A simple preparation method for apoaspartate aminotransferase from Escherichia coli B, and its application for the assay of pyridoxal and pyridoxamine 5'-phosphate

A simple and rapid preparation method for apoaspartate aminotransferase from Escherichia coli B was developed. A crude extract of the bacterial cells was treated batchwise with DEAE-cellulose. The enzyme fraction obtained was then applied to a pyridoxamine-Sepharose column. Apoaspartate aminotransferase was eluted with 50 mM potassium phosphate buffer (pH 7.0), and found to be electrophoretically homogeneous. The apoenzyme preparation thus obtained showed very low holoenzyme activity (only 0.4% of the activity seen in the fully saturated condition with pyridoxal 5'-phosphate) and was successfully used for assaying pyridoxal and pyridoxamine 5'-phosphate.     

Transaminations catalysed by brain glutamate decarboxylase.

In addition to normal decarboxylation of glutamate to 4-aminobutyrate, glutamate decarboxylase from pig brain was shown to catalyse decarboxylation-dependent transamination of L-glutamate and direct transamination of 4-aminobutyrate with pyridoxal 5'-phosphate to yield succinic semialdehyde and pyridoxamine 5'-phosphate in a 1:1 stoichiometric ratio. Both reactions result in conversion of holoenzyme into apoenzyme. With glutamate as substrate the rates of transamination differed markedly among the three forms of the enzyme (0.008, 0.012 and 0.029% of the rate of 4-aminobutyrate production by the alpha-, beta- and gamma-forms at pH 7.2) and accounted for the differences among the forms in rates of inactivation by glutamate and 4-aminobutyrate. Rates of transamination were maximal at about pH 8 and varied in parallel with the rate constants for inactivation from pH 6.5 to 8.0. Rates of transamination of glutamate and 4-aminobutyrate were similar, suggesting that the decarboxylation step is not entirely rate-limiting in the normal mechanism. The transamination was reversible, and apoenzyme could be reconstituted to holoenzyme by reverse transamination with succinic semialdehyde and pyridoxamine 5'-phosphate. As a major route of apoenzyme formation, the transamination reaction appears to be physiologically significant and could account for the high proportion of apoenzyme in brain.     

Dissociation of aspartate aminotransferase into subunits. Effect of ligands upon this dissociation.

Frontal and zonal analysis of the chromatography of aspartate aminotransferase (EC2.61.1), pig heart cytosolic enzyme, on Bio-Gel P150 shows that holo- and apoenzyme can dissociate at pH 8.3. Ultracentrifugation and fluorescence depolarization confirm this result. Kinetic analysis of the fluorescence depolarization experiments favors a biphasic phenomenon: a few minutes for the faster one and several hours for the slower one. The apparent dissociation constant is 0.8 muM for the apoenzyme and 0.18 muM for the pyridoxal 5'-phosphate form of the holoenzyme. In the presence of sucrose or 0.1 M L-aspartate or a mixture of 70 mM L-glutamate and 2 mM alpha-ketoglutarate, the holoenzyme is dimeric at concentrations higher than 5 nM. The addition of a mixture of the substrates L-glutamate and alpha-ketoglutarate to a monomeric holoenzyme leads to dimerization. The stability of the dimeric form is in the order: holoenzyme + substrates greater than apoenzyme.     

Activation of Glutamate Apodecarboxylase by Succinic Semialdehyde and Pyridoxamine 5''-Phosphate

Glutamate apodecarboxylase was activated by incubation with succinic semialdehyde and pyridoxamine 5'-phosphate. Activation required both compounds and was highly selective for succinic semialdehyde. Of 18 analogs tested, only glyoxylate, pyruvate, oxaloacetate, and 2-oxoglutarate activated the apoenzyme significantly, but much higher concentrations of these compounds than of succinic semialdehyde were required. In the presence of pyridoxamine 5'-phosphate, the concentration of succinic semialdehyde giving half-maximal activation of apoenzyme was 7 microM. In contrast, the Ki for succinic semialdehyde as a competitive inhibitor of glutamate decarboxylation was 1.2 mM, indicating that apoenzyme with bound pyridoxamine 5'-phosphate has a much higher affinity for succinic semialdehyde than does holoenzyme. The concentration of pyridoxamine 5'-phosphate giving half-maximal activation was 17 microM, which is more than an order of magnitude greater than the corresponding value for pyridoxal 5'-phosphate.     

Mitochondrial aspartate aminotransferase-independent function of the catalytic binding sites.

The enzyme mitochondrial aspartate aminotransferase from beef liver is a dimer of identical subunits. The enzymatic activity of the resolved enzyme is restored upon addition of the cofactor pyridoxal 5-phosphate. The binding of 1 molecule of cofactor restores 50% of the original enzymatic activity, whereas the binding of a 2nd molecule of cofactor brings about more than 95% recovery of the catalytic activity. Following addition of 1 mol of pyridoxal-5-P per dimer, three forms of the enzyme may exist in solution: apoenzyme-2 pyridoxal 5'-phosphate, apoenzyme-1 pyridoxal 5'-phosphate, and apoenzyme. The enzyme species are separated by affinity chromatography and the following distribution was found: apoenzyme-2 pyridoxal 5'-phosphate/apoenzyme-1 pytidoxal 5'-phosphate/apoenzyme, 2/6/2. Similar distribution was observed after reduction with NaBH4 of the mixture containing apoenzyme and pyridoxal-5-P at a mixing ratio of 1:1. Fluorometric titrations conducted on samples of apoenzyme and apoenzyme-1 pyridoxal 5'-phosphate reveal that the enzyme species display identical affinity towards the inhibitor 4-pyridoxic-5-P (KD equals 1.1 times 10- minus 6 M). It is concluded that the binding of the cofactor to one of the catalytic sites does not affect the affinity of the second site for the inhibitor. These results, obtained by two independent methods, lend strong support to the hypothesis that the two subunits of the enzyme function independently.     

Regulatory properties of brain glutamate decarboxylase

1. Glutamate decarboxylase is a focal point for controlling gamma-aminobutyric acid (GABA) synthesis in brain. Several factors that appear to be important in the regulation of GABA synthesis have been identified by relating studies of purified glutamate decarboxylase to conditions in vivo. 2. The interaction of glutamate decarboxylase with its cofactor, pyridoxal 5'-phosphate, is a regulated process and appears to be one of the major means of controlling enzyme activity. The enzyme is present in brain predominantly as apoenzyme (inactive enzyme without bound cofactor). Studies with purified enzyme indicate that the relative amounts of apo- and holoenzyme are determined by the balance in a cycle that continuously interconverts the two. 3. The cycle that interconverts apo- and holoenzyme is part of the normal catalytic mechanism of the enzyme and is strongly affected by several probable regulatory compounds including pyridoxal 5'-phosphate, ATP, inorganic phosphate, and the amino acids glutamate, GABA, and aspartate. ATP and the amino acids promote apoenzyme formation and pyridoxal 5'-phosphate and inorganic phosphate promote holoenzyme formation. 4. Numerous studies indicate that brain contains multiple molecular forms of glutamate decarboxylase. Multiple forms that differ markedly in kinetic properties including their interactions with the cofactor have been isolated and characterized. The kinetic differences among the forms suggest that they play a significant role in the regulation of GABA synthesis.     

Structural changes accompanying the removal of pyridoxal 5'-phosphate from phosphorylase b.

The changes in physical properties accompanying the removal of pyridoxal 5'-phosphate from glycogen phosphorylase b have been examined. The apoenzyme retains a high degree of structural rigidity, as determined from the time decay of anisotropy. The bulk of the secondary structure remains intact, although a significant change in circular dichroism indicates some degree of alteration. The mobility of a sulfhydryl-linked spin label increases. The restoration of pyridoxal 5'-phosphate reverses this effect, with indication of interaction between subunits. One or more new binding sites for 1-anilinonaphthalene-8-sulfonate appear for the apoenzyme. The kinetics of the recombination of pyridoxal 5'-phosphate with the apoenzyme, as monitored by difference spectra, indicate a high activation energy for the process. The apoenzyme is a reversibly associating system at 20-30 degrees C, pH 7.0.     

Structures of apo- and holo-tyrosine phenol-lyase reveal a catalytically critical closed conformation and suggest a mechanism for activation by K+ ions

Tyrosine phenol-lyase, a tetrameric pyridoxal 5'-phosphate dependent enzyme, catalyzes the reversible hydrolytic cleavage of L-tyrosine to phenol and ammonium pyruvate. Here we describe the crystal structure of the Citrobacter freundii holoenzyme at 1.9 A resolution. The structure reveals a network of protein interactions with the cofactor, pyridoxal 5'-phosphate, and details of coordination of the catalytically important K+ ion. We also present the structure of the apoenzyme at 1.85 A resolution. Both structures were determined using crystals grown at pH 8.0, which is close to the pH of the maximal enzymatic activity (8.2). Comparison of the apoenzyme structure with the one previously determined at pH 6.0 reveals significant differences. The data suggest that the decrease of the enzymatic activity at pH 6.0 may be caused by conformational changes in the active site residues Tyr71, Tyr291, and Arg381 and in the monovalent cation binding residue Glu69. Moreover, at pH 8.0 we observe two different active site conformations: open, which was characterized before, and closed, which is observed for the first time in beta-eliminating lyases. In the closed conformation a significant part of the small domain undergoes an extraordinary motion of up to 12 A toward the large domain, closing the active site cleft and bringing the catalytically important Arg381 and Phe448 into the active site. The closed conformation allows rationalization of the results of previous mutational studies and suggests that the observed active site closure is critical for the course of the enzymatic reaction and for the enzyme's specificity toward its physiological substrate. Finally, the closed conformation allows us to model keto(imino)quinonoid, the key transition intermediate.     

Purification and properties of a pyridoxal 5'-phosphate-dependent histidine decarboxylase from Morganella morganii AM-15

A pyridoxal 5'-phosphate-dependent histidine decarboxylase from Morganella morganii AM-15 was purified to homogeneity. The enzyme is a tetramer (Mr 170,000) of identical subunits and binds 4 pyridoxal-P/tetramer; it is resolved by dialysis against cysteine at pH 6.8. Between pH 6.2 and 8.8, the holoenzyme shows pH-independent absorbance maxima at 333 and 416 nm. Vmax/Km is highest at pH 6.5; this optimum reflects chiefly increased Km values for histidine at lower or higher pH values, whereas Vmax is highest at pH 5.0 and decreases only moderately between pH 5.0 and 8.0. The enzyme also decarboxylates beta-(2-pyridyl)alanine and N tau-methylhistidine (but not N pi-methylhistidine); arginine, lysine, and ornithine are neither substrates nor inhibitors. The hydrazine analogue of histidine, 2-hydrazino-3-(4-imidazolyl)propionic acid, is a very potent competitive inhibitor; other carbonyl reagents and a variety of carboxyl- or amino-substituted histidines also inhibit competitively. alpha-Fluoromethylhistidine is a potent irreversible inhibitor of the enzyme; alpha-methylhistidine is a competitive inhibitor/substrate that is decarboxylated slowly and undergoes a slow decarboxylation-dependent transamination that converts the holoenzyme to pyridoxamine-P and apoenzyme. Dithiothreitol and other simple thiols are mixed-type inhibitors that interact with pyridoxal-P at the active site to form complexes (lambda max congruent to 340 nm), presumably the corresponding thioalkylamines, without resolving the holoenzyme. This histidine decarboxylase (Vmax = 72 mumol X min-1 X mg-1) is much more active than "homogeneous" preparations of mammalian pyridoxal-P-dependent histidine decarboxylase (Vmax congruent to 1.0) and is about equal in activity to the pyruvoyl-dependent histidine decarboxylases from Gram-positive bacteria.     

Kinetics of the inactivation of Escherichia coli glutamate apodecarboxylase by phenylglyoxal.

The possible interaction of the phosphate moiety of pyridoxal phosphate with a guanidinium group in glutamate apodecarboxylase was investigated. The holoenzyme is not inactivated significantly by incubation with butanedione, glyoxal, methylglyoxal, or phenylglyoxal. However, the apoenzyme is inactivated by these arginine reagents in time-dependent processes. Phenylgloxal inactivates the apoenzyme most rapidly. The inactivation follows pseudo-first-order kinetics at high phenylglyoxal to apoenzyme ratios. The rate of inactivation is proportional to phenylglyoxal concentration, increases with increasing pH, and is also dependent on the type of buffer present. The rate of inactivation of the apoenzyme by phenylglyoxal is fastest in bicarbonate — carbonate buffer and increases with increasing bicarbonate — carbonate concentration. Phosphate, which inhibits the binding of pyridoxal phosphate to the apoenzyme, protects the apodecarboxylase against inactivation by phenylglyoxal. When the apodecarboxylase is inactivated with [¹⁴C]phenylglyoxal, approximately 1.6 mol of [¹⁴C]phenylglyoxal is incorporated per mol subunit. The phenylglyoxal is thought to modify an arginyl residue at or near the pyridoxal phosphate binding site of glutamate apodecarboxylase.     

Purification and biochemical characterization of some of the properties of recombinant human kynureninase.

Recombinant human kynureninase (L-kynurenine hydrolase, EC 3.7.1.3) was purified to homogeneity (60-fold) from Spodoptera frugiperda (Sf9) cells infected with baculovirus containing the kynureninase gene. The purification protocol comprised ammonium sulfate precipitation and several chromatographic steps, including DEAE-Sepharose CL-6B, hydroxyapatite, strong anionic and cationic separations. The purity of the enzyme was determined by SDS/PAGE, and the molecular mass verified by MALDI-TOF MS. The monomeric molecular mass of 52.4 kDa determined was > 99.99% of the predicted molecular mass. A UV absorption spectrum of the holoenzyme resulted in a peak at 432 nm. The optimum pH was 8.25 and the enzyme displayed a strong dependence on the ionic strength of the buffer for optimum activity. This cloned enzyme was highly specific for 3-hydroxykynurenine (Km = 3.0 microm +/- 0.10) and was inhibited by L-kynurenine (Ki = 20 microm), d-kynurenine (Ki = 12 microm) and a synthetic substrate analogue D,L-3,7-dihydroxydesaminokynurenine (Ki = 100 nm). The activity/concentration profile for kynureninase from this source was sigmoidal in all instances. There appeared to be partial inhibition by substrate, and excess pyridoxal 5'-phosphate was found to be inhibitory.     

Functional interactions between subunits of aspartate aminotransferase: Formation of monoliganded dimers during titration of the apoenzyme by pyridoxal 5′-phosphate

The interaction between apoaspartate aminotransferase and pyridoxal 5′-phosphate at either pH 8.3 (active form of holoenzyme) or pH 5.0 (inactive form) corresponds to a strong quenching of tryptophan fluorescence. The hybrid molecule containing one pyridoxal 5′-phosphate bound per dimer has been prepared both by electrofocusing and by ion exchange chromatography. At both pH values, the fluorescence of the hybrid is 80 to 85% of the arithmetic mean between the fluorescence of the symmetrical holoenzyme and apoenzyme. This is direct evidence of energy transfer from tryptophan residues of the subunit of apoenzyme to the coenzyme of the other subunit.Fluorescence intensity was used to determine the quantity of hybrid holoapoenzyme formed during titration of the apoenzyme by pyridoxal 5′-phosphate. At pH 8.3 a non-linear decrease in the fluorescence is observed, corresponding to 60% of hybrid for the point of half reactivation; this value corresponds to the percentage obtained by electrofocusing (Schlegel & Christen, 1974). At pH 5.0, the decrease in fluorescence is linear during pyridoxal binding; this indicates that at this pH the hybrid is never obtained at detectable concentrations. These results indicate strong interactions between subunits of aspartate aminotransferase corresponding to a weakly negative co-operativity at alkaline pH and a positive cooperativity at acidic pH for the binding of the coenzyme.     

Isolation and characterization of active N-terminal truncated apo- and holoenzyme of mammalian liver tyrosine aminotransferase.

Limited proteolysis was used to probe the structure of the apo- and holoenzyme of rat liver tyrosine aminotransferase. Both were subjected to trypsinolysis and the major fragments were isolated and characterized. Trypsin cleaves the apoenzyme after residues Arg57, Lys64, and Lys71 and the holoenzyme after Arg37 and Lys38. The difference in the accessibility of the enzyme deprived or associated with pyridoxal 5'-phosphate reflects two distinct conformations. The activity, the affinity for the ligands and the thermostability of the purified truncated enzyme forms are similar to those of the native apo- and holoenzyme. A model for the domain structure of mammalian tyrosine aminotransferase and a mechanism for its rapid turnover are proposed.     

Modification of mouse testicular lactate dehydrogenase by pyridoxal 5''-phosphate.

1. Mouse C4 lactate dehydrogenase treated in the dark with pyridoxal 5'-phosphate at pH8.7 and 25 degrees C loses activity gradually; 1mM-pyridoxal 5'-phosphate causes 83% inactivation, and higher concentrations of the reagent cause no further loss of activity. 2. The final extent of inactivation is very pH-dependent, greater inactivation occurring at the high pH values. 3. Inactivation may be fully reversed by addition of cysteine, or made permanent by reducing the enzyme with NaBH4. 4. The absorption spectrum of inactivated reduced enzyme indicates modification of lysine residues. Inactivation by 80% corresponds to modification of at least 1.8 mol of lysine/mol of enzyme subunit. 5. There is no loss of free thiol groups after inactivation with pyridoxal 5'-phosphate and reduction of the enzyme. 6. NAD+ or NADH gives complete protection against inactivation. protection studies with coenzyme fragments indicate that the AMP moiety is largely responsible for the protective effect. Lactate (10 mM) gives no protection in the absence of added nucleotides, but greatly enhances the protection given by ADP-ribose (1 mM). Thus ADP-ribose is able to trigger the binding of lactate. 7. Pyridoxal 5'-phosphate also acts as a non-covalent inhibitor of mouse C4 lactate dehydrogenase. The inhibition is non-competitive with respect to both NAD+ and lactate. 8. Km values for the enzyme at pH 8.0 and 25 degrees C, with the non-varied substrate saturating, are 0.3 mM-lactate and 5 microM-NAD+. 9. These results are discussed and compared with pyridoxal 5'-phosphate modification of other lactate dehydrogenase isoenzymes and related dehydrogenases.     

Effect of pH on conformation and kinetic properties of phosphorylase from bovine skeletal muscles

Interaction of phosphorylase with 8-anilino-1-naphthalene-sulfonate (ANS) results in the formation of an ANS-protein complex. The microenvironment of the protein-bound dye changes depending on pH. Using fluorimetric titration, the dissociation constants for the complex (Kd = 23 and 57 microM for pH 6.2 and 6.8, respectively) were determined. The mode of the enzyme inhibition by ANS also changes depending on pH. At pH 6.8, ANS competitively inhibits the enzyme with respect to AMP, but does not compete with the nucleotide at pH 6.2; the corresponding Ki values are equal to 160 and 26 microM. The protective effect of ligands from the inhibiting effect of ANS was studied. It was shown that at pH 6.2, the enzyme is protected from the inhibition only by the substrate, glucose-1-phosphate, whereas at pH 6.8--by the allosteric inhibitor, glucose-6-phosphate. These findings suggest that at pH 6.2 the conformation of the enzyme molecule is induced by the substrate, while at pH 6.8--by the allosteric inhibitor. ANS binding in the vicinity of the active or allosteric centers is due to the pH-dependent conformational transition. The data obtained suggest that the pH changes within the range of 6.2-6.8 are essential for the regulation of enzyme activity.     

Purification and characterization of aspartate aminotransferase from the thermoacidophilic archaebacterium Sulfolobus solfataricus

Aspartate aminotransferase from the archaebacterium Sulfolobus solfataricus, a thermoacidophilic organism isolated from an acidic hot spring (optimal growth conditions: 87 degrees C, pH 3.5) was purified to homogeneity. The enzyme is a dimer (Mr subunit = 53,000) showing microheterogeneity when submitted to chromatofocusing and/or isoelectric focusing analysis (two main bands having pI = 6.8 and 6.3 were observed). The N-terminal sequence (22 residues) does not show any homology with any stretch of known sequence of aspartate aminotransferases from animal and bacterial sources. The apoenzyme can be reconstituted with pyridoxamine 5'-phosphate and/or pyridoxal 5'-phosphate, each subunit binding 1 mol of coenzyme. The absorption maxima of the pyridoxamine and pyridoxal form are centered at 325 and 335 nm, respectively; the shape of the pyridoxal form band does not change with pH. The enzyme has an optimum temperature higher than 95 degrees C, and at 100 degrees C shows a half-inactivation time of 2 h. The above properties seem to be unique even for enzymes from extreme thermophiles (Daniel, R. M. (1986) in Protein Structure, Folding, and Design (Oxender, D. L., ed) pp. 291-296, Alan R. Liss, Inc., New York) and lead to the conclusion that aspartate aminotransferase from S. solfataricus is one of the most thermophilic and thermostable enzymes so far known.     

Mechanism of reactions catalyzed by selenocysteine beta-lyase

The reaction mechanism of selenocystine beta-lyase has been studied and it was found that elemental selenium is released enzymatically from selenocysteine, and reduced to H2Se nonenzymatically with dithiothreitol or some other reductants that are added to prepare selenocysteine from selenocystine in the anaerobic reaction system. 1H and 13C NMR spectra of L-alanine formed in 2H2O have shown that an equimolar amount of [beta-2H1]- and [beta-2H2]alanines are produced. The deuterium isotope effect at the alpha position was observed; kH/kD = 2.4. These results indicated that the alpha hydrogen of selenocysteine was removed by a base at the active site, and was incorporated into the alpha position of alanine, a product, without exchange of a solvent deuterium. When the enzyme was incubated with L-selenocysteine in the absence of added pyridoxal 5'-phosphate, the activity decreased with prolonged incubation time. However, the activity was recovered by addition of 5'-phosphate. The spectrophotometric study showed that the inactivated enzyme was the apo form. The apoenzyme was activated by a combination of pyridoxamine 5'-phosphate and various alpha-keto acids such as alpha-ketoglutarate and pyruvate. Thus, the enzyme is inactivated through transamination between selenocysteine and the bound pyridoxal 5'-phosphate to produce pyridoxamine 5'-phosphate and a keto acid derived from selenocysteine. The pyridoxal enzyme, an active form, is regenerated by addition of alpha-keto acids. This regulatory mechanism is analogous to those of aspartate beta-decarboxylase [EC 4.1.1.12], arginine racemase [EC 5.1.1.9], and kynureninase [EC 3.7.1.3] [K. Soda and K. Tanizawa (1979) Adv. Enzymol. 49, 1].     

Instability of the apo form of aromatic L-amino acid decarboxylase in vivo and in vitro: implications for the involvement of the flexible loop that covers the active site

Pyridoxine deficiency caused a decrease in the amount of aromatic L-amino acid decarboxylase (AADC) in PC12 cells to less than 5% of the control. The degree of the enzyme saturation with the coenzyme pyridoxal 5'-phosphate (PLP) was around 90% for both the control and the pyridoxine-deficient cells, contrary to earlier reports by others. Mathematical analysis of the result indicated that the AADC apoenzyme is degraded at least 20-fold faster than the holoenzyme in the cells. To determine the mechanism of the preferential degradation of the apoenzyme, in vitro model studies were carried out. AADC has a flexible loop that covers the active site. This loop was easily leaved by proteases at similar rates for both the holoenzyme and the apoenzyme. However, in the presence of the substrate analog, dopa methyl ester, the holoenzyme was not cleaved by proteases, while the apoenzyme was cleaved similarly. These results indicated that the ligand that forms a Schiff base (aldimine) with PLP is fixed to the active site and stabilizes the flexible loop. The structure of the rat AADC-dopa complex modeled on the crystal structure of pig AADC showed that the flexible loop can fit in the concave surface at the entrance of the active site, its aliphatic and aromatic residues forming hydrophobic interactions with the substrate catechol ring. It was postulated that the flexible loop of the holoenzyme is stabilized in vivo by taking a closed structure that holds the PLP-substrate aldimine, while the apoenzyme cannot bind the substrate and its flexible loop is easily cleaved, leading to the preferential degradation of the apoenzyme.     

Stability and Activation of Glutamate Apodecarboxylase from Pig Brain

The stability and activation of glutamate apodecarboxylase was studied with three forms of the enzyme from pig brain (referred to as the alpha, beta, and gamma forms). Apoenzyme was prepared by incubating the holoenzyme with aspartate followed by chromatography on Sephadex G-25. Apoenzyme was much less stable than holoenzyme to inactivation by heat (for beta-glutamate decarboxylase (beta-GAD) at 30 degrees C, t1/2 values of apo- and holoenzyme were 17 and greater than 100 min). ATP protected holoenzyme and apoenzyme against heat inactivation. The kinetics of reactivation of apoenzyme by pyridoxal-P was consistent with a two-step mechanism comprised of a rapid, reversible association of the cofactor with apoenzyme followed by a slow conversion of the complex to active holoenzyme. The reactivation rate constant (kr) and apparent dissociation constant (KD) for the binding of pyridoxal-P to apoenzyme differed substantially among the forms (for alpha-, beta-, and gamma-GAD, kr = 0.032, 0.17, and 0.27 min-1, and KD = 0.014, 0.018, and 0.04 microM). ATP was a strong competitive inhibitor of activation (Ki = 0.45, 0.18, and 0.39 microM for alpha-, beta-, and gamma-GAD). In contrast, Pi stimulated activation at 1-5 mM but inhibited at much higher concentrations. The results suggest that ATP is important in stabilizing the apoenzyme in brain and that ATP, Pi, and other compounds regulate its activation.