CBSX2 is required for the efficient oxidation of chloroplast redox‐regulated enzymes in darkness Authors: Yonghong Li, Lin Zhang, Yurou Shen, Lianwei Peng, Fudan Gao Published: 2023 ABSTRACT AbstractThiol/disulfide‐based redox regulation in plant chloroplasts is essential for controlling the activity of target proteins in response to light signals. One of the examples of such a role in chloroplasts is the activity of the chloroplast ATP synthase (CFoCF1), which is regulated by the redox state of the CF1γ subunit and involves two cysteines in its central domain. To investigate the mechanism underlying the oxidation of CF1γ and other chloroplast redox‐regulated enzymes in the dark, we characterized the Arabidopsis cbsx2 mutant, which was isolated based on its altered NPQ (non‐photochemical quenching) induction upon illumination. Whereas in dark‐adapted WT plants CF1γ was completely oxidized, a small amount of CF1γ remained in the reduced state in cbsx2 under the same conditions. In this mutant, reduction of CF1γ was not affected in the light, but its oxidation was less efficient during a transition from light to darkness. The redox states of the Calvin cycle enzymes FBPase and SBPase in cbsx2 were similar to those of CF1γ during light/dark transitions. Affinity purification and subsequent analysis by mass spectrometry showed that the components of the ferredoxin‐thioredoxin reductase/thioredoxin (FTR‐Trx) and NADPH‐dependent thioredoxin reductase (NTRC) systems as well as several 2‐Cys peroxiredoxins (Prxs) can be co‐purified with CBSX2. In addition to the thioredoxins, yeast two‐hybrid analysis showed that CBSX2 also interacts with NTRC. Taken together, our results suggest that CBSX2 participates in the oxidation of the chloroplast redox‐regulated enzymes in darkness, probably through regulation of the activity of chloroplast redox systems in vivo.Keywords: CBSX2, CF1γ subunit, NTRC, peroxiredoxin, photosynthesis, thioredoxin INTRODUCTION To ensure the efficiency of photosynthesis and other metabolic reactions, the activity of chloroplast enzymes needs to be flexibly controlled in response to changes in the light environment. Chloroplast redox regulation based on dithiol‐disulfide exchange is an important regulatory mechanism for rapid adaptation of chloroplasts to photometabolism (Cejudo et al., 2021). During photosynthesis, reducing power generated by the electron transport chain in the thylakoids can be transferred from ferredoxin (Fd) to Trx via Fd‐thioredoxins (Trx) reductase (FTR). Then, Trx transfers the reducing power to target enzymes such as the four Calvin–Benson cycle enzymes glyceraldehyde 3‐phosphate dehydrogenase (GAPDH), fructose 1,6‐bisphosphatase (FBPase), sedoheptulose 1,7‐bisphosphatase (SBPase) and phosphoribulokinase (PRK) as well as NADP‐malate dehydrogenase (MDH), glucose 6‐phosphate dehydrogenase (G6PDH) and chloroplast ATP synthase (Gütle et al., 2017). In darkness, these target enzymes are rapidly inactivated by oxidation through formation of disulfide bonds. Chloroplast ATP synthase on the thylakoid membrane belongs to the F‐type ATP synthases containing the perimembranous part CF1 and the intrinsic part CFo. During photosynthesis, accumulation of the protons in the thylakoid lumen leads to the generation of a proton motive force (pmf) across thylakoid membranes, which powers CFoCF1 to synthesize ATP in the chloroplast stroma (Hahn et al., 2018; Hisabori et al., 2013). Like other F‐type ATP synthases, chloroplast ATP synthase operates fully reversibly. Indeed, while chloroplast ATP synthase normally generates ATP during photosynthesis, it can also catalyze the hydrolysis of ATP when the pmf is insufficient to drive ATP synthesis (Hisabori et al., 2002). To prevent the waste of ATP in chloroplasts, regulation of chloroplast ATP synthase activity plays an important role during the normal operation of photosynthesis and in darkness (Davis & Kramer, 2020; Hahn et al., 2018). The activity of chloroplast ATP synthase is mainly regulated by changes of the redox state of the CF1γ subunit, which is thought to act as a “switch” that enables the activity of chloroplast ATP synthase to be rapidly switched off upon a transition from light to dark (Hahn et al., 2018; Nalin & McCarty, 1984). Redox regulation of the γ subunit of chloroplast ATP synthase is mediated by a stretch of 9‐amino acids containing two cysteine (Cys) residues (Cys199‐Cys205 in Arabidopsis) (Nalin & McCarty, 1984). Under light conditions, Trxs receive electrons from the photosynthetic electron transport chain through the FTR and then reduce the disulfide bond formed between the two Cys residues in the γ subunit, resulting in a reduced state of these Cys residues and in the activation of ATP synthase. Under dark conditions, the disulfide bond of the γ subunit is formed by oxidation to prevent the ATP synthase from hydrolyzing ATP (Konno et al., 2000; Yang et al., 2020). It is well established that upon illumination chloroplast Trxs mediate the transfer of reducing equivalents from the photosynthetic electron transfer chain through Fd and FTR to the target enzymes. Chloroplast Trxs include five subtypes of isoforms in Arabidopsis: m (four isoforms), f (two isoforms), y (two isoforms), x, and z (Cejudo et al., 2021; Nikkanen & Rintamäki, 2019; Thormählen et al., 2017). It has been shown that CF1γ can be reduced by both f‐type and m‐type Trxs (Schwarz et al., 1997; Sekiguchi et al., 2020). Chloroplasts also harbor an NADPH‐dependent redox pathway. NTRC (NADPH‐Trx reductase C) is unique to oxygenic photosynthetic organisms and consists of the NTR domain with a Trx at its C‐terminus (Serrato et al., 2004). As NTRC uses NADPH as a source of reducing power, it works independently of light, unlike the Trx pathway (Cejudo et al., 2021; Pérez‐Ruiz et al., 2006, 2017). It may contribute to reduce CF1γ under low light conditions (Carrillo et al., 2016). It is known that CF1γ is oxidized under dark conditions. Recently, two types of Trx‐like proteins, Trx‐like 2 (TrxL2) and atypical Cys His‐rich Trx (ACHT), were shown to be oxidation factors of Trx‐targeted proteins, such as fructose‐1,6‐bisphosphatase, Rubisco activase, and the CF1γ subunit of ATP synthase (Yokochi et al., 2021). Peroxiredoxins (Prx) are thiol peroxidases that are highly conserved throughout all biological kingdoms. Chloroplasts of plants contain three types of Prx, the classic 2‐cysteine peroxiredoxin (2‐Cys Prx), a bacteroferritin‐comigratory protein homolog Prx‐Q, and a type II peroxiredoxin named Prx‐II (Dietz, 2003). It has been shown that the 2‐Cys Prx A and B participate in the oxidation of thiol groups in the dark (Ojeda et al., 2018; Vaseghi et al., 2018; Yoshida et al., 2018). In the absence of photosynthetic electron transport, Trx may mediate the oxidation of reduced CF1γ and transfer reducing equivalents to hydrogen peroxide through the activity of 2‐Cys Prx (Cejudo et al., 2019). Cystathionine β‐synthase (CBS) domain‐containing proteins are widely distributed in bacteria, yeasts, animals, and plants (Ignoul & Eggermont, 2005; Kushwaha et al., 2009). The CBS domain consists of approximately 60 amino acids forming two α‐helices and three β‐strands. In the CBS‐containing proteins, the CBS domains are generally found as tandem repeats, particularly in pairs or quads (Bateman, 1997). Tandem paired CBS modules can bind adenosine derivatives, such as ATP, AMP, and SAM (S‐adenosyl‐L‐methionine), and regulate the function of related enzymes or other domains by binding these ligands (Baykov et al., 2011; Jeong et al., 2013; Scott et al., 2004). The CBSX subfamily of the CBS‐containing proteins contains a single CBS pair and likely regulates the function of their interacting targets (Kushwaha et al., 2009). The Arabidopsis genome encodes six CBSX proteins of which CBSX1 and CBSX2 are localized in chloroplasts (Ok et al., 2012). CBSX1 and CBSX2 interact with several Trxs, and both of them were shown to enhance the in vitro activity of chloroplast Trxs including isoforms of m2, f1, x, and y1 (Yoo et al., 2011). Recently, however, Baudry et al. found that CBSX2 selectively inhibits the activities of m‐type TRXs, and ATP can reverse this effect (Baudry et al., 2022). In addition, Murai et al. proposed that CBSX1 and CBSX2 may not function as Trx regulator for activation of the Calvin‐Benson cycle enzymes FBPase and SBPase in light conditions (Murai et al., 2021). Thus, it is still unclear how the chloroplast CBSX1 and CBSX2 modulate the activity of the Trx systems and their different targets. In this study, we found that the CBSX2‐knockout Arabidopsis mutant has altered non‐photochemical quenching (NPQ) induction during a dark–light transition. Further analysis showed that oxidation of the CF1γ subunit, Calvin–Benson cycle enzymes FBPase and SBPase are less efficient under these conditions in the mutant. According to the interactome of CBSX2, we propose that CBSX2 is involved in the oxidation of chloroplast redox‐regulated enzymes, probably through the regulation of Trxs, NTRC, and 2‐Cys Prx in the dark. MATERIALS AND METHODS Plant materials and growth conditions The mutant cbsx2 was isolated from the pSKI015 T‐DNA insertion Arabidopsis mutant pool (stock no.: CS31400). T‐DNA insertion mutant cbsx1 (SAIL_1271_E12) was obtained from NASC (Nottingham Arabidopsis Stock Centre). Plants were grown on soil at 23°C under long‐day conditions (16/8‐h light/dark at 50 μmol photons m−2 s−1). To complement the cbsx2 mutant, the cDNA of CBSX2 was fused with the sequence encoding the HA tag (influenza hemagglutinin protein epitope) and then cloned into the pBI121 vector. The binary plasmid (pBI121–35 s:CBSX2:HA) was used to transform cbsx2 mutant by floral dipping. Chlorophyll fluorescence measurements NPQ induction and PSII quantum yield Y (II) were measured with a default program with a FluorCAM 800‐O (Photon System Instruments, Zhang et al., 2016). Plants were dark‐adapted for at least 30 min. The maximum chlorophyll fluorescence in the dark was obtained by a saturating light pulse. During illumination with actinic light (AL, 50 μmol photons m−2 s−1), saturating light pulses were applied at 9, 36, 64, and 91 s. NPQ and Y (II) images were recorded and extracted using the system software. At least three leaves were selected to derive NPQ and Y (II) data for dynamic analysis upon illumination. Protein extraction and protein gel blot analyses Thylakoid membrane isolation and 2D‐PAGE analysis were performed as previously described (Zhang et al., 2018). Antibodies against D1 (PHY0103, PhytoAB), PsaA (PHY0342, PhytoAB), CF1α (PHY0312, PhytoAB), and CF1γ (PHY0313, PhytoAB) were obtained from PhytoAB (USA). AMS labeling of CF1γ, SBPase, and FBPase The redox state of the γ subunit was determined using the specific thiol‐labeling reagent 4‐acetoamido‐4′‐maleimidylstilbene‐2, 2′‐disulfonate (AMS), which only labels the free thiol groups of cysteine residues. The procedure was performed as previously described (Carrillo et al., 2016). Leaves with similar size were ground into powder in liquid nitrogen and then transferred into a 1.5 ml centrifuge tube containing 500 μL of ice‐cold 10% (v/v) trichloroacetic acid. The sediment was collected by centrifugation and washed twice with ice‐cold acetone. After drying, 40 μL of AMS labeling solution (8 mM AMS, 50 mM Tris–HCl, pH 8.0, and 1% w/v SDS) was added and incubated for 2 h at room temperature in the dark. Proteins in the solution were separated by nonreducing 12% SDS‐PAGE. The proteins were transferred to a nitrocellulose membrane and incubated with specific antibodies against the CF1γ subunit, SBPase (PHY0410S, PhytoAB), and FBPase (PHY3095A, PhytoAB). The immunoblot signal was captured by the chemiluminescence imaging system (LuminoGraph WSE‐6100; ATTO Technology). Measurement of ΔA515 relaxation kinetics Four‐week‐old plants were dark‐adapted overnight. The ΔA515–560 nm signal was detected during excitation of the adaxial leaf side by 200,000 μmol photons m−2 s−1 with a single turnover flash of 5 μs duration using the Dual‐PAM‐100 system equipped with the P515/535 module (Walz, Germany). Affinity chromatography and mass spectrometry analysis Freshly isolated intact chloroplasts from WT and cbsx2‐HA‐com plants were cross‐linked with 2.5 mM DSP (dithiobis [succinimidyl propionate]) as previous described (Zhang et al., 2016). After crosslinking, Tris–HCl pH 7.5 was added to a final concentration of 60 mM to stop the reaction and chloroplasts were osmotically ruptured in a buffer containing 20 mM HEPES‐KOH pH 7.6 and 10 mM MgCl2 supplemented with protein inhibitor cocktail (Roche). Thylakoids were removed by centrifugation twice at 18,000 g for 10 min at 4°C. The supernatant containing chloroplast stromal proteins was incubated with anti‐HA MicroBeads (Miltenyi Biotec) for 2 h at 4°C. The bound proteins were purified according to the manufacturer's instructions and separated by SDS‐PAGE. The gel strip was then directly used for liquid chromatography‐mass spectrometry (LC–MS) analysis. Yeast two‐hybrid (Y2H) Y2H analysis was performed following the MatchmakerTM Gold Y2H System procedure (Clontech, Mountain View, CA, USA). The sequences encoding the mature NTRC and CBSX2 proteins were fused into the GAL4 activation domain vector (pGADT7) and the GAL4 binding domain vector (pGBKT7), respectively. The resulting vectors were cotransformed into Y2HGold yeast strain by the small‐scale lithium acetate method. Positive colonies grown on double‐dropout medium (SD/‐Leu/‐Trp) were cultured on tripartite‐dropout (SD/‐Leu/‐Trp/‐His) and quadruple‐dropout (SD/‐Leu/‐Trp/‐His/‐ade) medium in the presence of X‐α‐Gal to investigate protein–protein interactions. RESULTS Isolation and identification of the cbsx2 mutant The mutant cbsx2 was identified within a pSKI015 T‐DNA insertion mutant pool of Arabidopsis (Figure 1a). This mutant shows different NPQ induction after illumination compared with wild‐type (WT) plants (Figure 1b,c). During transition from dark to light, activation of photosynthetic electron transport induces accumulation of protons in the thylakoid lumen. Acidification of the thylakoid lumen subsequently triggers the induction of the main component of NPQ, qE, which provides efficient photo‐protection (Niyogi, 1999). ATP synthase is inactivated in darkness and its full activation upon illumination occurs after a lag period. Thus, during transition from dark to light, NPQ was rapidly and transiently induced to 0.8 after illumination for 40 s in WT (Figure 1b,c). Proton efflux from the thylakoid lumen through ATP synthase occurs along with the activation of ATP synthase upon illumination. Thus NPQ was rapidly relaxed within 2 min of illumination of the WT (Figure 1b,c). In the cbsx2 mutant, the induction of NPQ was higher than in WT after illumination for 9 s. However, the NPQ level in cbsx2 was lower than in WT after illumination for approximately 36 and 64 s, indicating that relaxation of NPQ is more efficient in the cbsx2 mutant (Figure 1b,c). The cbsx2 mutant did not exhibit a visible growth phenotype under growth chamber conditions, and the maximum photochemical efficiency (Fv/Fm) was comparable to that of WT (Figure 1b) indicating that photosystem II (PSII) function in cbsx2 was not affected. However, we also found that the quantum yield of photosystem II [Y (II)] of cbsx2 was higher than in WT after illumination for 40 s (Figure 1b,d), suggesting that the photosynthetic electron transfer of cbsx2 was enhanced upon illumination. Phenotype of the cbsx2 mutant. (a) Schematic representation of CBSX2. The cbsx2 mutant has a deletion of 14 bases in CBSX2 (AT4G34120). (b) Growth, non‐photochemical quenching (NPQ), maximum photochemical efficiency (Fv/Fm), and PSII quantum yield Y (II) phenotypes of WT, cbsx2, and cbsx2 com. Four‐week‐old plants grown on soil at 50 μmol photons m−2 s−1 (left panels). Signal intensities for NPQ and Y (II) at various times upon illumination with AL 50 μmol photons m−2 s−1 and Fv/Fm are indicated according to the color scale (0 to 1.0) on the right. Time course of NPQ induction (c) and Y (II) (d) of WT, cbsx2, and cbsx2 com. Data are presented as the means ± SD (n = 3). Map‐based cloning and resequencing of the cbsx2 genome DNA revealed that 14 nucleotides are missing in the CBSX2 (AT4G34120) gene (Figure 1a). To confirm that the mutation of CBSX2 is responsible for the NPQ induction phenotype, we introduced the WT AT4G34120 cDNA sequence into cbsx2 (cbsx2 com), which fully complemented the cbsx2 phenotype (Figure 1). These results indicate that the loss of the 14 nucleotides of CBSX2 is responsible for the NPQ phenotype in the cbsx2 mutant. Arabidopsis cbsx1 mutant does not have the same NPQ phenotype as cbsx2 A total of six CBSX proteins were found to have only one pair of CBS domains in Arabidopsis, and only CBSX1 and CBSX2 are localized in chloroplasts (Ok et al., 2012). To test whether cbsx1 exhibits the same NPQ phenotype as cbsx2, we analyzed the NPQ induction upon illumination of the single cbsx1 mutant and the cbsx1cbsx2 double mutant. As shown in Figure 2, both the single cbsx1 mutant and the double cbsx1cbsx2 mutant did not exhibit a visible growth phenotype under growth chamber conditions (Figure 2a). The induction of NPQ in cbsx1 is similar to that of WT (Figure 2). Moreover, the NPQ induction curve in the double cbsx1cbsx2 mutant shows identical defects as the single cbsx2 mutant (Figure 2). These results indicate that in contrast to CBSX2, CBSX1 is not involved in the induction and relaxation of NPQ. Growth and non‐photochemical quenching (NPQ) phenotypes of WT and cbsx mutants. (A) Growth and NPQ phenotypes of WT, cbsx2, cbsx1 and cbsx1cbsx2. (B) NPQ induction curves of WT, cbsx2, cbsx1, and cbsx1cbsx2. The AL intensity is 50 μmol photons m−2 s−1, data are presented as the means ± SD (n = 4). Accumulation of the chloroplast ATP synthase is not affected in the cbsx2 mutant The above NPQ induction properties suggest that chloroplast ATP synthase activity or accumulation is affected in the cbsx2 mutant. To distinguish between these two possibilities, we first analyzed the accumulation of the ATP synthase in cbsx2. Immunoblot analysis showed that the levels of the chloroplast ATP synthase subunits CF1α and CF1γ are similar in cbsx1, cbsx2, and WT plants (Figure 3), indicating that accumulation of the chloroplast ATP synthase is not affected in the absence of the CBSX1 and CBSX2 proteins. Thylakoid protein gel blot analysis in cbsx1 and cbsx2. Freshly isolated thylakoid membrane proteins were separated by SDS‐urea‐PAGE and immunodetected with the antibodies against CF1α, CF1γ, D1, and PsaA. Oxidation of the CF1γ subunit in darkness is less efficient in the cbsx2 mutant Since the ATP synthase level was not altered in the cbsx2 mutant, CBSX2 may be involved in the reduction of the γ subunit of the chloroplast ATP synthase. To investigate the redox status of CF1γ in the WT and cbsx2 plants, we first analyzed the redox state of CF1γ in the dark and light. Reduced and oxidized CF1γ subunit could be separated and quantified by the AMS labeling method (Konno et al., 2012). Immunoblot analyses showed that CF1γ subunit from WT plants was completely oxidized in the dark, but a small amount of CF1γ subunit in cbsx2 remained in the reduced state (Figure 4a). Under light conditions, most of the CF1γ subunits were present in the reduced form in WT and cbsx2, and identical levels of the redox forms of the CF1γ subunit were detected in these two genotypes (Figure 4a). The redox state of CF1γ in WT and cbsx2 leaves. (a) The redox status of CF1γ in leaves under light (L, 50 μmol photons m−2 s−1) and dark conditions (D). Proteins were extracted from leaves dark‐adapted overnight at least for 8 h (D) and illuminated (L). The proteins were treated with AMS and separated by SDS‐PAGE for immunoblot analyses with an antibody against CF1γ. (b) Changes in redox status of CF1γ in leaves transferred from light to darkness (L‐D). The leaves subjected to light were dark‐adapted for various times indicated above and proteins were analyzed as in (a). (c) Changes in redox status of CF1γ in leaves transferred from darkness to light (D‐L). Overnight dark‐adapted leaves were illuminated for various times as indicated above and proteins were analyzed as in (a). (d) ΔA515 relaxation kinetics in WT and cbsx2 leaves. The changes in the ECS signal (ΔA515) was measured using Dual‐PAM‐100 with a P515/535 module. The signal was detected during excitation of the adaxial leaf side by 200,000 μmol photons m−2 s−1 with a single turnover flash of 5 μs duration. The redox regulation of CF1γ subunit in plants mainly occurs in the reciprocal transition between darkness and light, so we examined the redox changes of CF1γ subunit in WT and cbsx2 leaves during the transition between darkness and light. As shown in Figure 4b, CF1γ subunit of WT is completely oxidized within 4 min after transition from light to darkness. However, a substantial amount of reduced CF1γ is still present in cbsx2 even for 6 min after transition from light to darkness (Figure 4b). The CF1γ subunit can be rapidly reduced within a short period of time upon illumination to produce the active ATP synthase in light. No significant difference was detected in the reduction rate of CF1γ subunit between WT and cbsx2 (Figure 4c). These results indicate that the oxidation of CF1γ subunit in darkness is less efficient and the reduction of CF1γ in light is not affected in the absence of CBSX2. Plants must turn off the chloroplast ATP synthase at night to prevent the hydrolysis of ATP when the chloroplast redox potential cannot sustain ATP synthesis (Kühlbrandt, 2019). However, a substantial amount of reduced CF1γ is present in the cbsx2 leaves even after staying overnight in darkness (Figure 4a), suggesting that a portion of chloroplast ATP synthase is still active in darkness. To test this possibility, we measured the changes in the electrochromic pigment absorbance shift (ECS) signal (ΔA515) caused by a single flip saturation flash using the leaves dark‐adapted for at least 12 h. As shown in Figure 4d, the ΔA515 signal did not significantly relax within 80 ms in WT leaves, indicating that the ATP synthase is inactive after dark adaption overnight. In contrast, the ΔA515 signal relaxed slowly in the cbsx2 leaves (Figure 4d), suggesting that the protons generated by the saturation flash flow out the thylakoid lumen slowly through the active ATP synthase. This conclusion is consistent with the presence of the reduced form of CF1γ in the cbsx2 leaves dark‐adapted overnight (Figure 4a). Oxidation of the Calvin cycle enzymes SBPase and FBPase in darkness is also less efficient in the cbsx2 mutant We also examined redox changes of the Calvin cycle enzymes SBPase and FBPase in WT and cbsx2 leaves. Immunoblot analysis showed that FBPase and SBPase are completely oxidized in WT plants in the dark, but a small amount of FBPase and SBPase are reduced in cbsx2 (Figure 5a). After transition from light to darkness, SBPase and FBPase in WT plants are completely oxidized within 2 and 1 min, respectively (Figure 5b). In contrast, a large amount of reduced SBPase and a trace amount of FBPase are still present in cbsx2 even 6 min after transition from light to darkness (Figure 5b). During transition from dark to light, a part of FBPase and SBPase are rapidly reduced both in WT and cbsx2 plants (Figure 5c). These results indicate that CBSX2 is also involved in the oxidation of the Calvin cycle enzymes FBPase and SBPase in darkness. The redox state of SBPase and FBPase in WT and cbsx2 leaves. (a) The redox status of SBPase and FBPase in leaves under light (L, 50 μmol photons m−2 s−1) and dark conditions (D). Proteins were extracted from leaves dark‐adapted overnight at least for 8 h (D) and illuminated (L). The proteins were treated with AMS and separated by SDS‐PAGE for immunoblot analyses with antibodies against SBPase and FBPase. (b) Changes in redox status of SBPase and FBPase in leaves transferred from L‐D. (d) Changes in redox status of SBPase and FBPase in leaves transferred from D‐L. CBSX2 coimmunoprecipitates with components of several chloroplast redox systems CBSXs have been identified as redox regulators of the thioredoxin system through their interaction with several Trxs (Yoo et al., 2011). To investigate the potential interaction partners of CBSX2, coimmunoprecipitation (Co‐IP) experiments were performed. We complemented the cbsx2 mutant with a chimeric gene encoding the CBSX2 protein carrying a HA (influenza hemagglutinin protein epitope) tag. The CBSX2‐HA protein and its interacting proteins were affinity purified from CBSX2‐HA plants using the mMACS HA isolation kit (Miltenyi Biotec). Total purified protein was separated by gel electrophoresis and further analyzed by LC–MS/MS analysis. The results showed that in addition to the thioredoxin (Trxs) members (f1, m1, m4, f1, y2, and CDSP32/Trx L1), peroxiredoxins (2CPA, 2CPB, PrxIIE, and PrxQ) were identified in the CBSX2‐HA‐IP sample (Table 1). Besides these proteins, NADPH‐dependent thioredoxin reductase C (NTRC), ferredoxin‐dependent glutamate synthase (GLU1, GLU2), SIR (sulfite reductase), ferredoxin‐nitrite reductase (NIR1), and ferredoxin‐NADP reductase (RFNR1, RFNR2, LFNR1, and LFNR2) were also co‐purified (Table 1). These results suggest that CBSX2 interacts with a wide range of substrates in vivo and participates in the regulation of multiple chloroplast redox systems. TABLE 1Redox related proteins detected in the CBSX2‐HA‐IP sample by IP‐MS.Protein nameGene no.CoverageUnique peptidesTrx f1 Thioredoxin f1At3g027306.741Trx m1 Thioredoxin m1At1g0368013.972Trx m2 Thioredoxin m2AT4G0352056.943Trx m4 Thioredoxin m4At3g1536016.583Trx y2 Thioredoxin y2AT1G435604.791CDSP32/Trx L1Thioredoxin like proteinAT1G7608011.2632‐Cys Prx A/2CPA2‐Cys peroxiredoxin AAT3G1163052.2652‐Cys Prx B/2CPB2‐Cys peroxiredoxin BAt5g0629032.61PrxIIEPeroxiredoxin‐II‐EAT3G5296029.066PrxQPeroxiredoxin QAT3G2606011.112NTRCNADPH‐dependent thioredoxin reductase CAT2G4168044.6123GLU1Ferredoxin‐dependent glutamate synthase 1AT5G0414034.2835GLU2Ferredoxin‐dependent glutamate synthase 2AT2G4122029.129SIRSulfite reductaseAT5G0459036.9223NIR1Ferredoxin‐nitrite reductaseAT2G1562037.7122RFNR1Root‐type ferredoxin: NADP(H) oxidoreductase 1AT4G0539019.844RFNR2Root‐type ferredoxin: NADP(H) oxidoreductase 2AT1G305108.122LFNR1Leaf‐type ferredoxin: NADP(H) oxidoreductase 1AT5G66190103LFNR2Leaf‐type ferredoxin: NADP(H) oxidoreductase 2AT1G200203.251 Redox related proteins detected in the CBSX2‐HA‐IP sample by IP‐MS. CBSX2 interacts with NTRC A previous report showed that NTRC regulates the activity of the chloroplast ATP synthase through reduction of the γ subunit (Nikkanen et al., 2016). NTRC is a NADPH‐dependent Trx reductase with a joint Trx domain and was co‐purified with CBSX2 (Table 1). Nikkanen et al. also detected an interaction between CF1γ and NTRC by Co‐IP experiments (Nikkanen et al., 2016). These results suggest that CBSX2 may regulate chloroplast ATP synthase activity via NTRC. To investigate whether CBSX2 directly interacts with NTRC, we performed Y2H experiments. As expected, yeast expressing the NTRC and CBSX2 can growth on the tripartite‐dropout and quadruple‐dropout mediums (Figure 6), indicating a physical interaction between CBSX2 and NTRC in yeast. Yeast two‐hybrid assay for the interaction between CBSX2 and NTRC. AD‐NTRC, the mature NTRC protein fused to the GAL4 activation domain (AD). BD‐CBSX2, the mature CBSX2 protein fused into the GAL4 binding domain (BD). Cotransformation of pGBKT7‐53 with pGADT7‐T and pGBKT7‐Lam with pGADT7‐T were used as positive and negative controls, respectively. Cotransformation of AD with BD‐CBSX2 and BD with AD‐NTRC were used to verify no self‐activation of the prey or bait. DDO/X, TDO/X, and QDO/X represent double‐dropout, tripartite‐dropout, and quadruple‐dropout medium containing X‐α‐Gal, respectively. Numbers represent the dilution of the yeast culture from 1× to 10,000×. DISCUSSION The activity of CFoCF1‐ATP synthase is reversible as it catalyzes both the synthesis and hydrolysis of ATP. In the absence of the light reactions that produce pmf across the thylakoid membrane, the enzyme is inactivated, preventing futile ATP hydrolysis in the dark. When the CF1γ thiol is oxidized, a rather high threshold pmf is required to activate the enzyme, and the break of the disulfide bridge in the CF1γ subunit reduces the threshold pmf required for activation (Hisabori et al., 2013; Konno et al., 2012; Nikkanen et al., 2016). The pmf across the thylakoid membrane includes the proton gradient (∆pH) and the electric membrane potential (∆Ψ). The ∆pH is a key signal for triggering the induction of NPQ of PSII, which is crucial for the dissipation of excess absorbed light energy to protect against photochemical damage of PSII (Wang et al., 2022). The altered NPQ induction pattern upon illumination in cbsx2 can be explained by the redox state of CF1γ in this mutant (Figures 1 and 4). In darkness, a small amount of reduced CF1γ was detected in the cbsx2 mutants (Figure 4a). Thus, it is likely that ATP hydrolysis occurs, which in turn leads to a flow of protons from chloroplast stroma to thylakoid lumen, resulting in a higher level of ∆pH across the thylakoid membrane in cbsx2 than in WT in the dark. Consequently, a higher NPQ was induced in cbsx2 compared with WT in the first few seconds after the transition from dark to light (Figure 1b,c). However, upon continued illumination, proton efflux from the thylakoid lumen through ATP synthase was higher in cbsx2 compared with WT, probably due to the higher initial CFoCF1‐ATP synthase activity in cbsx2 than in WT and because full activation of CFoCF1‐ATP synthase upon illumination needs some time in WT. Therefore, NPQ is lower in cbsx2 than WT after light for about 35 s (Figure 1b,c). Similar levels of NPQ were detected for WT and cbsx2 upon illumination for 120 s (Figure 1b,c), which is consistent with the previous finding that there is no significant change of ETR and NPQ in cbsx2 during steady‐state photosynthesis under different light intensities (Murai et al., 2021). We also found that Y (II) is higher in cbsx2 than WT after the transition from dark to light for about 40 s (Figure 1b,d), indicating enhanced electron transport through PSII in cbsx2. Analyses of the redox state of Calvin‐Benson cycle enzymes SBPase and FBPase in WT and cbsx2 leaves showed that a small amount of reduced SBPase and FBPase was detected in the cbsx2 mutants in darkness (Figure 5a). These results suggest that the Calvin‐Benson cycle enzymes are not fully inactivated in the dark in the absence of CBSX2. Reducing power generated by photosynthetic electron transport may be more efficiently transferred to the Calvin‐Benson cycle in cbsx2 than in WT plants, thus enhancing the quantum yield of photosystem II. It is also possible that enhanced Y (II) resulted from the reduced NPQ level during the transition from dark to light (Figure 1b,c). Oxidation of CF1γ, FBPase, and SBPase is less efficient during a transition from light to darkness in the cbsx2 mutant (Figures 4 and 5), suggesting that CBSX2 generally participates in the oxidation of chloroplast redox‐regulated proteins including Calvin cycle enzymes and CF1γ in the dark and during a transition from light to dark. Our IP‐MS results showing that several components of the chloroplast redox systems including peroxiredoxin (2CPA, 2CPB, PrxIIE, and PRXQ), Trx f, Trx m, CDSP32/Trx L1 are coimmunoprecipitated with CBSX2 (Table 1). Among them, Trx f and Trx m are thought to be modulators of thiol enzymes in the Calvin‐Benson cycle (Michelet et al., 2013; Murai et al., 2021). An impaired reoxidation of redox‐regulated proteins during the light–dark transition was found in the Arabidopsis plants with decreased 2‐Cys Prx (2CP) levels (Yoshida et al., 2018). Moreover, about half of CF1γ remained in the reduced form in the 2cpa 2cpb double mutants (Yoshida et al., 2018). PRXQ and PRXIIE may act as Trx oxidases in regulating photosynthesis and chloroplast metabolism similar to 2CP (Telman et al., 2020). CDSP32/Trx L1 mediates electron transfer from Fd to 2CP in chloroplasts, playing a role in assisting 2CP to clear H2O2 (Broin & Rey, 2003). Recently, it has been shown that Trx f and two types of Trx‐like proteins, Trx‐L2 and ACHT (atypical Cys‐His‐rich Trx), are involved in the oxidation of different Trx‐targeted proteins, such as glyco‐1, 6‐diphosphatase, Rubisco activase, and the γ subunit of ATP synthase (Sekiguchi et al., 2022; Yokochi et al., 2021). In contrast to the FTR‐Trx system, NTRC appears to function primarily in the dark or under limited light, possibly because it derives its reducing potential from NADPH, which can be produced independently of photosynthesis through the pentose phosphate oxidation pathway (Carrillo et al., 2016). The NTRC activity was proposed to be modulated between light and dark conditions (Carrillo et al., 2016). Nikkanen et al. (2016) confirmed the interaction between CF1γ and NTRC by CoIP, and CBSX2 was shown to interact with NTRC by bimolecular fluorescence complementation assays (González et al., 2019). We also confirmed the direct interaction between CBSX2 and NTRC by IP‐MS and Y2H assays (Figure 6; Table 1). Similar with the cbsx2 mutant, a small amount of CF1γ is also present in the reduced form in plants with overexpressed NTRC (Nikkanen et al., 2016; Figure 4a). Recently, in vitro biochemical studies confirmed that CBSX protein acts as a negative regulator of NTRC in the presence of AMP, and even low concentration of CBSX protein can significantly inhibit the activity of NTRC (Tran et al., 2023). Thus, we propose that CBSX2 is involved in the oxidation of a large set of chloroplast redox‐regulated enzymes by modulating the activity of these chloroplast redox systems in vivo. AUTHOR CONTRIBUTIONS Lianwei Peng and Fudan Gao conceived the study and designed the experiments. Yonghong Li, Yurou Shen, and Lin Zhang performed the experiments and produced the figures. Yonghong Li, Lianwei Peng, and Fudan Gao wrote the manuscript. All authors analyzed the data. All authors contributed to the article and approved the submitted version. CONFLICT OF INTEREST STATEMENT The Authors did not report any conflict of interest. PEER REVIEW The peer review history for this article is available in the Supporting Information for this article. Supporting information Data S1. Peer Review. Click here for additional data file. ACKNOWLEDGMENTS We thank the NASC and ABRC for providing the mutant seeds. Li, Y. , Zhang, L. , Shen, Y. , Peng, L. , & Gao, F. (2023). CBSX2 is required for the efficient oxidation of chloroplast redox‐regulated enzymes in darkness. Plant Direct, 7(11), e542. 10.1002/pld3.542 Funding information This work was supported by the Shanghai Engineering Research Center of Plant Germplasm Resources (17DZ2252700) and Taishan University doctoral research start‐up fund project (5013230130). DATA AVAILABILITY STATEMENT All relevant data can be found within the manuscript. REFERENCES Associated Data Supplementary Materials Data S1. Peer Review. Click here for additional data file. Data Availability Statement All relevant data can be found within the manuscript.