Purification and Characterization of Fumarase from Corynebacterium

Biosci. Biotechnol. Biochem., 70 (5), 1102–1109, 2006

Purification and Characterization of Fumarase from Corynebacterium glutamicum

y Tomoko GENDA,1 Shoji WATABE,2 and Hachiro OZAKI1;

1Biological Institute, Faculty of Education, Yamaguchi University, 1677-1 Yoshida, Yamaguchi 753-8513, Japan 2Basic Laboratory Science, Faculty of Health Sciences, Yamaguchi University School of Medicine, 1-1-1 Minami-Kogushi, Ube 755-8505, Japan

Received August 2, 2005; Accepted December 10, 2005

Fumarase (EC 4.2.1.2) from Corynebacterium gluta- fumarases.9–12) micum (Brevibacterium flavum) ATCC 14067 was puri- Corynebacterium glutamicum is an industrially im- fied to homogeneity. Its amino-terminal sequence (res- portant microorganism widely used in the production idues 1 to 30) corresponded to the sequence (residues 6 of various amino acids. TCA-cycle members of the to 35) of the deduced product of the fumarase gene of enzymes of this bacterium, citrate synthase (EC C. glutamicum (GenBank accession no. BAB98403). The 4.1.3.7),13) isocitrate dehydrogenase (EC 1.1.1.42),14,15) molecular mass of the native enzyme was 200 kDa. The 2-oxoglutarate dehydrogenase (EC 1.2.4.2),16) malate protein was a homotetramer, with a 50-kDa subunit dehydrogenase (EC 1.1.1.37),17–19) and a membrane- molecular mass. The homotetrameric and stable prop- associated malate:quinone oxidoreductase (MQO) (EC erties indicated that the enzyme belongs to a family of 1.1.99.16)18,19) have been studied with respect to Class II fumarase. Equilibrium constants (Keq) for the metabolic regulation. As for fumarase of the bacterium, enzyme reaction were determined at pH 6.0, 7.0, and its amino-acid sequence deduced from the gene has been 8.0, resulting in Keq ¼ 6:4, 6.1, and 4.6 respectively in reported (GenBank accession no. BAB98403). Hence, phosphate buffer and in 16, 19, and 17 in non-phosphate we investigated the biochemical properties of fumarase buffers. Among the amino acids and nucleotides tested, of this bacterium. ATP inhibited the enzyme competitively, or in mixed- Recent studies on fumarase have been concentrated type, depending on the buffer. Substrate analogs, meso- upon unraveling the reaction mechanisms in the three- tartrate, D-tartrate, and pyromellitate, inhibited the dimensional structures of the active center using com- enzyme competitively, and D-malate in mixed-type. petitive inhibitors for fumarases of pig heart,20) E. coli,21) and yeast.22) This paper deals with purification Key words: fumarase; Corynebacterium glutamicum; and characterization of fumarase from C. glutamicum, Brevibacterium flavum; equilibrium con- including inhibition by substrate-analogs. stant; homotetramer Materials and Methods Fumarase or fumarate hydratase (EC 4.2.1.2), which catalyzes the reversible hydration of fumarate to L- Materials. Sodium fumarate, sodium L-malate, so- malate, is an integral part of the tricarboxylic acid dium L-tartrate, D-tartaric acid, meso-tartaric acid, urea, (TCA) cycle. Most studies have been concerned with DTT, and DTNB were products of Wako Pure Chemical mammalian fumarases.1–8) Fumarases from microorgan- Industries (Osaka, Japan). D-Malic acid was purchased isms have been studied less extensively. In Escherichia from Sigma (St. Louis, MO). L-Amino acids were coli, three kinds of fumarases, fumarase A (FumA), B products of Ajinomoto (Tokyo). Toyopearl-Phenyl- (FumB), and C (FumC), have been found.9) FumA and 650M was a product of Tosoh (Tokyo). Pyromellitic FumB are fumarases of a class I form, which are acid was a product of Tokyo Kasei Chemical (Tokyo). homodimeric, labile, and iron-dependent enzymes hav- All other chemicals used were the same as reported in ing a molecular mass of 120 kDa. FumC is of a class II the previous paper.17) form, which is a homotetrameric, stable, and iron- independent enzyme having a molecular mass of Microorganism and culture. A wild-type strain of 200 kDa. The biochemical properties of FumC have C. glutamicum ATCC14067 (Brevibacterium flavum been reported to be homologous to the mammalian no. 2247) was used. The nutrient agar medium (CM-2)

y To whom correspondence should be addressed. Fax: +81-83-933-5357; E-mail: [email protected] or [email protected] Abbreviations: DTT, 1,4-dithiothreitol; DTNB, 5,50-dithiobis(2-dinitrobenzoic acid); MES, 2-(N-morpholino)ethanesulfonic acid; TES, N- tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid Fumarase from C. glutamicum 1103 was composed, in g/l, of Polypepton, 10; yeast extract, of fumarase preparation properly diluted with 10 mM of 10; NaCl, 5; and agar, 20 (pH 7.0), and was sterilized at the buffer in the total volume of 1 ml, was used. The 120 C for 20 min. The medium for seed and main reaction was started by adding the enzyme. The initial culture (G30) was composed, in weight/l, of glucose, absorbance change at 250 nm was measured on a 36 g; urea, 10 g; KH2PO4, 1 g; MgSO4.4H2O, 0.4 g; spectrophotometer (Ultrospec 3000, Amersham Phar- FeSO4.7H2O, 10 mg; MnSO4.4H2O, 8.1 mg; thiamine. macia Biotech, Cambridge UK) at room temperature HCl, 100 mg; d-biotin, 30 mg; and 6 N HCl, 7 ml. For the (20–23 C). The activities were expressed in units (mmol 1 seed and main cultures, 50 ml and 3 liters of medium G30 fumarate formed min ) using an extinction coefficient were autoclaved in a 500-ml flask and a 5-liter jar of 1.45 mM1 cm1 at 250 nm.1) The reverse reaction respectively at 115 C for 10 min (final pH, 7.0). A was assayed in the same way except that 10 mM sodium loopful of cells of the bacterium grown on CM-2 at fumarate was used as substrate, and the absorbance 30 C for 24 h was inoculated into the flask and cultured change at 290 nm was measured. The activities were with shaking at 120 rpm at 30 C for 24 h as seed culture. calculated using an extinction coefficient of 0.109 The broth of one flask of the seed culture was inoculated mM 1 cm 1 at 290 nm.1) Protein was measured by the into the main-culture jar with 200 rpm of agitation and method of Lowry et al.,23) with bovine serum albumin as 1 liter/liter/min of aeration at 30 C for 20–24 h on a jar the standard. fermentor (model MD-300, Marubishi, Tokyo). The cells were collected by centrifugation, washed with Gel filtration. One ml of 50 mM Tris–HCl buffer, 0.2% of KCl, and stored at 20 C. pH 7.5, containing 3 mg of thyroglobulin (bovine, 670 kDa), 1 mg of urease (soybean, 480–490 kDa), 2 mg of Purification of fumarase. The washed cells (20 g wet catalase (bovine, 248 kDa), 0.5 mg of alcohol dehydro- weight) were suspended in 30 ml of 50 mM sodium genase (yeast, 150 kDa), 0.5 mg of peroxidase (horse phosphate buffer, pH 7.0 (buffer A) and disrupted in radish, 40 kDa), and 20 ml of purified fumarase was gel- a sonic oscillator (Kubota 201, 9 kHz, Kubota Shoji, filtered through a Toyopearl HW-60 column (1:9 Tokyo) under cooling (below 6 C) for 20 min. The 40 cm) using the same buffer. sonic extracts from 160 g of the cells were pooled and centrifuged at 100;000 g for 40 min. The supernatant Native polyacrylamide gel electrophoresis (PAGE). was treated with 2.0% (w/v) streptomycin sulfate for Native PAGE was done in 7.5% gel using the ATTO- 60 min, and the precipitate was removed by centrifuga- Compact PAGE system (Atto, Tokyo). The gel was tion (30;000 g for 20 min). The resulting supernatant stained with Coomassie Brilliant Blue R-250. was fractionated by ammonium sulfate precipitation and centrifugation. The pellet obtained between 40% and SDS–PAGE. SDS–PAGE was done in 11% gel using 60% was dissolved in buffer A and dialyzed against the ATTO-Compact PAGE system and a kit for mo- 10 mM Tris–acetate buffer, pH 7.3 (buffer B). The lecular weight purchased from Sigma (St. Louis, MO). dialyzed sample was chromatographed using a DEAE- Toyopearl 650M column (2:64 15 cm) and 500 ml of Amino-terminal sequence analysis. The amino-termi- buffer B containing 0.1 M NaCl and 500 ml of buffer B nal amino-acid sequence of the purified enzyme was containing 0.4 M NaCl. The fractions showing fumarase determined by use of a protein sequencer (ABI, Model activity were pooled and dialyzed against buffer B. The 476A). dialyzed sample was rechromatographed in the same way except for the use of a smaller column (1:9 Results and Discussion 11 cm) and 100 ml each of the buffer. The fumarase fractions obtained on the second chromatography were Purification of fumarase pooled, added to solid ammonium sulfate to give a final Fumarase was purified from strain no. 2247 by concentration of 2.0 M, and applied to a Phenyl-Toyo- ammonium sulfate precipitation, DEAE-Toyopearl, and pearl 650M column (1:5 10 cm) equilibrated with Phenyl-Toyopearl column chromatographies. A summa- 10 mM potassium phosphate buffer, pH 7.3 (buffer C), ry of the purification of the enzyme is shown in Table 1. containing 2 M ammonium sulfate. After the column was Fumarase was purified 252-fold, with a 12% yield. washed with the buffer, chromatography was performed PAGE of the enzyme preparation showed a single with a linear concentration gradient of ammonium protein band, as shown in Fig. 1B, indicating that the sulfate, 2 to 0 M in buffer C (100 ml each). Fumarase enzyme was electrophoretically homogeneous. fraction was rechromatographed in the same way except that the gradient elution was done with 1.0 to 0 M of Molecular mass, subunit composition, and amino- ammonium sulfate. terminal sequence of fumarase The molecular mass of native fumarase was 200 kDa Enzyme assay. The standard incubation mixture for by gel filtration, as shown in Fig. 2. When the protein fumarase activity, consisting of 100 mM sodium phos- was run on denaturing SDS–PAGE, it appeared as a phate buffer, pH 7.3, 50 mM sodium L-malate, and 20 ml single band of 50 kDa, as shown in Fig. 1A and C. Thus 1104 T. GENDA et al. Table 1. Purification of Fumarase from C. glutamicum

Activity Total volume Total protein Yield Purification Purification step (ml) (mg) Total Specific (%) fold (U) (U/mg) Crude extract 390 3342 4173 1.25 100 1 Streptomycin treatment 380 3244 3610 1.11 86 0.9 Ammonium sulfate 24 470 2698 5.74 64 4.5 1st DEAE-Toyopearl 65 65 2600 40.2 62 32 2nd DEAE-Toyopearl 22 11 1200 109 28 87 1st Phenyl-Toyopearl 10 4 932 233 22 186 2nd Phenyl-Toyopearl 4 1.6 504 315 12 252

Fig. 1. SDS–PAGE (A, C) and Native PAGE (B) of Purified C. glutamicum Fumarase. Molecular weight markers: 1, albumin, bovine (66 kDa); 2, albumin, egg (45 kDa); 3, glyceraldehyde-3-phosphate dehydrogen- Fig. 2. Gel Filtration of Purified C. glutamicum Fumarase through a ase (36 kDa); 4, carbonic anhydrase (29 kDa); 5, trypsinogen (24 Toyopearl HW-60 Column. kDa); 6, trypsin inhibitor (20.1 kDa); 7, -lactalbumin (14.2 kDa). Molecular weight markers: 1, thiogloblin; 2, urease; 3, catalase; The arrow indicates R for fumarase. f 4, alcohol dehydrogenase; 5, peroxidase. The arrow indicates the elution volume for fumarase. the enzyme appears to be a homotetramer. This observation was confirmed by determination of the amino-terminal sequence of the enzyme, which featured marase closely resembles fumarases purified from a single amino-acid stretch. The first 30 amino acids of mammalian sources,7) Saccharomyces cerevisiae,24) the amino-terminal sequence were TEQEFRIEHD TM- E. coli (FumC),10,11) Bacillus subtilis,25) Sulfolobus sol- GEVKVPAK ALWQAQTQRA. This sequence corre- fataricus,26) and Thermus thermophilus.27) These fumar- sponds to that (residues 6 to 35) deduced from the ases are grouped into Class II, being distinct from those nucleotide sequence of fumarase gene of C. glutamicum of Class I, which are characteristically labile, dimeric (GenBank accession no. BAB98403). The gap of five enzymes, represented by FumA and FumB.9–11) amino acids appears to have been occurred due to post- translational processing. It has been reported that E. coli Molecular stability and reaction with thiol reagents FumC is not processed.10–12) The fumarase of C. glutamicum has two cysteine In contrast to the presence of three fumarase genes, residues in each subunit (GenBank accession no. fumA, fumB, and fumC,inE. coli,9) one gene for BAB98403), that is, eight cysteine residues in the whole fumarase has been reported in C. glutamicum (GenBank molecule. First the purified enzyme was preliminarily accession nos. NC 003450 and BX927147), and no other incubated in 100 mM potassium phosphate buffer, genes or proteins homologous to the fumarase gene were pH 7.0 and pH 7.6, containing 3 mM DTNB, under the obtained in database searches. Since no extra-fumarase same conditions as described in legend to Fig. 3. No fraction was obtained in course of purification of the inactivation was observed at pH 7.0 (data not shown), enzyme, the purified enzyme in this study must be a and slight inactivation at pH 7.6, as shown in Fig. 3B. In product of the fumarase gene. this experiment, the enzyme was incubated in two pHs, With respect to homotetrameric structure and stability 7.6 and 8.5, for the sake of mixed-disulfide formation against oxygen (described below), C. glutamicum fu- with DTNB and the thiol groups of the enzyme protein. Fumarase from C. glutamicum 1105 remain intact in its protein conformation in neutral pH and the thiol groups are protected from oxygen. These stability characteristics of C. glutamicum fumarase are nearly the same as those of fumarase from pig heart5) and the FumC of E. coli,9–11) but are different from P. aeruginosa FumC.29) Teipel and Hill have reported that fumarase of pig heart dissociated into subunits with loss of activity at low concentrations of the enzyme, below pH 6, above pH 10, and in solutions of guanidine hydrochloride.30) A similar conformational change must have occurred to C. glutamicum fumarase in the experi- ments summarized in Fig. 3.

Substrate specificity and optimum pH The enzyme showed absolute specificity for the substrates fumarate and L-malate; no activity was observed in 10 or 100 mM Tris–acetate buffer, pH 7.3, with 1–50 mM of the compounds used for the inhibition test shown in Table 3. Pig-heart fumarase, however, catalyzes dehydration of L-tartrate.6,31) Pig-heart fumarase has been observed to be influenced by phosphate ions: the ions show activation in dilute Fig. 3. Stability Test of Fumarase in the Presence of Various concentrations and also inhibition at high concentra- 32) Compounds. tions. The optimum pH was determined using phos- The purified enzyme preparation (4 mg) was incubated in 100 mM phate buffers and non-phosphate buffers to detect the potassium phosphate buffer, pH 7.6 (A, B, C) or 100 mM Tris–HCl phosphate-ion effect on C. glutamicum fumarase. Figure buffer, pH 8.5 (D, E, F) (total volume, 1 ml) in the presence of 1 M 4A and B shows plots of enzyme activity vs. pH with ( ), 2 M ( ), or 4 M ( ) of urea; 1 mM ( )or3mM ( ) of DTNB; fumarate and L-malate respectively as substrate. The 2 M urea plus 2 mM DTT ( ); 2 mM DTT ( ); and 1 M urea plus 2mM DTT ( ) at room temperature (23–25 C) for the indicated presence of phosphate ions did not influence the periods. The control incubation (no addition) ( ) was done for all fumarase activity, as shown in Fig. 4B. The results experiments except C. At the indicated times, 0.02 ml of the with phosphate buffer showed optimum pHs of 6.5 and incubation mixture was withdrawn and the residual activity was 8.25 for fumarate and L-malate respectively, and those assayed. with Tris–HCl buffer showed optimum pHs of 7.5 and 8.5 respectively. The values of the optimum pH with phosphate buffer are nearly equal to those with the The enzyme preparation was stable at room temperature buffer of pig-heart fumarase2,3) and E. coli FumC.11) The at pH 7.6, as shown in Fig. 3A and B for the control difference in the optimum pH observed between the two experiment. But the enzyme at a lower concentration buffers (Fig. 4A) leads to the consideration which buffer became labile even in a buffer of neutral pH (data not is more physiological for the bacterium. shown). Stability tests were carried out in the presence of several chemicals at pH 7.6 (Fig. 3A–C) and pH 8.5 Equilibrium constant (Fig. 3D–F). DTT showed partial restoration from urea In a reversible one-substrate system catalyzed by an and alkaline inactivation, as shown in Fig. 3C and F. enzyme, the equilibrium constant (Keq) is given by DTNB showed slight inactivation at pH 7.6 (Fig. 3B), equation (1), which is known as the Haldane relation- and acceleration of alkaline inactivation (Fig. 3E). ship.33) These results suggest that active fumarase has thiol K ¼ V K =V K ð1Þ groups. eq F mM M mF The enzyme was incubated with 1 mM (as final where VF, VM, KmF, KmM are the maximum velocities concentration) of iodoacetate, iodoacetamide, N-ethyl- and Michaelis constants for fumarate and L-malate maleimide, and p-chloromercuribenzoate (0.5 mM)in respectively. 100 mM buffer B, but no inactivation was observed with In order to obtain the equilibrium constant (Keq) at the these thiol reagents. The above results indicate that the physiological pH and at the same pHs as reported by thiol groups are buried in the interior of the enzyme. Hill Alberty et al.,3) and to examine the influences of 28) and Teipel have reported that the thiol groups of phosphate ions on Keq, the Lineweaver-Burk plots for fumarase of pig heart might reside in or near the contact the two substrates by C. glutamicum fumarase were regions between subunits, and the location of the thiol carried out at pH 6.0, 7.0, and 8.0 using 100 mM groups on the enzyme molecule has been illustrated by phosphate buffer and 100 mM non-phosphate buffers. 20) Beeckmans and Driessche. The enzyme appears to All the plots were linear. The maximum velocities (VF, 1106 T. GENDA et al.

Fig. 4. Effect of pH on Fumarase Activity. The reaction mixtures were same as the standard assay mixture except that sodium phosphate buffer, pH 5.0–8.5 (100 mM for all); MES– NaOH buffer, pH 5.0–6.5; TES–NaOH buffer, pH 6.5–7.2; Tris–HCl buffer, pH 7.2–9.5; and glycine–NaOH buffer, pH 8.7–10.0 were used. , phosphate buffer; , non-phosphate buffers. Fumarate (A) and L-malate (B) were used as substrates, and 0.4 mg of fumarase was added to each.

Table 2. Catalytic Center Activities (kcat), Michaelis Constants (Km), and Equilibrium Constants (Keq) of Fumarase from C. glutamicum at Various pHs

Fumarate L-Malate Substrate Keq k K k K pH Buffer cat m cat m (s1) 102 (M) 103 (s1) 102 (M) 103 6.0 Phosphate 7.2 4.2 0.48 1.8 6.4 MES–NaOH 3.1 0.81 0.34 1.4 16 7.0 Phosphate 7.3 3.8 2.0 6.3 6.1 TES–NaOH 4.3 0.38 1.1 1.8 19 8.0 Phosphate 5.1 3.1 4.6 13 4.6 Tris–HCl 6.5 0.67 2.9 5.0 17

VM) and Michaelis constants (Km) for both substrates Inhibition were obtained from the plots, and the values of Keq were Effects on the enzyme activity of amino acids and calculated using equation (1). The results are shown in organic acids related to the TCA cycle were investigated Table 2. The values of Keq at the three pHs in the in 100 mM Tris–acetate buffer, pH 7.3, containing phosphate buffer were 6.4, 6.1, and 4.6 respectively, and 50 mML-malate. L-Aspartate (20 to 500 mM), L-gluta- the values in the non-phosphate buffers were 16, 19, and mate (20 to 500 mM), and organic acids (20 to 200 mM 17 respectively, indicating that phosphate ions influence citrate, 2-oxoglutarate, and succinate) did not affect the Keq mainly due to a large influence on the Km values, as enzyme activity significantly. The effect of inorganic shown in Table 2. The values of Km for both substrates compounds was also examined, using 100 to 500 mM and those of Keq obtained in phosphate buffer in this NaCl, KCl, and (NH4)2SO4, and 0.1 to 2 mM MnCl2 and study (Table 2) are comparable to those of fumarase MgCl2 in the same reaction system, resulting in no from pig heart (in 133 mM sodium phosphate buffers), as significant effect on the enzyme, but 2 mM MgCl2 had a 3) reported by Alberty et al., and Keq ¼ 4:3 (at pH 7.5) of small stimulatory effect on the activity in buffer B (data 24) fumarase from S. cerevisiae. As for kcat values, not shown). Alberty32) has reported that the turnover numbers for Among the nucleotides tested in buffer B, only ATP the mammalian enzyme were 2 103 and 2 102 per s inhibited the enzyme. The effect of ATP concentration per mol of enzyme for fumarate and L-malate respec- on enzyme activity was examined in the presence and 2 tively, corresponding to kcat values of 5 10 and 0:5 absence of 2 mM MgCl2 in buffer B. A slight stimulation 102 respectively. These values are nearly of the same of ATP inhibition was observed, as shown in Fig. 5. The order of magnitude as those obtained in this study same experiment was done in 100 mM TES–NaOH (Table 2). The nearly constant Keq values in a broad buffer, pH 7.0 (Fig. 5), but no stimulatory effect by range of pH indicate that fumarase would work well 2mM MgCl2 was observed (data not shown). In contrast without any affect by pH change for this bacterium. to this result, ATP inhibitions for yeast and pig-heart Fumarase from C. glutamicum 1107

Fig. 5. Effect of ATP on Fumarase Activity. Purified fumarase (0.4 mg) was incubated in 100 mM TES–NaOH buffer, pH 7.0 ( ), or in buffer B (total volume, 1 ml) containing 1mM fumarate, and the indicated concentrations of ATP in the presence ( ) and absence ( )of2mM MgCl2. fumarases were rather diminished by the presence of Mg2þ.34) The types of inhibition in buffer B containing 2mM MgCl2 were of fully mixed (Fig. 6B), and that in Fig. 6. Lineweaver-Burk Plots of Fumarase Activities in the Presence TES–NaOH buffer was fully competitive (Fig. 6A). The of ATP. 0 The reaction mixtures were the same as in the legend to Fig. 5, inhibition constants Ki and K of ATP were obtained i except that the indicated concentrations of fumarate were used. The graphically35) to be 0.3 mM and 1.6 mM respectively, as mixtures for A contained 0.1 M TES–NaOH buffer, pH 7.0, and shown in the inset to Fig. 6B, and Ki of 0.83 mM was 0mM ( ), 1 mM ( ), and 2 mM ATP ( ). The mixtures for B obtained, as shown in the inset to Fig. 6A. These values contained buffer B, 2 mM MgCl2, and 0 mM ( ), 0.5 mM ( ), and 0 of the inhibition constants of C. glutamicum fumarase 1mM ATP ( ). The inhibition constants Ki and Ki were obtained were much larger than those of yeast and pig-heart from the secondary plots (insets) for slope and intercept on the vertical axis, respectively. fumarases, 0.005 mM and 0.03 mM respectively as reported by Penner and Cohen.34) Therefore it appears unlikely that ATP inhibition works in feedback control Table 3. Inhibition Constants of Substrate Analogs for Fumarases of the TCA cycle to a significant extent in this from C. glutamicum, Pig Heart, and E. coli (FumC) bacterium. Purified fumarase from C. glutamicum (0.4 mg) was incubated in 100 mM Tris–acetate buffer, pH 7.3 (total volume, 1 ml), containing 2– Pyromellitic acid, the substrate analog, has been 20 mML-malate and the inhibitors listed in the table, and the initial reported to be a strong competitive inhibitor for pig-liver absorbance changes at 250 nm were measured, except that the reactions 36) fumarase (Ki ¼ 0:2 mM), pig-heart fumarase (Ki ¼ for the pyromellitate inhibition were measured at 290 nm using 4 mgof 11) 20) the enzyme. The reactions for maleate inhibition were measured at 1 mM and 0.6 mM ), and E. coli FumC (Ki ¼ 1:2 290 nm in the same incubation mixture, except that 10 mM fumarate as mM).11) It also inhibited C. glutamicum fumarase com- substrate, 10 and 20 mM maleate, and 0.8 mg of the enzyme were petitively, but only slightly (Ki ¼ 1:3 mM), as shown in included. Table 3. meso-Tartrate and D-tartrate, the competitive 0 inhibitors for pig-heart fumarase, inhibited C. glutami- Substrate Inhibition constants Ki (Ki )(mM) Reference cum fumarase competitively, with Ki values of 0.11 mM analogs C. glutamicum Pig heart E. coli no. and 16 mM respectively, these values being two orders Pyromellitate 1:3 103 1.0 1.2 11 37) larger than those of the pig-heart enzyme, as shown in 0.6 20 Table 3. Inhibition of yeast fumarase by meso-tartrate L-Tartrate NI 1:3 103 NR 37 4 2 has been reported to be competitive toward L-malate.38) D-Tartrate 1:6 10 3:4 10 NR 37 meso-Tartrate 1:1 102 2.9 NR 37 L-Tartrate, succinate, and glycine have been reported to 3 20,37) D-Malate 2:5 10 5:0 10 NR 20 inhibit pig-heart fumarase competitively, but these (1:2 104) compounds did not inhibit C. glutamicum fumarase. Citrate NI 2:2 10 CI 6, 21 Citrate, which is a competitive inhibitor for both pig- Succinate NI 3:2 102 NR 37 heart6) and E. coli21) fumarases, did not inhibit the Glycine NI 4:0 103 NR 20 Maleate NI CI NR 39 enzyme of this bacterium. D-Malate inhibited C. gluta- micum fumarase in mixed type with a Ki value of 2.5 mM NI, not inhibited. NR, not reported. CI, competitive inhibition 1108 T. GENDA et al. 0 (Ki value of 12 mM), being two orders larger than that 8) Kobayashi, K., and Tuboi, S., End group analysis of the 20) (Ki ¼ 0:05 mM, competitive) of the pig-heart emzyme. cytosolic and mitochondrial fumarases from rat liver. Maleate, the diastereoisomer of fumarate, a competitive J. Biochem., 94, 707–713 (1983). inhibitor for pig-heart fumarase,39) did not inhibit the 9) Woods, S. A., Schwartzbach, S. D., and Guest, J. R., C. glutamicum enzyme. Two biochemically distinct classes of fumarase in Escherichia coli. Biochim. Biophys. Acta, 954, 14–26 In conclusion, fumarase of C. glutamicum possesses (1988). the same characteristics as pig-heart fumarase and 10) Yumoto, N., and Tokushige, M., Characterization of E. coli FumC in many biochemical respects, but features multiple fumarase proteins in Escherichia coli. Biochem. no inhibition by citrate and slight inhibition by pyro- Biophys. Res. Commun., 153, 1236–1243 (1988). mellitic acid (a three-order larger Ki value). The point of 11) Ueda, Y., Yumoto, N., Tokushige, M., Fukui, K., and difference from the pig-heart enzyme is that C. gluta- Ohya-Nishiguchi, H., Purification and characterization micum fumarase was not inhibited by succinate, L- of two types of fumarase from Escherichia coli. J. tartrate, glycine, or maleate (Table 3). Such a difference Biochem., 109, 728–733 (1991). in kinetic properties must come from the steric variety of 12) Weaver, T. M., Levitt, D. G., and Banaszak, L. J., the active center of the enzyme molecule. With a view to Purification and crystallization of fumarase C from unraveling the reaction mechanism in the active center, Escherichia coli. J. Mol. Biol., 231, 141–144 (1993). 13) Shiio, I., Ozaki, H., and Ujigawa, K., Regulation of it was of interest to investigate substrate and substrate- citrate synthase in Brevibacterium flavum, a glutamate- analog binding with C. glutamicum fumarase by crys- producing bacterium. J. Biochem., 82, 395–405 (1977). tallographic, kinetic, and site-directed mutagenesis ex- 14) Shiio, I., and Ozaki, H., Concerted inhibition of periments, referring to recent studies with fumarases of isocitrate dehydrogenase by glyoxylate plus oxalacetate. pig heart,20,40) E. coli,21,41) and S. cerevisiae.22,42) J. Biochem., 64, 45–53 (1968). Concerning the slight activation of C. glutamicum 15) Ozaki, H., and Shiio, I., Regulation of the TCA and fumarase by Mg2þ, as described above, we noticed a glyoxylate cycles in Brevibacterium flavum. I. Inhibition suggestion by Weaver and Banaszak21) that a ‘‘Mg2þ of isocitrate lyase and isocitrate dehydrogenase by site’’ might be present at the same site of a bound water organic acids related to the TCA and glyoxylate cycles. molecule (W-26) in the region of the active center of J. Biochem., 64, 355–363 (1968). E. coli fumarase. 16) Shiio, I., and Ujigawa-Takeda, K., Presence and regu- lation of -ketoglutarate dehydrogenase complex in a glutamate-producing bacterium, Brevibacterium flavum. Acknowledgment Agric. Biol. Chem., 44, 1897–1904 (1980). 17) Genda, T., Nakamatsu, T., and Ozaki, H., Purification The authors thank Dr. Yoshimi Yamamoto of the and characterization of malate dehydrogenase from College of Agriculture of Yamaguchi University for Corynebacterium glutamicum. J. Biosci. Bioeng., 95, technical advice in enzyme preparation. 562–566 (2003). 18) Molenaar, D., van der Rest, M. E., and Petrovic, S., References Biochemical and genetic characterization of the mem- brane-associated malate dehydrogenase (acceptor) from 1) Bock, R. M., and Alberty, R. A., Studies of the enzyme Corynebacterium glutamicum. Eur. J. Biochem., 254, fumarase. I. Kinetics and equilibrium. J. Am. Chem. 395–403 (1998). Soc., 75, 1921–1925 (1953). 19) Molenaar, D., van der Rest, M. E., Drysch, A., and 2) Massey, V., and Alberty, R. A., Ionization constants of Yucel, R., Functions of the membrane-associated and fumarase. Biochim. Biophys. Acta, 13, 354–359 (1954). cytoplasmic malate dehydrogenases in the citric acid 3) Alberty, R. A., Massey, V., Frieden, C., and Fuhlbrigge, cycle of Corynebacterium glutamicum. J. Bacteriol., A. R., Studies of the enzyme fumarase. III. The 182, 6884–6891 (2000). dependence of the kinetic constants at 25 C upon the 20) Beeckmans, S., and Driessche, E. V., Pig heart fumarase concentration and pH of phosphate buffers. J. Am. Chem. contains two distinct substrate-binding sites differing in Soc., 76, 2485–2493 (1954). affinity. J. Biol. Chem., 273, 31661–31669 (1998). 4) Kanarek, L., and Hill, R. L., The preparation and 21) Weaver, T., and Banaszak, L., Crystallographic studies characterization of fumarase from swine heart muscle. of the catalytic and a second site in fumarase C from J. Biol. Chem., 239, 4202–4206 (1964). Escherichia coli. Biochemistry, 35, 13955–13965 5) Robinson, G. W., Bradshaw, R. A., Kanarek, L., and (1996). Hill, R. L., The thiol groups of fumarase. J. Biol. Chem., 22) Weaver, T., Lees, M., Zaitsev, V., Zaitseva, I., Duke, E., 242, 2709–2718 (1967). Lindley, P., McSweeny, S., Svensson, A., Keruchenko, 6) Teipel, J. W., Hass, G. M., and Hill, R. L., The substrate J., Keruchenko, I., Gladilin, K., and Banaszak, L., specificity of fumarase. J. Biol. Chem., 243, 5684–5694 Crystal structures of native and recombinant yeast (1968). fumarase. J. Mol. Biol., 280, 431–442 (1998). 7) Kobayashi, K., Yamanishi, T., and Tuboi, S., Physico- 23) Lowry, O. H., Rosebrough, N. J., Farr, A. L., and chemical, catalytic, and immunochemical properties of Randall, R. J., Protein measurement with the Folin fumarases crystallized separately from mitochondrial phenol reagent. J. Biol. Chem., 193, 265–275 (1951). and cytosolic fractions of rat liver. J. Biochem., 89, 24) Keruchenko, J. S., Keruchenko, I. D., Gladilin, K. L., 1923–1931 (1981). Zaitsev, V. N., and Chirgadze, N. Y., Purification, Fumarase from C. glutamicum 1109 characterization and preliminary X-ray study of fumar- 33) Dixon, M., Webb, E. C., Thorne, C. J. R., and Tipton, ase from Saccharomyces cerevisiae. Biochim. Biophys. K. F., Equilibrium constant in terms of V and Km: the Acta, 1122, 85–92 (1992). Haldane relationship. In ‘‘Enzymes’’ 3rd ed., eds. Dixon, 25) Moir, A., Feavers, I. M., and Guest, J. R., Character- M., and Webb, E. C., Longman Group, London, pp. 67– ization of the fumarase gene of Bacillus subtilis 168 70 (1979). cloned and expressed in Escherichia coli K 12. J. Gen. 34) Penner, P. E., and Cohen, L. H., Effects of adenosine Microbiol., 130, 3009–3017 (1984). triphosphate and magnesium ions on the fumarase 26) Puchegger, S., Redl, B., and Sto¨ffler, G., Purification and reaction. J. Biol. Chem., 244, 1070–1075 (1969). properties of a thermostable fumarate hydratase from the 35) Dixon, M., Webb, E. C., Thorne, C. J. R., and Tipton, K. archaeobacterium Sulfolobus solfataricus. J. Gen. Mi- F., Enzyme inhibition and activation. In ‘‘Enzymes’’ 3rd crobiol., 136, 1537–1541 (1990). ed., eds. Dixon, M., and Webb, E. C., Longman Group, 27) Mizobata, T., Fujioka, T., Yamasaki, F., Hidaka, M., London, pp. 332–353 (1979). Nagai, J., and Kawata, Y., Purification and character- 36) Beeckmans, S., and Kanarek, L., A new purification ization of a thermostable class II fumarase from Thermus procedure for fumarase based of affinity chromatogra- thermophilus. Arch. Biochem. Biophys., 355, 49–55 phy: isolation and characterization of pig-liver fumarase. (1998). Eur. J. Biochem., 78, 437–444 (1977). 28) Hill, R. L., and Teipel, J. W., Fumarase and crotonase. In 37) Wigler, P. W., and Alberty, R. A., The pH dependence of ‘‘The Enzymes’’ 3rd ed., ed. Boyer, P. D., Academic the competitive inhibition of fumarase. J. Am. Chem. Press, New York, pp. 539–571 (1971). Soc., 82, 5482–5488 (1960). 29) Hassett, D. J., Howell, M. L., Sokol, P. A., Vasil, M. L., 38) Rose, I. A., Restructuring the active site of fumarase for and Dean, G. E., Fumarase C activity is elevated in the fumarate to malate reaction. Biochemistry, 36, response to iron deprivation and in mucoid, alginate- 12346–12354 (1997). producing Pseudomonas aeruginosa: cloning and char- 39) Massey, V., Studies on fumarase. 4. The effects of acterization of fumC and purification of native FumC. inhibitors on fumarase activity. Biochem. J., 55, 172–177 J. Bacteriol., 179, 1442–1451 (1997). (1953). 30) Teipel, J. W., and Hill, R. L., The subunit interactions of 40) Rose, I. A., Warms, J. V. B., and Kuo, D. J., Proton fumarase. J. Biol. Chem., 246, 4859–4865 (1971). transfer in catalysis by fumarase. Biochemistry, 31, 31) Nakamura, S., and Ogata, H., Specificity of fumarate 9993–9999 (1992). hydratase. I. Formation of oxalacetate from unnatural 41) Rose, I. A., and Weaver, T. M., The role of the allosteric ()-tartrate by fumarate hydratase. J. Biol. Chem., 243, B site in the fumarase reaction. Proc. Natl. Acad. Sci. 528–532 (1968). USA, 101, 3393–3397 (2004). 32) Alberty, R. A., Fumarase. In ‘‘The Enzymes’’ 2nd ed., 42) Rose, I. A., How fumarase recycles after the malate! eds. Boyer, P. D., Lardy, H., and Myrbuck, K., fumarate reaction: insights into the reaction mechanism. Academic Press, New York, pp. 531–544 (1961). Biochemistry, 37, 17651–17658 (1998).