Two contributions to fluorescence lowering have been identified,- qZ, associated with reactions of the zeaxanthin cycle, activated by a low lumenal pH [1-4], and a second process, qE, dependent on the proton gradient but of unknown mechanism [5,6]. In intact plants, the contribution due to a change in zeaxanthin was sensitive to DTT, caused a decrease in F0, showed a slower kinetics of onset, and a higher light intensity requirement, allowing this component to be distinguished from qE. More recently, Horton and colleagues  have shown that the qE component in chloroplasts is dependent on the preillumination history; there was a correlation between increased qE and high levels of zeaxanthin, suggesting a potentiating effect of zeaxanthin. In their early studies, Wraight and Crofts [6,8,9] noted that the conditions giving rise to qE quenching were the same as those which led to enhanced delayed fluorescence (DF) in the ms time range. The latter effect was attributed to stabilization of the DF substrate P+QA- by pH through effects on the equilibria of secondary donors and acceptors with H+ in the aqueous phases on opposite sides of the membrane. Subsequent work showed the 1,0,1,2 pattern of proton release associated with the S-state transitions (starting at S0), and a matching period 4 pattern in the amplitude of DF . Since P+ is a quencher of fluorescence , the following working hypothesis provides the simplest explanation of fluorescence quenching (qE) associated with pH. As the lumenal pH drops on illumination, the reactions of the S-states are progressively inhibited in transitions associated with H+-release, especially S3-S0 [10,11]. These conditions would lead to accumulation of P+ which quenches fluorescence, allowing the plant to dump excess absorbed energy to the environment as heat, and to enhanced DF [8,9], and a lower photochemical yield. This minimal hypothesis has the added attraction of a natural connection to the irreversible quenching, qI, associated with photoinhibition. Under conditions in which P+ accumulates, secondary quenching species (Chl+ and Car+) are formed close to the reaction center [13,14], and it seems likely that one or both of these species are responsible for qI quenching.
In this paper, we test this minimal hypothesis, and examine the inhibition of qE quenching by antimycin. We show that antimycin sensitive quenching occurs in the absence of high potential oxidizing components on the donor side of PS II, but that the quenching is more extensive, and shows a more rapid onset, when the S-state intermediates are poised in transitions above the S1-state. We conclude that qE quenching involves a mechanism independent of the S-state transitions.
Fluorescence yield was measured using flashes of <0.2% actinic yield. The fluorimeter was equipped with flash (5 µs, 90% saturating) and continuous actinic (10 ms shutter) lamps. By AC coupling the detector, yield could be measured without interference from the continuous actinic source.
Following the protocol of Wraight and Crofts , we have studied the dependence of qE on the state of the oxygen evolving complex (OEC). Chloroplasts were clocked into either the S1 or the S3 state by illumination with 0 or 2 flashes after 10 min dark adaptation, and then inhibited with DCMU. A pH was generated by continuous illumination at actinic intensity, driving turn-over of PS I using DAD and methyl viologen as substrates. Illumination also activated a single turn over of PS II, so that the reactions were constrained to the transition:
Sn.Z.P.QA.DCMU ---> Sn.Z.P+.QA-.DCMU === Sn+1.Z.P.QA-.DCMU + mH+
Fig. 1. Fluorescence lowering with intact OEC. Left: Induction of qE; open squares, initially in S1; open triangles, initially in S3. Right: Decay of FV; closed squares, initially in S1; closed triangles, initially in S3. Chloroplasts (5 µg/ml chlorophyll (Chl)) were dark-adapted for 10 min, treated with 0 (S1) or 2 flashes (S3). DCMU (50 µM), diaminodurene (DAD) (1.5 mM), and methylviologen (MV) (150 µM) were added, and 20 s later, the shutter opened for 6 s (kinetics on left), then closed (right).
There was a marked fluorescence lowering on generation of pH when chloroplasts were trapped in either the S3-S0 transition or the S1-S2 transition (Fig. 1, left); the quenching in the S3-S0 transition was usually more marked, but this was a small and somewhat variable effect. The back-reaction of QA- (measured from the decline in FV) showed a half-time (1.5-2 s) (Fig.1, right) similar to that observed when the pH and qE quenching were inhibited by nigericin (Fig.2). These results are in part consistent with the above hypothesis, but there are some anomalous features. For a simple equilibrium, it would be expected from the reaction above that [P+], and hence the back-reaction rate, would vary strongly with lumenal pH, but this was not found. Some effect might also have been expected due to variation in the H+ stoichiometry (m in the equation) with S-state [8,11], but no obvious effect was apparent. Even when account is taken of the non-integer stoichiometry, the effect of pH on the proton yield pattern , and the relaxation of quenching which competes with the FV decline in the absence of nigericin, the lack of any clear S-state dependence is unexpected.
Fig. 2. Effect of nigericin on qE induction and back-reaction. Conditions as Fig. 1. Chloroplasts initially in S1. Open circles (left) and closed tringle (right); 5 µM nigericin present.
When chloroplasts were preincubated for 10 min with 50 µM NH2OH, sufficient to convert the OEC to the S-1-state, a very different behavior was seen. The chloroplasts showed a significant fluorescence lowering, although at a slower rate than when the S-states were poised in higher transitions
Fig. 3. Fluorescence lowering with disfunctional OEC. Chloroplasts (5 µg/ml Chl) were dark-adapted for 10 min in the presence of 50 µM NH2OH, then DCMU, DAD and MV were added, and qE and FV decay measured as in Fig. 1. Insert: Tris-washed chloroplasts were used, with no preincubation with NH2OH.
(Fig. 3, left). On turning off the actinic light, the fluorescence returned almost to the Fmax level rather than declined, indicating that no high potential component able to act as an acceptor in the backreaction from QA- was present on the donor side (Fig. 3, right). Presumably the single turn-over of PS II was unable to generate a high potential intermediate in the OEC. Similar results were obtained when Tris-washed chloroplasts were used, but with DCMU, DAD and methylviologen as the only additions (Fig. 2, insert). In this case, the oxidized donor side components (predominantly Z+) generated in the single turn over were presumably reduced by DAD. In all cases, the fluorescence lowering was inhibited by 20 µM antimycin (Fig. 4, A) , or by 5 µM nigericin (Fig. 2). Fluorescence lowering was almost completely abolished if chloroplasts were pre-treated with 10 mM NH2OH for 2 min before addition of DCMU (not shown). However, under these circumstances, the rise in FV on closing the shutter was much more rapid than with the other treatments, suggesting that most of the inhibition could be accounted for by a partial uncoupling of pH.
The results show that the generation of high potential intermediates in the OEC is not a necessary condition for qE quenching. The minimal hypothesis outlined above does not explain the quenching when the OEC is functionally or structurally deranged.
The inhibition of qE quenching by antimycin occurs at relatively low concentration (5-20 µM range) . In order to explain the inhibition in terms of the minimal hypothesis above, it would be necessary to show an effect on the S-state transitions or other donor side components in this concentration range. We have studied the effects of antimycin on the backreaction QA-YZ+ to QAYZ in DCMU-inhibited Tris-washed chloroplasts. Our results from fluorescence and EPR studies show that in the presence of antimycin, electron donation to YZ+ is accelerated (Fig. 5), suggesting an ADRY-like effect for antimycin. This effect could also be seen in preparations with an intact OEC. When antimycin was added in experiments like those of Fig. 1, at concentrations which inhibited qE, the decline in FV on closing the shutter was truncated, leaving a substantial fraction of centers (>50%) in the QA- state (Fig. 4, B). Addition of FCCP had a similar effect at much lower concentrations. Ghanotakis, Yerkes and colleagues (17,18) have suggested that ADRY reagents act by reducing oxidized components on the donor-side of PS II, suggesting a mechanism for inhibition in line with the above hypothesis. The ADRY behavior of antimycin does not explain the inhibition of qE quenching when the donor-side reactants are reduced.
Fig. 4. Inhibition of qE quenching by antimycin. Conditions as in Fig. 1, chlorplasts initially in S1. Left, open triangle, and right, closed triangles; 20 µM antimycin A present.
Fig. 5. ADRY-like effect of antimycin in Tris-washed chloroplasts . Left: EPR kinetic traces (average of 200 at 0.1 Hz, time full-scale 1 s) showing Z+. Chloroplasts (2 mg/ml Chl) suspended at pH 7.5 with 100 µM DCMU, 5 mM ferricyanide. Where shown, 200 µM antimycin was added. Right: Chloroplasts (5 µg/ml Chl) at pH 7.5, with no additions (circles at bottom), with 10 µM DCMU (diamonds), or with DCMU and 10 µM antimycin (triangles, circles, top), were illuminated with 1 or 2 (circles, top) flashes.