Structure of a thermostable methionine adenosyltransferase from Thermus thermophilus HB27 reveals a novel fold of the flexible loop

Abstract

Methionine adenosyltransferases (MATs) are the family of enzymes which synthesize S-adenosylmethionine (AdoMet), the major biological methyl donor. Three-dimensional structures have been reported for MATs from several different species, including bacteria, archaea and eukarya. A common feature in MAT structures is a flexible loop which is proposed to serve as a dynamic lid controlling the access of substrate to the active site. In this study, we solved the X-ray structure of a thermostable MAT from Thermus thermophilus HB27 (TtMAT) at 2.67 Å resolution. Both the tetrameric assembly and the residues in the active site of TtMAT are similar to those of MAT from Escherichia coli (EcMAT). However, the flexible loop in TtMAT is longer than that of EcMAT and well-defined in an open conformation, which is unusual among known MAT structures. Moreover, the loop of TtMAT is in a conformation different from those of other MATs which contain ordered/disordered loops. A further analysis of this loop suggested that it might explain the better thermostability of TtMAT.

---

1. Introduction

Methionine adenosyltransferase (S-adenosylmethionine synthase, MAT, EC 2.5.1.6) is a ubiquitous enzyme which catalyzes the reaction between methionine and ATP, leading to the synthesis of S-adenosylmethionine (SAM).^1,2 SAM, a cofactor abundant in all domains of life, plays important roles in different types of enzymatic reactions and could donate all the groups surrounding the sulfur atom.³ The major role of SAM is as a methyl group donor for C-, N-, S- and O-methylation, usually catalyzed by methyltransferases.⁴ Methyltransferases use the methyl group obtained from SAM to synthesize a broad range of compounds, such as phospholipids and neurotransmitters.³ In addition, SAM also contributes in the biosynthesis of polyamines, as well as an important source for 5-deoxyadenosyl radicals which are involved in a variety of reactions such as isomerization, insertion and elimination.⁵

The synthesis of SAM is carried out solely by MAT and occurs in a two-step reaction.

First, the adenosyl group is transferred from ATP to the sulfur atom of methionine through a S_N2 reaction. In the second step, the tripolyphosphate formed in the first step is further hydrolyzed into pyrophosphate and orthophosphate before SAM is released.⁶ To date, most of the structural studies have been carried out for the bacterial representatives such as Escherichia coli (EcMAT),^7–9 Burkholderia pseudomallei (BpMAT)¹⁰ and several Mycobacterium species, as well as those from eukaryotic origins such as Rattus norvegicus⁴ and Homo sapiens.¹¹ Meanwhile, structures are also available with various ligands in the active site, such as substrates (ATP and methionine), products (SAM and tripolyphosphate analogs) and inhibitors. These enzyme–ligand complex structures together contributed to the elucidation of the reaction sequence. A putative mechanism has been resolved based on the EcMAT structures with different ligand combinations. In this mechanism, a conserved histidine residue acts as an acid to cleave the C5′–O5′ bond of ATP (triphosphate-ribose bond), and then the S of methionine makes a nucleophilic attack on the C5′ to form SAM.⁹ However, a substantially different mechanism has been proposed for MAT from R. norvegicus, which involves extensive movements of the substrate ATP in the active site of the enzyme.⁴

Consistent with the high sequence homology among eukaryotic and bacterial MATs, structures of MAT homologs known to date all share both secondary structural elements and three-dimensional folds. Recently, structures of two archaeal MATs have been reported,^12,13 which further confirmed the conservation of their structural folds as well as reaction mechanisms with bacterial enzymes. All known MATs function as oligomeric proteins (tetramer or dimer), harbor a highly conserved three-domain fold,³ and require Mg²⁺ and K⁺ for full activity.¹⁴ The most important and common feature among the MAT enzymes is a flexible loop (98–114 amino acids in EcMAT), which has been proposed to serve as a dynamic lid that controls access to the active site. It has been suggested that the loop is in an open conformation when no substrate is in the active site of the enzyme, and enzymes crystallized in this conformation will have a “disordered” loop with no density in this region. However, when substrates bind to the active site in the open conformation, the loop will move extensively and folds into a helix to fully block the entrance to the active site and facilitate the synthesis reaction. Thus, a fully ordered loop could be resolved in the structures of MATs crystallized with SAM product in their active sites.^8,9

Enzymes derived from thermophiles and hyperthermophiles, have been the subject of intensive investigations during these years due to their ability in retaining their structure and function at very high temperature.¹⁵ Thermophilic enzymes are superior to the conventional catalysts in that they could catalyze reactions even under severe conditions, under which mesophilic proteins are completely inactive. In our previous work, we have identified a thermostable methionine adenosyltransferase from Thermus thermophilus HB27 (TtMAT) and characterized its kinetic parameters, optimal temperature and pH value, and ion preference.¹⁶ Thermus thermophilus HB27 is a Gram-negative aerobic bacterium, living at very high-temperate environment and the optimal temperature for its growth ranges from 50 °C to 82 °C. Since the genome sequence of Thermus thermophilus species was determined, a lot of enzymes in different pathways were studied.^17,18 To investigate the structural basis for the thermostability of TtMAT and shed light on its further application, we determined the crystal structure of apo-TtMAT at 2.67 Å resolution. Both the tetrameric assembly and the residues in the active site of TtMAT are similar to those of EcMAT. However, although there are no substrates or products in the structure of TtMAT, the flexible loop is in an open conformation and well-defined, contrary to the structures reported before.^8,9 To our knowledge, the structure of TtMAT represents the first structure of a bacterial MAT enzyme with an ordered loop in all the four subunits but no ligands in the active site. Comparison between the structures of TtMAT and the closed conformation of EcMAT provides insights into how this loop region works after substrates bind into the active sites. Moreover, the flexible loop region of TtMAT is much longer than those of other MAT enzymes with known structures (Fig. 1), which might explain the better thermostability of TtMAT.

Fig. 1 Sequence alignment of MAT homologs from different species. Sequences of MATs from Thermus thermophilus HB27 (TtMAT), Escherichia coli (EcMAT, PDB code: 1RG9), Homo sapiens (HsMAT, PDB code: 2P02), Burkholderia pseudomallei (BpMAT, PDB code: 3IML), Thermococcus kodakarensis (TkMAT, PDB code: 4L4Q) and Sulfolobus solfataricus (SsMAT, PDB code: 4HPV) are aligned. Residues with 100% homology, over 75% homology, and over 50% homology are shaded in dark blue, pink and light blue, respectively. The flexible loop region is marked with red box.

2. Materials and methods

2.1 Cloning, expression and purification

T. thermophilus HB27 cells were grown in LB-medium at 65 °C for 48 h. The cells were harvested by centrifugation. The genomic DNA was isolated using a Bacterial Genomic DNA Isolation Kit (Beijing Biomed Co., Ltd.) according to the manufacturer's instructions. The TtMAT gene (GI: 46197207 in the GenBank database) was amplified from the genomic DNA of T. thermophilus HB27 by PCR using the following primers: 5′-ggaattcCATATGCGCGCGTTGAGGCTGGTCA′ and 5′-cccaagcttAAGCCCGCTTCCCGCCT-3′. The PCR product was purified and digested with Nde I and Hind III restriction endonucleases. The digested PCR product was purified and cloned into pET22b vector (Novagen, Madison, WI, USA). The resulting expression plasmid was subsequently transformed into Escherichia coli BL21(DE3) cells. The cells were grown at 37 °C in LB medium to OD_{600 nm} of 1.0, and overexpression of TtMAT was induced with 0.5 mM isopropyl-β-D-thiogalactopyranoside (IPTG) at 16 °C for 12 h. The cells were harvested and suspended in the lysis buffer (50 mM Tris–HCl, pH 8.0, 300 mM NaCl, 20 mM MgCl₂, 40 mM KCl and 1 mM phenylmethylsulfonyl fluoride), and lysed by sonication on ice. The protein was purified with affinity chromatography using a Ni-NTA column (Qiagen) pre-equilibrated with the lysis buffer and eluted with the lysis buffer supplemented with 300 mM imidazole, which was followed by purification with gel filtration using a Superdex-200 (GE Healthcare) column equilibrated with a buffer containing 10 mM Tris–HCl pH 8.0, 200 mM NaCl, and 5 mM DTT. The purified protein was analyzed by SDS-PAGE. The fractions containing the target protein were pooled and concentrated to 20 mg ml⁻¹. The EcMAT gene was also inserted into pET22b vector, expressed and purified similarly.

2.2 Stability assay

Purified TtMAT and EcMAT enzymes were incubated in 40, 60, and 80 °C for the indicated time. At scheduled time, 20 μl of sample was analyzed for its enzyme activity as follows. The standard assay mixture contained 100 mM Tris–HCl buffer (pH 8.0), 50 mM KCl, 20 mM MgCl₂, 10 mM ATP, 10 mM methione, and 0.4 mg ml⁻¹ enzyme. One unit of MAT activity was defined as the amount of enzyme activity that produces 1 μmol of SAM per minute. After incubation at 70 °C for 20 min, 20% perchloric acid solution was added to stop the reaction. HPLC Shimadzu 2010 equipped with a C18 column (Kromasil, Φ 4.6 × 250 mm, 5 μm, 100 Å) was used for detection of the product. Mobile phase was at a flow rate of 1 ml·min⁻¹ containing 0.01 M sodium 1-hexanesulfonate and 10% acetic acid. The product was monitored under UV-light at 260 nm. The activity at 0 hour was defined as 100% for both enzymes.

2.3 Crystallization

The TtMAT protein was concentrated to 20 mg ml⁻¹ in 10 mM Tris–HCl pH 8.0, 200 mM NaCl and 5 mM DTT. Crystals were grown using the hanging-drop vapor diffusion method. Crystals of TtMAT were grown at 18 °C by mixing an equal volume of the protein (20 mg ml⁻¹) with reservoir solution containing 0.01 M iron(III) chloride hexahydrate, 0.1 M sodium citrate tribasic dehydrate pH 5.6 and 10% Jeffamine M-600.

2.4 Data collection, structural determination and refinement

All the data were collected at the Shanghai Synchrotron Radiation Facility (SSRF) beamline BL17U, integrated and scaled using the HKL2000 package.¹⁹ Further processing was carried out using programs from the CCP4 suite.²⁰ The structure of EcMAT (PDB code: 1RG9, chain A) was selected as the search model for molecular replacement, which was carried out with PHASER.²¹ The final model rebuilding was performed using COOT²² and the protein structure was refined with PHENIX²³ against the data using stereochemistry information and noncrystallographic symmetry (NCS) as restraints. Data-collection and refinement statistics are presented in Table 1. The interfaces of the TtMAT tetramer were analyzed using PDBePISA (http://www.ebi.ac.uk/msd-srv/prot_int/pistart.html).²⁴ All structural model figures were created with PyMOL.²⁵

Table 1 Statistics of data collection and refinement^a

	TtMAT
a Values in parentheses are for the highest resolution shell. , where Ih is the mean intensity of the i observations of symmetry related reflections of h. R = ∑|Fobs − Fcalc|/∑Fobs, where Fcalc is the calculated protein structure factor from the atomic model (Rfree was calculated with 5% of the reflections selected).
Data collection
Space group	C2
Unit cell (Å)	230.024, 77.053, 96.792
Wavelength (Å)	0.979
Resolution (Å)	2.667 (2.762–2.667)
Rmerge%	10.2 (60.8)
I/sigma	11.4 (2.0)
Completeness (%)	99.61 (96.86)
Redundancy	5.1 (5.3)
Wilson B factor (Å^2)	45.6
Refinement
R factor	0.1729
Rfree	0.2327
No. atoms	12 140 protein atoms
Overall B factors	51.90
RMSD bond lengths	0.006
RMSD bond angles	1.23
Ramachandran plot statistics (%)
In preferred regions	98
In allowed regions	1.87
Outliers	0.13

---

2.5 Gel filtration assay

TtMAT and EcMAT proteins purified as described above were subjected to gel filtration analysis (Superdex-200, GE Healthcare). Buffer containing 10 mM Tris pH 8.0, 100 mM NaCl was used for the assay. The assays were performed with a flow rate of 0.5 ml min⁻¹ and an injection volume of 7 ml at 4 °C. The protein fractions were verified by SDS-PAGE.

3. Results

3.1 Stability assay

In our previous studies, a thermostable MAT was identified from Thermus thermophilus HB27 (TtMAT).¹⁶ To further characterize this enzyme and evaluate its application of biotechnological interest, we moved on to test whether TtMAT exhibits better thermostability than the well-studied EcMAT. We incubated TtMAT and EcMAT at 40, 60, and 80 °C for 25 hours and tested the activity loss of both enzymes over time, at these three temperatures. At 40 °C and 60 °C, the activity of EcMAT underwent a quick decrease and dropped to almost 30% of the initial activity after 25 hours, whereas TtMAT showed almost unchanged (40 °C) and even higher activity (60 °C) than the initial activity during the same time (Fig. 2A and B). A possible reason for the increased activity of TtMAT during incubation might be that since the optimal temperature for the growth of Thermus thermophilus HB27 ranges from 50 °C to 82 °C, the TtMAT enzyme purified at low temperature (4 °C) might need some incubation time at its optimal temperature to restore its maximal activity. During this incubation period, the conformational change of the enzyme might organize its active site and surrounding regions to a fully accessible state for the substrate, thus increasing the activity of the enzyme. At 80 °C, the activities of TtMAT and EcMAT dropped to zero in 15 and 5 hours, respectively (Fig. 2C). In all, our results at three different temperatures all suggested that TtMAT is more thermostable than EcMAT and might have a better application in biological industry.

Fig. 2 TtMAT exhibits better thermostability than EcMAT. TtMAT and EcMAT enzyme were incubated in 40 °C (A), 60 °C (B), and 80 °C (C) for the indicated time. At scheduled time, 20 μl of sample was analyzed for its enzyme activity as mentioned in the methods. The activity at 0 hour was defined as 100% for both enzymes.

3.2 Overall structure of TtMAT

To investigate the underlying reason for its thermostability, we solved the crystal structure of apo-TtMAT at 2.67 Å resolution (Table 1). The overall structural topology of TtMAT is essentially the same as the other X-ray structures of MATs reported from other organisms. TtMAT consists of four identical subunits related by 222 symmetry (Fig. 3). Two subunits strongly interact with each other to form a tight dimer, and two dimers associate into a peanut-shape tetrameric enzyme. Supporting this assembly, TtMAT eluted at the same position as EcMAT in the gel filtration assay (Fig. S1†), and the size was equal to a tetramer (The size of a TtMAT monomer is ∼42 kD). For each monomer, the peptide chain folds into three domains (N-terminal, central and C-terminal domains) and is arranged in a disc-like structure (Fig. S2†). Each monomer contains ten β strands, twelve α helices and three 3₁₀ helices (Fig. 1 and S2†). Two such monomers are tightly packed as a dimer, with two active sites located in the interface of the dimer. The superimposition between TtMAT and EcMAT revealed a root-mean-squared displacement (RMSD) of 1.01 Å within 367 amino acids aligned (Fig. S3†), indicating that they exhibit high structural similarities to each other.

Fig. 3 Overall structure of TtMAT. The protein is shown in cartoon model, with the four subunits colored in different colors. All the structure figures were prepared with PyMOL.

3.3 The flexible loop of TtMAT

Although the overall structure of TtMAT exhibits close resemblance to that of EcMAT, an obvious difference exists between the flexible loop regions of the two structures. The flexible loop region was previously suggested to control access to the active site at the dimer interface.^3,9 In the structures with bound ligands in the active site, the loop is in the “closed” conformation as seen in EcMAT (PDB entries 1RG9 and 1P7L) and HsMAT (PDB code: 2P02). However, in the apo-enzyme structures the loop is in the “open” conformation and is disordered in most structures. The structure of EcMAT reported by Fu et al. (PDB code: 1FUG)⁷ was an exception in an open conformation but with an ordered loop, as it was determined with crystals crystallized at 4 °C, compared to 18–20 °C under which most crystals are crystallized. In contrast to the former view that the loop can only be defined in the “closed” structure, the structure of TtMAT has a well-defined loop in the corresponding region, but represents a fully open structure without any ligands in the active site (Fig. 4).

Fig. 4 Flexible loop regions are different in TtMAT and EcMAT (PDB code: 1RG9). (A and B) EcMAT whose loop region is colored yellow, is shown in cartoon model (A) and surface model (B), respectively. (C and D) TtMAT whose loop region is colored blue, is shown in cartoon model (C) and surface model (D), respectively. The oval-shaped region indicates the active site region in MAT enzymes.

Moreover, the flexible loop region of TtMAT is longer than that of EcMAT and most MAT enzymes by approximately 10 residues (Fig. 1). This region in TtMAT involves residues 100–126, extends mainly as a loop followed by two short α helices (α3 and α4) (Fig. 4 and S4†), whereas in the closed structure of EcMAT (PDB code: 1RG9) it folds into a helix. Comparison of the two structures suggest how the binding of the ligands might trigger the movement of the flexible loop: upon substrates bind to the active site near the flexible loop of the open structure, the flexible loop moves towards the substrate and at the same time packs against the other protomer in the tight dimer. In the other protomer, α6, η2, β8 (numbered as in TtMAT structure) and the loops connecting these secondary structures also move towards the flexible loop region, thus completely blocking the entrance to the active site.

3.4 The active site of TtMAT

MATs have two active sites located in the dimer interface, each of which is contributed by the residues from both protomers.^12,13 Komoto et al. used AMPPNP instead of ATP together with Met and trapped EcMAT in the structure containing products SAM and PPNP,⁹ since PPNP could not be hydrolyzed as tripolyphosphate. Superimposition of the structures of TtMAT and EcMAT (PDB code: 1RG9) revealed that the interacting residues with the two products SAM and PPNP exhibit distinct conservativeness. For PPNP and K⁺ and Mg²⁺ surrounding it, all the interacting residues from both protomers are essentially the same between TtMAT and EcMAT, and the spatial positions are also highly conserved between the two structures (Fig. 5). For the other product SAM, although the positions of several residues which form hydrogen bonds with O or N atoms of SAM are conserved, other residues are not superimposed well (Fig. 6). Typically, in EcMAT the carboxyl group of SAM is stabilized by the polar group of Q98 from the flexible loop, whereas the corresponding residue in TtMAT (Q100) is far from the carboxyl group due to the different conformations of the flexible loop in the two structures. In addition, the adenine amino group of SAM is stabilized by a hydrogen bond with the carbonyl of R229 in EcMAT. In the TtMAT structure, the carbonyl of R242 is 4.1 Å to the adenine amino group, a distance too far to form a hydrogen bond. Consistently, R242 is in the loop after β8, which is proposed to move towards the other protomer when the reaction proceeds.

Fig. 5 Superimposition of the PPNP binding sites in TtMAT and EcMAT (PDB code: 1RG9). The structures of TtMAT and EcMAT are shown as in Fig. 4. An enlarged view of the active site surrounding PPNP in EcMAT is shown left. The PPNP, and active site residues from both proteins are shown in sticks, colored orange (PPNP), cyan (TtMAT) and magenta (EcMAT), respectively. The number of the amino acids are marked with A or B in parentheses, indicating whether they are from the monomer in the top (B) or the bottom (A). K⁺ and Mg²⁺ are shown in green and gray balls, respectively.

Fig. 6 Superimposition of the SAM binding sites in TtMAT and EcMAT (PDB code: 1RG9). The structures of TtMAT and EcMAT are shown as in Fig. 4. The SAM, and active site residues from both proteins are shown in sticks, colored yellow (SAM), cyan (TtMAT) and magenta (EcMAT), respectively. The number of the amino acids are marked with A or B in parentheses, indicating whether they are from the monomer in the top (B) or the bottom (A).

4. Discussion

As a bacterial MAT, TtMAT shows high resemblance to EcMAT in structure, however, distinct features also exist between them, mainly in the flexible loop region, which is located near the active site. The flexible loop actually forms a dynamic lid to control the entrance of substrates into the active site, supported by the kinetic and X-ray crystallographic data. Consistent with its “dynamic” feature, this loop region was discovered in an ordered conformation in only a few of the reported crystal structures, those for MAT II (PDB code: 2P02), EcMAT (1FUG, 1P7L, and 1RG9), all of which are in the closed conformation with substrates in the active site except 1FUG which is solved by crystals crystallized under low temperature. However, the loop structure of TtMAT is also different from that of 1FUG, suggesting that the conformation caught in that structure might represent a half-open structure (Fig. S5A†), as one part of the flexible loop has moved down to the other protomer.

Previous to our study, the structures of two thermostable MATs have been reported, both of which are archaeal enzymes (MATs from Thermococcus kodakarensis, TkMAT, and from Sulfolobus solfataricus, SsMAT). To our knowledge, TtMAT reported in this study represent the first bacterial thermostable MAT. Although MATs from archaea, bacteria and eukaryotic organisms share the similar three-dimensional structures, supporting that these enzymes share an early common ancestor, archaeal enzymes show low sequence identity (less than 20%) with the bacterial and eukaryotic enzymes (Fig. 1), which indicated that they evolved independently from the common ancestor.

Although the overall three-dimensional structure of TtMAT resembles those of TkMAT and SsMAT, there are still remarkable differences among them. The RMSD is 2.63 Å between TtMAT and SsMAT, 2.53 Å between TtMAT and TkMAT, but 1.26 Å between TkMAT and SsMAT. It indicates that MATs from bacteria and eukaryotes are more similar to each other in structure, but less similar to those from archaea. We also compared the flexible loop regions of the three enzymes in the open conformations (Fig. S5B†). In contrast with the undefined loop in SsMAT, TkMAT also contains a well-defined flexible loop in an open conformation. In the structure of TkMAT, this loop folds into a 12-residue helix (144–155 amino acids), different from the loop in TtMAT (Fig. S5B†), which might be caused by the different domains of life Thermus thermophilus and Thermococcus kodakarensis belong to. Interestingly, TtMAT and TkMAT are the only two MATs ever reported with a well-defined flexible loop in an open conformation, and both of them are thermostable MATs. Thus, we suggest that the flexible loop might participate in the thermostability of MATs.

The important residues in the active site are all conserved between the EcMAT and TtMAT, suggesting that they catalyze following an identical mechanism. However, the amino acid residues in the flexible loop region are poorly conserved among TtMAT, EcMAT and various species (Fig. 1), which might explain their different catalytic efficiencies and thermostabilities. Further experiments are needed to test whether the composition of the loop region can affect the thermostability of MATs. Up to now, several studies have been conducted on different enzymes such as dihydrofolate reductase,^26,27 triosephosphate isomerase,²⁸ ribulose-1,5-bisphosphate carboxylase/oxygenase²⁹ and tryptophan synthetase³⁰ to study the importance of loop regions.⁷

5. Conclusions

Here in this study, we studied the thermostability of TtMAT and presented its three-dimensional structure. Although the overall structure and active site are very similar to those of bacterial and eukaryotic enzymes, the conserved flexible loop in MATs is longer in TtMAT and well-defined in an open structure without any ligands in the active site. The loop has been proposed as a flexible lid controlling the entrance of substrate to the active site. The conformation of the loop in TtMAT is also different from those of either EcMAT or TkMAT in the open conformation. We propose that the longer loop, composed of un-conservative residues, might be responsible for the better thermostability of TtMAT. Future studies should be conducted to study the function of this loop region in the thermostability of TtMAT. As thermostable enzymes have often been considered as excellent catalysts of biological industries, TtMAT has great potential for application in the enzymatic production of SAM. Our structural characterization of TtMAT in this study could shed light on the mechanism of its catalysis and thermostability, and facilitate its further industrial application.

Acknowledgements

We would like to thank the staff at beamline BL17U of the Shanghai Synchrotron Radiation Facility for their assistance with data collection. This work was supported by the National Nature Science Foundation of China (31400635 and 21436002), the Beijing Natural Science Foundation (5154031) and the National Basic Research Program of China (973 program) (2013CB733600).

References

1. S. C. Lu and J. M. Mato, J. Gastroenterol. Hepatol., 2008, 23, S73–S77.
2. M. Fontecave, M. Atta and E. Mulliez, Trends Biochem. Sci., 2004, 29, 243–249.
3. G. D. Markham and M. A. Pajares, Cell. Mol. Life Sci., 2009, 66, 636–648.
4. B. González, A. Pajares, J. A. Hermoso, D. Guillerm, G. Guillerm and J. Sanz-Aparicio, J. Mol. Biol., 2003, 331, 407–416.
5. P. A. Frey, A. D. Hegeman and F. J. Ruzicka, Crit. Rev. Biochem. Mol. Biol., 2008, 43, 63–88.
6. F. Wang, S. Singh, J. Zhang, T. D. Huber, K. E. Helmich, M. Sunkara, K. A. Hurley, R. D. Goff, C. A. Bingman and A. J. Morris, FEBS J., 2014, 281, 4224–4239.
7. Z. Fu, Y. Hu, G. D. Markham and F. Takusagawa, J. Biomol. Struct. Dyn., 1996, 13, 727–739.
8. F. Takusagawa, S. Kamitori, S. Misaki and G. D. Markham, J. Biol. Chem., 1996, 271, 136–147.
9. J. Komoto, T. Yamada, Y. Takata, G. D. Markham and F. Takusagawa, Biochemistry, 2004, 43, 1821–1831.
10. L. Baugh, L. A. Gallagher, R. Patrapuvich, M. C. Clifton, A. S. Gardberg, T. E. Edwards, B. Armour, D. W. Begley, S. H. Dieterich, D. M. Dranow, J. Abendroth, J. W. Fairman, D. Fox 3rd, B. L. Staker, I. Phan, A. Gillespie, R. Choi, S. Nakazawa-Hewitt, M. T. Nguyen, A. Napuli, L. Barrett, G. W. Buchko, R. Stacy, P. J. Myler, L. J. Stewart, C. Manoil and W. C. Van Voorhis, PLoS One, 2013, 8, e53851.
11. N. Shafqat, J. R. Muniz, E. S. Pilka, E. Papagrigoriou, F. von Delft, U. Oppermann and W. W. Yue, Biochem. J., 2013, 452, 27–36.
12. J. Schlesier, J. Siegrist, S. Gerhardt, A. Erb, S. Blaesi, M. Richter, O. Einsle and J. N. Andexer, BMC Struct. Biol., 2013, 13, 22.
13. F. Wang, S. Singh, J. Zhang, T. D. Huber, K. E. Helmich, M. Sunkara, K. A. Hurley, R. D. Goff, C. A. Bingman, A. J. Morris, J. S. Thorson and G. N. Phillips Jr, FEBS J., 2014, 281, 4224–4239.
14. J. Mato, L. Alvarez, P. Ortiz and M. A. Pajares, BMC Struct. Biol., 1997, 73, 265–280.
15. M. Porcelli, C. P. Ilisso, L. Mosca and G. Cacciapuoti, Bioengineered, 2015, 6, 184–186.
16. Y. Liu, Z. Wang, L. Liu and T. Tan, Front. Chem. Sci. Eng., 2016 DOI:10.1007/s11705-016-1566-2.
17. N. Ohtani, M. Tomita and M. Itaya, J. Bacteriol., 2010, 192, 5499–5505.
18. A. Henne, H. Brüggemann, C. Raasch, A. Wiezer, T. Hartsch, H. Liesegang, A. Johann, T. Lienard, O. Gohl and R. Martinez-Arias, Nat. Biotechnol., 2004, 22, 547–553.
19. M. W. Otwinowski Z, Macromolecular Crystallography, Part A, 1997, 307–326.
20. Collaborative Computational Project Number 4, Acta Crystallogr., Sect. D: Biol. Crystallogr., 1994, 50, 760–763.
21. A. J. McCoy, R. W. Grosse-Kunstleve, P. D. Adams, M. D. Winn, L. C. Storoni and R. J. Read, J. Appl. Crystallogr., 2007, 40, 658–674.
22. P. Emsley and K. Cowtan, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2004, 60, 2126–2132.
23. P. D. Adams, R. W. Grosse-Kunstleve, L. W. Hung, T. R. Ioerger, A. J. McCoy, N. W. Moriarty, R. J. Read, J. C. Sacchettini, N. K. Sauter and T. C. Terwilliger, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2002, 58, 1948–1954.
24. E. Krissinel and K. Henrick, J. Mol. Biol., 2007, 372, 774–797.
25. D. WL, 2002.
26. K. Gekko, K. Yamagami, Y. Kunori, S. Ichihara, M. Kodama and M. Iwakura, J. Biochem., 1993, 113, 74–80.
27. L. Li, P. E. Wright, S. J. Benkovic and C. J. Falzone, Biochemistry, 1992, 31, 7826–7833.
28. D. L. Pompliano, A. Peyman and J. R. Knowles, Biochemistry, 1990, 29, 3186–3194.
29. P. Curmi, D. Cascio, R. Sweet, D. Eisenberg and H. Schreuder, J. Biol. Chem., 1992, 267, 16980–16989.
30. W. Lim, S. Sarkar and J. Hardman, J. Biol. Chem., 1991, 266, 20205–20212.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra27938k

‡ These authors contributed equally to this work.

---