Photosystem II antenna complexes CP26 and CP29 are essential for nonphotochemical quenching in Chlamydomonas reinhardtii

Abstract Photosystems must balance between light harvesting to fuel the photosynthetic process for CO2 fixation and mitigating the risk of photodamage due to absorption of light energy in excess. Eukaryotic photosynthetic organisms evolved an array of pigment‐binding proteins called light harvesting complexes constituting the external antenna system in the photosystems, where both light harvesting and activation of photoprotective mechanisms occur. In this work, the balancing role of CP29 and CP26 photosystem II antenna subunits was investigated in Chlamydomonas reinhardtii using CRISPR‐Cas9 technology to obtain single and double mutants depleted of monomeric antennas. Absence of CP26 and CP29 impaired both photosynthetic efficiency and photoprotection: Excitation energy transfer from external antenna to reaction centre was reduced, and state transitions were completely impaired. Moreover, differently from higher plants, photosystem II monomeric antenna proteins resulted to be essential for photoprotective thermal dissipation of excitation energy by nonphotochemical quenching.

Gerotto, Giacometti, Bassi, & Morosinotto, 2010;Jansson, 1999). All antenna proteins are pigment-binding proteins encoded by nuclear genes and share a similar structure, with three transmembrane domains and two amphipathic α-helices (Matteo Ballottari, Girardon, Dall'Osto, & Bassi, 2012;Liu et al., 2004). In Chlamydomonas reinhardtii, PSII supercomplex structure has been observed by electron microscopy, revealing two layers of antenna proteins surrounding a homodimeric core complex (Tokutsu, Kato, Bui, Ishikawa, & Minagawa, 2012): The inner layer is composed of CP29, CP26, and S-LHCII (S, strongly bound), forming with the core the C 2 S 2 particle; the outer layer is made of M-LHCIIs (moderately bound) forming, together with C 2 S 2 , the C 2 S 2 M 2 complex. M-LHCIIs are expected to be connected to the core through CP29. In C. reinhardtii, an additional LHCII (N-LHCII) can be connected directly to the PSII core in the position occupied by CP24 in higher plants, forming the larger C 2 S 2 M 2 N 2 complexes (Tokutsu et al., 2012). PSII antenna proteins are also involved in the State 1-State 2 transition (Allen, 1992), which regulates redistribution of excitation energy pressure between the two photosystems by disconnecting part of the LHCII trimers from PSII, which are then possibly transferred to PSI (Tokutsu, Iwai, & Minagawa, 2009).
Photosynthetic organisms have evolved different mechanisms to protect themselves from harmful excess light; the fastest of which is the thermal dissipation of excitation energy by nonphotochemical quenching (NPQ), a dissipative mechanism triggered by lumen acidification when the photosynthetic apparatus is saturated (Genty, Briantais, & Baker, 1989). In C. reinhardtii, the light harvesting complex stress-related (LHCSR) subunits 1 and 3 are pigment-binding LHC-like proteins that sense luminal pH and trigger NPQ switching to a dissipative state (Girolomoni et al., 2019;Peers et al., 2009). In vascular plants, LHCSR is substituted by the PsbS subunit, which, however, does not bind pigments (Li et al., 2004). PsbS is also present in C.
reinhardtii, but it is accumulated only transiently upon high-light or ultraviolet exposure, and its contribution to NPQ is limited (Allorent et al., 2016;Correa-Galvis et al., 2016;Tibiletti, Auroy, Peltier, & Caffarri, 2016). Both LHCSR proteins are overexpressed on prolonged high light treatment, but the molecular mechanisms at the base of NPQ and the interaction of LHCSR subunits with PSII supercomplexes are still not clear: LHCSR3 has been suggested to be bound to CP26 and/or to an LHCII trimer (Semchonok et al., 2017) suggesting a possible key role of monomeric subunits in the energy pathway from PSII to LHCSR subunits during NPQ induction. Moreover, it has been recently reported that phosphorylation of CP29 in C. reinhardtii is linked to LHCSR3 phosphorylation, being this process likely related to PSII assembly and repair in high light (Scholz et al., 2019). A knock-out mutant without monomeric subunits was obtained in Arabidopsis thaliana, showing only a slower NPQ induction (Dall'Osto et al., 2017;Townsend et al., 2018). In microalgae, only RNA interference mutants of C. reinhardtii with individual monomeric subunits knocked down are reported in literature, revealing a peculiar role of CP29 in state transitions (Tokutsu et al., 2009), but the role on CP26 and CP29 in NPQ has never been investigated yet.
In this paper, we present the characterization of a C. reinhardtii double mutant, obtained with CRISPR-Cas9 technology, completely depleted of CP26 and CP29 subunits. This mutant presents reduced photosynthetic efficiency and impaired state transitions. Surprisingly, the lack of CP26 and CP29 completely abolishes NPQ induction even in the presence of LHCSR1 and LHCSR3 subunits, implying that LHCSR needs the interaction with monomeric PSII antenna in order to perform its quenching function.
2 | MATERIALS AND METHODS 2.1 | Strains and culture conditions C. reinhardtii wild type (Wt; CC503) and mutant strains were grown at 24°C in high-salts (HS) medium (Harris & Harris, 2008) on a rotary shaker in Erlenmeyer flasks under continuous illumination with white LED light at 100-μmol photons m −2 s −1 . High light acclimation was induced by growing cells for 2 weeks at 500-μmol photons m −2 s −1 in HS.

| Generation of C. reinhardtii mutants by CRISPR-Cas9 genome editing
Delivery of the DNA-free CRISPR-Cas9 RNP complex into cell was performed according to Baek et al. (2016) and Shin et al. (2019) with a few modifications. In brief, Cas9 protein (100 μg; ToolGen, South Korea) and in vitro transcribed sgRNA (70 μg), which was synthesized by using GeneArt™ Precision gRNA Synthesis Kit following the manufacturer's protocol (ThermoFisher, MA, USA), were mixed to form the RNP complex. The premixed RNP complex with and/or without aph7 DNA cassette was introduced to the cells by Biorad Gene Pulser Xcell™ electroporation system (the aph7 gene was prepared by PCR amplification from pChlamy3 vector with specific primer sets [F: ATGATTCCGCTCCGTGTAAATG, R: AGTACCATCAACTGACGTTAC ATTC]). After CRISPR-Cas9 transformation, cells were incubated in TAP liquid medium supplemented with 40-mM sucrose for 12 hr and harvested for genotype characterization or immediately diluted (2 × 10 3 cells) and plated on TAP medium containing 1.5% agar to obtain single colonies. In the case of k9 mutant strains, the colonies were screened by Fv/Fm fluorescence signal using a Walz Imaging PAM System (M-series). For the selection of CRISPR-Cas9-mediated antibiotic resistance knock-in-derived mutants, cells were plated on TAP medium with hygromycin (25 μg/ml). Putative mutants screened were further analysed by Sanger sequencing to confirm the indel mutations or antibiotic resistance insertion. In the case of knock-in mutants, single insertion events were confirmed by southern blot analysis.
2.3 | Isolation of genomic DNA and complementary DNA preparation and quantitative reverse transcription PCR Healthcare, Chicago, IL, USA), and cross-linked using shortwave ultraviolet light (254 nm). The probe was designed to cover 467 bp of the hygromycin resistance gene in pChlamy3 vector and was amplified from genomic DNA by PCR with the forward primer 5′-ATGATTCCT ACGCGAGCCTG-3′ and reverse primer 5′-ATCCGGCTCATCACCA GGTA-3′. The amplified probe was labelled with alkaline phosphatase using the Gene Images AlkPhos Direct Labeling and Detection System kit (GE Healthcare, Chicago, IL, USA). Labelling, hybridization, washing, and signal detection were conducted according to the manufacturer's protocol. Total RNA was extracted from cells in the exponential phase with RNeasy Plant Mini Kit (Qiagen, Hilden, Germany). Complementary DNA (cDNA) was synthesized from total RNA by using 2X reverse transcription master premix (ELPiS Biotech, Daejeon, Korea) and amplified using SYBR premix (Takara, Kusatsu, Japan) in a Thermal Cycler Dice Real Time System TP 8200 (Takara, Kusatsu, Japan). The gene for the "receptor for activated C kinase 1 (RACK1)" was used as a reference and was amplified with the forward primer 5′-GGCT GGGACAAGATGGTCAA-3′ and reverse primer 5′-GAGAAGCACAG GCAGTGGAT-3′.

| Pigment analysis
Chlorophyll and carotenoids content were analysed by highperformance liquid chromatography upon pigment extraction in 80% acetone as described in Lagarde, Beuf, and Vermaas (2000).

| Spectroscopy and fluorescence
Absorption spectra of pigments, extracted from intact cells using 80% acetone buffered with Na 2 CO 3 , were measured with Jasco V-550 ultaviolet-visible spectrophotometer. Spectra were fitted as previously described with different pigment absorption spectra (Croce, Canino, Ros, & Bassi, 2002). Low-temperature fluorescence spectra were performed on frozen samples with BeamBio custom device equipped with USB2000+ spectrometer (OceanOptics) and 475-nm LED light sources for excitation.

| Photosynthetic parameters and NPQ measurements
Photosynthetic parameters ΦPSII, qL, electron transport rate (ETR), and NPQ were obtained by measuring with a DUAL-PAM-100 fluorimeter (Heinz-Walz) chlorophyll fluorescence of intact cells, at room temperature in a 1 × 1-cm cuvette mixed by magnetic stirring. ΦPSII, qL, and ETR were measured and calculated according to Baker (2008) and Van Kooten and Snel (1990) at steady-state photosynthesis upon 20 min of illumination. NPQ measurements were performed on dark-adapted intact cells, with a saturating light of 4,000-μmol photons m −2 s −1 and actinic lights from 150-to 3,900-μmol photons m −2 s −1 . PSII functional antenna size was measured from fast chlorophyll induction kinetics induced with a red light of 11-μmol photons m −2 s −1 on dark-adapted cells (~2·10 6 cells/ml) incubated with 50-μM DCMU (Malkin, Armond, Mooney, & Fork, 1981). The reciprocal of time corresponding to two thirds of the fluorescence rise (τ 2/3 ) was taken as a measure of the PSII functional antenna size (Malkin et al., 1981). Proton motive force upon exposure to different light intensities was measured by electrochromic shift (ECS) with MultispeQ V2.0 (PhotosynQ) according to Kuhlgert et al. (2016). State transitions were induced with the protocol present in . State transitions were estimated as (Fm PSIS2 − Fm PSIS1 )/Fm PSIS2 where Fm PSIS1 and Fm PSIS2 are the maximal PSI fluorescence in State 1 and State 2, respectively. The oxygen evolution activity of the cultures was measured at 25°C with a Clark-type O 2 electrode (Hansatech), during illumination with light from a halogen lamp (Schott). Measurements were performed in 1-ml cell suspension concentrated at 5 × 10 6 cell/ml.

| 77-K fluorescence
Low-temperature quenching measures were performed according to Girolomoni et al. (2019). Cells were frozen in liquid nitrogen after being dark adapted or right after 10 min of illumination at 1,200-μmol photons m −2 s −1 , and fluorescence emission spectra were recorded. Green fluorescent protein was added to all samples as an internal standard for fluorescence emission spectra normalization.

| Statistics analysis
All the experiments herein described were performed at least on two independent lines for each mutant genotype in at least four independent biological experiments. Statistical analysis was performed by using two-sided Student's t test.

| Data and strains availability
The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information files.
All the strains described in this work are fully available upon request to the corresponding author.
Thylakoid membrane polypeptides isolated from Wt and k9 were separated on a denaturing gel ( Figure 1). k9 mutants lacked the band corresponding to CP29, and the complete absence of the protein was confirmed by western blot analysis using a specific antibody ( Figure 1d). The cp26 gene was subsequently targeted in Wt and k9.2 strains: Two lines selected (for hygromycin resistance) from Wt background were named k6 mutants (for knock-out CP26), whereas the lines obtained from k9.2 background were named k69 (for knock-out CP26 CP29). SDS-PAGE and western blot analysis confirmed that both k6 and k69 mutants were completely depleted of CP26 ( Figure 1). Surprisingly, both k6 and k69 lines also lacked the FIGURE 1 CP26-and CP29-targeted gene disruption in Chlamydomonas reinhardtii by CRISPR/Cas9 genome editing. (a) DNA sequence alignment of wild type (Wt) and mutants on cp26 (k6) and cp29 (k9) obtained by CRISPR/Cas9 genome editing. The 20-bp target sequence of sgRNA is reported before the red-coloured PAM sequence. In the case of k9 mutant, insertions (k9-1) or deletions were induced (k9-2) by nonhomologous repair, whereas in the case of k6 mutant, a Cas9-mediated hygromycin resistance cassette insertion was obtained in the target site. Double-mutant k69 was obtained by hygromycin resistance cassette insertion in a k9 background. (b) PCR amplification of cp26 and cp29 CDS sequences from Wt, k6, k9, and k69 cDNA. In the case of k6 and k69 mutants, the cp26 CDS cannot be amplified due to hygromycin cassette insertion. (c) SDS-PAGE analysis of Wt and mutant strains performed with the Tris-Tricine buffer system. Fifteen microgram of Chl was loaded in each lane. Molecular weight (MW) markers are indicated on the left. CP26 and CP29 bands are marked. (d) Immunoblot with specific antibodies directed against CP26 and CP29 on the same lanes of (c). Immunoblot against CP43 was added as control of the loading. (e) qRT-PCR on cp26 and cp29 genes in Wt and mutant strains. rack1 gene was used as control for qRT-PCR (see details in Figure S3) CP29 protein even if the gene was not targeted in k6 mutant strains (Figures 1 and S2). In order to investigate if the absence of CP29 in k6 mutant was related to a transcriptional or post-transcriptional event, transcription profiles of lhcb4 and lhcb5 genes were investigated in Wt and mutant lines. As reported in Figures 1e and S3, the transcription of cp29 and cp26 genes was not impaired in k6, k9, and k69 mutants compared with Wt, suggesting that the absence of CP26 and CP29 accumulation was related to translation or posttranslation events. Because k6 and k69 showed the same pattern of monomeric subunits and phenotype, the following characterization was focused on k9 and k69 strains.

| Pigments composition and stoichiometry of pigment-protein complexes
When grown in HS medium (Harris & Harris, 2008) at 100 μmol m −2 s −1 , k9 and k69 showed a slight but not significant reduction of chlorophyll (Chl) content compared with Wt; the Chl/carotenoid (Car) ratio was also similar in all the three genotypes (Table 1). Car composition determined by high-performance liquid chromatography analysis was similar in the three genotypes with an increase of neoxanthin and violaxanthin and a decrease of β-carotene in the k69 mutant (Table 1). In these growth conditions, no zeaxanthin accumulation was detected in any genotype.
The Chl a/b ratio was remarkably affected by the absence of monomeric subunits, being reduced in k9 and even further decreased in k69 in comparison with Wt ( Figure 2a and Table 1), indicating a possible increment of PSII antenna proteins. In order to confirm this, we determined the relative amount of protein components of the photosynthetic apparatus by immunoblotting on LHCII, LHCI, CP43 (subunit of PSII core), and PSAA (subunit of PSI core; Figures 2b and S4; Ballottari, Dall'Osto, Morosinotto, & Bassi, 2007). Deletion of monomeric subunits had a strong effect on both PSII/PSI and LHCII/PSII ratios: In the k9 mutant, there was a relative PSI increment of~23% with respect to Wt, whereas in the double-mutant k69, the amount of PSI was more than doubled. The effect of CP26 and CP29 mutations on LHCII/PSII was even stronger: In k9, the LHCII amount was doubled, whereas the k69 mutants showed an approximately fivefold increase. The absence of CP26 and CP29 did not affect the PSI antenna because the LHCI/PSI ratio was the same as the Wt (Figures 2b and S4).
The organization of pigment-binding complexes was further analysed by sucrose gradient upon solubilization of thylakoid membranes with dodecyl-α-D-maltoside (α-DM; Figure 2c). Different green bands were resolved; from top to bottom, the bands corresponded to LHC protein in monomeric state (Fraction 1), trimeric LHCII (Fraction 2), PSII core complex (Fractions 3 and 4), PSI-LHCI (Fractions 5 and 6), and PSII-LHCII supercomplex (Fraction 7; . In Wt, the upper band contained CP29, CP26, and LHCII, monomerized by solubilization, whereas in k69, the band contained only the latter. The densitometry analysis of the band of the gradient confirmed the data from the immunotitration ( Figure S5) with an increment of the bands corresponding to antenna proteins (Fractions 1 + 2) and a decrease of those corresponding to PSII core (Fractions 5 + 6). The bands corresponding to PSII-LHCII supercomplex were absent in k9 and k69, indicating that in mutants, these complexes were not assembled or they were easily dissociated by solubilization.
The functional antenna size of PSII in Wt and mutant strains was then determined by measuring fast chlorophyll florescence induction in the presence of DCMU (Malkin et al., 1981), being inversely proportional to the rise time. Mutants showed an antenna size increment of 1.5-fold in k9 and 2.5-fold in the k69 double mutant with respect to Wt (Table 1 and Figure S6), consistently with the increased LHCII content and decreased Chl a/b ratio measured in these strains.

| Photosynthetic efficiency and state transition
Fluorescence induction analysis in dark-adapted cells (Butler, 1973) revealed a significant decrease of PSII (Fv/Fm) in mutants compared with Wt (Table 1), suggesting that energy transfer from LHCII to PSII reaction centre was less efficient in the absence of monomeric subunits. In particular, on a Chl basis, a clear increase of F 0 was evident in the mutant strains ( Figure S8). Accordingly, 77-K fluorescence emission spectra (Figure 2d) revealed a blue shift in the first peak from 686 nm in Wt to 681 in k69 mutant, with a 680-nm shoulder evident also in k9 mutant: This 681-nm peak corresponds to light harvesting antennae unable to transmit their energy to PSII (Garnier, Maroc, & Guyon, 1986). In order to evaluate how this defective connection affects photosynthetic complexes, microalgae were illuminated at different light intensities, and PSII operating efficiency (ΦPSII), ETR, and photochemical quenching (1-qL) were recorded (Figure 3a with Wt at all light intensities tested, and the slope of light saturation curve initial linear increase was reduced (Table S1). These data confirmed that the quantum yield of photosynthetic apparatus is impaired in these genotypes, especially in double-mutant k69. Dark respiration was instead not affected by the absence of CP26 and CP29.
The roles of CP26 and CP29 in state transitions were then investigated by inducing State 1 and State 2 as previously described (Drop, Yadav K N, et al., 2014) using the stt7 mutant, unable to perform state transitions, as control (Depege, Bellafiore, & Rochaix, 2003  Immunotitration of thylakoid proteins using specific antibodies against PSAA, LHCII, and LHCA. Data were corrected for CP43 amount and normalized to Wt ratio. LHCA was corrected for PSAA amount. LHCII is shown as total amount (LHCII) or divided for three bands that are separated on the Tris-Tricine buffer system (see Figure S2). Data are expressed as mean ± standard deviation (n = 4).

*
Values that are significantly different (Student's t test, P < 0.05) from the WT. ** Data that are significantly different between k6 and k69. (c) Sucrose density gradient fractionation of WT, k9, and k69 solubilized with 0.8% dodecyl-α-D-maltoside (α-DM). Composition of the green bands is indicated on the left according to Semchonok et al. (2017); Wt in black, k9 in grey, and k69 in light grey. (d) Low-temperature fluorescence of Wt, k9, and k69 cells excited at 475 nm. Emission spectra are normalized at the maximum emission on PSII peak. All data were collected from algae grown in high salts at 100-μmol photons m −2 s −1 was 42 ± 12%: k69 double mutant was unable to perform state transitions as in the case of the negative control, stt7 mutant. Single-mutant k9 retained a small fraction of state transitions, but the mechanism was hampered (14 ± 5%) compared with Wt. The defects in state transitions observed in the k9 and k69 mutant strains ( Figure 4) were not related to a different accumulation of the phosphorylating enzyme STT7, which was similarly accumulated in mutant strains compared with Wt ( Figure S7). The treatment inducing State 2 in Wt caused in k9 mutant a shift of PSII emission from 686 to 681 nm (Figure 4), indicating that in k9 mutant, when state transitions are induced, LHCII detaches from PSII complex but only a part of it is able to shift to PSI. In State 2, the shape of PSII emission was identical to k69 double mutant with a peak at 681 nm, suggesting that most of the LHCII trimers were already disconnected from PSII in both States 1 and 2.

| Non-photochemical quenching
The specific role of different PSII antenna in NPQ mechanism has long been debated (Ahn et al., 2008;Dall'Osto et al., 2017;Ruban et al., 2007;Townsend et al., 2018). We thus investigated the ability of Wt, k9, and k69 to undergo quenching of Chl fluorescence upon exposure to excess light. C. reinhardtii needs to be adapted to high light conditions to fully activate NPQ mechanism; for this reason, Wt and mutant strains were grown for 2 weeks at 500-μmol photons m −2 s −1 in autotrophy. After adaptation to high light, reduced Chl/cell and Chl/Car ratios were evident with no significant difference between Wt and mutant strains (  were present in mutants, and their abundance, with respect to PSII core, was even higher than in Wt (Figures 5 and S10). The generation of an electrochemical proton gradient across thylakoid membranes was assessed using ECS-induced absorbance changes at 520 nm   where CP29 knock-out plants lacked also CP24: CP24 and CP29 cooperate in binding M-LHCII (moderately bound) to the core, forming the C 2 S 2 M 2 complex (de Bianchi et al., 2011). The fact that CP26 was still present in k9 mutant could indicate that CP26 could, by itself, maintain a stable connection between S-LHCII and the core even in the absence of CP29. Alternatively, because CP26 is in an external position in the PSII supercomplex (Drop, Webber-Birungi, et al., 2014), it could be stabilized by the interaction with antennas of interacting neighbour PSII complexes.  However, this strategy was only partially successful in the absence of CP29 and CP26: k69 mutant showed an approximately fivefold increase in LHCII/PSII stoichiometry, but the functional antenna size was only increased by approximately twofold with respect to Wt.
Moreover, it is worth to note that the functional antenna size of k9 and k69 might be overestimated as a consequence of the increased F 0 observed in these strains (Malkin et al., 1981). These findings thus demonstrate that in the absence of CP26 and CP29, a significant fraction of trimers was not energetically coupled with PSII.
Monomeric antenna proteins have thus a key role in C. reinhardtii in controlling excitation energy transfer from LHCII to the reaction centre, as previously reported in the case of higher plants (Townsend et al., 2018;. Values of PSII operating efficiency, relative ETR, ECS, and oxygen evolution were lower at all light intensities tested: The absence of CP29 or both CP26 and CP29 caused a partial reduction of photosynthetic efficiency, which could not be restored by the increased content of LHCII ( Figure 3).
Previous work showed that CP29 is involved also in state transitions (Tokutsu et al., 2009) and more specifically in the docking of LHCII to PSI (Drop, Yadav K N, et al., 2014;Kargul et al., 2005). The results herein reported demonstrate that in C. reinhardtii, both CP26 and CP29 are involved in state transitions (Tokutsu et al., 2009), with a no-state transitions phenotype for the k69 mutant. Indeed, it has been shown that both subunits in State 2 condition are phosphorylated and likely bind to PSI with LHCII trimers upon State 2 induction (Takahashi, Iwai, Takahashi, & Minagawa, 2006).
The results herein presented confirm the important role of monomeric antenna subunits in light harvesting and in preserving PSII photosynthetic efficiency. Similar evidence has already been reported in A. thaliana (Dall'Osto et al., 2017;Townsend et al., 2018), even if the effects were more extreme in C. reinhardtii with respect to higher plant. The stronger difference observed when comparing C. reinhardtii and A. thaliana mutants depleted of monomeric CP26 and CP29 subunits is the NPQ phenotype. In an A. thaliana mutant without monomeric antennae, NPQ activation was slower in dark-to-light transition, but after a few minutes of illumination, the quenching was identical to Wt (Dall'Osto et al., 2017;Townsend et al., 2018). In the case of C. reinhardtii, in the absence of CP29, an almost halved NPQ phenotype was detected, whereas the double-mutant k69 presented a no-NPQ phenotype even after 30 min of illumination at high light intensity ( Figure 5). In C. reinhardtii, NPQ is thus totally dependent on the presence of PSII monomeric antenna subunits. The NPQ phenotype observed in k9 and k69 strains is not related to a different accumulation of LHCSR subunits, the NPQ triggers, because these subunits were rather increased in the absence of CP26 and CP29.
Moreover, the absence of NPQ could not be due to a different LHCSR/antenna stoichiometry because in high light adapted cells, the difference in Chl a/b ratio between Wt and k69 was only~10%.
k69 mutants exhibited also an impairment in state transitions induction and a slight reduction of the amount of zeaxanthin accumulated after 2 weeks in high light (Table S2). However, these differences cannot be responsible for the no-NPQ phenotype observed in the absence of CP26 and CP29 subunits: Both stt7 and npq1 mutants, impaired in state transitions and zeaxanthin accumulation, respectively, were reported with a similar NPQ compared with Wt (Allorent et al., 2013;Girolomoni et al., 2019;Niyogi, Björkman, & Grossman, 1997). k69 mutant has reduced electron transport rates, oxygen evolution, and proton gradient generation, but these alterations could eventually explain a slower kinetic of NPQ activation but not its complete absence after prolonged illumination (Townsend et al., 2018). Besides, the highest actinic light used to induce NPQ was at least three times higher than the intensity needed to saturate photosynthesis, according to the oxygen evolution curves reported in Figure 3.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of the article. with specific antibody. Immunoblot against CP43 was added as control of the loading, immunoblot against CP26 to confirm that k6 strains were knock-out lines for CP26. Total protein extract from wild-type (Wt) and a k9 lines were added on the external lanes as control.     Table   1 normalized to the WT case, which was set to 100. Data are expressed as mean ± SD. Values that are significantly different (Student's t-test, P < 0.05) from the wild-type (WT) are marked with an asterisk (*). Date that are significantly different between k6 and k69 are marked with a circle (°).      Table S1. Photosynthesis and respiration rates Table S2. Pigment content of cell acclimated to 500 μmol photons m-2 s-1.