Purification and properties of NAD+-dependent glyceraldehyde-3-phosphate dehydrogenase from spinach leaves

An NAD⁺-dependent glyceraldehyde-3-phosphate dehydrogenase (d-glyceraldehyde-3-phosphate:NAD⁺ oxidoreductase (phosphorylating), EC. 1.2.1.12) has been purified from spinach leaves as a homogeneous protein of 150 000 daltons.Kinetic constants of 2.5·10⁻⁴ M and 4 · 10⁻⁴ M have been calculated for NAD⁺ and glyceraldehyde 3-phosphate, respectively.The amino acid composition is characterized by a cysteine content higher than that found in analogous enzymes.On sodium dodecyl sulphate gel electrophoresis, the native enzyme dissociates into two subunits of 37 000 and 14 000 daltons. The two subunits have been isolated in equimolar amounts by gel filtration; end-group analysis shows that alanine is the N-terminal residue of the large subunit, while serine is found at the N-terminus of the small subunit.Comparison of amino acid analyses and peptide maps shows that the two subunits have a different amino acid sequence. These results indicate that the NAD⁺-dependent glyceraldehyde-3-phosphate dehydrogenase, isolated from spinach leaves has an atypical oligomeric structure, the protomer being formed by two different subunits.     

NAD+ analogue binding to glyceraldehyde-3-phosphate dehydrogenase

A series of NAD+ analogues, modified on the pyridinium ring, have been tested for their enzymic properties in reactions with D-glyceraldehyde-3-phosphate dehydrogenase form sturgeon muscle, rabbit muscle and Bacillus stearothermophilus. The observed activity, inhibition and binding data are correlated to the structure of the enzyme and coenzyme analogue by model building on a Vector General interactive graphic display system using coordinates from the B. stearothermophilus holoenzyme structure. Most of the analogues with substituents in the pyridinium-3 position could be bound to glyceraldehyde-3-phosphate dehydrogenase, either in manner similar to NAD+ or in a completely different way with the substituted pyridinium ring rotated 110 degrees or more around the glycosidic bond. This indicates different possible modes of binding of NAD+ analogues within the pyridinium binding subsite. Analogues with substituents in the pyridinium-4 position are shown to be weakly bound to glyceraldehyde-3-phosphate dehydrogenase. This is explained by a strong interaction of the substituent in the 4 position with the residues Asn-313 and Cys-149.     

Purification and properties of NAD+-dependent glyceraldehyde-3-phosphate dehydrogenase from spinach leaves.

An NAD+-dependent glyceraldehyde-3-phosphate dehydrogenase (D-glyceraldehyde-3-phosphate:NAD+ oxidoreductase (phosphorylating), EC. 1.2.1.12) has been purified from spinach leaves as a homogeneous protein of 150,000 daltons. Kinetic constants of 2.5 . 10(-4) M and 4 . 10(-4) M have been calculated for NAD+ and glyceraldehyde-3-phosphate, respectively. The amino acid composition is characterized by a cysteine content higher than that found in analogous enzymes. On sodium dodecyl sulphate gel electrophoresis, the native enzyme dissociates into two subunits of 37,000 and 14,000 daltons. The two subunits have been isolated in equimolar amounts by gel filtration; end-group analysis shows that alanine is the N-terminal residue of the large subunit, while serine is found at the N-terminus of the small subunit. Comparison of amino acid analysies and peptide maps shows that the two subunits have a different amino acid sequence. These results indicate that the NAD+-dependent glyceraldehyde-3-phosphate, dehydrogenase, isolated from spinach leaves has an atypical oligomeric structure, the protomer being formed by two different subunits.     

Synthesis of spin-labeled photoaffinity derivatives of NAD+ and their interaction with lactate dehydrogenase

The synthesis of NAD+ derivatives spin-labeled at either N6 or C8 of the adenine ring is described, in which the carboxamide function of the nicotinamide moiety is replaced by a diazirine ring. Irradiation of these compounds at 350 nm generates a carbene which will react with any functional group in its vicinity including hydrocarbons. Both NAD+ derivatives form tight ternary complexes with lactate dehydrogenase and were covalently incorporated into this enzyme. They may be employed for ESR studies when non-covalent interactions are too weak for motionally restricted species to be observed.     

Interactions and spatial arrangement of spin-labeled NAD+ bound to glyceraldehyde-3-phosphate dehydrogenase. Comparison of EPR and X-ray modeling data

The spatial arrangement of coenzyme NAD+ in remote and adjacent binding sites in various stoichiometric complexes with tetrameric glyceraldehyde-3-phosphate dehydrogenase from rabbit muscle was examined via EPR spectroscopy. An adenosine N6-15N,2H17 spin-labeled derivative of coenzyme NAD+ (SL-NAD+) was chemically synthesized for this work. The spectral simplifications and narrow line widths afforded by 15N and 2H substitution enabled experimental EPR spectra to be deconvoluted into their three component spectra: (a) unbound coenzyme, (b) bound coenzyme without adjacent site occupied, and (c) bound coenzyme with adjacent site occupied. Binding of SL-NAD+ in adjacent active centers of R axis-related subunits resulted in resolved dipolar interactions which characterized intersubunit distances. Binding to distant subunits related by the P and Q axes gave no dipolar interaction. Once the first NAD+ site was occupied, EPR spectra at various stoichiometries provided evidence for nonpreferential spatial binding of SL-NAD+ to the three unoccupied sites. EPR spectral simulations indicated a separation of 12.8 A for the unpaired electrons of spin label moieties of R axis-related coenzymes. Molecular modeling based on x-ray crystallographic data predicted 11-13 A. The angles and distance relating to interacting spin-labels were calculated from atomic coordinates based on molecular modeling of both anti-anti and anti-syn (adenine-ribose) conformations of SL-NAD+. Computer-generated line shapes indicated best agreement with experimental EPR results when the anti-anti geometry was employed. Comparison of EPR spectra from soluble and ammonium sulfate-precipitated enzymes indicated that the NAD+-binding domains are positioned equivalently in the two physical states. Since the observed dipolar line shapes are critically dependent on the distance and geometry relating to the interacting SL-NAD+, these data provide direct evidence for a high degree of conservation of quaternary structure of the enzyme in the hydrated crystalline state. Studies on the enzyme isolated from human erythrocytes also indicated a close correlation with the rabbit muscle enzyme in both the arrangement of NAD+-binding domains and negative cooperativity of coenzyme binding.     

A disubstituted NAD+ analogue is a nanomolar inhibitor of trypanosomal glyceraldehyde-3-phosphate dehydrogenase

N6-Naphthalenemethyl-2'-methoxybenzamido-beta-NAD+, a derivative of a low micromolar first-generation inhibitor of trypanosomal glyceraldehyde phosphate dehydrogenase (GAPDH), was synthesized, taking advantage of methodology for the selective phosphitylation of nucleosides. The compound was found to be a poor alternate cosubstrate for GAPDH, but an extremely potent inhibitor. Although intended for use in crystallization trials, the analogue presents possibilities for further drug design.     

Mapping of the interaction site of CP12 with glyceraldehyde-3-phosphate dehydrogenase from Chlamydomonas reinhardtii. Functional consequences for glyceraldehyde-3-phosphate dehydrogenase

The 8.5 kDa chloroplast protein CP12 is essential for assembly of the phosphoribulokinase/glyceraldehyde-3-phosphate dehydrogenase (GAPDH) complex from Chlamydomonas reinhardtii. After reduction of this complex with thioredoxin, phosphoribulokinase is released but CP12 remains tightly associated with GAPDH and downregulates its NADPH-dependent activity. We show that only incubation with reduced thioredoxin and the GAPDH substrate 1,3-bisphosphoglycerate leads to dissociation of the GAPDH/CP12 complex. Consequently, a significant twofold increase in the NADPH-dependent activity of GAPDH was observed. 1,3-Bisphosphoglycerate or reduced thioredoxin alone weaken the association, causing a smaller increase in GAPDH activity. CP12 thus behaves as a negative regulator of GAPDH activity. A mutant lacking the C-terminal disulfide bridge is unable to interact with GAPDH, whereas absence of the N-terminal disulfide bridge does not prevent the association with GAPDH. Trypsin-protection experiments indicated that GAPDH may be also bound to the central alpha-helix of CP12 which includes residues at position 36 (D) and 39 (E). Mutants of CP12 (D36A, E39A and E39K) but not D36K, reconstituted the GAPDH/CP12 complex. Although the dissociation constants measured by surface plasmon resonance were 2.5-75-fold higher with these mutants than with wild-type CP12 and GAPDH, they remained low. For the D36K mutation, we calculated a 7 kcal.mol(-1) destabilizing effect, which may correspond to loss of the stabilizing effect of an ionic bond for the interaction between GAPDH and CP12. It thus suggests that electrostatic forces are responsible for the interaction between GAPDH and CP12.     

Study of the interaction of glyceraldehyde-3-phosphate dehydrogenase with DNA

Glyceraldehyde-3-phosphate dehydrogenase binds to homologous and heterologous single-stranded but not double-stranded DNA. Binding to RNA, poly(A) and poly(dA-dT) has also been observed. Enzyme binding to these nucleic acids leads to the formation of an insoluble complex which can be sedimented at low speed.The interaction of glyceraldehyde-3-phosphate dehydrogenase with DNA is strongly inhibited by NAD and NADH but not by NADP. Adenine nucleotides, which inhibit the dehydrogenase activity by competing with NAD for its binding site (Yang, S.T. and Deal, W.C., Jr. (1969) Biochemistry 8, 2806–2813), also inhibit enzyme binding to DNA, whereas glyceraldehyde-3-phosphate and inorganic phosphate are non-inhibitory. These results suggest that DNA interacts through the NAD binding sites of glyceraldehyde-3-phosphate dehydrogenase. In accordance with this idea, it was found that DNA also binds to lactate dehydrogenase, an enzyme containing a similar dinucleotide binding domain, and that this binding is inhibited by NADH.A study of the base specificity of the DNA-glyceraldehyde-3-phosphate dehydrogenase interaction using dinucleoside monophosphates shows that inhibition of DNA binding by the dinucleotides requires the presence of a 3′-terminal adenosine and is greater when the 5′-terminus contains a pyrimidine instead of a purine. These results suggest that the dinucleotides bind at the NAD site of the dehydrogenase and that the enzyme would interact preferentially with PypA dinucleotides present in the nucleic acid.     

Mechanism of glyceraldehyde-3-phosphate transfer from aldolase to glyceraldehyde-3-phosphate dehydrogenase

The catalytic interaction of glyceraldehyde-3-phosphate dehydrogenase with glyceraldehyde 3-phosphate has been examined by transient-state kinetic methods. The results confirm previous reports that the apparent Km for oxidative phosphorylation of glyceraldehyde 3-phosphate decreases at least 50-fold when the substrate is generated in a coupled reaction system through the action of aldolase on fructose 1,6-bisphosphate, but lend no support to the proposal that glyceraldehyde 3-phosphate is directly transferred between the two enzymes without prior release to the reaction medium. A theoretical analysis is presented which shows that the kinetic behaviour of the coupled two-enzyme system is compatible in all respects tested with a free-diffusion mechanism for the transfer of glyceraldehyde 3-phosphate from the producing enzyme to the consuming one.     

Interaction of glyceraldehyde-3-phosphate dehydrogenase with AMP as studied by means of a spin-labeled analog

The binding of a spin-labeled AMP analog to tetrameric glyceraldehyde-3-phosphate dehydrogenase from rabbit muscle is described. The spin label, perdeuterated and 15N-substituted 4-amino-2,2,6,6-tetramethylpiperidine-1-oxyl, was attached to C-8 of AMP (C8-SL-AMP). Up to 8 equivalents of C8-SL-AMP bind per enzyme tetramer, i.e., 2 per monomer. Combining sites are the adenine subsite of the coenzyme-binding domain and the phosphate site. Glyceraldehyde 3-phosphate causes a conformational change in the enzyme that brings C8-SL-AMP molecules bound to adjacent R-axis-related subunits closer to one another by 0.2-0.3 nm and allows for spin-spin interaction between the nitroxide radicals. Similar, but less pronounced structural changes take place upon lowering the pH from 8 to 7. Addition of a single equivalent of NAD+ to a complex of the enzyme with 7.6 equivalents of C8-SL-AMP leads to the release of almost 4 C8-SL-AMP molecules. This supports our previous findings that binding of just one NAD+ molecule induces conformational changes in all four subunits.     

Enzymatic properties of glyceraldehyde-3-phosphate dehydrogenase from fish muscle

1. Glyceraldehyde-3-phosphate dehydrogenase was isolated from the ordinary muscle of red sea bream Pagrus major, Pacific mackerel Scomber japonicus and carp Cyprinus carpio by ammonium sulfate fractionation, followed by DEAE-Sepharose CL-6B and DEAE-cellulose column chromatography and Sephadex G-150 gel filtration, and examined for enzymatic properties. 2. Their optimum pH values in the backward reaction ranged from 7.8 to 8.2, and Km values from 1.56 to 1.90 mM. 3. Irrespective of the species of fish, the enzymatic activity was non-competitively inhibited by inorganic phosphate in the backward reaction. Divalent metal ions were not necessary to activate these glyceraldehyde-3-phosphate dehydrogenases. In the presence of 1 mM Zn(2+), these enzymes showed relative activities of 42-64% the activities measured in the absence of those ions. 5. Thermal stability of carp enzyme was higher than those of red sea bream and Pacific mackerel; the enzyme activity of the latter two species was almost lost on incubation at 45 degrees C for 10-20 min, whereas carp enzyme retained half the activity even when incubated at 60 degrees C for 30 min.     

Crystal structure of two ternary complexes of phosphorylating glyceraldehyde-3-phosphate dehydrogenase from Bacillus stearothermophilus with NAD and D-glyceraldehyde 3-phosphate

The crystal structure of the phosphorylating glyceraldehyde-3-phosphate dehydrogenase (GAPDH) from Bacillus stearothermophilus was solved in complex with its cofactor, NAD, and its physiological substrate, D-glyceraldehyde 3-phosphate (D-G3P). To isolate a stable ternary complex, the nucleophilic residue of the active site, Cys(149), was substituted with alanine or serine. The C149A and C149S GAPDH ternary complexes were obtained by soaking the crystals of the corresponding binary complexes (enzyme.NAD) in a solution containing G3P. The structures of the two binary and the two ternary complexes are presented. The D-G3P adopts the same conformation in the two ternary complexes. It is bound in a non-covalent way, in the free aldehyde form, its C-3 phosphate group being positioned in the P(s) site and not in the P(i) site. Its C-1 carbonyl oxygen points toward the essential His(176), which supports the role proposed for this residue along the two steps of the catalytic pathway. Arguments are provided that the structures reported here are representative of a productive enzyme.NAD.D-G3P complex in the ground state (Michaelis complex).     

Differential binding of NAD+ to acyl glyceraldehyde-3-phosphate dehydrogenase and its role in the acyl group transfer reaction.

Enzyme protein fluorescence of di-furylacryloyl-glyceraldehyde-3-phosphate dehydrogenase (di-FA-GPDH:lambda max.excitation 290 nm, lambda max.emission 338 nm) is quenched about 28% on saturation with NAD+. Results of fluorometric titration of di-FA-GPDH with NAD+ suggest the presence of two tight and two loose coenzyme binding sites (Kdiss. 0.1 and 6.0 microM, respectively). Initial rates of the NAD(+)-dependent reaction of di-FA-GPDH with arsenate and phosphate and of mono-FA-GPDH with phosphate have been determined at varying coenzyme concentrations. The data suggest that binding of NAD+ at the tight sites does not activate the acyl group for its reaction with the acceptor (phosphate or arsenate). The group transfer reaction is dependent only on NAD+ binding to the loose sites, which carry the acyl group. The large difference in the NAD+ binding affinity to the two types of sites and their different effects on the group transfer reaction impart a sigmoidal shape to the rate versus [NAD+] plots. The sigmoidicity is abolished if the reactive SH groups at the unacylated sites are blocked by carboxymethylation.     

Isolation and characterization of glyceraldehyde-3-phosphate dehydrogenase from human erythrocyte membranes.

A rapid and convenient procedure for isolating human glyceraldehyde-3-phosphate dehydrogenase from erythrocytes has been developed and yields enzyme with a specific activity of 33–52. The physical and catalytic properties of the enzyme are similar to those of rabbit muscle enzyme. Reassociation of freshly isolated human glyceraldehyde-3-phosphate dehydrogenase with washed erythrocyte membranes increases the specific activity and stability of the enzyme suggesting that enzyme-membrane interactions may have an important effect on the conformation and catalytic activity. That the human enzyme behaves as a dimer of dimers, similar to the behavior or rabbit muscle glyceraldehyde-3-phosphate dehydrogenase, is suggested by its half-of-the-sites reactivity toward 4-iodoacetamido-1-naphthol. The human enzyme binds nicotinamide hypoxanthine dinucleotide, a structural analog of NAD⁺, with negative cooperativity, further indicating its similarity to rabbit muscle enzyme.