JOURNAL OF BACTERIOLOGY, Aug. 1975, p. 704-716 Copyright © 1975 American Society for Microbiology
Vol. 123, No. 2 Printed in U.S.A.
Purification and Properties of Malate Dehydrogenase from Pseudom onas testosteroni KWAN-SA YOU AND NATHAN 0. KAPLAN* Department of Chemistry, University of California at San Diego, La Jolla, California 92037
Received for publication 3 March 1975
Nicotinamide adenine dinucleotide-linked malate dehydrogenase has been purified from Pseudomonas testosteroni (ATCC 11996). The purification represents over 450-fold increase in specific activity. The amino acid composition of the enzyme was determined and found to be quite different from the composition of the malate dehydrogenases from animal sources as well as from Escherichia coli. Despite this difference, however, the data show that the enzymatic properties of the purified enzyme are remarkably similar to those of other malate dehydrogenases that have been previously studied. The Pseudomonas enzyme has a molecular weight of 74,000 and consists of two subunits of identical size. In addition to L-malate, the enzyme slowly oxidizes other four-carbon dicarboxylates having an a-hydroxyl group of S configuration such as meso- and (- )-tartrate. Rate-determining steps, which differ from that of the reaction involving L-malate, are discussed for the reaction involving these alternative substrates. Oxidation of hydroxymalonate, a process previously undetected with other malate dehydrogenases, is demonstrated fluorometrically. Hydroxymalonate and D-malate strongly enhance the fluorescence of the reduced nicotinamide adenine dinucleotide bound to the enzyme. The enzyme is A-stereospecific with respect to the coenzyme. Malate dehydrogenase is present in a single form in the Pseudomonas. The susceptibility of the enzyme to activation or inhibition by its substrates-particularly the favoring of the oxidation of malate at elevated concentrations-strongly resembles the properties of the mitochondrial enzymes. The present study reveals that whereas profound variations in chemical composition have occurred between the prokaryotic and eukaryotic enzymes, the physical and catalytic properties of malate dehydrogenase, unlike lactate dehydrogenase, are well conserved during the evolutionary process. The purification, properties, and catalytic mechanism of nicotinamide adenine dinucleotide-linked malate dehydrogenases (L-malate: NAD+ oxidoreductase, EC 1.1.1.37) from various sources, particularly animals, have been a subject of extensive study over recent years. The enzymes occur in two forms in eukaryotic organisms: one in the mitochondrion and the other in the cytoplasm. The enzymes from animal sources are composed of two subunits having an intact molecular weight of 65,000 to 74,000 (see reference 40), and are capable of reversibly oxidizing meso-tartrate, (-)-tartrate, a-hydroxyglutarate, and oxaloglycolate, in addition to L-malate (9). Ketomalonate was reported to be reduced irreversibly by the enzymes (13, 32). Data obtained from initial velocity (29), product inhibition (30), and isotopic ex704
change at equilibrium (33, 34) studies with animal enzymes were consistent with an "ordered Bi Bi" mechanism. This mechanism involves formation of a ternary complex, where binding of the coenzyme to the enzyme occurs prior to substrate binding and dissociation of product precedes that of the coenzyme produced, which is also the rate-determining step of the catalysis. The enzymes are A-stereospecific with respect to the coenzyme (10). Although malate dehydrogenases from animal sources are well-defined enzymes, no extensive studies comparable to those of the animal enzymes have been reported for the enzymes from bacterial origins. Available information seems to indicate that, unlike the animal malate dehydrogenases, the bacterial enzymes have diverse physical properties; the Bacillus subtilis
VOL. 123, 1975
MALATE DEHYDROGENASE FROM PSEUDOMONAS
enzyme is unusual and is shown to be a tetramer with a molecular weight of 117,000 to 148,000 (25, 41), whereas molecular weights of 60,000 to 67,000 and 43,000 were reported for the enzyme from Escherichia coli (consisting of two subunits) (25, 26) and Pseudomonas acidovorans (18), respectively. However, since the enzyme has been isolated from a small number of different kinds of bacteria, it is quite difficult to determine the prevalence of this diversity. The present studies on malate dehydrogenase from Pseudomonas testosteroni Stanier strain 78 (ATCC 11996) were undertaken to investigate the variations of the properties of the bacterial enzymes, and, also, to compare them with the well-studied animal enzymes. MATERIALS AND METHODS Chemicals. Chemicals used were purchased from the following sources: nicotinamide adenine nucleotide coenzymes, coenzyme analogues, and yeast alcohol dehydrogenase from P-L Biochemicals; oxaloacetic acid, L(-)- and D(+)-malic acid, (-)-, (+)-, and meso-tartaric acid, hydroxymalonic acid, ketomalonic acid, L( + )-aspartic acid, and succinate (disodium) from Calbiochem; D,L-a-hydroxybutyrate (sodium). L-a-hydroxyglutarate (disodium), horse cytochrome c, ovalbumin, bovine serum albumin, sodium dodecyl sulfate (SDS), and DEAE-Sephadex(A-25-120) from Sigma; Sephadex G-100 from Pharmacia; diethylaminoethyl (DEAE)-11 cellulose from W. & R. Balston; and trypsin inhibitor from Worthington. Growth of organism. The culture medium and conditions for the growth of P. testosteroni were the same as that described by Marcus and Talalay (22) except for the omission of testosterone. The culture was grown in a 500-liter fermenter (at the New England Enzyme Center), harvested, and washed with cold 0.05 M tris(hydroxymethyl)aminomethane-chloride containing 0.01 M ethylenediaminetetraacetate, pH 8.0, by centrifugation. Approximately 2.3 g of cells (wet weight) was obtained per liter. Enzyme assay. Malate dehydrogenase activity was assayed essentially by the spectrophotometric method described previously (25). For routine assays, a 3.0-ml assay mixture contained 0.13 mM reduced nicotinamide adenine dinucleotide (NADH), 0.47 mM oxaloacetate (neutralized with 0.1 N NaOH), 0.1 M Tris-chloride, pH 8.8, and enzyme. L-Malate oxidation was measured in the presence of 10 mM L-malate and 3.4 mM nicotinamide adenine dinucleotide (NAD+) in the same buffer. One enzyme unit was defined as the amount of enzyme required for oxidation/reduction of 1 umol of substrate per min. Assays in the presence of coenzyme analogues were conducted at the following wavelengths: 3-acetylpyridine-NAD+, 363 nm; thionicotinamide-NAD+, 395 nm; deamino-NAD+ (the hypoxanthine analogue of NADI), 338 nm. Molar extinction coefficients for coenzymes and coenzyme analogues were obtained
705
from "Ultraviolet absorption spectra of pyridine nucleotide coenzymes and coenzyme analogues" (28). Protein determinations. During the purification procedure, the protein concentrations were determined spectrophotometrically by the ratio of absorbance at 280 and 260 nm. The extinction coefficient Al% nm was determined with pure enzyme by the biuret method and used for protein determination of purified enzyme. The biuret reaction was standardized with bovine serum albumin solution of known concentration. Electrophoresis. The agar gel and starch gel electrophoresis were carried out by the methods described by Nerenberg (27) and Fine and Costello (12), respectively. Sedimentation and diffusion coefficients. Sedimentation velocity analysis was performed in a Spinco Model E ultracentrifuge at 67,770 rpm with schlieren optics and a 12-mm single-sector cell. The value of S,0.w was obtained from Sob. determined at five different enzyme concentrations. In this and other ultracentrifugation studies, enzyme solution in 0.05 M Tris-chloride, pH 8.0, was exhaustively dialyzed against the same buffer containing 0.1 M NaCl. The diffusion coefficient was determined from the rate of boundary spreading during a synthetic boundary run (31). The centrifugation was performed at 4,609 rpm at 11 C with a 12-mm double-sector centerpiece having a synthetic boundary capillary. Dob. determined at a protein concentration of 2.1 mg/ml was converted to D20, . Molecular weight determinations. Molecular weight determination by sedimentation equilibrium centrifugation was carried out according to Yphantis (42). Enzyme solutions of 0.370, 0.185, and 0.093 mg/ml were centrifuged at 34,630 rpm and 6 to 11 C. The partial specific volume was estimated from the amino acid composition by the method of Cohn and Edsall (8). The molecular weight of the subunit of the enzyme was estimated by SDS-polyacrylamide gel electrophoresis (38) with bovine serum albumin, ovalbumin, soybean trypsin inhibitor, and horse heart cytochrome c as authentic markers. An estimation of molecular weight of the enzyme was also determined by gel filtration as described by Andrews (2). An analytical Sephadex G-100 column (2.5 by 94 cm) equilibrated at 4 C with 0.05 M Tris-chloride, pH 7.5, containing 0.1 M KCl, was used. Amino acid analysis. Amino acid analysis was performed with the enzyme samples hydrolyzed for 24, 48, and 72 h according to Moore et al. (24) in a Beckman-Spinco automatic amino acid analyzer. Corrections for zero time hydrolysis were made for threonine and serine from semilogarithmic extrapolation. Values obtained with samples hydrolyzed for 72 h were used for valine and isoleucine. Cysteine and cystine were determined as cysteic acid after performic acid oxidation (23). Tryptophan was determined according to Bencze and Schmid (3). Immunological preparations. Antiserum against the enzyme was prepared by injecting a rabbit subcutaneously at weekly intervals with 0.08 mg of the
706
YOU AND KAPLAN
J. BACTERIOL.
enzyme emulsified in Freund incomplete adjuvant tive analogues, preliminary double reciprocal plots (BBL) over a period of 4 weeks. Two weeks after the were drawn manually. When these plots appeared to final subcutaneous injection, a total of 1.8 mg of the be linear, the data were then analyzed by the program enzyme in saline was administered intravenously at written to make least-squares fits to the hyperbola 2-day intervals. The antiserum was prepared 4 weeks equation: after the last intravenous injection and tested in the v = VA/(K + A) (1) Ouchterlony double diffusion plates. is the concentration A is velocity where v (Lmol/min), Fluorometric studies. All fluorometric measurements were carried out in an Aminco SPF 125 of varied substrate, K is the apparent Michaelis fluorometer at room temperature. Samples with a constant, and V the maximum velocity. This program final volume of 3 ml were excited at 340 nm, and gave the values of K and V (as well as the values of fluorescence emission was measured at 440 to 450 nm. K/V, 1/V, and V/K) including their respective stanThe absorbance of samples was kept lower than 0.10 dard errors. Inhibition experiments with substrate analogues in order to minimize concentration quenching. Coenzyme stereospecificity. [4-3H]NAD+ was were carried out at oxaloacetate concentrations where prepared essentially by the enzymatic procedure of no substrate inhibition was observed and L-malate activation took place Allison et al. (1). In this procedure [4B-'H]NADH concentrations where substrate of the enzyme activation possible to prevent in order was prepared by stereospecifically reducing NAD+ with glyceraldehyde-3-phosphate dehydrogenase in by substrate analogues under examination, if the the presence of [1-3H ]glyceraldehyde-3-phosphate, enzyme is not in its fully activated form. To deterwhich was generated from [1-3 H Idihydroxyacetone mine the inhibition constants, slopes and intercepts of inhibitors were replotted phosphate by triose phosphate isomerase (1). The obtained in the presence by hand against inhibitor concentrations. When the stereospecific labeling of NADH was then doubly replots were linear, the slope and intercept data were confirmed by reoxidizing it with fl-hydroxybutyrate statistically by the computer programs writanalyzed dehydrogenase from Rhodopseudomonas spheroides, (2), (3), a B-specific enzyme, and supernatant malate dehy- ten to make least-squares fits to equations drogenase from pig heart. fB-Hydroxybutyrate de- and (4) for competitive, uncompetitive, and noncomhydrogenase from R. spheroides (obtained from petitive inhibition, respectively. (2) v = VA/[K(1 + I/K,8) + A] Calbiochem), as well as the lecithin requiring mammalian enzyme from heart mitochondria, exhibits = v VA/[K + (1 + I/Kii)A] (3) B-stereospecificity with respect to NAD+ (K. You and L. J. Arnold, Jr., manuscript in preparation). (4) v = VA/[K(1 + I/K,8) + (1 + I/K,,)A] The former resulted in nonradioactive NAD+. This preparative procedure was performed by L. H. Bern- In these equations Ki, and K,, are inhibition constants stein in this laboratory and [4-3H]NAD+, the product representing uncompetitive (intercept) and competiof the supernatant malate dehydrogenase reaction, tive (slope) inhibition, respectively. Purification of malate dehydrogenase. The was used in the present experiment. [4- 'H INAD + (1.2 umol; specific radioactivity 1.5 washed cells were suspended in 1 volume of the same x 106 counts/min per Mmol) was converted to [4B- buffer used for the wash. Whole-cell acetone-dried 3H]NADH with yeast alcohol dehydrogenase in the powder was prepared by slowly pouring the cell presence of 10 Amol of ethanol in 0.01 M NH4HCO,. suspension into 5 to 10 volumes of cold (-20 C or When the reaction reached equilibrium, the mixture lower) acetone with vigorous agitation. The resultant was applied on a DEAE-Sephadex (A-25-120) column cell precipitate was collected by filtration through (1.5 by 20 cm) equilibrated with 0.01 M NH4HCO8, Whatman 3 MM paper. Step 1. Crude extract. Acetone-dried powder eluted according to the procedure of Allison et al. (1), and radioactivity and optical density at 260 and 340 (800 g) prepared from 3.1 kg of wet cells was susnm of the effluent were determined. [4B-3H]NADH pended in 10 volumes of 0.05 M Tris-chloride, con(0.62 lsmol; 9.4 x 105 counts/min) thus prepared was taining 0.01 M ethylenediaminetetraacetate, pH 8.0, then reacted with 3 ,umol of oxaloacetate in the and stirred for 18 h at 4 C, and then the insoluble presence of Pseudomonas enzyme in 0.01 M debris was removed by centrifugation. Step 2. Ammonium sulfate fractionation. Solid NH4HCO,. When oxidation of the [4B-3H]NADH ceased, the reaction mixture was lyophilized, the ammonium sulfate was added to the crude extract NAD + produced was isolated by the above chro- from step 1 to give 48% saturation. The degree of matographic procedure, and its radioactivity was saturation was based on the solubility of ammonium determined. sulfate at 25 C. The solution was stirred for 4 h at 4 C Kinetic data processing. The enzyme kinetic and centrifuged, and then more ammonium sulfate constants such as Michaelis constants, maximum was added to the supernatant to give 80% saturation. velocities, and inhibition constants were determined The precipitate was collected by centrifugation, disby employing the Fortran computer program devel- solved in 0.05 M Tris-chloride, pH 8.0, and slowly oped by Cleland (6, 7), with the use of a Control Data brought to 68% ammonium sulfate saturation by Corp. 3600 digital computer. dialyzing against an ammonium sulfate saturation by To determine Michaelis constants and maximum dialyzing against an ammonium sulfate solution in velocities for substrates, coenzymes, and their respec- 0.05 M Tris-chloride, pH 8.0, which gave the desired
VOL. 123, 1975
MALATE DEHYDROGENASE FROM PSEUDOMONAS
707
which corresponds to a catalytic constant of 2.00 x 105/mol of the enzyme on the basis of the molecular weight of 74,200 (below). The purified enzyme showed a single protein band coincident with the activity band in agar gel electrophoresis (Fig. 2). Figure 2 also presents the pattern of starch gel electrophoresis of the bacterial crude extract, which showed a single activity spot, thereby excluding the possibility of the presence of isoenzymes or conformers (17) in the organism. Homogeneity of the purified enzyme was further demonstrated by the following observations: a symmetrical boundary traversing to the bottom of the cell during a sedimentation velocity run (Fig. 3); a single precipitin band on the Ouchterlony plate between the antiserum against the enzyme and the purified enzyme or crude extract (Fig. 4); and a single protein band (subunit) in SDS-polyacrylamide gel electrophoresis (Fig. 5). Figure 3 also shows a slight increase in the sedimentation coefficient S20, of the enzyme upon dilution. Extrapolation of S20.W values to infinite dilution of protein concentration gave a sedimentation coefficient S 2o2wof 4.50 S. The diffusion coefficient D20,w of the enzyme at a fixed protein concentration of 2.1 mg/ml was determined to be 5.39 x 10-f cm2/s by the synthetic boundary run. The enzyme solution in 0.02 M potassium phosphate, pH 6.5, gave an extinction coefficient A" 2nm of 9.00, which corresponds to a molar extinction coefficient of 6.66 x 104 M-1 RESULTS cm-'. The absorbance ratio A280A260 of the Purification and physicochemical enzyme was 1.81 at pH 8.8. The following properties. The specific activities and the ac- information weight of the enzyme of in tivity recoveries the enzyme the purifica- was obtainedonbymolecular 74,200 by techniques: various tion steps are summarized in Table 1. Figure 1 sedimentation equilibrium runs at protein conshows the photograph of the enzyme crystals. centrations between 0.093 to 0.370 mg/ml; The specific activity of the enzyme under the 75,000 by the Svedberg equation employing the standard assay conditions was 2,700 Amol of sedimentation and diffusion coefficients deterNADH oxidized per min per mg of protein, mined at the protein concentration of 2.1 mg/ ml; 70,000 by the analytical Sephadex G-100 TABLE 1. Summary of purification of Pseudomonas column chromatography. The partial specific malate dehydrogenase volume of the enzyme was determined to be 0.732 cm3/g. Total Total Activity Figure 5 shows the banding pattern of the protein enzyte recovery Sp act Steps enzyme in SDS-polyacrylamide gel electropho(g) (X 10-3) resis along with authentic protein markers of 6 known molecular weight. The enzyme moved as 100 246 1. Crude extract 1,458 83 88 13.8 2. 48 to 68% am1,208 a single band with a mobility corresponding to a monium sulfate molecular weight of 36,000, a value approxi365 542 37 1.485 3. DEAE-11 cellumately half of that obtained from ultracentrifulose column 33 393 gation or gel filtration. It is therefore concluded 0.157 4. Sephadex G-100 2.506 column that the Pseudomonas enzyme is made up of 26 373 0.135 5. Crystallization 2,762 two subunits of identical size. Table 2 contains
saturation when equilibrium was reached. The precipitate was dissolved in 0.005 M Tris-chloride, pH 8.0, and dialyzed against the same buffer. Step 3. DEAE-ll cellulose column chromatography. The dialyzed solution from step 2 was applied on a DEAE-11 cellulose column (5.0 by 70 cm) previously equilibrated with 0.005 M Tris-chloride, pH 8.0. The column was washed with the equilibration buffer until the effluent did not show any appreciable absorption at 280 nm. The enzyme was then eluted with a linear gradient salt concentration consisting of 3 liters of the equilibration buffer in a mixing chamber and the same volume of the buffer containing 0.3 M NaCl in a reservoir chamber. Fractions containing the enzyme activity were pooled and proteins were precipitated by dialysis against saturated ammonium sulfate solution. The precipitate was dissolved in 0.05 M Tris-chloride, pH 8.0, and followed by a 53 to 68% ammonium sulfate fractionation by the dialysis method described in step 2. Step 4. Sephadex G-100 column chromatography. The proteins precipitated at 53 to 68% ammonium sulfate saturation were dissolved in a minimum volume of 0.05 M Tris-chloride, pH 8.0, layered on a Sephadex G-100 column (5 by 80 cm), equilibrated with 0.05 M Tris-chloride, pH 8.0, and eluted with the equilibration buffer. The fractions having the enzyme activity were pooled. Step 5. Crystallization. The pooled enzyme solution was brought to 53% ammonium sulfate saturation and the precipitated inactive proteins were removed by centrifugation. Additional ammonium sulfate was added a few crystals at a time until slight turbidity was observed. This solution was stored at 4 C, whereupon crystallization occurred in 12 h.
708
YOU AND KAPLAN
J. BACTERIOL.
4~~~ zf ,4
Vol. 123, No. 2 Printed in U.S.A.
Purification and Properties of Malate Dehydrogenase from Pseudom onas testosteroni KWAN-SA YOU AND NATHAN 0. KAPLAN* Department of Chemistry, University of California at San Diego, La Jolla, California 92037
Received for publication 3 March 1975
Nicotinamide adenine dinucleotide-linked malate dehydrogenase has been purified from Pseudomonas testosteroni (ATCC 11996). The purification represents over 450-fold increase in specific activity. The amino acid composition of the enzyme was determined and found to be quite different from the composition of the malate dehydrogenases from animal sources as well as from Escherichia coli. Despite this difference, however, the data show that the enzymatic properties of the purified enzyme are remarkably similar to those of other malate dehydrogenases that have been previously studied. The Pseudomonas enzyme has a molecular weight of 74,000 and consists of two subunits of identical size. In addition to L-malate, the enzyme slowly oxidizes other four-carbon dicarboxylates having an a-hydroxyl group of S configuration such as meso- and (- )-tartrate. Rate-determining steps, which differ from that of the reaction involving L-malate, are discussed for the reaction involving these alternative substrates. Oxidation of hydroxymalonate, a process previously undetected with other malate dehydrogenases, is demonstrated fluorometrically. Hydroxymalonate and D-malate strongly enhance the fluorescence of the reduced nicotinamide adenine dinucleotide bound to the enzyme. The enzyme is A-stereospecific with respect to the coenzyme. Malate dehydrogenase is present in a single form in the Pseudomonas. The susceptibility of the enzyme to activation or inhibition by its substrates-particularly the favoring of the oxidation of malate at elevated concentrations-strongly resembles the properties of the mitochondrial enzymes. The present study reveals that whereas profound variations in chemical composition have occurred between the prokaryotic and eukaryotic enzymes, the physical and catalytic properties of malate dehydrogenase, unlike lactate dehydrogenase, are well conserved during the evolutionary process. The purification, properties, and catalytic mechanism of nicotinamide adenine dinucleotide-linked malate dehydrogenases (L-malate: NAD+ oxidoreductase, EC 1.1.1.37) from various sources, particularly animals, have been a subject of extensive study over recent years. The enzymes occur in two forms in eukaryotic organisms: one in the mitochondrion and the other in the cytoplasm. The enzymes from animal sources are composed of two subunits having an intact molecular weight of 65,000 to 74,000 (see reference 40), and are capable of reversibly oxidizing meso-tartrate, (-)-tartrate, a-hydroxyglutarate, and oxaloglycolate, in addition to L-malate (9). Ketomalonate was reported to be reduced irreversibly by the enzymes (13, 32). Data obtained from initial velocity (29), product inhibition (30), and isotopic ex704
change at equilibrium (33, 34) studies with animal enzymes were consistent with an "ordered Bi Bi" mechanism. This mechanism involves formation of a ternary complex, where binding of the coenzyme to the enzyme occurs prior to substrate binding and dissociation of product precedes that of the coenzyme produced, which is also the rate-determining step of the catalysis. The enzymes are A-stereospecific with respect to the coenzyme (10). Although malate dehydrogenases from animal sources are well-defined enzymes, no extensive studies comparable to those of the animal enzymes have been reported for the enzymes from bacterial origins. Available information seems to indicate that, unlike the animal malate dehydrogenases, the bacterial enzymes have diverse physical properties; the Bacillus subtilis
VOL. 123, 1975
MALATE DEHYDROGENASE FROM PSEUDOMONAS
enzyme is unusual and is shown to be a tetramer with a molecular weight of 117,000 to 148,000 (25, 41), whereas molecular weights of 60,000 to 67,000 and 43,000 were reported for the enzyme from Escherichia coli (consisting of two subunits) (25, 26) and Pseudomonas acidovorans (18), respectively. However, since the enzyme has been isolated from a small number of different kinds of bacteria, it is quite difficult to determine the prevalence of this diversity. The present studies on malate dehydrogenase from Pseudomonas testosteroni Stanier strain 78 (ATCC 11996) were undertaken to investigate the variations of the properties of the bacterial enzymes, and, also, to compare them with the well-studied animal enzymes. MATERIALS AND METHODS Chemicals. Chemicals used were purchased from the following sources: nicotinamide adenine nucleotide coenzymes, coenzyme analogues, and yeast alcohol dehydrogenase from P-L Biochemicals; oxaloacetic acid, L(-)- and D(+)-malic acid, (-)-, (+)-, and meso-tartaric acid, hydroxymalonic acid, ketomalonic acid, L( + )-aspartic acid, and succinate (disodium) from Calbiochem; D,L-a-hydroxybutyrate (sodium). L-a-hydroxyglutarate (disodium), horse cytochrome c, ovalbumin, bovine serum albumin, sodium dodecyl sulfate (SDS), and DEAE-Sephadex(A-25-120) from Sigma; Sephadex G-100 from Pharmacia; diethylaminoethyl (DEAE)-11 cellulose from W. & R. Balston; and trypsin inhibitor from Worthington. Growth of organism. The culture medium and conditions for the growth of P. testosteroni were the same as that described by Marcus and Talalay (22) except for the omission of testosterone. The culture was grown in a 500-liter fermenter (at the New England Enzyme Center), harvested, and washed with cold 0.05 M tris(hydroxymethyl)aminomethane-chloride containing 0.01 M ethylenediaminetetraacetate, pH 8.0, by centrifugation. Approximately 2.3 g of cells (wet weight) was obtained per liter. Enzyme assay. Malate dehydrogenase activity was assayed essentially by the spectrophotometric method described previously (25). For routine assays, a 3.0-ml assay mixture contained 0.13 mM reduced nicotinamide adenine dinucleotide (NADH), 0.47 mM oxaloacetate (neutralized with 0.1 N NaOH), 0.1 M Tris-chloride, pH 8.8, and enzyme. L-Malate oxidation was measured in the presence of 10 mM L-malate and 3.4 mM nicotinamide adenine dinucleotide (NAD+) in the same buffer. One enzyme unit was defined as the amount of enzyme required for oxidation/reduction of 1 umol of substrate per min. Assays in the presence of coenzyme analogues were conducted at the following wavelengths: 3-acetylpyridine-NAD+, 363 nm; thionicotinamide-NAD+, 395 nm; deamino-NAD+ (the hypoxanthine analogue of NADI), 338 nm. Molar extinction coefficients for coenzymes and coenzyme analogues were obtained
705
from "Ultraviolet absorption spectra of pyridine nucleotide coenzymes and coenzyme analogues" (28). Protein determinations. During the purification procedure, the protein concentrations were determined spectrophotometrically by the ratio of absorbance at 280 and 260 nm. The extinction coefficient Al% nm was determined with pure enzyme by the biuret method and used for protein determination of purified enzyme. The biuret reaction was standardized with bovine serum albumin solution of known concentration. Electrophoresis. The agar gel and starch gel electrophoresis were carried out by the methods described by Nerenberg (27) and Fine and Costello (12), respectively. Sedimentation and diffusion coefficients. Sedimentation velocity analysis was performed in a Spinco Model E ultracentrifuge at 67,770 rpm with schlieren optics and a 12-mm single-sector cell. The value of S,0.w was obtained from Sob. determined at five different enzyme concentrations. In this and other ultracentrifugation studies, enzyme solution in 0.05 M Tris-chloride, pH 8.0, was exhaustively dialyzed against the same buffer containing 0.1 M NaCl. The diffusion coefficient was determined from the rate of boundary spreading during a synthetic boundary run (31). The centrifugation was performed at 4,609 rpm at 11 C with a 12-mm double-sector centerpiece having a synthetic boundary capillary. Dob. determined at a protein concentration of 2.1 mg/ml was converted to D20, . Molecular weight determinations. Molecular weight determination by sedimentation equilibrium centrifugation was carried out according to Yphantis (42). Enzyme solutions of 0.370, 0.185, and 0.093 mg/ml were centrifuged at 34,630 rpm and 6 to 11 C. The partial specific volume was estimated from the amino acid composition by the method of Cohn and Edsall (8). The molecular weight of the subunit of the enzyme was estimated by SDS-polyacrylamide gel electrophoresis (38) with bovine serum albumin, ovalbumin, soybean trypsin inhibitor, and horse heart cytochrome c as authentic markers. An estimation of molecular weight of the enzyme was also determined by gel filtration as described by Andrews (2). An analytical Sephadex G-100 column (2.5 by 94 cm) equilibrated at 4 C with 0.05 M Tris-chloride, pH 7.5, containing 0.1 M KCl, was used. Amino acid analysis. Amino acid analysis was performed with the enzyme samples hydrolyzed for 24, 48, and 72 h according to Moore et al. (24) in a Beckman-Spinco automatic amino acid analyzer. Corrections for zero time hydrolysis were made for threonine and serine from semilogarithmic extrapolation. Values obtained with samples hydrolyzed for 72 h were used for valine and isoleucine. Cysteine and cystine were determined as cysteic acid after performic acid oxidation (23). Tryptophan was determined according to Bencze and Schmid (3). Immunological preparations. Antiserum against the enzyme was prepared by injecting a rabbit subcutaneously at weekly intervals with 0.08 mg of the
706
YOU AND KAPLAN
J. BACTERIOL.
enzyme emulsified in Freund incomplete adjuvant tive analogues, preliminary double reciprocal plots (BBL) over a period of 4 weeks. Two weeks after the were drawn manually. When these plots appeared to final subcutaneous injection, a total of 1.8 mg of the be linear, the data were then analyzed by the program enzyme in saline was administered intravenously at written to make least-squares fits to the hyperbola 2-day intervals. The antiserum was prepared 4 weeks equation: after the last intravenous injection and tested in the v = VA/(K + A) (1) Ouchterlony double diffusion plates. is the concentration A is velocity where v (Lmol/min), Fluorometric studies. All fluorometric measurements were carried out in an Aminco SPF 125 of varied substrate, K is the apparent Michaelis fluorometer at room temperature. Samples with a constant, and V the maximum velocity. This program final volume of 3 ml were excited at 340 nm, and gave the values of K and V (as well as the values of fluorescence emission was measured at 440 to 450 nm. K/V, 1/V, and V/K) including their respective stanThe absorbance of samples was kept lower than 0.10 dard errors. Inhibition experiments with substrate analogues in order to minimize concentration quenching. Coenzyme stereospecificity. [4-3H]NAD+ was were carried out at oxaloacetate concentrations where prepared essentially by the enzymatic procedure of no substrate inhibition was observed and L-malate activation took place Allison et al. (1). In this procedure [4B-'H]NADH concentrations where substrate of the enzyme activation possible to prevent in order was prepared by stereospecifically reducing NAD+ with glyceraldehyde-3-phosphate dehydrogenase in by substrate analogues under examination, if the the presence of [1-3H ]glyceraldehyde-3-phosphate, enzyme is not in its fully activated form. To deterwhich was generated from [1-3 H Idihydroxyacetone mine the inhibition constants, slopes and intercepts of inhibitors were replotted phosphate by triose phosphate isomerase (1). The obtained in the presence by hand against inhibitor concentrations. When the stereospecific labeling of NADH was then doubly replots were linear, the slope and intercept data were confirmed by reoxidizing it with fl-hydroxybutyrate statistically by the computer programs writanalyzed dehydrogenase from Rhodopseudomonas spheroides, (2), (3), a B-specific enzyme, and supernatant malate dehy- ten to make least-squares fits to equations drogenase from pig heart. fB-Hydroxybutyrate de- and (4) for competitive, uncompetitive, and noncomhydrogenase from R. spheroides (obtained from petitive inhibition, respectively. (2) v = VA/[K(1 + I/K,8) + A] Calbiochem), as well as the lecithin requiring mammalian enzyme from heart mitochondria, exhibits = v VA/[K + (1 + I/Kii)A] (3) B-stereospecificity with respect to NAD+ (K. You and L. J. Arnold, Jr., manuscript in preparation). (4) v = VA/[K(1 + I/K,8) + (1 + I/K,,)A] The former resulted in nonradioactive NAD+. This preparative procedure was performed by L. H. Bern- In these equations Ki, and K,, are inhibition constants stein in this laboratory and [4-3H]NAD+, the product representing uncompetitive (intercept) and competiof the supernatant malate dehydrogenase reaction, tive (slope) inhibition, respectively. Purification of malate dehydrogenase. The was used in the present experiment. [4- 'H INAD + (1.2 umol; specific radioactivity 1.5 washed cells were suspended in 1 volume of the same x 106 counts/min per Mmol) was converted to [4B- buffer used for the wash. Whole-cell acetone-dried 3H]NADH with yeast alcohol dehydrogenase in the powder was prepared by slowly pouring the cell presence of 10 Amol of ethanol in 0.01 M NH4HCO,. suspension into 5 to 10 volumes of cold (-20 C or When the reaction reached equilibrium, the mixture lower) acetone with vigorous agitation. The resultant was applied on a DEAE-Sephadex (A-25-120) column cell precipitate was collected by filtration through (1.5 by 20 cm) equilibrated with 0.01 M NH4HCO8, Whatman 3 MM paper. Step 1. Crude extract. Acetone-dried powder eluted according to the procedure of Allison et al. (1), and radioactivity and optical density at 260 and 340 (800 g) prepared from 3.1 kg of wet cells was susnm of the effluent were determined. [4B-3H]NADH pended in 10 volumes of 0.05 M Tris-chloride, con(0.62 lsmol; 9.4 x 105 counts/min) thus prepared was taining 0.01 M ethylenediaminetetraacetate, pH 8.0, then reacted with 3 ,umol of oxaloacetate in the and stirred for 18 h at 4 C, and then the insoluble presence of Pseudomonas enzyme in 0.01 M debris was removed by centrifugation. Step 2. Ammonium sulfate fractionation. Solid NH4HCO,. When oxidation of the [4B-3H]NADH ceased, the reaction mixture was lyophilized, the ammonium sulfate was added to the crude extract NAD + produced was isolated by the above chro- from step 1 to give 48% saturation. The degree of matographic procedure, and its radioactivity was saturation was based on the solubility of ammonium determined. sulfate at 25 C. The solution was stirred for 4 h at 4 C Kinetic data processing. The enzyme kinetic and centrifuged, and then more ammonium sulfate constants such as Michaelis constants, maximum was added to the supernatant to give 80% saturation. velocities, and inhibition constants were determined The precipitate was collected by centrifugation, disby employing the Fortran computer program devel- solved in 0.05 M Tris-chloride, pH 8.0, and slowly oped by Cleland (6, 7), with the use of a Control Data brought to 68% ammonium sulfate saturation by Corp. 3600 digital computer. dialyzing against an ammonium sulfate saturation by To determine Michaelis constants and maximum dialyzing against an ammonium sulfate solution in velocities for substrates, coenzymes, and their respec- 0.05 M Tris-chloride, pH 8.0, which gave the desired
VOL. 123, 1975
MALATE DEHYDROGENASE FROM PSEUDOMONAS
707
which corresponds to a catalytic constant of 2.00 x 105/mol of the enzyme on the basis of the molecular weight of 74,200 (below). The purified enzyme showed a single protein band coincident with the activity band in agar gel electrophoresis (Fig. 2). Figure 2 also presents the pattern of starch gel electrophoresis of the bacterial crude extract, which showed a single activity spot, thereby excluding the possibility of the presence of isoenzymes or conformers (17) in the organism. Homogeneity of the purified enzyme was further demonstrated by the following observations: a symmetrical boundary traversing to the bottom of the cell during a sedimentation velocity run (Fig. 3); a single precipitin band on the Ouchterlony plate between the antiserum against the enzyme and the purified enzyme or crude extract (Fig. 4); and a single protein band (subunit) in SDS-polyacrylamide gel electrophoresis (Fig. 5). Figure 3 also shows a slight increase in the sedimentation coefficient S20, of the enzyme upon dilution. Extrapolation of S20.W values to infinite dilution of protein concentration gave a sedimentation coefficient S 2o2wof 4.50 S. The diffusion coefficient D20,w of the enzyme at a fixed protein concentration of 2.1 mg/ml was determined to be 5.39 x 10-f cm2/s by the synthetic boundary run. The enzyme solution in 0.02 M potassium phosphate, pH 6.5, gave an extinction coefficient A" 2nm of 9.00, which corresponds to a molar extinction coefficient of 6.66 x 104 M-1 RESULTS cm-'. The absorbance ratio A280A260 of the Purification and physicochemical enzyme was 1.81 at pH 8.8. The following properties. The specific activities and the ac- information weight of the enzyme of in tivity recoveries the enzyme the purifica- was obtainedonbymolecular 74,200 by techniques: various tion steps are summarized in Table 1. Figure 1 sedimentation equilibrium runs at protein conshows the photograph of the enzyme crystals. centrations between 0.093 to 0.370 mg/ml; The specific activity of the enzyme under the 75,000 by the Svedberg equation employing the standard assay conditions was 2,700 Amol of sedimentation and diffusion coefficients deterNADH oxidized per min per mg of protein, mined at the protein concentration of 2.1 mg/ ml; 70,000 by the analytical Sephadex G-100 TABLE 1. Summary of purification of Pseudomonas column chromatography. The partial specific malate dehydrogenase volume of the enzyme was determined to be 0.732 cm3/g. Total Total Activity Figure 5 shows the banding pattern of the protein enzyte recovery Sp act Steps enzyme in SDS-polyacrylamide gel electropho(g) (X 10-3) resis along with authentic protein markers of 6 known molecular weight. The enzyme moved as 100 246 1. Crude extract 1,458 83 88 13.8 2. 48 to 68% am1,208 a single band with a mobility corresponding to a monium sulfate molecular weight of 36,000, a value approxi365 542 37 1.485 3. DEAE-11 cellumately half of that obtained from ultracentrifulose column 33 393 gation or gel filtration. It is therefore concluded 0.157 4. Sephadex G-100 2.506 column that the Pseudomonas enzyme is made up of 26 373 0.135 5. Crystallization 2,762 two subunits of identical size. Table 2 contains
saturation when equilibrium was reached. The precipitate was dissolved in 0.005 M Tris-chloride, pH 8.0, and dialyzed against the same buffer. Step 3. DEAE-ll cellulose column chromatography. The dialyzed solution from step 2 was applied on a DEAE-11 cellulose column (5.0 by 70 cm) previously equilibrated with 0.005 M Tris-chloride, pH 8.0. The column was washed with the equilibration buffer until the effluent did not show any appreciable absorption at 280 nm. The enzyme was then eluted with a linear gradient salt concentration consisting of 3 liters of the equilibration buffer in a mixing chamber and the same volume of the buffer containing 0.3 M NaCl in a reservoir chamber. Fractions containing the enzyme activity were pooled and proteins were precipitated by dialysis against saturated ammonium sulfate solution. The precipitate was dissolved in 0.05 M Tris-chloride, pH 8.0, and followed by a 53 to 68% ammonium sulfate fractionation by the dialysis method described in step 2. Step 4. Sephadex G-100 column chromatography. The proteins precipitated at 53 to 68% ammonium sulfate saturation were dissolved in a minimum volume of 0.05 M Tris-chloride, pH 8.0, layered on a Sephadex G-100 column (5 by 80 cm), equilibrated with 0.05 M Tris-chloride, pH 8.0, and eluted with the equilibration buffer. The fractions having the enzyme activity were pooled. Step 5. Crystallization. The pooled enzyme solution was brought to 53% ammonium sulfate saturation and the precipitated inactive proteins were removed by centrifugation. Additional ammonium sulfate was added a few crystals at a time until slight turbidity was observed. This solution was stored at 4 C, whereupon crystallization occurred in 12 h.
708
YOU AND KAPLAN
J. BACTERIOL.
4~~~ zf ,4
