Oxidation of P700 Induces Alternative Electron Flow in Photosystem I in Wheat Leaves

Oxygen (O2)-evolving photosynthetic organisms oxidize the reaction center chlorophyll, P700, in photosystem I (PSI) to suppress the production of reactive oxygen species. The oxidation of P700 is accompanied by alternative electron flow in PSI (AEF-I), which is not required for photosynthetic linear electron flow (LEF). To characterize AEF-I, we compared the redox reactions of P700 and ferredoxin (Fd) during the induction of carbon dioxide (CO2) assimilation in wheat leaves, using dark-interval relaxation kinetics analysis. Switching on an actinic light (1000 μmol photons m−2 s−1) at ambient CO2 partial pressure of 40 Pa and ambient O2 partial pressure of 21 kPa gradually oxidized P700 (P700+) and enhanced the reduction rate of P700+ (vP700) and oxidation rate of reduced Fd (vFd). The vFd showed a positive linear relationship with an apparent photosynthetic quantum yield of PSII (Y[II]) originating at point zero; the redox turnover of Fd is regulated by LEF via CO2 assimilation and photorespiration. The vP700 also showed a positive linear relationship with Y(II), but the intercept was positive, not zero. That is, the electron flux in PSI included the electron flux in AEF-I in addition to that in LEF. This indicates that the oxidation of P700 induces AEF-I. We propose a possible mechanism underlying AEF-I and its physiological role in the mitigation of oxidative damage.


Introduction
Suppression of the carbon dioxide (CO 2 ) assimilation efficiency of photosynthesis under conditions of high light intensity, low/high temperature, or drought decreases the regeneration efficiency of the electron acceptor (NADP + ) in photosystem I (PSI) and increases the risk of the accumulation of electrons in PSI of thylakoid membranes [1][2][3][4][5]. The reaction center chlorophyll (Chl), P700, is the electron source in PSI and drives the electron transport reaction from plastocyanin (PC) to NADP + through ferredoxin (Fd). In the sunflower (Helianthus annuus), exposure of intact leaves to repetitive short-pulse high light intensity (rSP-illumination treatment) in the dark inactivates the PSI electron transport reaction [6]. Short-pulse illumination causes the accumulation of electrons on the acceptor side of PSI, which stimulates the production of reactive oxygen species (ROS) such as superoxide radicals and singlet oxygen [6][7][8]. By contrast, the rSP-illumination treatment under steady-state actinic light (AL) oxidizes P700, but does not lead to the accumulation of electrons or inactivation of PSI.

Results
2.1. Effect of Ambient CO 2 Partial Pressure on the Reduction Rates of PC + and P700 + and Oxidation Rate of Fd − To modulate the electron flux in the LEF in wheat leaves, we manipulated the partial pressure of ambient CO 2 (pCO 2 ). Generally, a reduction in pCO 2 decreases the electron flux in the LEF, thus oxidizing the reaction center Chl, P700, in PSI [17,18]. This situation is advantageous, as it enhances our understanding of the effects of P700 oxidation on the relationship between P700 + reduction rate and Fdoxidation rate.
We set pCO 2 at 40, 20, and 5 Pa in a stepwise manner, with a photon flux density of 1000 µmol photons m −2 s −1 and an ambient oxygen partial pressure (pO 2 ) of 21 kPa, in wheat leaves. The typical kinetics of the reduction of PC + and P700 + and oxidation of Fd − at the highest and lowest pCO 2 (40 and 5 Pa, respectively) were obtained using DIRK analysis after the electron flux in LEF reached a steady state, which was determined based on the stable values of quantum yield of PSII (Y[II]) ( Figure 1) [21,22]. oxidized P700 (P700 + ), oxidized PC (PC + ) and reduced Fd (Fd − ), reduction rates of P700 + and PC + and oxidation rate of Fd − under AL illumination were estimated [21]. The rate of decrease in P700 + and PC + and the reduction rate of P700 + and PC + at 40 Pa pCO2 ( Figure 1A) were higher than those at 5 Pa pCO2 ( Figure 1B). Similarly, the rate of decrease in Fd − and the oxidation rate of Fd − at 40 Pa pCO2 ( Figure 1A) was higher than that at 5 Pa pCO2 ( Figure 1B). These results indicate that the turnover rates of redox reactions of P700 + , PC + and Fd − at 40 Pa pCO2 are higher than those at 5 Pa pCO2. Figure 1. DIRK analysis of the decay of oxidized P700 (P700 + ), PC (PC + ), and reduced Fd (Fd − ) in wheat leaves after turning off actinic light (AL) illumination. Changes in the redox state of P700, PC, and Fd were monitored using a Dual/KLAS-NIR spectrophotometer [19,20]. To determine the reduction rates of P700 + and PC + and oxidation rate of Fd − in wheat leaves under the illuminated condition, AL was transiently turned off at time zero for 400 ms. The initial slope of the decrease in P700 + , PC + , and Fd − at time zero indicated the reduction rates of P700 + and PC + and oxidation rate of Fd − . The initial slope changes of P700 + , PC + , and Fd − were characterized by averaging 70 sets of measurements at 25 °C leaf temperature, 21 kPa pO2, and 1000 mol photons m −2 s −1 light intensity, with either 40 Pa pCO2 (A) or 5 Pa pCO2 (B) for 110 ms after the AL was turned off. These data were obtained at a steady state, which was confirmed by the achievement of stable Y(II).
We plotted the steady-state levels of P700 + , PC + , and Fd − against Y(II) (Figure 2 [17] and Miyake et al. [18]. Furthermore, the level of non-photochemical quenching of Chl fluorescence (NPQ) increased with the decrease in Y(II) ( Figure S1). The decrease in pCO2 induces lumen acidification of thylakoid membranes, which drives NPQ induction and P700 oxidation.  [19,20]. To determine the reduction rates of P700 + and PC + and oxidation rate of Fd − in wheat leaves under the illuminated condition, AL was transiently turned off at time zero for 400 ms. The initial slope of the decrease in P700 + , PC + , and Fd − at time zero indicated the reduction rates of P700 + and PC + and oxidation rate of Fd − . The initial slope changes of P700 + , PC + , and Fd − were characterized by averaging 70 sets of measurements at 25 • C leaf temperature, 21 kPa pO 2 , and 1000 mol photons m −2 s −1 light intensity, with either 40 Pa pCO 2 (A) or 5 Pa pCO 2 (B) for 110 ms after the AL was turned off. These data were obtained at a steady state, which was confirmed by the achievement of stable Y(II).
The steady-state level of P700 + under AL illumination at 40 Pa pCO 2 ( Figure 1A) was lower than that at 5 Pa pCO 2 ( Figure 1B). Additionally, the steady-state level of PC + at 40 Pa pCO 2 ( Figure 1A) was slightly lower than that at 5 Pa pCO 2 ( Figure 1B), whereas the steady-state level of Fd − at 40 Pa pCO 2 was higher than that at 5 Pa pCO 2 ( Figure 1B).
In the DIRK analysis, AL illumination was transiently turned off for 400 ms. From the decay of oxidized P700 (P700 + ), oxidized PC (PC + ) and reduced Fd (Fd − ), reduction rates of P700 + and PC + and oxidation rate of Fd − under AL illumination were estimated [21]. The rate of decrease in P700 + and PC + and the reduction rate of P700 + and PC + at 40 Pa pCO 2 ( Figure 1A) were higher than those at 5 Pa pCO 2 ( Figure 1B). Similarly, the rate of decrease in Fd − and the oxidation rate of Fd − at 40 Pa pCO 2 ( Figure 1A) was higher than that at 5 Pa pCO 2 ( Figure 1B). These results indicate that the turnover rates of redox reactions of P700 + , PC + and Fd − at 40 Pa pCO 2 are higher than those at 5 Pa pCO 2 .
We plotted the steady-state levels of P700 + , PC + , and Fd − against Y(II) ( Figure 2). Lowering the pCO 2 decreased Y(II), indicating the suppression of LEF activity. With the reduction in Y(II), the level of PC + increased (linear regression line: , which was consistent with the results of both Golding and Johnson [17] and Miyake et al. [18]. Furthermore, the level of non-photochemical quenching of Chl fluorescence (NPQ) increased with the decrease in Y(II) ( Figure S1). The decrease in pCO 2 induces lumen acidification of thylakoid membranes, which drives NPQ induction and P700 oxidation.
Next, we evaluated the reduction rates of both P700 + and PC + and the oxidation rate of Fd − under steady-state conditions, using DIRK analysis ( Figure 3). It is assumed that the initial rates of the redox changes in these components after the introduction of transient darkness reflect their rates of reduction (P700+ and PC+) or oxidation (Fd − ) during illumination right before the darkness [21]. The relative reduction rates of P700 + (vP700) and PC + (vPC) and relative oxidation rate of Fd − (vFd), obtained from the initial 5 ms changes after turning off AL, were expressed as the percentage relative to the maximal level of each component. Initial changes of Fd − oxidation was estimated under different pCO 2 values, while simultaneously measuring Y(II), and plotted against Y(II). The relative rate of Fd − oxidation clearly showed a positive linear relationship with Y(II) (linear regression line: . These results indicate that the turnover rate of Fd is determined by the photosynthetic LEF, which includes both CO 2 assimilation and photorespiration; these results are consistent with those of a previous study examining Arabidopsis thaliana leaves [22]. Next, we evaluated the reduction rates of both P700 + and PC + and the oxidation rate of Fd − under steady-state conditions, using DIRK analysis ( Figure 3). It is assumed that the initial rates of the redox changes in these components after the introduction of transient darkness reflect their rates of reduction (P700+ and PC+) or oxidation (Fd − ) during illumination right before the darkness [21]. The relative reduction rates of P700 + (vP700) and PC + (vPC) and relative oxidation rate of Fd − (vFd), obtained from the initial 5 ms changes after turning off AL, were expressed as the percentage relative to the maximal level of each component. Initial changes of Fd − oxidation was estimated under different pCO2 values, while simultaneously measuring Y(II), and plotted against Y(II). The relative rate of Fd − oxidation clearly showed a positive linear relationship with Y(II) (linear regression line: . These results indicate that the turnover rate of Fd is determined by the photosynthetic LEF, which includes both CO2 assimilation and photorespiration; these results are consistent with those of a previous study examining Arabidopsis thaliana leaves [22].  (Figure 3). Like Fd, the redox reaction of PC was mainly determined by LEF.

AEF-I Functions in the Induction of Photosynthesis
We showed that Fd-independent electron flow, which is referred to as AEF-I, functioned within PSI ( Figure 3). Next, we tried to detect the electron flux in AEF-I during the induction of photosynthesis. When the AL was switched on, both P700 and PC were oxidized with a lag time in the first phase (phase I) ( Figure 4A,B), reaching maximum oxidation (75% and 90%, respectively) at approximately 80 s in the second phase (phase II). In the third phase (phase III), the level of P700 + declined by approximately 40%, whereas that of PC + remained largely unchanged. Illumination using AL rapidly increased the level of Fd − to approximately 85% of that in phase I ( Figure 4C). On the other hand, approximately 45% of Fd − was oxidized at 80 s in phase II, and the level of Fd − gradually decreased in phase III.  We evaluated the changes in vP700, vPC, and vFd by DIRK analysis during the photosynthesis induction, where the redox changes of P700, PC and Fd showed three phases (phase I, II, and III: Figure 4). These facts suggested that each redox reaction rate would change in three phases ( Figure  5). On AL illumination, vP700 increased to a maximum at approximately 250 seconds, with a lag time in phase I ( Figure 5A). By contrast, vPC decreased considerably to levels approximating zero ( Figure  5B), probably because of the rapid oxidation of PC − by P700 + , the level of which was in excess of 20%, as described above [21,23]. Change in vFd exhibited a complex pattern; vFd decreased rapidly from 3 to 2 (%/5 ms) over 20 s in phase I after the AL was switched on, then increased from 2 to 5 (%/5 ms) in phase II, and continued to increase in phase III, reaching a maximum of 12 (%/5 ms) at 800 s ( Figure  5C). The pattern of increase in vFd resembled that observed in Y(II) during the induction of CO2 assimilation ( Figure S2) [24]. In the lag phase of the induction of CO2 assimilation (phases I and II), the behavior of Y(II) did not match that of the net CO2 fixation rate, where an increased Y(II) showed that photorespiration would function. Photorespiration would start when the AL was switched on and would drive the photosynthetic LEF as a major electron sink in the lag phase ( Figure 5C). A rapid start of photorespiration during the induction of CO2 assimilation has been demonstrated previously [25]. The increase in vFd in phase III was likely driven by the activated photosynthesis and photorespiration [24].
The behavior of the increase in vFd differed from that of vP700. Since vFd reflects the electron flux in LEF and vP700 can function independent on vFd (Figure 3), rapid increase of vP700 in phase II would show the activation of AEF-I. That is, during the induction of photosynthesis, AEF-I functions with the oxidation of P700 (Figure 4). We evaluated the changes in vP700, vPC, and vFd by DIRK analysis during the photosynthesis induction, where the redox changes of P700, PC and Fd showed three phases (phase I, II, and III: Figure 4). These facts suggested that each redox reaction rate would change in three phases ( Figure 5). On AL illumination, vP700 increased to a maximum at approximately 250 seconds, with a lag time in phase I ( Figure 5A). By contrast, vPC decreased considerably to levels approximating zero ( Figure 5B), probably because of the rapid oxidation of PC − by P700 + , the level of which was in excess of 20%, as described above [21,23]. Change in vFd exhibited a complex pattern; vFd decreased rapidly from 3 to 2 (%/5 ms) over 20 s in phase I after the AL was switched on, then increased from 2 to 5 (%/5 ms) in phase II, and continued to increase in phase III, reaching a maximum of 12 (%/5 ms) at 800 s ( Figure 5C). The pattern of increase in vFd resembled that observed in Y(II) during the induction of CO 2 assimilation ( Figure S2) [24]. In the lag phase of the induction of CO 2 assimilation (phases I and II), the behavior of Y(II) did not match that of the net CO 2 fixation rate, where an increased Y(II) showed that photorespiration would function. Photorespiration would start when the AL was switched on and would drive the photosynthetic LEF as a major electron sink in the lag phase ( Figure 5C). A rapid start of photorespiration during the induction of CO 2 assimilation has been demonstrated previously [25]. The increase in vFd in phase III was likely driven by the activated photosynthesis and photorespiration [24]. that photorespiration would function. Photorespiration would start when the AL was switched on and would drive the photosynthetic LEF as a major electron sink in the lag phase ( Figure 5C). A rapid start of photorespiration during the induction of CO2 assimilation has been demonstrated previously [25]. The increase in vFd in phase III was likely driven by the activated photosynthesis and photorespiration [24].
The behavior of the increase in vFd differed from that of vP700. Since vFd reflects the electron flux in LEF and vP700 can function independent on vFd (Figure 3), rapid increase of vP700 in phase II would show the activation of AEF-I. That is, during the induction of photosynthesis, AEF-I functions with the oxidation of P700 (Figure 4). The behavior of the increase in vFd differed from that of vP700. Since vFd reflects the electron flux in LEF and vP700 can function independent on vFd (Figure 3), rapid increase of vP700 in phase II would show the activation of AEF-I. That is, during the induction of photosynthesis, AEF-I functions with the oxidation of P700 (Figure 4).

Discussion
We examined the effect of P700 oxidation on the redox state of Fd and activation of AEF-I. Switching on the AL to induce CO 2 assimilation reduced Fd to approximately 90%; subsequently, the level of Fd − decreased to 45%, while that of P700 + increased to approximately 80% (Figures 4 and 5). These results indicate that the limitation of P700 turnover in PSI shifted from the acceptor side to the donor side [4,10,12,26]. The donor side limitation of P700 turnover was induced by enhanced LEF, as observed by the increase in vFd (phase II in Figures 4 and 5), which contributed to P700 oxidation. The oxidation of P700 then induced AEF-I ( Figures 5 and 6).
We presume that charge recombination in PSI drives AEF-I ( Figure 6). Once P700 is excited to P700*, it undergoes charge separation to produce P700 + and an electron [27][28][29][30]. The electrons released from P700* flow toward Fd via four electron-transfer cofactors: A 0A /A 0B , A 1A /A 1B , F x , and [F A /F B ] [31][32][33][34][35]. The electrons accumulated in these cofactors then flow toward P700 + ; this phenomenon is referred to as a charge recombination. During this phenomenon, P700 + functions as an electron sink in the PSI of thylakoid membranes. The reduced electron-transfer cofactors A 0A /A 0B , A 1A /A 1B , F x , and [F A /F B ] recombine with P700 + in approximately 30 ns, 20 µs, 0.5-2 ms, and 100 ms, respectively. In the present study, the half-time of the reduction of P700 + was in the same range as the recombination rate of P700 + with F x or [F A /F B ]. We propose that the AEF-I is driven by charge recombination of P700 + with the reduced form of F x or [F A /F B ] in PSI ( Figure 6).
These electron-transfer cofactors in PSI donate electrons to O 2 to produce superoxide radicals [6][7][8][36][37][38]. The O 2 reduction rate constants of these cofactors are in the order of 10 6 M −1 s −1 [32]. The apparent K m for O 2 in the photoreduction of O 2 by PSI in thylakoid membranes is approximately 20 µM, which is approximately 1/10 of the O 2 concentration in water equilibrated with atmospheric O 2 (20.95%) [39]. This means that the photoreduction of O 2 to superoxide radicals in thylakoid membranes is not limited by the availability of O 2 , but by that of electron donors on the reducing side of PSI. The accumulation of electrons in the electron-transfer cofactors, A 0A /A 0B , A 1A /A 1B , Fx, and [F A /F B ], is dangerous for chloroplasts, as this enables the production of ROS. The recombination of these cofactors with P700 + inhibits the interaction between PSI and O 2 . released from P700* flow toward Fd via four electron-transfer cofactors: A0A/A0B, A1A/A1B, Fx, and [FA/FB] [31][32][33][34][35]. The electrons accumulated in these cofactors then flow toward P700 + ; this phenomenon is referred to as a charge recombination. During this phenomenon, P700 + functions as an electron sink in the PSI of thylakoid membranes. The reduced electron-transfer cofactors A0A/A0B, A1A/A1B, Fx, and [FA/FB] recombine with P700 + in approximately 30 ns, 20 µs, 0.5-2 ms, and 100 ms, respectively. In the present study, the half-time of the reduction of P700 + was in the same range as the recombination rate of P700 + with Fx or [FA/FB]. We propose that the AEF-I is driven by charge recombination of P700 + with the reduced form of Fx or [FA/FB] in PSI ( Figure 6).

Figure 6.
Hypothetical pathway of the AEF-I. Photo-excited P700 (P700*) donates an electron to the first electron carrier, A0, to produce P700 + . Subsequently, P700 + accepts an electron from PC in PSII to regenerate P700. A0 donates the electron to the second electron carrier, A1. Thereafter, the electron flows to ferredoxin (Fd) through the third and fourth electron carriers, Fx and FA/FB, respectively. Empty arrows represent P700 turnover in the photo-oxidation reduction cycle. Solid arrows indicate electron flow. Dotted arrows indicate electron flow during charge recombination (please see text for further details). Charge recombination is one of the mechanisms of additional electron flow. This figure is a modification of original figures published previously [27][28][29][30][31][32][33][34][35].
These electron-transfer cofactors in PSI donate electrons to O2 to produce superoxide radicals [6][7][8][36][37][38]. The O2 reduction rate constants of these cofactors are in the order of 10 6 M −1 s −1 [32]. The apparent Km for O2 in the photoreduction of O2 by PSI in thylakoid membranes is approximately 20 µM, which is approximately 1/10 of the O2 concentration in water equilibrated with atmospheric O2 (20.95%) [39]. This means that the photoreduction of O2 to superoxide radicals in thylakoid The charge recombination reactions between electron-transfer cofactors and P700 + are exergonic. The mid-point potentials of these cofactors are lower than that of P700 + [32]. Therefore, these recombination reactions dissipate energy as heat, which contributes to the alleviation of PSI photoinhibition under excess light energy [40].
It has been suggested for a long time that photosynthetic organisms operate Fd-dependent cyclic electron flows around PSI [41][42][43][44]. It is possible that Fd is an electron donor for Fd-quinone oxidoreductase (FQR) and NADH dehydrogenase (NDH). FQR requires protein cofactors, such as PGR5/PGRL1. Recently, a positive linear relationship between vFd and photosynthetic LEF rate during steady-state photosynthesis was demonstrated in Arabidopsis mutants deficient in PGR5/PGRL1 and NDH [22]. This also supports the hypothesis that Fd turnover is mainly determined by CO 2 assimilation and photorespiration. Thus, both FQR-and NDH-dependent cyclic electron flows show negligible activity compared with photosynthetic LEF [45].
Furthermore, our conclusion that AEF-I is not driven by PGR5/PGRL1 or NDH is strongly supported by previous studies. Yamamoto et al. [46] clearly showed the recovery of AEF-I in PSI after the recovery of the P700 oxidation system. The Arabidopsis PGR5-deficient mutant did not maintain P700 in its oxidized state, nor did it show any AEF-I under fluctuating light [46]. A double mutant lacking PGR5 and overexpressing FLV recovered AEF-I. The FLV stimulated the photosynthetic LEF and oxidized P700 in PSI, which induced the electron flux in AEF-I in both wild-type and PGR5-deficient Arabidopsis plants. This was further confirmed by the introduction of FLV in PGR5 deficient rice (Oryza sativa) plants [47]. Based on these studies, we conclude that AEF-I is not driven by PGR5/PGRL1 or NDH.
We propose a new physiological function of P700 oxidation. The oxidation of P700 suppresses the accumulation of electrons in the electron-transfer cofactors on the acceptor side of PSI by stimulating charge recombination in PSI, as described above. The oxidation of P700 also suppresses electron flow to Fd, as observed in the induction of Fd oxidation. The Fd − donates electrons to O 2 to produce a superoxide radical [48]. The superoxide radical disproportionates to O 2 and H 2 O 2 , the latter of which