Investigation of a catalytic zinc binding site in Escherichia coli L-threonine dehydrogenase by site-directed mutagenesis of cysteine-38.

L-Threonine dehydrogenase catalyzes the NAD+-dependent oxidation of threonine forming 2-amino-3-ketobutyrate. Chemical modification of Cys-38 of Escherichia coli threonine dehydrogenase, whose residue aligns with the catalytic zinc-binding residue, Cys-46, of related alcohol/polyol dehydrogenases, inactivates the enzyme [B. R. Epperly and E. E. Dekker (1991) J. Biol. Chem. 266, 6086-6092; A. R. Johnson and E. E. Dekker (1996) Protein Sci., 382-390]. To probe its function, Cys-38 was changed to Ser, Asp, and Glu by site-directed mutagenesis. Mutants C38S and C38D were purified to homogeneity and found to be, like the wild-type enzyme, homotetrameric proteins containing one Zn2+ atom per subunit. The circular dichroism spectra of these mutants were essentially identical to that of the wild-type enzyme. Mutant C38S was catalytically inactive but mutant C38D had a specific activity of 0.2 unit/mg, a level approximately 1% that of the wild-type enzyme. After it was incubated with 1 mM Zn2+ and then assayed in the presence of 15 mM Zn2+, mutant C38S showed only a trace of enzymatic activity (i.e., 0.013 unit/mg). Preincubation of mutant C38D with 5 mM Zn2+, Co2+, or Cd2+ increased its activity 57-, 6-, or 3-fold, respectively; 1 mM Mn2+ halved and 0.5 mM Hg2+ abolished activity. Zn2+-stimulated mutant C38D showed these properties: apparent substrate activation at low threonine concentrations, a maximum activity of 27 units/mg with 20 mM threonine, and inhibition by high levels of substrate; an activation Kd = 3 mM Zn2+; and a pH optimum of 8.4 (in contrast to pH 10.3 for the wild-type enzyme). Without added Zn2+, mutant C38D is equally active with threonine and 2-amino-3-hydroxypentanoate, but Zn2+-activated mutant C38D is 10-fold more reactive with threonine than with 2-amino-3-hydroxypentanoate. In the absence of added metal ions, wild-type enzyme similarly uses substrates other than threonine and shows a dramatic increase in activity with only threonine when stimulated by either Cd2+ or Mn2+; added Zn2+ has no effect on activity with threonine. Cys-38 of threonine dehydrogenase, therefore, is located in an activating divalent metal ion-binding site. Having a negatively charged residue like Asp in this position allows the binding of a catalytic Zn2+ ion which enhances activity with threonine and reduces activity with substrate analogs. Whether Cys-38 of wild-type threonine dehydrogenase binds a catalytic metal ion (possibly Zn2+) in vivo remains to be established.

Site-directed mutagenesis of histidine-90 in Escherichia coli L-threonine dehydrogenase alters its substrate specificity.

Escherichia coli L-threonine dehydrogenase is a member of the Zn(2+)-containing alcohol/polyol dehydrogenase family. Methylation of His-90 of L-threonine dehydrogenase was recently found to cause total inactivation (J. P. Marcus and E. E. Dekker, 1995 Arch. Biochem. Biophys. 316, 413-420). Since His-90 is not conserved among the related dehydrogenases, this residue was changed to arginine, asparagine, and alanine by site-directed mutagenesis in order to probe its role. All three purified, homogeneous mutants, like wild-type enzyme, contained one Zn2+ atom/subunit and exhibited a sequential catalytic mechanism; the kcat value for each, however, was reduced approximately 10-fold. The K(m) value for threonine was elevated from 3 mM for wild-type enzyme to 31, 328, and 417 mM, respectively, for mutants H90R, H90N, and H90A. The activation energy of catalysis for mutant H90A was increased by 6.6 kcal/mol, suggesting that in the wild-type enzyme His-90 forms at least one crucial hydrogen bond in the transition state. Whereas wild-type enzyme catalyzed the oxidation of threonine amide (0.75 M) about twice as fast as this same concentration of threonine or 0.375 M L-2-amino-3-hydroxypentanoate, the reaction rate of mutant H90A with 0.75 M threonine amide or threonine methyl ester was 33- to 35-fold higher than with this level of threonine. Similarly, mutant H90N used 0.75 M threonine methyl ester or threonine amide as substrate 9- to 13-fold better than it used this concentration of threonine. Mutants H90A and H90N were more reactive with 0.225 M L-threonine hydroxamate than with 0.75 M threonine, but mutant H90A did not oxidize L-2-amino-3-hydroxypentanoate (0.375 M) and mutant H90N used this substrate poorly. The best substrates for mutant H90R were threonine methyl ester, threonine, and threonine amide (all tested at 0.75 M); 0.375 M L-2-amino-3-hydroxypentanoate was a poor substrate. The isolation and characterization of these first His-90 mutants of E. coli L-threonine dehydrogenase confirm the importance of this residue in catalysis and suggest that His-90 is an active-site residue which modulates the substrate specificity of L-threonine dehydrogenase.

Woodward's reagent K inactivation of Escherichia coli L-threonine dehydrogenase: increased absorbance at 340-350 nm is due to modification of cysteine and histidine residues, not aspartate or glutamate carboxyl groups.

L-Threonine dehydrogenase (TDH) from Escherichia coli is rapidly inactivated and develops a new absorbance peak at 347 nm when incubated with N-ethyl-5-phenylisoxazolium-3'-sulfonate (Woodward's reagent K, WRK). The cofactors, NAD+ or NADH (1.5 mM), provide complete protection against inactivation; L-threonine (60 mM) is approximately 50% as effective. Tryptic digestion of WRK-modified TDH followed by HPLC fractionation (pH 6.2) yields four 340-nm-absorbing peptides, two of which are absent from enzyme incubated with WRK and NAD+. Peptide I has the sequence TAICGTDVH (TDH residues 35-43), whereas peptide II is TAICGTDVHIY (residues 35-45). Peptides not protected are TMLDTMNHGGR (III, residues 248-258) and NCRGGRTHLCR (IV, residues 98-108). Absorbance spectra of these WRK-peptides were compared with WRK adducts of imidazole, 2-hydroxyethanethiolate, and acetate. Peptides III and IV have pH-dependent lambda max values (340-350 nm), consistent with histidine modification. Peptide I has pH-independent lambda max (350 nm) indicating that a thiol is modified. WRK, therefore, does not react specifically with carboxyl groups in this enzyme, but rather modifies Cys-38 in the active site of TDH; modification of His-105 and His-255 does not affect enzyme activity. These results are the first definitive proof of WRK modifying cysteine and histidine residues of a protein and show that enzyme inactivation by WRK associated with the appearance of new absorptivity at 340-350 nm does not establish modification of aspartate or glutamate residues, as has been assumed in numerous earlier reports.

Functional analysis of E. coli threonine dehydrogenase by means of mutant isolation and characterization.

The oxidation of L-threonine to 2-amino-ketobutyrate, as catalyzed by L-threonine dehydrogenase, is the first step in the major pathway for threonine catabolism in both eukaryotes and prokaryotes. Threonine dehydrogenase of E. coli has considerable amino-acid sequence homology with a number of Zn(2+)-containing, medium-chain alcohol dehydrogenases. In order to further explore structure/function interrelationships of E. coli threonine dehydrogenase, 35 alleles of tdh that imparted a no-growth or slow-growth phenotype on appropriate indicator media were isolated after mutagenesis with hydroxylamine. Within this collection, 14 mutants had single amino-acid changes that were divided into 4 groups: (a) amino-acid changes associated with proposed ligands to Zn2+; (b) a substitution of one of several conserved glycine residues; (c) mutations at the substrate or coenzyme binding site; (d) alterations that resulted in a change of charge near the active site. These findings uncover previously unidentified amino-acid residues that are important for threonine dehydrogenase catalysis and also indicate that the three-dimensional structure of tetrameric E. coli threonine dehydrogenase has considerable similarity with the dimeric horse liver alcohol dehydrogenase.

Identification of a second active site residue in Escherichia coli L-threonine dehydrogenase: methylation of histidine-90 with methyl p-nitrobenzenesulfonate.

Incubation of L-threonine dehydrogenase from Escherichia coli with methyl p-nitrobenzenesulfonate results in a time- and concentration-dependent loss of enzymatic activity. As the concentration of the methylating agent is increased, the rate of inactivation reaches a limiting value of 0.01 min-1 at pH 7.0 and 25 degrees C, suggesting that the inhibitor is binding at a specific site prior to reaction. Approximately one methyl group is incorporated per enzyme subunit inactivated. Reaction with [14C]methyl p-nitrobenzenesulfonate followed by amino acid analysis shows that greater than 90% of the radioactivity incorporated into the enzyme is associated with a peak that coelutes with 3-methyl-N-histidine. Tryptic digestion of the inactive enzyme adduct yields a radioactive peptide corresponding to residues 85-97 of the protein; the radioactivity is associated with histidine residue-90. The Zn2+ content of the inactivated and the native enzyme remains the same. The substrate, L-threonine, and substrate analogs, L-threonine methyl ester and L-threonine amide, provide about 60% protection against inactivation, whereas NAD+ has no effect. In contrast, NADH markedly enhances the rate of inactivation by this methylating agent, suggesting a possible conformational change in the vicinity of His-90 is induced by binding of the coenzyme.

Threonine formation via the coupled activity of 2-amino-3-ketobutyrate coenzyme A lyase and threonine dehydrogenase.

The enzymes L-threonine dehydrogenase and 2-amino-3-ketobutyrate coenzyme A (CoA) lyase are known to catalyze the net conversion of L-threonine plus NAD+ plus CoA to NADH plus glycine plus acetyl-CoA. When homogeneous preparations of these two enzymes from Escherichia coli were incubated together for 40 min at 25 degrees C with glycine, acetyl-CoA, and NADH, a 36% decrease in the level of glycine (with concomitant NADH oxidation) was matched by formation of an equivalent amount of threonine, indicating that this coupled sequence of enzyme-catalyzed reactions is reversible in vitro. Several experimental factors that affect the efficiency of this conversion in vitro were examined. A constructed strain of E. coli, MD901 (glyA thrB/C tdh), was unable to grow unless both glycine and threonine were added to defined rich medium. Introduction of the plasmid pDR121 (tdh+kbl+) into this strain enabled the cells to grow in the presence of either added glycine or threonine, indicating that interconversion of these two amino acids occurred. Threonine that was isolated from the total pool of cellular protein of MD901/pDR121 had the same specific radioactivity as the [14C]glycine added to the medium, establishing that threonine was formed exclusively from glycine in this strain. Comparative growth rate studies with several strains of E. coli containing plasmid pDR121, together with the finding that kcat values of pure E. coli 2-amino-3-ketobutyrate CoA lyase favor the cleavage of 2-amino-3-ketobutyrate over its formation by a factor of 50, indicate that the biosynthesis of threonine is less efficient than glycine formation via the coupled threonine dehydrogenase-2-amino-3-ketobutyrate lyase reactions.

Identity and some properties of the L-threonine aldolase activity manifested by pure 2-amino-3-ketobutyrate ligase of Escherichia coli.

2-Amino-3-ketobutyrate ligase catalyzes the reversible, pyridoxal 5'-phosphate-dependent condensation of glycine with acetyl CoA forming the unstable intermediate, 2-amino-3-ketobutyrate. Several independent lines of evidence indicate that the pure protein obtained in the purification of this ligase from Escherichia coli also has L-threonine aldolase activity. The evidence includes: (a), a constant ratio of specific activities (aldolase/ligase) at all stages of purifying 2-amino-3-ketobutyrate ligase to homogeneity; (b), the same rate of loss of aldolase and ligase activities during controlled heat inactivation of the pure protein at 60 degrees C in the absence, as well as in the presence of acetyl CoA, a protective substrate; (c), ratios of the two enzymatic activities that are not significantly different during slow inactivation by iodoacetamide, with and without L-threonine added; (d), coincident rates of loss and essentially identical rates of recovery of aldolase activity and ligase activity during resolution of the holoenzyme with hydroxylamine followed by reconstitution with pyridoxal 5'-phosphate. No aldolase activity is observed with D-threonine as substrate and L-allothreonine is about 25% as effective as L-threonine. Whereas ligase activity has a sharp pH optimum at 7.5, the aldolase activity of this pure protein is maximal at pH 9.0. Comparative apparent Km values for glycine (ligase) and L-threonine (aldolase) are 10 mM and 0.9 mM, respectively, whereas corresponding respective Vmax values were found to be 2.5 mumol of CoA released/min per mg vs. 0.014 mumol of acetaldehyde formed (NADH oxidized)/min per mg.

pH-dependent decarboxylation of 2-amino-3-ketobutyrate, the unstable intermediate in the threonine dehydrogenase-initiated pathway for threonine utilization.

2-Amino-3-ketobutyrate can be readily formed enzymatically by the action of L-threonine dehydrogenase. A convenient assay for determining the half-life of this beta-keto acid is afforded by its rapid and quantitative conversion to glycine (+ acetyl CoA), as catalyzed by 2-amino-3-ketobutyrate CoA lyase. Using this system, we have found the half-life of 2-amino-3-ketobutyrate varies with pH from 8.6 minutes at pH 5.9 to 140 minutes at pH 11.1 yielding a theoretical titration curve that predicts a pKa value of 8.15 for the alpha-amino group of this intermediate. These data are considered relevant to discussions pertaining to a threonine dehydrogenase/2-amino-3-ketobutyrate CoA lyase enzyme complex in the threonine utilization pathway and to mechanistic aspects of the 5-aminolevulinate synthase-catalyzed reaction where 2-amino-3-ketoadipate is involved.

Inactivation of Escherichia coli 2-amino-3-ketobutyrate CoA ligase by phenylglyoxal and identification of an active-site arginine peptide.

Treatment of homogeneous preparations of 2-amino-3-ketobutyrate CoA ligase from Escherichia coli, a pyridoxal 5'-phosphate-dependent enzyme, with phenylglyoxal, 4-(oxyacetyl)phenoxyacetic acid, 2,3-butanedione, or 1,2-cyclohexanedione results in a time- and concentration-dependent loss of enzymatic activity. Phenylglyoxal in 50 mM phosphate buffer (pH 7.0) is the most effective modifier, causing > 95% inactivation within 20 min at 25 degrees C. Controls establish that this inactivation is not due to modifier-induced dissociation or photoinduced nonspecific alteration of the ligase. The substrate, acetyl CoA, or the coenzyme, pyridoxal 5'-phosphate, gives > 50% protection against inactivation. Enzyme partially inactivated by phenylglyoxal has the same Km value for glycine but the Vmax decreases in proportion to the observed level of inactivation. Whereas the native apoligase shows good recovery of activity with time in parallel with an increase in 428-nm absorptivity when incubated with pyridoxal 5'-phosphate, no such effects are seen with the phenylglyoxal-modified apoligase. Reaction of the enzyme with [14C]phenylglyoxal allowed for the isolation of a peptide which, by amino acid composition and sequencing data, was found to correspond to residues 349-378 in the intact enzyme. These results indicate that arginine residue-366 and/or residue-368 in the primary structure of E. coli 2-amino-3-ketobutyrate ligase is at the active site.

2-Keto-4-hydroxyglutarate aldolase: purification and characterization of the homogeneous enzyme from bovine kidney.

2-Keto-4-hydroxyglutarate aldolase, which catalyzes the reversible cleavage of 2-keto-4-hydroxyglutarate, yielding pyruvate plus glyoxylate, has been purified from extracts of bovine kidney to apparent homogeneity as judged by polyacrylamide gel electrophoresis, gel filtration chromatography, sucrose density gradient centrifugation, and meniscus depletion sedimentation equilibrium experiments. The enzyme from this source has a native and a subunit mass of 144 and 36 kDa, respectively; the pH-activity optimum is 8.8. Rather than being stimulated, aldolase activity is inhibited to varying degrees by added divalent metal ions, whereas a number of metal ion-chelating agents have no effect. An absolute requirement for added thiol compounds could not be shown, but 2-mercaptoethanol enhances activity 2-fold, and added Hg2+ as well as p-mercuribenzoate or dithiodipyridine markedly inhibit catalysis. Incubation of the enzyme with either pyruvate or glyoxylate in the presence of NaBH4 causes extensive loss of aldolase activity concomitant with stable binding of approximately 1.0-1.5 mol of 14C-labeled substrate/mol of enzyme. The circular dichroism spectrum for native aldolase is characteristic of an alpha-helix; incubation of the enzyme with glyoxylate has no effect on this spectrum, but it is considerably altered by pyruvate. Bovine kidney aldolase shows no stereospecificity in catalyzing the aldol cleavage of the two optical isomers of 2-keto-4-hydroxyglutarate, and although it also catalyzes the beta-decarboxylation of oxalacetate, its decarboxylase/aldolase activity ratio is lower than that seen with the pure enzyme from either bovine liver or Escherichia coli.

Cloning, nucleotide sequence, overexpression, and inactivation of the Escherichia coli 2-keto-4-hydroxyglutarate aldolase gene.

Having previously determined the complete amino acid sequence of 2-keto-4-hydroxyglutarate aldolase from Escherichia coli (C. J. Vlahos and E. E. Dekker, J. Biol. Chem. 263:11683-11691, 1988), we amplified the gene that codes for this enzyme by the polymerase chain reaction using synthetic degenerate deoxyoligonucleotide primers. The amplified DNA was sequenced by subcloning the polymerase chain reaction products into bacteriophage M13; the nucleotide sequence of the gene was found to be in exact agreement with the amino acid sequence of the gene product. Overexpression of the gene was accomplished by cloning it into the pKK223.3 expression vector so that it was under control of the tac promoter and then using the resultant plasmid, pDP6, to transform E. coli DH5 alpha F'IQ. When this strain was grown in the presence of isopropyl beta-D-thiogalactopyranoside, aldolase specific activity in crude extracts was 80-fold higher than that in wild-type cells and the enzyme constituted approximately 30% of the total cellular protein. All properties of the purified, cloned gene product, including cross-reactivity with antibodies elicited against the wild-type enzyme, were identical with the aldolase previously isolated and characterized. A strain of E. coli in which this gene is inactivated was prepared for the first time by insertion of the kanamycin resistance gene cartridge into the aldolase chromosomal gene.

L-threonine dehydrogenase from Escherichia coli. Identification of an active site cysteine residue and metal ion studies.

Pure L-threonine dehydrogenase from Escherichia coli is a tetrameric protein (Mr = 148,000) with 6 half-cystine residues/subunit; its catalytic activity as isolated is stimulated 5-10-fold by added Mn2+ or Cd2+. The peptide containing the 1 cysteine/subunit which reacts selectively with iodoacetate, causing complete loss of enzymatic activity, has been isolated and sequenced; this cysteine residue occupies position 38. Neutron activation and atomic absorption analyses of threonine dehydrogenase as isolated in homogeneous form now show that it contains 1 mol of Zn2+/mol of enzyme subunit. Removal of the Zn2+ with 1,10-phenanthroline demonstrates a good correlation between the remaining enzymatic activity and the zinc content. Complete removal of the Zn2+ yields an unstable protein, but the native metal ion can be exchanged by either 65Zn2+, Co2+, or Cd2+ with no change in specific catalytic activity. Mn2+ added to and incubated with the native enzyme, the 65Zn2(+)-, the Co2(+)-, or the Cd2(+)- substituted form of the enzyme stimulates dehydrogenase activity to the same extent. These studies along with previously observed structural homologies further establish threonine dehydrogenase of E. coli as a member of the zinc-containing long chain alcohol/polyol dehydrogenases; it is unique among these enzymes in that its activity is stimulated by Mn2+ or Cd2+.

4-methyleneglutamine amidohydrolase from peanut leaves : preparation, catalytic properties, and immunological responses of a highly purified form of the enzyme.

4-Methyleneglutamine amidohydrolase has been extracted and purified over 1000-fold from 14-day-old peanut (Arachis hypogaea) leaves by modification of methods described previously. The purified enzyme shows two bands of activity and three to four bands of protein after electrophoresis on nondenaturing gels. Each of the active bands is readily eluted from gel slices and migrates to its original position on subsequent electrophoresis. Although they are electrophoretically distinct, the two forms of the enzyme are immunologically identical by Ouchterlony double-diffusion techniques and have similar catalytic properties. Activity toward glutamine that has a threefold lower V(max) and a four-fold higher K(m) value copurifies with MeGln aminohydrolase activity. 4-Methyleneglutamine and 4-methyleneglutamic acid inhibit the hydrolysis of glutamine while glutamine inhibits 4-methyleneglutamine hydrolysis, further indicating the identity of the activity toward both substrates. Amidohydrolase activity is stimulated up to threefold by preincubation with either ionic or non-ionic detergents (0.1%) and also by added proteins (0.5% bovine serum albumin or whole rabbit serum); it is inhibited 50% by 1 millimolar borate or the glutamine analog, albizziin (10 millimolar). Rabbit antiserum to the purified peanut enzyme cross-reacts with one or more proteins in extracts of some plants but not others; in no instance, however, was 4-methyleneglutamine amidohydrolase activity detected in other species. Overall, the results support the hypothesis that 4-methyleneglutamine supplies N, via its hydrolysis by the amidohydrolase, to the growing shoots of peanut plants, whereas glutamine hydrolysis is prevented by the prepon-derance of the preferred substrate. Some results also suggest that this amidohydrolase activity may be regulated by metabolites and/or by association with other cellular components.

Active-site residues of 2-keto-4-hydroxyglutarate aldolase from Escherichia coli. Bromopyruvate inactivation and labeling of glutamate 45.

Treatment of pure 2-keto-4-hydroxyglutarate aldolase from Escherichia coli, a "lysine-type," Schiff-base mechanism enzyme, with the substrate analog bromopyruvate results in a time- and concentration-dependent loss of enzymatic activity. Whereas the substrates pyruvate and 2-keto-4-hydroxyglutarate provide greater than 90% protection against inactivation by bromopyruvate, no protective effect is seen with glycolaldehyde, an analog of glyoxylate. Inactivation studies with [14C] bromopyruvate show the incorporation of 1.1 mol of 14C-labeled compound/enzyme subunit; isolation of a radioactive peptide and determination of its amino acid sequence indicate that the radioactivity is associated with glutamate 45. Incubation of the enzyme with excess [14C]bromopyruvate followed by denaturation with guanidine.HCl allow for the incorporation of carbon-14 at cysteines 159 and 180 as well. Whereas the presence of pyruvate protects Glu-45 from being esterified, it does not prevent the alkylation of these 2 cysteine residues. The results indicate that Glu-45 of E. coli 2-keto-4-hydroxyglutarate aldolase is essential for catalytic activity, most likely acting as the amphoteric proton donor/acceptor that is required as a participant in the overall mechanism of the reaction catalyzed.

2-Amino-3-ketobutyrate CoA ligase of Escherichia coli: stoichiometry of pyridoxal phosphate binding and location of the pyridoxyllysine peptide in the primary structure of the enzyme.

Pure 2-amino-3-ketobutyrate CoA ligase from Escherichia coli, which catalyzes the cleavage/condensation reaction between 2-amino-3-ketobutyrate (the presumed product of the L-threonine dehydrogenase-catalyzed reaction) and glycine + acetyl-CoA, is a dimeric enzyme (Mr = 84,000) that requires pyridoxal 5'-phosphate as coenzyme for catalytic activity. Reduction of the hololigase with tritiated NaBH4 yields an inactive, radioactive enzyme adduct; acid hydrolysis of this adduct allowed for the isolation and identification of epsilon-N-pyridoxyllysine. Quantitative determinations established that 2 mol of pyridoxal 5'-phosphate are bound per mol of dimeric enzyme. After the inactive, tritiated enzyme adduct was digested with trypsin, a single radioactive peptide containing 23 amino acids was isolated and found to have the following primary structure: Val-Asp-Ile-Ile-Thr-Gly-Thr-Leu-Gly-Lys*-Ala-Leu-Gly-Gly-Ala-Ser-Gly-Gly -Tyr-Thr-Ala-Ala-Arg (where * = the lysine residue in azomethine linkage with pyridoxal 5'-phosphate). This peptide corresponds to residues 235-257 in the intact protein; 10 residues around the lysine residue have a high level of homology with a segment of the primary structure of 5-aminolevulinate synthase from chicken liver.

The sulfhydryl content of L-threonine dehydrogenase from Escherichia coli K-12: relation to catalytic activity and Mn2+ activation.

When oxidized to cysteic acid by performic acid or converted to carboxymethylcysteine by alkylation of the reduced enzyme with iodoacetate, a total of six half-cystine residues/subunit are found in L-threonine dehydrogenase (L-threonine: NAD+ oxidoreductase, EC 1.1.1.103; L-threonine + NAD(+)----2-amino-3-oxobutyrate + NADH) from Escherichia coli K-12. Of this total, two exist in disulfide linkage, whereas four are titratable under denaturing conditions by dithiodipyridine, 5,5'-dithiobis(2-nitrobenzoic acid), or p-mercuribenzoate. The kinetics of enzyme inactivation and of modification by the latter two reagents indicate that threonine dehydrogenase has no free thiols that selectively react with bulky compounds. While incubation of the enzyme with a large excess of iodoacetamide causes less than 10% loss of activity, the native dehydrogenase is uniquely reactive with and completely inactivated by iodoacetate. The rate of carboxymethylation by iodoacetate of one -SH group/subunit is identical with the rate of inactivation and the carboxymethylated enzyme is no longer able to bind Mn2+. NADH (0.5 mM) provides 40% protection against this inactivation; 60 to 70% protection is seen in the presence of saturating levels of NADH plus L-threonine. Such results coupled with an analysis of the kinetics of inactivation caused by iodoacetate are interpreted as indicating the inhibitor first forms a reversible complex with a positively charged moiety in or near the microenvironment of a reactive -SH group in the enzyme before irreversible alkylation occurs. Specific alkylation of one -SH group/enzyme subunit apparently causes protein conformational changes that entail a loss of catalytic activity and the ability to bind Mn2+.

Inactivation of Escherichia coli L-threonine dehydrogenase by 2,3-butanedione. Evidence for a catalytically essential arginine residue.

Incubation of homogeneous preparations of L-threonine dehydrogenase from Escherichia coli with 2,3-butanedione, 2,3-pentanedione, phenylglyoxal, or 1,2-cyclohexanedione causes a time- and concentration-dependent loss of enzymatic activity; plots of log percent activity remaining versus time are linear to greater than 90% inactivation, indicative of pseudo-first order inactivation kinetics. The reaction order with respect to the concentration of modifying reagent is approximately 1.0 in each case suggesting that the loss of catalytic activity is due to one molecule of modifier reacting with each active unit of enzyme. Controls establish that this inactivation is not due to modifier-induced dissociation or photoinduced nonspecific alteration of the dehydrogenase. Essentially the same Km but decreased Vmax values are obtained when partially inactivated enzyme is compared with native. NADH (25 mM) and NAD+ (70 mM) give full protection against inactivation whereas much higher concentrations (i.e. 150 mM) of L-threonine or L-threonine amide provide a maximum of 80-85% protection. Amino acid analyses coupled with quantitative sulfhydryl group determinations show that enzyme inactivated 95% by 2,3-butanedione loses 7.5 arginine residues (out of 16 total)/enzyme subunit with no significant change in other amino acid residues. In contrast, only 2.4 arginine residues/subunit are modified in the presence of 80 mM NAD+. Analysis of the course of modification and inactivation by the statistical method of Tsou (Tsou, C.-L. (1962) Sci. Sin. 11, 1535-1558) demonstrates that inactivation of threonine dehydrogenase correlates with the loss of 1 "essential" arginine residue/subunit which quite likely is located in the NAD+/NADH binding site.

Specificity of aspartate aminotransferases from leguminous plants for 4-substituted glutamic acids.

Aspartate aminotransferase (glutamate-oxalacetate transaminase) was partially purified from extracts of germinating seeds of peanut (Arachis hypogaea), honey locust (Gleditsia triacanthos), soybean (Glycine max), and Sophora japonica. The ability of these enzyme preparations, as well as aspartate aminotransferase purified from pig heart cytosol, to use 4-substituted glutamic acids as amino group donors and their corresponding 2-oxo acids as amino group acceptors in the aminotransferase reaction was measured. All 4-substituted glutamic acid analogs tested were poorer substrates than was glutamate or 2-oxoglutarate. 2-Oxo-4-methyleneglutarate was least effective (lowest relative V(m)/K(m)) as a substrate for the enzyme from peanuts and honey locust, which are the two species studied that accumulate 4-methyleneglutamic acid and 4-methyleneglutamine. Of the different aminotransferases tested, the enzyme from honey locust was the least active with 2-oxo-4-hydroxy-4-methylglutarate, the corresponding amino acid of which also accumulates in that species. These results suggest that transamination of 2-oxo-4-substituted glutaric acids is not involved in the biosynthesis of the corresponding 4-substituted glutamic acids in these species. Rather, accumulation of certain 4-substituted glutamic acids in these instances may be, in part, the result of the inefficacy of their transamination by aspartate aminotransferase.