Different types of glutathionylation of hemoglobin can exist in intact erythrocytes

Glutathionylation of hemoglobin (Hb) was studied by incubation of intact human erythrocytes with 1 mM tert-butylhydroperoxide (tBHP). Electrophoresis of the membranes showed a time dependent increase of membrane-bound Hb alpha chain until 10 min, and immunoblotting study showed that membrane-bound Hb alpha chain reacted with anti-glutathione antibody only after 10 min. Concomitant with the Hb alpha chain, membrane associated actin, spectrin, and glyceraldehyde 3-phosphate dehydrogenase reacted with the antibody. Cytosolic Hb of the control erythrocytes reacted with anti-glutathione antibody. Together with our previous paper, the present study indicates that at least three different types of glutathionylation of Hb can exist in erythrocytes. The first type is a mixed disulfide bond between reduced glutathione (GSH) and normal Hb. The second type is a disulfide bond between the cysteine 93 of metHb beta chain and oxidized glutathione (GSSG), and the third type is a disulfide bond between the other cysteine residues of metHb alpha chain and/or metHb beta chain and GSSG.     

Prolonged oxygen-carrying ability of hemoglobin vesicles by coencapsulation of catalase in vivo

Hemoglobin (Hb) vesicles (particle diameter, ca. 250 nm) have been developed as Hb-based oxygen carriers in which a purified Hb solution is encapsulated with a phospholipid bilayer membrane. The oxidation of Hb to nonfunctional ferric Hb (metHb) was caused by reactive oxygen species, especially hydrogen peroxide (H(2)O(2)), in vivo in addition to autoxidation. We focused on the enzymatic elimination of H(2)O(2) to suppress the metHb formation in the Hb vesicles. In this study, we coencapsulated catalase with Hb within vesicles and studied the rate of metHb formation in vivo. The Hb vesicles containing 5.6 x 10(4) unit mL(-1) catalase decreased the rate of metHb formation by half in comparison with Hb vesicles without catalase. We succeeded in prolonging the oxygen-carrying ability of the Hb vesicle in vivo by the coencapsulation of catalase.     

Hemoglobin vesicles containing methemoglobin and L-tyrosine to suppress methemoglobin formation in vitro and in vivo

Hemoglobin (Hb) vesicles have been developed as cellular-type Hb-based O(2) carriers in which a purified and concentrated Hb solution is encapsulated with a phospholipid bilayer membrane. Ferrous Hb molecules within an Hb vesicle were converted to ferric metHb by reacting with reactive oxygen species such as hydrogen peroxide (H(2)O(2)) generated in the living body or during the autoxidation of oxyHb in the Hb vesicle, and this leads to the loss of O(2) binding ability. The prevention of metHb formation by H(2)O(2) in the Hb vesicle is required to prolong the in vivo O(2) carrying ability. We found that a mixed solution of metHb and L-tyrosine (L-Tyr) showed an effective H(2)O(2) elimination ability by utilizing the reverse peroxidase activity of metHb with L-Tyr as an electron donor. The time taken for the conversion of half of oxyHb to metHb (T(50)) was 420 min for the Hb vesicles containing 4 g/dL (620 microM) metHb and 8.5 mM L-Tyr ((metHb/L-Tyr) Hb vesicles), whereas the time of conversion for the conventional Hb vesicles was 25 min by stepwise injection of H(2)O(2) (310 microM) in 10 min intervals. Furthermore, in the (metHb/L-Tyr) Hb vesicles, the metHb percentage did not reach 50% even after 48 h under a pO(2) of 40 Torr at 37 degrees C, whereas T(50) of the conventional Hb vesicles was 13 h under the same conditions. Moreover, the T(50) values of the conventional Hb vesicles and the (metHb/L-Tyr) Hb vesicles were 14 and 44 h, respectively, after injection into rats (20 mL/kg), confirming the remarkable inhibitory effect of metHb formation in vivo in the (metHb/L-Tyr) Hb vesicles.     

Covalent binding of glutathione to hemoglobin. I. Inhibition of hemoglobin S polymerization

Thiol reagents react with cysteine beta 93 of hemoglobin and as a result increase the oxygen affinity of hemoglobin. In the present studies we have used a thiol-disulfide exchange between mixed disulfides of hemoglobin and reduced glutathione to attach intracellular glutathione to hemoglobin and to study its antisickling properties. The rates of production of glutathionyl hemoglobin (G-Hb) depend on the structure of the thiol reagent linked to cysteine beta 93. Up to 25% G-Hb can be produced in normal and sickle red cells because of the high intracellular concentration of reduced glutathione. This high level of G-Hb in normal cells increases the oxygen affinity by about 35% and reduces heme-heme interactions. In sickle cells the increased oxygen affinity is associated with an inhibition of sickling of about 70% at 21 mm Hg. Inhibition of polymerization of deoxy HbS is also due to a direct inhibition of intermolecular contacts in the fibers as demonstrated by the increased solubility and the increased delay time of G-HbS compared to deoxy HbS.     

Complete deoxygenation from a hemoglobin solution by an electrochemical method and heat treatment for virus inactivation

Hemoglobin (Hb) has been widely studied as a raw material for various types of oxygen carriers. In the purification of Hb from red blood cells including virus inactivation and denaturation of other proteins and the long-term storage of Hb vesicles (HbV), a deoxygenation process is one of the important processes because of the high stability of deoxygenated Hb to heating and metHb formation. Though an oxygenated Hb solution can be deoxygenated with an artificial lung, it is difficult to reduce the oxygen partial pressure of the Hb solution to less than 10 Torr. We developed an electrochemical system for complete deoxygenation of the Hb solution at the cathode compartment using hydrogen containing nitrogen gas at the anode compartment. Oxygen in the Hb solution was reduced to OH(-) at the cathode compartment within several minutes at a potential value of -1.67 V and was finally converted to water by neutralization with H(+) from the anode in the whole system. The resulting completely deoxygenated Hb could tolerate heat treatment at 62 degrees C for 10 h with no denaturation of deoxygenated Hb. The metHb formation rate of reoxygenated Hb at 37 degrees C was not changed after heat treatment. Furthermore, vesicular stomatitis virus (VSV) could be inactivated at an inactivation degree of more than 5.96 log by heat treatment.     

Radiation-induced changes of structural and functional properties of human hemoglobin

Gel filtration and SDS-PAGE separation of hemoglobin (Hb) irradiated under argon or N2O show formation of covalent-aggregated Hb molecules. The production of covalent bonds is attributed mainly to the action of hydroxyl radicals, because addition of ethanol, a scavenger of these radicals, suppresses this reaction to a great extent. The oxidized heme iron forming metHb or hemichromes is found in all the separated fractions of irradiated Hb. It is also found that the radiation-modified Hb molecules exhibit a decrease of co-operative binding of oxygen.     

The interaction of hemoglobin with hexadecyltrimethylammonium bromide

The interaction of hemoglobin (Hb) with hexadecyltrimethylammonium bromide (CTAB) is investigated by UV–vis absorption spectra and fluorescence spectra method. CTAB monomer can convert methemoglobin (metHb) to hemichrome, and CTAB molecular assemblies, such as micelle, microemulsion and lamellar liquid crystal, can induce heme monomer to leave the hydrophobic cavity of Hb. TEM results show that Hb maintains the spherical structure in CTAB microemulsions while it is unfolded in CTAB lamellar liquid crystals. The existence of proton in the above systems can increase the stability of metHb.     

The reaction of deoxy-sickle cell hemoglobin with hydroxyurea.

In addition to its capacity to increase fetal hemoglobin levels, other mechanisms are implicated in hydroxyurea's ability to provide beneficial effects to patients with sickle cell disease. We hypothesize that the reaction of hemoglobin with hydroxyurea may play a role. It is shown that hydroxyurea reacts with deoxy-sickle cell hemoglobin (Hb) to form methemoglobin (metHb) and nitrosyl hemoglobin (HbNO). The products of the reaction as well as the kinetics are followed by absorption spectroscopy and electron paramagnetic resonance (EPR) spectroscopy. Analysis of the kinetics shows that the reaction can be approximated by a pseudo-first order rate constant of 3.7x10(-4) (1/(s.M)) for the disappearance of deoxy-sickle cell hemoglobin. Further analysis shows that HbNO is formed at an observed average rate of 5.25x10(-5) (1/s), three to four times slower than the rate of formation of metHb. EPR spectroscopy is used to show that the formation of HbNO involves the specific transfer of NO from the NHOH group of hydroxyurea. The potential importance of this reaction is discussed in the context of metHb and HbNO being able to increase the delay time for sickle cell hemoglobin polymerization and HbNO's vasodilating capabilities through conversion to S-nitrosohemoglobin.     

Inactivation of alcohol dehydrogenase (ADH) by ferryl derivatives of human hemoglobin

In this paper, inactivation of alcohol dehydrogenase (ADH) by products of reactions of H2O2 with metHb has been studied. Inactivation of the enzyme was studied in two systems corresponding to two kinetic stages of the reaction. In the first system H2O2 was added to the mixture of metHb and ADH [the (metHb+ADH)+H2O2] system (ADH was present in the system since the moment of addition of H2O2 i. e. since the very beginning of the reaction of metHb with H2O2). In the second system ADH was added to the system 5 min after the initiation of the reaction of H2O2 with metHb [the (metHb+H2O2)5 min+ADH] system. In the first case all the products of reaction of H2O2 with metHb (non-peroxyl and peroxyl radicals and non-radical products, viz. hydroperoxides and *HbFe(IV)=O) could react with the enzyme causing its inactivation. In the second system, enzyme reacted almost exclusively with non-radical products (though a small contribution of reactions with peroxyl radicals cannot be excluded). ADH inactivation was observed in both system. Hydrogen peroxide alone did not inactivate ADH at the concentrations employed evidencing that enzyme inactivation was due exclusively to products of reaction of H2O2 with metHb. The rate and extent of ADH inactivation were much higher in the first than in the second system. The dependence of ADH activity on the time of incubation with ferryl derivatives of Hb can be described by a sum of three exponentials in the first system and two exponentials in the second system. Reactions of appropriate forms of the ferryl derivatives of hemoglobin have been tentatively ascribed to these exponentials. The extent of the enzyme inactivation in the second system was dependent on the proton concentration, being at the highest at pH 7.4 and negligible at pH 6.0. The reaction of H2O2 with metHb resulted in the formation of cross-links of Hb subunits (dimers and trimers). The amount of the dimers formed was much lower in the first system i. e. when the radical forms dominated the reaction of inactivation.     

Potential of peroxynitrite to promote the conversion of oxyhemoglobin to methemoglobin

Superoxide anion and NO can react to form the highly oxidizing species peroxynitrite (ONOO-)which can react directly with hemoglobin (Hb) even in the presence of physiological concentration CO:. Thisresearch was to determine the ONOO--mediated oxidation damage to the heme of oxyhemoglobin (oxyHb)under conditions expected in blood. Results showed that 8-10 mol ONOO- was needed to quickly andcompletely convert 1 mol oxyHb to methemoglobin (metHb). ONOO- (20-140 μM) caused raoid andextensive formation of metHb from oxyHb (50 μM) mainly occurring within first 5-20 min of incubation.The conversion efficiency reached 16%, 48%, 60%, 79% and 88% output of metHb after 90 min ofincubation at 0, 20, 40, 100, and 140 μM ONOO- respectively. 1 mM CO2 caused a small decrease in theability of ONOO- to oxidize oxyHb, and ONOO--promoted conversion of oxyHb to metHb increased whenpH decreased from 8.0 to 6.0. Relatively lower temperature in blood condition will inhibit this reaction insome degree. We postulate that ONOO- can mediate oxidation damage to the heme, and cause heme lossfrom the hydrophobic cavity of Hb when its concentration exceeded 90 μM. These results indicated thatONOO- could convert oxyHb to metHb under the conditions expected in blood, and this reaction wasregulated by CO2 concentration, reaction time, temperature and pH value.     

The quaternary structure of tetrameric hemoglobin regulates the oxygen affinity of polymerized hemoglobin

This study focuses on the effect of the initial quaternary structure of bovine hemoglobin (Hb) on the physical properties of glutaraldehyde polymerized Hb (PolyHb) solutions. Tense (T) state PolyHb was synthesized by maintaining the pO₂ of Hb before and after polymerization at 0 mm Hg. In contrast, relaxed (R) state PolyHb was generated by maintaining the pO₂ of Hb before and after polymerization to >749 mm Hg. PolyHb solutions were characterized by measuring the pO₂, methemoglobin (metHb) level, molecular weight distribution, O₂ affinity and cooperativity coefficient. The metHb level of all PolyHb solutions was low (<2%). Analysis of the molecular weight distribution of PolyHb solutions indicates that in general, the molecular weight of PolyHb solutions increased with increasing cross‐link density. T‐state PolyHb solutions exhibited lower O₂ affinity compared to unmodified Hb, whereas R‐state PolyHb solutions exhibited higher O₂ affinity compared to unmodified Hb. In addition, the polymerization reaction resulted in a significant decrease in cooperativity that was more pronounced at higher cross‐link densities. All of these results were explained in terms of the quaternary structure of Hb. Taken together, our results yield more insight into the importance Hb's quaternary structure plays in defining the physical properties of glutaraldehyde PolyHb solutions. This information will be useful in designing optimized glutaraldehyde PolyHb oxygen carriers for various applications in transfusion medicine. © 2009 American Institute of Chemical Engineers Biotechnol. Prog. 2009     

Differences in coordination states of substituted tyrosine residues and quaternary structures among hemoglobin M probed by resonance Raman spectroscopy

Among the four types of hemoglobin (Hb) M with a substitution of a tyrosine (Tyr) for either the proximal (F8) or distal (E7) histidine in the α or β subunits, only Hb M Saskatoon (βE7Tyr) assumes a hexacoordinate structure and its abnormal subunits can be reduced readily by methemoglobin (metHb) reductase. This is distinct from the other three M Hbs. To gain new insight into the cause of the difference, we examined the ionization states of E7 and F8 Tyrs by UV resonance Raman (RR) spectroscopy and Fe–O(Tyr) bonding by visible RR spectroscopy. Hb M Iwate (αF8Tyr), Hb M Boston (αE7Tyr), and Hb M Hyde Park (βF8Tyr) exhibited two extra UV RR bands at 1,603 cm⁻¹ (Y8a′) and 1,167 cm⁻¹ (Y9a′) arising from deprotonated (ionized) Tyr, but Hb M Saskatoon displayed the UV RR bands of protonated (unionized) Tyr at 1,620 and 1,175 cm⁻¹ in addition to those of deprotonated Tyr. Evidence for the bonding of both ionization states of Tyr to the heme in Hb M Saskatoon was provided by visible RR spectroscopy. These results indicate that βE7Tyr of Hb M Saskatoon is in equilibrium between protonated and deprotonated forms, which is responsible for facile reducibility. Comparison of the UV RR spectral features of metHb M with that of metHb A has revealed that metHb M Saskatoon and metHb M Hyde Park are in the R (relaxed) structure, similar to that of metHb A, whereas metHb M Iwate, metHb M Boston and metHb M Milwaukee are in the T (tense) quaternary structure.     

Covalent binding of glutathione to hemoglobin. II. Functional consequences and structural changes reflected in NMR spectra

Binding of glutathione by disulfide linkage to Cys-beta 93 of hemoglobin tetramers within sickle cells increases the oxygen affinity and significantly inhibits sickling at low partial oxygen pressure (Garel, M-C., Domenget, C., Caburi-Martin, J., Prehu, C., Galacteros, F., and Beuzard, Y. (1986) J. Biol. Chem. 261, 14704-14709). This article reports a characterization of the oxygen-binding properties of glutathionyl hemoglobin (G-Hb) in solution in the presence or absence of allosteric effectors. The studies reveal a nearly 6-fold increase in oxygen affinity compared to native HbA and a Hill coefficient at half-saturation (n50) of 1.50 compared to n50 of approximately 2.9 for HbA. The oxygen Bohr effect measured in the alkaline pH range is reduced by 38%. Addition of 2,3-diphosphoglycerate decreases the oxygen affinity of G-Hb and HbA to a similar extent and increases the Bohr effect, indicating that the binding sites for organic phosphates are not perturbed in G-Hb. The rate of autooxidation of G-HbO2 is slower than of HbAO2. Oxidation by ferricyanide of G-HbCO is also reduced and is biphasic, demonstrating a heterogeneous susceptibility of the hemes in G-Hb. Flash photolysis experiments indicate that the tetramer-dimer dissociation constant is 1 order of magnitude greater for G-HbCO than for HbACO. High resolution NMR spectra at 400 MHz show that in G-Hb: the tertiary structure of the beta heme pocket is significantly perturbed, particularly in the F helix and the EF corner; the formation of the salt bridge between His-beta 146 and Asp-beta 94, a feature of the deoxy state, is precluded; and a deoxy interchain (alpha 1 beta 2) contact between Asp beta 2 99 and Tyr alpha 1 42 is appreciably destabilized. The NMR data provide a structural basis for interpreting the high oxygen affinity, reduced cooperativity, and diminished polymerization of G-HbS.     

Mechanism of peroxynitrite interaction with ferric hemoglobin and identification of nitrated tyrosine residues. CO(2) inhibits heme-catalyzed scavenging and isomerization.

Hemoproteins are one of the major targets of peroxynitrite in vivo. It has been proposed that the bimolecular heme/peroxynitrite interaction results in both peroxynitrite inactivation (scavenging) and catalysis of tyrosine nitration. In this study, we used spectroscopic techniques to analyze the reaction of peroxynitrite with human methemoglobin (metHb). Although conventional differential spectroscopy did not reveal heme changes, our results suggest that, in the absence of bicarbonate, the heme in metHb reacts bimolecularly with peroxynitrite but is quickly back-reduced by the reaction products. This hypothesis is based on two indirect observations. First, metHb prevents the peroxynitrite-mediated nitration of a target dipeptide, Ala-Tyr, and second, it promotes the isomerization of peroxynitrite to nitrate. Both the scavenging and the isomerization activities of metHb were heme-dependent and inhibited by CO(2). Ferrous cytochrome c was an efficient scavenger of peroxynitrite, but in the ferric form did not show either scavenging or isomerization activities. We found no evidence of an increase in Ala-Tyr nitration with these hemoproteins. Peroxynitrite-treated metHb induced the formation of a long-lived radical assigned to tyrosine by spin-trapping studies. This radical, however, did not allow us to predict an interaction of peroxynitrite with heme. Hb was nitrated by peroxynitrite/CO(2) mainly in tyrosines beta 130, alpha 42, and alpha 140 and, to a lesser extent, alpha 24. The nitration of alpha chain tyrosines more exposed to the solvent (alpha 140 and alpha 24) was higher in CO-Hb and metHb, while nitration of alpha 42, the tyrosine nearest to the heme, was higher in oxyHb. We deduce that the heme/peroxynitrite interaction, which is inhibited in CO-Hb and metHb, affects alpha tyrosine nitration in two opposite ways, i.e., by protecting exposed residues and by promoting nitration of the residue nearest to the heme. Conversely, nitration of beta Tyr 130 was comparable in oxyHb, metHb, and CO-Hb, suggesting a mechanism involving only nitrating species formed during peroxynitrite decay.     

Haptoglobin phenotypes differ in their ability to inhibit heme transfer from hemoglobin to LDL

LDL oxidation plays a pivotal role in atherosclerosis. Excellular hemoglobin (Hb) is a trigger of LDL oxidation. By virtue of its ability to bind hemoglobin, haptoglobin (Hp) serves as an antioxidant. Oxidation of LDL by hemoglobin was analyzed to occur by heme displacement from methemoglobin lodged in LDL. The LDL-associated heme is disintegrated, and iron inserted this way in LDL triggers formation of lipid peroxides. The genetic polymorphism of haptoglobin was found to be a risk factor in the pathogenesis of atherosclerosis. Individuals with Hp2-2 have more vascular incidences as compared to those with Hp1-1. In the current study, oxidation of LDL by metHb was carried out at physiological pH without addition of external peroxides. Hb-derived oxidation of lipids and protein was found to be practically inhibited by Hp1-1 but only partially by Hp2-2. Heme transfer from metHb to LDL was almost completely omitted by Hp1-1 and only partially by Hp2-2. We concluded that partial heme transfer from the Hb-Hp2-2 complex to LDL is the reason for oxidation of LDL lipids as well as protein. These findings provide a molecular basis for Hp2-2 atherogenic properties.     

Oxidation of 3-hydroxykynurenine catalyzed by methemoglobin with hydrogen peroxide.

Methemoglobin (metHb) with H2O2 catalyzed the oxidation of 3-hydroxykynurenine (3-HKY) in the reaction mixture of metHb, 3-HKY, and H2O2. The spectrophotometric experiments suggest the following mechanism for the 3-HKY oxidation by metHb with H2O2. MetHb first reacts with H2O2 to form the ferryl complex of Hb. This species then oxidizes 3-HKY, while it returns to metHb. 3-HKY was more reactive with the ferryl complex than glutathione but less reactive than ascorbic acid. Scavengers of the hydroxyl radical, dimethyl sulfoxide and ethanol, scarcely inhibited the 3-HKY oxidation by metHb with H2O2. Desferrioxamine, a metal chelator, hardly suppressed the 3-HKY oxidation. These results indicate that the hydroxyl radical is not involved in the 3-HKY oxidation by metHb with H2O2.     

Membrane effects of nitrite-induced oxidation of human red blood cells.

The aim of our investigation was to study the red blood cell (RBC) membrane effects of NaNO(2)-induced oxidative stress. Hyperpolarization of erythrocyte membranes and an increase in membrane rigidity have been shown as a result of RBC oxidation by sodium nitrite. These membrane changes preceded reduced glutathione depletion and were observed simultaneously with methemoglobin (metHb) formation. Changes of the glutathione pool (total and reduced glutathione, and mixed protein-glutathione disulfides) during nitrite-induced erythrocyte oxidation have been demonstrated. The rates of intracellular oxyhemoglobin and GSH oxidation highly increased as pH decreased in the range of 7.5-6.5. The activation energy of intracellular metHb formation obtained from the temperature dependence of the rate of HbO(2) oxidation in RBC was equal to 16.7+/-1.6 kJ/mol in comparison with 12.8+/-1.5 kJ/mol calculated for metHb formation in hemolysates. It was found that anion exchange protein (band 3 protein) of the erythrocyte membrane does not participate significantly in the transport of nitrite ions into the erythrocytes as band 3 inhibitors (DIDS, SITS) did not decrease the intracellular HbO(2) oxidation by extracellular nitrite.     

Immunochemical detection of hemoglobin-derived radicals formed by reaction with hydrogen peroxide: involvement of a protein-tyrosyl radical

To investigate the involvement of a hemoglobin radical in the human oxyhemoglobin (oxyHb) or metHb/H2O2 system, we have used a new approach called "immuno-spin trapping," which combines the specificity and sensitivity of both spin trapping and antigen:antibody interactions. Previously, a novel rabbit polyclonal anti-DMPO nitrone adduct antiserum, which specifically recognizes protein radical-derived nitrone adducts, was developed and validated in our laboratory. In the present study, the formation of nitrone adducts on hemoglobin was shown to depend on the oxidation state of the iron heme, the concentrations of H2O2 and DMPO, and time as determined by enzyme-linked immunosorbent assay (ELISA) and by Western blotting. The presence of reduced glutathione or L-ascorbate significantly decreased the level of nitrone adducts on metHb in a dose-dependent manner. To confirm the ELISA results, Western blotting analysis showed that only the complete system (oxy- or metHb/DMPO/H2O2) generates epitopes recognized by the antiserum. The specific modification of tyrosine residues on metHb by iodination nearly abolished antibody binding, while the thiylation of cysteine residues caused a small but reproducible decrease in the amount of nitrone adducts. These findings strongly suggest that tyrosine residues are the site of formation of the immunochemically detectable hemoglobin radical-derived nitrone adducts. In addition, we were able to demonstrate the presence of hemoglobin radical-derived nitrone adducts inside red blood cells exposed to H2O2 and DMPO. In conclusion, our new approach showed several advantages over EPR spin trapping with the anti-DMPO nitrone adduct antiserum by demonstrating the formation of tyrosyl radical-derived nitrone adduct(s) in human oxyHb/metHb at much lower concentrations than was possible with EPR and detecting radicals inside RBC exposed to H2O2.     

Reduction of methemoglobin via electron transfer from photoreduced flavin: restoration of O2-binding of concentrated hemoglobin solution coencapsulated in phospholipid vesicles

Ferric methemoglobin is reduced to its ferrous form by photoirradiation either by direct photoexcitation of the heme portion to induce electron transfer from the surrounding media (Sakai at al. (2000) Biochemistry 39, 14595-14602) or by an indirect electron transfer from a photochemically reduced electron mediator such as flavin. In this research, we studied the mechanism and optimal condition that facilitates photoreduction of flavin mononucleotide (FMN) to FMNH(2) by irradiation of visible light, and the succeeding reduction of concentrated metHb in phospholipid vesicles to restore its O(2) binding ability. Visible light irradiation (435 nm) of a metHb solution containing FMN and an electron donor such as EDTA showed a significantly fast reduction to ferrous Hb with a quantum yield (Phi) of 0.17, that is higher than the method of direct photoexcitation of heme (Phi = 0.006). Electron transfer from a donor molecule to metHb via FMN was completed within 30 ns. Native-PAGE and IEF electrophoresis indicated no chemical modification of the surface of the reduced Hb. Coencapsulation of concentrated Hb solution (35 g/dL) and the FMN/EDTA system in vesicles covered with a phospholipid bilayer membrane (Hb-vesicles, HbV, diameter: 250 nm) facilitated the metHb photoreduction even under aerobic conditions, and the reduced HbV restored the reversible O(2) binding property. A concentrated HbV suspension ([Hb] = 8 g/dL) was sandwiched with two glass plates to form a liquid layer with the thickness of about 10 microm (close to capillary diameter in tissue, 5 microm), and visible light irradiation (221 mW/cm(2)) completed 100% metHb photoreduction within 20 s. The photoreduced FMNH(2) reacted with O(2) to produce H(2)O(2), which was detected by the fluorescence measurement of the reaction of H(2)O(2) and p-nitrophenylacetic acid. However, the amount of H(2)O(2) generated during the photoreduction of HbV was significantly reduced in comparison with the homogeneous Hb solution, indicating that the photoreduced FMNH(2) was effectively consumed during the metHb reduction in a highly concentrated condition inside the HbV nanoparticles.