Biochemical and molecular characterization of the isocitrate dehydrogenase with dual coenzyme specificity from the obligate methylotroph Methylobacillus Flagellatus

The isocitrate dehydrogenase (MfIDH) with unique double coenzyme specificity from Methylobacillus flagellatus was purified and characterized, and its gene was cloned and overexpressed in E. coli as a fused protein. This enzyme is homodimeric,—with a subunit molecular mass of 45 kDa and a specific activity of 182 U mg -1 with NAD+ and 63 U mg -1 with NADP+. The MfIDH activity was dependent on divalent cations and Mn2+ enhanced the activity the most effectively. MfIDH exhibited a cofactor-dependent pH-activity profile. The optimum pH values were 8.5 (NAD+) and 6.0 (NADP+).The Km values for NAD+ and NADP+ were 113 μM and 184 μM respectively, while the Km values for DL-isocitrate were 9.0 μM (NAD+), 8.0 μM (NADP+). The MfIDH specificity (kcat/Km) was only 5-times higher for NAD+ than for NADP+. The purified MfIDH displayed maximal activity at 60°C. Heat-inactivation studies showed that the MfIDH was remarkably thermostable, retaining full activity at 50°C and losting ca. 50% of its activity after one hour of incubation at 75°C. The enzyme was insensitive to the presence of intermediate metabolites, with the exception of 2 mM ATP, which caused 50% inhibition of NADP+-linked activity. The indispensability of the N6 amino group of NAD(P)+ in its binding to MfIDH was demonstrated. MfIDH showed high sequence similarity with bacterial NAD(P)+-dependent type I isocitrate dehydrogenases (IDHs) rather than with eukaryotic NAD+-dependent IDHs. The unique double coenzyme specificity of MfIDH potentially resulted from the Lys340, Ile341 and Ala347 residues in the coenzyme-binding site of the enzyme. The discovery of a type I IDH with double coenzyme specificity elucidates the evolution of this subfamily IDHs and may provide fundamental information for engineering enzymes with desired properties.

Insufficient data exist regarding IDHs from methylotrophic bacteria. Lloyd and Weitzman, demonstrated that an IDH from the obligate methylotroph M. methylotrophusis is NAD +linked [28]. Two isozymes, one that is NAD + -dependent and one that is NADP + -dependent, were partially purified from the obligate methylotroph Pseudomonas W6 [29]. The facultative methylotroph P. oleovorans possessed only an NAD + -specific IDH [30].
Methylobacillus flagellatus is an obligate methylotroph with the 2-keto-3-deoxy-6-phosphogluconate aldolase/transaldolase variant of the ribulose monophosphate (RuMP) pathway of formaldehyde fixation [31]. This organism has an incomplete TCA cycle, and lacksα-ketoglutarate dehydrogenase [32]. Thus, the IDH of M. flagellatus provides α-ketoglutarate, which participates in NH 4 + fixation following the reaction catalysed by NADP-dependent glutamate dehydrogenase. In M. flagellatum this enzyme as well as citrate synthase are not regulated by intermediary metabolites [33][34][35]. NH 4 + assimilation might be regulated only at the level of the reactions in the RuMP cycle or by the modulation of IDH activity [35]. Although only one gene encoding an IDH was found in the genome of M. flagellatus, we demonstrated that NAD + -IDH activity is up-regulated and NADP + -IDH activity is down-regulated in the N-limited cultures [33].
In this paper, we report the purification, cloning, heterologous expression, and characterization of the IDH from M. flagellatus (MfIDH). We provide experimental evidence that demonstrates that MfIDH is an enzyme with bona fide double cofactor specificity and its catalytic efficiency with NAD + (k cat /K m ) is comparable to the efficiency of prokaryotic NADP + -IDHs. This detailed enzymatic characterization of MfIDH adds a new and interesting member to the IDH family.

Bacterial strains and growth conditions
Methylobacillus flagellatus ATCC51484 was obtained from the laboratory collection and was aerobically grown at 42˚C in minimal medium, containing 2% (v/v) methanol [31]. The genomic DNA was isolated with a Promega Wizard kit.

Preparation of cell-free extracts and purification of the native enzyme
All procedures were carried out at 4˚C. Cells were harvested by centrifugation, washed with an equal volume of 50 mM potassium phosphate buffer (PPB), pH 7.5, and centrifuged again. The pellet was resuspended in an appropriate amount of the same buffer and disrupted by sonication. The cell debris was removed by centrifuging the sample for 30 min at 10 000×g. The supernatant was heat treated for 30 min at 55˚C and rapidly cooled. The precipitated protein was removed by centrifugation. Solid (NH 4 ) 2 SO 4 was added to the heat-treated extract until the solution reached 50% saturation. The protein that precipitated was removed by centrifugation and discarded. The solution concentration of (NH 4 ) 2 SO 4 was increased to 80% saturation, and the resulting protein precipitate was collected by centrifugation and redissolved in a minimal volume of 50 mM Tris-HCl buffer, pH 7.0 (buffer A). This enzyme solution was loaded onto an S-200 Sephacryl column (90×2.6 cm), that was equilibrated with buffer A. Active fractions, wich were eluted with buffer A, were pooled and loaded onto a Red Sepharose CL-6B (Pharmacia Biotech) column (15×1.6 cm), that had been previously equilibrated with buffer A. The enzyme was eluted with a linear KCl gradient (0-1 M during 15 column volumes). The active fractions were pooled and concentrated, and their buffer was changed to a 25 mM histidine-HCl buffer, pH 6.4 (buffer B) by using a Vivaspin 20 centrifugal concentrator. The enzyme solution was loaded onto a Polybuffer exchanger PBE94 (Pharmacia Biotech) column (30×1.0 cm), that had been previously equilibrated with buffer B. The column was washed with buffer B (one column volume) and then it contents were eluted with 11 column volumes of Polybuffer 74, pH 5.0. The active fractions were pooled and concentrated, and their buffer was changed to 25 mM Tris-HCl, pH 7.2, by using a centrifugal concentrator. The results of a typical purification procedure are reported in S1 Table. Plasmid construction Based on the genome sequence of M. flagellatus ATCC51484 (GenBank accession no. CP000284) [32], two specific primers were designed to amplify the complete IDH gene: sense primer 5 0 -GCGCGCCATGGGCAGCAGCCATCATCATCATCATCACAGCAGCGGCATGTCTACAAAA ATCAAAGTACCCACTACTG-3 0 (NcoI site underlined) and antisense primer 5 0 -CCATTGGATCCTGACATGTGCTTGACGATCTCCGCACCGAATTCTGCACTGC-3 0 (BamHI site underlined). The expression vector pET-15b was used for the heterologous expression of MfIDH. The PCR product containing the IDH gene was purified, digested and ligated into the NcoI/BamHI-digested multiple cloning site of pET-15b; in this manner, the plasmid region encoding the thrombin recognition site was eliminated, creating the recombinant plasmid pET-MfIDH. The presence of the MfIDH gene with a 6x His-tag coding sequence directly downstream of the start codon was confirmed by sequencing.

Overexpression and purification
The E. coli BL21 (DE3) strain harboring the pET-MfIDH plasmid was cultured overnight in Luria-Bertani (LB) medium supplemented with 150 μg/mL of ampicillin at 37˚C. The cells were then inoculated into 100 mL of fresh LB (with the same antibiotic) to a final OD 600nm of up of 0.1 and grown until the cell density reached an OD 600nm of 0.5-0.6. At this time, IPTG was added to the culture at a final concentration of 1 mM; the incubation continued for 3 more hours. The cells were harvested and resuspended in sonication buffer. The cell debris was then removed by centrifuging the sample at 12 000× g for 15 min at 4˚C. The recombinant MfIDH with the 6x-His -tag on its N-terminus was purified using Ni-NTA Affinity Resin (Clontech, La Jolla, CA) according to the manufacturer's instructions. The protein fractions were eluted with an imidazole gradient from 30 mM (in the binding buffer) to 500 mM (in the elution buffer). The fraction containing the recombinant MfIDH was dissolved in a buffer with 50 mM Tris-HCl (pH 7.5) and 10% glycerol. The purity of the recombinant enzyme was analyzed and confirmed using 12% SDS-PAGE. For the Western blot analysis, the SDS-PAGE gels were transferred to a nitrocellulose membrane by electroblotting. The membrane was blocked for 1 h at room temperature with 5% skim milk in a buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, and 0.2% Tween-20. His-tagged polyclonal antibody (Thermo Fisher Scientific, USA) and alkaline phosphatase-conjugated goat anti-rabbit IgG (Promega, USA) were applied to the blot, at the appropriate dilution. The chemiluminescence signal was visualized by exposing the blots to X-ray film.

Measurement of enzyme activity
The IDH activity was routinely measured by monitoring the reduction of NAD + (or NADP + ) at 340 nm. The reaction mixtures were incubated at 42˚C and contained 100 mM Tris-HCl buffer (pH 7.5), 2 mM MnCl 2 , 5.0 mM DL-isocitrate, 0.4 mM NAD + or NADP + and the enzyme, with a total volume of 1.0 mL. After determing their pH optima, NAD + -IDH activity was measured in 100 mM Tris-HCl buffer (pH 8.5), and NADP + -IDH activity was measured in 100 mM Bis-Tris-HCl buffer (pH 6.0). The increase in NAD(P)H concentration was determined by monitoring the absorbance at 340 nm with a thermostated Shimadzu-1800 UV-Vis spectrophotometer (Shimadzu Corp, Japan) and converting the absorbance to concentration using a molar extinction coefficient of 6.22 mM -1 cm -1 . One unit (U) of activity was defined as 1 μmol of NAD(P)H formed per min. The protein concentrations were determined using the Bio-Rad protein assay kit (Bio-Rad, USA) with bovine serum albumin as the standard. All the reported values are the means of at least three independent experiments.
The effects of pH and temperature on the native MfIDH activity were determined in the presence of Mn 2+ . To obtain its pH profile, the enzyme's activity was assayed in 100 mM buffer (Bis-Tris-HCl, pH 5.0-7.0, Tris-HCl, pH 7.5-9.0 or 2-amino-2-methyl-1,3-propanediol, pH 9.5-10.5). The effect of the temperature on the activity was determined for temperatures of up to 65˚C. The influence of temperature on protein stability was investigated by incubating the pure enzyme (0.07 mg/ml) in 50 mM PPB at different temperatures for 60 min. Next, the aliquots were immediately cooled on ice and then their residual activity was assayed. The kinetic parameters for the native MfIDH were determined by measuring the activity of native MfIDH at various concentrations of one substrate (isocitrate and NADP + or NAD + ) while at saturating concentrations of the other substrate. The apparent kinetic parameters were derived from a double-reciprocal Lineweaver-Burk plot.
The effects of different metal ions (2 mM MnCl 2 , 2 mM MgCl 2 , 2 mM CaCl 2 , 2 mM CuSO 4 , and 2 mM ZnSO 4 ), metabolites or cofactor analogs on the native/recombinant activity were determined using the standard assay protocol.
Polyacrylamide gel-electrophoresis 12% SDS-PAGE and non-denaturing gradient polyacrylamide (4-20%, w/v) electrophoresis were carried out as described elsewhere. The IDH activity after electrophoresis was measured by incubating the gel slices in a solution that, stains based on the enzyme activity: 100 mM Tris-HCl buffer (pH 7.5), 2 mM MnCl 2 , 5.0 mM DL-isocitrate, 2.0 mM NAD + or NADP + , 1.0 mM Nitroblue tetrazolium, and 0.5 mM phenazine methosulfate in 100 mM Tris-HCl buffer, pH 7.5. Isoelectric focusing experiments were performed in a horizontal slab gel. Pharmalyte 3-10 (GE Healthcare) was used to obtain a pH gradient.

Results and discussion Enzyme purification and characteristics
The native and the recombinant 6x-His-tagged MfIDHs have essentially the same biochemical characteristics; the native enzyme was used in this study unless otherwise noted. A 232-fold purification of native MfIDH was achieved with a 46% recovery. Polyacrylamide gel electrophoresis in denaturing and non-denaturing conditions showed that the native and recombinant MfIDH were purified to homogeneity. A single protein band with a molecular mass of approximately 45.0 kDa was observed on SDS-PAGE gels, was correlated well with the predicted value (44.6 kDa) (Fig 1A), which was also recognized as His-tag protein in the Western blot ( Fig 1B). After staining them for enzymatic activity, the gradient non-denaturing PAGE gels exhibited one protein band with a molecular mass of approximately 87 kDa, suggesting a homodimeric structure of the native enzyme ( Fig 1C). Size exclusion chromatography (SEC) also confirmed that MfIDH is a homodimer in solution. A single symmetric peak with a molecular mass of approximately 83.6 kDa was observed (Fig 1D) in the gel filtration experiment, while the calculated molecular mass of homodimeric MfIDH was 89.2 kDa.
An isoelectric focusing gel exhibited one protein band with a pI of 5.5, which was lower than the predicted (pI 5.9) (S1 Fig).
The specific activity of the purified MfIDH was 182 U mg -1 when using NAD + and 63 U mg -1 when using NADP + as substrates,demonstrating the enzyme possesses true double coenzyme specificity. The enzyme exhibits Michaelis-Menten kinetics. The enzyme's apparent K m value for DL-isocitrate was 9.0 μM when NAD + was used as a cosubstrate and 8.0 μM when NADP + was used. The kinetic analysis showed that the K m values of MfIDH for NAD + and NADP + were almost equal-113 μM and 184 μM, respectively. The MfIDH specificities [(k cat / K m )NAD/(k cat /K m )NADP] were only 5-times higher for NAD + than for NADP + (Table 1). Based on this observation, MfIDH showed slight preference for NAD + , indicating that MfIDH was wrongly annotated as an NADP + -dependent enzyme in GenBank. However, our results provide solid experimental evidence that this enzyme demonstrates almost equal cofactor specificity, with a slight preference toward NAD + .  According to Zhu et al, NAD + usage is an ancestral trait and NADP + dependency by prokaryotic IDHs emerged near the time that eukaryotic mitochondria first appeared, (some 3.5 billion years ago). The switch of the coenzyme specificity of prokaryotic IDH from NAD + to NADP + is an ancient adaptation to the anabolic demand for NADPH during growth on acetate [1]. The aerobic Gram-negative bacterium M. flagellatus which has an uncoupled TCA cycle contains an IDH that is specific for both NAD + and NADP + , which provides flexibility to use either available cofactor and generate NADH or NADPH.

Sequence analysis
The IDH gene in M.flagellatus (MfIDH) is 1242 bp in length and encodes a polypeptide of 413 amino acids. The overall GC content is approximately 56.17% (genome 55.7%), which is similar to those of the chromosomes of Methylophilaceae species (37-57%) [41]. The search for regions that are identical to the MfIDH gene indicated that the highest identity values were with IDHs from the following organisms: Methylobacillus glycogenes (96%), Methylovorus glucosotrophus (91%), Methylotenera mobilis (87%), Candidatus Methylopumilus turicensis (86%), Methylotenera versatilis (86%), and Methylophilus methylotrophus (86%). The amino acid identities of MfIDH with typical homodimeric NADP + -IDHs from E. coli and B. subtilis, and with NAD + -IDHs from A. thiooxidans and Z. mobilis were 66, 62, 58 and 56%, respectively. The 3D-structure of MfIDH was modeled using the AtIDH (2D4V) structure as a template. A secondary-structure-based alignment revealed that most structural elements that are, involved in the binding of the substrate and coenzyme are highly conserved within prokaryotic homodimeric type I IDHs (Fig 2).
The interactions between the 2'-phosphate of NADP + and the amino acid residues Lys344, Tyr345 and Val351 in EcIDH, and Lys350, Tyr351 and Val357 in BsIDH have been declared the determinants of NADP + cofactor specificity [1,42]. The possibility of switching the cofactor preference was shown experimentally by replacement of the original motif Lys350, Tyr351 and Val357 in NADP + -IDH of E.coli with the mutated motif Asp350, Ile351 and Ala357 in engineered NAD + -IDH of E.coli [42]. According to Dean and Golding, Imada et al; substitution of the Lys with an Asp, results in the formation of double hydrogen bonds with the 2 0 -and 3 0hydroxyl groups of the adenosine ribose of NAD + and the repelling of the negatively charged 2'-phosphate of NADP + through electrostatic repulsion, which together cause the NAD + cofactor specificity of AtIDH (Asp357, Ile358 and Ala364) and ZmIDH (Asp348, Ile349 and Ala355) [42,43]. Furthermore, the amino acid residues Asp328, Ile329 and Ala335 have been declared the determinants of NAD cofactor specificity in NAD-IDH from Pyrococcus furiosus [23]. The site-directed mutagenesis experiment that replaced Asp328 with Lys328 in the cofactor discrimination site of the NAD + -IDH from P. furiosus; led to a significant reduction in K m for NADP (~27fold), whereas the K m for NAD was unaltered and the specificity for NADP was increased five-fold compared with the wild-type enzyme. This motif-Lys328, Ile329 and Ala335 results in a double coenzyme specificity of chimeric P.furiosus IDH. The introduction of the double replacement of Asp-328-Lys/Ile-329-Tyr (motif Lys328, Tyr329 and Ala347) has not changed the efficiency of NADP-IDH, but rather slightly increased both K m and K cat for NADP. The k cat was unaltered compared with the single-mutated enzyme [23]. The structurebased alignment revealed that there is the same motif-Lys340, Ile341 and Ala347 in naturally occurred MfIDH. Thus, signature residues involved in substrate discrimination in MfIDH flagellatus and the ZmIDH model of the NAD-IDH from Z. mobilis were generated using the SWISS-MODEL modeling server with AtIDH structure as the template. The secondary structure of MfIDH is depicted above the alignment. The completely conserved amino acids are highlighted as shaded red boxes. The conserved residues involved in cofactor-(•) and substrate-binding (▲) are indicated, respectively. The conserved phosphorylation sites are indicated (■). The lysine residues that may be acetylated in MfIDH and EcIDH are highlighted with shades light-blue and pink boxes, respectively. The major cofactor specificity determinants are highlighted with shaded yellow boxes and indicated with stars ($). The alignment was drawn with ESPRIPT 3.0.
https://doi.org/10.1371/journal.pone.0176056.g002 appeared to be Lys340, Ile341 and Ala347 (Fig 2). The ability to use efficiently both NAD + and NADP + as cofactors is caused by the presence of these three key amino acids in the protein structure of MfIDH. The MfIDH was incorrectly annotated as NADP + specific isocitrate dehydrogenase. We suggest that MfIDH can be annotated as a homodimeric type I isocitrate dehydrogenase with dual coenzyme specificity.
Post-translational modifications are one of the most efficient biological mechanisms for regulating enzyme activity and cellular physiology. The activity of EcIDH is regulated by an IDH-kinase/phosphatase (aceK) that responds to changes in the metabolic environment [44]. Although phosphorylation sites are conserved in EcIDH (Ser113), AtIDH (Ser113), BsIDH (Ser104), ZmIDH (Ser102) and MfIDH (Ser114), no corresponding IDH-kinase/phosphatase gene was found in the genome of M. flagellatus.

Effects of pH and temperature
The effects of the pH on the MfIDH activity were determined for the NAD + -and NADP +linked reactions in the presence of Mn 2+ . Surprisingly, MfIDH exhibited a strict cofactordependent pH-activity profile, which has never been described in the literatur. The results demonstrate that the optimum pH is 8.5 with NAD + and 6.0 with NADP + (Fig 3A). For NAD +linked activity, this value is similar to those of the Z. mobilis (pH 8.5) [26] and A. thiooxidans NAD + -IDHs (pH 8.5) [21], but is lower than that of the H. thermophilus NAD + -IDH (pH 10.5) [24]. For the NADP + -linked activity, this pH value is rather similar to that of the IDH from the acidophilic fungus A. niger (pH 6.0-8.0) [50]. The temperature for maximum activity MfIDH is approximately 60˚C, which is similar to those of the B. longum IDH (60˚C) [51] and L. interrogan IDH (60˚C) [12], but higher than that of the E. coli IDH (50˚C) [22] (Fig 3B). Heat-inactivation studies revealed that the MfIDH is remarkably thermostable, retaining its full activity at 50˚C and losing ca. 50% of its activity after one hour of incubation at 75˚C (Fig 3C). Thus, the thermostability of MfIDH is closer to that of IDHs from thermophiles rather than mesophiles [52]. The increased thermostability of MfIDH may be explained by its possessing twofold fewer Cys residues than EcIDH does (0.70 and 1.40%, respectively); having fewer Cys residues is a common trend for thermophilic proteins [22]. The aromatic cluster in the clasp domain has previously been observed in the IDHs of hyperthermophilic A. fulgidus and A. pernix, and is believed to stabilize the interface [22]. The aromatic cluster of the A. fulgidus IDH contains Phe179, which is substituted by the nonpolar residue Met in typical mesophilic IDHs, e.g., Met183 (AtIDH), Met183 (EcIDH) and Met172 (ZmIDH). Interestingly, MfIDH has a polar residue (Gly184) at the same position (Fig 2), which is typical for methylotrophic IDHs, e.g., M. glycogenes, M. glucosotrophus, M. mobilis, M. versatilis and others. The role of Gly in the clasp stabilization at elevated temperature requires further investigation.

Effects of metal ions on MfIDH activity
The effects of different cations on the MfIDH activity were studied, and the results indicate that MfIDH retaines ca. 18% of its activity even without the addition of divalent ions ( Table 2). This behavior is unusual for most IDHs, whose actvities entirely depend on the binding of a divalent cation [3]. Mn 2+ was found to be the ion that most effectively enhances the enzyme's activity, although Mg 2+ can act as a significant substitute by providing up to 58-75% of the enzyme's maximal activity. Whereas the MfIDH activity is completely inhibited by Ca 2+ and Cu 2+ , it is entirely restored by the addition of 2 mM of Mn 2+ (data not shown). Although most IDHs are strongly inhibited by Zn 2+ , we observed a very interesting effect that Zn 2+ addition has on MfIDH activity. NAD + -linked activity at pH 8.5 was strongly inhibited by Zn 2+ but partially restored at pH 6.0.
In contrast, the NADP + -linked activity at pH 8.5 was fully activated by the presence of Zn 2+ but decreased by half at pH 6.0, similar to the NAD + -linked activity. Thus, the pH optimum of   (Fig 2), there is no plausible explanation for these phenomena.

Effects of analogous cofactors on the MfIDH activity
The effects of different cofactor analogs on the MfIDH activity were examined (Table 3). More than half of the NAD + -or NADP + -linked activity was retained when the amide group of the nicotinamide ring was replaced by the acetyl group in 3-acetylpyridine adenine dinucleotide (phosphate); thus, the amide group is not indispensable for binding. In contrast, the substitution of the N 6 amino group of the adenine ring with the oxo-group in nicotinamide hypoxanthine dinucleotide (phosphate) completely abolished NAD(P) + binding. To our knowledge, this is the first report that demonstrates the indispensability of the amino group of adenine in cofactor recognition. Imada et al. thoroughly studied amino acid residues that are involved in the recognition of the adenine and nicotinamide rings of the cofactor [43]. The adenine N 6 atom is hydrogen-bonded with the carbonyl oxygen of Asn-348 and has amino-aromatic hydrogen-bond interactions with the imidazole ring of His-335, which are conserved interactions in the type I IDHs (Fig 2).

Substrate specificity and inhibition
No appreciable effect on the activity of MfIDH was observed upon addition of the following compounds (at final concentrations of 5 mM, unless noted) to the reaction mixture: glutamate, glutamine, α-ketoglutarate, oxaloacetate, cis-aconitate, citrate, pyruvate, malate, fumarate, succinate, ADP (2 mM), AMP (2 mM), CoA, AcCoA, NADH, and NADPH (at a final concentration of 0.2 mM). ATP (2 mM) caused 50% inhibition of only the NADP + -linked MfIDH activity. Thus, MfIDH activity is not regulated at the metabolic level, as has been demonstrated for IDHs from organisms with a complete TCA cycle.

Conclusions
The isocitrate dehydrogenase from M. flagellatus was purified, overexpressed and characterized in the present study. Our data reveal that MfIDH exhibits unique double coenzyme specificity toward both NAD + and NADP + cofactors, and its activity is dependent on divalent cations. MfIDH exhibits a strict cofactor-dependent pH-activity profile. Our study also shows that MfIDH is remarkably thermostable and is not regulated at the metabolic level. We suggest the major amino acids in the protein structure of MfIDH that determine the double cofactor specificity. The enzymatic characterization of MfIDH can enrich our knowledge of type I IDHs and might be useful for the engineering of IDHs with desirable specificities.
Supporting information S1 Fig. Isoelectric focusing of