Kinetic studies of reduction of a 1:1 cytochrome c-flavodoxin complex by free flavin semiquinones and rubredoxin

The kinetics of reduction by free flavin semiquinones and reduced rubredoxin of the individual components of the 1:1 complex formed between horse heart cytochrome c and Clostridium pasteurianum flavodoxin have been studied. Complex formation did not affect the rate constant for reduction of flavodoxin by 5-deazariboflavin semiquinone, indicating that the accessibility of the flavin mononucleotide (FMN) of complexed flavodoxin is the same as in the free protein. Reduction of the complexed cytochrome c by the neutral flavin semiquinones of lumiflavin and riboflavin was significantly affected by complex formation (2-3-fold rate constant decrease), indicating that there are steric constraints on the accessibility of the cytochrome heme to small exogenous reductants. Reduction of complexed cytochrome c by the negatively charged semiquinones of FMN and Cl2FMN was also characterized. A repulsive electrostatic interaction between the reductants and complexed cytochrome was observed, whereas with free cytochrome an attractive interaction had previously been found. This is consistent with the presence of negative electrostatic potential at the protein interface due to uncompensated flavodoxin carboxylates, as predicted by Matthew et al. [Matthew, J. B., Weber, P. C., Salemme, F. R., & Richards, F. M. (1983) Nature (London) 301, 169-171]. Further, pseudo-first-order rate constants for the reduction of complexed cytochrome by these flavins had a nonlinear concentration dependence, rather than obeying simple second-order kinetics. This is interpreted by using a mechanism involving a rate-determining structural isomerization of the protein complex prior to the second-order electron-transfer step. The magnitude of the decrease in the rate constant for reduction of complexed cytochrome c by the negatively charged reduced rubredoxin was approximately the same as observed for free flavins. Furthermore, simple second-order kinetics were obtained, and the apparent electrostatic interaction between rubredoxin and the complex was attractive. These results suggest that flavodoxin was partially displaced from its complex with cytochrome c by a collisional interaction with rubredoxin. The effects of complexation on the kinetics have been correlated with a solvent-accessible surface representation of the computer-generated model of the flavodoxin-cytochrome c complex [Simondsen, R. P., Weber, P. C., Salemme, F. R., & Tollin, G. (1982) Biochemistry 21, 6366-6375]. The experimental observations are generally consistent with the structural model but clearly require the invocation of dynamic motions at the protein-protein interface.

Kinetics of electron transfer between cytochromes c' and the semiquinones of free flavin and clostridial flavodoxin

Rate constants have been measured for the reactions of a series of high-spin cytochromes c' and their low-spin homologues (cytochromes c-554 and c-556) with the semiquinones of free flavins and flavodoxin. These cytochromes are approximately 3 times more reactive with lumiflavin and riboflavin semiquinones than are the c-type cytochromes that are homologous to mitochondrial cytochrome c. We attribute this to the greater solvent exposure of the heme in the c'-type cytochromes. In marked contrast, the cytochromes c' are 3 orders of magnitude less reactive with flavodoxin semiquinone than are the c-type cytochromes. We interpret this result to be a consequence of the location of the exposed heme in cytochrome c' at the bottom of a deep groove in the surface of the protein, which is approximately 10-15 A deep and equally as wide. While free flavins are small enough to enter the groove, the flavin mononucleotide (FMN) prosthetic group of flavodoxin is apparently prevented by steric constraints from approaching the heme more closely than approximately 10 A without dynamic structural rearrangements. Most cytochromes c' are dimeric, but a few are monomeric. The three-dimensional structure of the Rhodospirillum molischianum cytochrome c' dimer suggests that the heme should be more exposed in the monomer than in the dimer, but no relationship is observed between intrinsic reactivity toward free flavin semiquinones and the aggregation state of the protein. Likewise, there is no evidence that the spin state or ligand field of the iron has any effect on intrinsic reactivity.(ABSTRACT TRUNCATED AT 250 WORDS)

Electron transfer to nitrogenase in Klebsiella pneumoniae. nifF gene cloned and the gene product, a flavodoxin, purified

The nifF gene of Klebsiella pneumoniae was cloned into a multicopy plasmid in order to construct a strain that synthesizes and retains an elevated concentration of the gene product relative to the wild-type strain. Characterization of the isolated flavodoxin, which serves as an electron donor to nitrogenase, shows unambiguously that it is the product of the nifF gene.

Kinetics of reduction of high redox potential ferredoxins by the semiquinones of Clostridium pasteurianum flavodoxin and exogenous flavin mononucleotide. Electrostatic and redox potential effects

We have measured the ionic strength dependence of the rate constants for the electron-transfer reactions of flavin mononucleotide (FMN) and flavodoxin semiquinones with 10 high redox potential ferredoxins (HiPIP's). The rate constants were extrapolated to infinite ionic strength by using a theoretical model of electrostatic interactions developed in our laboratory. In all cases, the sign of the electrostatic interaction was the same as the protein net charge, but the magnitudes were much smaller. The results are consistent with a model in which the electrical charges are approximately uniformly distributed over the HiPIP surface and in which there are both short- and long-range electrostatic interactions. An electrostatic field calculation for Chromatium vinosum HiPIP is consistent with this. The presumed site of electron transfer includes that region of the protein surface to which the iron-sulfur cluster is nearest and appears to be relatively hydrophobic. The principal short-range electrostatic interaction would involve the negative charge on the iron-sulfur cluster. For some net negatively charged proteins, this effect is magnified, and for net positively charged HiPIP's, it is counterbalanced. The rate constants extrapolated to infinite ionic strength can be correlated with redox potential differences between the reactants, as has previously been shown for cytochrome-flavin semiquinone reactions. Both electrostatic and redox potential effects are magnified for the flavodoxin semiquinone as compared to the FMN semiquinone-HiPIP reactions. This was also observed previously for the flavin semiquinone-cytochrome reactions.(ABSTRACT TRUNCATED AT 250 WORDS)

Carbon-13 and nitrogen-15 nuclear-magnetic-resonance investigation on Desulfovibrio vulgaris flavodoxin

Desulfovibrio vulgaris apoflavodoxin has been reconstituted with 15N and 13C-enriched riboflavin 5'-phosphate. For the first time all carbon atoms of the isoalloxazine ring of the protein-bound prosthetic group have been investigated. The reconstituted protein was studied in the oxidized and in the two-electron-reduced state. The results are interpreted in terms of specific interactions between the apoprotein and the prosthetic group, and the chemical structure of protein-bound FMN. In the oxidized state weak hydrogen bonds exist between the apoprotein and the N(5), N(3) and O(4 alpha) atoms of FMN. The N(1) and O(2 alpha) atoms of FMN form strong hydrogen bonds. The isoalloxazine ring of FMN is strongly polarized and the N(10) atom shows an increased sp2 hybridisation compared to that of free FMN in aqueous solution. The N(3)-H group is not accessible to bulk solvent, as deduced from the coupling constant of the N(3)-H group. In the reduced state the hydrogen bond pattern is similar to that in the oxidized state and in addition a strong hydrogen bond is observed between the N(5)-H group of FMN and the apoprotein. The reduced prosthetic group possesses a coplanar structure and is ionized. The N(3)-H and N(5)-H groups are not accessible to solvent water. Two-electron reduction of the protein leads to a large electron density increase in the benzene subnucleus of bound FMN compared to that in free FMN. The results are discussed in relation to the published crystallographic data on the protein.

Proton NMR study of the cytochrome c:flavodoxin electron transfer complex

The effects of complex formation with flavodoxin on the proton NMR spectrum of cytochrome c are to change the resonance frequencies and to increase the bandwidths of most of the low and high field heme, Met-80, and His-18 protons. These effects are, in general, more pronounced than has been reported for other cytochrome c complexes. The degree of line broadening for many heme related resonances suggests that complex formation induces changes in the cytochrome structure. These results provide the first spectroscopic evidence which corroborates the proposed model for the cytochrome c: flavodoxin complex (1-3).

A modified flavodoxin with altered redox potentials is less efficient in electron transfer to nitrogenase

The flavodoxins of the Azotobacter vinelandii wild-type and a mutant strain TZN 200 have been studied. Although the primary structure of the two proteins is the same, the ability of the mutant flavodoxin to donate electrons to nitrogenase is reduced by 75%. One reason may be the raised mid-point potential of -435 mV for the semiquinone/hydroquinone couple in the mutant flavodoxin. The respective redox potential for the wild-type flavodoxin was found to be -480 mV. As shown by paper chromatography and light absorption spectroscopy, the structure of FMN is modified in the TZN 200 flavodoxin.

The catalytic activity of nitrogenase in intact Azotobacter vinelandii cells

The influence of the growth conditions on the concentration of nitrogenase and on the nitrogenase activity, was studied in intact Azotobacter vinelandii cells. It was observed that whole cell nitrogenase activity could be enhanced in two ways. An increase of the growth rate of cells was accompanied by an increase in whole cell nitrogenase activity and by an increase in the concentration of nitrogenase in the cells. The molar ratio of Fe protein:MoFe protein was 1.47 +/- 0.17 and independent of the growth rate. Activity measurements in cell extracts showed that the catalytic activity of the nitrogenase proteins was independent of the growth rate of cells. The second way to increase whole cell nitrogenase activity was to expose cells to excess oxygen. Whole cells were exposed for 2.5 h to an enhanced oxygen-input rate. After this incubation nitrogenase activity was increased without an increase in protein concentration. It is calculated that the catalytic activity of the Fe protein in these cells was 6200 nmol C2H4 formed X min-1 X (mg Fe protein)-1. With these cells and with cells grown at a high growth rate, 50% of the whole cell activity is lost by preparing a cell-free extract. It will be demonstrated that this inactivation is partly caused by the activity measurements in vitro. When dithionite was replaced by flavodoxin as electron donor, a maximal catalytic activity of 4500 nmol C2H4 formed X min-1 X (mg Fe protein)-1 was measured in vitro for the Fe protein. The results are discussed in relation to the present model for nitrogenase catalysis.

On the intermolecular electron transfer between different redox states of flavodoxin from Megasphaera elsdenii. A 500-MHz 1H NMR study

The electron transfer reactions between molecules of flavodoxin from Megasphaera elsdenii in different redox states have been investigated by proton nuclear magnetic resonance techniques at 500 MHz. The electron transfer between molecules in the oxidized and semiquinone state is shown to be at least 350-times slower than that between molecules in the semiquinone and hydroquinone state. The latter reaction was studied at different ionic strengths and temperatures. The rate of electron transfer increases with increasing ionic strength, as expected for a reaction between molecules of identical charges. The electron transfer reaction is only slightly dependent on temperature suggesting an outer sphere reaction mechanism. The results indicate that the activation energy for the electron transfer reaction between the semiquinone and hydroquinone state is negligible in contrast to that between the oxidized and semiquinone state. It is suggested that this feature renders M. elsdenii flavodoxin to an exclusive one-electron donor/acceptor in the cell, thereby shuttling between the semiquinone and the hydroquinone state. Mechanistic implications of the findings are briefly discussed.

The thermodynamics of flavin binding to the apoflavodoxin from Azotobacter vinelandii

A thermodynamic study of the binding of flavins (FMN, FAD, 8-carboxylic acid-riboflavin) to the purified apoflavodoxin from Azotobacter vinelandii has been conducted. The binding of FMN was studied at a number of temperatures (10, 15, 20, 25, and 30 degrees C), pH's (6.0, 7.4, and 9.0), and buffer conditions. The binding of FAD was studied at pH 7.4 and 25 degrees C under a number of buffer conditions. The binding of 8-carboxylic acid-riboflavin to the apoflavodoxin and the binding of FMN to the dimeric form of the apoflavodoxin were investigated at pH 7.4 and 25 degrees C. Enthalpies of binding for FMN, FAD, and 8-carboxylic- acid-riboflavin were -28.3, -16.6, and -14.0 kcal mol-1, respectively. The enthalpy of binding of FMN to the dimeric form of the apoflavodoxin was -22.2 kcal mol of binding sites-1. Binding constants of about 10(8), 10(6), and 10(6) were obtained for the binding of FMN, FAD, and 8-carboxylic acid-riboflavin, respectively. Using established thermodynamic relationships free energy and entropy changes were calculated. The entropy data indicate that a large degree of ordering of the system occurs upon flavin binding. The pH data suggest that FMN may bind in both the mono- and dianion forms, and that binding doesn't change the pKa of any functional group in the system. It appears that the phosphate group is probably responsible for approximately half the binding enthalpy observed for the binding of FMN. The temperature-dependence data over the temperature range studied is biphasic, centered at 20 degrees C, indicating that flavin binding occurs to the protein in two thermodynamic states corresponding to the two heat capacities observed. These findings are used to discuss a model for flavin binding.

Redox potentials of algal and cyanobacterial flavodoxins

The redox potentials of flavodoxins from the cyanobacteria Synechococcus PCC 6301 (formerly Anacystis nidulans) and Nostoc strain MAC, and from the red alga Chondrus crispus, were determined by potentiometric titration. For the oxidized-semiquinone interconversion the potentials at pH 7.0 of the three flavodoxins were between -210 and -235 mV, and these were pH-dependent over the range pH 6.9-8.2. For the semiquinone-reduced interconversion the potentials of the cyanobacterial flavodoxins were close to -414 mV, and that for the algal flavodoxin, -370 mV, is the highest reported in this group of flavoproteins.

Characterization of ferredoxin, flavodoxin, and rubredoxin from Clostridium formicoaceticum grown in media with high and low iron contents

Ferredoxin, flavodoxin, and rubredoxin were purified to homogeneity from Clostridium formicoaceticum and characterized. Variation of the iron concentration of the growth medium caused substantial changes in the concentrations of ferredoxin and flavodoxin but not of rubredoxin. The ferredoxin has a molecular weight of 6,000 and is a four iron-four sulfur protein with eight cysteine residues. The spectrum is similar to that of other ferredoxins. The molar extinction coefficients are 22.6 X 10(3) and 17.6 X 10(3) at 280 and 390 nm, respectively. From 100 g wet weight of cells grown with 3.6 microM iron and with 40 microM iron, 5 and 20 mg offerredoxin were isolated, respectively. The molecular weight of rubredoxin is 5,800 and it contains one iron and four cysteines. The UV-visible absorption spectrum is dissimilar to those of other rubredoxins in that the 373 nm absorption peak is quite symmetric, lacking the characteristic 350-nm shoulder found in other rubredoxins. The flavodoxin is a 14,500-molecular-weight protein which contains 1 mol of flavin mononucleotide per mol of protein. It forms a stable, blue semiquinone upon light irradiation in the presence of EDTA or during enzymatic reduction. When cells were grown in low-iron medium, flavodoxin constituted at least 2% of the soluble cell protein; however, it was not detected in extracts of cells grown in high-iron medium. The rubredoxin and ferredoxin expressed during growth in low-iron and high-iron media are identical as judged by iron, inorganic sulfide, and amino acid analysis, as well as light absorption spectroscopy.

Transient kinetics of redox reactions of flavodoxin: effects of chemical modification of the flavin mononucleotide prosthetic group on the dynamics of intermediate complex formation and electron transfer

The effects of structural modifications of the flavin mononucleotide (FMN) prosthetic group of Clostridium pasteurianum flavodoxin on the kinetics of electron transfer to the oxidized form (from 5-deazariboflavin semiquinone produced by laser flash photolysis) and from the semiquinone form (to horse heart cytochrome c by using stopped-flow spectrophotometry) have been investigated. The analogues used were 7,8-dichloro-FMN, 8-chloro-FMN, 7-chloro-FMN, and 5,6,7,8-tetrahydro-FMN. The ionic strength dependence of cytochrome c reduction was not affected by chlorine substitution, although the specific rate constants for complex formation and decay were appreciably smaller. On the other hand, all of the chlorine analogues had the same rate constant for deazariboflavin semiquinone oxidation. The rate constants for tetrahydro-FMN flavodoxin semiquinone reduction of cytochrome c were considerably smaller than those for the native protein. The implications of these results for the electron-transfer mechanism of flavodoxin are discussed.

Structure of oxidized flavodoxin from Anacystis nidulans

The structure of oxidized flavodoxin from the cyanobacterium Anacystis nidulans has been determined at 2.5 A resolution with phases calculated from ethylmercury phosphate and dimercuriacetate derivatives. The determination of partial sequences, including a total of 85 residues, has assisted in the interpretation of the electron density. Preliminary refinement of a partial model (1072 atoms) has reduced R to 0.349 for the 10.997 reflections between 2.0 and 5.0 A with 1 greater than 2 sigma. The polypeptide backbone, which comprises 167 residues in the current model, adopts the familiar beta-alpha-beta conformation found in other flavodoxins and in the nucleotide-binding domains of the pyridine-nucleotide dehydrogenases, with five parallel strands in the central sheet. Comparison with flavodoxin from Clostridium MP (138 residues) shows that extra residues of A. nidulans flavodoxin are accommodated in a major insertion about 20 residues in length, which forms a lobe adjacent to the fifth strand of parallel sheet, and in additions to several external segments. Residues added between the fourth sheet strand and the start of the third helix alter the environment of the pyrimidine end of the flavin mononucleotide ring. The flavin mononucleotide phosphate binds to the start of helix 1, interacting with hydroxyamino acids and with main-chain amide groups. Two hydrophobic residues, both tentatively identified as Trp, enclose the isoalloxazine ring; the solvent-exposed Trp is nearly parallel to the flavin ring. The hydrophobic environment provided by these residues must be partly responsible for the pronounced vibrational resolution of the flavin spectrum near 450 nm. The flavin ring is tilted relative to its orientation in Clostridium MP flavodoxin. In addition, atoms N-3 and O-2 alpha of the isoalloxazine appear to form hydrogen bonds to the backbone at CO97 and NH99 in a conformation entirely different from that found in Clostridium MP flavodoxin but structurally analogous to Desulfovibrio vulgaris flavodoxin.

Electrostatic orientation during electron transfer between flavodoxin and cytochrome c

Various studies have shown that reaction rates between reversibly binding electron transfer proteins depend strongly on solution ionic strength. These observations suggest that intermolecular electrostatic interactions are important in facilitating the formation of a productive reaction complex. A recently examined system involves the reduction of vertebrate cytochrome c by bacterial flavodoxin. Although this is a nonphysiological reaction, it proceeds with rates typical for natural partners and is similarly inhibited at high ionic strengths. Here we describe computational studies which examine the role of electrostatics in the formation of a putative reaction complex between flavodoxin and cytochrome c. The results suggest that electrostatic interactions preorient the molecules before they make physical contact, facilitating the formation of an optimal reaction complex.

Transient kinetics of electron transfer reactions of flavodoxin: ionic strength dependence of semiquinone oxidation by cytochrome c, ferricyanide, and ferric ethylenediaminetetraacetic acid and computer modeling of reaction complexes

Electron transfer reactions between Clostridum pasteurianum flavodoxin semiquinone and various oxidants [horse heart cytochrome c, ferricyanide, and ferric ethylenediaminetetraacetic [horse heart cytochrome c, ferricyanide, and ferric ethylenediaminetetraacetic acid (EDTA)] have been studied as a function of ionic strength by using stopped-flow spectrophotometry. The cytochrome c reaction is complicated by the existence of two cytochrome species which react at different rates and whose relative concentrations are ionic strength dependent. Only the faster of these two reactions is considered here. At low ionic strength, complex formation between cytochrome c and flavodoxin is indicated by a leveling off of the pseudo-first-order rate constant at high cytochrome c concentration. This is not observed for either ferricyanide or ferric EDTA. For cytochrome c, the rate and association constants for complex formation were found to increase with decreasing ionic strength, consistent with negative charges on flavodoxin interacting with the positively charged cytochrome electron transfer site. Both ferricyanide and ferric EDTA are negatively charged oxidants, and the rate data respond to ionic strength changes as would be predicted for reactants of the same charge sign. These results demonstrate that electrostatic interactions involving negatively charged groups are important in orienting flavodoxin with respect to oxidants during electron transfer. We have also carried out computer modeling studies of putative complexes of flavodoxin with cytochrome c and ferricyanide, which relate their structural properties to both the observed kinetic behavior and some more general features of physiological electron transfer processes. The results of this study are consistent with the ionic strength behavior described above.

A photo-CIDNP study of the active sites of Megasphaera elsdenii and Clostridium MP flavodoxins

Megasphaera elsdenii and Clostridium MP flavodoxins have been investigated by photo-CIDNP techniques. Using time-resolved spectroscopy and external dyes carrying different charges it was possible to assign unambiguously the resonance lines in the NMR-spectra to tyrosine, tryptophan and methionine residues in the two proteins. The results show that Trp-91 in M.elsdenii and Trp-90 in Cl.MP flavodoxin are strongly immobilized and placed directly above the benzene subnucleus of the prosthetic group. The data further indicate that the active sites of the two flavodoxins are extremely similar.

Biochemistry of dissimilatory sulphate reduction

Extensive information is available on the enzymology of respiratory sulphate reduction and the structure of electron transfer proteins isolated from the sulphate-reducing bacteria; however, it has not yet been possible to delineate satisfactorily the function of these electron transfer proteins in terms of the enzymes involved in respiratory sulphate reduction. New information about differences in pyrophosphate metabolism by Desulfovibrio and Desulfotomaculum, cellular localizations of electron transfer proteins and enzymes, and the concepts of vectorial electron transfer plus hydrogen cycling suggest that previous data on the function of electron transfer proteins must be re-evaluated and new experimental approaches designed before the problem is resolved. New information on the enzymology of lactate dehydrogenase, pyruvate dehydrogenase, adenylyl sulphate reductase, bisulphite reductase and hydrogenase is presented and discussed in the context of enzyme localization and specifically for electron transfer proteins. The function of cytochrome c3 (Mr = 13000) in the mechanism of the periplasmic hydrogenase and the role of the new [3Fe-3S] non-haem iron centres in electron transfer is emphasized.

Routes of flavodoxin and ferredoxin reduction in Escherichia coli. CoA-acylating pyruvate: flavodoxin and NADPH: flavodoxin oxidoreductases participating in the activation of pyruvate formate-lyase

Flavodoxin and ferredoxin become reduced in Escherichia coli cells by oxidoreductase reactions which use pyruvate and NADPH as electron donor substrates. The two enzymes, which are minor proteins of this organism, were measured through the reduced flavodoxin-dependent activation of pyruvate formate-lyase. The NADPH-dependent enzyme, obtained homogeneously through Procion-red affinity chromatography, was identified as the flavoprotein 'component R' described previously by Fujii and Huennekens [J. Biol. Chem. 249, 6745-6753 (1974)]. The pyruvate-dependent enzyme was identified as CoA-acetylating pyruvate:flavodoxin (ferredoxin) oxidoreductase. Its catalytic properties in the forward, reverse, and the 14CO2-pyruvate exchange reaction are reported. The dihydro form of flavodoxin was characterized as the particular species involved in the activation of pyruvate formate-lyase. The activation process still occurs with 70% of maximal efficiency when the ratio [NADPH]/([NADP] + [NADPH]) is fixed at the intracellular 'anabolic reduction charge' value of 0.45, in conjunction with the NADPH-dependent enzyme. The [2Fe-2S] ferredoxin, though being readily used as electron acceptor of both oxidoreductases and having a redox potential similar to flavodoxin, proved incompetent in mediating the activation of pyruvate formate-lyase.

Transient kinetics of electron-transfer reactions of flavodoxins

Stopped-flow and laser photolysis methods have been used to investigate the rates of electron-transfer reactions of fully reduced riboflavin and the three oxidation states of Clostridium pasteurianum flavodoxin. Both normal and 7,8-dichloroflavin analogues were studied. Redox reagents included oxygen, ferricyanide, ferric EDTA, and several c-type cytochromes as oxidants and the semiquinone of 5-deazariboflavin as a reductant. The dependence of the rate of oxidation of the semiquinone form of the dichloro analogue flavodoxin upon oxidant concentration has provided clear evidence for the existence of a complex in the reaction pathway. Rate constant comparisons demonstrate that dichloro substitution decreases the rate of flavodoxin semiquinone oxidation by at least 1-2 orders of magnitude. The limiting first-order rate constants were found to be dependent on the redox potential of the oxidant, as would be predicted by theory if these were reflecting the actual electron-transfer reaction. Rate constant decreases upon chlorine substitution were also observed for the reduction of both oxidized and semiquinone forms of flavodoxin by deazariboflavin semiquinone. These results, considered in conjunction with the redox potential shift of the flavodoxin produced by the chlorine substitution, provide support for the hypothesis that electron transfer to and from the semiquinone form of the flavodoxin involves direct participation of the dimethylbenzene ring of the flavin. A comparison of oxidation rate constants for free and protein-bound fully reduced flavin suggests that the protein environment does not markedly influence coenzyme reactivity in this oxidation state.

Structure-function relations in flavodoxins

Flavodoxins are low molecular weight, FMN containing, proteins which function as electron transfer agents in a variety of microbial metabolic processes, including nitrogen fixation. Utilizing structural information obtained from x-ray crystal analysis, it has been possible to derive some new and important insights into the relationships which exist between flavin properties and protein environment by comparing the spectroscopic, thermodynamic and kinetic behavior of the flavodoxins with that of free flavin. Thus, for example, a qualitative understanding of the contribution of the protein to flavin redox potentials, semiquinone reactivity and mechanism of electron transfer is beginning to emerge. The highly negative redox potential required for the biochemical activity of the flavodoxins is accomplished by stabilizing the semiquinone via a hydrogen bond to the N-5 position of the flavin and destabilizing the fully-reduced form by constraining it to assume an unfavorable planar conformation. The reactivity of the semiquinone form is lowered by the aforementioned hydrogen bond, as well as by an interaction with a tryptophan residue in the binding site. Electron transfer is accomplished through the exposed dimethylbenzene ring of the bound coenzyme. Although it is not possible at present to determine the extent to which this understanding can be generalized to other flavoproteins, it is clear that a study of the flavodoxins will provide us with at least some of the principles which biological systems have used to modify flavin properties to fulfill a biochemical need.

Efficiency of ferredoxins and flavodoxins as mediators in systems for hydrogen evolution

1. The efficiencies of ferredoxins and flavodoxins from a range of sources as mediators in systems for hydrogen evolution were assessed. 2. In supporting electron transfer from dithionite to hydrogenase of the bacterium Clostridium pasteurianum, highest activity was shown by the ferredoxin from the cyanobacterium Chlorogloeopsis fritschii and flavodoxin from the bacterium Megasphaera elsdenii. The latter was some twenty times as active as comparable concentrations of Methyl Viologen. Ferredoxins from the cyanobacterium Anacystis nidulans and the red alga Porphyra umbilicalis also showed high activity. 3. In mediating electron transfer from chloroplast membranes to Clostridium pasteurianum hydrogenase the flavodoxin from Anacystis nidulans proved the most active with Nostoc strain MAC flavodoxin and Porphyra umbilicalis ferredoxin also being appreciably more active than other cyanobacterial and higher plant ferredoxins. 4. In both hydrogenase systems the ferredoxin and flavodoxin from the red alga Chondrus crispus and the ferredoxin from another red alga Gigartina stellata showed very low activity. 5. There appeared to be no apparent correlation of efficiency in supporting hydrogenase activity with midpoint redox potential (Em) of the mediators, though some correlation of Em with the efficiency of the mediators in supporting NADP+ photoreduction by chloroplasts, or pyruvate oxidation by a Clostridium pasteurianum system, was evident. 6. Activity of the mediators in the hydrogenase systems therefore primarily reflects differences in tertiary structure conferring differing affinities for the other components of the systems.