Dual role of FMN in flavodoxin function: electron transfer cofactor and modulation of the protein-protein interaction surface.

Flavodoxin (Fld) replaces Ferredoxin (Fd) as electron carrier from Photosystem I (PSI) to Ferredoxin-NADP(+) reductase (FNR). A number of Anabaena Fld (AnFld) variants with replacements at the interaction surface with FNR and PSI indicated that neither polar nor hydrophobic residues resulted critical for the interactions, particularly with FNR. This suggests that the solvent exposed benzenoid surface of the Fld FMN cofactor might contribute to it. FMN has been replaced with analogues in which its 7- and/or 8-methyl groups have been replaced by chlorine and/or hydrogen. The oxidised Fld variants accept electrons from reduced FNR more efficiently than Fld, as expected from their less negative midpoint potential. However, processes with PSI (including reduction of Fld semiquinone by PSI, described here for the first time) are impeded at the steps that involve complex re-arrangement and electron transfer (ET). The groups introduced, particularly chlorine, have an electron withdrawal effect on the pyrazine and pyrimidine rings of FMN. These changes are reflected in the magnitude and orientation of the molecular dipole moment of the variants, both factors appearing critical for the re-arrangement of the finely tuned PSI:Fld complex. Processes with FNR are also slightly modulated. Despite the displacements observed, the negative end of the dipole moment points towards the surface that contains the FMN, still allowing formation of complexes competent for efficient ET. This agrees with several alternative binding modes in the FNR:Fld interaction. In conclusion, the FMN in Fld not only contributes to the redox process, but also to attain the competent interaction of Fld with FNR and PSI.

Tuning of the FMN binding and oxido-reduction properties by neighboring side chains in Anabaena flavodoxin.

Contribution of three regions (phosphate-binding, 50's and 90's loops) of Anabaena apoflavodoxin to FMN binding and reduction potential was studied. Thr12 and Glu16 did not influence FMN redox properties, but Thr12 played a role in FMN binding. Replacement of Trp57 with Glu, Lys or Arg moderately shifted E(ox/sq) and E(sq/hq) and altered the energetic of the FMN redox states binding profile. Our data indicate that the side chain of position 57 does not modulate E(ox/sq) by aromatic stacking or solvent exclusion, but rather by influencing the relative strength of the H-bond between the N(5) of the flavin and the Asn58-Ile59 bond. A correlation was observed between the isoalloxazine increase in solvent accessibility and less negative E(sq/hq). Moreover, E(sq/hq) became less negative as positively charged residues were added near to the isoalloxazine. Ile59 and Ile92 were simultaneously mutated to Ala or Glu. These mutations impaired FMN binding, while shifting E(sq/hq) to less negative values and E(ox/sq) to more negative. These effects are discussed on the bases of the X-ray structures of some of the Fld mutants, suggesting that in Anabaena Fld the structural control of both electron transfer steps is much more subtle than in other Flds.

Molecular determinants for FMN-binding in Desulfovibrio gigas flavoredoxin.

Flavoredoxin participates in Desulfovibrio gigas thiosulfate reduction pathway. Its 3-dimensional model was generated allowing the oxidized riboflavin-5'-phosphate (FMN) site to be predicted. Residues likely to be involved in FMN-binding were identified (N29, W35, T56, K92, H131 and F164) and mutated to alanine. Fluorescence titration with apoprotein showed that FMN is strongly bound in the wild-type protein. Comparison of K(d) values for mutants suggests that interactions with the phosphate group of FMN, contribute more to binding than the interactions with the isoalloxazine ring. The redox potential of bound FMN determined for wild-type and mutants revealed shifts to less negative values. These findings were correlated with the protein structure in order to contribute to a better understanding of the structure-function relationships in flavoredoxin.

Identification and properties of an inducible phenylacyl-CoA dehydrogenase in Pseudomonas putida KT2440.

A novel acyl-CoA dehydrogenase that initiates beta-oxidation of the side chains of phenylacyl-CoA compounds by Pseudomonas putida was induced by growth with phenylhexanoate as carbon source. It was identified as the product of gene PP_0368, which was cloned and overexpressed in Escherichia coli. This phenylacyl-CoA dehydrogenase was found to be dimeric with a subunit molecular mass of 66 kDa, to contain FAD and to be active with phenylacyl-CoA substrates having side chains from four to at least 11 carbon atoms. The same enzyme was induced by the aliphatic alkanoate octanoate. The optimal aliphatic substrates for the enzyme were palmitoyl-CoA and stearoyl-CoA, a property shared with mammalian very-long-chain acyl-CoA dehydrogenases. The FAD in the enzyme was reduced by aromatic and aliphatic substrates, with changes to the oxidation-reduction potential. Chemical reduction by dithionite ion and oxidation by ferricyanide ion showed that the enzyme can accept four electrons: two to reduce the flavin and two to slowly reduce an unknown acceptor, which in its reduced form interacts with the oxidized flavin in a charge-transfer complex. The experiments identify for the first time an acyl-CoA dehydrogenase that oxidizes the activated forms of aromatic acids similar to those used to first demonstrate the biological beta-oxidation of fatty acids.

The ferredoxin-NADP(H) reductase from Rhodobacter capsulatus: molecular structure and catalytic mechanism.

The photosynthetic bacterium Rhodobacter capsulatus contains a ferredoxin (flavodoxin)-NADP(H) oxidoreductase (FPR) that catalyzes electron transfer between NADP(H) and ferredoxin or flavodoxin. The structure of the enzyme, determined by X-ray crystallography, contains two domains harboring the FAD and NADP(H) binding sites, as is typical of the FPR structural family. The FAD molecule is in a hairpin conformation in which stacking interactions can be established between the dimethylisoalloxazine and adenine moieties. The midpoint redox potentials of the various transitions undergone by R. capsulatus FPR were similar to those reported for their counterparts involved in oxygenic photosynthesis, but its catalytic activity is orders of magnitude lower (1-2 s(-)(1) versus 200-500 s(-)(1)) as is true for most of its prokaryotic homologues. To identify the mechanistic basis for the slow turnover in the bacterial enzymes, we dissected the R. capsulatus FPR reaction into hydride transfer and electron transfer steps, and determined their rates using stopped-flow methods. Hydride exchange between the enzyme and NADP(H) occurred at 30-150 s(-)(1), indicating that this half-reaction does not limit FPR activity. In contrast, electron transfer to flavodoxin proceeds at 2.7 s(-)(1), in the range of steady-state catalysis. Flavodoxin semiquinone was a better electron acceptor for FPR than oxidized flavodoxin under both single turnover and steady-state conditions. The results indicate that one-electron reduction of oxidized flavodoxin limits the enzyme activity in vitro, and support the notion that flavodoxin oscillates between the semiquinone and fully reduced states when FPR operates in vivo.

The protein coded by the PP2216 gene of Pseudomonas putida KT2440 is an acyl-CoA dehydrogenase that oxidises only short-chain aliphatic substrates.

A gene (PP2216) that codes for an acyl-CoA dehydrogenase was cloned from Pseudomonas putida strain KT2240 and over-expressed in Escherichia coli, and the recombinant enzyme purified and characterised. The enzyme is tetrameric with one FAD per subunit of molecular mass 40,500 Da. An anaerobic titration with sodium dithionite showed that the enzyme accepts two electrons. A similar titration with butyryl-CoA showed that reduction by this substrate was incomplete with 4.5 mol butyryl-CoA added per mol enzyme FAD; the equilibrium was used to calculate that the oxidation-reduction potential of the enzyme at pH 7 and 25 degrees C is 5+/-5 mV versus the standard hydrogen electrode. The enzyme shows catalytic activity with butyryl-CoA, valeryl-CoA and hexanoyl-CoA, and very low activity with heptanoyl-CoA and octanoyl-CoA; it fails to oxidise propionyl-CoA. These properties resemble those of short-chain acyl-CoA dehydrogenases from other sources. The enzyme is inactive with the CoA derivatives of all phenylalkanoates that were tested (side chains 3-8 carbon atoms) indicating that in contrast to an earlier suggestion, the enzyme is not involved in the beta-oxidation of aromatic compounds.

Role of neighboring FMN side chains in the modulation of flavin reduction potentials and in the energetics of the FMN:apoprotein interaction in Anabaena flavodoxin.

Flavodoxins (Flds) are electron transfer proteins that carry a noncovalently bound flavin mononucleotide molecule (FMN) as a redox active center. A distinguishing feature of these flavoproteins is the dramatic change in the E(sq/rd) reduction potential of the FMN upon binding to the apoprotein (at pH 8.0, from -269 mV when free in solution to -438 mV in Anabaena Fld). In this study, the contribution of three neighboring FMN residues, Thr56, Asn58, and Asn97, and of three negatively charged surface residues, Glu20, Asp65, and Asp96, to modulate the redox properties of FMN upon its binding to the apoprotein has been investigated. Additionally, the role of these residues in the apoflavodoxin:FMN interaction has been analyzed. Concerning the redox potentials, the most noticeable result was obtained for the Thr56Gly mutant. In this Fld variant, the increased accessibility of FMN leads to an increase of +63 mV in the E(sq/rd) value. On the other hand, a correlation between the electrostatic environment of FMN and the E(sq/rd) has been observed. The more positive residues or the less negative residues present in the surroundings of the FMN N(1) atom, then the less negative the value for E(sq/rd). With regard to FMN binding to apoflavodoxin, breaking of hydrophobic interactions between FMN and residues 56, 58, and 97 seems to increase the K(d) values, especially in the Thr56Gly Fld. Such results suggest that the H-bond network in the FMN environment influences the FMN affinity.

Cofactor-apoprotein hydrogen bonding in oxidized and fully reduced flavodoxin monitored by trans-hydrogen-bond scalar couplings.

Hydrogen bonding plays a key role in the tight binding of the FMN cofactor and the regulation of its redox properties in flavodoxins. Hydrogen bonding interactions can be directly observed in solution by multidimensional heteronuclear NMR spectroscopy through the scalar couplings between donor and acceptor nuclei. Here we report on the detection of intermolecular trans-hydrogen-bond couplings ((h)J) between the flavin ring system and the backbone of Desulfovibrio vulgaris flavodoxin in the oxidized and the two-electron reduced states. For this purpose, experiments are adapted from pulse sequences previously applied to determining (h)J coupling constants in nucleic acid-base pairs and proteins. The resulting (h2)J(N,N), (h4)J(N,N), (h3)J(C,N), and (h1)J(H,N) couplings involve the (15)N(1), (13)C(2), and (15)N(3) nuclei of the pyrimidine moiety of FMN, whereas no such interactions are detectable for (13)C(4) and (15)N(5). Several long-range (15)N-(15)N, (13)C-(15)N, and (1)H-(15)N J-coupling constants within the flavin are obtained as "by-products". The magnitudes of both (h)J and regular J couplings are found to be dependent on the redox state. In general, good correlations between (h)J coupling constants and donor-group (1)H chemical shifts and also crystallographic donor-acceptor distances are observed.

Role of the C-terminal tyrosine of ferredoxin-nicotinamide adenine dinucleotide phosphate reductase in the electron transfer processes with its protein partners ferredoxin and flavodoxin.

The catalytic mechanism proposed for ferredoxin-NADP(+) reductase (FNR) is initiated by reduction of its flavin adenine dinucleotide (FAD) cofactor by the obligatory one-electron carriers ferredoxin (Fd) or flavodoxin (Fld) in the presence of oxidized nicotinamide adenine dinucleotide phosphate (NADP(+)). The C-terminal tyrosine of FNR, which stacks onto its flavin ring, modulates the enzyme affinity for NADP(+)/H, being removed from this stacking position during turnover to allow productive docking of the nicotinamide and hydride transfer. Due to its location at the substrate-binding site, this residue might also affect electron transfer between FNR and its protein partners. We therefore studied the interactions and electron-transfer properties of FNR proteins mutated at their C-termini. The results obtained with the homologous reductases from pea and Anabaena PCC7119 indicate that interactions with Fd or Fld are hardly affected by replacement of this tyrosine by tryptophan, phenylalanine, or serine. In contrast, electron exchange is impaired in all mutants, especially in the nonconservative substitutions, without major differences between the eukaryotic and the bacterial FNR. Introduction of a serine residue shifts the flavin reduction potential to less negative values, whereas semiquinone stabilization is severely hampered, introducing further constraints to the one-electron-transfer processes. Thus, the C-terminal tyrosine of FNR plays distinct and complementary roles during the catalytic cycle, (i) by lowering the affinity for NADP(+)/H to levels compatible with steady-state turnover, (ii) by contributing to the flavin semiquinone stabilization required for electron splitting, and (iii) by modulating the rates of electron exchange with the protein partners.

Crystallographic investigation of the role of aspartate 95 in the modulation of the redox potentials of Desulfovibrio vulgaris flavodoxin.

The side chain of aspartate 95 in flavodoxin from Desulfovibrio vulgaris provides the closest negative charge to N(1) of the bound FMN in the protein. Site-directed mutagenesis was used to substitute alanine, asparagine, or glutamate for this amino acid to assess the effect of this charge on the semiquinone/hydroquinone redox potential (E(1)) of the FMN cofactor. The D95A mutation shifts the E(1) redox potential positively by 16 mV, while a negative shift of 23 mV occurs in the oxidized/semiquinone midpoint redox potential (E(2)). The crystal structures of the oxidized and semiquinone forms of this mutant are similar to the corresponding states of the wild-type protein. In contrast to the wild-type protein, a further change in structure occurs in the D95A mutant in the hydroquinone form. The side chain of Y98 flips into an energetically more favorable edge-to-face interaction with the bound FMN. Analysis of the structural changes in the D95A mutant, taking into account electrostatic interactions at the FMN binding site, suggests that the pi-pi electrostatic repulsions have only a minor contribution to the very low E(1) redox potential of the FMN cofactor when bound to apoflavodoxin. Substitution of D95 with glutamate causes only a slight perturbation of the two one-electron redox potentials of the FMN cofactor. The structure of the D95E mutant reveals a large movement of the 60-loop (residues 60-64) away from the flavin in the oxidized structure. Reduction of this mutant to the hydroquinone causes the conformation of the 60-loop to revert back to that occurring in the structures of the wild-type protein. The crystal structures of the D95E mutant imply that electrostatic repulsion between a carboxylate on the side chain at position 95 and the phenol ring of Y98 prevents rotation of the Y98 side chain to a more energetically favorable conformation as occurs in the D95A mutant. Replacement of D95 with asparagine has no effect on E(2) but causes E(1) to change by 45 mV. The D95N mutant failed to crystallize. The K(d) values of the protein FMN complex in all three oxidation-reduction states differ from those of the wild-type complexes. Molecular modeling showed that the conformational energy of the protein changes with the redox state, in qualitative agreement with the observed changes in K(d), and allowed the electrostatic interactions between the FMN and the surrounding groups on the protein to be quantified.

A comparison of the urea-induced unfolding of apoflavodoxin and flavodoxin from Desulfovibrio vulgaris.

The kinetics and thermodynamics of the urea-induced unfolding of flavodoxin and apoflavodoxin from Desulfovibrio vulgaris were investigated by measuring changes in flavin and protein fluorescence. The reaction of urea with flavodoxin is up to 5000 times slower than the reaction with the apoprotein (0.67 s(-1) in 3 m urea in 25 mm sodium phosphate at 25 degrees C), and it results in the dissociation of FMN. The rate of unfolding of apoflavodoxin depends on the urea concentration, while the reaction with the holoprotein is independent of urea. The rates decrease in high salt with the greater effect occurring with apoprotein. The fluorescence changes fit two-state models for unfolding, but they do not exclude the possibility of intermediates. Calculation suggests that 21% and 30% of the amino-acid side chains become exposed to solvent during unfolding of flavodoxin and apoflavodoxin, respectively. The equilibrium unfolding curves move to greater concentrations of urea with increase of ionic strength. This effect is larger with phosphate than with chloride, and with apoflavodoxin than with flavodoxin. In low salt the conformational stability of the holoprotein is greater than that of apoflavodoxin, but in high salt the relative stabilities are reversed. It is calculated that two ions are released during unfolding of the apoprotein. It is concluded that the urea-dependent unfolding of flavodoxin from D. vulgaris occurs because apoprotein in equilibrium with FMN and holoprotein unfolds and shifts the equilibrium so that flavodoxin dissociates. Small changes in flavin fluorescence occur at low concentrations of urea and these may reflect binding of urea to the holoprotein.