Distinct Roles of Two Heme Centers for Transmembrane Electron Transfer in Cytochrome b 561 from Bovine Adrenal Chromaffin Vesicles as Revealed by Pulse Radiolysis*

The reaction of monodehydroascorbate (MDA) radical with purified cytochrome b 561 from bovine adrenal chromaffin vesicles was investigated by the technique of pulse radiolysis. Radiolytically generated MDA radical oxidized rapidly the reduced form of cytochrome b 561 to yield the oxidized form. Subsequently the oxidized form of cytochromeb 561 was re-reduced by ascorbate in the medium. The second-order rate constants of the reaction of MDA radical were increased with decreasing pH, whereas a maximum of the second-order rate constant for the reaction with ascorbate was obtained around pH 6.8. At excess MDA radical to cytochrome b 561concentration, only half of the heme in cytochromeb 561 was oxidized, indicating that only one of the two heme centers can react with MDA radical. On the other hand, when the reactions were examined using cytochromeb 561 pretreated in a mild alkaline condition in the oxidized state, the cytochrome b 561 could not be oxidized with MDA radical, suggesting that the heme center specific for the electron donation to MDA radical is selectively modified upon the alkaline treatment. These results suggest that the two heme b centers have distinct roles for the electron donation to MDA radical and the electron acceptance from ascorbate, respectively.

In neurosecretory vesicles, such as adrenal chromaffin vesicles and pituitary neuropeptide secretory vesicles, intravesicular ascorbate (AsA Ϫ ) 1 functions as the electron donor for copper-containing monooxygenases such as dopamine ␤-monooxygenase and peptidyl-glycine ␣-amidating monooxygenase (1). Upon these monooxygenase reactions, monodehydroascorbate (MDA) radical is produced by univalent oxidation of AsA Ϫ (2,3). It is believed that the intravesicular MDA radical is reduced back to AsA Ϫ by membranebound cytochrome b 561 , and subsequently cytochrome b 561 is reduced by extravesicular AsA Ϫ (4-7). Thus, cytochrome b 561 is likely to serve as an electron shuttle, maintaining the AsA Ϫ concentration inside the vesicles.
Cytochrome b 561 is a highly hydrophobic hemoprotein with a molecular mass of ϳ28 kDa and contains five or six transmem-brane ␣-helices (8,9). It had been widely accepted that cytochrome b 561 contains only one b-type heme per molecule, by analyses with pyridine hemochrome and Western blotting methods for quantitation of heme and apoprotein, respectively (10 -13). However, very recently, we have established a new purification procedure of cytochrome b 561 from bovine adrenal chromaffin vesicles (14). We found that the purified cytochrome b 561 contained two b-type hemes per molecule. In addition, each heme b center exhibited independent EPR signals in the oxidized state. Based on these results and comparison of the amino acid sequences of cytochrome b 561 from various species, we have proposed that the two heme prosthetic groups are located on both sides of the membrane in close contact with AsA Ϫ and MDA binding sites, respectively, to facilitate the electron transfer across the membranes (14,15). However, the role of each of the two heme b centers has not been elucidated.
Because of the instability of the MDA radical, which disproportionates rapidly to dehydroascorbate and AsA Ϫ , the electron transfer reaction with cytochrome b 561 was measured indirectly in the presence of AsA Ϫ and ascorbate oxidase (5, 6). However, it is possible to investigate directly the reaction of MDA radical using a pulse radiolysis technique (16 -20). We could observe electron donations to MDA radical from hepatic NADH-cytochrome b 5 reductase (19) and from MDA reductase purified from cucumber chloroplasts (20). In particular, MDA reductase was shown to be a good electron donor for MDA radical (20). The present study describes a successful application of the pulse radiolysis technique to investigate the reaction of MDA radical with cytochrome b 561 purified from bovine adrenal chromaffin vesicles. We obtained clear evidence that the two heme b centers in cytochrome b 561 have distinct roles in the reaction with MDA radical and AsA Ϫ , respectively.

MATERIALS AND METHODS
Purification of Cytochrome b 561 -Cytochrome b 561 was purified to a homogeneous state, as reported previously (14). The purity of cytochrome b 561 was analyzed with visible absorption spectra and SDSpolyacrylamide gel electrophoresis. Before use, cytochrome b 561 was passed through a Sephadex G-25 column equilibrated with 10 mM potassium phosphate buffer (pH 7.0) containing 1.0% (w/v) ␤-octyl glucoside and 1.0 mM AsA Ϫ . The passed fractions of cytochrome b 561 were concentrated in the Amicon concentrator. All other reagents were commercially obtained as the analytical grade. The concentration of cytochrome b 561 was determined using a millimolar extinction coefficient of 267.9 mM Ϫ1 cm Ϫ1 at 427 nm in the reduced state (14).
Pulse Radiolysis-Samples of cytochrome b 561 for pulse radiolysis were prepared as follows. Solutions containing 10 mM potassium phosphate buffer (pH 5-8), 1% ␤-octyl glucoside, and 5 mM AsA Ϫ were bubbled with N 2 O gas for 5 min. Then, a concentrated solution of cytochrome b 561 was added to the solution to make an appropriate final concentration as indicated in the figure legends.
Pulse radiolysis experiments were performed with an electron linear accelerator at the Institute of Scientific and Industrial Research, Osaka * This work was supported by Grants-in-aid for Scientific Research on Priority Areas (Molecular Biometallics) 08249104 (to K. K.) and 08249234 (to M. T.) and Grant-in-aid 08680727 from the Japanese Ministry of Education, Science, Sports and Culture (to M. T.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
University (19 -23). The pulse width and energy were 8 ns and 27 MeV, respectively. The sample was placed in a quartz cell with an optical path length of 1 cm. The temperature of the sample was maintained at 20°C. The light source for a spectrophotometer was a 150-watt halogen lamp. After passing through an optical path, the transmitted light intensities were analyzed and monitored by a fast spectrophotometric system composed of a Nikon monochromator, an R-928 photomultiplier, and a Unisoku data analyzing system. For each measurement, a fresh sample was used even though pulse radiolysis did not cause any damage to the sample as judged by its visible absorption spectrum. The concentration of MDA radical generated by pulse radiolysis was estimated from the absorbance at 360 nm, using a molar extinction coefficient of 3300 M Ϫ1 cm Ϫ1 (18). This concentration could be adjusted by varying the dose of electron beams.
Treatment in Alkaline pH in the Oxidized Form-Purified cytochrome b 561 (ϳ100 M) in 10 mM potassium phosphate buffer (pH 7.0 and 8.4) containing 1.0% ␤-octyl glucoside was oxidized with stepwise additions of potassium ferricyanide (100 mM) solution. The fully oxidized cytochrome b 561 samples were kept in the dark overnight on ice. The samples thus obtained were passed through Sephadex G-25 equilibrated with 10 mM potassium phosphate buffer (pH 7.0) containing 1.0% (w/v) ␤-octyl glucoside and 1.0 mM AsA Ϫ . Optical absorption spectra were recorded with a UVIKON 922 (Kontron), a UV-2200 A (Shimadzu), or a Hitachi U-3000.

RESULTS
A transient spectrum of MDA radical with an absorbance maximum at 360 nm was observed 100 ns after pulse radiolysis of N 2 O-saturated aqueous solutions in the presence of 5 mM AsA Ϫ and 1.0% ␤-octyl glucoside. Under the conditions employed here, the primary species (hydrated electron (e aq Ϫ ), OH ⅐ , H ⅐ ) generated by pulse radiolysis of aqueous solutions were efficiently converted to MDA radical at an approximate concentration of 20 -30 M (17-20). The MDA radicals thus formed reacted rapidly with the reduced form of cytochrome b 561 . A decrease at 430 nm and an increase at 405 nm reflected this reaction (Fig. 1A). The kinetic difference spectrum obtained 20 ms after the pulse is similar to that of the difference spectrum of the oxidized minus reduced forms of cytochrome b 561 (Fig.  1B). It is therefore concluded that MDA radical reacts with the reduced form of cytochrome b 561 to produce the oxidized form, as shown in Reaction 1.
For determination of the rate constant of Reaction 1, the concentration of MDA radical was lowered to 1-2 M MDA radical, and the cytochrome b 561 concentration was varied between 46 and 100 M. Fig. 2 shows the concentration dependence of cytochrome b 561 on the apparent rate constants. From the slope of Fig. 2, the second-order rate constant of the reaction was calculated to be 2.6 ϫ 10 6 M Ϫ1 s Ϫ1 at pH 7.0. This value is in good agreement with the one reported previously (1.2 ϫ 10 6 M Ϫ1 s Ϫ1 at pH 7.0) in which a steady-state kinetics method was used employing chromaffin vesicle membranes and an AsA Ϫ -ascorbate oxidase system (7).
Subsequently, the initial changes in absorbance reversed in the time range of seconds (Fig. 3), indicating that re-reduction of cytochrome b 561 occurred. The rate constant of this process increased with increases in the concentration of AsA Ϫ (data not shown). This indicates that the reduction process is a consequence of a bimolecular reaction of AsA Ϫ with the oxidized form of cytochrome b 561 , as shown in Reaction 2.
The second-order rate constant of the reaction was calculated to be 8 ϫ 10 2 M Ϫ1 s Ϫ1 at pH 7.0. The value is also good agreement with the value reported previously (4.5 ϫ 10 2 M Ϫ1 s Ϫ1 ) in which a stopped-flow method was employed to analyze the electron transfer reaction from AsA Ϫ to oxidized cytochrome b 561 in chromaffin vesicle membranes (7,24).
The effects of pH on the oxidation rates of cytochrome b 561 with MDA radical and for reduction with AsA Ϫ were examined. It is evident that pH profiles for the oxidation and the reduction reactions are different, as shown in Fig. 4. The rate constant for the oxidation of cytochrome b 561 increased with decreasing pH. The maximum reaction rate of 4.3 ϫ 10 6 M Ϫ1 s Ϫ1 was obtained at pH 5.5. The pH-dependent change could be fitted to a single deprotonation process with a pK a value of 6.7. In contrast, the rate constants of the reduction of cytochrome b 561 increased with increasing pH in the range of 5 to 6.5 and then decreased with increasing pH in the range of 7 to 8. The maximum reaction rate constant of 8 ϫ 10 2 M Ϫ1 s Ϫ1 was obtained around pH 6.8.
To elucidate the contribution of two heme b centers in the electron transfer reaction, MDA radical concentration dependence was examined. Under the experimental conditions employed, the concentration of MDA radical could be varied between 2 and 20 M by attenuating the dose of the electron beam whereas the concentration of the reduced cytochrome b 561 was maintained constant at 12 M. As shown in Fig. 5, the oxidized In a previous report, we showed that incubation of oxidized cytochrome b 561 in a mild alkaline condition specifically depletes the electron-accepting ability from AsA Ϫ for about onehalf of the heme centers (15). Fig. 6A shows the visible spectra of cytochrome b 561 (pre-treated at pH 8.2 and 7.0, respectively, in the oxidized state) in the presence of AsA Ϫ . Addition of 5 mM AsA Ϫ to the alkaline treatment of cytochrome b 561 caused a reduction of only about half of the heme centers, whereas all the heme centers of cytochrome b 561 treated at pH 7.0 (control sample) could be reduced. Fig. 6, B and C, shows the absorb-ance changes at 430 nm after pulse radiolysis of these samples. The absorbance changes of the control sample after the pulse were not affected. For the alkaline-treated sample, on the other hand, the decrease in absorbance at 430 nm after the pulse was very small, although about half of the heme b centers were in the reduced state. The slower second phase (i.e. the absorbance increase at 430 nm due to the re-reduction of oxidized cytochrome b 561 with AsA Ϫ ) was completely lost. These results suggest that one of the heme centers, responsible for electron donation to the MDA radical, is selectively modified upon the alkaline treatment. DISCUSSION The present study clearly shows that one of the heme b centers in cytochrome b 561 specifically reacts with MDA radical, whereas the other does not. This is verified by the MDA radical concentration dependence on the oxidization of heme b after the pulse as shown in Fig. 5. Almost stoichiometric oxidation of heme b was observed under the condition of [MDA radical] Ͻ [1/2 of cytochrome b 561 ], whereas only one-half of heme b was oxidized at excess MDA radical concentrations. In addition, the distinct functions of two heme b centers were revealed by pulse radiolysis experiments on the alkalinetreated sample. Our previous report showed that one of the heme centers is very labile to the alkaline treatment, whereas the other heme retains the ability to accept electrons from AsA Ϫ (15). It is likely that one of the two heme centers, which is very labile to the alkaline treatment, participates in the electron donation to MDA radical, because very slight oxidation of the heme center was observed after pulse radiolysis of the alkaline-treated sample.
EPR spectra of the purified cytochrome b 561 in the oxidized state showed the presence of two distinct heme b species (14). One of them shows usual low spin signals and is very similar to those of microsomal cytochrome b 5 (25), chloroplast cytochrome b 559 (26), and cytochrome b of bo-type ubiquinol oxidase (27), all of which are known to have bisimidazole ligands. The other species shows a highly anisotropic low spin signal (g z ϭ 3.70) with a lower redox potential and is very similar to those of cytochrome b of the mitochondrial complex III (28,29) and chloroplast cytochrome b 6 (30). The presence of two independent heme centers was supported by the observation of two potentiometrically different forms of cytochrome b 561 determined by an optical potentiometric technique (31). We could not identify definitely which heme center is responsible for the electron donation to MDA radical, but it is very likely that the heme center with the lower redox potential (the g z ϭ 3.70 species) participates in the electron-accepting reaction from the extravesicular AsA Ϫ . On the other hand, the usual low spin heme species (the g z ϭ 3.14 species) with a higher redox potential is responsible for the electron donation to MDA radical. This assumption is reasonable since the intramolecular electron transfer of cytochrome b 561 occurs from extravesicular to intravesicular to sides. This is also consistent with the result of the alkaline treatment experiment. Indeed, the heme center having the g z ϭ 3.14 signal can be converted to another form (the g z ϭ 2.84 species) upon elevation of pH, whereas the other heme center (the g z ϭ 3.70 species) showed only a slight pHdependent spectral change (14).
Following the electron donation to MDA radical from the heme with a higher redox potential, an intramolecular electron transfer to the oxidized heme from the other heme with a higher redox potential should take place. The expected intramolecular electron transfer, however, could not be followed directly by the present method, because the two heme centers have indistinguishable visible spectra (14,31). If the rate of the intramolecular electron transfer (k i ) is much faster and the rate-determining step is the oxidation of the heme center with MDA radical (k i Ͼ k 1 [MDA radical]), the two hemes would be oxidized simultaneously. In the present work, however, only half of the heme in cytochrome b 561 was oxidized under excess MDA radical concentration. This fact indicates that the intramolecular electron transfer must occur later than 1 ms region. In our previous studies, the intramolecular electron transfer in the region of milliseconds was observed for coppercontaining nitrite reductase (21,22) and cytochrome cd 1 nitrite reductase (23). However, the two redox centers in these proteins are only 10 to 20 Å apart (32,33). In cytochrome b 561 , the two heme centers are expected to be located on both sides of vesicular membranes (15), and therefore the distance between these two redox centers might be 40 to 50 Å. Thus, an intramolecular electron transfer in the range of seconds may be quite reasonable.
The pH dependence of the rate constants of the electron transfer reactions (Fig. 4) provides further insight into the function of cytochrome b 561 . The optimal pH of the oxidation (pH ϭ 5.5) and the reduction (pH 6.8) of the heme centers of cytochrome b 561 correspond to the physiological pH at the intra-and extravesicular sides, respectively. Recently, we have proposed a plausible structural model of cytochrome b 561 on the basis of a comparison of the deduced amino acid sequences of seven species (15). In the model, there are two fully conserved regions in the sequences; the first conserved sequence ( 69 ALLVYRVFR 77 ) is located on the extravesicular side of ␣-helical segment, and the second one ( 120 SLHSW 124 ) is located the oxidized state were prepared as described in the text. Solid sodium ascorbate was added (5 mM), and the spectra were recorded immediately. Other conditions were the same as in Fig. 1. in an intravesicular loop connecting two ␣-helical segments. Since these conserved sequences are likely to form the binding sites for extravesicular AsA Ϫ and intravesicular MDA, respectively, the present results support our proposal that the two heme b centers are located on both sides of the vesicular membranes in close contact with the AsA Ϫ -and MDA-binding sites (15).
It is important to note that this is the first direct observation of electron transfer between a ferrous hemoprotein and an MDA radical generated by pulse radiolysis. In the previous report, we could not observe such electron transfer reactions with MDA radical for several ferrous hemoproteins (19), although the reactions are expected to be energetically favorable on the basis of redox potential differences between MDA . /AsA Ϫ (E m,7 ϭ 330 mV) (34) and hemoproteins (35). It is evident that, for biological molecules, the occurrence of electron transfer with MDA radical cannot be predicted simply in terms of the redox potentials (19). The highest rate constant for the reaction of MDA radical with biological molecules was obtained for MDA reductase (2.6 ϫ 10 8 M Ϫ1 s Ϫ1 ) (20). We have proposed previously that the high rate constant of the enzymatic reaction might be governed by both the specific geometry of substrate within the active site and the redox potential difference. The facilitated electron transfer reaction in MDA reductase is likely to be provided by several cationic amino acid residues near the active site, which may have a role in electrostatic guidance of the anionic MDA radical substrate to the active center (20). It is noteworthy that, in the intravesicular side of cytochrome b 561 , there are several conserved positively charged residues located in the loop connecting helices 3 and 4, and the fully conserved sequence ( 120 SLHSW 124 ) follows this region immediately (15).
In conclusion, the following unique properties of cytochrome b 561 have now become evident. First, an efficient transmembrane electron transfer is catalyzed by this hemoprotein, using AsA Ϫ as a physiological electron supply. Second, the two heme b centers have distinct roles for the electron donation to MDA radical and the electron acceptance from AsA Ϫ . Further studies are in progress to reveal the structural properties of this cytochrome.