APX2 Is an Ascorbate Peroxidase–Related Protein that Regulates the Levels of Plastocyanin in Chlamydomonas

Abstract The function of ascorbate peroxidase–related (APX-R) proteins, present in all green photosynthetic eukaryotes, remains unclear. This study focuses on APX-R from Chlamydomonas reinhardtii, namely, ascorbate peroxidase 2 (APX2). We showed that apx2 mutants exhibited a faster oxidation of the photosystem I primary electron donor, P700, upon sudden light increase and a slower re-reduction rate compared to the wild type, pointing to a limitation of plastocyanin. Spectroscopic, proteomic and immunoblot analyses confirmed that the phenotype was a result of lower levels of plastocyanin in the apx2 mutants. The redox state of P700 did not differ between wild type and apx2 mutants when the loss of function in plastocyanin was nutritionally complemented by growing apx2 mutants under copper deficiency. In this case, cytochrome c6 functionally replaces plastocyanin, confirming that lower levels of plastocyanin were the primary defect caused by the absence of APX2. Overall, the results presented here shed light on an unexpected regulation of plastocyanin level under copper-replete conditions, induced by APX2 in Chlamydomonas.


Introduction
Ascorbate peroxidases (APXs) (EC 1.11.1.11)are heme b-containing enzymes belonging to class I peroxidases and catalyze the reduction of H 2 O 2 to H 2 O using ascorbate as an electron donor.Several nucleus-encoded APXs are present in the different compartments of plants and algae (Caverzan et al. 2012).In chloroplasts, for example, APXs protect cells against oxidative damage (Shigeoka et al. 2002, Foyer andNoctor 2011).
Their primary function is carried out on the stromal side of the chloroplasts, where the superoxide anion produced by photoreduction of O 2 at the level of photosystem I (PSI) is converted into H 2 O 2 by superoxide dismutase (Asada 1999).
Beside the classic APXs, a new class of APX-related (APX-R) enzymes was identified (Lazzarotto et al. 2021a).This class is present in all the green photosynthetic eukaryotes (Lazzarotto et al. 2015) and typically encoded by a single locus (Dunand et al. 2011).It is characterized by the absence of the amino acids typical for ascorbate binding, while the amino acids for heme binding and catalysis are conserved (Lazzarotto et al. 2021a).Most of the information regarding APX-Rs comes from studies in Arabidopsis thaliana (hereafter Arabidopsis), where the chloroplast localization of the APX-R enzyme (APX6) has been observed (Lazzarotto et al. 2021b).In vitro studies showed that APX6 does not use ascorbate to reduce H 2 O 2 (Lazzarotto et al. 2021b).In vivo, APX6 expression is increased during seed maturation, germination and leaf senescence (Chen et al. 2014(Chen et al. , 2021) ) and responds to stresses, such as high light combined with high temperature (Giraud et al. 2008).In overexpressing transgenic plant lines, APX6 follows a degradative path from chloroplast plastoglobuli to cytoplasm during photomorphogenesis (Lazzarotto et al. 2021b).APX-R expression was also studied in a few other plants.In Triticum aestivum, expression of APX-R is down-or upregulated during drought, heat stress or their combination (Tyagi et al. 2020).In Brassica rapa, the expression of APX-R is differentially expressed according to stress exposure to drought and heat (Verma et al. 2022).In Oryza sativa, the expression is upregulated during drought or cold stress and in the presence of aluminum and UV light.In addition, mutant lines defective for APX-R display a growth delay (Lazzarotto et al. 2011).
Overall, no detailed information regarding the physiology and the photosynthetic activity of the abovementioned land plants has been provided so far.
Previous work from our groups recently showed that an APX-R of the green alga Chlamydomonas, namely, APX2, is chloroplast-localized and exhibits a twin-arginine translocation (TAT) motif for targeting the lumen of the thylakoids and a MxxM motif, typical for metal binding.In addition, the recombinant protein does not use ascorbate as an electron donor for reduction of H 2 O 2 , confirming its classification to the APX-R family.Additionally, it binds both copper and heme, and it modulates the copper-binding capabilities of plastocyanin in vitro (Caccamo et al. 2023).
With the aim of elucidating the link between APX2 and plastocyanin, we decided to investigate the physiology of apx2 insertional mutants.Our detailed analysis of the photosynthetic machinery revealed that the mutants were not affected at the level of photosystem II (PSII).However, when transferred from low light to a sudden increase of light, apx2 mutants exhibited a different redox state of P700, the PSI primary electron donor.This defect was caused by reduced levels of the photosynthetic electron carrier, plastocyanin.This phenotype was rescued under copper-deficient conditions, in which cytochrome c 6 functionally replaces plastocyanin, confirming that plastocyanin deficiency was the sole responsible factor for the mutant phenotype.Overall, these results provide support for the role of APX2 in regulating plastocyanin levels under copper-replete conditions in Chlamydomonas.

The apx2 mutants display undetectable levels of APX2
The APX2 gene model (Cre06.g285150)as described in Phytozome (https://phytozome-next.jgi.doe.gov/)contains seven exons and six introns, has a length of 3,568 bp and codes for a protein of 337 amino acids (Fig. 1A).Two insertional mutants of the APX2 gene, containing the mutagenesis cassette in exon 2 (apx2-1, LMJ.RY0402.180063) and exon 1 (apx2-2, LMJ.RY0402.095128),have been ordered from the Chlamydomonas Library Project (CLiP) library (Fig. 1A).To confirm the APX2 gene disruption, we first checked the integration site of the cassette by amplifying the cassette junctions by PCR followed by sequencing of the amplified products (Supplementary Table S1).We were able to map both sides for the apx2-1 mutant and only the 3 ′ side for the apx2-2 mutant (Fig. 1A).The reason of failure for the apx2-2 mutant could be the presence of a genomic rearrangement at the 5 ′ insertion site of the cassette (Li et al. 2016(Li et al. , 2019)), which could prevent the sequencing of the flanking region.Polyclonal antibodies were raised against the purified recombinant protein expressed in Escherichia coli.Their sensitivity was assessed by a dilution series of the wild type (wt).For that purpose, 30 μg of wt-soluble extracts was diluted by a factor two (100%, 50, 25, 12.5, 6.25, Empty blocks are the 5 ′ UTR and the 3 ′ UTR, and filled blocks are the exons.(B) Immunoblot using soluble protein extracts with a serial dilution of a factor two for wt (100% is 30 μg of loaded extracts, 50% is 15 μg, 25% is 7.5 μg, 12.5% is 3.25 μg, 6.25% is 1.625 μg and 3.125% is 0.8125 μg) and the two apx2 mutants (100% with 30 μg) labeled with antibodies against APX2.The Coomassie staining of the gel is shown as control of loading.(C) Immunoblot using chloroplast fractions (15 μg per lane) from wt and the apx2-1 mutant labeled with antibodies against APX2 and FNR (Ferredoxin-NADP + reductase) (shown as loading control) together with the Coomassie staining of the gel.
3.125%) and the immunoblot was labeled with the APX2 antibodies.A signal corresponding to the expected size of the mature form (24 kDa, Caccamo et al. 2023) was detected in wtsoluble extracts and absent in the apx2 mutants.The detection limit was between 12.50 and 6.25%, indicating that the level of APX2 in the apx2 mutants was below 12.5% of that of wt (Fig. 1B, Supplementary Fig. S1A).Since our previous work demonstrated that APX2 is chloroplast-localized and exhibits a  TAT motif for targeting the lumen of the thylakoids (Caccamo et al. 2023), chloroplast fractions of the wt and of one of the apx2 mutants (the apx2-1 allele) were isolated and analyzed on immunoblot (Fig. 1C, Supplementary Fig. S1B).The 24-kDa band was not detected in the chloroplasts of the apx2-1 mutant, while it was clearly visible in the chloroplast fraction of wt.

apx2 mutants display normal growth and PSII activity under low light but with a potential PSI activity defect
To evaluate the impact of the loss of APX2 on the physiology of Chlamydomonas, the growth rates of the apx2 mutants were determined in phototrophy (minimal medium, low light intensity) and mixotrophy (acetate as organic carbon source, low light intensity).The growth rates of the mutants were similar to those of wt cells except for a minor difference for the apx2-2 mutant in mixotrophy (Table 1).These results suggest that the loss of APX2 is not responsible for physiological modifications that affect growth under low light conditions.Next, we evaluated the photosynthetic activity.In order to maximize the use of photosynthesis, the cells were grown under phototrophic conditions.Under these cultivation conditions, the maximum efficiency of PSII was not affected, except for a small decrease (8%) for the apx2-2 mutant (Fig. 2A).The initial phases (O-J, Fig. 2B) of the rapid rise in chlorophyll a fluorescence in PSII, associated with photochemistry (Kitajima and Butler 1975), showed similar trends.Additionally, the relative electron transport rate (rETR) of PSII remained unaffected at low light intensities (Fig. 2C).All these results indicated that the PSII activity in low light is barely affected in the apx2 mutants, aligning with the calculated growth rates (Table 1).
Nevertheless, the decline in rETR of PSII under high light conditions (i.e.beyond 100 μmol photons m −2 s −1 , Fig. 2C) was observed in the apx2 mutants.Besides, the faster increase in the J-I phase during the analysis of the rapid rise in chlorophyll a fluorescence suggested a donor site limitation of PSI (Zivcak et al. 2015) (Fig. 2B, inset).Lastly, the ratio between the active PSI and PSII centers (PSI/PSII) was lower in the mutants, indicating a potential deficiency of PSI (Fig. 2D).

Reduced levels of plastocyanin in apx2 mutants lead to rapid oxidation of P700
To further investigate PSI activity, we monitored P700 oxidation during 1 s of saturating light (Fig. 3A).apx2 mutants exhibited intracellular level in wt and apx2 mutants measured using spectroscopic analyses (n = 3).(E) Immunoblot with a serial dilution for wt (100% is 30 μg of loaded extracts, 50% is 15 μg, 25% is 7.5 μg and 10% is 3 μg) labeled with antibodies against plastocyanin (PC) and Cytf.Coomassie bluestained gel with the same loadings is shown as control.(F) Ratio between [Cytf ] and [P700] (changes in absorption values calculated considering the respective molar extinction coefficients of 18 mM −1 cm −1 at 554 nm for Cytf and of 50 mM −1 cm −1 at 705 nm for P700) (n = 4).Statistical differences were calculated with one-way ANOVA followed by Dunnett's test for (D) and (F), **** P < 0.0001.a more oxidized P700 redox state, during the first 100 ms of the saturating light intensity, indicating a donor side limitation of PSI.To determine the underlying cause of the observed effect in the mutant cells, we conducted a single-turnover flash experiment to induce a single charge separation in active photosystems.It is worth mentioning that the electron transfer between plastocyanin and P700 + occurs rapidly, with time constants of approximately 1-3 μs (Drepper et al. 1996).Considering this fast electron transfer, the first detection in the P700 monitoring system (200 μs after the single-turnover flash) could not detect the oxidized state in the wt (Fig. 3B), but it was surprising to see that apx2 mutant cells retained the oxidized state for a longer period with a constant time for P700 + re-reduction of approximately 1 ms (50% of P700 + re-reduced).Such a difference suggested an impairment at the donor side of PSI.
To assess the size of the PSI electron donor pool (i.e.mainly plastocyanin), the P700 redox state was monitored in vivo in response to a train of 10 consecutive single-turnover flashes and in the presence of inhibitors of PSII and Cyt b 6 f.In wt, the oxidation level of P700 was increased with consecutive flashes, indicating a progressive withdrawal of electrons from the plastocyanin pool to PSI.The maximum P700 oxidation level was obtained after seven flashes.In contrast, apx2 mutant cells required only two flashes to reach the maximum P700 oxidation level (Fig. 3C).This result suggests a significant reduction in the size of the plastocyanin pool in the absence of APX2.Spectrophotometric quantification of plastocyanin on soluble protein extracts demonstrated a substantial decrease in plastocyanin content in the apx2 mutants compared to wt cells (Fig. 3D).Lower abundance was further confirmed by immunoblots labeled with plastocyanin antibodies, showing that the quantity of plastocyanin in the apx2 mutants is less than 10% of that in wt cells (Fig. 3E, Supplementary Fig. S1C).In contrast, the molar stoichiometry of Cytf + to P700 + remained consistent between wt and the apx2 mutants (Fig. 3F) as well as the level of Cytf, as indicated by an immunoblot analysis (Fig. 3E).

Proteomic analyses of the apx2-1 mutant validate the reduced abundance of plastocyanin and uncover deficiencies in PSI subunits, as well as copper-related enzymes
Since plastocyanin represents the major sink of copper in Chlamydomonas under phototrophic conditions (Kropat et al. 2015), we wanted to investigate the whole proteome to identify any variation of copper-related proteins.For this reason, tandem-mass-tag (TMT) labeling (Zecha et al. 2019) was used to semi-quantitatively analyze the whole cell proteome of Chlamydomonas in phototrophy for the wt and the apx2-1 mutant only, since the two apx2 mutants showed the same phenotype.Using this technique, we could identify and quantify 2,502 and 2,255 proteins, respectively (Dataset 1, Supplementary material).Variations in abundance were considered significant when there was an increase or a decrease of 30% compared to the wt, with an adjusted P value of <0.05 (log 2 fold change ≥0.37 and ≤−0.5) (Dataset 1, Supplementary material, Fig. 4A).Twentyone proteins were found to be significantly decreased, and 42 were significantly increased.As expected, there was a notable decrease in the quantity of plastocyanin (log 2 fold change −1.4).We successfully identified six subunits of PSI, six subunits of cytochrome b 6 f and nine subunits of PSII (Dataset 1, Supplementary material).No significant differences were observed for these proteins except for some subunits of PSI.Indeed, four PSI core subunits (i.e.PsaB, PsaD, PsaF and PsaL) exhibited a significant decrease in abundance.These findings were validated through immunoblot analysis using antibodies specifically targeting PsaF and PsaD subunits (Fig. 4B, Supplementary Fig. S1D).These results support the finding of a reduced PSI/PSII ratio and the absence of any PSII defects (Fig. 2) in these apx2 mutants.
Regarding other copper-related proteins beyond plastocyanin, we did observe a significant difference in abundance for two of them, ferroxidase-like 1 (FOX1) and a cupredoxin multicopper enzyme (Cre12.g537250).FOX1, reported as a multicopper periplasmic oxidase involved in iron uptake (Terzulli and Kosman 2010), showed a log 2 fold change of −0.7 in the absence of APX2, whereas Cre12.g537250 was found to be more this enzyme in the apx2-1 mutant might suggests a function as a copper storage.However, no information is available about Cre12.g537250 and its localization, and the hypothesis on its function in this context remains speculative.The reduction in FOX1 protein was verified through immunoblot analysis for the two apx2 mutants (Fig. 4C, Supplementary Fig. S1E).On the contrary, there was no observed alteration in protein levels of several other copper proteins: copper transporters CTR1 and CTR3, copper chaperone ATX1 and one of the copper-binding subunits of the mitochondrial cytochrome c oxidase (COX2B) (Fig. 4A, Dataset 1, Supplementary material).Altogether, these findings suggest that the pattern of copper metabolism could be modified in the mutants, although not to a significant extent.

Copper deficiency induces a similar response in wt and apx2 mutants
To investigate further the impact of APX2 gene disruption at the level of the PSI electron donor, we decided to evaluate the functional replacement of plastocyanin by cytochrome c 6 induced by copper deficiency (Li and Merchant 1995).Previously, it has been demonstrated that Chlamydomonas cells utilize cytochrome c 6 (Cre16.g651050)as a PSI electron donor below concentrations of 20 nM of copper in the culture medium (Hill and Merchant 1992).Therefore, cells were grown under phototrophic conditions in the presence of 1 nM CuSO 4 .We observed that after 10 days of copper deficiency, cytochrome c 6 took over the role of plastocyanin, as evidenced by the quantification of its characteristic absorption peak at 552 nm in soluble protein extracts (Fig. 5A) and confirmed by immunodetection using antibodies specific to cytochrome c 6 (Fig. 5B, Supplementary Fig. S1F).Cytochrome c 6 seems to be more abundant in the mutants.Further experiments will be required to understand the mechanisms underlying this increased level.Interestingly, under these conditions, the oxidation level of P700 induced by a series of 10 consecutive singleturnover flashes displayed identical kinetics in both the wt and apx2 mutant cells (Fig. 5C).Additionally, we do not observe any changes in the P700 kinetics between wt and apx2 mutants (Fig. 5D).These observations indicate that the pool of electron donors to PSI becomes comparable between the mutants and the wt cells under copper-limited conditions.This would also suggest that the levels of the PsaF and PsaD subunits are also comparable between the wt and the apx2 mutants under copper-limited conditions, although this hypothesis was not verified.In addition, even if variations in the rETR are observed (between 7 and 14%; Fig. 5E), PSII activity is restored even upon high light intensities, contrary to what is observed in the presence of copper (Fig. 2C).We concluded that the absence of APX2 adversely affects plastocyanin, without compromising the ability to substitute plastocyanin with cytochrome c 6 during copper deficiency.

Transcriptomics analyses reveal post-transcriptional regulation under copper-replete conditions and no modification of the copper response regulator 1 targets under copper-deficient conditions
From the aforementioned results, it can be concluded that the main differences between wt and apx2 mutants are present under copper-replete conditions and not under copper-deficient conditions.To further understand the cellular regulation, a transcriptomics study was undertaken by comparing apx2-1 and wt cells grown in phototrophy under both copper-replete (6 μM CuSO 4 ) and copper-deficient (1 nM CuSO 4 ) conditions.To achieve this, RNA was extracted from three biological replicates for each cultivation condition and RNA-seq data were acquired.After quality check, approximately 20 million reads could be uniquely mapped, indicating the high quality of the libraries (Dataset 2, Supplementary material).The generated raw data are available under the project number PRJNA994826.
In copper deficiency, both the wt [wt-wt (1 nM Cu)] and the apx2-1 mutant [apx2-apx2 (1 nM Cu)] display higher abundance of some transcripts governed by the copper response regulator 1 (CRR1) transcription factor (Cre09.g390023), such as Cyt c 6 and FDX5-encoding ferredoxin 5 (Supplementary Fig. S2A).This prompted us to look at the abundance of the 58 target genes of CRR1 (Kropat et al. 2015).Out of these targets, one gene was excluded from our analysis (Cre07.g346900)due to insufficient mapped reads.The remaining 57 genes showed consistent differential expression in both strains (Fig. 6B).However, for some of them, the log 2 fold change was not significant, in contrast to previous observations by Kropat et al. (2015).This discrepancy is likely attributed to variations in growth conditions used in the two studies.In conclusion, even though there are differences between wt and apx2 mutants in copper deficiency, they do not seem to be linked to the CRR1 targets and will be complicated to interpret due to the absence of a phenotype in this condition.

Discussion
APX-R enzymes form a recently identified class within the APX family, present in the eukaryotic green lineage (Lazzarotto et al. 2015).Previous studies on loss-of-function mutant lines in Arabidopsis have suggested that the APX-R enzyme plays a role during stress periods, such as seed maturation, germination and leaf senescence (Chen et al. 2014(Chen et al. , 2021)), although the precise underlying mechanisms remain unknown.Despite the chloroplast localization of APX-R, no interaction with the photosynthetic machinery has been explored in any photosynthetic organism.
Our previous in vitro work demonstrated that recombinant APX2 binds copper and that copper-loaded APX2 interferes with the copper-binding capabilities of plastocyanin (Caccamo et al. 2023).The detailed photosynthetic analysis performed in this study reveals that the main defect of the apx2 mutants is a significant reduction of their plastocyanin levels.Therefore, the relation between plastocyanin and APX2 is confirmed in vivo in Chlamydomonas.
The diminished plastocyanin levels in the apx2 mutants raise the question of how the electron transport can still take place.In Arabidopsis, there are two genes encoding plastocyanin, PETE1 and PETE2 (Abdel-Ghany 2009).While PETE2 is the predominant isoform, PETE1 is less abundant and substitutes PETE2 in copper deficiency (Abdel-Ghany 2009).Mutants of pete1 and pete2 exhibit reduced plastocyanin content, and null pete2 mutants, with 90% reduction of plastocyanin content, only show a slight effect on growth and photosynthesis (Pesaresi et al. 2009).Double mutants are not viable in Arabidopsis (Weigel et al. 2003).Consequently, a low level of plastocyanin is not limiting under optimal growth conditions, being under day/night cycles, suggesting another role for the protein that remains to be deciphered (Pesaresi et al. 2009).Similarly, based on the spectroscopic and immunoblot analyses, the results presented here show that a plastocyanin level less than 10-15% of the wt does not affect the growth under low light.Considering the effect that the lack of APX2 exerts on the plastocyanin content, a question about the nature of unveiled interactors of plastocyanin in Chlamydomonas arises.
Apart from Arabidopsis, other land plants such as poplar, parsley, tobacco, rice and the moss Physcomitrella patens express two plastocyanin isoforms (as described in Pesaresi et al. 2009).In contrast, green algae and cyanobacteria, such as Chlamydomonas and Synechocystis, respectively, possess only one plastocyanin gene (Merchant and Bogorad 1987, Briggs et al. 1990, Pesaresi et al. 2009).This raises the question of whether APX-R of Arabidopsis (APX6) would have a similar impact on plastocyanin levels.
The TAT signal motif for lumen translocation consists of a twin arginine (RR) followed by a hydrophobic stretch and an AxA cleavage site (Cline 2015).When comparing Chlamydomonas and Arabidopsis APX-R sequences, we could notice the absence of the AxA cleavage site in APX6 (Supplementary Fig. S3).This suggests that APX6, despite being chloroplastlocalized (Lazzarotto et al. 2021b), might be not directed to the lumen.In addition, APX6 is transcribed under copperreplete condition, while in copper deficiency, its transcription is repressed by miR398, which is regulated by SQUAMOSApromoter-binding protein-like 7 (SPL7), an ortholog of the CRR1 of Chlamydomonas (Chen et al. 2021).In Chlamydomonas, the transcriptomics analysis presented here shows that APX2 is transcribed under both copper-replete and copperdeficient conditions in wt (Dataset 2, Supplementary material, Supplementary Fig. S2B).Thus, the function of the APX-R might differ between the green algae and the land plants.Nevertheless, both APX-R contain the MxxM motif, typical for metal binding (Caccamo et al. 2023) (Supplementary Fig. S3), suggesting that their dependence on copper could be conserved.
In conclusion, this study demonstrates that APX2 regulates the plastocyanin level under a standard copper concentration in Chlamydomonas.This discovery paves the way for future experiments that could unveil the yet undiscovered mechanisms underlying the relation between APX2 and plastocyanin and, more broadly, the biogenesis of this photosynthetic component.

Growth measurements
Algal growth was determined two times per day by measuring the optical density (OD) at 750 nm until the stationary phase was reached, using a 265 UV/Vis spectrophotometer (LAMBDA TM 265, PerkinElmer, Belgium).The growth rate  (h − 1) of the cell cultures was calculated as in (Jia et al. 2015),  (h − 1) = [ln(OD2) − ln(OD1)]/(t 2 − t 1 ), in which  is the growth rate (h − 1), OD is the optical density of cells at the end (OD2) and beginning (OD1) of the exponential phase, t 2 is the time in h at the end of the exponential phase and t 1 is the time in h at the beginning of the exponential phase.

Total DNA extraction and PCR amplification
DNA extraction was performed by following the protocol described in Newman et al. (1990).The DNA was used for PCR amplifications to confirm the presence of the paromomycin cassette in apx2 mutants.PCR amplifications were performed as suggested on the CLiP library website, using the couple of primers apx2-1_F/OMJ944 and apx2-1_R/OMJ913 to map the 3 ′ end and the 5 ′ end of the cassette, respectively, of the apx2-1 mutant and primers apx2-2_F/OMJ944 to map the 3 ′ side of the cassette of the apx2-2 mutant.Primer sequences are presented in Supplementary Table S2.Sequencing was made at Eurofins genomics.

Total RNA extraction
Cultures of wt and apx2-1 mutant were grown in triplicates in TMP and TMP + 1 nM CuSO 4 under continuous low light at 30 μmol m −2 s −1 .Cultures of TMP + 1 nM CuSO 4 were started by washing the pellets of the precultures grown in TMP-CuSO 4 .The pellets were suspended in TMP-CuSO 4 and the cultures launched in TMP + 1 nM CuSO 4 .After 4 d, the cultures were refreshed with new TMP + 1 nM CuSO 4 medium until plastocyanin was completely substituted by cytochrome c 6 .The cultures were then collected at the mid-exponential phase, the pellets were resuspended in 0.5 ml of 2% sodium dodecyl sulfate (SDS), 400 mM NaCl, 40 mM EDTA, 100 mM Tris/HCl, pH 8.0 and dimethyl pyrocarbonate to lyse the cells and total RNA was extracted using the NucleoSpin® RNA Plant and Fungi kit as described in the Macherey-Nagel protocol (Ref. 740120.50).The concentration and the purity (260/280 ratio 2.0-2.1) of RNA were determined using NanoDrop (Agilent BioTek Synergy Mx Monochromator-Based Multi-Mode Reader with Time-resolved Fluorescence) and using gel electrophoresis.

Transcriptomic analyses
Sample collection and preparation.RNA quantification and quality assessment (Novogene company).RNA integrity was assessed using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, California, USA).
Library preparation for transcriptome sequencing (Novogene experimental department).Total RNA was used as input material for the library preparations, which was carried out by using Novogene NGS RNA Library Prep Set (PT042) according to the manufacturer's protocol.In order to select cDNA fragments of preferentially 370-420 bp in length, the library fragments were purified using an AMPure XP system (Beckman Coulter, Beverly, Massachusetts, USA).Then, PCR was performed with Phusion High-Fidelity DNA polymerase, Universal PCR primers and Index (X) Primer.At last, PCR products were purified (AMPure XP system) and library quality was assessed on the Agilent Bioanalyzer 2100 system.

Clustering and sequencing (Novogene experimental department).
The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumina, San Diego, California, USA) according to the manufacturer's instructions.After cluster generation, the library preparations were sequenced on an Illumina Novaseq platform and 150 bp paired-end reads were generated.
Transcriptomic analyses.Reads were quality trimmed using Trimmomatic software (Bolger et al. 2014) removing adapters and bases with the average score lower than 30 over 10 bp sliding window.Afterward, they were aligned and counted using Trinity software (Grabherr et al. 2011).Chlamydomonas reinhardtii v5.5 transcriptome data were downloaded from Phytozome (Goodstein et al. 2012) and indexed with bowtie2 (Langmead and Salzberg 2012).Transcripts, including isoforms and abundances, were obtained using RNA-Seq by expectation maximization (Li and Dewey 2011).The sequencing depth is presented in Dataset 2, Supplementary material (section Mapped_reads).The normalization and differential expression analysis were performed using DESeq2 R package (Love et al. 2014).The reported differentially expressed (DE) transcripts were obtained by selecting the results with log 2 FC > 1 or log 2 FC < −1 and false discovery rate (FDR)-adjusted P values ≤0.05.

Chlorophyll content
Pigments were extracted from whole cells using cold pure methanol, following a protocol adapted from Emonds-Alt et al.

Isolation of soluble and chloroplast-enriched fractions
Soluble fractions were prepared as in Li and Merchant (1992).Cultures of wt and apx2 mutants were harvested in the mid-exponential phase and washed with sodium phosphate buffer 10 mM pH 7.0, 2000 g for 3 min at room temperature.The pellet was suspended in the same buffer at 2 × 10 8 cells ml −1 and frozen at −80 ∘ C. Five cycles of freeze-thaw for 1 h each were performed to extract soluble proteins.After the last cycle, the cells were centrifuged for 15 min at 4 ∘ C at 12,000 g, the supernatant and the pellet were collected and the protein content was determined using a Bradford assay, as in Ernst and Zor (2010).Chloroplast-enriched fractions were isolated according to Mason et al. (2006).
A blue light-emitting diode probe of 10 μs pulses was used to measure the chlorophyll a fluorescence under dark conditions (Fo) and after a saturating pulse of red light (Fm, 3000 μmol photons m −2 s −1 ) to reach the maximum PSII quantum yield (Fv/Fm = Fm − Fo/Fm).Samples were then acclimated to different photosynthetic photon flux densities (PPFD) for 3 min, and fluorescence values were measured (Fs).A saturating pulse was then given to fully reduce PSII acceptors and reach the maximal fluorescence yield (Fm').These parameters were used to calculate the PSII quantum yield (ΦPSII = (Fm' − Fs)/Fm') and the relative electron transfer rate (rETR = ΦPSII × PPFD) (Maxwell and Johnson 2000).

Fast fluorescence increase
The fast chlorophyll a fluorescence rises in response to a saturating pulse of light of 3,000 μmol photons m −2 s −1 and was obtained using a plant efficiency analyzer fluorometer (Handy PEA, Hansatech, Kings Lynn, UK) after a short dark incubation in 1-cm microcuvettes.The chlorophyll a fluorescence increases were normalized to variable fluorescence (i.e. between the values of Fo and Fm) (Strasser and Govindjee 1992).

Activity of PSI
PSI activity was monitored using the Joliot-Type-Spectrophotometer (JTS-10, BioLogic France) by the changes in absorption in the near infra-red region of the P700, the primary electron donor at the PSI reaction center.The signal at 705 nm was corrected by the absorption at 725 nm.After a short period of dark incubation, a saturating red actinic light of 3000 μmol photons m −2 s −1 was given for 1 s.The measurements were normalized to the maximum change of absorption due to P700 oxidized in the presence of 10 μM of 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) and 10 μM of 2,5dibromo-3-methyl-6-isopropylbenzoquinone (DBMIB).The signal/level of oxidized cytochrome f (Cytf + ) from the cytochrome b 6 f complex was deconvoluted from the signal at 554 nm and a baseline between 546 nm and 573 nm in the presence of 10 μM of DCMU and 10 μM DBMIB.The relative abundance of Cytf + and P700 + (Cytf + /P700 + ) was calculated from molar extinction coefficients of 18 mM −1 cm −1 at 554 nm for Cytf (Pierre et al. 1995) and of 50 mM −1 cm −1 at 705 nm for P700 (Heinnickel et al. 2013).Electron transfer to P700 + was monitored after a single-turnover flash, provided by a Nd-YAG laser (Minilite II, Continuum) generating one charge separation per photosystem.A train of 10 consecutive single-turnover flashes, triggered every 100 ms, was used to oxidize P700 in the presence of DCMU and DBMIB (Drepper et al. 1996).

Electrochromic shift assay
The electrochromic shift (ECS), used to monitor the electric field of the thylakoid membranes in response to photosynthetic activity, was measured using the JTS-10.Absorption changes were monitored at 520 nm and corrected by measurements at 546 nm.A baseline of absorption in darkness was made, and afterward, a single-turnover flash was applied.The measurements were repeated with the addition of 10 μM DCMU and 1 mM hydroxylamine, preventing charge separations at PSII.The change in absorption corresponding to PSII and PSI or only to PSI charge separations was used to calculate the relative abundance of active PSI and PSII (Bailleul et al. 2010).

Plastocyanin and cytochrome c 6 quantification
Soluble proteins were extracted as described in Li and Merchant (1992).Cultures of wt and apx2 mutants were harvested in the mid-exponential phase and washed with sodium phosphate buffer 10 mM pH 7.0, 2000 g for 3 min at room temperature.Soluble extracts were prepared as described earlier.Each sample was split into two, one oxidized with 1 mM potassium ferricyanide and the second one reduced with 1 mM of ascorbic acid.Absorbance was recorded between 400 and 800 nm.The plastocyanin concentration was calculated using a molar extinction coefficient for the oxidized form of 4.9 mM −1 cm −1 at 597 nm (Sommer et al. 2002).The actual absorbances of plastocyanin at 597 nm were calculated by subtracting the absorption value of the reduced sample to the oxidized one (Li and Merchant 1995) and correcting the signal by removing that one at 700 nm.The cytochrome c 6 concentration was calculated using a molar extinction coefficient for the reduced form of 20 mM −1 cm −1 at 552 nm (Sommer et al. 2002).The actual absorbances of cytochrome c 6 at 552 nm were calculated by subtracting the absorption value of the oxidized samples from the reduced one (Howe and Merchant 1992) by correcting the signal to that one at 530 nm.

Proteomic analyses
Cells were grown in phototrophy under low light conditions until the midexponential phase.A total of 10 8 cells were harvested by centrifugation for 3 min at 1000 g, at room temperature.The supernatant was discarded, and the pellet was suspended in 1 ml of RadioImmunoPrecipitation Assay bufferlike buffer: triethylammonium bicarbonate (TEAB) 100 mM, NaCl 150 mM, octylphenoxypolyethoxyethanol 1%, deoxycholic acid sodium salt 0.5% SDS 0.1%, n-dodecyl-β-d-maltoside 0.2% and phenylmethylsulfonyl fluoride 0.5 mM EDTA, vortexed for 10 min at room temperature and centrifuged for 10 min at 4 ∘ C at 21,000 g.The protein concentration of the supernatant was determined using a BCA assay (Pierce™ BCA Protein Assay Kit).Reduction of 200 μl of protein solution at 1.5 μg μl −1 was performed with 10 mM of 1,4-dithiothreitol at

Fig. 1
Fig. 1 Mutants of APX2 exhibit levels of the APX2 protein that are beyond detection.(A) Scheme of the APX2 gene (Cre06.g285150)and position of the paromomycin cassette in the apx2-1 and apx2-2 mutants.The numbers indicate the position of the cassette from the ATG start codon.The arrows indicate the orientation of the cassette.Empty blocks are the 5 ′ UTR and the 3 ′ UTR, and filled blocks are the exons.(B) Immunoblot using soluble protein extracts with a serial dilution of a factor two for wt (100% is 30 μg of loaded extracts, 50% is 15 μg, 25% is 7.5 μg, 12.5% is 3.25 μg, 6.25% is 1.625 μg and 3.125% is 0.8125 μg) and the two apx2 mutants (100% with 30 μg) labeled with antibodies against APX2.The Coomassie staining of the gel is shown as control of loading.(C) Immunoblot using chloroplast fractions (15 μg per lane) from wt and the apx2-1 mutant labeled with antibodies against APX2 and FNR (Ferredoxin-NADP + reductase) (shown as loading control) together with the Coomassie staining of the gel.

Fig. 2
Fig. 2 apx2 mutants display no defect in PSII activity under low light but a possible defect around PSI. (A) Maximum quantum yield of PSII (Fv/Fm) measured after 1 min of dark incubation (n = 6).(B) Fast chlorophyll a fluorescence increase was measured during 1 s of saturating light after 1 min of dark cell incubation [r.u.: relative units of normalized fluorescence between O (20 μs) and P levels].Inset in (B): Zoom on the increase of the J to the intermediate I phase for the apx2 mutants (n = 6).(C) Light-response curve of photosynthetic rETR at increasing light intensities with measurements of 30 s per each light step (n = 3).(D) Ratio between active PSI and PSII centers calculated from ECS measurements in the presence of PSII inhibitors (n = 9).Values are shown as averages with standard deviations.Statistical differences were calculated using one-way ANOVA for (A) and (D), or two-way ANOVA for (C) followed by Dunnett's test, *** P < 0.001, * P < 0.05.

Fig. 3
Fig. 3 The electron transfer to P700 was impaired in the apx2 mutant cells, accompanied by a reduced level of plastocyanin.(A) P700 redox kinetics measured during 1 s of saturating light after 1 min of dark incubation.Values are normalized to maximum oxidizable P700 in the presence of DCMU and DBMIB, inhibitors of PSII and Cyt b 6 f, respectively (n = 3).(B) Monitoring of the P700 redox state upon a single-turnover flash on wt and apx2 mutants (values normalized to the maximum oxidizable value of P700) (n = 3).(C) Monitoring of the P700 redox state during a train of 10 singleturnover flashes (every 100 ms) in wt and apx2 mutants previously incubated in PSII and Cyt b 6 f inhibitors (values normalized to the maximum oxidizable value of P700 after adding 10 μM DCMU and 10 μM DBMIB inhibitors of PSII and Cyt b 6 f, respectively) (n = 3).(D) Plastocyanin (PC)intracellular level in wt and apx2 mutants measured using spectroscopic analyses (n = 3).(E) Immunoblot with a serial dilution for wt (100% is 30 μg of loaded extracts, 50% is 15 μg, 25% is 7.5 μg and 10% is 3 μg) labeled with antibodies against plastocyanin (PC) and Cytf.Coomassie bluestained gel with the same loadings is shown as control.(F) Ratio between [Cytf ] and [P700] (changes in absorption values calculated considering the respective molar extinction coefficients of 18 mM −1 cm −1 at 554 nm for Cytf and of 50 mM −1 cm −1 at 705 nm for P700) (n = 4).Statistical differences were calculated with one-way ANOVA followed by Dunnett's test for (D) and (F), **** P < 0.0001.

Fig. 4
Fig.4Comparative proteomic analysis reveals variations in protein levels in the apx2-1 mutant, which are confirmed by immunoblot in both mutants.(A) Volcano plots of the proteomic data.Red dots indicate significantly different proteins in the apx2-1 mutant compared to wt (log2 fold change > 0.37 and < −0.5).Black dots indicate the nonsignificant proteins.In the plot plastocyanin, PSI subunits (PSAB, PSAD, PSAF and PSAL), copper enzymes (FOX1, cupredoxin) and subunits of the cytochrome c oxidase (COX2A and COX2B) are indicated.The PSII and cytochrome b 6 f subunits are shown in the non-significant portion of the plot.(B) Immunoblot with a serial dilution for wt (100% is 30 μg of loaded extracts, 50% is 15 μg, 25% is 7.5 μg and 10% is 3 μg) labeled with antibodies against PSAF and PSAD.Coomassie bluestained gel with the same loadings is shown as control.(C) Immunoblot with a serial dilution for wt (100% is 30 μg of loaded extracts, 50% is 15 μg, 25% is 7.5 μg and 10% is 3 μg) labeled with antibodies against FOX1.Coomassie blue-stained gel with the same loadings is shown as control.

Fig. 6
Fig. 6 Transcriptomics analyses reveal post-transcriptional differences between wt and the apx2-1 mutant under copper-replete conditions and similar response to CRR1 master regulator under copper-deficient conditions.(A) Expression of PCY1, PsaF, PsaD, PsaL, FOX1 and cupredoxin is not significantly different between wt and apx2-1 mutant.The color code at the right shows the difference in log 2 fold change.The adjusted P value is shown in the boxes corresponding to each comparison.(B) Comparison between wt and apx2-1 mutant in response to copper deficiency.The color code at the right shows the difference

Fig. 6 (
Fig. 6 (Continued)  in log 2 fold change.The adjusted P value is shown in the boxes corresponding to each comparison with significantly differential expressed genes indicated with asterisks.

Table 1
The growth rate of both wt and apx2 mutants is comparable Chlamydomonas wt and apx2-1 cells were grown in triplicate under the two cultivation conditions (12 samples) and collected for RNA extraction as described in the Total RNA extraction section.The samples were sent for sequencing to the Novogene Company (Novogene Company Limited, Cambridge, UK).Raw data are available at the National Center for Biotechnology Information with the BioProject ID number PRJNA994826.