Unfolding and refolding of hen egg-white riboflavin binding protein

The unfolding and refolding of riboflavin-binding protein (RfBP) from hen egg-white induced by addition of guanidinium chloride (GdnHCl), and its subsequent removal by dialysis have been studied by c.d. and fluorescence for both the native and reduced protein. The reduction of its nine disulphide bonds causes a reduction in the secondary structure (alpha-helix plus beta-sheet) from 63% to 33% of the amino acid residues. Unfolding of the native protein occurred in two phases; the first involving a substantial loss of tertiary structure, followed by a second phase involving loss of secondary structure at higher GdnHCl concentrations. By contrast this biphasic behaviour was not discernible in the reduced protein. The loss of ability to bind riboflavin occurred after the first phase of unfolding. Comparison of unfolding of the holoprotein and apoprotein suggested that riboflavin has only a small stabilizing effect on the unfolding process. After removal of GdnHCl, the holoprotein, apoprotein and reduced protein assumed their original conformation. The significance of the results in relation to various models for protein folding is discussed.     

Evaluation of riboflavin binding protein domain interaction using differential scanning calorimetry

Riboflavin binding (or carrier) protein (RfBP) is a monomeric, two-domain protein, originally purified from hens' egg white. RfBP contains nine disulfide bridges; as a result, the protein forms a compact structure and undergoes reversible three-state thermal denaturation. This was demonstrated using a differential scanning calorimetry (DSC) method [Wasylewski M. (2000) J. Prot. Chem. 19(6), 523-528]. It has been shown that the RfBP complex with riboflavin denaturates in a three-state process which may be attributed to sequential unfolding of the RfBP domains. In case of apo RfBP, the ligand binding domain denaturates at a lower temperature than the C-terminal domain. Ligand binding greatly enhances the thermostability of the N-terminal domain, whereas the C-terminal domain thermostability is only slightly affected and, in case of the examined holo RfBPs, the denaturation peaks of both domains merge or cross over. The magnitude of the changes depends on ligand structure. A detailed study of protein concentration effects carried out in this work allowed to estimate not only the thermostability of both domains but also the strength of domain interactions. The DeltaCp, of denaturation was found for C-terminus and N-terminus of RfBP-riboflavin complex to amount to 2.5 and -1.9 kcal mol(-1), respectively. The calculated domain interaction free energy, DeltaGCN, was estimated to be approximately -1580 cal mol(-1) at 67.0 degrees C. This value indicates that the interdomain interaction is of medium strength.     

The interaction of riboflavin with a protein isolated from hen's egg white: a spectrofluorimetric study.

The interaction of riboflavin with a protein isolated from egg white has been studied spectrofluorimetrically at different pH values. In 0.1 M phosphate buffer pH 7.0; 1:1 complex formation occurs with the association constant Ka = 7.7-10(7) M-1. In the presence of 0.033% sodium dodecyl sulphate, the complex dissociated with a rate constant of 4-10(-2) sec-1 at 29 degrees C. The binding was sensitive to pH and to the antibodies produced against the protein. On lowering the pH from 7 to 4 the binding affinity decreased approximately 100-fold and below pH 4, the binding could not be detected at all. These data, together with those obtained by measuring the fluorescence intensities of riboflavin in presence of N-bromosuccinimide oxidized- and disulphide reduced apoprotein, suggest that carboxyl functions, 1-2 tryptophan residues and 2-3 disulphide bridges are essential for binding. The emission spectra of the protein under different conditions upon excitation at 280 and 295 nm were analyzed to calculate the quantum yield (Q) and the efficiency of energy transfer (e) from tyrosine to tryptophan residues. From these data it was concluded that the energy transfer did not occur with equal efficiency under all conditions and that the tryptophan residues responsible for the riboflavin binding are more accessible to N-bromosuccinimide oxidation than others.     

Isolation of riboflavin carrier proteins from pregnant human and umbilical cord serum: Similarities with chicken egg riboflavin carrier protein

A riboflavin carrier protein has been purified from human pregnancy and umbilical cord sera by affinity chromatography and fast protein liquid chromatography. This protein has a similar molecular weight to the chicken egg riboflavin carrier protein and shares other physicochemical properties, such as pI and riboflavin binding characteristics, with the avian counterpart. A high degree of immunological cross-reactivity is observed between the human and avian riboflavin binding proteins indicating the extensive conservation of this protein throughout evolution.     

Flavin binding to the high affinity riboflavin transporter RibU

The first biochemical and spectroscopic characterization of a purified membrane transporter for riboflavin (vitamin B(2)) is presented. The riboflavin transporter RibU from the bacterium Lactococcus lactis was overexpressed, solubilized, and purified. The purified transporter was bright yellow when the cells had been cultured in rich medium. We used a detergent-compatible matrix-assisted laser desorption ionization time-of-flight mass spectrometry method (Cadene, M., and Chait, B. T. (2000) Anal. Chem. 72, 5655-5658) to show that the source of the yellow color was riboflavin that had been co-purified with the transporter. The method appears generally applicable for substrate identification of purified membrane proteins. Substrate-free RibU was produced by expressing the protein in cells cultured in chemically defined medium. Riboflavin, FMN, and roseoflavin bound to RibU with high affinity and 1:1 stoichiometry (K(d) for riboflavin is 0.6 nM), but FAD did not bind to the transporter. The absorption spectrum of riboflavin changed dramatically when the substrate bound to RibU. Well resolved bands appeared at 441, 464, and 486 nm, indicating a hydrophobic binding pocket. The fluorescence of riboflavin was almost completely quenched upon binding to RibU, and also the tryptophan fluorescence of the transporter was quenched when flavins bound. The results indicate that riboflavin is stacked with one or more tryptophan residues in the binding pocket of RibU. Mutagenesis experiments showed that Trp-68 was involved directly in the riboflavin binding. The structural properties of the binding site and mechanistic consequences of the exceptionally high affinity of RibU for its substrate are discussed in relation to soluble riboflavin-binding proteins of known structure.     

Study of interaction of cytochrome P450 2B4 with riboflavin by fluorescence spectrometry

Fluorescence quenching of riboflavin by cytochrome P450 2B4 was used to probe the ligand--enzyme binding interaction ((lambda ex = 385 nm, lambda em = 520 nm). Riboflavin is a component of a flavoprotein NADPH dependent cytochrome P450 reductase, an essential electron carrier during cytochrome P450 catalysis. Fluorescence titration measurements revealed that cytochrome P450 2B4 and riboflavin formed a complex with an apparent Kd = 8.8 +/- 1 microM. The fluorescence intensity of riboflavin decreased upon the addition of cytochrome P450 2B4, which may be caused by the resonance excitation energy transfer from the fluorescent donor riboflavin to the cytochrome P450 2B4 heme acceptor. These data suggest that there may exist specific sites of binding of riboflavin with the protein globule of cytochrome P450 2B4.     

Variations in riboflavin binding by human plasma: identification of immunoglobulins as the major proteins responsible

Riboflavin binding by plasma proteins from healthy human subjects was examined by equilibrium dialysis using a physiological concentration of [2-14C]riboflavin (0.04 microM). Binding ranged from 0.080 to 0.917 pmole of riboflavin/mg of protein (with a mean +/- SD of 0.274 +/- 0.206), which corresponded to 4.14 to 49.4 pmole/ml of plasma (15.5 +/- 11.0) (N = 34). Males and females yielded similar results. Upon fractionation of plasma by gel filtration, the major riboflavin-binding components eluted with albumin and gamma-globulins. Albumin was purified and found to bind riboflavin only very weakly (Kd = 3.8 to 10.4 mM), although FMN and photochemical degradation products (e.g., lumiflavine and lumichrome) were more tightly bound. Binding in the gamma-globulin fraction was attributed to IgG and IGA because the binding protein(s) and immunoglobulins copurified using various methods were removed by treatment of plasma with protein A-agarose, and were coincident upon immunoelectrophoresis followed by autoradiography to detect [2-14C]riboflavin. Differences among the plasma samples correlated with the binding recovered with the immunoglobulins. Binding was not directly related to the total IgG or IgA levels of subjects. Hence, it appears that the binding is due to a subfraction of these proteins. These findings suggest that riboflavin-binding immunoglobulins are a major cause of variations in riboflavin binding in human circulation, and may therefore affect the utilization of this micronutrient.     

Competitive binding assays for riboflavin and riboflavin-binding protein

A competitive binding procedure that can be used to determine either riboflavin or riboflavin-binding protein has been developed. Riboflavin-binding protein from chicken egg white binds tightly to DEAE-cellulose while free riboflavin does not. Stock [2-¹⁴C]riboflavin solutions, diluted with varying amounts of a standard unlabeled riboflavin solution or an unknown sample, are mixed with aporiboflavin-binding protein and washed through small DEAE-cellulose columns. The protein-bound riboflavin is batch eluted into scintillation vials, counted, and the unknown samples compared to a standard curve. This is a simple, rapid method for assaying riboflavin by isotope dilution. By a slight modification of the incubation conditions of this procedure, the degree of saturation and amount of riboflavin-binding protein can be determined. Data from both assays can be represented by linear plots in which slopes or intercepts correspond to unknown values. The principles presented here have been extended to the assay of biotin and avidin and should apply to other vitamins and vitamin-binding proteins.     

Effect of antibody on interaction between vitamin B2 and riboflavin apoprotein.

1. Dissociation of riboflavin from flavoprotein and from the flavoprotein-antibody complex occurs under the same conditions. 2. The precipitated apoprotein-antibody complex retains 15% of the apoprotein capacity to bind riboflavin. After solubilization of the complex in 0.3 M-KCl or 1 M-urea, the binding of riboflavin amounts to 80 - 90% of its capacity. 3. The apoprotein modified by oxidation of 50% of tryptophan residues loses the ability to bind riboflavin but its immunological reactivity with the anti-flavoprotein antibody is similar to that of native apoprotein. The apoprotein with all tryptophan residues oxidized shows much lower immunoreactivity. 4. The obtained results suggest that in riboflavin flavoprotein the region around the riboflavin-binding site does not show the properties of an antigenic determinant.     

Multi-spectroscopic and molecular modelling approach to investigate the interaction of riboflavin with human serum albumin

Riboflavin (RF) plays an important role in various metabolic redox reactions in the form of flavin adenine dinucleotide and flavin mononucleotide. Human serum albumin (HSA) is an important protein involved in the transportation of drugs, hormones, fatty acid and other molecules which determine the biodistribution and physiological fate of these molecules. In this study, we have investigated the interaction of riboflavin RF with HSA under simulative physiological conditions using various biophysical, calorimetric and molecular docking techniques. Results demonstrate the formation of riboflavin–HSA complex with binding constant in the order of 10⁴ M^?1. Fluorescence spectroscopy confirms intermediate strength having a static mode of quenching with stoichiometry of 1:1. Experimental results suggest that the binding site of riboflavin mainly resides in sub-domain IIA of HSA and that ligand interaction increases the α-helical content of HSA. These parameters were further verified by isothermal titration calorimetry ITC which confirms the thermodynamic parameters obtained by fluorescence spectroscopy. Molecular docking was employed to suggest a binding model. Based on thermodynamic, spectroscopic and computational observations it can be concluded that HSA-riboflavin complex is mainly stabilized by various non-covalent forces with binding energy of ?7.2 kcal mol^?1.     

Evolution of vitamin B2 biosynthesis: riboflavin synthase of Arabidopsis thaliana and its inhibition by riboflavin

A synthetic gene specifying the catalytic domain of the Arabidopsis thaliana riboflavin synthase was expressed with high efficiency in a recombinant Escherichia coli strain. The recombinant pseudomature protein was shown to convert 6,7-dimethyl-8-ribityllumazine into riboflavin at a rate of 0.027 s-1 at 25 degrees C. The protein sediments at a rate of 3.9 S. Sedimentation equilibrium analysis afforded a molecular mass of 67.5 kDa, indicating a homotrimeric structure, analogous to the riboflavin synthases of Eubacteria and fungi. The protein binds its product riboflavin with relatively high affinity (Kd =1.1 microM). Product inhibition results in a characteristic sigmoidal velocity versus substrate concentration relationship. Characterization of the enzyme/product complex by circular dichroism and UV absorbance spectroscopy revealed a shift of the absorption maxima of riboflavin from 370 and 445 to 399 and 465 nm, respectively. Complete or partial sequences for riboflavin synthase orthologs were analyzed from 11 plant species. In each case for which the complete plant gene sequence was available, the catalytic domain was preceded by a sequence of 1-72 amino acid residues believed to function as plastid targeting signals. Comparison of all available riboflavin synthase sequences indicates that hypothetical gene duplication conducive to the two-domain architecture occurred very early in evolution.     

A study of flavin-protein and flavoprotein-ligand interactions. Binding aspects and spectral properties of D-amino acid oxidase and riboflavin binding protein.

To study flavin-protein and flavoprotein-ligand interaction, the absorption, CD and MCD spectra of riboflavin, FAD, roseoflavin, the complexes of riboflavin and roseoflavin with riboflavin binding protein(RBP),D-amino acid oxidase(D-AO) and its complexes with ligands were observed in the spectral region of 310-600 nm and the binding properties of D-AO with di-substituted benzoate derivatives and of RBP with roseoflavin were also measured. The dimer of D-amino acid oxidase has a higher affinity for di-substituted benzoate derivatives than the monomer. The change in the absorption of FAD in D-AO caused by the binding of the first ligand to the dimer, which can bind two ligands, was similar to that caused by the binding of the second ligand. Roseoflavin could bind to RBP in a 1 : 1 ratio and the dissociation constant was 3.8 x 10(-8)M. The protein fluorescence of RBP was quenched by about 86% due to complex formation with roseoflavin. The MCD spectra showed similar patterns for all molecular complexes of riboflavin and FAD, with two negative extrema of ellipticity which probably correspond to the Faraday B-term, but the Faraday A-term could not be observed, suggesting that there was no degeneracy in the excited state of flavins. It is also suggested, based on a comparison of the absorption, CD and MCD spectra, that the vibronic structure of flavin was modified differently by each flavin-protein or flavoprotein-ligand interaction. Comparison of the absorption, CD and MCD spectra(310-600 nm) for roseoflavin and the roseoflavin-RBP complex revealed that there were five spectral components around 320, 340, 400, 500, and 550 nm in roseoflavin.     

Biosynthesis of riboflavin: an unusual riboflavin synthase of Methanobacterium thermoautotrophicum.

Riboflavin synthase was purified by a factor of about 1,500 from cell extract of Methanobacterium thermoautotrophicum. The enzyme had a specific activity of about 2,700 nmol mg(-1) h(-1) at 65 degrees C, which is relatively low compared to those of riboflavin synthases of eubacteria and yeast. Amino acid sequences obtained after proteolytic cleavage had no similarity with known riboflavin synthases. The gene coding for riboflavin synthase (designated ribC) was subsequently cloned by marker rescue with a ribC mutant of Escherichia coli. The ribC gene of M. thermoautotrophicum specifies a protein of 153 amino acid residues. The predicted amino acid sequence agrees with the information gleaned from Edman degradation of the isolated protein and shows 67% identity with the sequence predicted for the unannotated reading frame MJ1184 of Methanococcus jannaschii. The ribC gene is adjacent to a cluster of four genes with similarity to the genes cbiMNQO of Salmonella typhimurium, which form part of the cob operon (this operon contains most of the genes involved in the biosynthesis of vitamin B12). The amino acid sequence predicted by the ribC gene of M. thermoautotrophicum shows no similarity whatsoever to the sequences of riboflavin synthases of eubacteria and yeast. Most notably, the M. thermoautotrophicum protein does not show the internal sequence homology characteristic of eubacterial and yeast riboflavin synthases. The protein of M. thermoautotrophicum can be expressed efficiently in a recombinant E. coli strain. The specific activity of the purified, recombinant protein is 1,900 nmol mg(-1) h(-1) at 65 degrees C. In contrast to riboflavin synthases from eubacteria and fungi, the methanobacterial enzyme has an absolute requirement for magnesium ions. The 5' phosphate of 6,7-dimethyl-8-ribityllumazine does not act as a substrate. The findings suggest that riboflavin synthase has evolved independently in eubacteria and methanobacteria.     

Domain structure of riboflavin synthase.

Riboflavin synthase of Escherichia coli is a homotrimer of 23.4 kDa subunits catalyzing the formation of the carbocyclic ring of the vitamin, riboflavin, by dismutation of 6,7-dimethyl-8-ribityllumazine. Intramolecular sequence similarity suggested that each subunit folds into two topologically similar domains. In order to test this hypothesis, sequence segments comprising amino-acid residues 1-97 or 101-213 were expressed in recombinant E. coli strains. The recombinant N-terminal domain forms a homodimer that can bind riboflavin, 6,7-dimethyl-8-ribityllumazine and trifluoromethyl-substituted 8-ribityllumazine derivatives as shown by absorbance, circular dichroism, and NMR spectroscopy. Most notably, the recombinant domain dimer displays the same diastereoselectivity for ligands as the full length protein. The minimum N-terminal peptide segment required for ligand binding comprises amino-acid residues 1-87. The recombinant C-terminal domain comprising amino-acid residues 101-213 is relatively unstable and was shown not to bind riboflavin but to differentiate between certain diastereomeric trifluoromethyl-8-ribityllumazine derivatives. The data show that a single domain comprises the intact binding site for one substrate molecule. The enzyme-catalyzed dismutation requires two substrate molecules to be bound in close proximity, and each active site of the enzyme appears to be located at the interface of an N-terminal and C-terminal domain.     

Biosynthesis of riboflavin: 6,7-dimethyl-8-ribityllumazine synthase of Schizosaccharomyces pombe.

A cDNA sequence from Schizosaccharomyces pombe with similarity to 6,7-dimethyl-8-ribityllumazine synthase was expressed in a recombinant Escherichia coli strain. The recombinant protein is a homopentamer of 17-kDa subunits with an apparent molecular mass of 87 kDa as determined by sedimentation equilibrium centrifugation (it sediments at an apparent velocity of 5.0 S at 20 degrees C). The protein has been crystallized in space group C2221. The crystals diffract to a resolution of 2.4 A. The enzyme catalyses the formation of 6,7-dimethyl-8-ribityllumazine from 5-amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione and 3,4-dihydroxy- 2-butanone 4-phosphate. Steady-state kinetic analysis afforded a vmax value of 13 000 nmol.mg-1.h-1 and Km values of 5 and 67 microm for 5-amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione and 3,4-dihydroxy-2-butanone 4-phosphate, respectively. The enzyme binds riboflavin with a Kd of 1.2 microm. The fluorescence quantum yield of enzyme-bound riboflavin is < 2% as compared with that of free riboflavin. The protein/riboflavin complex displays an optical transition centered around 530 nm as shown by absorbance and CD spectrometry which may indicate a charge transfer complex. Replacement of tryptophan 27 by tyrosine or phenylalanine had only minor effects on the kinetic properties, but complexes of the mutant proteins did not show the anomalous long wavelength absorbance of the wild-type protein. The replacement of tryptophan 27 by aliphatic amino acids substantially reduced the affinity of the enzyme for riboflavin and for the substrate, 5-amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione.     

Intramolecular electron and proton transfer in proteins: CO2- reduction of riboflavin binding protein and ribonuclease A

The formate radical (CO2-) reacts with ribonuclease A to form the cystine disulfide radical as one of the products. CO2- reacts with the riboflavin binding protein of chicken egg white with the ultimate product being the neutral flavin semiquinone. Formation of the disulfide radical in ribonuclease is slower than the reaction between protein and CO2-; formation of the flavin semiquinone in the riboflavin binding protein is slower than the protein-CO2- reaction. We conclude for both proteins that CO2- must reduce an as yet unidentified group or groups, which in turn reduce(s) the disulfide of RNase or the flavin of riboflavin binding protein. This conclusion is supported in the case of ribonuclease by the observation of a transient, broad absorption band centered between 350 and 370 nm. The CO2--initiated reductions of the disulfide in ribonuclease and the flavin in the riboflavin binding protein are mixed first- and second-order processes. We propose that the transfer of an electron from the unknown intermediate(s) to the final product involves both inter- and intramolecular paths between groups that may not be in van der Waals contact. With the hydrated electron, in contrast to CO2-, as reductant of the riboflavin binding protein, the anionic semiquinone is observed as an intermediate. The anionic semiquinone is then rapidly protonated, yielding the stable neutral semiquinone. From the reaction kinetics and protein concentration dependence, we conclude that a group or groups on the protein donate(s) a proton to the anionic semiquinone by both inter- and intramolecular paths.     

The structure of the N-terminal domain of riboflavin synthase in complex with riboflavin at 2.6A resolution

Riboflavin synthase of Escherichia coli is a homotrimer with a molecular mass of 70 kDa. The enzyme catalyzes the dismutation of 6,7-dimethyl-8-(1'-D-ribityl)-lumazine, affording riboflavin and 5-amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione. The N-terminal segment (residues 1-87) and the C-terminal segment (residues 98-187) form beta-barrels with similar fold and a high degree of sequence similarity. A recombinant peptide comprising amino acid residues 1-97 forms a dimer, which binds riboflavin with high affinity. Here, we report the structure of this construct in complex with riboflavin at 2.6A resolution. It is demonstrated that the complex can serve as a model for ligand-binding in the native enzyme. The structure and riboflavin-binding mode is in excellent agreement with structural information obtained from the native enzyme from Escherichia coli and riboflavin synthase from Schizosaccharomyces pombe. The implications for the binding specificity and the regiospecificity of the catalyzed reaction are discussed.     

Uptake and binding of riboflavin by membrane vesicles of bacillus subtilis

Riboflavin uptake and membrane-associated riboflavin-binding activity have been investigated in Bacillus subtilis. The uptake and binding activity of the vitamin were found to be repressed coordinately by riboflavin present in the growth medium. The uptake of riboflavin has been shown to have properties of a carrier-mediated process, and membrane vesicles have been shown to demonstrate riboflavin counterflow and exchange. The membrane-associated binding activity for riboflavin has been solubilized with detergents, and a procedure for the partial purification of this component is described. The partially purified riboflavin-binding component has properties expected for a carrier involved in riboflavin uptake, as it shows saturation kinetics and is inhibited by riboflavin analogues. Evidence is also presented showing that reduced riboflavin binds to a greater extent than oxidized riboflavin, and the possible role of the reduced riboflavin in riboflavin uptake is discussed.     

Fluorescent assay for riboflavin binding to cytochrome P450 2B4

The interactions between the hemoprotein cytochrome P450 2B4 (CYP 2B4) and riboflavin - a low molecular weight component of the flavoprotein NADPH-dependent cytochrome P450 reductase - were investigated by fluorescence spectroscopy. Riboflavin fluorescence quenching by cytochrome P450 2B4 was used to probe the ligand-enzyme binding (lambda(ex)=385 nm, lambda(em)=520 nm). Fluorescence titration experiments showed formation of a complex between cytochrome P450 2B4 and riboflavin with an apparent dissociation constant value, K(d)=8.8+/-1 microM. The fluorescence intensity of riboflavin was decreased with increasing the cytochrome P450 2B4 concentration, indicating the transfer of resonance excitation energy from riboflavin (energy donor) to the cytochrome P450 2B4 heme (energy acceptor). The data obtained are suggestive of the existence of riboflavin binding site(s) on the hemeprotein molecule.