The 1.9 A resolution crystal structure of phosphono-CheY, an analogue of the active form of the response regulator, CheY.

To structurally characterize the activated state of the transiently phosphorylated signal transduction protein CheY, we have constructed an alpha-thiophosphonate derivative of the CheY D57C point mutant and determined its three-dimensional structure at 1.85 A resolution. We have also characterized this analogue with high-resolution NMR and studied its binding to a peptide derived from FliM, CheY's target component of the flagellar motor. The chemically modified derivative, phosphono-CheY, exhibits many of the chemical properties of phosphorylated wild-type CheY, except that it is indefinitely stable. Electron density for the alpha-thiophosphonate substitution is clear and readily interpretable; omit refinement density at the phosphorus atom is greater than 10sigma. The molecule shows a number of localized conformational changes that are believed to constitute the postphosphorylation activation events. The most obvious of these changes include movement of the side chain of the active site base, Lys 109, and a predominately buried conformation of the side chain of Tyr 106. In addition, there are a number of more subtle changes more distant from the active site involving the alpha4 and alpha5 helices. These results are consistent with our previous structural interpretations of other CheY activation mutants, and with our earlier hypotheses concerning CheY activation through propagation of structural changes away from the active site.

Structural changes induced in p21Ras upon GAP-334 complexation as probed by ESEEM spectroscopy and molecular-dynamics simulation.

BACKGROUND: The means by which the protein GAP accelerates GTP hydrolysis, and thereby downregulates growth signaling by p21Ras, is of considerable interest, particularly inasmuch as p21 mutants are implicated in a number of human cancers. A GAP "arginine finger," identified by X-ray crystallography, has been suggested as playing the principal role in the GTP hydrolysis. Mutagenesis studies, however, have shown that the arginine can only partially account for the 10(5)-fold increase in the GAP-accelerated GTPase rate of p21. RESULTS: We report electron spin-echo envelope modulation (ESEEM) studies of GAP-334 complexed with GMPPNP bound p21 in frozen solution, together with molecular-dynamics simulations. Our results indicate that, in solution, the association of GAP-334 with GTP bound p21 induces a conformational change near the metal ion active site of p21. This change significantly reduces the distances from the amide groups of p21 glycine residues 60 and 13 to the divalent metal ion. CONCLUSIONS: The movement of glycine residues 60 and 13 upon the binding of GAP-334 in solution provides a physical basis to interpret prior mutagenesis studies, which indicated that Gly-60 and Gly-13 of p21 play important roles in the GAP-dependent GTPase reaction. Gly-60 and Gly-13 may play direct catalytic roles and stabilize the attacking water molecule and beta,gamma-bridging oxygen, respectively, in p21. The amide proton of Gly-60 could also play an indirect role in catalysis by supplying a crucial hydrogen bonding interaction that stabilizes loop L4 and therefore the position of other important catalytic residues.

Kinetic characterization of CheY phosphorylation reactions: comparison of P-CheA and small-molecule phosphodonors.

In the chemotaxis system of Escherichia coli, phosphorylation of the CheY protein plays an important role in regulating the swimming pattern of the cell. In vitro, CheY can be phosphorylated either by phosphotransfer from phospho-CheA or by acquiring a phosphoryl group from any of a variety of small, high-energy phosphodonor molecules such as acetyl phosphate. Previous work explored the rapid kinetics of CheY phosphorylation by CheA. Here we extend that work and examine the kinetics of CheY phosphorylation by several small-molecule phosphodonors, including acetyl phosphate, benzoyl phosphate, carbamoyl phosphate, 2-methoxybenzoyl phosphate, and phosphoramidate. Our results indicate that these phosphodonors bind to CheY with relatively low affinity (Ks values ranging from 10 to 600 mM) and that the rate constant (kphos) for phosphotransfer at saturating phosphodonor concentrations is relatively slow (values ranging from 0.05 to 0.5 s-1). By contrast, under identical conditions, phosphorylation of CheY by phospho-CheA occurs much more rapidly (kphos approximately 800 s-1) and reflects CheY binding to phospho-CheA considerably more tightly (Ks approximately 60 microM) than it does to the small-molecule phosphodonors. In comparing CheA-mediated phosphorylation of CheY to small-molecule-mediated phosphorylation of CheY, the large difference in kphos values suggests that phospho-CheA makes significant contributions to the catalysis of CheY phosphorylation. The effects of pH and ionic strength on CheY phosphorylation kinetics were also investigated. For CheA-->CheY phosphotransfer, increasing ionic strength resulted in increased Ks values while kphos was unaffected. For CheY phosphorylation by small-molecule phosphodonors, increasing ionic strength resulted in decreasing Ks values and increasing kphos values. The significance of these effects is discussed in relation to the catalytic mechanism of CheY phosphorylation by phospho-CheA and small-molecule phosphodonors.

Synthesis and biochemical characterization of an analogue of CheY-phosphate, a signal transduction protein in bacterial chemotaxis.

CheY is a signal transduction protein of the bacterial chemotaxis system that acts as a molecular switch to alter the swimming behavior of the bacterium. When CheY becomes phosphorylated at Asp57, CheY-Pi interacts with flagellar motor proteins, including FliM, to increase the likelihood that the flagellar motor will change its sense of rotation, increasing the frequency of tumbling. The structure of CheY in its dephosphorylated (inactive) state has been intensively investigated. The short lifetime ( approximately 20 s) of the aspartyl phosphate has precluded the complete structural determination of CheY-Pi. We have synthesized an analogue of CheY-Pi by alkylating an aspartate-to-cysteine mutant at position 57 of CheY to add a phosphonomethyl group at Cys57. This analogue, phosphono-CheY, is stable for months. Phosphono-CheY binds to two of the targets of CheY-Pi, FliM and CheZ, in a manner similar to that of CheY-Pi and much better than either unphosphorylated CheY or the unmodified form of D57C CheY. Phosphono-CheY also binds Mg(II) with a dissociation constant of approximately 6 mM at neutral pH and moderate salt level. These observations indicate that phosphono-CheY is a good biochemical analogue of CheY-Pi.

The effects of cryoprotection on the structure and activity of p21 ras: implications for electron spin-echo envelope modulation spectroscopy.

Electron spin-echo envelope modulation (ESEEM) spectroscopy is widely used to investigate the active sites of biological molecules in frozen solutions. Various cryoprotection techniques, particularly the addition of co-solvents, are commonly employed in the preparation of such samples. In conjunction with ESEEM studies of Mn(II) guanosine nucleotide complexes of p21 ras, we have investigated the effects of cryoprotection on the spectroscopy, the structure, and the activity of this protein. Echo decay times, which typically govern ESEEM spectral resolution, were found to vary linearly with the concentration of glycerol or methyl alpha-D-glucopyranoside (MG), with both additives equally effective on a per-mole basis. The effect of glycerol and MG on the ESEEM amplitudes of various protein nucleiwas studied in ras p21.Mn(II). 5'guanylylimido-diphosphate(p21.Mn(II)-GMPPNP) complexes: these additives did not alter the distances of these nuclei from the Mn(II) ion. In particular, in p21 incorporating [2H-3]Thr, the Mn(II)-[2H-3]Thr35 distance was found to be unaffected by the concentration of cryoprotectant or the rate of freezing. The proximity of the cryoprotectants to the Mn(II) ion was probed by 2H ESEEM in solutions made with d5-glycerol and d7-methyl alpha-D-glucopyranoside (d7-MG). In p21.Mn(II)GMPPNP, the large deuterium modulations from the d5-glycerol exhibit saturation behavior with increasing d5-glycerol concentration, implying that glycerol, a widely used cryoprotectant, replaces the aquo ligands of the Mn(II) ion. The interaction between the Mn(II) ion of p21 and MG, however, is less intimate: the deuterium ESEEM amplitudes are much smaller for samples prepared with d7-MG than with d5-glycerol. Several polyhydroxylic compounds were found to have essentially no effect on the ability of the guanosine 5'-triphosphate (GTP) hydrolysis activating protein, GAP334, to catalyze hydrolysis of p21. guanosine 5'-triphosphate. This observation implies that the introduction of cryoprotectant does not significantly perturb the structure of p21 and gives insight into the mechanism of the GTPase reaction.

The frozen solution structure of p21 ras determined by ESEEM spectroscopy reveals weak coordination of Thr35 to the active site metal ion.

BACKGROUND: The G protein p21 ras is a molecular switch in the signal transduction pathway for cellular growth and differentiation. Hydrolysis of tightly bound GTP alters the conformation of p21, terminating the signal. The coordination of the p21 residue Thr35 to Mg2+ in its active site, which has been observed in the crystal structure of p21 in complex with a GTP-analog GMPPNP but not with GDP, has been proposed to drive the conformational change accompanying nucleotide substitution and may have a role in the GTP hydrolysis reaction itself. However, previous electron spin-echo envelope modulation (ESEEM) studies of selectively 2H beta-threonine and 15N-threonine labeled p21.Mn2+ GMPPNP suggest that Thr35 only weakly coordinates the metal ion in the growth-active GTP-bound state of p21. RESULTS: A 13C beta-Thr35 to Mn2+ distance of 4.3 +/- 0.2 A and a 15N epsilon-Lys16 to Mn2+ distance of 5.3 +/- 0.2 A were determined from ESEEM spectra of the selectively 13C beta-Thr and 15N epsilon-Lys labeled p21.Mn2+ GMPPNP frozen solution structure. The 13C beta-Thr35 to Mn2+ distance is greater than that (3.16 A) observed in the crystal structure. In contrast, the 15N epsilon-Lys16 to Mn2+ distance is in good agreement with the 5.1 A crystal structure distance. CONCLUSIONS: The 13C beta of Thr35 is more distant from the active site Mn2+ in the frozen solution structure than in the crystal structure of p21.Mg2+ GMPPNP, indicating that Thr35 only weakly coordinates the metal ion in frozen solution. Thr35 coordination of the metal ion is therefore unlikely to drive the conformational change between GTP- and GDP-bound states of p21. Thr35 may be essential for GTPase-activating protein (GAP)-stimulated GTP hydrolysis and/or signal transduction for other reasons.

A low-barrier hydrogen bond in subtilisin: 1H and 15N NMR studies with peptidyl trifluoromethyl ketones.

The N delta 1 proton of His 64 forms a hydrogen bond with Asp 32, as part of the catalytic triad in serine proteases of the subtilisin family. His 64 in subtilisin has been studied by 1H and 15N NMR spectroscopy in the presence and absence of peptidyl trifluoromethyl ketones (TFMKs) that are transition state analog inhibitors. For subtilisin Carlsberg, the downfield resonance of the imidazolium N delta 1 proton is approximately 18.3 ppm and the D/H fractionation factor is 0.55 +/- 0.04 at pH 5.5 (11 degrees C), and 0.63 +/- 0.04 (5 degrees C) and 0.68 +/- 0.04 at pH 6 (11 degrees C). In the complex between subtilisin Carlsberg and Z-L-leucyl-L-leucyl-L-phenylalanyltrifluoromethyl ketone (Z-LLF-CF3) at pH values between 6.5 and 10.6, His 64 remains positively charged, and the D/H fractionation factor of its N delta 1 proton is 0.85 +/- 0.05. In the complex between a subtilisin variant from Bacillus lentus and Z-LLF-CF3, the proton resonance at 18.8 ppm is correlated with a 15N resonance at 197.6 ppm downfield from liquid NH3 with a 1JNH of 81 Hz. The chemical shifts of subtilisin complexes with peptidyl TFMKs are among the most downfield shifts reported for any protein. At pH 9.5, His 64 is neutral and the D/H fractionation factor increases to 1.2 with a chemical shift of 15.0. His 64 is positively charged in the free enzyme at low pH, the inhibitor hemiketal complex at neutral pH, and the transition state for amide bond hydrolysis. These data thus provide indirect evidence for the presence of a low-barrier hydrogen bond in the catalytic mechanism of subtilisin proteases.

High frequency (139.5 GHz) electron paramagnetic resonance characterization of Mn(II)-H2(17)O interactions in GDP and GTP forms of p21 ras.

As a molecular switch, the ras protein p21 undergoes structural changes that couple recognition sites on the protein surface to the guanine nucleotide-divalent metal ion binding site. X-ray crystallographic studies of p21 suggest that coordination between threonine-35 and the divalent metal ion plays an important role in these conformational changes. Recent ESEEM studies of p21 in solution, however, place threonine-35 more distant from the metal and were interpreted as weak or indirect coordination of this residue. We report high frequency (139.5 GHz) EPR spectroscopy of p21.Mn(II) complexes of two guanine nucleotides that probes the link between threonine-35 and the divalent metal ion. By analysis of high-frequency EPR spectra, we determine the number of water molecules in the first coordination sphere of the manganous ion to be four in p21.Mn(II).GDP, consistent with prior low-frequency EPR and X-ray crystallographic studies. In the complex of p21 with a GTP analog, p21.Mn(II).GMPPNP, we determine the hydration number to be 2, also consistent with crystal structures. This result rules out indirect coordination of threonine-35 in the solution structure of p21.Mn(II).GMPPNP, and implicates direct, weak coordination of this residue as suggested by Halkides et al. [(1994) Biochemistry 33,4019]. The 17O hyperfine coupling constant of H2(17)O is determined as 0.25 mT in the GDP from and 0.28 mT in the GTP form. These values are similar to reported values for 17O-enriched aquo ligands and some phosphato ligands in Mn(II) complexes. The high magnetic field strength (4.9 T) employed in these 139.5 GHz EPR measurements leads to a narrowing of the Mn(II) EPR lines that facilitates the determination of 17O hyperfine interactions.

High frequency (139.5 GHz) electron paramagnetic resonance spectroscopy of the GTP form of p21 ras with selective 17O labeling of threonine.

Electron paramagnetic resonance spectroscopy at 139.5 GHz has been used to study p21 ras complexed with Mn(II) and guanosine 5'-(beta, gamma-imidotriphosphate), an analog of GTP. The p21 sample studied was selectively labeled with [17O gamma]threonine to a final enrichment of 30%. A Mn(II)-17O hyperfine interaction was observed, but the value of the coupling constant, 0.11 +/- 0.04 mT, is the smallest such value yet reported. Ab initio calculations indicate that this value is consistent with direct coordination of the threonine hydroxyl group and provide an estimate for the Mn(II)-17O bond length of 2.7 A. The measured hyperfine coupling constant and associated bond length starkly contrast with typical values for Mn(II)-17O coordination complexes, namely, approximately 0.25 mT and approximately 2.2 A, respectively. This contrast underscores the peculiar weakness of this Mn(II)-O interaction in p21 and persuasively argues that the nucleotide-induced conformational change, which is known to encompass the region of p21 involving Thr35, is not driven by Mn(II) coordination of the Thr35 hydroxyl group.

The effect of 17O on the relaxation of an amide proton within a hydrogen bond.

The relaxation rates of the multiple-quantum coherence for the amide hydrogen of Gly13 in ras p21.GDP were determined in the presence and absence of 17O labeling in the beta-phosphate of GDP. No significant difference could be observed between labeled and unlabeled samples, in spite of the fact that the hydrogen bond from the amide group of Gly13 to the beta-phosphate is shorter than is typical, based on its chemical shift. For macromolecules in which an oxygen atom is the acceptor of a hydrogen bond, dipolar coupling between 17O and hydrogen is unlikely to be observable, except for extremely short H-bonds.

Characterization of the active site of p21 ras by electron spin-echo envelope modulation spectroscopy with selective labeling: comparisons between GDP and GTP forms.

Selectively labeled samples of human H- or N-ras p21 ligated to MnIIGDP or MnIIGMPPNP were studied by electron spin-echo envelope modulation spectroscopy in order to define the protein environment around the divalent metal. We incorporated [4-13C]-labeled Asx into p21.MnIIGDP and found that the distance from the carboxyl 13C of Asp57 to MnII is approximately 4.1 A. Our result is consistent with indirect coordination of this residue to the metal. From a [2-2H]Thr-labeled sample, we estimate that the distance from the MnII ion to the 2H of Thr35 is at least 5.8 A. Thus, the only protein or nucleotide ligands to the metal appear to be Ser17 and the beta-phosphate of GDP, as previously reported [Larsen, R. G., Halkides, C. J., Redfield, A. G., & Singel, D. J. (1992b) J. Am. Chem. Soc. 114, 9608-9611]. In the 5'-guanylylimido diphosphate (GMPPNP) form of p21, Thr35 has been reported by X-ray crystallography to be a ligand of the metal via its hydroxyl group, and this residue appears to play a key role in the biologically important conformational change upon nucleotide substitution [Pai, E. F., Krengel, U., Petsko, G., Goody, R. S., Kabsch, W., & Wittinghofer, A. (1990) EMBO J. 9, 2351-2359]. The ESEEM spectrum of p21.MnIIGMPPNP labeled with [2-2H]Thr yields a MnII-2H distance of 4.9 A, a distance inconsistent with strong coordination. A sample of p21 in which the Thr residues were fully labeled with 13C and 15N yielded a value of 5.0 A for the distance from MnII to the amide nitrogen of Thr35, while the 13C signal is much smaller than expected if Thr35 were coordinated. A [15N]serine/glycine-labeled sample gives a distance to the amide 15N of Ser17 of 3.9 A, consistent with the X-ray structure; a [4-13C]-labeled Asx sample of p21 gives a distance of approximately 4 A between MnII and the label of Asp57, again implying indirect coordination. Both of these values are very similar to those found for the GDP form of the protein. The results for Thr35, however, reveal a structural difference between the GDP and GTP forms in the region of Thr35. In addition, the position of this residue is found to be different from the crystal structure and in a manner suggesting that the metal ligation of Thr35 does not drive the conformational change that accompanies nucleotide substitution.

An NMR comparison of the changes produced by different guanosine 5'-triphosphate analogs in wild-type and oncogenic mutant p21ras.

We have used nuclear magnetic resonance spectroscopy to compare the conformational changes produced by replacement of bound GDP by the GTP analogs guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S) and guanylyl (beta, gamma-imido)diphosphate (GMPPNP) in wild-type p21ras as well as the oncogenic mutant (G12D)p21ras. We have used isotope-edited nuclear magnetic resonance spectroscopy to observe the amide resonances of selectively [15N]glycine and [15N]isoleucine labeled p21ras-nucleotide complexes. We find that eight of the nine resonances that respond strongly to GTP gamma S and GMPPNP binding are the same but that the nature of the effect appears different. With GTP gamma S, seven new resonances replace the eight resonances specifically associated with GDP-p21ras, but in GMPPNP-p21ras only two resonances replace the GDP-specific resonances that are lost. The resonance of Gly 60 is clearly shown to be responsive to replacement of GDP by GMPPNP, in addition to glycines 10, 12, 13, 15, and 75 and isoleucines 36, 21, and one other, that were found to respond to GTP gamma S by Miller et al. [Miller, A.-F., Papastavros, M. Z., & Redfield, A.G. (1992) Biochemistry 31, 10208-10216). The two GMPPNP-specific resonances observed appear in positions similar to GTP gamma S-specific resonances, and the GTP gamma S-specific resonances, although not lost altogether, are weaker than the GDP-specific resonances they replace. Thus, the two GTP analogs have similar effects on the spectrum of p21ras, suggesting that the effects are due to features common to both analogs. We propose that active site resonance intensities are specifically attenuated when GTP analogs are bound because interactions with the gamma-phosphate of GTP analogs couple the flexible loops 2 and 4 to the rigid loop 1 of the active site. The conformational heterogeneity and dynamics of loops 2 and 4 would be constrained by loop 1 but also transmitted to it. Coupled conformational exchange on a common intermediate time scale could explain the simultaneous loss of resonances from all three loops in the active site. In our comparison of wild-type and (G12D) GDP-p21ras, we find that the resonance of Ile 36 is not visible in (G12D)p21ras. In (G12D)p21ras, replacement of GDP by GTP gamma S causes the resonances of glycines 10, 13, 15, 60, and 75 and isoleucine 21 and four others to shift from their GDP-specific positions. GTP gamma S-specific resonances are observed for all but two of these.(ABSTRACT TRUNCATED AT 400 WORDS)

Interactions of metal ions with mu-monothiopyrophosphate.

mu-Monothiopyrophosphate (MTP) binds monovalent and divalent metal ions with dissociation constants (Kd) similar to those for pyrophosphate (PPi). The values of Kd for metal-MTP complexes are the following, as measured kinetically in the hydrolysis of MTP (microM): Mg2+, 32 +/- 4; Mn2+, 5.4 +/- 1.4; and Co2+, 27 +/- 15. The thermodynamically measured (EPR) values for Mg2+ and Co2+ are 28 +/- 13 microns and 11 +/- 4 microM, respectively; and the Kd for the complex MnPPi is 3.4 +/- 0.5 microM. The metal-MTP complexes undergo hydrolysis at rates modestly faster or slower than the rate at which MTP itself reacts. The complexes MgMTP2-, CoMTP2-, and MnMTP2- undergo hydrolytic cleavage with release of thiophosphate with observed first-order rate constants of 1.6 x 10(-2) min-1, 2.3 x 10(-2) min-1, and 0.6 x 10(-2) min-1, respectively, at 35 degrees C, compared with 1.1 x 10(-2) min-1 for MTP4- under the same conditions. Alkali metal cations also stimulate or retard the hydrolysis of MTP. At 25 degrees C and pH 12.2, the observed rate constant for tetramethylammonium MTP4- is 2.1 x 10(-3) min-1, and the estimated rate constants (min-1) for saturating alkali metals under the same conditions are as follows: Li+, 0.25 x 10(-3); Na+, 3.9 x 10(-3), K+, 6.7 x 10(-3); and Cs+, 6.7 x 10(-3). Divalent metal ions markedly retard the hydrolysis of MTP at pH 7 and 8 because complexation shifts the pH rate profile more than 2 pH units toward the acid side.(ABSTRACT TRUNCATED AT 250 WORDS)

The substrate reactivity of mu-monothiopyrophosphate with pyrophosphate-dependent phosphofructokinase: evidence for a dissociative transition state in enzymatic phosphoryl group transfer.

mu-Monothiopyrophosphate (MTP), an analogue of pyrophosphate (PPi) with sulfur in place of oxygen in the bridge position, is a substrate for the enzyme pyrophosphate-dependent phosphofructokinase. At pH 9.4 and 6 degrees C, the maximal velocity for the phosphorylation of fructose 6-phosphate (F6P) by MgMTP is about 2.8% of that with MgPPi as the phosphoryl donor. The kinetic mechanism is equilibrium random with rate-limiting transformation of the substrate ternary complex to the product when either MgMTP or MgPPi is the phosphoryl donor. This is known from independent studies to be kinetic mechanism at pH 8.0 and 25 degrees C [Bertagnolli, B. L., & Cook, P. F. (1984) Biochemistry 23, 4101-4108]. The dissociation constant of MgPPi is 14 microM, that of MgMTP is 64 microM, and that of F6P from the enzyme is about 5 mM. The Km values for MgPPi and MgMTP are 14.5 and 173 microM, respectively. MgMTP competes with MgPPi for binding to the enzyme. The values of kcat are 3.4 s-1 and 140 s-1 for MgMTP and MgPPi, respectively, at pH 9.4 and 6 degrees C. The estimated rate enhancement factors are 3.6 x 10(5) and 1.4 x 10(14) for the reactions of MgMTP and MgPPi, respectively. Therefore, MgMTP is a reasonably good substrate for PPi-dependent PKF, on the basis of comparisons of kcat. However, the rate enhancement factors show that the enzyme is a poor catalyst for the reaction of MgMTP. Lesser enzymatic catalysis in the reaction of MgMTP compared with MgPPi is largely compensated for by the greater intrinsic reactivity of MgMTP. Thus, the larger substrate MgMTP is well accommodated in the active site, and the dissociative reaction of MgMTP is well accommodated in the transition state. The results are interpreted to indicate a dissociative transition state for phosphoryl group transfer by PPi-dependent PFK. A modified synthesis and purification of MTP are described, in which (trimethylsilyl)trifluoromethanesulfonate and tetra-N-butylammonium iodide are used in place of iodotrimethylsilane to dealkylate tetramethyl-MTP.

UDP-galactose 4-epimerase. Phosphorus-31 nuclear magnetic resonance analysis of NAD+ and NADH bound at the active site.

The phosphorus atoms of NAD+ bound within the active site of UDP-galactose 4-epimerase from Escherichia coli exhibit two NMR signals, one at delta = -9.60 +/- 0.05 ppm and one at delta = -12.15 +/- 0.01 ppm (mean +/- standard deviation of four experiments) relative to 85% H3PO4 as an external standard. Titration of epimerase.NAD+ with UMP causes a UMP-dependent alteration in the chemical shifts of the resulting exchange-averaged spectra, which extrapolate to delta = -10.51 ppm and delta = -11.06 ppm, respectively, for the fully liganded enzyme, with an interconversion rate between epimerase.NAD+ and epimerase.NAD+.UMP of at least 490 s-1. Conversely, the binding of 8-anilinonaphthalene-1-sulfonate, which is competitive with UMP, causes a significant sharpening of the epimerase.NAD+ resonances but very little alteration in their chemical shifts, to delta = -9.38 ppm and delta = -12.16 ppm, respectively. UMP-dependent reductive inactivation by glucose results in the convergence of the two resonances into a single signal of delta = -10.57 ppm, with an off-rate constant for UMP dissociation from the epimerase.NADH.UMP complex estimated at 8 s-1. Reductive inactivation by borohydride under anaerobic conditions yields a single, broad resonance centered at about delta = -10.2 ppm. The data are consistent with, and may reflect, the activation of NAD+ via a protein conformational change, which is known from chemical studies to be driven by uridine nucleotide binding. Incubation of epimerase.NAD+ with UMP in the absence of additional reducing agents causes a very slow reductive inactivation of the enzyme with an apparent pseudo-first-order rate constant of 0.013 +/- 0.001 h-1, which appears to be associated with liberation of inorganic phosphate from UMP.

Synthesis and properties of 2-acetylthiamin pyrophosphate: an enzymatic reaction intermediate.

The synthesis of 2-acetylthiamin pyrophosphate (acetyl-TPP) is described. The synthesis of this compound is accomplished at 23 degrees C by the oxidation of 2-(1-hydroxyethyl)thiamin pyrophosphate using aqueous chromic acid as the oxidizing agent under conditions where Cr(III) coordination to the pyrophosphoryl moiety and hydrolysis of both the pyrophosphate and acetyl moieties were prevented. Although the chemical properties exhibited by acetyl-TPP are similar to those of the 2-acetyl-3,4-dimethylthiazolium ion examined by Lienhard [Lienhard, G.E. (1966) J. Am. Chem. Soc. 88, 5642-5649], significant differences exist because of the pyrimidine ring in acetyl-TPP. Characterization of acetyl-TPP by ultraviolet spectroscopy, 1H NMR, 13C NMR, and 31P NMR provided evidence that the compound in aqueous solution exists as an equilibrium mixture of keto, hydrate, and intramolecular carbinolamine forms. The equilibria for the hydration and carbinolamine formation reactions at pD 1.3 as determined by 1H NMR are strongly dependent on the temperature, showing an increase in the hydrate and carbinolamine forms at the expense of the keto form with decreasing temperature. The concentration of keto form also decreases with increasing pH. Acetyl-TPP is stable in aqueous acid but rapidly deacetylates at higher pH to form acetate and thiamin pyrophosphate. Trapping of the acetyl moiety in aqueous solution occurs efficiently with 1.0 M hydroxylamine at pH 5.5-6.5 to form acetohydroxamic acid and to a much smaller extent with 1.0 M 2-mercaptoethanol at pH 4.0 and 5.0 to form thio ester. Transfer of the acetyl group to 0.5 M dihydrolipoic acid at pH 5.0 and 1.0 M phosphate dianion at pH 7.0 is not observed to any significant extent in water. The kinetic and thermodynamic reactivity of acetyl-TPP with water and other nucleophiles is compatible with a hypothetical role for acyl-TPPs as enzymatic acyl-transfer intermediates.