ROS-mediated enhanced transcription of CYP38 promotes the plant tolerance to high light stress by suppressing GTPase activation of PsbO2

As a member of the Immunophilin family, cyclophilin38 (CYP38) is discovered to be localized in the thylakoid lumen, and is reported to be a participant in the function regulation of thylakoid membrane protein. However, the molecule mechanisms remain unclear. We found that, CYP38 plays an important role in the process of regulating and protecting the plant to resist high light (HL) stress. Under HL condition, the gene expression of CYP38 is enhanced, and if CYP38 gene is deficient, photochemistry efficiency, and chlorophyll content falls distinctly, and excessive reactive oxygen species synthesis occurs in the chloroplast. Western blot results showed that the D1 degradation rate of cyp38 mutant plants is faster than that of wide type plants. Interestingly, both gene expression and activity of PsbO2 were drastically enhanced in cyp38 mutant plants and less changed when the deleted gene of CYP38 was restored under HL treatment. This indicates that CYP38 may impose a negative regulation effect on PsbO2, which exerts a positive regulation effect in facilitating the dephosphorylation and subsequent degradation of D1. It is also found that, under HL condition, the cytoplasmic calcium ([Ca2+]cyt) concentration and the gene expression level of calmodulin 3 (CaM3) arose markedly, which occurs upstream of CYP38 gene expression. In conclusion, our results indicate that CYP38 plays an important role in plant strengthening HL resistibility, which provides a new insight in the research of mechanisms of CYP38 protein in plants.


Introduction
The light-dependent reactions and subsequent oxygen evolution in photosynthesis are performed by four multi-subunit protein complexes within the thylakoid membranes of chloroplast. They are Photosystem I (PSI) and Photosystem II (PSII), cytochrome b 6 f complex, and CFo-CF 1 complex (Shi and Schröder, 2004;Barber, 2006;Nelson and Yocum, 2006). Among these multisubunit protein complexes, PSII serves as a light-driven water-plastoquinone oxidoreductase in the thylakoid membranes of oxygenic photosynthetic organisms, included in the process of oxygenevolution and a membrane-embedded light-harvesting chlorophyll a/b complex (LHCII) that directs the light energy to the PSII core. As the essence of photosynthesis, light also can lead to inactivation of PSII and causes damage to the D1 protein in the reaction center, especially triggered by high light (HL) conditions. Due to such light-induced damage, D1 turns over at a higher rate than all other thylakoid polypeptides (Mattoo et al., 1981). Photodamage to PSII and attendant turnover of D1 protein for PSII repairing occur at all degrees of light intensity. It requires a number of processes to repair photodamaged PSII, which include (i) phosphorylation of the damaged D1 protein and migration from the grana to the stroma membrane regions of thylakoid, (ii) dephosphorylation of p-D1 (p-D1 instead of phosphorylated D1), and then its degradation, (iii) reassembly of new D1 protein into functional PSII complexes (Adir et al., 1990;Barbato et al., 1992;Prasil et al., 1992;Aro et al., 1993;Rintamäki et al., 1996). However, these processes need much more precise molecular regulations which have not yet been completely unveiled.
Initially, immunophilins are found to be cellular receptor proteins for immunosuppressive drugs: FK506 and cyclosporin A (Schreiber, 1991). The receptor proteins for FK506 and cyclosporin A, respectively, called FK506-binding proteins (FKBPs) and cyclophilins (CYPs), are indicated as immunophilins collectively. Now it is known that these proteins widely exist in organisms, from bacteria and fungi to animals and plants. Genes encoding 52 putative immunophilins have been identified in Arabidopsis, including 29 CYPs, and 23 FKBPs (He et al., 2004). Thus far, the group is the largest family of immunophilin identified in any organism. One feature of Arabidopsis immunophilin family that arrests attention is that a large fraction is confined to a particular locality of the chloroplast and that up to 11 FKBPs and 5 CYPs were predicted to exist in the thylakoid lumen. The sequences divergence implies that they may have highly specialized functions in thylakoid lumen (Romano et al., 2005;Edvardsson et al., 2007).
The CYP38 protein is a complex immunophilin residing in the Arabidopsis thylakoids lumen. Its ortholog from Spinacia, TLP40, is also discovered to be located in the thylakoids lumen (Fulgosi et al., 1998), which is intended to function in the phosporylation of thylakoid proteins (Vener et al., 1999;Rokka et al., 2000). Recently, it is shown that CYP38 takes up a vital role in the assembly and maintenance of PSII supercomplexes in Arabidopsis (Fu et al., 2007), and its crystal structure is also found (Vasudevan et al., 2012). CYP38 deficiency is found to be able to severely decrease the in vivo phosphorylation of PSII core proteins (Sirpiö et al., 2008). Sirpiö et al. (2008) also suggests that during assembly of PSII, CYP38 protein directs the appropriate folding of D1 (and CP43) into PSII, thereby making the accurate assembly of the water-splitting Mn 4 -Ca cluster able, even with high PSII turnover. It is still a void in many studies, however, when it comes to a mechanistic analysis of plant response to HL stress, especially to how the signaling pathway of CYP38 protects the PSII.
PsbO, referred to as the 'manganese-stabilizing' protein, has been proved to bind with high affinity to GTP (Spetea et al., 2004) and function as a GTPase (Lundin et al., 2007a). From cyanobacteria to higher plants this extrinsic subunit is conserved. There are two different protein isoforms, PsbO1 and PsbO2, have been discovered in Arabidopsis, and they differ only by 10 amino acids in the mature form. It is found in further study that the PsbO2 component demonstrated considerably higher GTPase activity than PsbO1 protein (Lundin et al., 2008). The different functions of PsbO1 or PsbO2 are understood with the help of two T-DNA insertion mutant lines of Arabidopsis deficient in either isoform: it seems that PsbO1 mainly supports PSII activity, whereas the PsbO2 regulates the dephosphorylation and turnover of D1 protein (Lundin et al., 2007b). However, studies on the molecular mechanisms that underlie the roles of PsbO2 in response to abiotic stresses are poor.
It is believed that reactive oxygen species (ROS) and calcium (Ca 2+ ) are quite essential signaling messengers in plant cells (Bowler and Fluhr, 2000;Romeis et al., 2001). calmodulin (CaM), a ubiquitous second messenger, serves as a crucial sensor for Ca 2+ in signal transduction in plants. CaM in Arabidopsis varies in different isoforms and they can interact with their particular targets upon different exogenous stimuli (Bowler and Fluhr, 2000;Romeis et al., 2001;Li et al., 2012). However, analysis of the ROS and Ca 2+ -CaM cascade response to HL, especially the signaling pathway, is still lacking. This paper probes into the possible molecular mechanisms of the protective roles of CYP38 on PSII under the HL stress process. The result of the study showed that CYP38 gene transcription was up regulated in HL condition. It is also discovered that PsbO2 may a target inhibited by CYP38, and the role of CYP38 in HLinduced defense response is important.

High Light Treatment
Arabidopsis plants were fully exposed to HL (2000 µmol photons m −2 s −1 ) supplied by light-emitting diode panels (Photon System Inst.) (Zhao et al., 2014). Protoplasts at a chlorophyll concentration of 0.2 mg ml −1 were illuminated with white light at 2000 µmol photons m −2 s −1 at 20 • C (Herbstová et al., 2012). Control samples of Protoplasts were kept in the dark or normal light, and otherwise identical conditions. Isolated thylakoid membranes (0.1 mg Chl ml −1 ) were illuminated for up to 1 h with 2000 µmol photons m −2 s −1 at 20 • C. The Arabidopsis plants used for qRT-PCR were illuminated with 2000 µmol photons m −2 s −1 at 20 • C for up to 15 h.

DAB Staining
For ROS detection, the leaves from wide type (WT) plants and cyp38 mutant plants were incubated in DAB solution (1 mg ml −1 ) in the dark for 8 h (Liu et al., 2007). The leaves were then cleared using 95% ethanol for 10 min at 80 • C to remove the chlorophyll completely.

Measurement of Chlorophyll Content and Photochemical Efficiency
Chlorophyll from leaves, Protoplasts or thylakoid membranes was extracted by 95% ethanol at 80 • C. The chlorophyll content in each unit mass or unit volume of sample was determined as described by Lichtenthaler (1987), and then counting or concentration adjusting was performed, so that it would be more convenient to release the data or perform the next stage of the experiment. Integrity of PSII affected the ratio of maximum variable fluorescence to maximum yield of fluorescence (Fv/Fm), which was measured by a PAM fluorometer (Walz GmbH, Effeltrich, Germany) (Ralph et al., 2005).
The content was determined spectrophotometrically using the formula Chl (a + b) = 5.24 A 664.2 + 22.24 A 648.6 , the content of Chlorophyll (%) = Chl (a + b) × V × 10 −3 /m, where Chl (a + b) equaled the chlorophyll concentration in in µg ml −1 , A represented the absorption, V was the volume of the 95% ethanol and m symbolized the mass of individual leaves.

Total RNA Extraction and Quantitative Reverse Transcript-PCR (qRT-PCR)
Following the protocol of the manufacturer of TRIzol reagent (Sigma), total RNA was extracted, at times as indicated subject to different treatments, from detached leaves utilizing the said reagent. The RNA concentration and purity were determined, with OD at 260 nm. Total RNAs were transcribed reversely into cDNAs using an M-MLV kit (Takara), and the cDNAs were adopted as polymerase chain reactions templates. For qRT-PCR assay, SYBR green RT-PCR amplifications were conducted using a total volume of 20 µl that contains 10 µl of SYBR Green real-time PCR Master mix (Toyobo), 5 µl of cDNA, 0.5 µl each of forward and reverse primers (10 µM), and 4 µl of distilled water. The Actin (Zhou et al., 2013) was amplified as an endogenous control. PCR amplification was performed with the light cycler 2.0 instrument (Roche Applied Science). The gene-specific primers for RT-PCR analysis were employed as depicted in the Supplementary Table S1. Each measurement of gene expression profiles was carried out at least three times with independent experimental replicates.

Treatment with CsA
Before HL treatment, the Arabidopsis leaves, protoplasts or thylakoid membranes from different samples were pre-sprayed with solutions containing CsA. CsA was used at a final concentration of 10 µM.

Isolation of Thylakoid Membranes and GTP Hydrolysis Assay
Thylakoid membranes from the WT, cyp38, psbo1, and psbo2 were isolated as previously described (Robinson et al., 1981).
To assess the GTP hydrolysis activity of PsbO protein, PSII membranes were isolated from different type plants finally at a same chlorophyll content for each sample as previously described (Berthold et al., 1981;Lundin et al., 2007a). GTP hydrolysis assay using a QuantiChrom TM GTPase Assay Kit (DATG-200, Hongyue, China).

Laser Confocal Scanning Microscopy (LCSM)
Reactive oxygen species production was determined by detecting the fluorescence of DCF, the product of oxidation of H 2 DCFDA, as described previously (Gao et al., 2008). Ca 2+ content of the protoplasts was determined by detecting the fluorescence of Fluo-3, the product of Fluo-3-AM (Zhang et al., 1998). With or without HL treatment, the Arabidopsis protoplasts were incubated with H 2 DCFDA or Fluo-3-AM at a final concentration of 5 µM. The intracellular ROS production and distribution, as well as the chloroplast autofluorescence and Fluo-3 fluorescence, were visualized under the Zeiss LSM 510. The fluorescence intensity at 525 nm was used to determine the relative DCF and Fluo-3, and chloroplast autofluorescence was visualized at 650 nm with a long pass filter, all this with an excitation wavelength of 488 nm.

Flow Cytometry Analysis
Singlet oxygen ( 1 O 2 ) react with the SOSG to produce SOSG endoperoxides (SOSG-EP), as described previously (Shen et al., 2011). To the changes in DCF, Fluo-3, and SOSG-EP, flow cytometry FACScanto II (Becton Dickinson, Mountain View, CA, USA) equipped with BD FACSDiva acquisition soft-ware was used. Protoplasts were incubated with 10 µM of SOSG, 5 µM H 2 DCFDA, or 5 µM Fluo-3-AM for 30 min at room temperature and then analyzed by flow cytometry. The fuorescence of SOSG-EP was excitated with a 504 nm laser. Autofuorescence of the unstained protoplasts under the same physiological conditions were used as controls. Histograms were processed with FCS Express software. Fluorescent effux was measured by counting cells in the H region of the plot (Chen et al., 2006).

HL Causes Inhibition of Photosynthesis in Arabidopsis
It is reported in literatures that deletion of CYP38 causes stunted growth of the plants and the cyp38 mutant plants is highly sensitive to HL (Fu et al., 2007;Sirpiö et al., 2008). When Arabidopsis plants were treated under HL condition, as shown in the Figure 1A, cyp38 mutant plants soon wilt and finally die. However, when the CYP38 gene was restored, the wilting and death of the plants (cyp38-C plants) were eased off (data not shown). Further DAB staining showed that, under HL condition, as the cyp38 mutant was dying, its leaves accumulated more ROS in comparison with WT ( Figure 1B). Stunted growth, as well as high sensitivity to HL of the cyp38 mutant, illustrate that, because of CYP38 deletion, the photosynthetic activity of plants is more likely to be damaged by oxidation induced by HL. For further data analysis, we have simultaneously detected the chlorophyll content of leaves and photochemistry efficiency of PSII. As shown in the Figures 1C,D, under HL treatment, there was dramatic decline in chlorophyll content and photochemistry efficiency of the cyp38 mutant. After HL treatment for 3 days, the chlorophyll contents in WT plants and cyp38 mutant plants, respectively, fell down to 80 and 28% of the original levels, and the photochemistry efficiencies, respectively, fell down to 83 and 30% of the original levels.

CYP38 is Involved in Response to HL Stress Condition in Arabidopsis
The data given above illustrates that, in terms of protecting photochemistry efficiency and HL resistance of plant, CYP38 plays an important role. In order to further confirm with experiment that CYP38 essentially participates in the protection of photosynthetic activity under HL, we detected the change of CYP38 gene expression level in WT plant after HL treatment. qRT-PCR results showed that, after HL treatment, CYP38 gene expression arose, and rose to a level four times higher than that in the control group at 6 h (Figure 2). To verify the authenticity of Result of RT-PCR, we added Rbcl gene (Kress and Erickson, 2007) and UBQ gene (Miki et al., 2005), respectively, as control groups (Supplementary Figure S4), and found that CYP38 gene expression levels are consistent with the Actin group.  On the contrary, the gene expression under low light (LL) is not significantly different from that of control genes (Supplementary Figure S1). In sum, CYP38 may, as an important regulation molecule, participate in the protection of Arabidopsis plant under HL condition.

Lack of CYP38 Gene Increases HL-Induced Chloroplast ROS Production
Under HL condition, the process of damage to photosynthetic activity of cyp38 mutant plants, as well as death of the plants later on, was characterized by outbreak of excessive ROS in leaves ( Figure 1B). These illustrate that CYP38 plays a positive role in maintaining the redox state of the leaves. Cells are the most fundamental unit for plants' function. In order to carry out deeper research on relevant mechanisms, it is necessary to do ROS distribution analysis at cellular or even sub-cellular level. The imaging results showed that, after 30 min of HL treatment, there was generation of ROS in protoplasts derived from WT plants; after 60 min, significance in quantity was observed ( Figure 3A). The inhibition of photosynthetic chain can induce 1 O 2 to synthesize. To clarify the type of ROS, it is found after flow cytometry analysis that, lots of 1 O 2 was synthesized in the cell during the early stage of HL processing (Figures 3B,C). As 1 O 2 is of highly oxidation activity, high cell synthesis is the cause of overdue impairment of cells. Under HL condition, the cell generated ROS, indicating that HL causes change of the redox state of cells. In comparison to this, a higher outbreak of protoplast ROS derived from cyp38 mutant plants occurred as early as 30 min after HL treatment, while DCF fluorescence imaging and chloroplasts were subject to colocalization ( Figure 3A). Furthermore, CsA, as the non-selective combinator and inhibitor of CYPs (Handschumacher et al., 1984;Steinmann et al., 1991), can be used for pre-treatment of protoplasts derived from WT plants (Supplementary Figure S2). It was discovered that, after 30 min of illumination (HL) in the pre-treatment, the ROS which was larger than that nonpretreated group. The outbreak of ROS which is subject to colocalization with chloroplasts in CYP38-deletant under HL may be an important reason for lack of photosynthetic efficiency and poor HL resistibility in cyp38 mutant plant.

D1 Protein is Drastically Decreased in CYP38-Deletion Plants Under HL Stress Condition
Oxygen-evolving complex (OEC), as an important constituent of PSII, can become a major damaged point (Hakala et al., 2005), whether under biotic or abiotic stress, leading to the decrease in activity of oxygen evolution. Studies also proved that, D1, as one of the core proteins of PSII, boasts the fastest turnover rate and is the part which is most likely to be damaged by oxidation (Greenberg et al., 1989;Jegerschoeld et al., 1990). Our research proved that CYP38 may plays a certain role in maintaining the redox state of chloroplast, and deletion of this protein induce excessive generation of chloroplast ROS (Figure 3). In order to offer more in-depth information, the thylakoid membranes of WT and cyp38 mutant were extracted in this study, and treated under HL up to 60 min, in order to detect the oxygen evolution activity and remaining D1 protein content. As shown in the Figure 4A, under HL treatment, the activity of oxygen evolution of cyp38 mutant fell distinctly. Instead, the WT showed a lag phase of 15 min, and reached similar level as in cyp38 mutant after 45 min of HL treatment. Both the D1 contents of WT and cyp38 mutant declined, yet there was distinctive difference between them. In 60 min, the inventories of D1 protein in the extracted membranes of WT and cyp38 mutant, respectively, fell down to 70 and 15% of the original levels (Figures 4B,C). The detection of the level of remaining protein after HL treatment showed that D1 protein degraded rapidly in cyp38 mutant. This may be caused by over-oxidation of chloroplasts caused by CYP38 deletion. The mechanism behind the more significant reduction in D1 proteins in cyp38 mutant as compared with WT was further investigated.

HL-Induced PsbO2 Transcription and Activation Can be Enhanced by CsA
D1 dephosphorylation is a step of the degradation damaged D1, and degradation is impossible if dephosphorylation is abnormal (Rintamäki et al., 1996). PsbO2, as a membrane protein of PSII, is proved to can positively regulate D1 dephosphorylation in recent research report (Lundin et al., 2007b). Our data given above indicates that CYP38 participates in the turnover regulation of D1 (Figure 4). In addition, the results of Sirpiö et al. (2008) indicate that CYP38 imposes regulation effect on D1 phosphorylation. In order to elucidate if there is an interrelation between CYP38 and PsbO2, CsA, the common combinator and inhibitor of CYPs (Handschumacher et al., 1984;Steinmann et al., 1991), was used. First, the change in transcript level of PsbO2 was detected using qRT-PCR. Results showed a significant increase in PsbO2 expression in detached Arabidopsis leaves pretreated with CsA at the 6 h and 9 h under HL treatment, compared to that of the nonpretreated samples (Figure 5A). While the transcription of PsbO1 has no much change under HL or CsA pretreated conditions. The change of PsbO activity under HL was analyzed using two T-DNA insertion mutant lines, psbo1 and psbo2. In CsA-pretreated PSII membranes, the distinctive increase of PsbO activity was found at the 6 h after HL treatment in psbo1 mutant compared to that WT and psbo2 mutant ( Figure 5B). These data indicates that CYPs cascades are involved in HL-induced PsbO activation of and transcription, and PsbO2 may be the main target.

CYP38 is Required for Inhibiting HL-Induced PsbO2 Activity
Data given above illustrates that CYPs may participate in regulation of HL-induced gene expression and activity of PsbO2 (Figure 5). However, further clarification is needed in order to confirm whether CYP38 participates in the process. In order to explain whether CYP38 involves regulation activity of PsbO2, cyp38 mutant and cyp38-C were utilized simultaneously in the experiment. First of all, under HL condition, we detected PsbO2 gene expressions in cyp38 mutant and cyp38-C ( Figure 6A). The results showed that PsbO2 gene expression in cyp38 mutant was higher than that of WT and cyp38-C, and such distinction became even larger after HL treatment. However, contrary to this, there was no distinctive change of PsbO1 gene expression was observed either in cyp38 mutant or cyp38-C (Supplementary Figure S3). The detection and analysis of PsbO activity was carried out at the same time. In cyp38 mutant, PsbO activity rose distinctly after HL treatment ( Figure 6B). In cyp38-C, PsbO activity was similar to that of WT. In sum, CYP38 has an inhibiting effect on PsbO activity and influence of its transcription, the main target is PsbO2.

Gene Expression of CYP38 is Related to Ca 2+ -CaM3
Reactive oxygen species, as a widely studied oxidative stress molecular, can also function as cellular second messenger. Recently, it is proved that ROS plays an important role in cell signal transduction, one target being Ca 2+ (Mori and Schroeder, 2004;Bose et al., 2014). The results showed that, there was a significant increase in the level of ROS in the plant cells under HL conditions, subsequent to the cytoplasmic calcium [Ca 2+ ] cyt , GTPase activity analysis using CsA-pretreated PSII membranes of WT, psbo1 and psbo2 after HL treatment. Detached Arabidopsis thaliana leaves or PSII membranes were pre-incubated with or without CsA (10 µM), and then treated with HL. WT GTPase activity at control light was set to 1. Different letters indicate statistically signifcant differences between treatments (Duncan's multiple range test: P < 0.05). Asterisks ( * ) indicate a significant difference from the Control at * P < 0.05, * * P < 0.01. content increased gradually; after the removal of ROS by AsA, the level of Ca 2+ significantly dropped (Figures 7A,B). These results indicated that ROS may be an important signal molecular to increase the concentration of the [Ca 2+ ] cyt . Ca 2+ , as an important second messenger in plant cells, can function under abiotic stresses (Bowler and Fluhr, 2000;Li et al., 2012). We detected an increase in [Ca 2+ ] cyt and upregulation of the CAM3 gene transcript level under HL condition (Figures 7A,C). In order to verify whether this change associated with the function of CYP38, the AsA, BAPTA-AM , CAP and cam3 mutant was used. Under HL condition,  WT,cyp38 and cyp38-C with or without HL treatment. (B) GTPase activity analysis using PSII membranes extracted from WT, cyp38 and cyp38-C, then treated with or without HL for subsequent test. WT GTPase activity at control light was set to 1. Error bars are ± SD values for three replicates. Asterisks ( * ) indicate a significant difference from the control at * P < 0.05, * * P < 0.01. either scavenging [Ca 2+ ] cyt with BAPTA-AM or abolishing ROS by treatment with AsA dramatically declined CYP38 gene expression (Figures 7D,E). The CYP38 gene expression was also droped in cam3 mutant plants. On the contrary, rapid [Ca 2+ ] cyt elevation was induced by CPA can increase its transcription. The results above indicate that gene expression of CYP38 is related to cell redox signal induced [Ca 2+ ] cyt -CaM3 cascade.

Discussion
This study is an attempt to understand the possible molecular mechanisms in order to explain how CYP38 plays an important role in protection of PSII in Arabidopsis under HL treatment.
The results shown in this paper provide evidence for the cellular signaling cascade of the function of CYP38 in Arabidopsis in response to HL stress.
Recent studies have reported that deletion of CYP38 in Arabidopsis results in stunted growth and CYP38-deletant is highly sensitive to HL environment (Fu et al., 2007;Sirpiö et al., 2008; Figure 1A). The present study observed that, under HL condition, cyp38 leaves were more likely to be damaged by oxidation than WT leaves (Figure 1B). At the same time, experiment at the sub-cellular level showed that, in the protoplasts derived from cyp38 mutant, there was excessive outbreak of ROS which mainly subject to co-localization with chloroplasts (Figure 3). Further detected the change of CYP38 gene expression under HL, and discovered that its gene expression was increasing under HL condition compared with controlled and low light condition (Figure 2; Supplementary Figure S1), indicating that CYP38 plays an important role in the protection of the chloroplasts in HL response. Sirpiö et al. (2008) have demonstrated that CYP38 deletion drastically decreases the in vivo phosphorylation of D1 protein in Arabidopsis. This research discovered that, under HL treatment, the degradation rate of D1 protein in cyp38 mutant was distinctly higher than that of WT (Figure 4). Literature reports in the past pointed out that, first of all, the damaged D1 undergos phosphorylation (Koivuniemi et al., 1995;Rintamäki et al., 1995a). As a signal for location change, the D1 undergoing phosphorylation is guided and migrated from grana membranes to stroma lamellae, where D1 undergos dephosphorylation and later on be degraded (Tikkanen et al., 2008). To sum up, CYP38 deletion causes the inhibiting effect on thylakoid membrane phosphatase to disappear, and correspondingly facilitate dephosphorylation of p-D1, and therefore strengthen degradation of the damaged D1. PsbO2 also has been showed to have positive regulation effect on dephosphorylation and turnover of the PSII reaction center D1 protein (Lundin et al., 2007b). CYP38 and PsbO2 play important roles in D1 turnover process, which indicates that there may be some relevance between them (see the discussion below).
CYP38 protein is identified to be present in the thylakoid lumen. However, the research of Sirpiö et al. (2008) show that considerable amount of CYP38 protein is also associated with PSII monomer. As far as we know, these associated proteins of CYP38 with PSII monomer may participate in relevant regulation of membrane protein, rather than merely being a structural composition. Our data presented that PsbO2 gene expression and its GTPase activation were increased by CsA pro-treatment in WT plants compared with control ( Figure 5). Further detection in cyp38 mutant discovered similar results. However, in CYP38 complementation line PsbO2 gene expression and its GTPase activation fell down (Figure 6). We speculate that, the change of PsbO2 transcription mainly because of the redox diversification of the plant cell related to CYP38. All these indicated that CYP38 may, through inhibition of PsbO2 activity and then influence the dephosphorylation rate of the damaged D1.
Previous studies discovered that, under HL condition, the degradation rate of p-D1 is lower than that of D1 (Rintamäki et al., 1995b). Therefore, as a mode of existence, p-D1 may FIGURE 7 | Effect of ROS and Ca 2+ -CaM3 cascade on CYP38 gene expression. (A) Sub-cellular localization and content of HL-induced Ca 2+ production. Protoplasts were treated with or without HL for the indicated time period, then incubated with Fluo-3-AM for 30 min at room temperature, and observed using LCSM as described in materials and methods. Chloroplast autofluorescence is false colored red, and Fluo-3 fluorescence is false colored green. Scale bars = 10 µm.  be more stable than D1. We consider that the high gene expression of CYP38 under HL and inhibition of PsbO2 prevent dephosphorylation-based degradation of p-D1 protein, and as a result they transform D1 to p-D1 which is relatively more stable. In this way, the excessive degradation loss caused by HL is avoided. Other reports note that PSII existing in the form of p-D1 may no longer have the activity for transformation of light energy, however, the energy dissipation of such PSII can be enhanced (Critchley and Russell, 1994), which is beneficial for the adaptation to HL stress. From another perspective, CYP38 expression drops down to certain levels under LL conditions compared with HL (Figure 2; Supplementary Figure S1), which may release inhibition of PsbO2 and therefore enhance turnover of the damaged D1, facilitate PSII repair and increase efficiency of photosynthetic activity.
When plants are under abiotic threatening circumstances (such as excessive illumination), concentration of chloroplsts ROS grows and Ca 2+ levels in the cytosol increases as well (Baxter et al., 2014). If the balance between light absorption and the use of light energy is disturbed, photosynthesis will be further inhibited (Asada, 2006). In addition, the lack of light energy consumption leads to photoinhibition of PSII and enhances generation of 1 O 2 , which results in sequential reduction to superoxide anion radical (O • 2 −), hydrogen peroxide (H 2 O 2 ) and hydroxyl radicals ( • OH; Foyer and Noctor, 2000). This study shows that a large amount of 1 O 2 is generated in the cells under excessive light. DAB-stained leaves also tell that a great amount of H 2 O 2 is produced under the same condition (Figures 1 and 3). Recently, some articles claim that H 2 O 2 'signatures' , including antioxidant defense signal and well-known cytosolic calcium 'signatures' Ozgur et al., 2014;Pottosin et al., 2014), may operate in plant signaling networks. Ca 2+ , as an important messenger in plant cells (Hashimoto and Kudla, 2011), can function under various stimuli (Bowler and Fluhr, 2000;Romeis et al., 2001;Li et al., 2012). Under HL stress, we found that concentration of [Ca 2+ ] cyt and CAM3 gene expression increased, and inhibited [Ca 2+ ] cyt by BAPTA-AM or deficient CaM3, which drastically decreased CYP38 gene expression in Arabidopsis under HL condition, however, pre-incubated with CPA, which induced rapid elevation of [Ca 2+ ] cyt , and drastically increased CYP38 gene expression. Additionally, the level of [Ca 2+ ] cyt and CYP38 gene expression were significantly inhibited after removal of ROS by AsA, indicating that ROS generation may be a upstream signaling in these process (Figure 7). Our data present a cellular signaling cascade, composed of ROS production, [Ca 2+ ] cyt increase and the upward regulation of CaM3 transcript level, which function in the upstream of CYP38 gene expression in response to HL.

Conclusion
Our data showed that CYP38-mediated inhibition of PsbO2 activity modulated the tolerance of Arabidopsis to HL stress. According to our experimental results, a potential cascade of cellular events under HL stress was given (Figure 8): Firstly, HL stress induced ROS production in chloroplast, causing change to chloroplastic redox state. This signal might cause the increase of [Ca 2+ ] cyt and the activation of Ca 2+ -CaM3 cascade, and subsequently facilitated CYP38 expression. The increase in CYP38 greatly inhibited PsbO2 activity, which weakened the dephosphorylation of D1 protein and kept it in a relatively stable phosphorylation state to avoid degradation. These results present a possible mechanism for plants to respond to external HL stress, and shed a new insight into the cellular signaling cascade of CYP38.