Changes in the transthylakoid proton gradient are caused by the movement of phycobilisomes in the cyanobacterium Synechocystis sp. strain PCC 6803

Phycobilisomes (PBSs) are the main accessory light-harvesting complexes in cyanobacteria and their movement between photosystems (PSs) affects cyclic and respiratory electron transport. However, it remains unclear whether the movement of PBSs between PSs also affects the transthylakoid proton gradient (ΔpH). We investigated the effect of PBS movement on ΔpH levels in a unicellular cyanobacterium Synechocystis sp. strain PCC 6803, using glycinebetaine to immobilize and couple PBSs to photosystem II (PSII) or photosystem I (PSI) by applying under far-red or green light, respectively. The immobilization of PBSs at PSII inhibited decreases in ΔpH, as reflected by the slow phase of millisecond-delayed light emission (ms-DLE) that occurs during the movement of PBSs from PSII to PSI. By contrast, the immobilization of PBSs at PSI inhibited the increase in ΔpH that occurs when PBSs move from PSI to PSII. Comparison of the changes in ΔpH and electron transport caused by the movement of PBSs between PSs indicated that the changes in ΔpH were most likely caused by respiratory electron transport. This will further improve our understanding of the physiological role of PBS movement in cyanobacteria.

A balanced distribution of the excitation energy absorbed by light-harvesting complexes to the complexes of photosystems I and II (PSI and II) is considered to be one of the most important factors for ensuring optimal photosynthetic efficiency [1][2][3]. In response to fluctuating light conditions, the photosynthetic machinery regulates the distribution of excitation energy between the 2 photosystems (PSs) [4]. This dynamic and rapid process of achieving energy balance is called "state transition" [5,6]. Because phycobilisome (PBS) movement is a prerequisite for cyanobacterial state transitions [7,8], "mobile PBSs" are believed to play a key role in allowing state transitions in cyanobacteria. A recent study indicated that the movement of PBSs between PSII and PSI affects both cyclic and respiratory electron transport [9]. However, it remains unclear whether the movement of *Corresponding author (email: wma@shnu.edu.cn) PBSs between PSs also affects transthylakoid proton gradient (ΔpH) levels in cyanobacterial cells.
The aim of this study was to investigate the effect of PBS movement between the 2 PSs on ΔpH levels in cyanobacteria. It is difficult to probe ΔpH levels in vivo as the system relaxes rapidly [10,11]. Millisecond-delayed light emission (ms-DLE) originates from the reverse reaction of the photoact in PSII, and is composed of one fast (within 0.1 s of the onset of measuring flashes) and one slow phase [12,13]. Crofts and his colleagues concluded that the intensities of the fast and slow phases of ms-DLE are stimulated by the membrane potential (ΔE) and the ΔpH, respectively [14][15][16]. Further characteristics and properties of ms-DLE were studied and confirmed in subsequent research [17][18][19][20][21]. It was found that ΔpH levels can be probed in vivo by measuring the slow phase of ms-DLE.
The movement of PBSs is inhibited by glycinebetaine (GB) in the cyanobacterium Synechocystis sp. strain PCC 6803 (hereafter Synechocystis 6803) regardless of changes in spectral quality and light intensity [9]. We applied GB to cells to immobilize and couple PBSs to PSII or PSI under light which preferentially excites PSI or PSII, respectively. Comparison of the ΔpH levels in untreated and GB-treated cells enabled us to study the physiological role of PBS movement in cyanobacteria.

Immobilization of PBSs
Cells cultured for 4 d (A 730 = 0.6-0.8) that showed the highest photosynthetic activity [23] were harvested by centrifugation (5000×g for 5 min at 25°C) and suspended in fresh BG-11 medium buffered with Tris-HCl (5 mmol/L, pH 8.0) at a chlorophyll a concentration of 20 μg mL -1 . After exposure to far-red light (FR, 5.2 μE m -2 s -1 ; Ditric Optics 705 nm long pass filter) or green light (GL, 15 μE m -2 s -1 ; Ditric Optics 520 nm long pass and 546 nm short pass filter) for 20 min, the cells were treated with 1 mol/L GB under the same FR or GL illumination for 20 min to immobilize and couple PBSs to the stromal surface of the PSII or PSI core complex in the thylakoid membrane.

Measurements of chlorophyll fluorescence yield
The yield of chlorophyll fluorescence was monitored at several indicated time points with a pulse-amplitudemodulated (PAM) chlorophyll fluorometer (Walz, Effeltrich, Germany) and an emitter-detector-cuvette assembly (ED-101US) with a 101ED unit, as described elsewhere [9,24]. Cells were subjected to darkness for 2 min before measurements were made. The intensity of the modulated measurement light was lower than 0.2 μE m -2 s -1 to avoid photosynthetic electron transport. Maximum fluorescence levels (F m or F m ′) were determined using saturating light pulses (3200 μE m -2 s -1 , 600 ms duration), which were produced by a KL-1500 lamp (Schott, Germany). The intensities of FR and GL were 5.2 (>705 nm, through a home-made filter) and 15 μE m -2 s -1 (~540 nm, through another home-made filter), respectively.

Measurements of ms-DLE
The ms-DLE signal was measured using a lab-made phosphoroscope. A sample, in a polymethylmethacrylate cuvette, was irradiated with light passing through a 2-cm thick layer. The holes on the rotating wheels were arranged such that the measuring process might be divided into a series of 5.6 ms cycles for the excitation measurement, with 1-ms excitation with light of 1500 μE m -2 s -1 followed by 4.6-ms darkness. The delayed light observed between 2.8 and 3.8 ms after each flash was measured with an EMI9558B photomultiplier with a red glass filter. The signal passing through an amplifier was recorded continuously with a SC-16 light beam oscillograph [17].

The slow phase of ms-DLE reflects in vivo ΔpH
It has been previously reported that the intensity of fast and slow phases of ms-DLE closely correlates with ΔE and ΔpH levels, respectively [14][15][16]. To confirm the relationship between ΔpH and the slow phase of ms-DLE, an uncoupler, nigericin (Nig), which decreases ΔpH by stimulating exchanges of H + and K + between the two sides of the thylakoid, was used. As shown in Figure 1, Nig treatment resulted in a considerable decrease in the slow phase of ms-DLE but the same effect was not observed for the fast phase, indicating that the slow phase of ms-DLE indeed reflects changes of ΔpH in vivo in the unicellular cyanobacterium Synechocystis 6803.

Light-induced state transitions are dependent on PBS movement
The small molecule, GB, has previously been used to lock PBSs into one position, rendering them unable to diffuse through the thylakoid membrane surface [9,25,26]. In this study, we used 1 mol/L GB to immobilize and couple PBSs to PSII or PSI in Synechocystis 6803. Figures 2 and 3 show the changes in chlorophyll fluorescence yield at room temperature in untreated and GB-treated cells of Synechocystis Figure 3 Effect of switching from GL to FR on state transition and ms-DLE in untreated and GB-treated cells. PAM fluorescence kinetic traces of untreated cells (a) and GB-treated cells in which PBSs were immobilized and coupled to PSI (c) under GL. Sample in (a) was first illuminated with GL (arrow 1), which was switched to FR as indicated by arrow 2. During the switch from GL to FR, the ms-DLE was determined in untreated (b) and GB-treated cells (d) at several indicated time points. The changes in the slow phase intensity of ms-DLE were then analyzed (e) and standard errors were calculated from at least six independent experiments. 6803 upon switching from FR (5.2 μE m -2 s -1 , >705 nm) to GL (15 μE m -2 s -1 , 540 nm; Figure 2(a) and (c)) or from GL to FR (Figure 3(a) and (c)). The transition from state 1 (2) to state 2 (1) was clearly observed in untreated cells by switching from FR (GL) to GL (FR), and the process was complete within 1-2 min (Figures 2(a) and 3(a)). The transition was strongly inhibited when the PBSs were immobilized by GB (Figures 2(c) and 3(c)). These results clearly show that light-induced state transitions depend predominantly on the movement of PBSs [7,25].

Changes in the slow phase of ms-DLE are caused by the movement of PBSs
To investigate the effect of PBS movement between PSs on ΔpH levels, the intensity of ms-DLE was compared between untreated and GB-treated cells of Synechocystis 6803. Switching from FR to GL considerably decreased the intensity of the slow phase of ms-DLE in untreated cells but not in GB-treated cells (Figure 2(b), (d) and (e)). By contrast, switching from GL to FR increased the amplitude of the slow phase of ms-DLE in untreated cells but only slightly in GB-treated cells (Figure 3(b), (d) and (e)). Further, the movement of PBSs between PSs did not significantly influence the fast phase of ms-DLE (Figures 2 and 3). Therefore, the movement of PBSs between PSs affects the slow phase of ms-DLE but has no effect on the fast phase. Taking these combined results together, we conclude that the intensity of the slow phase of ms-DLE is high when PBSs are associated with PSII, but low when PBSs are associated with PSI.

Discussion
It has been shown previously that the speed of PBS movement depends on the intensity of the inducing light [27]. Under 5.2 μE m -2 s -1 FR and 15 μE m -2 s -1 GL, the movement of PBSs occurred rapidly and was complete within 1-2 min (Figures 2 and 3). No significant changes in linear electron transport were observed during this period [9]. However, the activity of respiration was high when cells were in state 1 (PBSs associated with PSII) and low when cells were in state 2 (PBSs associated with PSI). By contrast, the activity of cyclic electron transport was high when PBSs were coupled to PSI and low when coupled to PSII [9]. The results of this study indicated that ΔpH levels, as reflected by the intensity of the slow phase of ms-DLE, were high when PBSs were associated with PSII but low when PBSs were associated with PSI (Figures 2 and 3), which is consistent with the changes observed in respiration, and not cyclic, electron transport caused by the movement of PBSs [9]. These findings suggest that changes in ΔpH, caused by the movement of PBSs between PSs, most likely depend on the rate of respiratory electron transport.
It has been reported previously that NADPH dehydrogenase (NDH-1)-mediated respiration is absent in ndh gene inactivation mutants ΔndhB (M55) and ΔndhD1/D2 (D1/D2) [28][29][30]. However, NDH-1-dependent cyclic electron transport was enhanced in D1/D2 cells and reduced in M55 cells, when compared to the wild-type Synechocystis 6803 (WT) [28][29][30]. Under low photon flux densities (5-6 μE m -2 s -1 ), ΔpH levels, as determined by the intensity of the slow phase of ms-DLE, declined considerably in both M55 and D1/D2 cells, when compared to those of WT cells (data not shown). Under these light conditions, linear electron transport significantly declined, and was therefore not a major contributor for the establishment of ΔpH. The NDH-1-dependent respiratory electron transport was therefore found to be responsible for the establishment of ΔpH, which is consistent with previous observations [31]. This result supports the contribution of respiration to the changes in ΔpH during the movement of PBSs between the two PSs.
In the cyanobacterium Synechocystis 6803, the respiratory electron transport that is mainly mediated by NDH-1 plays an important role in donating electrons to the plastoquinone (PQ) pool, thereby establishing the ΔpH [31,32]. It therefore seems likely that high respiration rates induce a rise in H + concentration on the lumen side, resulting in a high ΔpH level, and vice versa.
In higher plants, ΔpH levels, as reflected by ms-DLE measurements, show biphasic behavior during the transition from state 1 to state 2, with the ΔpH transiently increasing to its maximal level within 0.5 min, and then gradually decreasing from the maximal value [33,34]. However, in cyanobacteria, biphasic behavior was not observed during the movement of PBSs between PSII and PSI (Figures 2 and  3). This may be attributed to the different mechanisms of state transition in cyanobacteria and higher plants.
Under FR-and GL-preillumination, the addition of 1 mol/L GB to the cultures reduced markedly the intensity of the slow phase of ms-DLE (Figures 2 and 3). Similarly, treatments with 1 mol L -1 phosphate, sucrose and potassium chloride solutions, which are also inhibitors of PBS movement [7,35], also significantly reduced the intensity of the slow phase of ms-DLE (data not shown). Taken together, it appears that the movement of PBSs on the surface of the thylakoid membrane is responsible for the effective establishment of ΔpH in cyanobacteria.
In conclusion, our results, which were obtained with a unicellular cyanobacterium, increase our knowledge of the physiological role of PBS movement, indicating that ΔpH levels are high when PBS associates with PSII but low when PBS associates with PSI. The changes in ΔpH caused by the movement of PBSs between PSs most likely depend on the rate of respiratory electron transport. While the working mechanism underlying such changes remains uncertain, our findings will help in understanding the physiological role of PBS movement in cyanobacteria.