Redox properties of electron-transferring flavoprotein from Megasphaera elsdenii.

Electron-transferring flavoprotein (ETF) from the anaerobic bacterium Megasphaera elsdenii catalyzes electron transfer from NADH or D-lactate dehydrogenase to butyryl-CoA dehydrogenase. As a basis for understanding the interactions of ETF with its substrates, we report here on the redox properties of ETF alone. ETF exhibited reversible, two-electron transfer during electrochemical reduction in the presence of mediator dyes. The midpoint redox potentials of the FAD cofactor were -0.185 V at pH 5.5, -0.259 V at pH 7.1 and -0.269 +/- 0.013 V at pH 8.4, all versus the standard hydrogen electrode In the presence of the indicator dye 1-deazariboflavin, the Nernst slopes were 0.029 V and 0.026 V at pH 5.5 and pH 7.1, respectively, compared with an expected value of 0.028 V at 10 degrees C. At pH 8.4, in the presence of 2-hydroxy-1,4-naphthoquinone or phenosafranine, the Nernst slope varied from 0.021 V to 0.041 V. In the experiments at pH 8.4, equilibration was very slow in the reductive direction and a difference of as much as 30 mV was observed between reductive and oxidative midpoints. ETF exhibited no thermodynamic stabilization of the radical form of the FAD cofactor during electrochemical reduction at pH 5.5, 7.1 or 8.4. However, up to 93% of kinetically stable, anionic radical was produced by dithionite titration at pH 8.5. Molar absorptivities of ETF radical were 17,000 M-1 X cm-1 at 365 nm and 5100 M-1 X cm-1 at 450 nm. The four ETF preparations used here contained less than 7% 6-OH-FAD. However, two of the preparations contained significant amounts (up to 30%) of flavin which stabilized radical and reduced at potentials 0.2 V more positive than those required for reduction of the major form of ETF. This is referred to as the B form of ETF. The proportion of ETF-FAD in the B form was increased by incubation with free FAD or by a cycle of reduction and reoxidation. These treatments caused marked changes in the absorption spectrum of oxidized ETF and decreases of 20-25% in ETF units/A450.

Determination of the redox potential of deazariboflavin by equilibration with flavins.

The redox potential of deazariboflavin has been determined for pH values from 5.5 to 9.2 by equilibration with riboflavin and lumiflavin 3-acetate. The position of the equilibrium with riboflavin was measured spectrophotometrically and fluorimetrically; the equilibrium potential with lumiflavin 3-acetate was measured spectrophotometrically and potentiometrically. The Em7 for deazariboflavin was found to be--0.273 +/- 0.003 V against the standard hydrogen electrode. Equilibrium with flavodoxin at pH 9.5 and 10.0 was also used to determine the redox potential of deazariboflavin at high pH values. The pK of dihydrodeazariboflavin was found from the break in the potential vs. pH diagram and from spectrophotometric pH titration. The pK value obtained by both methods is 7.00 +/- 0.05. We found that borate, a product of the reducing agent borohydride, complexed with the ribityl sidechain of deazariboflavin, causing a shift in the pK for the reduced form to values of about 8.

Redox potential-pH properties of the flavoprotein L-amino-acid oxidase.

The redox potential-pH characteristics of the enzyme L-amino-acid oxidase (L-amino-acid: oxygen oxidoreductase (deaminating), EC 1.4.3.2) have been measured in the pH range 6.2 to 8.3 at 4 degrees C. All potentials are reported versus the standard hydrogen electrode. Consistent with the protonation states proposed for the anionic red semiquinone and anionic dihydroquinone forms of the flavin coenzyme, the first electron potential is independent of pH (-0.056 +/- 0.006 V), while the second electron transfer potential is pH-dependent, exhibiting a 0.060 V/pH unit slope. At all pH values investigated, the percentage of semiquinone species observed matches closely that calculated from measured potential separations. The semiquinone species is thermodynamically stable, as indicated by formation of semiquinone, similarity of redox potentials in oxidative and reductive directions, and by the slope of Nernst plots.

The effects of reversible freezing inactivation and inhibitor binding on redox properties of L-amino-acid oxidase.

We measured the redox potentials of frozen inactivated L-amino-acid oxidase (L-amino-acid:oxygen oxidoreductase (deaminating), EC 1.4.3.2) and inhibitor-bound (anthranilic acid) enzyme, and compared these redox properties to those of active L-amino-acid oxidase and benzoate-bound D-amino-acid oxidase (EC 1.4.3.3), respectively. The redox properties of the inactive enzyme are similar to the properties of free flavin; the potential is within 0.015 V of free flavin and no radical stabilization is seen. This corresponds to the loss of most interactions between apoprotein and flavin. In contrast, the anthranilic acid lowers the amount of radical stabilized from 85% to 35%. The potentials are still 0.150 V positive of free flavin, indicating that in the presence of inhibitor, many flavin-protein interactions remain intact. The difference between this behavior and that of D-amino-acid oxidase bound to benzoate, where the amount of radical declined from 95% to 5%, is explained on the basis of the relative tightness of binding of apoprotein to FAD. D-Amino-acid oxidase apoprotein has a relatively low Ka (10(6)) for FAD, and benzoate has a relatively high Ka (10(5)) for the enzyme. Therefore, the binding of benzoate increases the tightness of FAD binding to apo-D-amino-acid oxidase (10(11)), indicating significant changes in flavin-protein interactions. In contrast, apo-L-amino-acid oxidase binds flavin tightly (the Ka is greater than 10(7)) and the enzyme binds to anthranilate much less tightly, with a Ka of 10(3). The L-amino-acid oxidase apoprotein binding to FAD is tight initially, and the binding of anthranilate changes it only slightly.(ABSTRACT TRUNCATED AT 250 WORDS)

Influence of two substrate analogues on thermodynamic properties of medium-chain acyl-CoA dehydrogenase.

The objective of this work was to identify the key structural functionalities of substrate or product that modulate the thermodynamic properties of medium-chain acyl-CoA dehydrogenase (MCAD). In order to achieve this, two classes of substrate analogues, acetyl-CoA and thioether-CoAs, were complexed with MCAD and their effects on the redox properties of MCAD were measured. A pH dependence study of the redox potential of uncomplexed MCAD allowed us to compare redox properties between complexed and uncomplexed MCAD and to calculate the dissociation constants of the analogues to the three redox states of MCAD. The results from this work indicate that these analogues are not influencing the thermodynamic behavior of MCAD in the same way as natural substrate. Thus, we propose that the following two key structural features of the binding ligand are necessary for mimicking the thermodynamic effects natural substrate has on MCAD: a thioester carbonyl on carbon 1 and a fatty acyl-CoA chain length around 8 carbon units. Furthermore, with the advent of structural knowledge, insights into the interactions of these structural features with MCAD and their influence on MCAD's highly regulated dehydrogenation mechanism are discussed.

Oxidation-reduction properties of trimethylamine dehydrogenase: effect of inhibitor binding.

The redox potentials of trimethylamine dehydrogenase from the bacterium W3A1 have been determined by means of uv-visible spectroelectrochemistry. In the presence of the inhibitor tetramethylammonium chloride a shift of +0.2 V was observed in the midpoint redox potential for conversion of the oxidized 6-S-cysteinyl-FMN to the flavin radical form. The pH-independent value was +0.23 V vs the standard hydrogen electrode. The pH-dependent conversion of this radical to fully reduced flavin was shifted negative by 0.1 V in the presence of the inhibitor to -0.05 V at pH 7.0 and -0.15 V at pH 8.4. Tetramethylammonium chloride also caused moderate negative shifts (0.03-0.05 V) in the midpoint redox potential for the Fe4S(+2)4/Fe4S(+1)4 couple of trimethylamine dehydrogenase. The midpoint potentials are +0.06 V at pH 7.0 and +0.04 V at pH 8.4. Therefore, in the presence of tetramethylammonium chloride, electron transfer from the flavin radical to the Fe4S(+2)4 group is energetically unfavorable and trimethylamine dehydrogenase is trapped in the flavin radical state. The redox potential changes provide a thermodynamic basis for inhibition by tetramethylammonium chloride. Spectroelectrochemical titrations of trimethylamine dehydrogenase which had been inactivated by phenylhydrazine revealed heterogeneity in the redox behavior which had not been observed in other laboratories. The reason for this heterogeneity was not determined, but the midpoint redox potential for the Fe4S(+2)4/Fe4S(+1)4 couple of the main fraction of the inactivated enzyme was the same as that of active trimethylamine dehydrogenase.

Oxidation-reduction properties of short-chain acyl-CoA dehydrogenase: effects of substrate analogs.

Redox potentials of short-chain acyl-CoA dehydrogenase from the anaerobe, Megasphaera elsdenii, have been determined by means of uv-visible spectroelectrochemistry in the presence of substrate analogs. During redox titrations in the presence of 2-azabutyryl-CoA, up to 85% anionic FAD semiquinone was stabilized with a molar absorbance at 387 nm of 19 mM-1 cm-1. Despite a slow reduction of short-chain acyl-CoA dehydrogenase by 2-azabutyryl-CoA (< 2% reduction/h), a dissociation constant of 0.7 microM was measured and redox potentials, E1(0') and E2(0'), of -0.07 and -0.17 V, respectively, were determined at pH 7.0 for the first and second electrons in reduction of the FAD of short-chain acyl-CoA dehydrogenase. The analog, 2-azoctanoyl-CoA, did not reduce short-chain acyl-CoA dehydrogenase, bound with a dissociation constant of 2 microM, stabilized up to 47% anionic FAD semiquinone, and gave values of -0.08 and -0.11 V for E1(0') and E2(0') at pH 6.9. In contrast to 2-aza-acyl-CoA, the thioethers, butyl-CoA, octyl-CoA, and allyl-CoA, and the thioester, acetyl-CoA, did not bind strongly (Kd > or = 50 microM) and caused no significant change in the redox properties of short-chain acyl-CoA dehydrogenase. The two-electron redox potential, Em, remained at -0.08 V at pH 7.0 and there was no stabilization of FAD semiquinone in the presence of these analogs. These results show that no single feature of substrate structure, the thioester carbonyl, the presence of 2,3-unsaturation, or a fatty alkyl chain of appropriate length, can account for the 0.06-V positive change in redox potential which is observed in the presence of the substrate couple, crotonyl-CoA/butyryl-CoA (M. T. Stankovich and S. Soltysik (1987) Biochemistry 26 2627-2632). As outlined above, 2-azabutyryl-CoA or 2-azaoctanoyl-CoA did cause marked changes in the redox properties of short-chain acyl-CoA dehydrogenase, but the preferential stabilization of FAD semiquinone and negative change in Em distinguish the effects of 2-azaacyl-CoA from those of the substrate couple.

Regulation of the butyryl-CoA dehydrogenase by substrate and product binding.

Until now, workers in the field of fatty acid metabolism have suggested that the substrates are isopotential with the enzymes and that the reactions are forced to completion by the formation of charge-transfer complexes [Gustafson, W. G., Feinberg, B. A., & McFarland, J. T. (1986) J. Biol. Chem. 261, 7733-7741]. To date, no experimental evidence for this hypothesis exists. The work presented here shows that the butyryl-CoA/crotonyl-CoA couple is not isopotential with the enzymes with which it interacts. The potential of the butyryl-CoA/crotonyl-CoA couple (E ' = -0.013 V) is significantly more positive than the potential of either of the enzymes with which it interacts, bacterial butyryl-CoA dehydrogenase (E ' = -0.079 V) and mammalian general acyl-CoA dehydrogenase (E ' = 0.133 V). These data imply that the regulation of enzyme potential is essential for any electron transfer from substrate to enzyme to occur in mammalian or bacterial systems. In support of this assertion, a significant shift in potential for bacterial butyryl-CoA dehydrogenase (an analogue of the mammalian enzyme) in the presence of butyryl-CoA and crotonyl-CoA is reported. The potential is shifted positive by 60 mV. Larger potential shifts will undoubtedly be observed with the mammalian enzyme, which would be consistent with the catalytic direction of electron transfer.

Product binding modulates the thermodynamic properties of a Megasphaera elsdenii short-chain acyl-CoA dehydrogenase active-site mutant.

Previous work has shown that the redox properties of Megasphaera elsdenii short-chain acyl-CoA dehydrogenase (SCAD) are specifically modulated upon the binding of the substrate/product couple, allowing the reaction to proceed thermodynamically [Stankovich, M.T., & Soltysik, S. (1987) Biochemistry 26, 2627-2632]. The focus of this study on the Glu367Gln SCAD mutant protein is to gain an understanding of this phenomenon. The active-site mutant Glu367Gln SCAD inactivates the reductive and oxidative pathways and allows the effects of substrate (butyryl-CoA) and product (crotonyl-CoA) binding on the redox properties of the Glu367Gln SCAD mutant protein to be determined separately. Red anionic semiquinone was found to be thermodynamically stabilized in coulometric/potentiometric reductions of both butyryl-CoA- and crotonyl-CoA-complexed Glu367Gln SCAD. Reduction potential measurements showed that butyryl-CoA binding has little effect on the reduction potential of Glu367Gln SCAD. Crotonyl-CoA complexation, however, shifted the reduction potential of the Glu367Gln SCAD mutant protein by 30 mV in the positive direction. This modulation is similar to the 60-mV positive shift observed in native M. elsdenii SCAD upon complexation with the substrate/product couple [Stankovich, M.T., & Soltysik, S. (1987) Biochemistry 26, 2627-2632]. Thus, product binding and not substrate binding, thermodynamically regulates M. elsdenii SCAD. We propose that this observation is best explained by assuming that the product resembles an intermediate in the catalytic mechanism that is responsible for facilitating isopotential electron transfer from the substrate to the enzyme.

Oxidation-reduction potentials of butyryl-CoA dehydrogenase.

In order to obtain butyryl-CoA dehydrogenase from Megasphaera elsdenii in pure enough form to perform redox studies, the existing purification procedures first had to be modified and clarified [Engel, P. (1981) Methods Enzymol. 71, 359-366]. These modifications are described, and the previously unpublished spectral properties of the electrophoretically pure CoA-free butyryl-CoA dehydrogenase are presented. In our spectral reductive titration of pure enzyme, we show that although blue neutral flavin radical is stabilized in nonquantitative amounts in dithionite titrations (19%) or in electrochemical reductions mediated by methylviologen (5%), it is not thermodynamically stabilized; therefore, only a midpoint potential for butyryl-CoA dehydrogenase is obtained. The electron-transfer behavior from pH 5.5 to pH 7.0 indicates reversible two-electron transfer accompanied by one proton: EFlox + 2e- + H+ = EFlredH- Em7 = -0.079 V vs. SHE where EFlox is oxidized butyryl-CoA dehydrogenase, EFlredH- is two electron reduced enzyme, and Em7 is the midpoint potential at pH 7.0 at 25 degrees C. Redox data and activity data both indicate that the enzyme loses activity rapidly at pH values above 7.0. The Em7 of the butyryl-CoA dehydrogenase is 40 mV positive of the Em7 of the butyryl-CoA/crotonyl-CoA couple [Gustafson, W. G., Feinberg, B. A., & McFarland, J. T. (1986) J. Biol. Chem. 261, 7733-7741]. Binding of substrate analogue acetoacetyl-CoA caused the potential of butyryl-CoA dehydrogenase to shift 100 mV negative of the free enzyme. The negative shift in potential makes electron transfer from enzyme to substrate more probable, which is consistent with the direction of electron transfer in the bacterial system.(ABSTRACT TRUNCATED AT 250 WORDS)

Determination of glucose oxidase oxidation-reduction potentials and the oxygen reactivity of fully reduced and semiquinoid forms.

The oxidation-reduction potential values for the two electron transfers to glucose oxidase were obtained at pH 5.3, where the neutral radical is the stable form, and at pH 9.3, where the anion radical is the stable form. The midpoint potentials at 25 degrees were: pH 5.3 EFl1ox + e- H+ equilibrium EFlH. Em1 = -0.063 +/- 0.011 V EFlH. + e- + H+ equilibrium EFlredH2 Em2 = -0.065 +/- 0.007 V pH 9.3 EFlox + e- EFi- Em1 = -0.200 +/- 0.010 V EFi- + e- + H+ equilibrium EFlredH- Em2 = -0.240 +/- 0.005 V All potentials were measured versus the standard hydrogen electrode (SHE). The potentials indicated that glucose oxidase radicals are stabilized by kinetic factors and not by thermodynamic energy barriers. The pK for the glucose oxidase radical was 7.28 from dead time stopped flow measurements and the extinction coefficient of the neutral semiquinone was 4140 M-1 cm-1 at 570 nm. Both radical forms reacted with oxygen in a second order fashion. The rate at 25 degrees for the neutral semiquinone was 1.4 X 10(4) M-1 s-1; that for the anion radical was 3.5 X 10(4) M-1 s-1. The rate of oxidation of the neutral radical changed by a factor of 9 for a temperature difference of 22 degrees. For the anion radical, the oxidation rate changed by a factor of 6 for a 22 degrees change in temperature. We studied the oxygen reactivity of the 2-electron reduced form of the enzyme over a wide wavelength range and failed to detect either oxygenated flavin derivatives or semiquinoid forms as intermediates. The rate of reoxidation of fully reduced glucose oxidase at pH 9.3 was dependent on ionic strength.

Spectral and electrochemical properties of glutaryl-CoA dehydrogenase from Paracoccus denitrificans.

Studies of the spectral (UV/vis and resonance Raman) and electrochemical properties of the FAD-containing enzyme glutaryl-CoA dehydrogenase (GCD) from Paracoccus denitrificans reveal that the properties of the oxidized enzyme (GCDox) appear to be invariant from those properties known for other acyl-CoA dehydrogenases such as mammalian general acyl-CoA dehydrogenase (GACD) and butyryl-CoA dehydrogenase (BCD) from Megasphaera elsdenii. However, when either free or complexed GCD is reduced, its spectral and electrochemical behavior differs from that of both GACD and BCD. Free GCD does not stabilize any form of one-electron-reduced GCD, but when GCD is complexed to its inhibitor, aceto-acetyl-CoA, the enzyme stabilizes 20% of the blue neutral radical form of FAD (FADH.) upon reduction. Like GACD, when crotonyl-CoA- (CCoA) bound GCD is reduced, the red anionic form of FAD radical (FAD.-) is stabilized, and excess reduction equivalents are necessary to effect full reduction of the complex. A comproportionation reaction is proposed between fully reduced crotonyl-CoA-bound GCD (GCD2e-CCoA) and GCDox-CCoA to partially explain the stabilization of GCD-bound FAD.- by CCoA. When GCD is reduced by its optimal substrate, glutaryl-CoA, a two-electron reduction is observed with concomitant formation of a long-wavelength charge-transfer band. It is proposed that the ETF specific for GCD abstracts one electron from this charge-transfer species and this is followed by the decarboxylation of the oxidized substrate. At pH 6.4, potential values measured for free GCD and GCD bound to acetoacetyl-CoA are -0.085 and -0.129 V, respectively. Experimental evidence is given for a positive shift in the reduction potential of GCD when the enzyme is bound to a 1:1 mixture of butyryl-CoA and CCoA. However, significant GCD hydratase activity is observed, preventing quantitation of the potential shift.

Regulation of the redox potential of general acyl-CoA dehydrogenase by substrate binding.

Significant thermodynamic changes have been observed for general acyl-CoA dehydrogenase (GAD) upon substrate binding. Spectroelectrochemical studies of GAD and several of its substrates have revealed that these substrates are essentially isopotential for chain lengths of C-4 to C-16 (E 0' =-0.038 to -0.045 V vs SHE). When GAD is bound by these substrates, a dramatic shift in the midpoint potential of the enzyme is observed (E 0' = -0.136 V for ligand-free GAD and -0.026 V for acyl-CoA-bound GAD), thus allowing a thermodynamically favorable transfer of electrons from substrate to enzyme. This contrasts with values reported elsewhere. From these data an isopotential scheme of electron delivery into the electron-transport chain is proposed.

Studies of electron-transfer properties of salicylate hydroxylase from Pseudomonas cepacia and effects of salicylate and benzoate binding.

The pH dependence of the redox behavior of salicylate hydroxylase from Pseudomonas cepacia as well as the effects of salicylate, benzoate, and chloride binding is described. At pH 7.6 in 0.02 M potassium phosphate buffer E1(0')(EFl ox/EFl.-) is -0.150 V and E2(0')(EFl.-/EFl red H-) is -0.040 V versus the standard hydrogen electrode (SHE). A maximum of 5% of FAD anion semiquinone is thermodynamically stabilized under these conditions. However, in coulometric and dithionite titrations more semiquinone is kinetically formed, indicating slow transfer of the second electron. The potential/pH dependence is consistent with a two-electron, one-proton transfer. Upon salicylate binding the midpoint potential is shifted 0.020 V negative from -0.094 to -0.114 V vs SHE at pH 7.6. A maximum of 7% of the neutral semiquinone is stabilized both in potentiometric and coulometric titrations. This small potential shift indicates that the substrate is bound nearly to the same extent to all three oxidation states of the enzyme. It is clear that the substrate binding does not make the reduction of the flavin thermodynamically more favorable. In contrast to salicylate, the potential shift caused by the effector, benzoate, is much more significant. (A maximum potential shift of -0.07 V is calculated.) Benzoate binds most tightly to the oxidized form and is least tightly bound to the two-electron-reduced form of the enzyme. For the reduction of the free enzyme the transfer of the second electron or the transfer of the proton is rate limiting, as is shown by the kinetic formation of the anionic semiquinone.(ABSTRACT TRUNCATED AT 250 WORDS)

Measurement of the oxidation-reduction potentials for one-electron and two-electron reduction of electron-transfer flavoprotein from pig liver.

Potentiometric titrations of pig liver electron-transfer flavoprotein (ETF) were performed at pH 7.5 and 4 degrees C, both in the reductive and oxidative directions. Reduction of ETF to the hydroquinone form required a total of two reducing equivalents/mol of ETF with the formation of sub-stoichiometric amounts of anionic semiquinone as an intermediate. The oxidation-reduction potentials for the two one-electron couples, oxidized ETF/ETF semiquinone and ETF semiquinone/fully reduced ETF, are +4 mV and -50 mV respectively. The overall midpoint potential for the two-electron couple (oxidized ETF/fully reduced ETF) is -23 mV.

Thermodynamic control of D-amino acid oxidase by benzoate binding.

The redox properties of D-amino acid oxidase (D-amino-acid: O2 oxidoreductase (deaminating) EC1.4.3.3) have been measured at 18 degrees C in 20 mM sodium pyrophosphate, pH 8.5, and in 50 mM sodium phosphate, pH 7.0. Over the entire pH range, 2 eq are required per mol of FAD in D-amino acid oxidase for reduction to the anion dihydroquinone. The red anion semiquinone is thermodynamically stable as indicated by the separation of the electron potentials and the quantitative formation of the semiquinone species. The first electron potential is pH-independent at -0.098 +/- 0.004 V versus SHE while the second electron potential is pH-dependent exhibiting a 0.060 mV/pH unit slope. The redox behavior of D-amino acid oxidase is consistent with that observed for other oxidase enzymes. On the other hand, the behavior of the benzoate-bound enzyme under the same conditions is in marked contrast to the thermodynamics of free D-amino acid oxidase. Spectroelectrochemical experiments performed on inhibitor-bound (benzoate) D-amino acid oxidase show that benzoate binding regulates the redox properties of the enzyme, causing the energy levels of the benzoate-bound enzyme to be consistent with the two-electron transfer catalytic function of the enzyme. Our data are consistent with benzoate binding at the enzyme active site destroying the inductive effect of the positively charged arginine residue. Others have postulated that this positively charged group near the N(1)C(2) = O position of the flavin controls the enzyme properties. The data presented here are the clearest examples yet of enzyme regulation by substrate which may be a general characteristic of all flavoprotein oxidases.

Electrochemical properties of the diiron core of uteroferrin and its anion complexes.

The reduction potentials (Em) of the purple acid phosphatase from porcine uterus, uteroferrin (Uf), and its phosphate, arsenate, and molybdate complexes were determined by coulometric methods at various pH values. The midpoint potential of Uf at the pH value for optimal enzyme activity (pH 5) was found to be +367 mV versus a normal hydrogen electrode (NHE), while at pH 6.01 Uf exhibits a reduction potential of +306 mV. At pH 6.01 molybdate was found to shift the potential of Uf more positive by 192 mV, while phosphate and arsenate shift the potential of Uf more negative by 193 and 89 mV, respectively. These shifts are consistent with the different susceptibilities of Uf to aerobic oxidation in the presence of these anions. Comparison of the reduction potential of Uf at pH 7.0 with those reported for other dinuclear non-heme iron enzymes and various (mu-oxo)diiron model complexes suggest that the potential of Uf is too positive to be consistent with a mu-oxo-bridge in Ufo. The pH dependence of the reduction potentials of Uf (60 mV/pH unit) and the fact that the electron transfer rate increases with decreasing pH indicate a concomitant participation of a proton during the oxidation-reduction process. This process was assigned to the protonation of a terminally bound hydroxide ligand at the Fe(II) center upon reduction of Ufo. Structural implications provided by the electrochemical data indicate that molybdate affects the dinuclear core in a manner that differs from that of phosphate and arsenate. This observation is consistent with previous spectroscopic and biochemical studies.(ABSTRACT TRUNCATED AT 250 WORDS)

Three-dimensional structure of butyryl-CoA dehydrogenase from Megasphaera elsdenii.

The crystal structure of butyryl-CoA dehydrogenase (BCAD) from Megasphaera elsdenii complexed with acetoacetyl-CoA has been solved at 2.5 A resolution. The enzyme crystallizes in the P422 space group with cell dimensions a = b = 107.76 A and c = 153.67 A. BCAD is a bacterial analog of short chain acyl-CoA dehydrogenase from mammalian mitochondria. Mammalian acyl-CoA dehydrogenases are flavin adenine dinucleotide (FAD)-containing enzymes that catalyze the first step in the beta-oxidation of fatty acids. Although specific for substrate chain lengths, they exhibit high sequence homology. The structure of BCAD was solved by the molecular replacement method using the atomic coordinates of pig liver medium chain acyl-CoA dehydrogenase (MCAD). The structure was refined to an R-factor of 19.3%. The overall polypeptide fold of BCAD is similar to that of MCAD. E367 in BCAD is at the same position and in a similar conformation as the catalytic base in MCAD, E376. The main enzymatic differences between BCAD and MCAD are their substrate specificities and the significant oxygen reactivity exhibited by BCAD but not by MCAD. The substrate binding cavity of BCAD is relatively shallow compared to that of MCAD, as consequences of both a single amino acid insertion and differences in the side chains of the helices that make the binding site. The si-face of the FAD in BCAD is more exposed to solvent than that in MCAD. Therefore solvation can stabilize the superoxide anion and considerably increase the rate of oxidation of reduced flavin in the bacterial enzyme.

Electron transfer properties of the R2 protein of ribonucleotide reductase from Escherichia coli.

The enzyme ribonucleotide reductase from Escherichia coli consists of two proteins, R1 and R2. The active R2 protein contains two dinuclear iron centers and the catalytically essential tyrosyl radical. We have explored the redox properties of the tyrosyl radical and estimate an apparent redox potential of +1000 +/- 100 mV (vs SHE) on the basis of the behavior of numerous mediators. The inability of most of these mediators to equilibrate with the tyrosyl radical supports the notion that the radical exists in an extremely protected hydrophobic pocket that prevents most radical scavengers from interacting with the radical, resulting in its unusual stability. The formal midpoint potential of the diiron clusters of the R2 protein was determined to be -115 +/- 2 mV at pH 7.6 and 4 degrees C. This reduction is a two-electron transfer process, making the R2 protein the first of the nonheme diiron proteins not to stabilize a mixed valence intermediate at ambient temperature. The formal midpoint potential of the dinuclear iron centers is pH dependent, exhibiting a 30 mV/pH unit variance, which indicates that one proton is accepted from the solvent per two electrons transferred to the dinuclear iron center upon reduction. The midpoint potential of the site-directed mutant Y122F R2 protein was also investigated under the same conditions, and this midpoint potential was determined to be -178 mV, providing the first direct evidence that the presence of the Y122 residue modulates the redox properties of the diiron clusters. The redox potentials of both the wild type and Y122F proteins experience cathodic shifts when measured in the presence of azide or the R1 protein. For the latter, the midpoint potentials were determined to be -226 mV for the wild type protein and -281 mV for the Y122F mutant protein, representing a negative shift of over 100 mV for both proteins. These results indicate that the presence of the Y122 residue does not influence the effect of R1 binding, that the R1 protein preferentially binds the oxidized form of R2, and that the binding of R1 acts as a regulatory control mechanism to prevent unnecessary turnover of the dinuclear iron centers.