Redox Changes of Cytochrome b 559 in the Presence of Plastoquinones*

We have found that short chain plastoquinones effectively stimulated photoreduction of the low potential form of cytochrome b 559 and were also active in dark oxidation of this cytochrome under anaerobic conditions in Triton X-100-solubilized photosystem II (PSII) particles. It is also shown that molecular oxygen competes considerably with the prenylquinones in cytochrome b 559 oxidation under aerobic conditions, indicating that both molecular oxygen and plastoquinones could be electron acceptors from cytochromeb 559 in PSII preparations. α-Tocopherol quinone was not active in the stimulation of cytochrome photoreduction but efficiently oxidized it in the dark. Both the observed photoreduction and dark oxidation of the cytochrome were not sensitive to 3-(3,4-dichlorophenyl)-1,1-dimethylurea. It was concluded that both quinone-binding sites responsible for the redox changes of cytochrome b 559 are different from either the QA or QB site in PSII and represent new quinone-binding sites in PSII.

We have found that short chain plastoquinones effectively stimulated photoreduction of the low potential form of cytochrome b 559 and were also active in dark oxidation of this cytochrome under anaerobic conditions in Triton X-100-solubilized photosystem II (PSII) particles. It is also shown that molecular oxygen competes considerably with the prenylquinones in cytochrome b 559 oxidation under aerobic conditions, indicating that both molecular oxygen and plastoquinones could be electron acceptors from cytochrome b 559 in PSII preparations. ␣-Tocopherol quinone was not active in the stimulation of cytochrome photoreduction but efficiently oxidized it in the dark. Both the observed photoreduction and dark oxidation of the cytochrome were not sensitive to 3-(3,4-dichlorophenyl)-1,1-dimethylurea. It was concluded that both quinone-binding sites responsible for the redox changes of cytochrome b 559 are different from either the Q A or Q B site in PSII and represent new quinone-binding sites in PSII.
Cytochrome b 559 (cyt b 559 ) 1 is an integral component of the photosystem II (PSII) reaction center (RC) complex, whose function has been the subject of very extensive studies within recent years (1,2). It was found that it has a crucial role in PSII assembly, and its electron transport properties are attributed to protection of PSII against photoinhibition (1,(3)(4)(5). It is found in thylakoid membranes and in PSII preparations in high potential (HP) and low potential (LP) forms differing in redox potentials (E m ϭ 330 -400 and 20 -80 mV, respectively) (2), and both these forms are believed to be at least partially interconvertible (1,4,6). In native, fully functional PSII complexes, the LP form is probably a minor component whose levels increase under photoinhibitory conditions. Under strong light illumination (acceptor side inhibition), the LP form of cyt b 559 was found to accept electrons from the photoreduced pheophytin, preventing PSII from photoinhibition (3,5,7). However, the natural electron acceptor from the cyt b 559 LP form is unknown. Among the potential candidates, plastoquinone and molecular oxygen have been suggested (4). The second possibility is supported by autoxidation of the LP form of cyt b 559 (8).
Under the impairment of the oxygen-evolving complex, when electron donation from the oxygen-evolving complex to P680 is inhibited (donor side inhibition), the electrons from the HP form of cyt b 559 are transferred to P680 via the redox active chlorophyll Z (Chl z ) (1). Moreover, it was suggested (1) that under such conditions the HP form is converted by an unknown molecular switch to the LP cyt b 559 form, which after its autoxidation cannot donate electrons to P680. This state, in turn, causes oxidation of Chl z , which in a radical form efficiently quenches excitation energy, preventing RC from overexcitation (1). It is known that the electron transport via cyt b 559 is relatively slow (1), and to fulfill its photoprotective function, it should be relatively fast. Therefore, it is of interest to investigate the factors accelerating reduction and oxidation of the cyt b 559 LP form. It was found that in isolated reaction centers, where only the LP cyt b 559 form is present (9), the photoreduction of cyt b 559 was strongly stimulated by artificial quinones, such as DBMIB (10) or decyl-PQ (9). It was suggested that these quinones substitute for the Q A site and mediate cytochrome reduction via the semiquinone form. However, it is not certain whether the interactive site of the quinones is in fact the Q A site (11). Therefore, it is of interest to investigate in more detail the specificity of cyt b 559 photoreduction by different natural and artificial quinones, to investigate whether the quinone-binding site(s) are different from the Q A site, and to consider whether this quinone-mediated cytochrome reduction could take place in vivo.
In the present study, we have performed measurements of the photoreduction and dark oxidation of the cyt b 559 LP form in Triton X-100-solubilized PSII particles under the influence of different plastoquinone homologues in anaerobic and aerobic conditions. We have also included for these studies ␣-tocopherol quinone (␣-TQ), a natural thylakoid prenylquinone of unknown function in plants (12) that is an effective quencher of PSII fluorescence (13) and also affects the cyclic electron flow around PSII (14).

MATERIALS AND METHODS
PSII particles were isolated according to the method of Berthold et al. (15). PQ-1-PQ-4 homologues were a kind gift of H. Koike, PQ-9 was a gift from Hoffmann-La Roche, PQ-decyl was from Sigma, and vitamin K 3 (menadione) was from Aldrich. ␣-TQ was prepared from ␣-tocopherol as described by Kruk (16). Menadiol (K 3 H 2 , MD) was obtained by reduction of menadione with NaBH 4 in methanol, and its concentration was determined using the millimolar extinction coefficient ⑀ ϭ 20.1 at 245 nm in ethanol. Plastoquinone concentrations were calculated using ⑀ 255 ϭ 17.94 mM cm Ϫ1 (17) in absolute ethanol. All measurements of cyt b 559 redox changes on PSII particles were made in 50 mM Hepes buffer, pH 7.5, 10 mM KCl, and 5 mM MgCl 2 at a chlorophyll concentration of 50 g/ml and in the presence of 0.2% (w/v) Triton X-100. Additionally, photoreduction of cyt b 559 was measured in the presence of 1 mM MnCl 2 as an electron donor to PSII. The cyt b 559 concentration was determined using ⌬⑀ 559 -570 ϭ 20 mM cm Ϫ1 (18). Unless otherwise indicated, all kinetic measurements were performed under anaerobic conditions us-* This work was supported by Grant 6 P04A 028 19 from the Committee for Scientific Research (KBN) of Poland. 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.
ing an oxygen trap composed of glucose oxidase (50 units/ml ϭ 1.5 mg/ml), catalase (500 units/ml ϭ 23.8 g/ml), and 1 mM glucose in a tightly closed cuvette at a volume of 2.5 ml. Both aerobic and anaerobic samples were incubated for 10 min before the measurements under continuous stirring. That time was sufficient for oxygen consumption by the oxygen trap and for the stabilization of the base line after Triton X-100 addition. The presence of 0.2% Triton X-100 had no influence on the efficiency of the oxygen trap when checked by the Clark oxygen electrode. The absorption spectra were measured in a split beam mode, and the kinetic measurements were made in a dual wavelength mode at 559 -570 nm using an SLM Aminco DW2000 spectrophotometer. The prenylquinones were added to the reaction mixtures as solutions in ethanol of which the final concentration did not exceed 1%. Photoreduction experiments were performed using light from a halogen projector passed through a 630-nm cutoff filter. To protect the photomultiplier from the scattered light, an interference filter ( max ϭ 564 nm, T max ϭ 30%, full width at half-maximum ϭ 12 nm) was placed in front of the photomultiplier. All the measurements were performed at room temperature.

Absorption Spectra of Different cyt b 559
Forms-To observe only redox changes of the LP form of cyt b 559 in PSII particles and to avoid possible absorption changes due to the HP form, we have performed our measurements in the presence of Triton X-100, which is known to convert the HP form into the LP form of cyt b 559 (19). This detergent was also found to increase the ascorbate-reducible fraction of the cyt b 559 as compared with the dithionite-reducible cyt b 559 in PSII RC (20), probably by disaggregation of the RC preparations, which in turn improves penetration of added redox compounds to the cytochrome. This detergent should also increase the solubility of the added hydrophobic prenylquinones. The difference absorption spectra in the ␣-band region of cyt b 559 absorption show ( Fig. 1) that our PSII preparation solubilized in 0.2% Triton X-100 does not contain any detectable cyt b 559 HP redox form, whereas the ascorbate reduces the cytochrome to practically the same level as that produced by MD, which is a stronger reductant than ascorbate. Dithionite reduced the cytochrome to the same level as that produced by MD and ascorbate (data not shown), indicating that Triton X-100-solubilized PSII particles contain only the low-potential, totally ascorbate-reducible cyt b 559 form. The presence of Triton X-100 also improves the optical properties of the sample.
Photoreduction of cyt b 559 in the Presence of Plastoquinones-Because the Triton X-100 treatment inactivates the oxygenevolving complex, 1 mM MnCl 2 was added to all light-treated samples as an electron donor to PSII particles. Fig. 2 shows that under anaerobic conditions in the presence of even very low PQ-2 concentration (1.73 mol of PQ-2/mol of cyt b 559 ), the cytochrome is photoreduced at a fast rate and oxidized back to the original level after turning off the light. The second illumination of the same sample gives both faster photoreduction and photooxidation rates. Interestingly, the same sample in the presence of ambient oxygen shows no detectable redox changes of the cytochrome. This might indicate that the cytochrome is oxidized so quickly by molecular oxygen that there is no detectable steady-state level of the reduced cytochrome. The control sample, with no added quinone, shows almost no redox changes of the cytochrome. PQ-9 was also active in catalyzing the photoreduction and subsequent dark oxidation of the cytochrome. However, much higher amounts of PQ-9 are required to obtain significant levels of the cytochrome photoreduction. Even though the photoreduction is considerably less sensitive to PQ-9 than to PQ-2 in our measurements, it is much more sensitive than in analogous experiments performed on PSII RCs (20), where the cytochrome could be photoreduced to not more than 62% of the total cytochrome present in the preparation and where this saturation was reached at over 300 mol of PQ-9/mol of cyt b 559 . In our case, the reduction level of cyt b 559 is 65 and 40% for 110 PQ-9 and 55 PQ-9 samples (Fig. 2), respectively. It is also evident for samples with low PQ-9 content (55 PQ-9 and 28 PQ-9) that there is a residual fast photoreduction and dark oxidation absorbance change that is still present in the control sample (with no added quinone) but is not observed under aerobic conditions. These cytochrome redox changes could originate from the residual plastoquinone present in PSII preparations or a low efficient electron transport via cyt b 559 without the participation of the plastoquinone. In contrast to plastoquinones, ␣-TQ was not active in stimulation of the cytochrome photoreduction even at high concentrations (Fig. 2). None of the cytochrome redox changes shown in Fig. 2 was inhibited by DCMU, indicating that the Q B site is not involved in the electron transport to and from cyt b 559 under our conditions. Fig. 3 shows that the cytochrome photoreduction rate is proportional to the PQ-2 content and is already observed at very low (substoichiometric) amounts of PQ-2 in relation to cyt b 559 , indicating a very tight binding site of PQ-2 in the vicinity of the cytochrome. The cytochrome is totally reduced already at 7 molecules of PQ-2 per heme of the cytochrome within 20 s of the measurements (compare Figs. 1 and 3). The photoreduction of cyt b 559 in the presence of 14 mol of PQ-2/mol of cyt b 559 at different light intensities under anaerobic conditions is presented in Fig. 4. It is evident that the photoreduction rate is proportional to the light intensity but is not saturated even at the highest light intensity applied. On the other hand, full reduction is already reached at 100 mol/m 2 /s within the time of measurement. The time-dependent cyt b 559 photoreduction in the presence of plastoquinones with the different side chain length and DBMIB at 1.73 mol of quinone/mol of cyt b 559 under anaerobic conditions shows (Fig. 5) that with decreasing the side chain length of PQ the photoreduction rate increases. PQ-9 was not active in the photoreduction at this concentration. DBMIB, which was shown in some studies to catalyze the cyt b 559 photoreduction (10,11,21), was nearly inactive at the concentration applied in Fig. 5 but showed stimulation of the reaction at higher DBMIB/cyt b 559 proportions (data not shown). It was suggested that the unusually high potential of the cyt b 559 HP form is caused by the localization of the cyt b 559 heme ring within the very hydrophobic interior of the membrane and that the transformation of the HP form to the LP form is the result of heme exposure to the more polar surroundings (2,22). The results shown in Fig. 5 confirm that the heme of the cytochrome LP form is localized in polar surroundings, because the photoreduction is catalyzed most efficiently by relatively polar, short chain plastoquinones.
Dark Autoxidation of cyt b 559 -To analyze the possible oxidation routes of the photoreduced cyt b 559 (Fig. 2), we used chemical reduction of the cytochrome with a small excess of MD and followed the cytochrome oxidation under molecular oxygen (in air-saturated buffer) and exogenously added prenylquinones. Addition of MD under anaerobic conditions results in fast dark reduction of the cytochrome, which remains reduced for minutes (Fig. 6A). If MD is added under aerobic conditions, after transient reduction, the cytochrome is gradually oxidized by molecular oxygen (Fig. 6B). When the cytochrome is reduced under anaerobic conditions and H 2 O 2 is added, which immediately releases molecular oxygen on contact with the catalase, which is a component of the oxygen trap, the cytochrome oxidation is observed (Fig. 6C). The second H 2 O 2 addition, which should give a 200 M oxygen concentration in the reaction mixture, causes a higher cytochrome oxidation rate than that under aerobic conditions, where the oxygen concentration should be about 250 M. The reason for this could be that under the conditions illustrated in Fig. 6B there is an excess of MD left after the cytochrome reduction that slows down its autoxidation. In the case of the reaction illustrated in Fig. 6C, the first H 2 O 2 addition probably at least partially removes the MD excess. Dark Prenylquinone-mediated Oxidation of cyt b 559 - Fig. 7 illustrates PQ-2-induced dark cytochrome oxidation after its prereduction with MD. PQ-2 induces cytochrome oxidation at rates proportional to the quinone concentration, but under anaerobic conditions the extent of the cytochrome reduction decreases with increasing PQ-2 concentration (Table I). In the presence of oxygen, the cytochrome oxidation rate is considerably increased by the simultaneous cytochrome autoxidation. In this case, the cytochrome is completely oxidized. PQ-9 and ␣-TQ were also active in the oxidation of cyt b 559 (Fig. 8), but the oxidation rates in the presence of PQ-9 and ␣-TQ are considerably lower than those caused by PQ-2. In contrast to PQ-2, the oxidation extent is considerably higher for PQ-9 and ␣-TQ (Table I). The oxidation rates under aerobic conditions are evidently higher, and the rates correspond approximately to the oxidation rates under anaerobic conditions increased by the cytochrome autoxidation rate (Table I). Comparing the oxidation rates during dark periods in the photoreduced samples (Fig. 2) and the rates obtained from Figs. 7 and 8, it can be seen that the rates for the dark cytochrome oxidation obtained from experiments on cyt b 559 prereduced by MD would not be high enough to account for the observed cytochrome oxidation rates prereduced by light at the same quinone concentrations. The reason for this could be partially connected with the excess of MD that remains after the cytochrome reduction. This probably inhibits the oxidation of the cytochrome by the added quinones, which can be deduced from comparison of the ob- served rates of the cytochrome autoxidation by ambient oxygen and the oxygen released from H 2 O 2 (Fig. 6, B and C). Nevertheless, the above experiments show that cyt b 559 can be effectively oxidized by prenylquinones or molecular oxygen in our system. DCMU had no influence on the observed dark oxidation of the cyt b 559 .

DISCUSSION
The observed effects described above of stimulation of cyt b 559 photoreduction by different plastoquinones are similar to those observed with DBMIB, PQ-decyl, or PQ-9 on the isolated PSII reaction centers (9,10,20), but in our case the photoreduction proceeds with higher quinone sensitivity and to a higher cytochrome reduction extent. The efficient photoreduction of the cytochrome at very low PQ-2 concentrations (Fig. 3) suggests a very tight binding site close to the cytochrome heme. The question arises as to whether it is one of the known quinone-binding sites in PSII (Q A or Q B ) or another high affinity quinone-binding site in PSII. The lack of DCMU sensitivity of the cytochrome redox changes and the fact that the Q B site is detergent-sensitive (23) and easily damaged indicate that the binding site is not Q B . Moreover, to efficiently reduce the cyt b 559 LP form, quinone molecules must operate via the semiquinone form, similarly to Q A , and it is known that the Q B site operates via a two-electron gate. However, involvement of Q A also seems unlikely. It has been suggested that in isolated reaction centers, where the Q A and Q B plastoquinone molecules are lost (21,24), the added quinones substitute for the Q A site (10,11,20). However, in our case this seems unlikely, because we used PSII membranes. It was shown that it is very difficult to remove or substitute the Q A molecule in isolated PSII core preparations (25). It requires high (1%) Triton X-100 concen-trations in the presence of phenanthroline and hours of incubation time (25). In our experiments the samples were incubated only with 0.2% Triton X-100 for 10 min before the measurements. Moreover, we have found that the photoreduction changes were already observed within a shorter time but that the base-line stability and signal to noise ratio were poor in that case (data not shown). Interestingly, the Q A reconstitution experiments in PQ-9-depleted PSII core preparations (26) showed that this site shows a high structural specificity, and only PQ-9 or UQ-10 efficiently substituted for Q A function, in contrast to PQ-2, DBMIB, or other short chain quinones. A similar specificity, only for PQ-9 or UQ-9, was observed in PSII core complexes where Q A was exchanged by 1% Triton X-100phenanthroline incubation (25). On the other hand, rather low structural quinone specificity is observed for cyt b 559 quinonecatalyzed photoreduction (9). These data favor the view that the quinone-binding site active in cyt b 559 LP photoreduction is different from the Q A and Q B sites. It is also evident that the Q A site shows specificity for hydrophobic quinones (PQ-9, UQ-9 or UQ-10), whereas the cyt b 559 quinone-binding site shows preference for polar plastoquinones or other artificial quinones, which is connected with the exposure of the heme of the cyt b 559 LP form to the polar surroundings.
The question arises as to whether this cyt b 559 LP quinonebinding site is occupied and functional in vivo in the native PSII. Some indication may come from analyses of the PQ-9 content of PSII membranes and PSII core complexes. Such analyses usually give about three or more PQ-9 molecules per RC in PSII membranes (27,28); two of these molecules are strongly bound (non-extractable by hexane) (27), and about two are tightly bound PQ-9 molecules for PSII core complexes (28 -30). In both preparations, one molecule is certainly the Q A molecule, but the identity of the second PQ-9 in PSII core complexes is not certain (29,30). Probably it is not a Q B molecule, because this site is easily damaged by detergent treatment during the PSII core isolation procedure, and the Q B molecule is lost (23, 28 -30). This also would explain poor or no DCMU sensitivity of the electron transport in PSII core complexes (29,30). It was originally thought that the additional, tightly bound PQ-9 molecule to PSII membranes is Z (27), the primary electron donor to P680, but it turned out that it was tyrosine. This data may suggest that the second, tightly bound PQ-9 molecule in PSII core complexes is the PQ-9 molecule participating in cyt b 559 photoreduction via the semiquinone form and is different from the Q A molecule.
The site of quinone interaction with the cyt b 559 causing its oxidation seems to be a low affinity quinone-binding site that is also DCMU-insensitive and is probably in contact and exchange with the PQ pool in thylakoid membranes. An argument supporting the concept of different photoreduction and dark oxidation sites of cyt b 559 LP is the inefficiency of ␣-TQ in the stimulation of the cytochrome photoreduction; but at the same concentration ␣-TQ efficiently oxidizes the reduced cytochrome.
The model of electron transport pathways in PSII showing the PQ-binding sites (Q A and Q B ) together with the high (Q C ) and low (Q) affinity PQ-binding sites involved in cyt b 559 redox changes is shown in Fig. 9. The possible physiological role of the quinone-stimulated cyt b 559 photoreduction (via the Q C site) and its reoxidation (via the Q site) could be the acceleration of the electron transport from pheophytin to the PQ pool or to molecular oxygen via cyt b 559 LP upon acceptor side inhibition ( Fig. 9) when the electron route via Q A 3 Q B 3 PQ pool is saturated or inhibited and electron donation to P680 from the oxygen-evolving complex is active. When PSII is fully functional, electron transport via Q A 3 Q B dominates, and the cyt b 559 exists mainly in the HP form, which is kept reduced by the plastoquinol of the PQ pool. Under donor side inhibition, P680 is reduced by the cyt b 559 HP form via the redox active Chl Z molecule (1). Moreover, a ␤-carotene molecule probably mediates electron transfer between Chl Z and P680 (31,32). The hypothetical molecular switch transforms the HP to the LP form, which is further oxidized by molecular oxygen or a neighboring PQ molecule (Q in Fig. 9), and the oxidized LP form is no longer able to donate electrons to P680. Then, Chl Z is oxidized, giving Chl Z ϩ , which effectively quenches the excitation energy, preventing the photosynthetic apparatus against overexcitation (1). The electron route via Q C upon acceptor side inhibition also requires the HP3 LP molecular switch to give a sufficient cyt b 559 LP level through which electrons could outflow from pheophytin via Q C and the cyt b 559 LP form to O 2 or PQ pool, preventing charge recombination in the RC, leading to P680 triplet state formation and consequently RC degradation (2,21).