The Redox-controlled Light-harvesting Chlorophyll a/b Protein Kinase DEACTIVATION BY SUBSTITUTED QUINONES*

The deactivation of the redox-controlled light-har- vesting chlorophyll a/b protein kinase of Acetabularia acetabulum and pea thylakoids was studied. Substi- tuted benzoquinone, naphthoquinone, and anthraquinone analogs including mono-, di-, and trihalogenated and/or alkylated quinones, which are known to inhibit the cytochrome bO/f activity, deactivate the kinase in the dark, and prevent its activation in the light. Analogs halogenated at positions 2- or 3- are the most effective deactivators. Increasing the size of the alkyl side chain and/or the number of rings lowers the deac- tivation effect. The activated state of the pea kinase decays with a tH of 15 min, while the Acetabularia enzyme retains its active state for at least 2 h. The midpoint potential for Acetabularia kinase activity in the dark is 120 2 10 mV and is compatible with the involvement of plastoquinone in the kinase activation via reduction of the cytochrome complex. Deactivation of kinase by the analogs inhibiting cytochrome b,/f complex activity and the kinase copurification with the cytochrome be/f fraction obtained from the Acetabularia

tal system, it was possible to demonstrate that a putative quinone binding site may be directly involved in the inhibition of the basal kinase activity. The process of kinase activation by interaction with the cytochrome bs/f complex increased the affinity of the kinase toward LHCII (10). However, the degree of activation exhibited by this system was relatively low as compared with intact thylakoids.
We have previously reported that the LHCII kinase present in prochlorophytes, Prochloron and Prochlorothrix hollandica, retains its activity in the dark in vivo or in isolated thylakoids in vitro (11,12). A similar situation was also found in chloroplasts of Acetabularia acetabulum (previously referred to as A. mediterranea), a unicellular green alga of the Dasycladales. The LHCII kinase of these cells is redox-controlled (9). The activity of Acetabularia LHCII kinase persisted in isolated thylakoids in the dark significantly longer than that of higher plants' thylakoids, such as pea or spinach. In the present work, we have used this experimental system, which offers the possibility of testing the role of cytochrome b6/f in the redox control of the enzyme by use of different cytochrome b6/f inhibitors.
The results of these experiments demonstrate the role of cytochrome bs/f in the deactivation of LHCII kinase and characterize the properties of the quinone analog(s) binding to the site(s) involved in the deactivation process.

Biological Material
Acetabularia cells were grown at the Max-Planck Institute as previously described (13). Cells 2-3 months old were transferred to plastic bottles (10,000-40,000 cells/transport) and were carried to the biological chemistry department in Jerusalem within 6-8 h. The cells

Effect of Quinone Analogs on
Deactivation of LHCII Kinase 25909 were transferred to deep Petri dishes containing 500-1000 cells in 200 ml of artificial seawater (13) and kept at 22 "C with a 12-h light, 12-h dark regime for up to 2 weeks. During this period, experiments on living cells were performed, and from the remaining cells, thylakoids were prepared and frozen until further use. A total of 150,000 cells was used for this work. Pea plants (Pisurn satiuum, Dan hybrid) were grown under a similar light regime as previously described (14).

Preparation of Thylakoids
Acetabularia thylakoids were prepared as described (9). The cells (approximately 4-5 cm long) were washed in fresh medium and aligned in parallel bundles of 100 cells. The stalks were tied with a sewing string close above the nucleus. The "hats" were removed by cutting while the cell bundles were submerged, and the bundles were hung with the cut stalks pointing down in test tubes containing 2 ml of an isotonic buffer consisting of 20 mM potassium phosphate, 1 mM MgClz, 2 mM EDTA, 0.6 M sorbitol, 20 mM NaC1, 2.5% w/v Ficoll (Pharmacia LKB Biotechnology Inc.), and 50 mM MES buffer, pH 6.1. The tubes were spun at 3,000 X g for 10 min at 4 "C to release the cell content. The chloroplast pellet was resuspended in hypotonic buffer (5 mM MgClz, 50 mM Tricine, pH 8.01, frozen in liquid nitrogen, and thawed to ensure chloroplast breaking. After centrifugation (15,000 X g for 15 min) the thylakoid membrane pellet was resuspended in the same buffer to a final chlorophyll concentration of 0.4 mg . ml" and stored in ice until use.
Pea thylakoids were prepared as previously described (14) and were resuspended at a final chlorophyll concentration of 0.4 mg. ml" in 50 mM Tricine, pH 7.8, 5 mM MgClz, 10 mM NaCl (TMN buffer). The thylakoids were stored in ice in the dark for up to 2 h. For long-term storage, both Acetabularia and pea thylakoids were immediately frozen in liquid nitrogen and stored for up to 3 months at -80 "C without significant loss of kinase activity.

Assays
Phosphorylation of Thylakoid LHCZZ and Histone ZZZ-s-LHCII phosphorylation in isolated thylakoids of Acetabularia was assayed as previously reported (9). The assay mixture in a final volume of 100 pl consisted of 50 mM HEPES buffer, pH 7.5, 10 mM MgC12, 5 mM NaF, 0.1 mM [ T -~~P ] A T P (0.1 rnCi.ml-'), and thylakoids equivalent to 2 pg of chlorophyll. Incubation was in white light (100 watts/m-*) or in the dark at 25 "C. Duroquinol(O.1 mM) was reduced with sodium borohydride as described (15). The quinone analogs, at concentrations as indicated, were dissolved in methanol or dimethyl sulfoxide (MezSO), and the final solvent concentration did not exceed 2% v/v. The reaction was linear with time and was terminated after 10 min by centrifugation at 13,000 X g in an Eppendorf microcentrifuge. The pellet was dissolved in the electrophoresis sample buffer (16) and heated for 2 min at 80 "C. Samples were then analyzed by SDS-PAGE and their phosphorylation pattern monitored by autoradiography. Quantitation of the radioactivity of LHCII bands was carried out by densitometry using a SL-TRFF scanner equipped with a soft laser beam.
Phosphorylation assay of pea thylakoids was similar except that the buffer used was 50 mM MES, pH 6.5. The choice of the phosphorylation assay buffer and pH was based on measurements of the pH optimum for each type of thylakoid (data not shown).
Histone 111-s phosphorylation assays were carried out in a 100-pl final volume containing the same reaction mixture as above except that the buffer was 50 mM Tris-C1, pH 8.0, and contained 5 mM dithiothreitol and 0.25 mg of substrate. Histone phosphorylation was used as an assay for the presence of kinase activity at various stages of cytochrome be/f complex purification. The amount of fraction added was equivalent to 5 pg of chlorophyll or 10-15 pg of total protein, and the reaction was terminated by cooling on ice. Samples (60-pl) were drawn and absorbed on Whatman No. 3MM paper discs followed by precipitation with cold 10% (w/v) trichloroacetic acid as described (9) and counting in a scintillation counter.
Electrophoretic Resolution of Polypeptides-Thylakoid membrane polypeptides and cytochrome b6/f fractions were resolved by SDS-or lithium dodecyl sulfate-PAGE using the method of Laemmli (16) and the Hoeffer minigel system. The gels were stained with Coomassie Brilliant Blue R-250, dried, and if radioactively labeled, exposed to x-ray film. For Western blotting, the resolved polypeptides were electrotransferred to nitrocellulose paper and immunodetected using phatase as described (17).
"'1-iodinated protein A or goat anti-rabbit conjugated alkaline phos-Measurements of Photosynthetic Actiuity-Electron flow via photosystem I1 was measured using isolated thylakoids with water or 1,5,-diphenylcarbazide as electron donors and DCIP as an electron acceptor as previously described (18). The light-dependent DCIP reduction was followed spectrophotometrically using a Kontron Uvikon-860 spectrophotometer. The reaction mixture containted 5 pg of chlorophyll, 20 mM MES buffer, pH 6.5, 10 mM NaCI, and 5 mM MgClZ at 25 "C. Light excitation was provided through a Schott 665 cut-off filter, and the photomultiplier was protected by a Corning 6-90 blue filter.
The Mehler reaction using duroquinol reduced by ascorbate as an electron donor and methyl viologen as an electron acceptor was according to White et al. (19).
Variable fluorescence kinetics was measured using a home-built fluorometer as previously described (20).
Isolation and Qunntitation of Cytochrome b6/f Complex Cytochrome bs/f complex was isolated from Acetabularia cells using the method described by Wynn et al. for isolation of this complex from green algae (22). For spectrophotometric quantitation, extinction coefficients of t554 18 mM".cm" for cytochrome f and 14 mM".cm" for cytochrome bs were used, respectively (23). Heme staining was carried out as described (24).
Protein and chlorophyll concentrations were determined according to Lowry et al. (25) and Arnon (26), respectively.
All chemicals used in this work were of analytical grade.

RESULTS
Stability of LHCII Kinase-activated State-The thylakoidbound LHCII kinase can be reversibly activated by redox reagents such as plastoquinol or duroquinol (15, 30, 31). The resulting active state can be deactivated by oxidizing reagents such as plastoquinone or ferricyanide (31). LHCII kinase can also be irreversibly inactivated by alkylating agents (21). To distinguish between enzyme deactivation (yet preserving the potential for reactivation by reducing agents) and loss of activity due to irreversible inactivation, we shall refer below to the kinase as being in different activation states: deactivated or having a basal activity, an activated state, and an irreversibly inactive state (Scheme 1).
To test the stability of the activated state of the Acetubu- FIG. 1. Deactivation of thylakoid LHCII kinase during storage in the dark. Isolated thylakoids (20 pg of chlorophyll.ml") of Acetabularia and pea kept at -80 "C were thawed in the dark and stored in ice. Pea thylakoid samples were illuminated for 30 s to activate the kinase and further stored in the dark in ice. At times as indicated, samples of the dark-incubated Acetabularia and preactivated pea thylakoids were assayed for LHCII phosphorylation in the dark or light for 10 min at 25 "C. The phosphorylation activity was estimated by excising the LHCII polypeptide bands resolved by SDS-PAGE and counting the incorporated radioactivity. The ratios of the dark activity to that measured in the light are plotted, while the light activity was taken as 100%. The 100% values were 0.75 nmol 32P. mg" (Acetabularia) and 2.1 nmol 32P.mg" (pea) incorporated into the 27-kDa LHCII band for the light-driven reaction. The amount of LHCII was estimated as in Ref. 9. laria kinase in the dark, frozen thylakoids were thawed and further kept in ice in the dark for up to 2 h. At times as indicated, samples were taken and the phosphorylation of LHCII was assayed in the dark or light. The results of such an experiment ( Fig. 1) show that the fraction of the activated kinase in the dark remained constant relative to the lightactivated kinase for at least 2 h. For comparison, the persistence of the activated state of pea thylakoid LHCII kinase was also measured. The LHCII kinase of isolated pea thylakoids was found in the deactivated or basal state. Following an initial exposure to the light (30 s) to activate the enzyme, the thylakoids were further stored in ice, in the dark, and at times as indicated, LHCII phosphorylation was assayed in the dark or light. The results ( Fig. 1) show that in the dark-stored pea thylakoids, the LHCII kinase-activated state deactivates rapidly ( tlA, 15 min) as compared with the kinase of Acetabularia thylakoids (tlA, >2 h).
Measurements of fluorescence induction kinetics using either pea (32) or Acetabularia thylakoids (data not shown) demonstrate that following light-driven reduction, the plastoquinone pool is reoxidized in the dark within less than 2 min. Thus, the persisting activated state of the kinase in both organisms for times longer than 15 min ( Fig. 1) cannot be due simply to the plastoquinol/plastoquinone ratio of the plastoquinone pool.
Deactivation of Acetabularia LHCII Kinase by Quinone Analogs-The nature of the kinase activator, plastoquinol or a reduced component of the cytochrome b6/f complex, can be disclosed by measuring the redox midpoint potential of the activation process. The midpoint potential for the activation/ deactivation of the Acetabularia enzyme was measured. The results of these experiments are shown in Fig. 2. The midpoint potential in the dark is c, , 120 f 10 mV, at pH 8.0. However, it was not possible to determine unequivocally whether n = 2 or 1. This em value is closer to the reported midpoint potential of plastiquinol (c, 80-100 mV, pH 7.5 (33)(34)(35)) than to that of the bl, and bH components of bs ( e , -146 and -50 mV, pH 7.0, respectively) so far reported (36,37). Thus, a reduced quinol binding site seems to be involved in the Acetabularia LHCII kinase activation as well, Information on the properties of the binding site involved in the activation/deactivation process could be obtained by assaying the effect of various substituted quinone analogs on the enzyme deactivation in the dark. Under these conditions the quinone analogs may not compete for the binding site with the plastoquinol pool reduced via light-driven photosystem I1 electron flow. Furthermore, the possible inhibitory effect of the analogs on the reduction of plastoquinone by PSII may not interfere with the effect of the analogs on the binding site.
The extent of deactivation of the Acetabularia LHCII kinase in the dark or light by substituted benzoquinones, naphthoquinones, and anthraquinones is shown in Table I. The p150 values for this deactivation are given in Table I1 together Tables I and 111, only the percent inactivation at a concentration of 0.5 FM was tested for all the analogs, since at this concentration, one could compare the effect of all the compounds tested both in the dark and in the light. Although this approach does not resolve in detail differences between the analogs causing extensive deactivation (285%) at the concentration used, it allows a comparison of the effects of those of lower potency.
The kinase was deactivated in the dark (in absence of added activators) by DBBB or DIBB with a value of 0.1 U M Potentiometric titration of LHCII phosphorylation in Acetabularia thylakoids. Acetabularia thylakoids a t a final concentration equivalent to 50 pg of chlorophyll. ml" were incubated under a N, stream in a reaction mixture containing the redox buffers as indicated under "Materials and Methods." Following adjustment of the redox potential as described, the phosphorylation reaction was initiated by addition of [Y-~'P]ATP and terminated after 10 min by addition of cold trichloroacetic acid (10% w/v). The redox potential was maintained during the phosphorylation reaction by addition of potassium ferricyanide or sodium dithionite. The degree of LHCII phosphorylation was estimated as in Fig. 1. The 100% value was 0.7 nmol 32P. mg" incorporated into the 27-kDa LHCII polypeptide. The curve shown was fitted by a computer program. The standard error was k9.37 mV.    (Fig. 3). When assayed in the light or in the dark with addition of duroquinol, the pI50 values were 1 and 3 PM, respectively, as expected, since the inhibitor competes with reduced plastoquinol for the binding site. The effect of the quinone analogs on the activity of photosystem I1 by measurements of DCIP reduction with water as an electron donor and the inhibition of kinase activity in the light were measured as well ( Table  I).

50-
The results presented in Fig. 4 and Tables I and I1 clearly demonstrate that the inhibitory effect of the quinone analogs is drastically reduced with the increase in the ring size. The pI60 values were 0.05, 1.0, and 3 PM for benzoquinone, napthoquinone, and anthraquinone, respectively. The lower activity of the trichloroanthraquinone is ascribed to the ring size rather than to the presence of an additional chlorine at position 4 as compared with benzoquinone and napthoquinone derivatives, since no significant difference was found between the activity of di-, tri-and tetrahalogenated benzoquinones (Table I). However, even the halogenated-substituted anthraquinone ring is still a potent inhibitor at 0.5 p~ concentration. The kinase deactivation is not related to the quinone binding sites associated with photosystem I1 (Table 11). Furthermore, the pIs0 values for the inactivation of electron flow from duroquinol to methyl viologen are similar to those of the LHCII kinase deactivation, indicating that these inhibitors interact with the cytochrome b6/f complex (Table 11).
The high inhibitory activity in the dark of some benzoquinone analogs was significantly diminished when the assay was carried out in the light. This was the case mostly for those compounds that had only a small inhibitory effect on the activity of photosystem 11, such as that of unsubstituted benzoquinone or of the tribromomethyl benzoquinone. This could be due to competition for the binding site with plastoquinol generated by electron flow via photosystem 11. Compounds such as 2-bromo-5-t-butyl-benzoquinone and DBMIB significantly inhibit the activity of photosystem I1 but only partially deactivate the kinase in the light (Table I). This may indicate that the relatively low plastoquinol concentration formed in the presence of these compounds in the light compete better with these analogs for the kinase activation site than with other analogs such as 2,3-dichloro-5-t-butylbenzoquinone. The latter deactivates the kinase in the light despite a higher rate of photosystem I1 activity, which could generate a higher plastoquinol/plastoquinone ratio in the thylakoid membranes.
Substituted quinone analogs may be reduced by light-incubated thylakoids (38, 39). The possibility that the reduced analogs may serve partially as activators could explain their lower inhibitory effect in the light. It is possible that such an activating effect may occur at the lower concentrations of the inhibitors as observed in Figs. 3 and 4.
The nonsubstituted naphthoquinone ring had no effect on the Acetabularia kinase activity in the dark or in the light. However, it partially inhibited DCIP reduction. The effect of the naphthoquinone analogs on kinase deactivation in the dark is drastically enhanced by halogen substitution a t position 2-and a further increase in the effect is obtained by addition of a second halogen at position 3- (Table I, compare 2-chloro-with 2,3-dichloro-, 2,3-dibromo, or 2,3-diiodonaphthoquinones). The effect of all these compounds is significantly lower in the light as expected, since photosystem I1 is only partially inhibited, and thus, the level of reduced plastoquinone is sufficient to compete with the inhibitors for the binding site.
It is also noteworthy that anthraquinones have a significant inhibitory effect on DCIP reduction and only a limited deactivation of the kinase in the dark (13-32%), yet they exhibit a higher inhibitory activity of the kinase in the light (50-66%, Table I).
Deactivation of Pea Thylakoids LHCII Kinase by Quinone Analogs-The activated state of pea LHCII kinase is less stable than that of Acetabularia (cf. Fig. l), and thus, measurements of kinase activity in the dark in the presence of the various quinone analogs is technically more difficult. Nevertheless, it was possible to demonstrate a similar behavior toward substituted quinone analogs for the pea LHCII kinase in the dark (Table 111). The effect of the substituted benzoquinone and especially naphthoquinone halogenated a t position 2,3-is higher as compared with the nonhalogenated  compounds and seems to be specific for the deactivation of the kinase. This conclusion is supported by both the difference in the effect of the analogs on DCIP reduction (Table 111) and the respective p150 values for this reaction (Table 11). Increasing the size of the analog molecule by addition of a third ring as in the case of anthraquinone further reduces the deactivat-  Presence of Kinase Activity in Cytochrome bs/f Complex Preparations of Acetabularia Thylakoids-The question arose as to whether the Acetabularia kinase activity also copurifies with the cytochrome bs/f complex. Such a preparation could indeed be isolated from thylakoids obtained from about 50,000 Acetabularia cells. A 7-fold enrichment in the cytochrome complex and 8-fold in that of the kinase activity relative to the content of the thylakoids before extraction was obtained (Table IV),

Effect of Quinone Analogs on Deactivation of LHCII Kinase
The electrophoretic polypeptide pattern of the cytochrome bs/f-enriched complex showed the presence of all four components (Fig. 5B) as identified by immunoblotting and the presence of both cytochromes b, and f as identified by heme staining (data not shown). The isolated fraction contained, as well, a polypeptide band of about 64 kDa (Fig. 5A ) corresponding to the reported kinase band isolated from spinach (42), which reacted with the anti-kinase antibodies raised against the spinach 64-kDa polypeptide.

DISCUSSION
The Nature of the Kinase Activator Molecule-The thylakoid-bound LHCII kinase of both Acetabularia and pea, once   Table   I. The pea-isolated thylakoids were preactivated by 30-s illumination before assay in the dark as described in Fig. 1, and the respective 100% DhosDhorvlation values were as in Fig. 1.  75  7  0  95  73  70  96  85  77  95   96   93  74  5  0  83  40  0  83  3  65  15  37  95  84  70   30  9  26  46  45  9  2  18  9  88  59  33  88  35  22  84  80  57   86  98  86  79  88  79  79  69  79 activated, can remain in this state for prolonged periods of time ( t M 2 15 min), while the concentration of the putative activator (ie. plastoquinol) diminishes below its activating level and that of the putative inhibitor (plastoquinone) rises to the inhibitory level much faster. It was previously suggested that a component of the cytochrome b,/f complex that remains in the reduced state may be responsible for maintaining the A c e~a b~~r i a kinase in the active state for prolonged periods of time (9). The results of the potentiometric titration of the Acetabularia kinase activity in the dark (~~H s . 0 , 120 & 10 mV) exclude this possibility. The kinase in Acetabularia copurifies with the cytochrome bG/f complex as previously reported for higher plants (10,43) and green algae (3). Thus, it is possible that in both Acetabularia and pea LHCII, phosphorylation may be due to the plastoquinol-mediated interaction of the cytochrome complex with the kinase resulting in an increase in the kinase affinity for the LHCII substrate as previously demonstrated for the spinach enzyme (10). However, the kinetics of the activation and particularly of the deactivation processes cannot simply be explained by the changes in the plastoquinol/plastoquinone ratio.
The Properties of the Quinone Binding Site-The degree of kinase deactivation by various quinone analogs suggests that the binding site responsible for the deactivation process preferentially accommodates a benzoquinone ring halogenated at position 2,3-similar to the case of the quinone binding site of PSI1 (38,39). Increasing the size of the analog by addition of side chains or of a second or third benzene ring, or increasing the bulk of the alkyl side chain, considerably reduces the inhibition efficiency. The interpretation of these results should take into consideration that the relative concentrations of the various analogs in the thylakoid membrane are significantly influenced by the partition coefficient of the analogs between the water and the membrane lipid phase and the polarity ascribed by the substitution of the benzene ring by halogen residues. The lipophilicity of the analogs increases with the increase in the number of benzene rings and/or size of the alkyl side chain (38), while the inhibitory activity decreases. This supports the conclusion that the binding site has a higher affinity for the bishalogenated benzoquinone component. Double halogenation increases the effectiveness of the benzoquinone analog, while addition of a third or fourth halogen does not significantly change the inhibitory effect. This indicates that the orientation of the analog at the binding site is such as not to be hindered by halogenation at position 5,6-. This conclusion is further suggested by the fact that addition of t-butyl or isopropyl residues at positions 5-and 6of the benzoquinone ring, respectively, does not reduce the effectiveness of the inhibitor in deactivating the LHCII kinase in the dark. However, the addition of the second ring to benzoquinone generating the naphthoquinone ring while blocking carbons 5 and 6 of the benzoquinone ring, as well as of the third ring blocking carbons 2 and 3 of the benzoquinone 2.40 4.8 "The ratio is given as nmol of cytochrome flkinase activity X The yield of both cytochrome and kinase is in the range of 2% as compared with 20-30% reported for the purification of similar preparations from spinach (10). Due to the low yield, the last step in the purification consisting of sucrose density centrifugation could not be performed.
The solubilized membrane fraction is the supernatant after membrane solubilization in the detergent mixture and centrifugation (22). to form the anthraquinone, significantly diminished the inhibitory effect of the analog. The most effective analogs inhibiting the LHCII kinase are those that bind to the Q, site of the cytochrome b6/f complex such as DBMIB, DIBB, and DBBB (38, 44), suggesting that this complex is one of the components involved in the activation/deactivation processes.
Some of the quinone analogs used in this work were reported to bind to nucleophilic groups (38). Since it was demonstrated that the LHCII kinase can be inactivated by alkylating reagents (21) one could expect that at least part of the inhibitory effects observed in this work may be due to their interaction with the kinase " S H groups. If this were the case, one would expect that the inhibition would be practically irreversible. To test this possibility, the effect of DBMIB and tetrabromo-1,4-benzoquinone (Bromanil) known to interact with " S H groups was tested and found to be reversible (data not shown). Similar results were also found by Coughlan et al. (40) for the inhibitory activity of DBMIB on thylakoidbound LHCII kinase from spinach.
It was previously reported that reduced DBMIB does not bind to the cytochrome b6/f complex (45). Since DBMIB can be reduced by electron flow from PSII, one may consider that DBMIB may interact with the kinase proper. Assay of the pIm of the kinase deactivation by DBMIB-H2 was similar to that of oxidized DBMIB (data not shown). This may indicate that the LHCII kinase had a distinct quinone binding site as previously suggested (11). It was reported that the isolated 64-kDa LHCII kinase appears to be a hydrophilic protein despite its strong binding to the thylakoid membrane (40). Thus, the presence of a quinone binding site in the kinase deserves further investigation.
Factors Affecting the Rate of Kinase ActiuationlDeactiuation-The slow deactivation process of the kinase in darkincubated thylakoids contrasts with the relatively fast activation by addition of plastoquinol or by light-driven electron flow. However, even the activation process is slower than that of the plastoquinone pool reduction in light-exposed thylakoids, which is achieved within seconds when isolated thylakoids are illuminated under conditions inducing LHCII kinase activation. The reoxidation of the pool in the dark occurs within 1-2 min. Thus, the kinetics of kinase activation and deactivation do not correlate in a simple way with the changes in the plastoquinol/plastoquinone ratio. The activation time of the kinase in the light was reported to be about 2-5 min as measured by LHCII phosphorylation in pea or Acetabularia (9,30,31). ATP-induced fluorescence quenching and state transition resulting from the phosphorylation of LHCII can be detected within 5-10 min (15). However, in this case the measured time includes the process of LHCII dissociation from PSII and its lateral diffusion within the membrane plane. Phosphorylation of LHCII can be activated at 0 "C (47), while the dissociation of LHCII and the associated loss of energy transfer to the reaction center are not affected at this temperature. Since binding of a quinone ligand to a specific binding site occurs on a time scale of seconds (38,39, 44,45), one must conclude that the activation of the kinase by plastoquinol or its deactivation by plastoquinone is mediated by an additional slower step(s) besides the quinone binding per se.
Two possible models could be considered to explain this aspect of the kinase activation/deactivation process. In both, the cytochrome-kinase interaction may be the slow step. In both models, cytochrome b6/f is reduced by plastoquinol (fast phase) and binds to the kinase, which becomes activated (slow phase). The cytochrome-kinase complex remains associated and active until the cytochrome is oxidized and dissociates from the kinase (slow phase) or until a quinone binding site on the cytochrome complex, the kinase proper, or both, is occupied by plastoquinone or an appropriate analog causing the kinase deactivation via dissociation of the cytochrome be/ f complex from the kinase (slow phase). Occupancy of the Qz site of the cytochrome complex by quinone analogs preventing its reduction by plastoquinol, or occupancy of the putative quinone binding site on the kinase by such analogs, may prevent this association or accelerate the dissociation process.
The possibility that self-phosphorylation of the kinase and phosphorylation of the cytochrome b. 5 within the complex may be part of the activation/deactivation processes should also be considered as suggested by recent results demonstrating that both the kinase and the cytochrome b.5 are phosphorylated when the isolated cytochrome b6/f-kinase complex is incubated with ATP (48).
An additional factor controlling the rate of these two processes may be the structural organization of the thylakoid. We have demonstrated before that LHCII kinase is specifically localized at the edges of the grana stacks (10). It was reported before that a fraction of the cytochrome b6/f complex is localized at the region between the appressed and nonappressed thylakoid membranes (48). Recent results demonstrate the coexistence of the LHCII kinase in such a fraction possibly associated with some of the cytochrome b6/f complex.* The possibility of the transient existence of "supercomplexes" of LHCII including PSII and cytochrome b6/f complex interconnected by confined plastoquinone molecules has been proposed as a control mechanism of the linear electron flow from PSII (49), while specific diffusion of cytochrome b,/f complex from the grana1 to the stromal domain during state transition has been implied to be related to the state transition and activation of the photosystem I cyclic electron flow (50). It is thus possible that slow, diffusion-controlled, structural interactions between the cytochrome bs/f complex and the kinase leading to the formation of an active supercomplex may control in a similar way the processes of activation/ deactivation of the LHCII kinase. The