Consequences of the reduction of the Photosystem II antenna size on the light acclimation capacity of Arabidopsis thaliana

Abstract In several systems, from plant's canopy to algal bioreactors, the decrease of the antenna size has been proposed as a strategy to increase the photosynthetic efficiency. However, still little is known about possible secondary effects of such modifications. This is particularly relevant because the modulation of the antenna size is one of the most important light acclimation responses in photosynthetic organisms. In our study, we used an Arabidopsis thaliana mutant (dLhcb2), which has a 60% decrease of Lhcb1 and Lhcb2, the two main components of the major Photosystem II antenna complex. We show that the mutant maintains the photosynthetic and photoprotective capacity of the Wild Type (WT) and adapts to different light conditions by remodelling its photosynthetic apparatus, but the regulatory mechanism differs from that of the WT. Surprisingly, it does not compensate for the decreased light‐harvesting capacity by increasing other pigment‐protein complexes. Instead, it lowers the ratio of the cytochrome b6f and ATP synthase to the photosystems, regulating linear electron flow and maintaining the photosynthetic control at the level of these complexes as in the WT. We show that targeting the reduction of two specific antenna proteins, Lhcb1 and Lhcb2, represents a viable solution to obtain plants with a truncated antenna size, which still maintain the capacity to acclimate to different light conditions.


| INTRODUCTION
To meet the demand of the growing world population, the food production needs to be doubled until the end of this century, knowing that there are scarce prospects for the expansion of the cultivatable area (UN, 2015;Kline et al., 2017). To reach this goal, an increase in photosynthetic capacity of crops is needed (reviewed in Long, Marshall-Colon, & Zhu, 2015).
A theoretical upper limit for the operational efficiency of plant photosynthesis from light capture to carbohydrate synthesis has been calculated to be about 4.6% for C 3 plants (Zhu, Long, & Ort, 2008;Zhu, Long, & Ort, 2010). A decrease of the light-harvesting capacity of the plants has been proposed as one of the possible strategies to maximize photosynthetic efficiency. It enables a more even distribution of light within the canopy, decreasing energy dissipation from the over-illuminated top leaf layers and increasing the fraction of light energy reaching the under-illuminated leaf layers deeper in the canopy (Melis, 2009;Ort et al., 2015;Ort, Zhu, & Melis, 2011;Song, Wang, Qu, Ort, & Zhu, 2017). The same principle can be applied to bioreactors, where a smaller antenna size of the photosystems decreases the over-absorption of light by the cells closest to the light source, enabling deeper sunlight penetration into the culture. An increase of biomass production in mutants with truncated antenna size has been already validated for cyanobacteria (Kirst, Formighieri, & Melis, 2014;Nakajima & Ueda, 1997;Nakajima & Ueda, 1999), Chlamydomonas reinhardtii (Kirst et al., 2012, b;Mussgnug et al., 2007;Polle, Kanakagiri, & Melis, 2003), species of the genus Botryococcus (reviewed in Melis, 2012) and tobacco plants (Kirst et al., 2018;Kirst, Gabilly, Niyogi, Lemaux, & Melis, 2017).
In plants, the transformation of light energy into chemical energy is performed by four major transmembrane protein complexes embedded in the thylakoid membrane: Photosystem II (PSII), cytochrome b 6 f, Photosystem I (PSI), and ATP synthase. PSII and PSI are multiprotein complexes composed of pigment-binding proteins involved in light harvesting (antennae) and charge separation (reaction centre [RC]).
Cytochrome b 6 f mediates and regulates the electron transfer between PSII and PSI. It oxidizes plastoquinol (PQH 2 ) to plastoquinone (PQ) releasing the protons in the lumen and reduces the electron carrier that links cytochrome b 6 f with PSI: Plastocyanin (PC). It regulates the redox state of PQ and PC and contributes to the creation of the ΔpH across the thylakoid membrane (reviewed in Schöttler, Tóth, Boulouis, & Kahlau, 2015). Due to the lumen-pH-dependent rate of reoxidation of PQH 2 (the so-called "photosynthetic control"), the electron flow (EF) through the complex changes substantially under different conditions (Joliot & Johnson, 2011;Stiehl & Witt, 1969).
Finally, the fourth complex, the ATP synthase, utilizes the protonmotive force (pmf) to synthesize ATP. The process is highly regulated (Kanazawa & Kramer, 2002), and the dissipation rate of the pH gradient can influence the EF through photosynthetic control.
Truncated antenna mutants are available since a long time in several species (see Melis, 1991). The reduction of the antenna size can be obtained in several ways: (a) by targeting proteins involved in the import in the chloroplast (e.g., Kirst et al., 2018) or (b) in the assembly of the antennae proteins with pigments (e.g., Ghirardi, McCauley, & Melis, 1986;Hansson, Kannangara, Wettstein, & Hansson, 1999;Sakowska et al., 2018) or (c) by targeting individual LHCs (e.g., Andersson et al., 2003;Dall'Osto et al., 2017;Dall'Osto, Ünlü, Cazzaniga, & Van Amerongen, 2014;de Bianchi et al., 2011;de Bianchi, Dall'Osto, Tognon, Morosinotto, & Bassi, 2008;Pietrzykowska et al., 2014). Although all those methods produced plants with a truncated antenna, the effect on the biomass production varied (e.g., Kirst et al., 2018;Sakowska et al., 2018). A possible reason for these apparently contradictory results is that those mutations induce pleiotropic effects hindering the capacity of the plants to perform and/or regulate photosynthesis. For example, mutants lacking the antennae completely were shown to become damaged in high light (Ramel et al., 2013). Mutants lacking the minor antennae have a bad connection between PSII and the core, which decreases the PSII trapping efficiency (Dall'Osto et al., 2014;Van Oort et al., 2010). The partial reduction of the antenna achieved in an untargeted way also produced contrasting results (Kirst et al., 2017;Slattery, Vanloocke, Bernacchi, Zhu, & Ort, 2017), suggesting that the difference can be due to secondary effects induced by the mutation.
A way to get around these issues is to specifically target a protein directly composing the LHCIIs, which are located at the periphery of the supercomplex, and the absence of which should thus not influence the efficiency of energy delivery to the RC. However, because LHCII modulation is a major acclimation strategy of plants (Anderson et al., 1995;Ballottari et al., 2007;Bielczynski et al., 2016;Wientjes, Van Amerongen, & Croce, 2013a), it was unclear if the absence of a part of the LHCIIs would influence the acclimation capacity of the plant. To answer this question, we challenged a mutant having a smaller pool of LHCII  to grow under different light conditions. Combining biochemical and functional measurements, we show that a targeted reduction of the LHCII pool is a viable strategy to decrease the antenna size without introducing secondary effects that negatively impact its performance.

| Leaf absorption measurements
Leaf absorption was measured using a Cary 4000 UV-Vis spectrophotometer with a mounted integrating sphere. Each leaf was carefully placed to ensure the same light cross-section was obtained and that the whole measuring beam was passing through the leaf.

| Thylakoid isolation
The isolation procedure was described in Robinson, Sharp, and Yocum (1980) with modifications from Caffarri et al. (2009). The isolation was performed on over-night dark-adapted plants. Isolated thylakoid membranes were resuspended in 20-mM HEPES, pH 7.5, 0.4-M sorbitol 15-mM NaCl, 5-mM MgCl 2 and stored until further use at −80 C, after rapid freezing in liquid nitrogen.

| Pigment isolation
The Chl a/b ratio and Chl/carotenoid ratio were determined from the absorption spectra of 80% acetone extracts. The absorption spectra were fitted with the spectra of individual pigments in the same solvent, as described in Croce, Canino, Ros, and Bassi (2002). The quantification of different carotenoids was performed by HPLC using a System Gold 126 Solvent module and 168 Detector (Beckman Coulter) as described by Gilmore and Yamamoto (1991) with the modification reported in Xu, Tian, Kloz, and Croce (2015).

| Functional antenna size of PSII and PSI
PSII functional antenna size measurements were performed as in Dinç, Ceppi, Tóth, Bottka, and Schansker (2012) with the modifications described in Bielczynski et al. (2016). PSI functional antenna size was measured as described by Takahashi, Clowez, Wollman, Vallon, and Rappaport (2013) with some modifications. After infiltrating the leaves with 200-μM DCMU (3-(3,4-dichlorophenyl)-1,1-dimethylurea) and 1-mM HA and effectively inhibiting the PSII (F V /F M < 0.1), the changes in the absorption at 520-546 nm, from electrochromic shift (ECS) generated by only PSI RCs, were measured in continuous red light (emission peak at 630 nm) of 300 μmol photons m −2 s −1 with a JTS-10 (BioLogic). The slope of the first 10 ms of illumination was used to estimate the functional antenna size of PSI.

| ATP synthase activity
The ATP synthase activity was monitored by measuring the proton conductivity as in Cruz, Sacksteder, Kanazawa, and Kramer (2001).
We used a JTS-10 (BioLogic) and measured the ECS signal decay (520-546 nm) following a 50 s red light (630 nm) illumination (300 μmol photons m −2 s −1 ). Subsequently, the signal was double normalized. The quasi-steady-state level at the end of the illumination period was defined as zero and the signal level 400 ms after the lightto-dark transition was defined as one.

| Cytochrome b 6 f resistance measurements
Cytochrome b 6 f resistance (Harbinson & Hedley, 1989;Harbinson & Hedley, 1993;Ott, Clarke, Birks, & Johnson, 1999) was measured with a Dual-PAM-100 fluorometer (Walz). The redox changes of P700 were followed by recording the difference between the absorption changes at 830 and 875 nm. At the beginning of the measurement, ΔA max was measured from the absorption changes before and during the SP according to Klughammer and Schreiber (1994). After 8-min illumination with red light (630 nm) of 1,000 μmol photons m −2 s −1 , the PQ pool was mostly reduced, and the PSI RCs were mostly oxidized. After switching off the light the electrons were transferred from the PQ pool via cytochrome b 6 f to oxidized PC and P700, the RC of PSI. Because re-oxidation of PQH 2 by cytochrome b 6 f is the rate-limiting step, the rate of re-reduction of P700 becomes an estimate of the cytochrome b 6 f resistance. The P700 + re-reduction kinetics was normalized to the absorption minimum at 50 ms after the AL was switched off, and rescaled to ΔA max , to allow a comparison of the kinetics of the different measurements.

| Quenching analysis
Chl fluorescence and P700 absorption (830-875 nm) changes were measured simultaneously on intact leaves, at room temperature, with a DUAL-PAM-100 fluorometer (Walz). The saturating pulse (SP) intensity was 5,000 μmol photons m −2 s −1 and the measuring light (ML) intensity 3 μmol photons m −2 s −1 . The F 0 was assessed with a weak ML after dark acclimation and F M as the peak-value during a 500-ms long SP. After a subsequent 15-s far-red (FR) illumination (89 μmol photons m −2 s −1 ), the ΔA max was measured from the absorption changes before and during the SP according to Klughammer and Schreiber (1994). To characterize the dark-to-light transition, the plants were light acclimated for 8 min with a selected actinic light (AL) intensity (200, 600, or 1,000 μmol photons m −2 s −1 ), where every 20 s, an SP with a 5 s FR pre-illumination was applied to the sample to determine F M 0 and a 2-s period of darkness to measure and calculate ΔA sat . Before each SP, F t was measured (ML + AL). After switching off AL, the light-to-dark transition was recorded. During the recovery period, every 50 s, an SP was applied to the sample, and the same parameters were determined. From the fluorescence measurements, calculated, based on Butler and Kitajima (1975), Genty, Briantais, and Baker (1989) and Klughammer and Schreiber (2008), as follows

| Irradiance curve
During irradiance curve measurements, as well as during quenching analysis measurements, we used two-channel mode and the settings for ML and SP were kept the same. However, dark-acclimated plants were first pre-illuminated for 10 min with 53 μmol photons m −2 s −1 AL to activate Ferredoxin NADP+ oxidoreductase (FNR). Afterward, the plants were acclimated for 60 s to 10 different AL intensities, in the range from 0 to 1,288 μmol photons m −2 s −1 . After each acclimation period, an SP was applied to the samples, and the same parameters as during the quenching analysis were determined. Additionally, ETR II , qP, and P700 red were calculated. The qP parameter (Genty et al., 1989) is based on the qQ parameter from Schreiber, Schliwa, and Bilger (1986).
where F 0 0 is calculated according to the Oxborough and Baker (1997) approximation.
As for ETR II , we calculated it according to Genty et al. (1989) with some modifications: where I is the incident light intensity, δ is the correction factor for the fraction of light that did not reach PSII as it got absorbed by PSI and α is the correction for the antenna size of PSII.
The P700 red is calculated in the following manner: where Φ PSI and Φ NA are the quantum yields of PSI and acceptor side limitation of PSI, respectively.

| WT and dLhcb2 comparison
When grown under standard conditions (GL200), the dLhcb2 plants showed a higher Chl a/b and a lower Chl/Car ratio than the WT plants. The mutant also contained around 20% less Chls per fresh weight than the WT; it had a slightly lower F V /F M value and a lower NPQ level. These data are similar to those observed before for the same mutant .

| Thylakoid membrane composition
In the mutant, the silencing of Lhcb2 should results in a decrease in the absorption cross-section of PSII. However, it was shown by The PSI/PSII (core) ratio was the same in the WT and in the mutant (p = .6), whereas the Lhcb1 and Lhcb2 content dropped to 40% of the WT value in the mutant (p < .001; Figure 1b). Surprisingly, the amount of these two proteins in the mutant was far higher than reported previously by Andersson et al. (2003) and Ruban et al. (2003), probably due to the instability of the RNA silencing. Interestingly, we also observed a decrease of cytochrome b 6 f and ATP synthase as compared with the WT (to 60% and 80%, respectively, with p < .05; Figure 1b).
In agreement with the decrease in the content of Lhcb1 and Lhcb2, the native gel showed that the amount of PSII supercomplexes was lower in the mutant. Indeed, in the WT, most of the PSII was organized in supercomplexes (~70%) and the rest in the form of core monomers and dimers (Figure 1c), in dLhcb2, those proportions were inverted: <40% of PSII was in the form of supercomplexes, >40% in the form of core monomers, and around 20% in the form of core dimers. However, the protein composition of the supercomplexes was the same in WT and mutant and contained the antennae CP24, CP26, CP29, Lhcb3, and Lhcb1/Lhcb2.

| PSII antenna composition
It was previously shown that the absence of Lhcb1 and Lhcb2 was partially compensated by an increase of CP26 . To determine the PSII antenna composition, we separated the thylakoid membrane proteins in 1D-PAGE (TT-SDS PAGE; Figure 2). As some of the protein bands were slightly overlapping in the gel, we used nonlinear least squares to fit the integrated optical density of the lanes with a sum of Gaussians.
This analysis confirms the decrease of Lhcb1 and Lhcb2 in dLhcb2 as compared with the WT (Table 2). We verified it while normalizing the number of antenna proteins to CP29: Lhcb3 also decreased, but all three major components of LHCII were still present in the plants.
The decrease in Lhcb1 and Lhcb2 was not compensated by an increase of CP26 because this protein did not differ significantly in the WT and in the dLhcb2 mutant.

| Energy partitioning in PSII
To observe how the changes in the protein composition influenced the photosynthetic and the photoprotective capacity of the mutant, we performed a quenching analysis (Figure 3). The effective quantum yield of PSII (Φ PSII ) was hardly affected in the mutant. The fast induction phase of NPQ was identical to that of the WT whereas the second phase was slightly slower in the mutant (Figure 3a). A small  T A B L E 2 Quantification of PSII antenna proteins. Based on the Gaussian fit performed on TT-SDS PAGE IOD profiles of isolated thylakoids from WT and dLhcb2 (as shown in Figure 1), we estimated the amount of CP24, CP26, Lhcb1 + 2, and Lhcb3. The amount of antennae was normalized to the amount of CP29. The quantification was performed based on three or five rounds of PAGE (for dLhcb2 and WT, respectively) with four technical replicates each. With triple asterisks are shown statistically significant groups which means they are different from the WT, based on the post hoc Tukey HSD test paler compared with the WT, in agreement with a decreased Chl content ( Figure 4 and Table 3).

| Pigment composition
In the WT, the exposure to higher light intensities led to a decrease in Chl content. In dLhcb2, the concentration of Chls still decreased with increasing GL intensity (Table 3), although less than in the WT.
The Chl a/b ratio increased in the WT at higher light intensities as observed before (

| Thylakoid membrane composition
To check for changes in the composition of the thylakoid membrane of the mutant upon acclimation to a different light intensity, we performed BN-PAGE (Figure 5a) followed by a 2D-PAGE analysis ( Figure S1, supporting information) as described above.
For the WT, it was shown before (Bielczynski et al., 2016) that the amount of Lhcb1 and Lhcb2 per PSII core decreased when the plants were grown under increasing light intensities (p < .05). At variance with this, in dLhcb2, under all light intensities, the amount of Lhcb1 and Lhcb2 per PSII core remained the same (p = .97), and it was always lower than in the WT, even at very high light intensities.
T A B L E 3 Pigment analysis of WT and dLhcb2 grown under different light conditions. Total chlorophylls (Chls) were quantified by fitting the absorption spectra of individual pigments to the spectrum of the 80% acetone extracts from a leaf from five different plants (n = 5) grown under GL200, 600, and 1800, and normalized to the fresh weight. The Chl a/b ratio and chlorophyll/carotenoid (Chl/Car) ratio were determined in the same way from isolated thylakoid membranes in three repetitions (n = 3). The same extracts were used for the quantification by HPLC of the carotenoids: neoxanthin (Neo), violaxanthin (Vio), lutein (Lut), and β-carotene (β-Car

| Functional antenna size of the photosystems
Because no changes in the LHCII/PSII ratio at the protein level in the mutant grown in different light intensities were observed, we then looked for possible changes in the functional antenna size (Figure 6b). The PSII functional antenna size of the mutant was around 40-50% smaller than that of the WT at GL200 (Figure 6c). Moreover, although in the WT, the functional antenna size decreased with light intensity, in the mutant, the differences were small (~10%), in line with the protein composition.
We also tested the PSI functional antenna size in WT and mutant after acclimation to the three different light intensities (Figure 6a).  Figure 6c). In the mutant, no changes were detected.

| Cytochrome b 6 f resistance
To determine if the large changes in the relative ratio of cytochrome b 6 f and PSI observed during acclimation in the dLhcb2 mutant had functional consequences for the EF, we measured the cytochrome b 6 f resistance. This was done by following P700 + re-reduction during a light-to-dark transition (Figure 7a). The starting absorption level was slightly different, especially in the dLhcb2 grown under different light intensities. However, the kinetics of WT and dLhcb2 plants was the same for all growth conditions. We verified this hypothesis by fitting the traces with mono-exponential decay function ( Figure S2A) and comparing the values of the decay parameter ( Figure S2B). Additionally, we have checked the P700 red after illumination under different AL intensities ( Figure S3). We can conclude that differences in the amount of cytochrome b 6 f between WT and mutant did not influence the EF.
F I G U R E 5 Quantification of thylakoid complexes of dLhcb2 and Wild Type (WT) plants long-term acclimated to different light intensities.
(a) The BN-polyacrylamide gel electrophoresis of thylakoid membranes isolated from dLhcb2 plants grown under GL200, GL600, and GL1800, solubilized with 1% n-dodecyl-α-D-maltoside (α-DDM). WT grown under GL200 was used as a control. (b) Relative quantification of the light phase photosynthetic complexes from a 2D-polyacrylamide gel electrophoresis based on three independent repetitions (n = 3) on WT and dLhcb2 (Panels a and b, respectively) grown under GL200, GL600, and GL1800 (yellow, orange, and brown fill, respectively). The data were normalized to PSII and to WT GL200. Numbers above the bars are the p-values of ANOVA's F test on a specific group. Red signifies rejection of the null hypothesis that the response to all growth lights is the same (α < .05), black its acceptance. GL, growth light; IOD, integrated optical density; LHC, light-harvesting complex; PSI, Photosystem I; PSII, Photosystem II [Color figure can be viewed at wileyonlinelibrary.com]

| ATP synthase activity
To assess the effect of the changes in the amount of ATP synthase relative to PSI in the dLhcb2 plants, we measured the ATP synthase conductivity by monitoring the dissipation of the pmf by following the disappearance of the ECS signal during a light-to-dark transition ( Figure 7b). During illumination, the proton and ion gradients across the thylakoid membrane create a pmf that is released through the ATP synthase (Cruz et al., 2001;Witt, 1979). The ECS decays did not differ between the WT and dLhcb2, and between different GL intensities.

| Short-term responses (irradiance curve)
Next, we investigated the photosynthetic and photoprotective capaci- First, to compare the EF threshold needed for NPQ activation, we followed the changes of NPQ as a function of EF through PSII ( Figure 8a). For NPQ and EF, we monitored Φ NPQ and ETR II , respectively. To get slightly closer to the reality, instead of using the simplified equation for ETR II (e.g., White & Critchley, 1999), we performed two additional corrections: (a) for the light interception by PSI and PSII, we used the relative values of PSI/PSII RC ratio from the biochemical data. On the basis of Hogewoning et al.
(2012), we assumed 0.73 as the absolute ratio for WT GL200; (b) we used the relative interception area of PSII (from the functional antenna size normalized to WT GL200) as a multiplicative scaling factor.
The kinetics was always sigmoidal as there is an EF threshold level at which ΔpH is large enough or the pH in the lumen is low enough to trigger NPQ. The ETR II value necessary to trigger NPQ increased for plants grown at higher light intensities.
When dLhcb2 was grown under 200 or 600 μmol photons m −2 s −1 , the EF threshold of NPQ was similar to that of the WT. Only for HL-grown plants, the threshold of the mutant was shifted towards higher ETR II compared with the WT.
F I G U R E 6 Functional antenna size of PSII and PSI. (a) Relationship between fluorescence intensity at 300 μs normalized to PPFD, and PPFD fitted with linear regression. The fitted lines with their standard error are shown as lines with shadows. Individual data points are from the WT (data from Bielczynski et al., 2016) and dLhcb2 (top and bottom panels respectively) plants grown under GL200, GL600, and GL1800 (yellow, orange, and brown, respectively). The functional antenna size of PSII was measured on 10 different leaves (n = 10). The asterisk indicates data from an experiment carried out in parallel on WT described in Bielczynski et al. (2016). whereas P700 red remained high, this was not the case in the mutant.
In general, in dLhcb2, the linear relationship was maintained, suggesting a comparable EF through both PSs.

| DISCUSSION
Antenna truncation has been proposed as a strategy to increase crop productivity because it would allow a better light distribution in the canopy (Melis, 2009;Ort et al., 2011;Ort et al., 2015). Although this is a promising strategy, it remained to be seen if the reduction of the antenna had negative effects on the functionality and acclimation capacity of the photosynthetic apparatus. Indeed, most of the truncated antenna mutants analysed so far show secondary effects due to a lack of connectivity between the remaining antennae and the core or decreased photoprotection, which counterbalanced the positive effects (van Oort et al., 2010;Miloslavina et al., 2011;Ramel et al., 2013;Dall'Osto et al., 2014). These data clearly show that the absence of some of the minor antenna complexes or of the complete peripheral antenna pool do not represent viable solutions. Mutants showing a partial reduction of the antenna were also analysed (Kirst et al., 2017;Kirst et al., 2018;Slattery et al., 2017). Some of those (Kirst et al., 2017;Kirst et al., 2018) showed an increase in biomass production, whereas others did not (Slattery et al., 2017). These mutants were generated with an untargeted approach (Kirst et al., 2017;Slattery et al., 2017) or targeting a protein that affect chloroplast import (Kirst et al., 2018), which results in the downregulation of several chloroplast components (e.g., Kawata & Cheung, 1990). These F I G U R E 7 Resistance of cytochrome b 6 f and ATP synthase conductivity. (a) The resistance of the cytochrome b 6 f was measured by the P700 + re-reduction (absorption at 830-875 nm), during the light-to-dark transition. Absorption changes were measured on intact leaves, coming from five different plants (n = 5) of WT and dLhcb2 (red and blue traces, respectively), grown under growth light (GL)200, GL600, and GL1800 (respectively the top, middle, and bottom panels). Traces were normalized to the minimum at 50 ms of decay and rescaled to the maximum absorption change measured at the beginning of the measurement (details in the Section 2). Shadows show the standard deviation of the measurements. (b) ATP synthase activity was followed by measuring the ECS decay kinetics (at 520 and 546 nm) after a 50-s red light illumination on leaves coming from five different plants (n = 5) of WT and dLhcb2 (red and blue traces, respectively) grown under GL200, GL600, and GL1800 (respectively the top, middle, and bottom panels). Traces were double normalized (details in the Section 2). Shadows show the standard deviation of measurements [Color figure can be viewed at wileyonlinelibrary.com] results suggest that a more targeted approach, which permits a better control of the changes in chloroplast proteins can be advantageous.
For example, the selective reduction of the peripheral antenna complexes (LHCII), the absence of which should not influence excitation energy transfer efficiency to the core, seems to be a promising target. Interestingly, although in the mutant, the PSI/PSII ratio was the same as in the WT, the amounts of both cytochrome b 6 f and ATP synthases decreased. These complexes are crucial for maintaining linear EF and for the photoprotective regulation of the photosynthetic apparatus (reviewed in Colombo et al., 2016). The reduction of both complexes in the mutant seems thus to be a strategy to maintain the photosynthetic control at the level of the cytochrome b 6 f in the presence of a smaller absorption cross-section of the photosystems. The lower amount of ATP synthase permits a slower dissipation of ΔH + across the membrane (reviewed in Schöttler et al., 2015) and as a consequence enables the chloroplasts to keep high levels of NPQ.