarate is similar to that for the wild-type enzyme, but each discrete step is affected by the mutations. Their significantly reduced k2 suggests that the conserved residues are also involved in modulating the conformational change from E-Int* to E-Int by controlling the orientation of their C-succinyl groups. These conserved residues likely exert a similar effect on the reactivity of the pre-decarboxylation intermediate, although this effect may be obscured by the rate-limiting step in the stopped-flow experiments.

R413A is different from the wild-type enzyme and other mutant proteins in the single-turnover kinetics. Its kapp1 was marginally increased and its k2 was decreased 1000-fold (Table 2), indicating a major shift in the mode of its interaction with α-ketoglutarate. Considering the presence of the non-native-binding site that stabilizes the C2-succinyl group of the inactive ThDP-bound intermediate in a totally different conformation (Figure 3C), this distinct kinetic behavior of R413A is proposed to stem from a non-native mode of interaction of α-ketoglutarate with the protein at a non-native site. This non-native site should include the residues interacting with the ThDP cofactor moiety and the conserved non-native C2-succinyl-binding site (Figures 3C and 6A). At this non-native site, the terminal carboxylate of the α-ketoacid substrate is oriented in an alternate position relative to the ThDP cofactor for reaction with the activated cofactor, forming the pre-decarboxylation intermediate with the succinyl group also taking a non-native orientation as found in the post-decarboxylation intermediates. This non-native pre-decarboxylation intermediate could also undergo decarboxylation like the corresponding intermediate in the native active site to directly form the non-native post-decarboxylation intermediate as observed in R413A, without a conformational change similar to the E-Int* to E-Int change for the wild-type enzyme. Obviously, activity of the enzyme at this non-native site leads to a catalytic dead end because the resulting post-decarboxylation intermediate is inactive as observed for the R413A mutant enzyme.

The non-native substrate-interacting site is always present in the enzyme active site and competes with the native site for the α-ketoglutarate substrate with a higher rate to form the non-native post-decarboxylation intermediate (the increased kapp1 in Table 2). However, reactions leading to the post-decarboxylation intermediates are not dominated by the non-native site in the wild-type enzyme. One explanation for this contradiction is that the two sites are not truly parallel but overlapped, so that the α-ketoglutarate substrate has to pass the native site before binding to the non-native site. In the wild-type enzyme, the intact arginine residues form strong interaction with the ketoacid substrate to prevent it from binding to the non-native site and thus set the normal reaction course. In support of this explanation, the non-native-binding site for the C2-succinyl group indeed appears to connect to the bulk solution only through the native site where the conserved arginine residues (Arg395 and Arg413) fix the succinyl group in a different orientation in the wild-type enzyme (Figure 6B). Owing to this overlap, the non-native site would not inhibit the enzymatic catalysis under normal conditions. Inhibition occurs only when the binding interaction at the native site is significantly weakened relative to the non-native site, as observed for R413A. In consistent with this view, computational modeling found that α-ketoglutarate was bound to the non-native site in R413A (Figure 6C), in which the binding interaction is significantly weakened at the native-binding site. Finding of this non-native site provides an opportunity for design of substrate analogs or other inhibitors to develop new antibacterial agents against the essential enzyme in vitamin K biosynthesis.

In summary, using a combination of site-directed mutagenesis, protein crystallography, and steady-state and single-turnover kinetics, we have gained novel insights into the roles of the conserved Arg395 and Arg413 in the new ThDP-dependent catalysis of E. coli MenD. Arg413 is found to exert a stronger effect than Arg395, but both contribute significantly to integrity of the intermediates by forming strong ionic hydrogen bonding to the terminal carboxylate group of the C2-succinyl group. Mutation of either residue to alanine still allows the formation of a post-decarboxylation intermediate, but loosens the C2-succinyl group to partially or completely take a non-native conformation to result in the reduction or elimination of the catalytic activity. In addition, these conserved residues also play a crucial role in binding and orienting both substrates in the catalysis, of which Arg413 is more important for the binding and reactivity of α-ketoglutarate, while Arg395 plays a more pronounced role in controlling the reactivity of isochorismate. More importantly, a non-native intermediate-binding site is found in the enzyme that allows faster formation of an inactive post-decarboxylation intermediate with a different C stereochemistry and a non-native conformation for the C2-succinyl group. This non-native site leads to a catalytic dead end and provides a new target for the development of new antibiotics against the essential menaquinone biosynthetic enzyme.

Abbreviations

     
  • 2KG

    α-ketoglutarate

  •  
  • CD

    circular dichroism

  •  
  • IP

    iminopyrimidine

  •  
  • RMSD

    root-mean-square deviation

  •  
  • SEPHCHC

    (1R, 2S, 5S, 6S)-2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexene-1-carboxylate

  •  
  • ThDP

    thiamine diphosphate

Author Contribution

Z.G. conceived and co-ordinated the study. Z.G. and M.Q. wrote the paper. M.Q. and H.S. designed, performed, and analyzed experiments in all figures. M.Q., H.S., X.D. and Y.C. purified the proteins and carried out protein crystallization, data collection, and structural determination. All authors reviewed the results and approved the final version of the manuscript.

Funding

This work was supported by GRF601413 and N_HKUST621/13 from the Research Grants Council and SBI14SC05 from the University Grants Council of the HKSAR Government.

Acknowledgments

We thank the staffs from beamline BL17U1 of SSRF and beamline BL19U1 of National Facility for Protein Science Shanghai (NFPS) for on-site technical assistance during data collection. Co-ordinates and structure factors have been deposited in the Protein Data Bank with accession numbers 5EJM, 5Z2P, 5Z2R, and 5Z2U.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Nowicka
,
B.
and
Kruk
,
J.
(
2010
)
Occurrence, biosynthesis and function of isoprenoid quinones
.
Biochim. Biophys. Acta, Bioenerg.
1797
,
1587
1605
2
Dairi
,
T.
,
Kuzuyama
,
T.
,
Nishiyama
,
M.
and
Fujii
,
I.
(
2011
)
Convergent strategies in biosynthesis
.
Nat. Prod. Rep.
28
,
1054
1086
3
Jiang
,
M.
,
Chen
,
X.
,
Guo
,
Z.-F.
,
Cao
,
Y.
,
Chen
,
M.
and
Guo
,
Z.
(
2008
)
Identification and characterization of (1R, 6R)-2-succinyl-6-hydroxy-2, 4-cyclohexadiene-1-carboxylate synthase in the menaquinone biosynthesis of Escherichia coli
.
Biochemistry
47
,
3426
3434
4
Chen
,
M.
,
Ma
,
X.
,
Chen
,
X.
,
Jiang
,
M.
,
Song
,
H.
and
Guo
,
Z.
(
2013
)
Identification of a hotdog fold thioesterase involved in the biosynthesis of menaquinone in Escherichia coli
.
J. Bacteriol.
195
,
2768
2775
5
Jiang
,
M.
,
Cao
,
Y.
,
Guo
,
Z.F.
,
Chen
,
M.
,
Chen
,
X.
and
Guo
,
Z.
(
2007
)
Menaquinone biosynthesis in Escherichia coli: identification of 2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexene-1-carboxylate (SEPHCHC) as a novel intermediate and re-evaluation of MenD activity
.
Biochemistry
46
,
10979
10989
6
Jiang
,
M.
,
Chen
,
M.
,
Cao
,
Y.
,
Yang
,
Y.
,
Sze
,
K.H.
,
Chen
,
X.
et al. 
(
2007
)
Determination of the stereochemistry of 2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexene-1-carboxylic acid, a key intermediate in menaquinone biosynthesis
.
Org. Lett.
9
,
4765
4767
7
Forsyth
,
R.A.
,
Haselbeck
,
R.J.
,
Ohlsen
,
K.L.
,
Yamamoto
,
R.T.
,
Xu
,
H.
,
Trawick
,
J.D.
et al. 
(
2002
)
A genome-wide strategy for the identification of essential genes in Staphylococcus aureus
.
Mol. Microbiol.
43
,
1387
1400
8
Kobayashi
,
K.
,
Ehrlich
,
S.D.
,
Albertini
,
A.
,
Amati
,
G.
,
Andersen
,
K.K.
,
Arnaud
,
M.
et al. 
(
2003
)
Essential Bacillus subtilis genes
.
Proc. Natl Acad. Sci. U.S.A.
100
,
4678
4683
9
Akerley
,
B.J.
,
Rubin
,
E.J.
,
Novick
,
V.L.
,
Amaya
,
K.
,
Judson
,
N.
and
Mekalanos
,
J.J.
(
2002
)
A genome-scale analysis for identification of genes required for growth or survival of Haemophilus influenzae
.
Proc. Natl Acad. Sci. U.S.A.
99
,
966
971
10
Kurosu
,
M.
,
Narayanasamy
,
P.
,
Biswas
,
K.
,
Dhiman
,
R.
and
Crick
,
D.C.
(
2007
)
Discovery of 1, 4-dihydroxy-2-naphthoate prenyltransferase inhibitors: new drug leads for multidrug-resistant gram-positive pathogens
.
J. Med. Chem.
50
,
3973
3975
11
Dhiman
,
R.K.
,
Mahapatra
,
S.
,
Slayden
,
R.A.
,
Boyne
,
M.E.
,
Lenaerts
,
A.
,
Hinshaw
,
J.C.
et al. 
(
2009
)
Menaquinone synthesis is critical for maintaining mycobacterial viability during exponential growth and recovery from non-replicating persistence
.
Mol. Microbiol.
72
,
85
97
12
Tian
,
Y.
,
Suk
,
D.-H.
,
Cai
,
F.
,
Crich
,
D.
and
Mesecar
,
A.D.
(
2008
)
Bacillus anthracis o-succinylbenzoyl-CoA synthetase: reaction kinetics and a novel inhibitor mimicking its reaction intermediate
.
Biochemistry
47
,
12434
12447
13
Fang
,
M.
,
Toogood
,
R.D.
,
Macova
,
A.
,
Ho
,
K.
,
Franzblau
,
S.G.
,
McNeil
,
M.R.
et al. 
(
2010
)
Succinylphosphonate esters are competitive inhibitors of MenD that show active-site discrimination between homologous α-ketoglutarate-decarboxylating enzymes
.
Biochemistry
49
,
2672
2679
14
Matarlo
,
J.S.
,
Lu
,
Y.
,
Daryaee
,
F.
,
Daryaee
,
T.
,
Ruzsicska
,
B.
,
Walker
,
S.G.
et al. 
(
2016
)
A methyl 4-oxo-4-phenylbut-2- enoate with in vivo activity against MRSA that inhibits MenB in the bacterial menaquinone biosynthesis pathway
.
ACS Infect. Dis.
2
,
329
340
15
Song
,
H.
,
Dong
,
C.
,
Qin
,
M.
,
Chen
,
Y.
,
Sun
,
Y.
,
Liu
,
J.
et al. 
(
2016
)
A thiamine-dependent enzyme utilizes an active tetrahedral intermediate in vitamin K biosynthesis
.
J. Am. Chem. Soc.
138
,
7244
7247
16
Qin
,
M.
,
Song
,
H.
,
Dai
,
X.
,
Chen
,
Y.
,
Chan
,
C.-K.
,
Chan
,
W.
et al. 
(
2018
)
Single turnover kinetics reveal a distinct mode of thiamine diphosphate-dependent catalysis in vitamin K biosynthesis
.
ChemBioChem
19
,
1514
1522
17
Dawson
,
A.
,
Fyfe
,
P.K.
and
Hunter
,
W.N.
(
2008
)
Specificity and reactivity in menaquinone biosynthesis: the structure of Escherichia coli MenD (2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexadiene-1-carboxylate synthase)
.
J. Mol. Biol.
384
,
1353
1368
18
Dawson
,
A.
,
Chen
,
M.
,
Fyfe
,
P.K.
,
Guo
,
Z.
and
Hunter
,
W.N.
(
2010
)
Structure and reactivity of Bacillus subtilis MenD catalyzing the first committed step in menaquinone biosynthesis
.
J. Mol. Biol.
401
,
253
264
19
Jirgis
,
E.N.M.
,
Bashiri
,
G.
,
Bulloch
,
E.M.M.
,
Johnston
,
J.M.
and
Baker
,
E.N.
(
2016
)
Structural views along the Mycobacterium tuberculosis MenD reaction pathway illuminate key aspects of thiamin diphosphate-dependent enzyme mechanisms
.
Structure
24
,
1
11
20
Fiedler
,
E.
,
Thorell
,
S.
,
Sandalova
,
T.
,
Golbik
,
R.
,
König
,
S.
and
Schneider
,
G.
(
2002
)
Snapshot of a key intermediate in enzymatic thiamin catalysis: crystal structure of the α-carbanion of (α,β-dihydroxyethyl)-thiamin diphosphate in the active site of transketolase from Saccharomyces cerevisiae
.
Proc. Natl Acad. Sci. U.S.A.
99
,
591
595
21
Nakai
,
T.
,
Nakagawa
,
N.
,
Maoka
,
N.
,
Masui
,
R.
,
Kuramitsu
,
S.
and
Kamiya
,
N.
(
2004
)
Ligand-induced conformational changes and a reaction intermediate in branched-chain 2-oxo acid dehydrogenase (E1) from Thermus thermophilus HB8, as revealed by X-ray crystallography
.
J. Mol. Biol.
337
,
1011
1033
22
Wille
,
G.
,
Meyer
,
D.
,
Steinmetz
,
A.
,
Hinze
,
E.
,
Golbik
,
R.
and
Tittmann
,
K.
(
2006
)
The catalytic cycle of a thiamin diphosphate enzyme examined by cryocrystallography
.
Nat. Chem. Biol.
2
,
324
328
23
Wagner
,
T.
,
Barilone
,
N.
,
Alzari
,
P.M.
and
Bellinzoni
,
M.
(
2014
)
A dual conformation of the post-decarboxylation intermediate is associated with distinct enzyme states in mycobacterial KGD (α-ketoglutarate decarboxylase)
.
Biochem. J.
457
,
425
434
24
Chabrière
,
E.
,
Vernède
,
X.
,
Guigliarelli
,
B.
,
Charon
,
M.H.
,
Hatchikian
,
E.C.
and
Fontecilla-Camps
,
J.C.
(
2001
)
Crystal structure of the free radical intermediate of pyruvate: ferredoxin oxidoreductase
.
Science
294
,
2559
2563
25
Amara
,
P.
,
Fontecilla-Camps
,
J.C.
and
Field
,
M.J.
(
2007
)
The enamine intermediate may not be universal to thiamine catalysis
.
Angew. Chem. Int. Ed. Engl.
119
,
9177
9180
26
Meyer
,
D.
,
Neumann
,
P.
,
Koers
,
E.
,
Sjuts
,
H.
,
Lüdtke
,
S.
,
Sheldrick
,
G.M.
et al. 
(
2012
)
Unexpected tautomeric equilibria of the carbanion-enamine intermediate in pyruvate oxidase highlight unrecognized chemical versatility of thiamin
.
Proc. Natl Acad. Sci. U.S.A.
109
,
10867
10872
27
Suzuki
,
R.
,
Katayama
,
T.
,
Kim
,
B.J.
,
Wakagi
,
T.
,
Shoun
,
H.
,
Ashida
,
H.
et al. 
(
2010
)
Crystal structures of phosphoketolase thiamine diphosphate-dependent dehydration mechanism
.
J. Biol. Chem.
285
,
34279
34287
28
Paul
,
M.
,
Breugst
,
M.
,
Neudörfl
,
J.M.
,
Sunoj
,
R.B.
and
Berkessel
,
A.
(
2016
)
Keto–enol thermodynamics of Breslow intermediates
.
J. Am. Chem. Soc.
138
,
5044
5051
29
Patel
,
H.
,
Nemeria
,
N.S.
,
Brammer
,
L.A.
,
Meyers
,
C.L.F.
and
Jordan
,
F.
(
2012
)
Observation of thiamin-bound intermediates and microscopic rate constants for their interconversion on 1-deoxy-d-xylulose 5-phosphate synthase: 600-fold rate acceleration of pyruvate decarboxylation by d-glyceraldehyde-3-phosphate
.
J. Am. Chem. Soc.
134
,
18374
18379
30
Berthold
,
C.L.
,
Gocke
,
D.
,
Wood
,
M.D.
,
Leeper
,
F.
,
Pohl
,
M.
and
Schneider
,
G.
(
2007
)
Structure of the branched-chain keto acid decarboxylase (KdcA) from Lactococcus lactis provides insights into the structural basis for the chemoselective and enantioselective carboligation reaction
.
Acta Crystallogr. D Biol. Crystallogr. D
63
,
1217
1224
31
Versees
,
W.
,
Spaepen
,
S.
,
Wood
,
M.D.
,
Leeper
,
F.J.
,
Vanderleyden
,
J.
and
Steyaert
,
J.
(
2007
)
Molecular mechanism of allosteric substrate activation in a thiamine diphosphate-dependent decarboxylase
.
J. Biol. Chem.
282
,
35269
35278
32
Pei
,
X.Y.
,
Titman
,
C.M.
,
Frank
,
R.A.
,
Leeper
,
F.J.
and
Luisi
,
B.F.
(
2008
)
Snapshots of catalysis in the E1 subunit of the pyruvate dehydrogenase multienzyme complex
.
Structure
16
,
1860
1872
33
Fang
,
M.
,
Macova
,
A.
,
Hanson
,
K.L.
,
Kos
,
J.
and
Palmer
,
D.R.J.
(
2011
)
Using substrate analogues to probe the kinetic mechanism and active site of Escherichia coli MenD
.
Biochemistry
50
,
8712
8721
34
Jiang
,
M.
and
Guo
,
Z.
(
2007
)
Effects of macromolecular crowding on the intrinsic catalytic efficiency and structure of enterobactin-specific isochorismate synthase
.
J. Am. Chem. Soc.
129
,
730
731
35
Guo
,
Z.-F.
,
Jiang
,
M.
,
Zheng
,
S.
and
Guo
,
Z.
(
2010
)
Structural change of the enterobactin synthetase in crowded solution and its relation to crowding-enhanced product specificity in nonribosomal enterobactin biosynthesis
.
Bioorg. Med. Chem. Lett.
20
,
3855
3858
36
Wang
,
Q.-S.
,
Yu
,
F.
,
Huang
,
S.
,
Sun
,
B.
,
Zhang
,
K.-H.
,
Liu
,
K.
et al. 
(
2015
)
The macromolecular crystallography beamline of SSRF
.
Nuclear Sci. Technol.
26
,
010102
37
Otwinowski
,
Z.
and
Minor
,
W.
(
1997
)
Processing of X-ray diffraction data collected in oscillation mode
.
Methods Enzymol.
276
,
307
326
38
Battye
,
T.G.G.
,
Kontogiannis
,
L.
,
Johnson
,
O.
,
Powell
,
H.R.
and
Leslie
,
A.G.W.
(
2011
)
iMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM
.
Acta Crystallogr. D Biol. Crystallogr.
67
,
271
281
39
Evans
,
P.R.
and
Murshudov
,
G.N.
(
2013
)
How good are my data and what is the resolution?
Acta Crystallogr. D Biol. Crystallogr.
69
,
1204
1214
40
Mccoy
,
A.J.
,
Grosse-Kunstleve
,
R.W.
,
Adams
,
P.D.
,
Winn
,
M.D.
,
Storoni
,
L.C.
and
Read
,
R.J.
(
2007
)
Phaser crystallographic software
.
J. Appl. Crystallogr.
40
,
658
674
41
Emsley
,
P.
,
Lohkamp
,
B.
,
Scott
,
W.G.
and
Cowtan
,
K.
(
2010
)
Features and development of Coot
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
486
501
42
Adams
,
P.D.
,
Afonine
,
P.V.
,
Bunkoczi
,
G.
,
Chen
,
V.B.
,
Davis
,
I.W.
,
Echols
,
N.
et al. 
(
2010
)
PHENIX: a comprehensive Python-based system for macromolecular structure solution
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
213
221
43
Schuttelkopf
,
A.W.
and
van Aalten
,
D.M.
(
2004
)
PRODRG: a tool for high-throughput crystallography of protein-ligand complexes
.
Acta Crystallogr. D Biol. Crystallogr.
60
,
1355
1363
44
Moriarty
,
N.W.
,
Grosse-Kunstleve
,
R.W.
and
Adams
,
P.D.
(
2009
)
Electronic ligand builder and optimization workbench (eLBOW): a tool for ligand coordinate and restraint generation
.
Acta Crystallogr. D Biol. Crystallogr.
65
,
1074
1080
45
Laskowski
,
R.A.
,
Macarthur
,
M.W.
,
Moss
,
D.S.
and
Thornton
,
J.M.
(
1993
)
Procheck — a program to check the stereochemical quality of protein structures
.
J. Appl. Crystallogr.
26
,
283
291
46
Chen
,
V.B.
,
Arendall
,
W.B.
,
Headd
,
J.J.
,
Keedy
,
D.A.
,
Immormino
,
R.M.
,
Kapral
,
G.J.
et al. 
(
2010
)
Molprobity: all-atom structure validation for macromolecular crystallography
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
12
21
47
DeLano
,
W.L.
(
2002
)
The PyMOL Molecular Graphics System
,
DeLano Scientific
,
San Carlos, CA
48
Krissinel
,
E.
and
Henrick
,
K.
(
2007
)
Inference of macromolecular assemblies from crystalline state
.
J. Mol. Biol.
372
,
774
797
49
Dolinsky
,
T.J.
,
Nielsen
,
J.E.
,
McCammon
,
J.A.
and
Baker
,
N.A.
(
2004
)
PDB2PQR: an automated pipeline for the setup, execution, and analysis of Poisson-Boltzmann electrostatics calculations
.
Nucleic Acids Res.
32
,
W665
W667
50
The UniProt Consortium
. (
2017
)
Uniprot: the universal protein knowledgebase
.
Nucleic Acids Res.
45
,
D158
D169
51
Sievers
,
F.
,
Wilm
,
A.
,
Dineen
,
D.
,
Gibson
,
T.J.
,
Karplus
,
K.
,
Li
,
W.Z.
et al. 
(
2011
)
Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega
.
Mol. Syst. Biol.
7
,
539
52
Robert
,
X.
and
Gouet
,
P.
(
2014
)
Deciphering key features in protein structures with the new ENDscript server
.
Nucleic Acids Res.
42
,
W320
W324
53
Morris
,
G.M.
,
Huey
,
R.
,
Lindstrom
,
W.
,
Sanner
,
M.F.
,
Belew
,
R.K.
,
Goodsell
,
D.S.
et al. 
(
2009
)
Autodock4 and AutoDockTools4: automated docking with selective receptor flexibility
.
J. Comput. Chem.
30
,
2785
2791
54
Müller
,
M.
,
Gocke
,
D.
and
Pohl
,
M.
(
2009
)
Thiamin diphosphate in biological chemistry: exploitation of diverse thiamin diphosphate-dependent enzymes for asymmetric chemoenzymatic synthesis
.
FEBS J.
276
,
2894
2904
55
Nemeria
,
N.S.
,
Chakraborty
,
S.
,
Balakrishnan
,
A.
and
Jordan
,
F.
(
2009
)
Reaction mechanisms of thiamin diphosphate enzymes: defining states of ionization and tautomerization of the cofactor at individual steps
.
FEBS J.
276
,
2432
2446
56
Paulikat
,
M.
,
Wechsler
,
C.
,
Tittmann
,
K.
and
Mata
,
R.A.
(
2017
)
Theoretical studies of the electronic absorption spectra of thiamin diphosphate in pyruvate decarboxylase
.
Biochemistry
56
,
1854
1864
57
Kluger
,
R.
and
Tittmann
,
K.
(
2008
)
Thiamin diphosphate catalysis: enzymic and nonenzymic covalent intermediates
.
Chem. Rev.
108
,
1797
1833
58
Abell
,
L.M.
and
Schloss
,
J.V.
(
1991
)
Oxygenase side reactions of acetolactate synthase and other carbanion-forming enzymes
.
Biochemistry
30
,
7883
7887

Author notes

*

Present address: Division of Structural Biology, Wellcome Centre of Human Genetics, University of Oxford, Roosevelt Drive, Oxford OX3 7BN, U.K.

Present address: Center for Cancer Research, National Cancer Institute, Building 37, Room 3118, Bethesda, MD 20892, U.S.A.