Higher order photoprotection mutants reveal the importance of ΔpH-dependent photosynthesis-control in preventing light induced damage to both photosystem II and photosystem I

Although light is essential for photosynthesis, when in excess, it may damage the photosynthetic apparatus, leading to a phenomenon known as photoinhibition. Photoinhibition was thought as a light-induced damage to photosystem II; however, it is now clear that even photosystem I may become very vulnerable to light. One main characteristic of light induced damage to photosystem II (PSII) is the increased turnover of the reaction center protein, D1: when rate of degradation exceeds the rate of synthesis, loss of PSII activity is observed. With respect to photosystem I (PSI), an excess of electrons, instead of an excess of light, may be very dangerous. Plants possess a number of mechanisms able to prevent, or limit, such damages by safe thermal dissipation of light energy (non-photochemical quenching, NPQ), slowing-down of electron transfer through the intersystem transport chain (photosynthesis-control, PSC) in co-operation with the Proton Gradient Regulation (PGR) proteins, PGR5 and PGRL1, collectively called as short-term photoprotection mechanisms, and the redistribution of light between photosystems, called state transitions (responsible of fluorescence quenching at PSII, qT), is superimposed to these short term photoprotective mechanisms. In this manuscript we have generated a number of higher order mutants by crossing genotypes carrying defects in each of the short-term photoprotection mechanisms, with the final aim to obtain a direct comparison of their role and efficiency in photoprotection. We found that mutants carrying a defect in the ΔpH-dependent photosynthesis-control are characterized by photoinhibition of both photosystems, irrespectively of whether PSBS-dependent NPQ or state transitions defects were present or not in the same individual, demonstrating the primary role of PSC in photoprotection. Moreover, mutants with a limited capability to develop a strong PSBS-dependent NPQ, were characterized by a high turnover of the D1 protein and high values of Y(NO), which might reflect energy quenching processes occurring within the PSII reaction center.


Results
Arabidopsis plants devoid of short-term regulatory mechanisms highlight the primary importance of ΔpH-dependent photosynthesis-control for optimal pSi activity. In order to dissect the interconnections and the relative importance of short-term photoprotective mechanisms, mutations that abolish PSBS-dependent NPQ (npq4-1, lacking the PSBS subunit of photosystem II), ΔpH mutants (pgr5 and pgrl1a pgrl1b, henceforth referred to as pgrl1ab, devoid of the PGR5-PGRL1 protein complex) and thylakoid protein phosphorylation (stn7 stn8, lacking the thylakoid-associated STN kinases) have been combined with the aim to obtain both higher order mutants (npq4-1 pgrl1ab, npq4-1 pgr5) and the sextuple mutant, lacking the entire set of short-term regulatory mechanisms, hereafter referred to as ΔSTeM (Fig. 1A). Worth to note that about 25% of PGRL1 protein is still detectable in pgr5 and npq4-1 pgr5 thylakoids (Fig. 1B), whereas no accumulation of PGR5 protein is observable in pgrl1ab and npq4-1 pgrl1ab (see also Dal Corso et al. 27 ). As shown in Fig. 1A,C, a marked reduction in the growth rate is observed in the stn7 stn8 and in the ΔSTeM mutants when grown under optimal conditions (growth light intensity of 100 μmol photons m −2 s −1 , over a photoperiod of 16 h light/8 h dark). In particular, stn7 stn8 and ΔSTeM growth rates appear comparable to Col-0 until 14 days after sowing (DAS), whereas they diverge at 18 DAS resulting in rosettes with a decreased size at 24 DAS (Fig. 1A,C). No major differences are, instead, observed in the total chlorophyll content (Chl a + b) and Chl a/b ratio between the entire set of mutants and Col-0 (Fig. 1D), with the exception of pgr5 leaves that accumulate less chlorophyll without changing the Chl a/b ratio. To confirm the main role of PGR5-PGRL1 protein complex in the formation of the proton motive force (pmf), Col-0, npq4-1, pgr5, npq4-1 pgr5, stn7 stn8 and ΔSTeM plant lines were subjected to the kinetic analysis of the electrochromic pigment absorbance shift (ECS) (Fig. 2). Leaf material adapted to moderate-light (50 μmol photons·m −2 ·s −1 ) was exposed for 5 min to red actinic light (LED, 500 μmol photons m −2 s −1 ), then relaxed in the dark for 50 seconds ( Fig. 2A), and the ECS relaxation kinetic was measured during the light-to-dark transition (Fig. 2B). As described in Fig. 2A,B, npq4-1 and stn7 stn8 mutants showed an ECS kinetic similar to Col-0 control, generating a comparable pmf. On the contrary, pgr5-containing plant lines showed a marked difference in ECS relaxation kinetics, revealing a significantly reduced (~30% drop) capability of pgr5, npq4-1 pgr5 and ΔSTeM in generating pmf, when acclimated to moderate-light conditions (50 μmol photons m −2 s −1 ). These findings are in line to what previously reported 22,23,28 . An identical analysis was performed after 4 h exposure to 500 μmol photons m −2 s −1 white light, and comparable differences between pgr5-containg plants and Col-0 were observed (Fig. S1A,B).
In addition, the effective quantum yield of PSII [Y(II), see Fig. 3A], measured after dark-adaptation and at increasing actinic light intensities (0 to 829 μmol·photons m −2 ·s −1 ) was generally comparable among the different genotypes, despite the PSBS-dependent NPQ [Y(NPQ)] resulted to be completely abolished in the npq4-1-containing genotypes and dramatically decreased in pgrl1ab and pgr5 mutants, in which reached 40% of Col-0 level, after exposure to high light (Fig. 3B). In agreement with these observations, values of the Y(NO) parameter, indicating the energy quenching processes occurring within the PSII reaction center 29,30 , rapidly raised as light intensity increased in the npq4-1-containing mutants, whereas ΔpH mutants, showed a peculiar kinetic, characterized by a npq4-like behavior, at moderate light intensities, and lower values at higher light intensity (>400 μmol·m −2 ·s −1 ), as the mutants succeeded to establish a proton gradient in the lumen and to induce the PsbS-dependent NPQ (Fig. 3C, see also Tikkanen et al. 31 ,). In agreement with that, the fraction of open PSII centers 29 (qL, Fig. 3D) resulted to be higher in Col-0 and stn7 stn8 leaves than the rest of the genetic backgrounds, particularly at moderate light intensities (100-350 μmol·photons m −2 ·s −1 ).
An identical experimental set-up was used to assess PSI activity in the different genetic backgrounds (Fig. 3E-G). In particular, the quantum yield of PSI, Y(I), was relatively high in Col-0, npq4-1 and stn7 stn8 mutant plants even at high light intensities (till around 500 μmol·m −2 ·s − 1), whereas single and multiple mutants devoid of the PGR5-PGRL1 protein complex showed a marked drop of PSI yield at moderate-to-high light intensities (>100 μmol·photons m −2 ·s −1 ), as a consequence of their inability to efficiently oxidize the P700 chlorophyll pair (Fig. 3E, Tikkanen et al. 31 , Grieco et al. 32 ). As a matter of fact, the kinetic of Y(NA) parameter, i.e. the quantum yield of non-photochemical energy dissipation in PSI due to acceptor side limitation (Fig. 3F), was similar in npq4-1, stn7 stn8 and Col-0 leaves, reaching peaks at low light intensities (around 20-40 μmol photons·m −2 ·s −1 ) and showing Figure 1. Phenotypes of Col-0 and mutant plants lacking short-term photoprotective mechanism, such as the npq4-1 mutant lacking the PSBS subunit responsible of NPQ, the pgrl1ab and pgr5 mutants devoid of the PGR5-PGRL1 protein complex that contributes to the formation of the ΔpH transthylakoidal gradient, the stn7 and stn8 mutants lacking the thylakoid STN kinases and the sextuple ΔSTeM mutant with no short-term regulatory mechanisms. (A) Images of Col-0 and mutant plants grown under long-day conditions in a growth chamber for 24 days. The size bar corresponds to 1 cm. (B) Immunoblots of fractionated total proteins from Col-0 and mutant leaves probed with antibodies specific for PSBS, PGRL1A, PGR5, STN7 and STN8 proteins. (C) Growth rate measurements of plants grown under long-day conditions in a growth chamber for 24 days. Leaf area is expressed as cm 2 (DAS, Days after sowing). (D) Chlorophyll content expressed as μg mg −1 leaf fresh weight (histogram) and ratio between Chl a and Chl b (curve). Pigments were extracted from adult plants grown under long-day conditions in a growth chamber for 24 days. Bars indicate the standard deviation and the asterisks represents the statistical significance (**p-value < 0,01), as evaluated by ANOVA test and Student t-test. a strong decrease at higher light conditions, as soon as the photosynthetic control is engaged. On the other hand, PGR-devoid mutants showed a rapid increase of Y(NA) values, reaching a plateau at light intensity values higher than 300 μmol·photons m −2 ·s −1 , as a consequence of the over-reduction of PSI acceptor side. Similarly, the Y(ND) values, i.e. the non-photochemical PSI quantum yield of donor-side limited heat dissipation (Fig. 3G), showed the incapability of PGR-devoid mutants to accumulate P700 in the oxidized form, as previously described 24,33,34 . Overall, our findings highlight the primary importance of proton gradient regulation (PGR)-dependent photosynthesis-control with respect to PSI yield, especially under moderate actinic light intensities. photoinhibition of pSii is phenomenologically linked to the lack of ΔpH-dependent photosynthesis-control and the consequent over-reduction of pSi reaction centers.
Photoinhibition of Col-0 and mutant plants was evaluated via the maximum quantum yield of PSII (Fv/Fm), measured after 2 hours of exposure to 130, 500 or 1000 μmol·m −2 ·s −1 of light (Fig. 4A). The results clearly show that mutants lacking either the STN kinases and, surprisingly, even the PSBS protein behaved very similarly to Col-0. On the contrary, single and higher order mutants devoid of the ability to form a full pmf (see also Fig. 2) were strongly photoinhibited, with no major differences between ΔSTeM and the pgrl1ab and pgr5 mutants. Thus, it seems that the impairment of ΔpH-dependent photosynthesis-control confers enhanced light sensitivity, irrespectively of whether NPQ or state transitions are developed or not. In particular, when the negative slopes of trend lines obtained from  (Fig. 4D) values, a clear association with photoinhibition was only displayed by the lines containing the pgr mutations. Indeed, stn7 stn8 and ΔSTeM leaves, both devoid of the State Transitions mechanism, had very different slope values: much smaller, therefore indicating higher photoinhibition, in ΔSTeM with respect to stn7 stn8 (Fig. 4B). Similarly, a higher slope value, i.e. less photoinhibition, was observed in npq4-1 leaves with respect to ΔSTeM plants lacking both ΔpH-dependent photosynthesis-control and NPQ (Fig. 4C). Thus, taking these findings together, it appears clear that the ΔpH-dependent photosynthesis-control plays a major role in photoprotection.
To further characterize the functionality of PSII in the different genetic backgrounds, fluorescence decay measurements in the 10 −4 -10 2 sec time-range were performed on dark-adapted and HL-treated plants, irradiated with high-light (500 μmol·photons m −2 ·s −1 ) for either 2 or 4 hours (Fig. 5). A single-turnover saturating flash was used to trigger the reduction of Q A with a single electron, extracted from the donor side of PSII, leading to increased fluorescence yield. The subsequent dark-induced re-oxidation of Q A − resulted in the relaxation of fluorescence yield and exhibited three main decay phases: fast, middle and slow 35 . For each phase, amplitude and decay time constant were determined, as reported in Table 1. In the case of dark-adapted Col-0 (Fig. 5A), the fast phase, raised from re-oxidation of Q A − by plastoquinone bound to Q B site in the dark, contributed to 82% of total amplitude, with a time constant (T 1 ) of 309 µs. The middle phase, originated from re-oxidation of Q A − by . Bars indicate the standard deviation and the asterisks represents the statistical significance (**p-value < 0,01; ***p-value < 0,001) as evaluated by ANOVA test and Student t-test. Note, that we preferred to report the pmf rather than ΔpH values, since the real partitioning of the pmf between its two components (ΔΨ and ΔpH) is still debated 58,59 .
Scientific RepoRtS | (2020) 10:6770 | https://doi.org/10.1038/s41598-020-62717-1 www.nature.com/scientificreports www.nature.com/scientificreports/ plastoquinone molecules in reaction centers with empty Q B site at the time of the flash light, displayed 9,5% of total amplitude with a time constant (T 2 ) of 17 ms. Finally, the slow phase that arises from a back-reaction of the S 2 state of the water-oxidizing complex with Q A − , which is populated via the equilibrium between Q A − Q B and Q A Q B − , had a 8,8% relative amplitude with a time constant (T 3 ) of 4.2 sec. Comparable amplitude values for the three phases could also be observed for all the dark-adapted mutant genotypes, although clear differences were present in the time constants of the middle phase, ranging from a minimum value of 13 ms, observed in npq4-1 pgr5, to 65 ms calculated for ΔSTeM, indicating that under standard growth conditions PSII is working in a similar way in Col-0 and mutant thylakoids. However, when plants were exposed to high light for 2 and 4 hours a totally different scenario appeared. In particular, Col-0 leaves irradiated with high light for 2 or 4 hours www.nature.com/scientificreports www.nature.com/scientificreports/ decreased the total amplitude of 4 and 11%, respectively, as a result of a reduction of the fast phase and the concomitant increase of the middle and slow phase ( Fig. 5A and Table 1). T 1 and T 2 remained in the order of 0,3 ms and 15-30 ms, irrespectively of the irradiation length, whereas a marked shortening of T 3 was observed upon high light exposure. A similar situation was observed in npq4-1 (Fig. 5B) and stn7 stn8 ( Fig. 5C) leaves, but major differences were detected in mutants defective in ΔpH formation and, therefore, the photosynthesis-control regulatory mechanism ( Fig. 5D-H). First of all, a loss of a total amplitude between 25% and 30% is observed in all genotypes after 4 hours of high light exposure, due to a marked loss of the fast phase and the increase of middle and slow phase (Table 1). In addition, while T 1 remained in the order of 0,3 ms both T 2 and T 3 values showed marked drops, further confirming the over-reduction of the electron transport chain in the absence of a normal ΔpH transthylakoidal gradient.
Thylakoid protein phosphorylation does not have a major impact on D1 protein turnover and pSii photoinhibition. The level of PSII photoinhibition was also measured as the maximum PSII quantum yield (Fv/Fm), in dark-adapted leaves and in leaves exposed to either optimal growth light (GL, 100 μmol photons m −2 s −1 ) or stressful high light (HL, 500 μmol photons m −2 s −1 ) for 60, 120 and 240 minutes, in absence or presence of Lincomycin (Lin) (Fig. 6). Under GL condition, where there is no effect of light on Fv/Fm, the addition of Lin leads to a partial loss of PSII activity, linked to the inhibition of the de novo synthesis of D1 and, more in general, of plastid-encoded proteins (not shown). Under HL conditions, in the absence of Lin, a general decrease of Fv/Fm values after 60 minutes of HL exposure could be appreciated, more obvious in pgrl1ab and npq4-1 pgr-l1ab mutants (Fig. 6A). A similar trend was observed after 120 minutes of HL treatment, whereas the exposure to HL for 240 min indicated that photosynthesis-control devoid mutants are more susceptible to photoinhibition than Col-0, stn7 stn8 and npq4-1, confirming the data reported in Fig. 4A. As expected, in presence of Lin, all genotypes displayed a similar trend as in GL but with a much larger susceptibility to the HL treatment, as revealed by the considerable decrease of Fv/Fm after 120 and 240 min of HL exposure (Fig. 6B). In particular, Col-0 Fv/Fm was reduced of about 40% after HL treatment for 240 min in presence of Lin, whereas npq4-1 pgr5 and ΔSTeM mutants showed the largest PSII photoinhibition, with Fv/Fm values reduced by 80 and 75%, respectively, in comparison to dark-adapted samples.
In addition to PSII photoinhibition evaluated by fluorescence-based methods (PAM, single turnover flash), the same samples were also analyzed for the ability to accumulate D1 protein by immunoblot analyses under GL and HL conditions with and without Lin treatments (Fig. 7). In all tested genotypes, D1 accumulation was not www.nature.com/scientificreports www.nature.com/scientificreports/ affected under GL conditions in the absence of Lin treatment, whereas leaves incubated overnight with 2,3 mM of Lin and then exposed to GL for 240 min, showed decreased D1 amount similar to the levels observed after HL exposure for 240 min in the absence of Lin. However, when the HL exposure was combined with the Lin treatment, differences became evident. In particular, D1 accumulation was markedly decrease in npq4-1, pgr5, pgrl1ab, npq4-1 pgrl1ab, npq4-1 pgr5 and ΔSTeM with respect to Col-0 amount. Interestingly, no additive effects were observed when the accumulation of D1 protein in npq4-1 thylakoids was compared with npq4-1 pgrrl1ab and npq4-1 pgr5 and the ΔSTeM sextuple mutant, in agreement with the Fv/Fm values reported in Figs. 4A and 6B. On the contrary, Col-0 and the stn7 stn8 mutant suffered a comparable and marginal decrease of D1 amount after 240 min of HL.
The fact that Col-0 and the stn7 stn8 double mutant do not show major differences with respect to PSII yield under HL stress conditions with and without Lin, points to a marginal role of thylakoid protein phosphorylation with respect to PSII photoprotection. To investigate further this aspect, the thylakoid phosphorylation pattern   www.nature.com/scientificreports www.nature.com/scientificreports/ was monitored in plants devoid of either NPQ or ΔpH-dependent photosynthesis-control and in the corresponding mutants where both mechanisms are inactivated (Fig. 8). In agreement with previous observations, the exposure of Col-0 leaves to GL led to a general increase in phosphorylation of all main phosphoproteins, i.e. LHCII, D1 and D2, over time (0-to-240 min), whereas CP43 was already strongly phosphorylated in our experimental conditions. The addition of Lin increased the phosphorylation level of PSII-core proteins even in the absence of light and a comparable accumulation of PSII-core phosphoproteins was maintained until 120 min of GL exposure. On the contrary, P-LHCII signal reached its peak at 60 min, markedly decreased at 120 min and disappear after 240 min of GL with Lin treatment. HL exposure in the absence of Lin maintained a relatively high accumulation of CP43, D1 and D2 phosphoproteins, comparable to what observed at 120-240 min of GL conditions, throughout the tested time points. However, LHCII phosphorylation was barely detectable after 30 min and disappeared after 60 min exposure to high light. The addition of Lin to HL conditions resulted in a gradual loss of phosphorylation levels. In general, the PSII-core phosphorylation pattern observed in mutant plants (see Fig. 8) was very similar to Col-0 under the different light regimes in presence or absence of Lin, although the accumulation of PSII core phosphoproteins was markedly higher in npq4-1 thylakoids and clearly reduced in pgrl1ab in comparison to Col-0. Notably, the D1 phosphoprotein was barely detectable in pgrl1ab thylakoids even after 240 min of GL exposure. On the contrary, the LHCII phosphorylation pattern, upon HL illumination, was markedly different between Col-0 and mutant plants devoid PGR proteins. In particular, LHCII phosphorylation was retained in pgr mutants upon high light treatment, unlike Col-0 and npq4-1 thylakoids where high light exposure suppressed LHCII phosphorylation. This phosphorylation pattern resembles the one of tap38 mutant 36 and is certainly the consequence of the high reduction state of the thylakoid electron transport carriers, including Cyt b 6 f, upon depletion of the ΔpH-dependent photosynthesis control. pSii photoinhibition guarantees pSi integrity. In order to evaluate the impact of PSII photoinhibition on PSI integrity, Fv/Fm and Pm (the maximal change of the P700 signal upon quantitative transformation of P700 from the fully reduced to the fully oxidized state) parameters were measured from dark-adapted and HL-treated (2 and 4 hours) plants, in either absence or presence of Lin ( Fig. S3 and Table 2). In the dark, Fv/Fm did not shown any marked difference among genotypes (p < 0,05), whereas the Pm parameter was higher in Col-0, npq4-1, stn7 stn8, than all pgr-containing mutants (p < 0,05). The Fv/Fm values slightly decreased by increasing the length of exposure to HL, while addition of Lin led to a marked drop of Fv/Fm values (see Table 2). On the other hand, Pm values remained higher than 0,40 at the different HL regimes, irrespectively of the presence or absence of Lin. On the contrary, PGR-devoid mutants were highly sensitive to high light conditions, displaying Fv/Fm values in the range of 0,65-0,31, much lower than 0,78 observed in Col-0 ( Fig. S3 and Table 2). In addition, PSI activity was found to be lower than 0,15 in pgr5, npq4-1 pgrl1ab and npq4-1 pgr5 thylakoids under the same conditions. Interestingly, the addition of Lin to the high-light treatment restored PSI activity in PGR-devoid mutants, while PSII efficiency dropped to values even lower than 0,30, as in the case of pgr5, npq4-1 pgr5 and ΔSTeM leaves. www.nature.com/scientificreports www.nature.com/scientificreports/ Overall, these findings indicate that in the absence of the PGR-dependent photosynthesis control, a marked inhibition of PSII activity is beneficial to prevent PSI inactivation, highlighting further the primary importance of photosynthesis control in photoprotection of PSI.

Discussion
Light induced inactivation of PSII causes enhanced degradation of the D1 protein, while the PSII recovery relies on de novo synthesis of D1. Under PSII photoinhibitory conditions (high light), activity and stability of PSI is not affected, unless high light treatment is performed in cold environment 10 or in mutant backgrounds lacking the PGR5/PGRL1 complex, in which the ability to form a normal ΔpH and activate the photosynthesis-control is not working properly 11,12 .
A large number of molecular processes have been suggested to function as protection mechanisms against an excess of light. Among those, the most relevant consists in the formation of the PSBS-dependent component of NPQ. Nevertheless, several authors argued that NPQ could have only a little role in the direct photoprotection of PSII, while could be important for the PSII recovery 14 . Our data from PSBS-depleted mutants are in line with these findings, as npq4-1 plants show sensitivity to high light similar to wild type. Accordingly, in a very recent study is reported that npq4-1 mutant, after 10 h irradiation with 1500 μmol photons m −2 s −1 showed a Fv/Fm ratio of about 0.48-0.50 whereas for the wild type the ratio was about 0.58-0.60 37 . However, the fact that high light irradiation, combined with Lincomycin treatment, led to the enhanced degradation of D1 in npq4-1 mutant, indicated that, in mutant background devoid of PSBS, the turnover of D1 is constitutively higher. A similar effect on D1 turnover in PSBS-less mutant was previously reported 38 . Thus, the loss of PSII activity is not observed as long as the rate of damage does not exceed the rate of repair 5,6 , indicating that the absence of PSBS-dependent NPQ is compensated by up-regulation of PSII repair. From these observations, we can suggest that the ability to engage a full NPQ might actually act as a signal aimed to regulate the D1 turnover, as well as a direct photoprotection mechanism meant to prevent D1 degradation. From a redox point of view, a reduced level of NPQ correlates with a higher accumulation of centers with a reduced Q A . As Q A − Q B is in equilibrium with Q A Q B − , it could be expected that in PSBS-less mutants a higher fraction of centers could accumulate a semi-reduced secondary quinone acceptor, which, according to previous works 39,40 , could play a role as a photosensitizer for enhanced degradation of the D1 protein. This could be the mechanism by which the turnover of D1 protein in the npq4-1 mutant is constitutively higher with respect to the wild-type, although it has to be taken into account that the high turnover rate of D1 and the accumulation of reduced Q A is not usually associated with reduced NPQ in wild-type leaves under physiological conditions. Furthermore, it should be noted that Y(NO) reflects the energy quenching processes occurring within PSII reaction center with Q A in a reduced state, and that the reduction of Q A has been suggested to be a major requirement and a prerequisite for an efficient PSII reaction centre quenching [41][42][43] . Therefore, the substantial increase of Figure 8. Thylakoid protein phosphorylation pattern. Thylakoid membranes were isolated from Col-0, single and multiple mutants, fractionated onto SDS-PAGE, transferred onto nitrocellulose membranes and probed with a polyclonal anti-phosphothreonine antibody. Levels of phosphorylation of CP43, D2, D1 and LHCII are shown over time (0-to-240 min) upon exposure to optimal growth light (GL) and high-light (HL) conditions. Lincomycin treatment was performed overnight in the dark where indicated (+Lin). One representative immunoblot (n = 3) for each genotype is shown. Y(NO) observed in the npq4-1-containing mutants, and in pgr mutants at moderate light intensities (see Fig. 3), also suggest the activation of PSII reaction center quenching, as a compensatory mechanism for an effective photoprotection, although this aspect is still debated.
In addition to that, mutants with defect in building up proper trans-thylakoidal pH gradient, such as pgr5 and pgrl1ab, show enhanced degradation of D1, similarly to PSBS-less mutant. In particular, in PGR-devoid mutants, the treatment with high light caused a strong inactivation of PSII, even in the absence of lincomycin. As they are able to engage about 40% of the NPQ observed in wild type and are much more sensitive to light than PSBS-less plants (where the extent of NPQ is near zero), we conclude that, at least in our experimental conditions, the PGR-dependent photosynthesis-control act as an efficient photoprotection mechanism.
Furthermore, unlike wild type and npq4-1 plants, the photosynthesis-control depleted mutants are not able to photo-accumulate P700 + , as their Y(ND) is near zero at any light intensity due to the low values of both thylakoid proton gradient (ΔpH) and proton motive force (pmf) they can develop 44 . At the same time, they are characterized by the overreduction of PSI acceptors, observed as an increase of Y(NA). Thus besides PSII, PSI is also photodamaged in these mutants, likely because of impairment of iron-sulfur clusters 13 . As no additive phenotypic effects are observed between the photosynthetic characteristics of ΔpH mutants (pgr5 and pgrl1ab) and the ones of higher order mutants (npq4-1 pgr5, npq4-1 pgrl1ab, ΔSTeM), it can be concluded that the short-term light adaptation is highly depending on the photosynthesis-control regulatory mechanism. Accordingly, mutants lacking of NDH-dependent CET such as crr2-2, crr-3, crr4-2 23 but still able to photo-accumulate P700 + , are more light resistant than the pgr5 mutant, deficient in CET and unable to photoaccumulate P700 in the oxidized form 25 .
It is noteworthy that PSI photodamage in PGR-depleted mutants can be markedly reduced through the inhibition of PSII activity, as a consequence of the fact that the amount of electrons injected in the intersystem transport chain is decreased. This indicates that the photosynthesis-control is the main regulator of photosynthetic electron transport and that PSII photoinhibition is the very last option to reduce PSI photodamage 15 . On the other hand, damages to PSI are relevant in inducing inactivation of PSII: the acceptor side of PSII becomes over-reduced and this, in turn, increases the rate of charge recombination with formation of 3 P680 35 and PSII inactivation. In addition, the absence of ΔpH-dependent photosynthesis-control affects the value of pmf 44 , and this could alter the electron transfer between Q A and Q B , promoting PSII inactivation.
Overall, it appears clear that the ΔpH-dependent photosynthesis control is essential for safeguarding the entire photosynthetic electron transport chain in the thylakoid membrane, and its failure induces a rapid and coordinated inactivation of both PSII and PSI. ΔpH-dependent photosynthesis-control thus maintains the optimal balance between the two main power-units of the photosynthetic apparatus.
Chlorophyll a fluorescence. In vivo chlorophyll a fluorescence and P700 absorbance were measured at different light intensities using the Dual-PAM 100 (Walz, http://www.walz.com/) as previously described 45 .The PSII Fv/Fm, Y(II), Y(NO), Y(NPQ) parameters, together with PSI yield [Y(I)], Pm, donor side Y(ND) and acceptor side Y(NA) limitations, were calculated as reported 53,54 . In the case of Fv/Fm parameter, measurements were performed after 20 min of dark adaptation. State transitions measurements were performed as previously described 55 .
In particular, state transitions were monitored on detached leaves with the DualPAM-100 fluorometer. Leaves were exposed to a 800 ms flash of saturating white light to determine F m , and subsequently illuminated for 15 min with 25 µmol photons m −2 s −1 red light (PSII light) directly in the PAM fluorometer. Far-red (PSI) light (intensity step 15) was turned on, and after 15 min the maximum fluorescence yield in state 1 (F m1 ) was determined. The far-red light was then switched off and the fluorescence recorded for 15 min, after which the maximum fluorescence yield in state 2 (F m2 ) was determined. The relative change in fluorescence was calculated as F r = [(F i′ -F i ) − ( F ii′ -F ii )]/(F i′ -F i ), where F i and F ii designate fluorescence in the presence of PSI light in state 1 and state 2, respectively, while F i′ and F ii′ designate fluorescence in the absence of PSI light in state 1 and state 2, respectively. Decay of flash-induced chlorophyll fluorescence was measured by the double modulation fluorometer FL-3500 (PSI, Brno, Czech Republic) and data were analyzed as described 35 . Multicomponent deconvolution of the measured curves was performed by using a fitting function with two exponential components and one hyperbolic component: F (t) -F 0 = A 1 exp(−t/T 1 ) + A2exp(−t/T 2 ) + A 3 /(1 + t/T 3 ) + A 0 . F (t) is the variable fluorescence yield, F 0 is the basic fluorescence before the flash, A 0 to A 3 are the amplitudes and T 1 to T 3 are the time constants. Very slowly decaying fluorescence is described by a constant A 0 amplitude. All measurements were performed after 20 min dark adaptation. electrochromic pigment absorbance shift measurements. In vivo electrochromic pigment absorbance shift (ECS) analyses were performed with a JTS-10 spectrophotometer (Biologic, France) on detached leaves, adapted to moderate-light (50 μmol photons·m −2 ·s −1 ) or treated with high-light (500 μmol photons·m −2 ·s −1 ) for 240 min. Leaf material was exposed to red actinic light (500 μmol photons m −2 s −1 ) for 5 min and ECS relaxation was measured during the light-to-dark transition. Data were collected as the difference between the signals at 520 and 546 nm as described by Cruz et al. 56 and Avenson et al. 57 . The amplitude of the ECS signal was normalized to the signal corresponding to one PSI + PSII charge separation, calculated after the application of xenon-induced ECS signals. The pmf was evaluated following ECS relaxation kinetics after actinic light switch off. Data analysis. Data (statistics analysis, data fitting) were analysed by using the software package OriginPro 9.0 (Microcal SR2; Northampton MA01060 USA).