Refinement and comparisons of the crystal structures of pig cytosolic aspartate aminotransferase and its complex with 2-methylaspartate.

Two high resolution crystal structures of cytosolic aspartate aminotransferase from pig heart provide additional insights into the stereochemical mechanism for ligand-induced conformational changes in this enzyme. Structures of the homodimeric native structure and its complex with the substrate analog 2-methylaspartate have been refined, respectively, with 1.74-A x-ray diffraction data to an R value of 0.170, and with 1.6-A data to an R value of 0.173. In the presence of 2-methylaspartate, one of the subunits (subunit 1) shows a ligand-induced conformational change that involves a large movement of the small domain (residues 12-49 and 327-412) to produce a "closed" conformation. No such transition is observed in the other subunit (subunit 2), because crystal lattice contacts lock it in an "open" conformation like that adopted by subunit 1 in the absence of substrate. By comparing the open and closed forms of cAspAT, we propose a stereochemical mechanism for the open-to-closed transition that involves the electrostatic neutralization of two active site arginine residues by the negative charges of the incoming substrate, a large change in the backbone (phi,psi) conformational angles of two key glycine residues, and the entropy-driven burial of a stretch of hydrophobic residues on the N-terminal helix. The calculated free energy for the burial of this "hydrophobic plug" appears to be sufficient to serve as the driving force for domain closure.

Use of 1H-15N heteronuclear multiple-quantum coherence NMR spectroscopy to study the active site of aspartate aminotransferase.

Aspartate aminotransferase from Escherichia coli, an 88 kDa enzyme, was uniformly and selectively enriched with 15N and was studied by heteronuclear multiple-quantum coherence NMR spectroscopy in H2O. Good resolution was obtained for the downfield region (above 9.5 ppm chemical shift in the 1H dimension) for NH protons in the amide, indole, imidazole, and guanidinium group regions and several resonances were tentatively assigned. Two downfield resonances, at 12.6 and 11.36 ppm, appear to belong to oxygen- or sulfur-bound protons. The most downfield amide resonance at 11.78 ppm was assigned to the active site cysteine 192 whose peptide proton is 2.9 A away from the negatively charged carboxyl group of aspartate 199. Large downfield shifts (up to 1.15 ppm) of the indole NH resonance of the active site tryptophan 140 were observed upon binding of dicarboxylic inhibitors to the pyridoxal 5'-phosphate (PLP) form and of inorganic dianions to the pyridoxamine 5'-phosphate (PMP) form of the enzyme. We discuss these striking differences in the light of the available crystallographic data. Active sites of proteins, as well as specific inhibitory molecules, often contain negatively charged groups. These may be able to form hydrogen-bonds to NH groups and to shift the NH resonances downfield into a less crowded and therefore more readily observable region for many large proteins. Our approach, which makes use of both HMQC spectroscopy and NOE observations, should be widely applicable.

NMR studies of 1H resonances in the 10-18-ppm range for cytosolic aspartate aminotransferase.

Continuing a previous investigation (Kintanar, A., Metzler, C. M., Metzler, D. E., and Scott, R. D. (1991) J. Biol. Chem. 266, 17222-17229), we have recorded 1H NMR spectra at 500 MHz in the 10-18-ppm range for the 93-kDa porcine cytosolic aspartate aminotransferase and for four specific mutant forms of the enzyme in which histidine 68 has been replaced by lysine or histidine 143, 189, or 193 has been replaced by glutamine. We have correlated resonances for apoenzyme, pyridoxamine and pyridoxal phosphate forms, and dicarboxylate complexes and have assigned imidazole NH resonances of active site histidines. The chemical shifts of several resonances undergo pH-dependent changes around the pKa of the Schiff base proton at the active site. Other resonances shift upon binding of dicarboxylates or other ligands. Phosphate or carboxylate ions, which can also occupy the site of the substrate's alpha-carboxylate, cause rapid exchange of the Schiff base proton. Although most resonances in the 10-18-ppm range disappear rapidly in D2O, a few are retained for months in the presence of the dicarboxylate inhibitor glutarate. We demonstrate that changes in chemical shifts and in exchange rates are sensitive indicators of electronic interactions of the enzyme with ligands and of conformational change. Nuclear Overhauser effects from NH protons have allowed us to identify resonances of CH protons of the imidazole rings of histidines 143, 189, and 193. Observed and predicted chemical shifts have been compared. We conclude that the net charge on this histidine cluster is zero but that some negative charge from the aspartate 222 carboxylate is donated inductively into the histidine 143 ring. Studies of the related enzyme from Escherichia coli are provided in an accompanying paper (Metzler, D. E., Metzler, C. M., Scott, R. D., Mollova, E. T., Kagamiyama, H., Yano, T., Kuramitsu, S., Hayashi, H., Hirotsu, K., and Miyahara, I. (1994) J. Biol. Chem. 269, 28027-28033). Our approach should be applicable to the study of active sites of a broad range of relatively large proteins.

NMR studies of 1H resonances in the 10-18-ppm range for aspartate aminotransferase from Escherichia coli.

We have recorded 500-MHz 1H NMR spectra in the 10-18-ppm range for aspartate aminotransferase from Escherichia coli and for three specific mutant forms. Histidine 143 has been replaced by either alanine or asparagine. In the third mutant, tryptophan 140 has been replaced by phenylalanine. The NMR spectrum of the native enzyme is very similar to that of porcine cytosolic aspartate aminotransferase in the most downfield region. However, the resonances of the proton on the ring nitrogen of the pyridoxal 5'-phosphate (peak A) and on the His-143 imidazole ring (peak B) of the E. coli enzyme are broader and more readily lost at low pH or higher temperatures than those of the porcine enzyme. The possible role of tautomerism in promoting such broadening is discussed. In the histidine mutant proteins, peak A of the pyridoxal 5'-phosphate form is too broad to see under most conditions but is clearly present in the pyridoxamine phosphate form. Peak B is missing in the 2 histidine mutants. Observation of nuclear Overhauser effects further confirms the identity of B as the resonance of HN epsilon 2 of His-143 and that of peak D at approximately 11.8 ppm as HN epsilon 2 of His-189. The mutant spectra also provide insight into electronic interactions between groups in and near the active site which confirm and supplement conclusions drawn from spectra of porcine cAspAT. While no clear loss of a peak was observed for the Trp-140 mutant in its free form, the spectrum of the succinate complex lacked a strong band at 11.26 ppm. This may represent the Trp-140 indole NH proton which has been shifted downfield by binding to a succinate carboxylate group. While our results confirm the basic similarity of cytosolic aspartate aminotransferase and E. coli aspartate aminotransferase 1H NMR spectra, they also point out differences that may be useful in identifying resonances. A large number of mutant proteins have been prepared for the E. coli enzyme. The present results provide essential information for future study of these mutants and for study of NMR spectra of isotopically labeled enzyme.

Enzymatically active truncated cat brain glutamate decarboxylase: expression, purification, and absorption spectrum.

The DNA encoding the sequence for glutamate decarboxylase from cat brain was recloned into the Escherichia coli expression vector pET11a. The N-terminal 77- to 84-amino acid residues encoded by the cloned gene had been deleted from the protein which was purified to near homogeneity in 20-mg batches. The truncated protein is a dimer with a subunit molecular mass of about 59 kDa. This protein is enzymatically active and has a Km for L-glutamate of 1.37 mM and a turnover number of 7 s-1 at its optimal pH of 6.6. The absorption spectrum, resulting from the bound coenzyme, pyridoxal phosphate, showed pH-dependent bands at 338 and 420 nm with an isosbestic point at 356 nm. A spectrophotometric pKa value of 6.92 was evaluated for the bound coenzyme. The pH-dependent kinetic data suggest the presence of two dissociable groups in the free enzyme with pKa values of about 6.45 and 7.05 and pKa value of the enzyme-substrate complex of about 6.82 in phosphate buffer. Structures for the coenzyme in the active site of brain glutamate decarboxylase are proposed.

NMR spectroscopy of exchangeable protons of glucoamylase and of complexes with inhibitors in the 9-15-ppm range.

1H-NMR spectra have been recorded for glucoamylases I and II from Aspergillus awamori var. X100 and from A. niger in the 9-15-ppm region. At least 17 distinct peaks, many of them arising from single protons, are observed. These are designated A-Q, A being the furthest downfield. At least 9 of these are lost rapidly by exchange when the enzyme is placed in D2O. Peaks A, B, E and H undergo distinct shifts with pH change in the pH region 3-7. Several others undergo smaller shifts. Small differences are also seen between the enzymes from the two different sources. Binding of the pseudotetrasaccharide inhibitor acarbose leads to a 0.50-ppm downfield shift of peak B, other smaller changes, and retention of two additional protons in D2O. delta-D-gluconolactone induces shifts in peaks E, H, and L. The slow substrate maltitol causes peak A to broaden and shift, peaks J and K to shift and a new or greatly shifted resonance to appear at 15.4 ppm. It disappears as the maltitol is hydrolyzed. Treatment with iodoacetamide or diethyl pyrocarbonate leads to disappearance of peak D at 12.3 ppm. When this peak was irradiated strong nuclear Overhauser effects (NOE) were observed at 8.01 ppm and 7.22 ppm, positions expected for the C epsilon 1 and C delta 2 protons of an uncharged imidazole ring. We identify D as arising from the N epsilon 2 proton of His254 which is uncharged except at the lowest pH values. Other NOE and two-dimensional NOE spectra have provided additional information. Three mutant forms of the A. niger enzyme, in which tryptophan residues have been replaced by phenylalanine, have been examined. Because of shifts induced by changes in ring current and other environmental effects it is hard to make a direct identification of the resonances from the replaced indole NH protons. However, on the basis of a distinct NOE between peaks E and H we have identified these resonances as arising from the indole NH protons of Trp52 and Trp120. Other possible assignments are considered. The NMR spectra of the glucoamylases I, which have a starch binding domain of about 104 residues at the carboxyl terminus, show four sharp resonances in the 9.7-10.6-ppm range that are not present in the glucoamylases II, which lack this domain. These resonances no doubt represent the four indole NH ring protons from Trp543, Trp562, Trp590 and Trp615. Three of these are very sharp suggesting a high mobility of this domain.

Starch-binding domain of Aspergillus glucoamylase-I. Interaction with beta-cyclodextrin and maltoheptaose.

The characterization is reported of two peptide fragments (SBD106 and SBD122) containing the starch-binding domain (SBD) of Aspergillus sp. glucoamylase I. The starch-binding peptides were produced in Escherichia coli as fusion proteins of the maltose-binding protein (MBP). SBD106 (11.9 kDa) and SBD122 (13.8 kDa) were purified from the factor Xa digest of MBP fusion proteins. The amino acid compositions were similar to those deduced from their amino acid sequences. The interactions of beta-cyclodextrin and maltoheptaose with purified SBD peptides were investigated by UV difference spectroscopy. SBD106 and SBD122 bound specifically beta-cyclodextrin with a dissociation constant (Kd) of 34 microM and 23.5 microM, respectively. Maltoheptaose binding to SBD106 and SBD122 was weaker than that of beta-cyclodextrin; dissociation constants were 0.57 and 0.50 mM, respectively. The results indicate that the intramolecular disulfide bonding is not required for the domain functioning and that O-glycosylation is not critical for the functioning of the starch-binding domain, but may affect its conformation and dynamics.

Role of an active site residue analyzed by combination of mutagenesis and coenzyme analog.

Asp222 of aspartate aminotransferase is an active-site residue which interacts with the pyridine nitrogen of the coenzyme, pyridoxal 5'-phosphate (PLP). The roles of Asp222 in the catalytic mechanism of Escherichia coli aspartate aminotransferase have previously been explored by site-directed mutagenesis. These studies confirmed that a negatively charged residue at position 222 is essential for catalysis, but the reason for this remained speculative. In the present studies, the roles of Asp222 were clarified experimentally by analyzing the mutant D222A enzyme (Asp222 replaced by Ala) reconstituted with the coenzyme analog N(1)-methylated PLP (N-MePLP). Spectroscopic and kinetic analyses showed that Asp222 stabilizes the protonated N(1) of PLP, raising the pKa value of N(1) by more than five units, in the active site of AspAT. The positive charge at N(1) accelerates abstraction of the alpha-proton from the amino acid substrate, stabilizing the transition state by 1.4 to 4.5 kcal.mol-1 in the reaction with aspartate. X-ray crystallographic (2.0 A resolution) and CD spectroscopic studies suggest that the coenzyme analog is not held in a proper orientation within the active site of D222A (N-MePLP). This may account for the finding that the catalytic activity was recovered only partially by the reconstitution of D222A with N-MePLP. These results fully support the following postulated role of Asp222: the negative charge of Asp222 stabilizes the positive charge at N(1) of PLP and thereby enhances the function of PLP as an electron sink.

Is pyridoxal 5'-phosphate an affinity label for phosphate-binding sites in proteins?: The case of bovine glutamate dehydrogenase.

The effects of pyridoxal 5'-phosphate (PalP) on ox liver glutamate dehydrogenase (94% inactivation by 1.8 mM reagent at pH 7 and 25 degrees C) have been compared with those of three analogues, 5'-deoxypyridoxal (96% inactivation), pyridoxal 5'-sulphate (97%) and pyridoxal 5-methylsulphonate (94%), in order to establish whether PalP acts as an affinity label for this enzyme. Like PalP and unlike pyridoxal, which is a much less potent inactivator, none of the analogues has a free 5'-OH group to cyclize with the aldehyde function. The result with 5'-deoxypyridoxal shows that a negative charge, such as that of the phosphate group, is not required for efficient inactivation. With all four reagents, addition of an excess of cysteine or lysine led to 90-100% re-activation over 3-20 h. Dialysis also caused reactivation to a similar extent. A combination of 2.15 mM NADH, 1 mM GTP and 10 mM 2-oxoglutarate gave complete protection against PalP, but only partial protection against the analogues. 5'-Deoxypyridoxal still caused 20-25% inactivation in the presence of the protection mixture. Absorbance measurements after reduction with NaBH4 show the characteristic features of a reduced Schiff's base and allowed estimation of the extent of reaction. With all the reagents the protection mixture decreased incorporation by about 1 mol/mol, but levels of incorporation without protection varied from about 2 mol/mol for PalP up to about 5 mol/mol for 5'-deoxypyridoxal. The labelling at additional sites may explain the residual inactivation in the presence of potent protecting agents.

NMR observation of exchangeable protons of pyridoxal phosphate and histidine residues in cytosolic aspartate aminotransferase.

Observation of the 93-kDa cytosolic aspartate aminotransferase by 500-MHz 1H NMR spectroscopy in H2O has revealed a series of resonances in the 10-18 ppm range arising from exchangeable protons. One of these (peak A) has been assigned to the proton bound to the ring nitrogen of the coenzyme pyridoxal 5'-phosphate. A second (peak B) is assigned to H143 which participates in a chain of hydrogen bonds that includes also the coenzyme-bound proton. There is a mutual nuclear Overhauser effect between these two resonances. Peaks A and B respond to changes in pH and to interaction of the enzyme with coenzyme derivatives and inhibitors. Peak A moves from 15.4 to 17.4 ppm as the pH is lowered, while peak B moves in the opposite direction from 14.7 to 13.7 ppm, both with an apparent pKa of 6.15. This pKa is associated with deprotonation of the imine nitrogen at the Schiff base linkage of the coenzyme with K258 of the enzyme. In spectra of enzyme containing pyridoxamine 5'-phosphate, peak A is observed at 16.5 ppm and peak B is at 13.9 ppm over a broad pH range. Peaks A and B are found at 17.8 and 14.0 ppm, respectively, for the enzyme complex with glutarate. When alpha-methylaspartate is added to the enzyme several new resonances appear in the spectrum, which are attributed to formation of the external aldimine. The position of peak A in spectra of various forms of the enzyme is interpreted to reflect the electronic distribution in the coenzyme ring. Several other peaks in this region of the spectrum also are sensitive to changes in pH or the addition of inhibitors. Some possible assignments of these resonances are discussed.

NMR spectra of exchangeable protons of pyridoxal phosphate-dependent enzymes.

We have recorded 1H NMR spectra in H2O for exchangeable protons of four pyridoxal phosphate-dependent enzymes: D-serine dehydratase, aspartate aminotransferase, tryptophan: indole-lyase and glutamate decarboxylase. The molecular masses range from 48-250 kDa. In every case there are downfield peaks which are lost when the apoenzyme is formed. In most cases some peaks shift in response to interactions with substrates and inhibitors and with changes in pH. We associate one downfield resonance with the proton on the ring nitrogen of the coenzyme and others with imidazole groups that interact with coenzyme or substrates. The chemical shift for the coenzyme-bound proton differs for free enzyme, substrate Schiff base or quinonoid forms.

Equilibria and absorption spectra of tryptophanase.

Tryptophanase (tryptophan: indole-lyase) from Escherichia coli has been isolated in the holoenzyme form and its absorption spectra and acid-base chemistry have been reevaluated. Apoenzyme has been prepared by dialysis against sodium phosphate and L-alanine and molar absorptivities of the coenzyme bands have been estimated by readdition of pyridoxal 5'-phosphate. The spectrophotometric titration curve, whose midpoint is at pH 7.6 in 0.1 M potassium phosphate buffers, indicates some degree of cooperativity in dissociation of a pair of protons. Resolution of the computed spectra of individual ionic forms of the enzyme with lognormal distribution curves shows that band shapes are similar to those of model Schiff bases and of aspartate aminotransferase. Using molar areas from the latter we estimated amounts of individual tautomeric species. In addition to ketoenamine and enolimine or covalent adduct the high pH form also appears to contain approximately 18% of a species with a dipolar ionic ring (protonated on the ring nitrogen and with phenolate -O-). We suggest that this may be the catalytically active form of the coenzyme in tryptophanase. The equilibrium between tryptophanase and L-alanine has also been reevaluated.

Reactions of phosphonate analogs of pyridoxal phosphate with apo-aspartate aminotransferase.

We have investigated reactions of the 5-phosphonoethyl and 5-phosphonoethenyl analogs of pyridoxal 5'-phosphate in the coenzyme site of cytosolic aspartate aminotransferase. Acid dissociation constants and equilibrium constants for hydration and for tautomerization have been evaluated for these compounds. In confirmation of previous results, both compounds are partially active. They bind to apoenzyme well and undergo conversion in the presence of glutamate to amine forms which show induced circular dichroism comparable to that of native enzyme. A normal "external" Schiff base is evidently formed with 2-methylaspartate, but the amounts of quinonoid intermediate formed with erythro-3-hydroxyaspartate are less than those formed with pyridoxal phosphate. The pKa of the imine group of the enzyme reconstituted with the phosphonoethyl analog is more than two units lower than that in the native enzyme. Binding of the dicarboxylates glutarate, 2-oxoglutarate, and succinate shifts the pKa upward. The absorption spectra of the resulting complexes indicate the existence of at least three low pH species. A shift of 2.3 to 2.9 ppm to a lower frequency was observed for the 31P NMR signal upon binding of these dicarboxylates or of 2-methylaspartate. Enzyme containing the analogs crystallizes. Polarized absorption spectra suggest that the coenzyme has an orientation similar to that of pyridoxal phosphate in the native enzyme.

Correlation of polarized absorption spectroscopic and X-ray diffraction studies of crystalline cytosolic aspartate aminotransferase of pig hearts.

Absorption spectra of large, well-formed crystals of cytosolic aspartate aminotransferase have been recorded using plane polarized light. Making use of measurements of crystal thickness we have calculated extinction coefficients with the electric vector of the light parallel to both the a and c axes of the crystals of the enzyme in space group P2(1)2(1)2(1). The spectra have been resolved into components with lognormal distribution curves and the resulting integrated intensities have been used to calculate the c/a polarization ratios for the absorption bands of the bound co-enzyme pyridoxal 5'-phosphate. From the polarization ratio and the co-ordinates of the co-enzyme ring atoms, provided by X-ray crystallography, we have assigned principal molecular directions of the transition dipole moment within the plane of the co-enzyme ring. Of two possible orientations, only one predicts the correct crystal extinction coefficients for the 436 nm band. In this orientation, when viewed from the B face of the ring (i.e. looking into the active site of the enzyme), the transition moment is related to the N-1-C-4 axis of the ring by counterclockwise rotation by 27 degrees. A tentative assignment of the principal molecular directions of the transition moment has also been made for the 368 nm band of the high pH form of the enzyme. In each case, the plane of the co-enzyme ring was located from the atomic co-ordinates of the ring atoms and of those atoms attached directly to the ring. The projection of the N-1 to C-4 axis on to this plane was used to evaluate the orientation of the transition moment, which was presumed to lie precisely within the plane of the ring. We have tilted this plane systematically to evaluate the error in transition moment direction resulting from uncertainties in the atomic co-ordinates. When 2-methylaspartate is diffused into the crystals if forms a Schiff base with the co-enzyme in which the ring has tilted about 32 degrees from its original position and the polarization ratio of the 436 nm band drops from 1.6 in the free enzyme to about 0.38. On the assumption that the orientation of the transition moment within the co-enzyme does not change during this rotation, this value of the polarization ratio is within experimental error of that predicted from X-ray structures on the two forms. The 2-methylaspartate binds only to subunit 1.(ABSTRACT TRUNCATED AT 250 WORDS)

Spectroscopic studies of quinonoid species from pyridoxal 5'-phosphate.

To establish the state of protonation of quinonoid species formed nonenzymically from pyridoxal phosphate (PLP) and diethyl aminomalonate, we have studied absorption spectra of the rapidly established steady-state mixture of species. We have evaluated the formation constant and the spectrum of the mixture of Schiff base and quinonoid species. For N-methyl-PLP a singly protonated species with a peak at 464 nm is formed from the unprotonated aldehyde and the conjugate acid of diethyl aminomalonate with a formation constant Kf of 240 M-1. The very intense absorption band with characteristic vibrational structure (most evident as a shoulder at 435 nm) is accompanied by a weaker, structured band at about 380 nm and a weak, broad band at 330 nm. We suggest that the 380-nm band may represent a tautomeric form of the quinonoid compound. Protonation of the phosphate group appears to affect the spectrum only slightly. The corresponding mixture of Schiff base and quinonoid species formed from PLP has a very similar spectrum at pH 6-7. It has a formation constant Kf of 230 M-1 and a pKa of 7.8, which must be attributed to the ring nitrogen atom. The dissociated species, which may be largely carbanionic, has a strong structured absorption band at 430 nm and a weaker one, again possibly a tautomer, in the 330-nm region. The analysis establishes that in all species a proton remains on either the phenolic oxygen or the imine nitrogen. Proton NMR spectroscopy, under some conditions, reveals only two components: free PLP and what appears to be Schiff base. However, we suggest that the latter may, in fact, be a quinonoid form, either alone or in rapid equilibrium with the Schiff base. Absorption spectra of quinonoid species formed in enzymes are analyzed and compared with the spectra of the nonenzymic species.

Quantitative description of absorption spectra of a pyridoxal phosphate-dependent enzyme using lognormal distribution curves.

Ultraviolet-visible absorption spectra of cytosolic aspartate aminotransferase of pig hearts have been analyzed by resolution with lognormal distribution curves. These have been compared with spectra of reference Schiff bases of pyridoxal 5'-phosphate. Spectra of the free enzyme in two different states of protonation and of complexes with monoanions, dicarboxylates, the substrates L-glutamate, L-aspartate, and L-erythro-3-hydroxyaspartate, and the quasi-substrate 2-methylaspartate have been analyzed. Relative amounts of three tautomeric species have been estimated, as have amounts of various enzyme-substrate intermediates. Bandshape parameters which can be used as a guide to analysis of spectra of other pyridoxal phosphate-dependent enzymes are tabulated. Some formation constants and pKa values, which were evaluated at the same time as the spectra of the complexes, are also reported.

Reactions of 3'-O-methylpyridoxal 5'-phosphate in aspartate aminotransferase.

The reaction of 3'-O-methylpyridoxal 5'-phosphate bound into the active site of aspartate aminotransferase with the substrate L-aspartate has been investigated. This methylated coenzyme is a very poor catalyst but it does function slowly to produce normal products of a transamination half-reaction. At pH 8.5 and above the characteristic absorption band of a quinonoid intermediate appears rapidly and becomes very intense when the aspartate concentration is raised to 2 M. At pH 6 the quinonoid band is not seen, but the conversion of the methylated coenzyme into 3'-O-methylpyridoxamine 5'-phosphate is about 7 times faster than at high pH with the pH dependence being determined by an apparent pKa of 8.1 at 30 degrees C. We suggest that the active site containing the methylated coenzyme carries a net charge 1 unit more positive than that of native enzyme. This causes a loss of some other proton from the active site and could leave the catalytic lysine-258 deprotonated in the quinonoid species. This may explain its inability to react rapidly. We have measured the spectral band shapes of the quinonoid species studied here and have compared it with that seen with native enzyme. Because of the close similarity we conclude that during normal transamination the proton bound to the imine nitrogen probably shifts onto the phenolic oxygen prior to or synchronously with the formation of the observed quinonoid species.

Studies of 6-fluoropyridoxal and 6-fluoropyridoxamine 5'-phosphates in cytosolic aspartate aminotransferase.

The chemical and spectroscopic properties of 6-fluoropyridoxal 5'-phosphate, of its Schiff base with valine, and of 6-fluoropyridoxamine 5'-phosphate have been investigated. The modified coenzymes have also been combined with the apo form of cytosolic aspartate aminotransferase, and the properties of the resulting enzymes and of their complexes with substrates and inhibitors have been recorded. Although the presence of the 6-fluoro substituent reduces the basicity of the ring nitrogen over 10 000-fold, the modified coenzymes bind predominately in their dipolar ionic ring forms as do the natural coenzymes. Enzyme containing the modified coenzymes binds substrates and dicarboxylate inhibitors normally and has about 42% of the catalytic activity of the native enzyme. The fluorine nucleus provides a convenient NMR probe that is sensitive to changes in the state of protonation of both the ring nitrogen and the imine or the -OH group of free enzyme and of complexes with substrates or inhibitors. The NMR measurements show that the ring nitrogen of bound 6-fluoropyridoxamine phosphate is protonated at pH 7 or below but becomes deprotonated at high pH around a pKa of 8.2. The bound 6-fluoropyridoxal phosphate, which exists as a Schiff base with a dipolar ionic ring at high pH, becomes protonated with a pKa of approximately 7.1, corresponding to the pKa of approximately 6.4 in the native enzyme. Below this pKa a single 19F resonance is seen, but there are two light absorption bands corresponding to ketoenamine and enolimine tautomers of the Schiff base. The tautomeric ratio is altered markedly upon binding of dicarboxylate inhibitors. From the chemical shift values, we conclude that during the rapid tautomerization a proton is synchronously moved from the ring nitrogen (in the ketoenamine) onto the aspartate-222 carboxylate (in the enolimine). The possible implications for catalysis are discussed.