Isocitrate Dehydrogenase from Streptococcus mutans: Biochemical Properties and Evaluation of a Putative Phosphorylation Site at Ser102

Isocitrate deyhdrogenase (IDH) is a reversible enzyme in the tricarboxylic acid cycle that catalyzes the NAD(P)+-dependent oxidative decarboxylation of isocitrate to α-ketoglutarate (αKG) and the NAD(P)H/CO2-dependent reductive carboxylation of αKG to isocitrate. The IDH gene from Streptococcus mutans was fused with the icd gene promoter from Escherichia coli to initiate its expression in the glutamate auxotrophic strain E. coli Δicd::kanr of which the icd gene has been replaced by kanamycin resistance gene. The expression of S. mutans IDH (SmIDH) may restore the wild-type phenotype of the icd-defective strain on minimal medium without glutamate. The molecular weight of SmIDH was estimated to be 70 kDa by gel filtration chromatography, suggesting a homodimeric structure. SmIDH was divalent cation-dependent and Mn2+ was found to be the most effective cation. The optimal pH of SmIDH was 7.8 and the maximum activity was around 45°C. SmIDH was completely NAD+ dependent and its apparent K m for NAD+ was 137 μM. In order to evaluate the role of the putative phosphorylation site at Ser102 in catalysis, two “stably phosphorylated” mutants were constructed by converting Ser102 into Glu102 or Asp102 in SmIDH to mimick a constitutively phosphorylated state. Meanwhile, the functional roles of another four amino acids (threonine, glycine, alanine and tyrosine) containing variant size of side chains were investigated. The replacement of Asp102 or Glu102 totally inactivated the enzyme, while the S102T, S102G, S102A and S102Y mutants decreased the affinity to isocitrate and only retained 16.0%, 2.8%, 3.3% and 1.1% of the original activity, respectively. These results reveal that Ser102 plays important role in substrate binding and is required for the enzyme function. Also, Ser102 in SmIDH is a potential phosphorylation site, indicating that the ancient NAD-dependent IDHs might be the underlying origin of “phosphorylation mechanism” used by their bacterial NADP-dependent homologs.


Introduction
Isocitrate deyhdrogenase (IDH) catalyzes the NAD(P) + -dependent oxidative decarboxylation of isocitrate to a-ketoglutarate (aKG) and the NAD(P)H/CO 2 -dependent reductive carboxylation of aKG to isocitrate using NAD + (EC 1.1.1.41) or NADP + (EC 1.1.1.42) as a cofactor. Phylogenetic analysis reveals that NAD + use by IDH is an ancestral phenotype and NADP + use by prokaryotic IDH arose on or about the time that eukaryotic mitochondria first appeared, some 3.5 billion years ago, in order to synthesize NADPH for bacterial adaptation on acetate [1]. As a better phenotype acquired through evolution, most IDHs exhibit the NADP + dependence, causing NAD + dependence comparatively uncommon among the bacterial IDHs. A small number of NAD + -IDHs have been characterized in some bacteria and archaea [2][3][4][5][6][7]. A general feature shared by these organisms is that they have an incomplete tricarboxylic acid (TCA) cycle due to the absence of one or more TCA cycle enzymes [3]. Thus, these NAD + -IDHs have been proposed to be reminiscent of the enzyme that participates in CO 2 fixation as well as glutamate biosynthesis [2,8]. Eukaryotic NAD + -IDHs localize exclusively in the mitochondria and play a central catabolic role in energy production. The enzyme is structural complex and is rate-limiting in the TCA cycle as its affinity for substrate is allosterically regulated by ADP [9].
Eukaryotes have several types of NADP + -dependent IDH isoenzymes, distributed in mitochondrial matrix, cell cytosol and peroxisome, respectively [10,11]. These enzymes share low sequence identity with prokaryotic counterparts, typified by NADP + -IDH from Escherichia coli (EcIDH). Eukaryotic NADP + -IDHs constitute a single clade in the phylogenetic tree, suggesting that they have evolved independently [5]. Mitochondrial isoenzymes provide auxiliary source of a-ketoglutarate and mitochondrial NADPH while the other two isoenzymes function in cellular defense against oxidative damage, detoxification of reactive oxygen species, and providing reducing power and carbon skeleton for fatty acids and amino acids biosynthesis [12][13][14][15]. Human cytosolic NADP + -IDH1 has recently been reported to be involved in tumorigenesis [16][17][18]. The IDH1 Arg132 mutation impairs the oxidative IDH activity of the enzyme, but acquires a new reduction function of converting a-ketoglutarate to 2-hydroxyglutarate, the resulting 2-hydroxyglutarate accumulation then induces the formation and malignant progression of tumors [19,20].
The prokaryotic NADP + -IDHs have been extensively studied. EcIDH lies at the critical juncture between TCA cycle and the glyoxylate bypass, a pathway needed for growth on nonfermentative carbon sources such as acetate and ethanol. Under these stressful conditions, ,75% of EcIDH is completely inactivated by phosphorylation at Ser113 catalyzed by the bifunctional IDH kinase/phosphatase (IDH K/P), thereby partitioning most of the isocitrate through the glyoxylate shunt [21]. Briefly, isocitrate is hydrogen-bonded to the c-hydroxyl of Ser113 in the active, dephosphorylated enzyme. The transfer of the c-phosphate from ATP to Ser113 prevents isocitrate binding by eliminating this hydrogen bond and by introducing a source of electrostatic repulsion and steric hindrance with the c-carboxylate of isocitrate [22][23][24][25]. This active serine in substrate binding is highly conserved in both prokaryotic and eukaryotic IDHs and the putative phosphorylation role of the equivalent serine in several NADP + -IDHs have been discussed [10,11,[26][27][28]. Although the NADP + -IDH of Bacillus subtilis (BsIDH) is not regulated by phosphorylation in vivo due to the absence of the gene encoding IDH K/P in this organism, it serves as the substrate for E. coli IDH K/P in vitro, and the phosphorylation does occur at the expected serine and inhibits IDH activity [27]. Despite the high conservation of the analogous serine in NAD + -IDHs, no studies have been reported yet to evaluate the possibility of the phosphorylation regulatory mechanism in these NAD + -IDHs.
The NAD + -dependent IDH of S. mutans (SmIDH) has a Ser102 at the position equivalent to Ser113 of EcIDH. In order to elucidate the function of Ser102, we generated two ''stably phosphorylated'' (pseudophosphorylated) mutants of SmIDH by converting Ser102 into glutamate or aspirate [29,30]. Meanwhile, another four mutations at site 102 were analyzed. These mutated enzymes were firstly screened by phenotypic complementation of icd deletion strain E. coli Dicd::kan r and then purified to homogeneity. The kinetic parameters of the wild-type together with the mutated enzymes were determined.

Strains, media and reagents
The E. coli strain Dicd::kan r was presented by Antony M. Dean's laboratory (BioTechnology Institute, University of Minnesota, MN 55108, USA), which was constructed by replacing the entire icd cistron of E. coli strain CGSC6300 with a kanamycin cassette [1]. In our study, this auxotrophic strain was used as the host strain for the expression of SmIDH and its mutants. LB and MD media were prepared and supplemented with 100 mg/ml ampicillin and/ or 30 mg/ml kanamycin as required. Plates contain 15 g/L agar. PrimeStar TM HS DNA polymerase was obtained from TaKaRa (Dalian, China). Restriction enzymes and protein molecular weight standards were purchased from Fermentas (Shanghai, China).

Plasmid construction
The 800-bp upstream region of the icd gene in the wildtype E. coli strain CGSC6300 was amplified with the following primers: S-Ec icdp , 59-atagatacctcgagCCATTGGCAAGATTATCCAAA-GAGT-39 (XhoI site (underlined letters) and additional bases are indicated by lowercase letters); R-Ec icdp , 59-CCCTTCTTCAAAACTTACTTTTTCTGCCATT-CACCTCTCCTTCGAGCGCTACTGGTTTGC-39, which contains the promoter sequence of E. coli icd gene. The S. mutans citC gene was amplified with the following primers: S-Sm icd , 59-GCAAACCAGTAGCG CTCGAAGGAGAGGTGAATGGCA-GAAAAAGTAAGTTTTGAAGAAGGG-39; R-Sm icd , 59-ta-tattctctgcagCTAGTGGTGGTGGTGGTGGTGTAAA-TAAGTCAATAGAAC-39 (PstI site (underlined letters) and additional bases are indicated by lowercase letters) using plasmid pKM49 [31] as template. The 39 end of R-Ec icdp and the 59 end of S-Sm icd were designed to be reverse and complementary to each other so that the two PCR products can be fused together by overlap extension using S-Ec icdp and R-Sm icd as primers. The resulting 2.0-kb PCR product containing S. mutans citC gene preceded by the promoter of E. coli icd gene was then ligated into XhoIand PstI-digested pSP72 (Promega) to create pWT.

Site-directed mutagenesis
All mutants (S102T, S102G, S102A, S102Y, S102D and S102E) were constructed by site-directed mutagenesis from pWT. The synthetic oligonucleotide primers are shown in Table 1. The mutations were introduced by sequential steps of PCR. In the first round, two reactions (I and II) were performed with the following primers: S-Ec icdp and one of the antisense primers containing the desired mutation (reaction I); one of the sense primers containing the desired mutation and R-Sm icd (reaction II). The purified two overlapping fragments were used as templates in the final amplification step with primers S-Ec icdp and R-Sm icd . All the final PCR products were cloned into pSP72 to obtain plasmids pS102T, pS102G, pS102A, pS102Y, pS102D and  pS102E. All mutated genes were confirmed by sequencing in both directions.

Protein expression and purification
All recombinant plasmids were transformed into E. coli strain Dicd::kan r , respectively. The resulting strains were grown overnight at 37uC in LB medium with 100 mg/ml ampicillin, and then inoculated (1:50) into 500 ml of MD medium with the same antibiotic and grown for two days. The cells were harvested by centrifugation, resuspended in wash buffer (10 mM KH 2 PO 4 (pH 7.7), 500 mM NaCl, 2 mM MgCl 2 and 2 mM b-mercaptoethanol) and disrupted by sonication. Then, cell debris was removed by centrifugation at 11,000 rpm for 10 min. The 6Histagged wild-type SmIDH and its mutated enzymes were purified using BD TALON Metal Affinity Resin (Clontech, LaJolla, CA, USA) according to the manufacturer's instructions.

Enzyme purity and western blotting
Enzyme purity was determined by SDS-PAGE. For western blotting analysis, SDS-PAGE gels were transferred to nitrocellu- lose membranes by electroblotting and blocked for 1 h at room temperature in TBS-T (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.2% Tween-20) containing 5% nonfat milk and then washed with TBS-T for three times. His-tag polyclonal antibody (Cell Signaling Technology Inc., Beverly, MA, USA) was applied to the blots for 1 h at room temperature. After three 10-min washes with TBS-T, the blots were incubated for 1 h with alkaline phosphatase conjugated anti-rabbit IgG (Promega, Madison, WI, USA ). The blots were washed three times in TBS-T, and bound conjugate was revealed by incubation with the alkaline phosphatase substrate. The chemiluminescence signal corresponding to the specific antibody-antigen reaction was visualized by exposing the blots to X-ray film for 15 minutes in the dark room.

Gel filtration chromatography
The molecular mass of SmIDH was estimated by gel filtration chromatography on a HiLoad TM  Circular dichroism spectroscopy of the wild-type and mutant enzymes Circular dichroism (CD) spectroscopy was conducted using a Jasco model J-810 spectropolarimeter. The ellipticity measurements as a function of wavelength were performed as described previously [32]. Purified protein samples (0.3 mg/ml) were prepared in 50 mM sodium phosphate and 60 mM NaCl

Enzyme assays and kinetic studies
The enzyme activity was assayed by a modification of the method described previously [33]. Reaction mixtures were incubated at 37uC in 1 ml volume containing 35 mM Tris-HCl buffer (pH 7.5), 3.5 mM MnCl2, 2.5 mM DL-isocitrate, 0.5 mM NAD + or 5 mM NADP + . The increase in NADPH or NADH was monitored at 340 nm with a thermostated Cary 300 UV-Vis spectrophotometer (Varian, CA, USA) using a molar extinction coefficient of 6220 M 21 cm 21 . One unit (U) of activity was defined as 1 mmol NADPH or NADH formed per minute. The apparent kinetic parameters were calculated by nonlinear regression using the program Prism 5.0 (Prism, GraphPad Software, CA, USA). All kinetic parameters were obtained from at least three measurements. The concentrations of the purified enzymes were estimated by absorbance measurements at 280 nm, using an extinction coefficient (e 280 = 51,255) calculated by the method of Pace et al. [34].

Effects of pH and temperature
The enzyme was assayed in 35 mM Tris-HCl buffer between pH 7.2 and 9.2. The optimum temperature was determined by the standard activity assay at different temperatures from 25uC to 55uC. To estimate thermal stability, enzymes were incubated for 20 min between 25 and 55uC in a water bath. Aliquots were withdrawn at periodic intervals, cooled in an ice bath and assayed as described above.

Expression and purification of recombinant SmIDH
As the recombinant SmIDH can not be produced in E. coli under the control of the lac promoter, a unique IDH expression method was applied in this study. The promoter sequence of E. coli icd gene was firstly fused to the N-terminus of S. mutans citC gene and then subcloned together into the vector pSP72, creating pWT. The recombinant plasmid pWT was transformed into the glutamate auxotrophic E. coli strain Dicd::kan r (icd-defective strain). The expression of S. mutans citC gene was examined by evaluating the strain growth on minimal medium containing glucose as the sole carbon source without glutamate. The transformants that grew well on the minimal medium were isolated, indicating that S. mutans citC gene can be expressed normally and function well by restoring the growth of E. coli strain Dicd::kan r .
SmIDH was produced as a 6-histidine fusion protein that had a calculated molecular mass of 43 kDa. A single band at around 43 kDa was visible on the SDS-PAGE gel ( Figure 1A), which was confirmed by Western blotting using anti-His antibody ( Figure 1B).
Gel filtration chromatography was performed to determinate the oligomerization status of SmIDH in solution. The native molecular mass of SmIDH estimated by gel filtration was 70 kDa (Figure 1C), suggesting a homodimeric structure similar to E. coli IDH, B. subtilis IDH and A. thiooxidans IDH [8,26,37].

Characterization of the enzymatic properties of SmIDH
The effects of pH on SmIDH activity were performed in the pH range of 7.2-9.2. SmIDH was found to have an optimal pH range of 7.5-8.5, with the optimal pH at 7.8 (Figure 2A), lower than that of the NAD + -IDHs from A. thiooxidans (pH 8.5) [2] and Hydrogenobacter thermophilus (pH 10.5) [38]. When compared with the broad optimum pH range of NADP + -IDHs from other sources, such as Streptomyces lividans (pH 8.5-10.0) [39], Fomitopsis palustris (pH 8.0-10.0) [40] and Aspergillus niger (pH 6.0-8.0) [41], it was narrower for SmIDH, suggesting that SmIDH was sensitive to pH changes.
SmIDH had a temperature optimum for activity at 40-53uC and its maximum activity was observed at 45uC ( Figure 2B). Although this mesophilic enzyme is stable at room temperature, 50% loss of activity and almost complete inactivation were observed after incubation at 43uC and 50uC for 20 min, respectively ( Figure 2C). This data was quite different from some known NAD + -IDHs while most of them were found to be thermostable, such as A. thiooxidans IDH (stable up to 55uC) [2], Methylococcus capsulatus IDH (optimum for activity at 55-60uC) [4], H. thermophilus IDH (half-inactivation at 88.7uC) [3] and most distinguished P. furiosus IDH (with a melting temperature of 103.7uC) [5]. The significant difference in thermostability among NAD + -IDHs could be the temperature adaptation of enzymes to their niches.
Kinetic studies revealed that the apparent K m of SmIDH displayed for NAD + and isocitrate was 154 mM and 75 mM, respectively ( Figure 3). The K m value of SmIDH for NAD + was higher than those determined for P. furiosus IDH (68.3 mM) and M. capsulatusbut IDH (122 mM), but lower than that observed for NAD + -IDHs from A. thiooxidans (180 mM), Streptococcus suis (233 mM), Zymomonas mobilis (312 mM) and H. thermophilus IDH (357 mM) [2][3][4][5][6][7]. SmIDH was completely NAD + dependent as shown and no enzyme activity was observed for SmIDH when NAD + was substituted by NADP + in concentrations up to 5 mM. The cofactor discrimination of IDH was determined by only a few amino acids [42]. In some NADP + -IDHs, such as EcIDH and BsIDH, Lys 344/350 and Tyr 345/351 are the major NADP +specificity determinants (Figure 4), which are substituted by the conserved Asp and Ile in all known NAD + -IDHs such as Asp 322/ 357 and Ile 323/358 in SmIDH and AtIDH (Figure 4). These major specificity determinants can be used as reliable landmarks to predict the coenzyme specificity of new IDHs. The effects of nine different metal ions on SmIDH activity were examined (Table 2). SmIDH activity was entirely dependent on the presence of a divalent cation, and Mn 2+ was found to be the most favored one although Mg 2+ can partially replace it. In the presence of Mn 2+ , the addition of 2 mM Co 2+ , Cu 2+ , Zn 2+ and Ni 2+ reduced SmIDH activity to about 11%, 12%, 2% and 1% of control value, respectively (Table 2). However, the dramatic inhibition of Ca 2+ on SmIDH activity was not observed in our study. This was quite different from some reports on NADP + -IDHs, such as E. coli IDH and S. lividans IDH, that Ca 2+ may reduce the activity at a large scale [39,43]. Neither Na + nor K + affected SmIDH activity in the presence of Mn 2+ .

Performances of Ser102 mutants
IDH is the first bacterial enzyme shown to be regulated by phosphorylation/dephosphorylation [44]. The modulation of IDH activity enables E. coli to make rapid shifts between TCA and glyoxalate bypass pathways, whereas the phosphorylation state of IDH determines its activity [45][46][47]. Phosphorylation on a serine residue of EcIDH by IDH K/P inactivates the enzyme by preventing NADP + binding, and dephosphorylation reactivates it [44,46,48]. Although an analogous serine has been found to be highly conserved in all known NAD + -IDHs, which is corresponding to the phosphorylation site of Ser113 in EcIDH, there were no reports on evaluating the role of the analogous serine in the regulatory mechanism of NAD + -IDH activity yet.
A putative phosphorylation site Ser102 was present in the NAD + -dependent SmIDH. Given the highly structural similarity of SmIDH to EcIDH (Figure 4), the same phosphorylation regulatory mechanism might be used by SmIDH, even though phosphorylation in vivo has not been found yet. In this study, we mimicked the phosphorylation state of Ser102 by mutating Ser102 to Asp or Glu (S102D or S102E). The icd-defective strain E. coli Dicd::kan r harboring the recombinant plasmid pS102D or pS102E did not show any growth on MD medium without glutamate, indicating that the mutations of S102D or S102E caused total loss of SmIDH activity. Similar results were reported by Thorsness et al. [24] and Matsuno et al. [31], and they found that both EcIDH and BsIDH were nearly abolished by replacing Ser with Asp in the active site. Given the possibility that a small amount of activity can be retained by S102D and S102E mutants, as observed in S113D and S113E mutants of EcIDH in the previous studies [22,23,25,49], the residual activities of the mutants were so low that the icd-defective E. coli was not able to survive on MD plates without glutamate.
Phylogenetic analysis reveals that NAD + use by IDH is an ancestral phenotype and NADP + use by prokaryotic IDH arose on or about 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 anabolic demand for NADPH during growth on acetate [1]. The anaerobic, Gram-positive bacterium S. mutans has an IDH with ancient NAD + -dependency, suggesting that S. mutans might be an ancient prokaryote and not be selected by poor carbon sources (i.e. two carbon compounds) through its evolutionary history. Mimicking phosphorylation by replacing Ser102 with Asp or Glu inactivates SmIDH, implying that Ser102 is a potential phosphorylation site and the ancient NAD-dependent IDHs might be the underlying origin of ''phosphorylation mechanism'' used by their bacterial NADP-dependent homologs.
The effects of amino acids with different size of side chains, such as Thr, Gly, Ala and Tyr, at the site 102 were also investigated. E. coli Dicd::kan r strains containing pS102T, pS102G, pS102A and pS102Y showed different growth performances in minimal media, respectively, and the four mutant enzymes were then purified to homogeneity. CD spectra of the wild-type and mutant SmIDH were measured to evaluate whether there is any change in secondary structure of these mutant enzymes. The CD spectra of S102T, S102G, S102A, and S102Y mutants are very similar to that of the wild-type enzyme ( Figure 5), suggesting that the mutations did not cause any appreciable conformational change. Four mutants decreased the affinity to isocitrate with 2-to 5-fold K m values of the wild-type enzyme. Consequently, the catalytic efficiency (k cat /K m ) of them was remarkably reduced about 10-to 412-fold, and S102T, S102G, S102A and S102Y only retained 16.0%, 2.8%, 3.3% and 1.1% of the original activity, respectively ( Figure 6). In S102T mutant enzyme, the Thr residue has a chydroxyl side chain as Ser that can bind isocitrate by a hydrogenbond and thus making S102T retain a relative high activity. The activities of S102Y, S102G and S102A were mainly depended on the size of the side chain of replaced residues, as the bulky side chain in the active site is sterically unfavorable for isocitrate and cofactor binding [22,25]. For example, the replaced Tyr in S102Y, an amino acid with large side chain, caused almost 99% activity loss of SmIDH ( Figure 6). The dramatic loss of S102Y activity was mainly caused by its decreased substrate-and cofactor-binding abilities. As shown in Table 3, the K m values for isocitrate and NAD + of S102Y were increased 5.1-fold (from 75 to 385 mM) and 10.1-fold (from 154 to 1560 mM), respectively.
Taken together, six mutations were introduced into SmIDH at Ser102 respectively. Two mutant enzymes lost the activity, and the other four mutants significantly decreased the activity. Apparently, SmIDH activity is sensitive to the substitution at position 102, the conserved Ser102 plays important role in substrate binding and is required for maintaining the proper structure of the active site and responsible for the enzyme function. It implies that SmIDH may have the similar quaternary structure and catalytic mechanism to the well-known NADP-IDHs, such as EcIDH and BsIDH. In addition, the completely conservation of the so-called ''phosphorylation loop'' and the missing of the ''insert-region'' (Figure 4), which restricts the access of E. coli IDH K/P to the phosphorylation site in BsIDH [26], suggest that SmIDH could be a better substrate for E. coli IDH K/P than BsIDH.