Photorespiration Alleviates Photoinhibition of Photosystem I under Fluctuating Light in Tomato

Fluctuating light (FL) is a typical natural light stress that can cause photodamage to photosystem I (PSI). However, the effect of growth light on FL-induced PSI photoinhibition remains controversial. Plants grown under high light enhance photorespiration to sustain photosynthesis, but the contribution of photorespiration to PSI photoprotection under FL is largely unknown. In this study, we examined the photosynthetic performance under FL in tomato (Lycopersicon esculentum) plants grown under high light (HL-plants) and moderate light (ML-plants). After an abrupt increase in illumination, the over-reduction of PSI was lowered in HL-plants, resulting in a lower FL-induced PSI photoinhibition. HL-plants displayed higher capacities for CO2 fixation and photorespiration than ML-plants. Within the first 60 s after transition from low to high light, PSII electron transport was much higher in HL-plants, but the gross CO2 assimilation rate showed no significant difference between them. Therefore, upon a sudden increase in illumination, the difference in PSII electron transport between HL- and ML-plants was not attributed to the Calvin–Benson cycle but was caused by the change in photorespiration. These results indicated that the higher photorespiration in HL-plants enhanced the PSI electron sink downstream under FL, which mitigated the over-reduction of PSI and thus alleviated PSI photoinhibition under FL. Taking together, we here for the first time propose that photorespiration acts as a safety valve for PSI photoprotection under FL.


Introduction
Growth light significantly affects photosynthetic performance in plants. Plants usually modulate their biochemical composition and leaf morphology to acclimate to the specific growth light conditions [1][2][3][4][5]. In general, plants grown under high light (HL-plants) have higher content of proteins and enzymes involving in photosynthetic electron flow and the Calvin-Benson cycle than plants grown under low light [6,7]. These characteristics favors the higher photosynthetic capacity in HL-plants. Concomitantly, the rate of ribulose-1,5-bisphosphate (RuBP) oxygenation is also increased in HL-plants due to the higher ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) content [8]. Photorespiration is essential for the normal photosynthesis ambient CO 2 and oxygen [9]. A stronger electron flow for photorespiration can protect photosystem II (PSII) by consuming the excess light energy [9,10]. However, the role of photorespiration in protecting photosystem I (PSI) under fluctuating light is not well known.
In natural habitats, leaves usually experience fluctuations of illumination owing to cloud, wind and changing leaf sun angle [11,12]. Under fluctuating light (FL), light absorption and PSII electron flow rapidly increased after an abrupt increase in light intensity [13,14]. Meanwhile, stomatal opening and the activation of the Calvin-Benson cycle where R d was measured after incubation in darkness for 20 min.
After adequate photosynthetic induction, the response of CO 2 assimilation rate to incident intercellular CO 2 concentration (A/C i ) curves were measured by decreasing the CO 2 concentration to a lower limit of 50 µmol mol −1 and then increasing stepwise to an upper limit of 1500 µmol mol −1 . For each CO 2 concentration, photosynthetic measurement was completed in 3 min. Using the A/C i curves, the maximum rates of electron flux (J max ) and RuBP carboxylation (V cmax ) were calculated [49].

PSI and PSII Measurements
We used a Dual-PAM 100 measuring system (Heinz Walz, Effeltrich, German) to measure PSI and PSII parameters under atmospheric CO 2 condition. After illumination at 1455 µmol photons m −2 s −1 for 5 min to activate photosynthetic electron sinks, leaves were exposed to FL alternating between low light (59 µmol photons m −2 s −1 , 2 min) and high light (1455 µmol photons m −2 s −1 , 1 min). PSI parameters were calculated as follows: Y(I) = (P m ' − P)/P m ; Y(ND) = P/P m ; Y(NA) = (P m − P m )/P m . Y(I), the quantum yield of PSI photochemistry; Y(ND), the quantum yield of PSI non-photochemical energy dissipation due to donor side limitation; Y(NA), the quantum yield of PSI nonphotochemical energy dissipation due to acceptor side limitation. PSII parameters were calculated as follows: the effective quantum yield of PSII photochemistry; NPQ, non-photochemical quenching in PSII; Y(NO), the quantum yield of non-regulatory energy dissipation in PSII. The relative photosynthetic electron transport rate through PSI and PSII were calculated as: rETRI = PPFD × Y(I) × 0.84 × 0.5; rETRII = PPFD × Y(II) × 0.84 × 0.5. rETRI minus rETRII is assumed to be the rate of CEF.

Electrochromic Shift Measurement
A Dual-PAM 100 equipped with a P515/535 emitter-detector module (Heinz Walz) was used measure the electrochromic shift (ECS) signals. After light adaptation at 1455 µmol photons m −2 s −1 for 5 min, leaves were illuminated at 59 µmol photons m −2 s −1 for 2 min. Afterwards, light intensity was changed to 1455 µmol photons m −2 s −1 , and ECS dark interval relaxation kinetics (DIRK ECS ) were recorded after this light transition for 10 s or 60 s. The proton gradient (∆pH) component of proton motive force were calculated using DIRK ECS [50,51]. The chloroplast ATP synthase activity (g H + ) was estimated as the inverse of the decay time constant of the first-order ECS relaxation [52].

Statistical Analysis
We determined whether significant differences existed between HL-and ML-plants using t-test (α = 0.05). The software SigmaPlot 10.0 was used for graphing and fitting.

Results
Upon the light intensity changing from LL (59 µmol photons m −2 s −1 ) to HL (1455 µmol photons m −2 s −1 ), the quantum yield of PSI photochemistry (Y(I)) rapidly decreased in both types of leaves ( Figure 1A). Under high-light phases, HL-plants had higher Y(I) values than ML-plants ( Figure 1A). At LL, the quantum yield of energy dissipation due to donor-side limitation, Y(ND), was similar in HL-and ML-plants. However, after the transition from LL to HL for 10 s, Y(ND) increased more quickly in HL-plants ( Figure 1B). Concomitantly, the quantum yield of energy dissipation due to acceptor-side limitation, Y(NA), increased to a much higher level in ML-plants ( Figure 1C), indicating that the PSI over-reduction under FL was aggravated in ML-plants. Similar to the performance of Y(I), the effective quantum yield of PSII photochemistry, Y(II), rapidly decreased in both types of leaves by transitioning from LL to HL ( Figure 2A). Furthermore, HL-plants displayed higher Y(II) values under high-light phases ( Figure 2A). After transition from LL to HL for 10 s, non-photochemical quenching in PSII, NPQ, rapidly increased to approximately 80% of the maximum level in both HL-and ML-plants ( Figure 2B), diminishing the quantum yield of non-regulatory energy dissipation in PSII (Y(NO)) ( Figure 2C). Furthermore, ML-plants had higher Y(NO) than HL-plants, indicating more excess light energy in ML-plants.
By transitioning from LL to HL, the relative PSI electron flow (rETRI) progressively increased in HL-plants ( Figure 3A). By comparison, rETRI first increased and then decreased in ML-plants. The relative PSII electron flow (rETRII) progressively increased in both types of leaves, and HL-plants showed much higher rETRI and rETRII values in high-light phases ( Figure 3B). In HL-plants, the rETRII values after this light transition for 60 s were much higher than those for 10 s. By comparison, rETRII just increased slightly in ML-plants. These results indicated that, within the initial 10 s after this light transition, the CO 2 assimilation was strongly restricted in HL-plants. In order to evaluate the performance of CEF under fluctuating light, we analyzed the change in rETRI-rETRII during FL treatment. HLand ML-plants showed similarly low values of rETRI-rETRII under LL. Interestingly, the changing pattern of CEF under FL varied between HL-and ML-plants ( Figure 3C). After the transition from LL to HL, rETRI-rETRII rapidly increased to the maximum value in 10 s and maintained stable over time in HL-plants. By comparison, rETRI-rETRII firstly increased and subsequently rapidly decreased in ML-plants. In the first four cycles of LL/HL, the increase of CEF to the maximum value in ML-plants needed 30 s, indicating the delayed CEF activation in ML-plants. After the abrupt increase in illumination, the rETRI/rETRII ratio first increased and then decreased, suggesting the transient CEF stimulation under FL ( Figure S1). Furthermore, the contribution of CEF to total photosynthetic electron flow was enhanced in ML-plants ( Figure S1). for 10 s, non-photochemical quenching in PSII, NPQ, rapidly increased to appr 80% of the maximum level in both HL-and ML-plants ( Figure 2B), diminishing tum yield of non-regulatory energy dissipation in PSII (Y(NO)) ( Figure 2C). Fur ML-plants had higher Y(NO) than HL-plants, indicating more excess light ener plants. Y(I), the quantum yield of PSI photochemistry; (B) Y(ND), the quantum yield of PSI n chemical energy dissipation due to the donor side limitation; (C) Y(NA), the quantum y non-photochemical energy dissipation due to the acceptor side limitation. Data are show ± SE (n = 5). Green bars indicate low light (59 μmol photons m −2 s −1 ); yellow bars indicate (1455 μmol photons m −2 s −1 ). Asterisk indicates a significant difference between HL-and M  rETRI-rETRII firstly increased and subsequently rapidly decreased in ML-plants. In the first four cycles of LL/HL, the increase of CEF to the maximum value in ML-plants needed 30 s, indicating the delayed CEF activation in ML-plants. After the abrupt increase in illumination, the rETRI/rETRII ratio first increased and then decreased, suggesting the transient CEF stimulation under FL ( Figure S1). Furthermore, the contribution of CEF to total photosynthetic electron flow was enhanced in ML-plants ( Figure S1). Because ∆pH significantly regulates PSI redox state and NPQ induction under high light [53,54], we next examined the change in ∆pH under fluctuating light. After transition from LL to HL for 10 s, the ∆pH value was slightly lower than that for 60 s in HL-plants ( Figure 4A), suggesting that the ∆pH formation within the first 10 s was almost sufficient for photosynthetic regulation in HL-plants. By comparison, the ∆pH value for 10 s was significantly lower than that for 60 s in ML-plants ( Figure 4A), indicating that the ∆pH formation within the first 10 s was insufficient to regulate photosynthetic apparatus in ML-plants. After this light transition, the chloroplast ATP synthase activity (g H + ) gradually increased in HL-plants but was maintained stable in ML-plants ( Figure 4B). Because a decrease in g H + can enhance the formation of ∆pH, the flexibility of g H + upon a sudden transition from LL to HL in HL-plants likely favored the rapid generation of ∆pH.
Plants 2022, 11, 195 8 of 16 significantly lower than that for 60 s in ML-plants ( Figure 4A), indicating that the ΔpH formation within the first 10 s was insufficient to regulate photosynthetic apparatus in ML-plants. After this light transition, the chloroplast ATP synthase activity (gH + ) gradually increased in HL-plants but was maintained stable in ML-plants ( Figure 4B). Because a decrease in gH + can enhance the formation of ΔpH, the flexibility of gH + upon a sudden transition from LL to HL in HL-plants likely favored the rapid generation of ΔpH. After fluctuating light treatment for eight cycles of low/high light, ML-plants showed a larger decrease in Pm than HL-plants ( Figure 5A). Thus, FL induced a stronger PSI photoinhibition in ML-plants. Furthermore, we found a tight positive relationship between PSI photoinhibition and Y(NA) after transition from LL to HL for 10 s (Y(NA)10s) ( Figure 5B). Therefore, the greater PSI photoinhibition in ML-plants induced by FL was mainly caused by the transient PSI over-reduction. After fluctuating light treatment for eight cycles of low/high light, ML-plants showed a larger decrease in Pm than HL-plants ( Figure 5A). Thus, FL induced a stronger PSI photoinhibition in ML-plants. Furthermore, we found a tight positive relationship between PSI photoinhibition and Y(NA) after transition from LL to HL for 10 s (Y(NA) 10s ) ( Figure 5B). Therefore, the greater PSI photoinhibition in ML-plants induced by FL was mainly caused by the transient PSI over-reduction.
The PSI redox state under FL is significantly affected by the electron sink downstream of PSI [38,55,56]. Compared with ML-plants, HL-plants not only displayed higher capacities for the maximum rates of RuBP carboxylation (V cmax ) and regeneration (J max ), and photosynthetic CO 2 assimilation rate (A sat ), but also showed higher electron flow for photorespiration (J O ) when CO 2 assimilation was restricted at low CO 2 concentrations ( Figure 6). Therefore, the electron sink downstream of PSI was significantly enhanced in HL-plants. Within the first 60 s after transition from 50 to 1500 µmol photons m −2 s −1 , the gross CO 2 assimilation rate did not differ between HL-and ML-plants ( Figure 6), indicating that the variation of Y(NA) 10s between HL-and ML-plants was caused by photorespiration rather than the Calvin-Benson cycle. Furthermore, we found the maximum J O was negatively correlated to the PSI over-reduction under FL and FL-induced PSI photoinhibition (Figure 7), indicating that the enhancement of photorespiration in HL-plants alleviated the PSI over-reduction under FL and thus mitigated PSI photoinhibition.  The PSI redox state under FL is significantly affected by th stream of PSI [38,55,56]. Compared with ML-plants, HL-plants not capacities for the maximum rates of RuBP carboxylation (Vcmax) a and photosynthetic CO2 assimilation rate (Asat), but also showed h

Discussion
After an abrupt increase in illumination, the slow kinetics of CO 2 assimilation led to the lack of NADP + and made PSI to be over-reduced owing to the restriction of electron flow from PSI to NADP + . Some previous studies reported that FL induced a marked PSI photoinhibition in some angiosperms [20,21,57,58]. Furthermore, in Arabidopsis thaliana (Arabidopsis) and Erigeron annuus, FL-induced PSI photoinhibition was stronger in LLplants than in HL-plants [59]. Therefore, growth light can significantly affect the response of PSI to FL. However, the underlying mechanisms are not clear. Here we demonstrated that, upon transition from LL to HL, tomato ML-plants significantly showed a stronger PSI over-reduction than HL-plants, ultimately causing higher PSI photoinhibition in ML-plants. Therefore, ML-plants can fine-turn the redox state of PSI under FL to a much lesser extent. Plants grown under high light might display slight damage of PSI FeS cluster after exposure to high light [60]. In this study, we measured all photosynthetic parameters in the morning (a.m. 9:00-12:00), which might exclude the possibility of PSI photodamage in HL-plants.
The PSI redox state under FL is largely affected by donor and acceptor side regulation [21]. During donor side regulation, a high ∆pH can down-regulates the plastoquinone (PQ) oxidation at the Cyt b 6 /f complex, which restricts the electron transport from PQ to plastocyanin (PC) and, thus, avoids excess electron flow to PSI [61]. If ∆pH formation under high light was impaired, PSI would be over-reduced, leading to uncontrolled PSI photoinhibition [62,63]. Therefore, a sufficient ∆pH is indispensable to PSI photoprotection under high light. The ∆pH formation under high light is mainly affected by photosynthetic electron flow and chloroplast ATP synthase activity [53,64]. After transition from LL to HL, HL-plants displayed high levels of photosynthetic electron transport rates and a relatively low chloroplast ATP synthase activity, contributing to the sufficient ∆pH formation. By comparison, the ML-plants generated an insufficient ∆pH after this light transition for 10 s. Therefore, HL-plants reinforced the PSI donor side regulation under FL.
In PSI acceptor side regulation, reducing power in PSI is consumed by linear electron flow, photo-reduction of O 2 mediated by the Mehler reactions and Flvs [33,65,66]. Flvdependent alternative electron flow was lost in angiosperms during evolution [37]. The O 2 photo-reduction mediated by the Mehler reactions, called water-water cycle, could rapidly consume excess electrons in PSI and thus avoided PSI over-reduction under FL [56,67,68]. In the present study, a transient PSI over-reduction under FL indicated that water-water cycle was not significant in tomato leaves. The excess PSII electron flow is the prerequisite for PSI over-reduction [36,63,69,70]. Although HL-plants displayed much higher PSII electron flow than ML-plants, HL-plants showed a rapid oxidation of PSI upon the sudden transition from LL to HL. These results suggested that HL-plants have the ability to rapidly consume the large amounts of reducing power in PSI through acceptor side regulation.
After transition from LL to HL for 10 s, the value of rETRII was much higher in HL-plants than ML-plants ( Figure 3B), whereas the gross CO 2 assimilation rate showed no significant difference between them ( Figure 6). These results indicated that, upon a sudden increase in illumination, the difference in rETRII between HL-and ML-plants was not attributed to the Calvin-Benson cycle but was caused by the change in photorespiration. Accordingly, within the first 10 s after this light transition, electron flow for photorespiration was the major component of PSII electron flow in HL-plants. Furthermore, photorespiration could rapidly consume the excess reducing power in PSI [42]. An increase in photorespiratory pathway can accelerate the consumption of NADPH and subsequently elevates the NADP + /NADPH ratio. Therefore, the enhanced photorespiratory pathway in HL-plants facilitates the outflow of electrons from PSI to NADP + under FL and eventually alleviates PSI over-reduction at acceptor side. Furthermore, the higher electron flow to photorespiration in HL-plants helped the rapid formation of ∆pH after transition from low to high light, which alleviated PSI over-reduction under FL at donor side. Taking together, the enhancement of photorespiration in HL-plants protected PSI against photoinhibition under FL at donor and acceptor sides.
The photorespiratory pathway starts with the oxygenation of RuBP, which leads to the formation of 3-phosphoglycerate (3-PGA) and 2-phosphoglycolate (2-PG) [71]. 2PG is a dead-end intermediate of photosynthesis, which cannot directly be used and strongly impairs plant carbon metabolism [72][73][74]. To prevent the accumulation of 2-PG, photorespiratory pathway converts 2-PG into 3-PGA to replenish the Calvin-Benson cycle [75,76]. Therefore, acceleration of the photorespiratory pathway can significantly boost CO 2 assimilation rate and plant growth via optimized RuBP regeneration and preventing 2-PG mediated down-regulation of the Calvin-Benson cycle [43,[77][78][79]. Thus, the enhancement of photorespiratory pathway has the potential to improve plant growth and crop yield under field FL conditions. Under FL conditions, avoiding PSI photoinhibition is a premise of the maintenance of CO 2 assimilation rate [31,80]. This is in good agreement with the presented results and the hypothesis that that the photorespiratory pathway acted as a safety valve to protect PSI under FL.
In previous studies on PSI photoprotection under FL, CEF is thought to be the main factor affecting the PSI redox state. During CEF, electrons are transferred from ferredoxin to PSI, which is accompanied with the proton translocation [81][82][83]. The CEF-dependent formation of ∆pH is critical for NPQ induction and PSII photoprotection under high light [84]. Moreover, once CEF activity was decreased by mutation of PGR5-and/or NDHpathway, the PSI over-reduction under high light would be aggravated, leading to severe PSI photoinhibition [62,63,80,85]. After transition from LL to HL, CEF first increased and then rapidly decreased in ML-plants, which was consistent with previous results reported in Arabidopsis and tobacco plants grown under low light [20,22]. The first transient CEF stimulation helped the formation of ∆pH. The latter decrease in CEF prevented an overacidification of thylakoid lumen, which maximized the photosynthetic light use efficiency. However, we found that the CEF performance in HL-plants was different from ML-plants. After the same light transition, the rate of CEF rapidly increased and then was maintained stable in HL-plants. Furthermore, within the first 30 s after transition from LL to HL, ML-plants displayed higher rETRI/rETRII ratios than HL-plants ( Figure S1). These results indicated that ML-plants showed a larger contribution of CEF to total photosynthetic electron transport under FL. Therefore, to prevent uncontrolled over-reduction of PSI under FL, ML-plants enhanced the stimulation of CEF to compensate for the shortage of photorespiration.

Conclusions
We here showed that ML-plants displayed a greater PSI over-reduction under FL than HL-plants in tomato. As a result, FL induced a stronger PSI photoinhibition in MLplants. Furthermore, the up-regulation of photorespiration in HL-plants enhanced the outflow of electrons from PSI under FL and thus alleviated PSI photoinhibition. Therefore, photorespiration acts as an important valve for PSI protection under FL. In ML-plants, CEF was highly activated to compensate for the lower capacity of photorespiration, which prevented uncontrolled PSI photoinhibition. Taking together, HL-and ML-plants likely use different strategies to protect their photosynthetic apparatus under FL. Future studies involving defined photorespiration mutants are needed to shed more light on these aspects.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.