Self-protection of cytosolic malate dehydrogenase against oxidative stress in Arabidopsis

Self-protection of cytosolic malate dehydrogenase against oxidative stress in Arabidopsis Abstract Plant malate dehydrogenase (MDH) isoforms are found in different cell compartments and function in key metabolic pathways. It is well known that the chloroplastic NADP-dependent MDH activities are strictly redox regulated and controlled by light. However, redox dependence of other NAD-dependent MDH isoforms have been less studied. Here, we show by in vitro biochemical characterization that the major cytosolic MDH isoform (cytMDH1) is sensitive to H2O2 through sulfur oxidation of cysteines and methionines. CytMDH1 oxidation affects the kinetics, secondary structure, and thermodynamic stability of cytMDH1. Moreover, MS analyses and comparison of crystal structures between the reduced and H2O2-treated cytMDH1 further show that thioredoxin-reversible homodimerization of cytMDH1 through Cys330 disulfide formation protects the protein from overoxidation. Consistently, we found that cytosolic thioredoxins interact specifically with cytMDH in a yeast two-hybrid system. Importantly, we also show that cytosolic and chloroplastic, but not mitochondrial NAD-MDH activities are sensitive to H2O2 stress in Arabidopsis. NAD-MDH activities decreased both in a catalase2 mutant and in an NADP-thioredoxin reductase mutant, emphasizing the importance of the thioredoxin-reducing system to protect MDH from oxidation in vivo. We propose that the redox switch of the MDH activity contributes to adapt the cell metabolism to environmental constraints. CytMDH1, dimerization, H2O2-triggered oxdation, overoxidation, sulfenic acid, thioredoxin Introduction NAD(P)-dependent malate dehydrogenases (MDHs, l-malate:NAD oxidoreductases; EC 1.1.1.37) reversibly catalyze the oxidation (dehydrogenation) of l-malate to oxaloacetate (OAA) while reducing NAD(P)+ to NAD(P)H. In plants, MDH isoforms play a role in central metabolism and redox homeostasis between organellar compartments (Faske et al., 1995; Scheibe, 2004; Scheibe and Dietz, 2012; Selinski et al., 2014). As in other plants, Arabidopsis MDH isoforms can be classified into different groups based on their localizations: three cytosolic NAD-dependent MDHs, cytMDH1, cytMDH2, and cytMDH3; two mitochondrial NAD-dependent MDHs, mitMDH1 and mitMDH2; one chloroplastic NAD-dependent MDH, chlMDH; two peroxisomal NAD-dependent MDHs, perMDH1 and perMDH2; and one NADP-dependent MDH (NADP-MDH) (see Supplementary Fig. S1 at JXB online). NADP-MDH, localized in the chloroplast, is redox regulated through two sets of redox-sensitive cysteine pairs, located at the N- and C-terminus, which form two intramolecular disulfides (Cys77–Cys82 and Cys418–Cys430) that inhibit the enzymatic activity (Issakidis et al., 1992, 1994; Faske et al., 1995; Ruelland et al., 1998; Johansson et al., 1999; Carr et al., 1999). The inactive disulfide-containing state of NADP-MDH can be reduced to its active form through thiol/disulfide exchange with specific classes of disulfide reductases, namely thioredoxins (TRXm and TRXf), which themselves are reduced by a light-dependent ferredoxin-TRX reductase (reviewed by Jacquot et al., 1997; Lemaire et al., 2007; Meyer et al., 2009). Active NADP-MDH oxidizes NADPH to convert OAA into malate. Malate is then transported to the cytosol, where the cytMDH converts it back to OAA, while reducing NAD+ to NADH in the process. This so-called ‘malate valve’ is the first step in the redistribution of chloroplastic reducing equivalents to other parts of the plant cell (Scheibe, 2004; Hossain and Dietz, 2016). Two decades ago, chlMDH was reported as a non-redox-regulated protein (Berkemeyer et al., 1998), and recently mitMDH was shown to be redox independent (Yoshida and Hisabori, 2016). As the redox-sensitive cysteine pairs of the NADP-MDH are absent in cytMDH isoforms, it had not been considered as a redox-sensitive enzyme. However, cytMDH was captured on aTRX affinity chromatography column (Yamazaki et al., 2004), and found to be S-nitrosothiolated upon gaseous nitric oxide treatment in Arabidopsis (Lindermayr et al., 2005). More recently, cytMDH was detected in a sulfenylated state (–SOH) in Medicago truncatula (Oger et al., 2012), and in Arabidopsis cell suspensions (Waszczak et al., 2014; Akter et al., 2015). In an in vitro study, CuCl2 treatment was shown to inactivate cytMDH and a disulfide-linked homodimer was formed. Moreover, this disulfide bond was efficiently reduced by a cytosolic thioredoxin h1 (TRXh1) (Hara et al., 2006). In Arabidopsis, cytosolic TRXs are encoded by multigenic families for which TRXh members are the major isoforms (reviewed by Meyer et al., 2009). Most of these members are reduced by NADPH-dependent TRX reductases (NTRA and NTRB) that are dually located in the cytosol and mitochondria (Laloi et al., 2001; Reichheld et al., 2005). Some discrete TRXh members are alternatively reduced by the glutathione reductase (GR)/glutaredoxin (GRX) pathway, another major cytosolic thiol reduction pathway dependent on NADPH (Gelhaye et al., 2003). Here, we re-examined the redox dependence of cytMDH by combining in vitro biochemical, structural, and biophysical approaches. We reveal that oxidation triggers overoxidation of several cysteine residues through sulfenylation. Moreover, we show that under oxidative stress, cytMDH switches from a non-covalent dimer to a covalent disulfide-linked dimer, probably affecting the structure, the kinetics, and the stability of the enzyme. We indicate that the dimerization via Cys330–Cys330 disulfide acts as a redox switch, protecting cytMDH1 from overoxidation, and can be reduced by TRXs. Importantly, we show that such regulation also occurs in planta and might be conserved in other cytMDH isoforms. Therefore, based on both in vitro and in planta data, we propose a mechanism of how cytosolic NAD-MDH is redox regulated. Materials and methods Plant material and growth conditions Arabidopsis thaliana plants were grown in soil in a controlled growth chamber (180 µE m–2 s–1, 16 h day/8 h night, 22 °C 55% relative humidity day, 20 °C 60% relative humidity night) for up to 3 weeks. Plant mutant lines ntra ntrb, cat2, gr1, and gr1 cat2 were previously described (Queval et al., 2007; Reichheld et al., 2007; Mhamdi et al., 2010a). In vitro protein-based complementation and TRX activity assays For in vitro protein-based complementation assays, 5 µg of protein extracts (200 ng µl–1) were incubated with 1 mM NADPH, 4.59 μM TRXh3, and 3.12 µMNTRA in 25 µl 50 mM Tris–HCl (pH 7.5) for 2 h on ice. This reaction mixture was diluted 40 times in the same buffer, and the NAD-MDH activity assay was performed as described above. Yeast two-hybrid assay For yeast expression, full-length TRXh and MDH cDNA were respectively cloned in the pDEST32 and pDEST22 plasmids using the Gateway cloning system. Yeast two-hybrid assays were performed according to the method described previously (Vignols et al., 2005). Briefly, the yeast strain CY306 (a Δtrx1Δtrx2 mutant) was co-transformed with both pDEST22 and pDEST32 constructs and plated on YNB medium with the required amino acids and bases (His, Ura, Lys, Ade, and Met). Transformed colonies were subcultured and diluted using dilution series (diluted at 5 × 10–2, 5 × 10–3, and 5 × 10–4 at an optical density of 600 nm) prior to dotting on YNB –His –Trp –Leu in Petri dishes containing 20 mM 3-amino-1,2,4-triazole (3AT). Three independent transformations per binary assay were performed, of which 2–4 colonies were assayed. Images were taken 4 d post-dotting. Cloning, expression, and purification of recombinant cytMDH1 The cytMDH1-coding sequence (At1g04410), with attL and PreScission (EVLFQ/GP) sites at the N-terminal end, was synthesized by the Gen9 company. The gene fragment was inserted into the pDEST17 vector (Life Technologies). The constructs were transferred into Escherichia coli C41(DE3) strain. The expression and purification of recombinant His-tagged cytMDH1 was done as previously described (Hara et al., 2006). The purified His-tagged cytMDH1 was incubated with PreScission HRV3C enzyme (20:1, w/w) at 4 °C overnight to remove the His tag. The cleaved protein sample was loaded onto an Ni2+-Sepharose column equilibrated with 50 mM Tris–HCl, pH 7.5, 200 mM NaCl. The flow-through was collected and evaluated on an SDS–polyacrylamide gel as pure cleaved recombinant protein. Purified protein samples were pooled and flash-frozen in liquid nitrogen and then stored at –80 °C. Protein extracts from Arabidopsis plants and cytMDH1 enzymatic assay Purified cytosolic, mitochondrial, and chloroplastic protein extracts were prepared as described before (Daloso et al., 2015). NAD-MDH activity was measured from 3–5 µg of protein extracts at 25 °C in a reaction medium containing 50 mM Tris–HCl (pH 7.5), 250 µM NADH, and 2.5 mM OAA. Catalase activity was measured for 20 µg of protein extracts at 25 °C in a reaction medium containing 50 mM potassium phosphate buffer (pH 7.0) and 10 mM H2O2. The activity was monitored as a change in absorbance at 240 nm due to breakdown of H2O2. Purified recombinant cytMDH1 protein was incubated with 20 mM DTT at room temperature for 1 h to obtain reduced cytMDH1 (cytMDH1Red). cytMDH1Red (8 μM) was incubated with 10 mM H2O2 at 30 °C for 45 min to obtain oxidized cytMDH1 (cytMDH1Ox). For oxidation by diamide, 8 µM cytMDH1Red was incubated with 40 µM or 1 mM diamide at 30 °C for 45 min. For oxidation of the iodoacetamide (IAM)-alkylated cytMDH1, 20 µM cytMDH1Red was incubated with 10 mM IAM at room temperature for 20 min. Then 8 μM cytMDH1IAM was incubated with 10 mM H2O2 at 30 °C for 45 min. After each treatment, the protein samples were passed through a micro bio-spin column (Bio-Rad) to remove chemicals. The concentration of modified cytMDH1 was measured by absorbance at 280 nm. For cytMDH1 enzymatic assay, the reaction was performed in a 96-well plate at 30 °C with a final volume of 200 µL (ThermoFisher Scientific), containing 50 mM Tris–HCl pH 7.5, 2.5 mM OAA, 0.25 mM NADH, and 2.5 nM cytMDH1. The decrease in NADH absorbance at 340 nm was monitored using a SpectraMax 340PC spectrophotometer (Molecular Devices). The molar extinction coefficient for NADH of 6220 M–1 cm–1 was used for the calculation. To obtain the KM for OAA or NADH, progress curves were recorded using varying concentrations of OAA (0–2.5 mM) or NADH (0.0125–0.25 mM). The initial velocity (vi) for each substrate concentration was measured, and the vi/E0 values were plotted and fitted with the Michaelis–Menten equation to obtain the kinetic parameters. Three independent replicates of vi were measured for each substrate concentration. For the time course experiment with H2O2 incubation, 4.5 μM cytMDH1Red was incubated with 50 mM Tris–HCl (pH 7.5)/10 mM DTT/10 mM H2O2 at 30 °C for 2 h. CytMDH1 was diluted to 2.5 nM to measure its activity. For MDH reactivation by TRXh, cytMDH1Ox was incubated with the indicated concentrations of recombinant TRXh in the presence of NADPH (0.125 mM) and NTRA (0.5 µM). After 30 min incubation, the mix was diluted 40-fold and used for the NAD-MDH activity assay, as described above. In vitro sulfenylation Dimedone (500 mM; Sigma-Aldrich) was prepared freshly in DMSO. Then 8 µM cytMDH1Red was incubated with 2 mM dimedone in the presence of 10 mM DTT or 10 mM H2O2 at 30 °C for 45 min. The excess was removed on an equilibrated Micro Bio-Spin® Column Bio-Gel® P-6. The eluted samples were treated with 2 mM IAM at room temperature for 20 min. The samples were run on an SDS–polyacrylamide gel and the protein was transferred onto a polyvinylidene difluoride (PVDF) membrane. The western blot was done with 1:10 000 dilution rabbit anti-Cys-SOH antibody (Millipore); the result was visualized using Western Lightning Plus-ECL, Enhanced Chemiluminescence Substrate (PerkinElmer). LC–MS/MS The dimedone-treated cytMDH1Red and cytMDH1Ox samples were examined by SDS–PAGE and the bands were excised after Coomassie Brilliant Blue (CBB) staining. After destaining, the protein samples were digested by trypsin and subjected to LC-MS/MS analysis. The peptides were analyzed by LC-MS/MS as described (Pyr Dit Ruys et al., 2012). The mass spectrometer was operated in the data-dependent mode and switched automatically between MS, Zoom Scan for charge state determination and MS/MS for the most abundant ion. Each MS scan was followed by a maximum of five MS/MS scans using a collision energy of 30%. Dynamic exclusion was enabled to allow analysis of co-eluting peptides. The following parameters were used: trypsin was selected with cleavage only after lysine and arginine; the number of internal cleavage sites was set to 1; the mass tolerance for precursor and fragment ions was 1.1 Da and 1.0 Da, respectively; and the considered dynamic modifications on cysteine residues were +138.0 Da for sulfenic-dimedone, +32.0 Da for sulfinic, +48.0 Da for sulfonic, and +57.0 Da for carbamidomethyl modifications. For peptide identification, peak lists were generated using the application spectrum selector in the Proteome Discoverer 1.4 package. The resulting peak lists were searched using Sequest against an Arabidopsis protein database. Peptide matches were filtered using Percolator within Proteome Discoverer and manually validated. The mixed disulfide peptide between C330 and C330 of MDH was identified by the use of DBond software (Choi et al., 2010). All the annotated MS/MS spectra of modified peptides were manually evaluated. Melting and overall secondary structure assessed by circular dichroism (CD) The far CD spectra for cytMDH1Red and cytMDH1Ox were recorded in a JASCO J-715 spectropolarimeter equipped with a Peltier temperature control unit. The measurements were carried out in a 0.1 cm path length cuvette with an average of three scans for each spectrum, a 100 nm min–1 scan speed, band width of 2 nm, data pitch of 4 nm s–1, and the recorded range was from 190 nm to 260 nm. The buffer used was 50 mM ammonium bicarbonate, pH 7.4, and the protein concentration was in the range of 3–4.5 µM. Thermal unfolding was carried out at 4.5 μM protein concentration by increasing the temperature of the cuvette from 15 °C to 75 °C, and monitoring the intensity peak shift at 222 nm. Thermodynamic parameters and melting temperature of cytMDH1Red and cytMDH1Ox unfolding were calculated as described previously (Suh and Savisky, 2011). The melting temperature was calculated assuming that the unfolding process occurred following a simple two-state model. The peak shift at 222 nm over the specified temperature range was fitted into the following dose–response equation: y=A+Z−A1+1 0(logX0−X).h (1) The terms of Equation 1 are as follows: A is the folded state, Z is the unfolded state, logX0 is the melting temperature, and h is the Hill coefficient. The thermodynamic parameters were estimated based on the following equations: ΔG=2.303×R×h×(T×Tm–T2) (2) ΔH=2.303×R×T2×h (3) ΔG=ΔH–TΔS (4) where ΔG is the free energy Gibbs variation, ΔH is the enthalpy variation, ΔS is the entropy variation and T is the temperature (30 °C). Protein reduction and oxidation were performed as described before. Crystallization, X-ray data collection, and structure solution Crystals of native cytMDH1 were grown by hanging-drop vapor diffusion at 20 °C in droplets with 1:1 addition of 18 mg ml–1 cytMDH1 to a precipitant of 0.18 M ammonium sulfate, 0.09 M sodium acetate trihydrate, pH 4.6, 27% polyethylene glycol monomethyl ether 2000, and 10% glycerol. Crystals intended for oxidation were also obtained by hanging-drop vapor diffusion, with a crystallization solution of 1.5 M ammonium sulfate, 0.1 M Tris–HCl, pH 8.0. Crystallization droplets were composed of 1 µl of cytMDH1 and 1 µl of precipitant solution, with NAD+ present in 10× molar excess. For X-ray data collection, native cytMDH1 crystals were cryoprotected by supplementation of 80% mother liquor with 20% ethylene glycol. Diffraction data were collected using an in-house MicroMax-007HF X-ray generator with a Saturn 944+ detector. Crystals of cytMDH1 were oxidized through inclusion of 30 mM H2O2 in the cryoprotectant of 0.75 M ammonium sulfate, 50 mM Tris–HCl, pH 8.0, 16% ethylene glycol, 10% glycerol, 10% 1,2-propanediol, with a soak duration of 5 min. Diffraction data from these H2O2-treated crystals were collected at the Proxima 2a beamline at SOLEIL synchrotron (https://www.synchrotron-soleil.fr/en). Indexing and integration of reflections was performed in XDS (Kabsch, 2010a, b), and scaling and merging using AIMLESS (Evans and Murshudov, 2013). Initial phases were determined in Phaser (McCoy et al., 2007) of the PHENIX software suite (Adams et al., 2002, 2010) with Sus scrofa MDH as search model (PDB ID: 4MDH, 62% sequence identity). Manual model building was performed in COOT (Emsley and Cowtan, 2004), with maximum-likelihood refinement in Phenix.refine (Afonine et al., 2012), and REFMAC5 (Murshudov et al., 2011) of the CCP4 suite (Collaborative Computational Project, 1994). Structural stereochemistry was checked in MolProbity (Chen et al., 2010). The native and peroxide-treated crystal structures of cytMDH1 have been submitted to the PDB under the accession codes 5NUF and 5NUE, respectively. Statistics of data reduction and refinement are summarized in Supplementary Table S3. Biomolecular interfaces were analyzed, and biological assemblies assigned using the PISA web service (Krissinel and Henrick, 2007). All structural figures were prepared in PyMOL. Accession numbers Assigned accession numbers for the genes used in this work are as follows: At1g04410 (cytMDH1), At5g43330 (cytMDH2), At5g56720 (cytMDH3), At1g53240 (mtMDH1), At3g15020 (mtMDH2), At2g22780 (perMDH1), At5g09660 (perMDH2), At3g47520 (chlMDH), At5g58330 (NADP-MDH). Reduced and oxidized cytMDH1 structures were deposited at the PDB database (www.PDBe.org) with the respective IDs: 5NUF and 5NUE. The RNA sequencing (RNA-Seq) (Waszczak et al., 2016) and microarray data (Willems et al., 2016) discussed in this article have been deposited in the GEO repository (http://www.ncbi.nlm.nih.gov/geo/) and are accessible through the GEO accession nos GSE77017 and GSE80200, respectively. Results H2O2-treated cytMDH1 forms a disulfide-linked homodimer In order to characterize redox modifications of cytMDH1, we produced cytMDH1 in E. coli, and purified the protein to homogeneity. To validate cysteine sulfenylation in vitro, we applied dimedone, a chemical compound that forms a thioether bond with the electrophilic sulfur of sulfenylated proteins. CytMDH1Red migrated as an ~37 kDa band, which is very close to the expected size of 35.7 kDa based on the amino acid sequence. However, we detected an additional ~70 kDa band of cytMDH1Ox (10 mM H2O2) on the non-reducing CBB-stained SDS–PAGE gel (Fig. 1A), suggesting that cytMDH1 forms a covalently linked dimer upon oxidation. On immunoblots with anti-dimedone antibodies (Fig. 1A), we only observed sulfenylation of the ~37 kDa band of cytMDH1Ox. Remarkably, the dimer was not sulfenylated, suggesting that the covalent dimerization protected cytMDH1Ox from cysteine sulfenylation. We further studied whether the covalent dimer and the sulfenylated form were reduced by DTT and/or the NADPH-dependent TRX system (NTS). While the covalent homodimer was readily reduced either by DTT or NTS, consistent with previous data of Hara et al. (2006), the sulfenylated MDH monomer was only could be reduced by DTT (Fig. 1B). Fig. 1. View largeDownload slide cytMDH1Ox is partially sulfenylated in vitro and dimerizes via a TRX-reducible intermolecular disulfide bond. (A) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence of 2 mM dimedone. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed on a western blot with an anti-cysteine sulfenic acid antibody. (B) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence or absence of the NTS system. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed by immunoblot with an anti-cysteine sulfenic acid antibody. Fig. 1. View largeDownload slide cytMDH1Ox is partially sulfenylated in vitro and dimerizes via a TRX-reducible intermolecular disulfide bond. (A) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence of 2 mM dimedone. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed on a western blot with an anti-cysteine sulfenic acid antibody. (B) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence or absence of the NTS system. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed by immunoblot with an anti-cysteine sulfenic acid antibody. To confirm the molecular size of the reduced and oxidized cytMDH1, we evaluated their elution position on a size exclusion chromatography (SEC) column (Supplementary Fig. S2). Both cytMDH1Red and cytMDH1Ox eluted as a single peak of ~70 kDa, indicating that both cytMDH1Red and cytMDH1Ox are in a dimeric form in solution. We further examined cytMDH1 dimerization by migrating cytMDH1Red and cytMDH1Ox on native gels (Supplementary Fig. S3). Both cytMDH1Red and cytMDH1Ox migrated at an apparent molecular mass of ~100 kDa in either non-reducing or reducing conditions. The observed molecular mass is slightly higher than that expected for a dimer (~72 kDa), which could be due to the charge of the protein. However, consistent with the SEC data, no monomeric forms were observed under reducing or oxidizing conditions, supporting the fact that both dimeric forms co-exist (covalent and non-covalent) depending on the redox environment of the protein. cytMDH1Red dimer is probably organized by non-covalent interactions, while cytMDH1Ox is likely to be a mixture of non-covalent and covalent dimers (Fig. 1; Supplementary Figs S2, S3). We also detected several bands close to each other in cytMDH1Ox on the native gel (Supplementary Fig. S3), probably due to oxidation of other cysteines, which gives extra charges to the protein. Notably, the upper band (in between 100 kDa and 130 kDa) almost completely disappeared under reducing conditions. Whether this form corresponds to the covalent dimer cannot be fully confirmed. Collectively, our data suggest that cytMDH1 undergoes a partial conformational change from a non-covalent homodimer to a disulfide-linked homodimer under H2O2 stress. Cysteine and methionine sulfurs are sensitive to oxidation To confirm the sulfenylation of cytMDH1 in vitro and the disulfide bond formation of cytMDH1Ox, we analyzed the purified recombinant protein treated with or without H2O2 in the presence of dimedone by MS. After treatment with 10 mM H2O2 and 2 mM dimedone, we blocked all free thiols by IAM, and the cytMDHOx sample was loaded on a non-reducing SDS–polyacrylamide gel. The 37 kDa and 70 kDa bands (Fig. 1A) were in-gel trypsin digested prior to MS analysis. In the 37 kDa band, a dimedone adduct on the Cys330-containing peptide was found, resulting in a mass increase of 138 Da (Fig. 2A), showing that Cys330 was sulfenylated. Cys330 was also partially present as a free thiol (carbamidomethyl modification by IAM) (Supplementary Table S1) and overoxidized to a sulfonic acid (Fig. 2B). Furthermore, Cys79 was also sulfenylated and overoxidized to a sulfinic acid and sulfonic acid (Supplementary Table S1). All other cysteines were identified as free thiols (Supplementary Table S1). In the 70 kDa band, an intermolecular disulfide bond between Cys330 was found (Fig. 2C), as described previously (Hara et al., 2006). Besides these cysteine modifications, we also observed that methionines were oxidized to a methionine sulfoxide in both 35 kDa (Met24/30/56/87/313) and 70 kDa (Met24/30/41/56/87/270/313) bands (Supplementary Table S1). Among these oxidized methionines, Met313 had been identified previously as a methionine sulfoxide in the cat2 mutant when compared with wild-type Arabidopsis plants (Jacques et al., 2015). All together, we showed that several Cys and Met residues of cytMDH1 are sensitive to oxidation. Fig. 2. View largeDownload slide Identification of oxidative cysteine modifications of cytMDH1Ox by LC-MS/MS. (A) Sulfenylation identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=1009.3. The y and b fragments indicate that Cys330 is present as a dimedone group-modified sulfonic acid (+138.0681 Da). (B) Sulfonic acid identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=919.3. The y and b fragments indicate that Cys330 is present as a sulfonic acid (+48 Da). (C) Disulfide lined peptide identification by using Dbond software. A doubly charged parent ion of [M+2H]2+=871.5 Da shows fragmentation characteristics of a disulfide linkage between two Cys330s of MDH. The y and b series of ions allow the exact localization of the disulfide bridge, p* one strand of the dipeptide. Fig. 2. View largeDownload slide Identification of oxidative cysteine modifications of cytMDH1Ox by LC-MS/MS. (A) Sulfenylation identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=1009.3. The y and b fragments indicate that Cys330 is present as a dimedone group-modified sulfonic acid (+138.0681 Da). (B) Sulfonic acid identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=919.3. The y and b fragments indicate that Cys330 is present as a sulfonic acid (+48 Da). (C) Disulfide lined peptide identification by using Dbond software. A doubly charged parent ion of [M+2H]2+=871.5 Da shows fragmentation characteristics of a disulfide linkage between two Cys330s of MDH. The y and b series of ions allow the exact localization of the disulfide bridge, p* one strand of the dipeptide. Kinetics are affected by H2O2 and TRXh reactivates cytMDH1 We next investigated the effect of cytMDH1 oxidation on its enzymatic activity. The cytMDH1 activity progressively decreased upon H2O2 treatment. No decrease in activity could be observed for the DTT-treated sample over the same time range and temperature, indicating that oxidation of cytMDH1 is the major factor leading to loss of activity (Fig. 3A). To understand how oxidation changes the activity of cytMDH1, we compared the kinetic parameters of cytMDH1Red and cytMDH1Ox (Fig. 3B, C; Table 1). The kcat is 2-fold higher for cytMDH1Red than that of cytMDH1Ox, indicating a decrease of the rate of substrate conversion when the enzyme is oxidized. Interestingly, the KM of cytMDH1Ox for OAA is 143 ± 15 µM, which is 40% lower than that of 238 ± 21 µM for cytMDH1Red. Similarly, the KM of cytMDH1Ox for NADH (50 ± 5 µM) is also 30% lower than that for cytMDH1Red (72 ± 7 µM) (Fig. 3B, C; Table 1), indicating small increases in affinity for both substrates when cytMDH1 is oxidized. Overall, the kcat/KM values for OAA are similar for cytMDH1Red and cytMDH1Ox. The kcat/KM value of cytMDH1Ox for NADH is 6.9 × 106 M–1 s–1, which is in the same range as the 9.0 × 106 M–1 s–1 value obtained for cytMDH1Red. This suggests that a KM adjustment occurs, to compensate for the kcat decrease between cytMDH1Red and cytMDH1Ox. Fig. 3. View largeDownload slide CytMDH1 is sensitive to oxidation and reactivated by TRXh in vitro. (A) The time course of cytMDH1 activity was measured under reducing or oxidizing conditions. cytMDH1Red at 8 μM was incubated with buffer (cytMDH1Red), with 10 mM DTT (cytMDH1Red+DTT), or with 10 mM H2O2 (cytMDH1Red+H2O2) at 30 °C. The plot of k (s–1) versus time of treatment is shown. Data presented are means ±SD (n >3). (B and C) CytMDH1 activity was measured at different concentrations of substrates: (B) OAA (0–2.5 mM) or (C) NADH (0–0.25 mM). The initial velocity (vi) for each substrate concentration was measured, and the k (s–1) values versus [OAA] or [NADH] were plotted and fitted with the Michaelis–Menten equation to obtain the kinetic parameters KM and specific activity. At least three independent replicates of k were measured for each substrate concentration. (D) CytMDH1Ox was incubated with increasing concentrations of recombinant TRXh3 in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). (E) Reduced or H2O2-oxidized cytMDH1 was incubated with different TRXhs (2 µM) in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). An asterisk indicates values significantly different from cytMDH1Red by the Student’s t-test at 5‰ (*P<5 × 10–3). Fig. 3. View largeDownload slide CytMDH1 is sensitive to oxidation and reactivated by TRXh in vitro. (A) The time course of cytMDH1 activity was measured under reducing or oxidizing conditions. cytMDH1Red at 8 μM was incubated with buffer (cytMDH1Red), with 10 mM DTT (cytMDH1Red+DTT), or with 10 mM H2O2 (cytMDH1Red+H2O2) at 30 °C. The plot of k (s–1) versus time of treatment is shown. Data presented are means ±SD (n >3). (B and C) CytMDH1 activity was measured at different concentrations of substrates: (B) OAA (0–2.5 mM) or (C) NADH (0–0.25 mM). The initial velocity (vi) for each substrate concentration was measured, and the k (s–1) values versus [OAA] or [NADH] were plotted and fitted with the Michaelis–Menten equation to obtain the kinetic parameters KM and specific activity. At least three independent replicates of k were measured for each substrate concentration. (D) CytMDH1Ox was incubated with increasing concentrations of recombinant TRXh3 in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). (E) Reduced or H2O2-oxidized cytMDH1 was incubated with different TRXhs (2 µM) in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). An asterisk indicates values significantly different from cytMDH1Red by the Student’s t-test at 5‰ (*P<5 × 10–3). Table 1. Steady-state parameters for reduced and oxidized cytMDH1 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 View Large Table 1. Steady-state parameters for reduced and oxidized cytMDH1 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 View Large Previously, it has been shown that TRXh1 reactivates cytMDH1Ox by reducing a disulfide bond at the Cys330 residue (Hara et al., 2006). The Arabidopsis genome encodes five cytosolic TRX isoforms (TRXh1–TRXh5) that might reduce cytMDH1. In order to test if TRXhs could reduce cytMDH1, we first tested protein interactions between recombinant TRXh1–TRXh5 and MDH isoforms in a yeast two-hybrid system (Vignols et al., 2005). In order to stabilize redox-sensitive interactions, the resolving cysteine of the TRX active site was mutated to a serine. All five TRXh isoforms interacted with cytMDH1. However, no interaction between the cytosolic TRXhs and mitMDH1, chlMDH, and NADP-MDH was observed, indicating a specificity of the TRXhs to interact with cytMDH1 (Supplementary Fig. S4). To test if MDHs are able to interact with TRXs other than TRXh, we performed a yeast two-hybrid test using the chloroplastic TRXf1 and f2 as bait. All tested MDHs (mitMDH1, chlMDH, cytMDH1, and NADP-MDH) were able to interact with TRXf1 and f2, suggesting that TRXfs have a larger interaction spectrum than TRXhs for MDH isoforms (Supplementary Fig. S5). We further tested whether recombinant TRXhs are able to restore the activity of H2O2-oxidized cytMDH1. All TRXh1–TRXh5 isoforms identified as interactors of cytMDH1 by yeast two-hybrid assay were able to restore the activity of cytMDH1Ox to a level similar to that of cytMDH1Red (Fig. 3D, E), implying that the TRXh–MDH interaction occurs in a redox-dependent manner. The thermodynamic stability of cytMDH1 is affected by H2O2 In order to understand how H2O2 affects the activity of cytMDH1, the secondary structures of reduced and oxidized cytMDH1 were compared with far UV-CD (Fig. 4A). cytMDH1Red and cytMDH1Ox are characterized by a typical α-helical and β-sheet secondary structure with negative peaks around 210–220 nm and a positive peak around 190 nm. However, the spectrum of cytMDH1Ox shows a pronounced modification of the secondary structure, indicating a more disordered protein. We further determined the influence of oxidation on the protein stability by following the unfolding of oxidized and reduced cytMDH1 at 222 nm from 15 °C to 75 °C by CD. We obtained a melting temperature (Tm) of cytMDH1Red of 54.74 °C, while the Tm of cytMDH1Ox dropped by almost 5 °C to 49.88 °C (Fig. 4B), indicating that cytMDH1Ox is less stable than cytMDH1Red. Thermodynamic processes, such as the thermal unfolding of cytMDH1, are characterized by a variation in entropy (ΔS), enthalpy (ΔH), and Gibbs free energy (ΔG) (Murphy and Freire, 1992). Thermal unfolding parameters specify a higher need for energy in order to disrupt intramolecular interactions such as Van der Waals forces, hydrogen bonds, and ionic salt bridges for cytMDH1Red, based on the ΔΔH value of 8.6 kcal mol–1 (Table 2). These interactions are well known as the main intramolecular forces that keep proteins structured, indicating that the amount of energy to unfold a protein is directly proportional to the interactions previously mentioned (O’Brien and Haq, 2004). The more flexible a protein is, the higher are the number of possible conformations. This effect is represented by conformational entropy (O’Brien and Haq, 2004). The ΔΔS value of 23 cal mol–1 KJ–1 suggests that cytMDH1Red undergoes a larger conformational change on the way to reach its unfolded conformation than its oxidized counterpart (Table 2). This result indicates that before thermal denaturation, cytMDH1Red is in a more restricted number of conformations maintained by a larger amount of interactions when compared with cytMDH1Ox. Fig. 4. View largeDownload slide CD of reduced and oxidized cytMDH1. (A) The far CD spectra for the cytMDH1Red and cytMDH1Ox were obtained with an average of three scans for each spectrum with a range from 190 nm to 260 nm. (B) Thermal unfolding monitored at 222 nm was carried out for cytMDH1Red and cytMDH1Ox by increasing the temperature from 15 °C to 75 °C. Fig. 4. View largeDownload slide CD of reduced and oxidized cytMDH1. (A) The far CD spectra for the cytMDH1Red and cytMDH1Ox were obtained with an average of three scans for each spectrum with a range from 190 nm to 260 nm. (B) Thermal unfolding monitored at 222 nm was carried out for cytMDH1Red and cytMDH1Ox by increasing the temperature from 15 °C to 75 °C. Table 2. Tm and thermodynamic parameters of reduced and oxidized cytMDH1 thermal denaturation ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔΔ stands for reduced minus oxidized. a Thermodynamic constants calculated at 30 °C. View Large Table 2. Tm and thermodynamic parameters of reduced and oxidized cytMDH1 thermal denaturation ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔΔ stands for reduced minus oxidized. a Thermodynamic constants calculated at 30 °C. View Large Crystal structure of cytMDH1 In an effort to gain structural insights into the effects of oxidative stress on cytMDH1, the crystal structures of cytMDH1Red and H2O2-soaked cytMDH1 were determined to a resolution of 1.8 Å and 1.35 Å, respectively (www.PDBe.org, ID: 5NUF and 5NUE). CytMDH1 crystallizes as a homodimer with an interfacing area of 1817 Å2 as determined by PISA analysis, and exhibited the structural fold common to the cytosolic MDH from porcine heart (Fig. 5) (Birktoft et al., 1989). NAD+ was modeled into the available density in the mFO–DFC omit map at the nucleotide-binding cleft of cytMDH1. A sulfate ion was found to occupy the malate/OAA-binding site proximal to the nicotinamide head of NAD+, binding with five water molecules and the side-chain groups of residues Asn132, Arg163, His188, and Ser243. Fig. 5. View largeDownload slide Crystal structures of cytMDH1 and sulfoxidation. (A) The conserved dimeric structural arrangement of MDH places the C-termini at opposite ends. Here in the crystal structure of Arabidopsis cytMDH1, each protomer is colored separately, and the side chain of Cys330 and NAD+ are colored orange and green, respectively. (B) Met97 is located on a flexible loop proximal to the OAA/malate-binding site, which in the crystal structure is occupied by a sulfate ion. Superposition of two subunits of the cytMDH1 homodimer shows that this flexible loop can be in either a closed (pink) or an open (blue) conformation independent of the loop conformation of the neighboring dimer subunit. Sulfoxidation of Met97 observed in the crystal structure of cytMDH1 is indicated, though it should be noted that this flexible loop region can adopt a closed/open conformation irrespective of Met97 sulfoxidation. Fig. 5. View largeDownload slide Crystal structures of cytMDH1 and sulfoxidation. (A) The conserved dimeric structural arrangement of MDH places the C-termini at opposite ends. Here in the crystal structure of Arabidopsis cytMDH1, each protomer is colored separately, and the side chain of Cys330 and NAD+ are colored orange and green, respectively. (B) Met97 is located on a flexible loop proximal to the OAA/malate-binding site, which in the crystal structure is occupied by a sulfate ion. Superposition of two subunits of the cytMDH1 homodimer shows that this flexible loop can be in either a closed (pink) or an open (blue) conformation independent of the loop conformation of the neighboring dimer subunit. Sulfoxidation of Met97 observed in the crystal structure of cytMDH1 is indicated, though it should be noted that this flexible loop region can adopt a closed/open conformation irrespective of Met97 sulfoxidation. Superposition of the individual protomers of the cytMDH1 homodimer revealed significant differences in the conformation of a flexible loop of Pro92–Val102 at the opening to the substrate-binding site. In one conformation, this loop is in a ‘closed’ conformation, and co-ordinates a sulfate ion through the backbone amide and side-chain head of Arg99. The alternative ‘open’ conformation of this flexible loop region was found to be more disordered, with elevated B-factors and relatively poor coverage of electron density. No significant structural differences were observed between the structures of cytMDH1 and the H2O2-soaked cytMDH1 crystal, both structures aligning with a root mean square deviation (RMSD) of 0.266 Å, and the largest structural deviations arising in the flexible loop region described above (Fig. 5A). Evidence of partial sulfoxidation was identified on Met56 and Met97. Sulfoxidation of Met56 appears to have no direct effect on the local structural environment. The Met97 is located on the flexible loop proximal to the malate/OAA-binding site, which has also been previously identified to form methionine sulfoxide in Arabidopsis plants (Jacques et al., 2015). Sulfoxidation of Met97 could conceivably affect mobility of the Pro92–Val102 loop region, and thereby possibly influence substrate delivery to the active site (Fig. 5B). However, we found no significant difference in activity after we oxidized the IAM-pretreated cytMDH1 with 10 mM H2O2 (Supplementary Fig. S6), indicating that methionine oxidation does not play a major role in the effect of oxidation on the enzymatic capacity of cytMDH1. From a mFO–DFC difference map, additional positive density adjoining the sulfur of Cys330 was observed in two of the three molecules of the asymmetric unit of cytMDH1. Into this extra density, partial-occupancy sulfenylation or sulfinylation at Cys330 was modeled for the respective molecules of cytMDH1 (Fig. 6). The partial occupancy of the cysteine sulfenylation and sulfinylation may be due to the short duration of the H2O2 crystal soak, resulting in an unequal distribution of peroxide throughout the crystal unit cells. Fig. 6. View largeDownload slide Cysteine oxidation observed in the crystal structure of H2O2-treated cytMDH1. The presence of additional density arising in the FO–FC map at Cys330 of chain B (A) and chain C (C) prompted modeling of a partial-occupancy sulfinic and sulfenic acid (B and D, respectively). Shown are 2mFO–DFC (gray) and mFO-DFC (green) electron density maps contoured at 0.33 e Å–3 and 0.34 e Å–3, respectively, for (C), and 0.41 e Å–3 and 0.295 e Å–3 for both (B) and (D). The 2mFO–DFC map for (A) is contoured at 0.25 e Å–3. Fig. 6. View largeDownload slide Cysteine oxidation observed in the crystal structure of H2O2-treated cytMDH1. The presence of additional density arising in the FO–FC map at Cys330 of chain B (A) and chain C (C) prompted modeling of a partial-occupancy sulfinic and sulfenic acid (B and D, respectively). Shown are 2mFO–DFC (gray) and mFO-DFC (green) electron density maps contoured at 0.33 e Å–3 and 0.34 e Å–3, respectively, for (C), and 0.41 e Å–3 and 0.295 e Å–3 for both (B) and (D). The 2mFO–DFC map for (A) is contoured at 0.25 e Å–3. Redox sensitivity of NAD-MDHs varies in different plant cell compartments To obtain an overview of the redox sensitivity of NAD-MDH in the plant cell, we measured the NAD-MDH activity in purified cytosol, mitochondrion, and chloroplast fractions before and after the 10 mM H2O2 treatment. Both the cytosolic and chloroplastic NAD-MDH activities were significantly decreased after the H2O2 treatment (40% and 20%, respectively) (Fig. 7A). No significant difference of the mitochondrial NAD-MDH activities between the non-treated and treated samples was observed. Fig. 7. View largeDownload slide NAD-MDH activity in different cell compartments and mutants. (A) The NAD-MDH activity was measured for different cell compartments in wild-type (Col-0) plant extracts. A 3 µg aliquot of protein extracts was used for activity assays. Data presented are means ±SD (n ≥3). An asterisk indicates values significantly different from the wild type by the Student’s t-test (*P<0.05, **P<0.01). (B) The NAD-MDH activity was measured in wild-type (Col-0) and different mutant plants grown in soil for 2 weeks (white bars) or 3 weeks (black bars). A 5 µg aliquot of protein extracts was used for activity assays. (C) Protein-based complementation assays were performed in the same mutants as in (A). A 5 µg aliquot of protein extracts was untreated (black bars) or incubated in 1 mM NADPH in the presence of 4.59 μM TRXh3 and 3.12 µM NTRA for 2 h on ice (white bars). This reaction mixture was diluted 40 times before performing the NAD-MDH activity assay. Data presented are means ±SD (n=4). An asterisk indicates values significantly different from the wild type by the Student’s t-test at 5‰ (*P<5 × 10–3). Fig. 7. View largeDownload slide NAD-MDH activity in different cell compartments and mutants. (A) The NAD-MDH activity was measured for different cell compartments in wild-type (Col-0) plant extracts. A 3 µg aliquot of protein extracts was used for activity assays. Data presented are means ±SD (n ≥3). An asterisk indicates values significantly different from the wild type by the Student’s t-test (*P<0.05, **P<0.01). (B) The NAD-MDH activity was measured in wild-type (Col-0) and different mutant plants grown in soil for 2 weeks (white bars) or 3 weeks (black bars). A 5 µg aliquot of protein extracts was used for activity assays. (C) Protein-based complementation assays were performed in the same mutants as in (A). A 5 µg aliquot of protein extracts was untreated (black bars) or incubated in 1 mM NADPH in the presence of 4.59 μM TRXh3 and 3.12 µM NTRA for 2 h on ice (white bars). This reaction mixture was diluted 40 times before performing the NAD-MDH activity assay. Data presented are means ±SD (n=4). An asterisk indicates values significantly different from the wild type by the Student’s t-test at 5‰ (*P<5 × 10–3). In order to study the impact of H2O2 on the MDH gene expression, we analyzed the microarray and RNA-Seq data for all MDH genes (Supplementary Fig. S7). Microarray analyses were performed on 7-day-old wild-type (Col-0) Arabidopsis seedlings without and with 10 mM H2O2 treatment for 24 h (Willems et al., 2016). RNA-Seq analysis was done on 21-day-old Col-0 and cat2 mutant Arabidopsis seedlings under photorespiratory stress (Waszczak et al., 2016). No major change was found for mRNA levels of all MDHs upon stresses from both data sets. From the RNA-Seq data set, similar mRNA levels of MDHs were found between the Col-0 wild-type and cat2 mutant Arabidopsis plants. Collectively, these data indicate that H2O2 has little influence on the transcriptional level of MDHs, and that the impact of H2O2 on NAD-MDH activities is probably due to post-translational modifications. Mutations in cat2 and ntra ntrb lower the in vivo NAD-MDH activity We further evaluated the NAD-MDH activity in a cat2 loss-of-function mutant, in which the major catalase isoform in leaves is inactivated, retaining only 8% of extractible catalase activity (Mhamdi et al., 2010b) (Supplementary Fig. S8). The NAD-MDH activity was significantly decreased in the cat2 mutant when compared with wild-type plants (25–33%), indicating that the NAD-MDH activity is affected by increased in vivo H2O2 levels (Fig. 7B). As we showed that the reduction in cytMDH1Ox depends on cytosolic TRXh, we also measured the NAD-MDH activity in the TRX reductase double mutant ntra ntrb in which TRXhs are partially inactive (Reichheld et al., 2007). As previously observed (Daloso et al., 2015), we also found a significant decrease (17–20%) in this mutant. However, decreasing the NTR gene expression with RNAi in cat2 (cat2 ntra ntrb) had no additional impact on the NAD-MDH activity, suggesting that the residual NTR activity in the RNAi lines is efficient enough to reduce the oxidized MDH (Fig. 7B; Supplementary Figs S8, S9). We further tested whether the glutathione-reducing system could be part of the cytMDH redox regulation mechanism. Therefore, the NAD-MDH activity was measured in a cytosolic glutathione reductase mutant (gr1) with a more oxidized glutathione redox state (Supplementary Table S2). No decrease in NAD-MDH activity was observed, and in the gr1 cat2 double mutant, the decrease in NAD-MDH activity was similar to that observed in the cat2 mutant. Finally, to confirm that the decreased NAD-MDH activities in cat2 and the ntra ntrb double mutants are linked to TRX reduction, we performed NAD-MDH activity assays for the plant extracts in the presence of recombinant cytosolic NTRA and TRXh3 proteins (Fig. 7C). After the addition of the NTR/TRXh3 system in the wild-type plant extracts, the NAD-MDH activity remained the same. However, the activity was significantly increased in the plant extracts from cat2 and ntra ntrb mutants, confirming a TRX-dependent activity of the MDH pool. Discussion Plant NAD(P)-dependent MDHs are part of a multigenic family (Supplementary Fig. S1). The redox regulation of chloroplastic NADP-MDH has been extensively studied in the last decades. However, a redox regulation of other isoforms of NAD-MDH is unlikely to act in a similar way. Accordingly, the mitochondrial NAD-MDH isoforms that are involved in the tricarboxylic acid (TCA) cycle are not prone to redox regulation (Yoshida and Hisabori, 2016), though other enzymes of the TCA cycle have been shown to be redox regulated (Daloso et al., 2015). Our NAD-MDH activity assays performed on purified subcellular fractions also support redox-independent mitochondrial NAD-MDH activities. However, the H2O2 sensitivity of the chloroplastic NAD-MDH activity might indicate a redox dependence of chloroplastic NAD-MDH in vivo, which was not previously found in vitro (Berkemeyer et al., 1998). A potential redox regulation of the chloroplastic NAD-MDH activity will need to be further explored. In clear contrast to mitochondrial isoforms, different arguments are in favor of a redox regulation of the Arabidopsis cytMDH1 (Hara et al., 2006; Hara and Hisabori, 2013; this work): (i) cytMDH1 was identified as a sulfenylated protein under oxidative stress and during symbiosis (Waszczak et al., 2014; Akter et al., 2015), which has been validated in vitro in our study (Fig. 1; Supplementary Fig. S10); (ii) not only cysteine sulfenylation, but also cysteine overoxidation and methionine sulfoxidations occur under oxidizing conditions (Figs 3, 6, 7; Supplementary Table S2); (iii) modifications in enzymatic kinetics, secondary structure, and folding characteristics are observed under H2O2 treatment (Figs 3–5); and (iv) NAD-MDH activities are affected in H2O2-supplied purified cytosolic fractions, and in mutants accumulating high levels of H2O2 (cat2) or impaired in the TRX reduction systems (ntra ntrb) (Fig. 7). Therefore, our present work extends in vitro data demonstrating a redox regulation of cytMDH1, and we provide clear evidence that this regulation occurs in vivo. All eukaryotic cytosolic NAD-MDHs exist as homodimers (Berkemeyer et al., 1998; Birktoft et al., 1989), which was confirmed for the cytMDH1 isoform by SEC (Hara et al., 2006). Our structural data also confirm this observation (Fig. 6). Moreover, we show that a major conformational change occurs between cytMDH1Red and dimeric cytMDH1Ox. Whereas the cytMDH1Red dimer is mainly stabilized by electrostatic and hydrophobic interactions, as visualized by SEC and native gels (Supplementary Figs S2, S3), oxidation causes the formation of a disulfide-linked dimer (Figs 2, 3). It has been hypothesized that an intermolecular Cys330–Cys330 bond co-ordinates a tetrameric structure (Hara et al., 2006), but not supported by our SEC and native gel profiles, because no tetrameric structure was observed (Supplementary Figs S2, S3), possibly due to different oxidation conditions between the two studies: we used H2O2-induced oxidation, while CuCl2-induced oxidation was used in the previous study (Hara et al., 2006). The H2O2-mediated cytMDH1 oxidation inducing a conformational modification is reminiscent of the conformational change reported for oxidized chloroplastic NADP-MDH (Ruelland et al., 1998). The formation of the C-terminal intramolecular disulfide bond upon oxidation of this enzyme triggers the binding of the C-terminal domain to OAA. In contrast, cytMDH1 oxidation has a minor effect on the affinity for the substrate, even slightly decreasing the KM for OAA and NADH. Therefore, a similar redox regulatory mechanism between the two isoforms can be ruled out. However, the observed decrease in the turnover rate (kcat) of the cytMDH1Ox suggests that this modification of the conformation affects the catalytic activity of the enzyme. Surprisingly, the comparison of the crystal structures of reduced and oxidized cytMDH1 does not show extensive changes in the global conformational fold of both structures. However, we should note that the crystal structure of oxidized cytMDH1 is a crystal of cytMDH1Red incubated with H2O2. Therefore, the constraints of the crystal lattice preclude the possibility of large-scale conformational changes. As both cytMDH1Red and cytMDH1Ox form homodimers, it is difficult to separate cytMDH1Ox from the cytMDH1Red for cytMDH1Ox crystallization trials. Although Met97 at the substrate-binding site was sensitive to oxidation (Fig. 6), the enzyme kinetics show no major impact of oxidation on the affinity of cytMDH1 towards OAA and NADH (Table 1). The activity of alkylated cytMDH1 is not sensitive to H2O2 oxidation, whereas cytMDH1Red is very sensitive to diamide (Supplementary Fig. S6), a reagent that introduces disulfide bonds. Altogether, our results indicate that the decrease in activity is not due to methionine oxidation, but rather to disulfide bond formation. Among the cysteine residues of cytMDH1 undergoing oxidative modifications, Cys79 is prone to overoxidation, while Cys330 tends preferentially to form a Cys330–Cys330 disulfide bond (Supplementary Table S2), suggesting that disulfide formation protects Cys330 from overoxidation (Fig. 7). We also demonstrated that sulfenylation of Cys does not by itself impact the activity of cytMDH1 (Supplementary Fig. S10). Therefore, we conclude that in contrast to another sulfenylated residue, Cys79, the sulfenylated Cys330 forms a disulfide bond with free Cys330 thiol (Fig. 8), which causes the reversible conformational change and a decrease of the activity. Fig. 8. View largeDownload slide How cytMDH1 copes with H2O2 oxidation. Oxidation of homodimeric cytMDH1 leads to sulfenylation of Cys330 and the formation of an intermolecular Cys330–Cys330 disulfide. Therefore, a major conformational change is required that leads to the formation of a thermodynamically less stable homodimer with an altered kinetic behavior but with similar kcat/KM values. The S–S linked dimer protects the Cys330 from irreversible overoxidation and becomes a substrate for the TRX pathway. Methionine sulfoxidation and sulfenylation of cysteines other than Cys330 do not affect the kinetic behavior. Fig. 8. View largeDownload slide How cytMDH1 copes with H2O2 oxidation. Oxidation of homodimeric cytMDH1 leads to sulfenylation of Cys330 and the formation of an intermolecular Cys330–Cys330 disulfide. Therefore, a major conformational change is required that leads to the formation of a thermodynamically less stable homodimer with an altered kinetic behavior but with similar kcat/KM values. The S–S linked dimer protects the Cys330 from irreversible overoxidation and becomes a substrate for the TRX pathway. Methionine sulfoxidation and sulfenylation of cysteines other than Cys330 do not affect the kinetic behavior. Intriguingly, our crystal structure data show that Cys330 residues from two interacting subunits are located at opposite sides in the cytMDH1 dimeric molecule (Fig. 6). A hypothesis is that upon H2O2 oxidation, a major conformational change brings Cys330 of both subunits in close proximity to permit Cys330–Cys330 disulfide bond formation. As mentioned above, our oxidized crystal structure does not allow the visualization of such conformational changes and we would need to obtain a crystal structure after oxidation of the protein in solution, but have failed as yet. Another more likely hypothesis is that the disulfide bond is formed between Cys330 of two individual non-covalent dimers. As a tetrameric cytMDH1 state was not detected, we suggest that Cys330–Cys330 formation between dimers leads to structural changes that disrupt the native non-covalent dimer, leading to a population of covalent homodimers (Fig. 8). This scenario is supported by the CD analyses demonstrating that cytMDH1Ox is significantly more disordered (Fig. 4A) than the cytMDH1Red, comprising a larger number of possible states prior to thermal denaturation (Fig. 4B; Table 2), and that cytMDH1Red is in a more restricted number of possible conformations. Additionally, ΔΔG and Tm values also indicate that cytMDH1Red is significantly more stable than cytMDH1Ox (Table 2). CD and steady-state kinetic data reinforce the final conclusion that oxidation of cytMDH1 induces a conformational disorder and increases instability of the protein, possibly caused both by Cys330–Cys330 disulfide formation and overoxidation. Furthermore, the disulfide bond dimer can be efficiently reactivated by TRXh proteins, providing full reversibility to the system and further demonstrating the role of the disulfide bond in protecting the enzyme activity (Fig. 8). Interestingly, the Cys330 is also conserved in the other two cytMDH isoforms encoded by the Arabidopsis genome (Supplementary Fig. S1) as well as in cytMDH orthologs of other plants, suggesting that this mechanism is conserved in the green lineage. It is also likely that the decrease in NAD-MDH activities observed in the cat2 mutant is reporting the inactivation of all three cytosolic isoforms. Altered activity is also found in the ntra ntrb mutants, where the TRX system is partially inactivated (Reichheld et al., 2007). Decreased NAD-MDH activity is probably due to the lower capacity of the ntra ntrb mutant to regenerate oxidized MDH in vivo. The Arabidopsis genome harbors multiple cytosolic TRXs (Meyer et al., 2012). We have tested five isoforms of TRXh with cytMDH1 as substrate, and all of them are able to restore the cytMDH1 activity after oxidation. It is likely that these TRX isoforms are redundant, which also explain the decreased NAD-MDH activity that is only observed in the ntra ntrb mutant, and not with several TRXh single mutants (trxh1, trxh2, trxh3, trxh4, and trxh5) tested so far (data not shown). Interestingly, our yeast two-hybrid test showed that cytosolic TRXhs interact specifically with cytMDH1. As Cys330 is only found in cytosolic MDHs, it is tempting to speculate that TRXh/cytMDH forms a mixed disulfide between the catalytic Cys of TRXh and Cys330 of cytMDH1, although other structural features could mediate this specific interaction. Cytosolic MDH plays a key role in equilibrating reducing equivalents between the organelles as well as in providing an important pool of OAA for further metabolic activities. Therefore, its activity needs to be preserved under stress conditions leading to reduced availability of its substrates. Our data clearly suggest that the redox regulation of cytMDH is playing an efficient protective role to avoid irreversible overoxidation of the enzyme. Interestingly, our kinetic data show that the cytMDH1Ox catalytic efficiency (kcat/KM) is not profoundly perturbed compared with that of cytMDH1Red. While the turnover rate (kcat) is 2-fold decreased, the affinity (KM) for the substrates is slightly increased, suggesting that cytMDH1Ox is adapting its activity to low availability of the substrate. Such a buffering mechanism as well as the reversibility of the regulation by the TRX system probably acts as a safety mechanism to minimize oxidative inactivation of cytMDH1 under stress conditions. Such a fine-tuned redox regulation is also consistent with the fact that the cytMDH1 protein is abundantly maintained under oxidative stress, in order to adapt the cell metabolism under changing environmental conditions. Supplementary data Supplementary data are available at JXB online. Fig. S1. Alignment of Arabidopsis thaliana MDHs. Fig. S2. Dimeric forms of reduced and oxidized cytMDH1 identified by size exclusion chromatography. Fig. S3. Reduced and oxidized cytMDH1 in native gel. Fig. S4. Cytosolic TRXhs interact with cytMDH1. Fig. S5. Plastidial TRXf1 and TRXf2 interact with members of the MDH family. Fig. S6. Redox sensitivity of cytMDH1 is regulated by TRXh on the disulfide bridge. Fig. S7. Transcriptional analysis of MDHs under stress conditions. Fig. S8. Catalase activity in mutants. Fig. S9. NTR accumulation is decreased in RNAi lines. Fig. S10. Sulfenylation does not affect the activity of cytMDH1. Table S1. Mass spectral summary of post-translational oxidation modifications on cysteines and methionines. Table S2. Glutathione levels in the gr1 mutant. Table S3. Crystallographic data collection and refinement statistics Abbreviations: Abbreviations: CD circular dichroism chlMDH chloroplastic NAD-dependent MDH cytMDH cytosolic NAD-dependent MDH IAM iodoacetamide MDH malate dehydrogenase mitMDH mitochondrial NAD-dependent MDH NADP-MDH NADP-dependent MDH NTRA/B NADPH-dependent thioredoxin reductase A/B OAA oxaloacetate perMDH peroxisomal NAD-dependent MDH SEC size exclusion chromatography –SOH sulfenic acid TCA tricarboxylic acid TRX thioredoxin. Acknowledgements This work was supported by grants from (i) the Centre National de la Recherche Scientifique and the Agence Nationale de la Recherche (ANR-Blanc Cynthiol 12-BSV6-0011); (ii) VIB (to JM); (iii) the SPR34 project of the Vrije Universiteit Brussel (VUB) (to JM); (iv) Research Foundation Flanders (grant no. G0D7914N) (to JM and FVB); and (v) Flanders Hercules Foundation (grant no. HERC16) for the purification platform (to JM). We thank the SOLEIL beamline scientists at Proxima 2a for their technical assistance in data collection; Inge Van Molle from VIB-VUB Center for Structure Biology for help with structural data collection; Barbara De Smet and Valerie Van Ruyskensvelde from VIB-UGent Center for Plant Systems Biology for the genotype and phenotype screening; Jing Yang from Beijing Proteome Research Center (BPRC) for the mass spectrometry support; and Patrick Willems from VIB-UGent Center for Plant Systems Biology for bioinformatics analysis. References Adams PD , Afonine PV , Bunkóczi G , et al. 2010 . PHENIX: a comprehensive Python-based system for macromolecular structure solution . Acta Crystallographica. Section D, Biological Crystallography 66 , 213 – 221 . Google Scholar CrossRef Search ADS PubMed Adams PD , Grosse-Kunstleve RW , Hung LW , et al. 2002 . PHENIX: building new software for automated crystallographic structure determination . Acta Crystallographica. Section D, Biological Crystallography 58 , 1948 – 1954 . Google Scholar CrossRef Search ADS PubMed Afonine PV , Grosse-Kunstleve RW , Echols N , et al. 2012 . Towards automated crystallographic structure refinement with phenix.refine . Acta Crystallographica. Section D, Biological Crystallography 68 , 352 – 367 . Google Scholar CrossRef Search ADS PubMed Akter S , Huang J , Bodra N , et al. 2015 . DYn-2 based identification of Arabidopsis sulfenomes . Molecular and Cellular Proteomics 14 , 1183 – 1200 . Google Scholar CrossRef Search ADS PubMed Berkemeyer M , Scheibe R , Ocheretina O . 1998 . A novel, non-redox-regulated NAD-dependent malate dehydrogenase from chloroplasts of Arabidopsis thaliana L . Journal of Biological Chemistry 273 , 27927 – 27933 . Google Scholar CrossRef Search ADS PubMed Birktoft JJ , Rhodes G , Banaszak LJ . 1989 . Refined crystal structure of cytoplasmic malate dehydrogenase at 2.5-Å resolution . Biochemistry 28 , 6065 – 6081 . Google Scholar CrossRef Search ADS PubMed Carr PD , Verger D , Ashton AR , Ollis DL . 1999 . Chloroplast NADP-malate dehydrogenase: structural basis of light-dependent regulation of activity by thiol oxidation and reduction . Structure 7 , 461 – 475 . Google Scholar CrossRef Search ADS PubMed Chen VB , Arendall WB 3rd , Headd JJ , Keedy DA , Immormino RM , Kapral GJ , Murray LW , Richardson JS , Richardson DC . 2010 . MolProbity: all-atom structure validation for macromolecular crystallography . Acta Crystallographica. Section D, Biological Crystallography 66 , 12 – 21 . Google Scholar CrossRef Search ADS PubMed Choi S , Jeong J , Na S , Lee HS , Kim HY , Lee KJ , Paek E . 2010 . New algorithm for the identification of intact disulfide linkages based on fragmentation characteristics in tandem mass spectra . Journal of Proteome Research 9 , 626 – 635 . Google Scholar CrossRef Search ADS PubMed Collaborative Computational Project, Number 4 . 1994 . The CCP4 suite: programs for protein crystallography . Acta Crystallographica. Section D, Biological Crystallography 50 , 760 – 763 . CrossRef Search ADS PubMed Daloso DM , Müller K , Obata T , et al. 2015 . Thioredoxin, a master regulator of the tricarboxylic acid cycle in plant mitochondria . Proceedings of the National Academy of Sciences, USA 112 , E1392 – E1400 . Google Scholar CrossRef Search ADS Emsley P , Cowtan K . 2004 . Coot: model-building tools for molecular graphics . Acta Crystallographica. Section D, Biological Crystallography 60 , 2126 – 2132 . Google Scholar CrossRef Search ADS PubMed Evans PR , Murshudov GN . 2013 . How good are my data and what is the resolution ? Acta Crystallographica. Section D, Biological Crystallography 69 , 1204 – 1214 . Google Scholar CrossRef Search ADS PubMed Faske M , Holtgrefe S , Ocheretina O , Meister M , Backhausen JE , Scheibe R . 1995 . Redox equilibria between the regulatory thiols of light/dark-modulated chloroplast enzymes and dithiothreitol: fine-tuning by metabolites . Biochimica et Biophysica Acta 1247 , 135 – 142 . Google Scholar CrossRef Search ADS PubMed Gelhaye E , Rouhier N , Jacquot J-P . 2003 . Evidence for a subgroup of thioredoxin h that requires GSH/Grx for its reduction . FEBS Letters 555 , 443 – 448 . Google Scholar CrossRef Search ADS PubMed Hara S , Hisabori T . 2013 . Kinetic analysis of the interactions between plant thioredoxin and target proteins . Frontiers in Plant Science 4 , 508 . Google Scholar CrossRef Search ADS PubMed Hara S , Motohashi K , Arisaka F , Romano PG , Hosoya-Matsuda N , Kikuchi N , Fusada N , Hisabori T . 2006 . Thioredoxin-h1 reduces and reactivates the oxidized cytosolic malate dehydrogenase dimer in higher plants . Journal of Biological Chemistry 281 , 32065 – 32071 . Google Scholar CrossRef Search ADS PubMed Hossain MS , Dietz KJ . 2016 . Tuning of redox regulatory mechanisms, reactive oxygen species and redox homeostasis under salinity stress . Frontiers in Plant Science 7 , 548 . Google Scholar PubMed Issakidis E , Miginiac-Maslow M , Decottignies P , Jacquot J-P , Crétin C , Gadal P . 1992 . Site-directed mutagenesis reveals the involvement of an additional thioredoxin-dependent regulatory site in the activation of recombinant sorghum leaf NADP-malate dehydrogenase . Journal of Biological Chemistry 267 , 21577 – 21583 . Google Scholar PubMed Issakidis E , Saarinen M , Decottignies P , Jacquot J-P , Crétin C , Gadal P , Miginiac-Maslow M . 1994 . Identification and characterization of the second regulatory disulfide bridge of recombinant sorghum leaf NADP-malate dehydrogenase . Journal of Biological Chemistry 269 , 3511 – 3517 . Google Scholar PubMed Jacques S , Ghesquière B , De Bock PJ , Demol H , Wahni K , Willems P , Messens J , Van Breusegem F , Gevaert K . 2015 . Protein methionine sulfoxide dynamics in Arabidopsis thaliana under oxidative stress . Molecular and Cellular Proteomics 14 , 1217 – 1229 . Google Scholar CrossRef Search ADS PubMed Jacquot J-P , Lancelin JM , Meyer Y . 1997 . Thioredoxins: structure and function in plant cells . New Phytologist 136 , 543 – 570 . Google Scholar CrossRef Search ADS Johansson K , Ramaswamy S , Saarinen M , Lemaire-Chamley M , Issakidis-Bourguet E , Miginiac-Maslow M , Eklund H . 1999 . Structural basis for light activation of a chloroplast enzyme: the structure of sorghum NADP-malate dehydrogenase in its oxidized form . Biochemistry 38 , 4319 – 4326 . Google Scholar CrossRef Search ADS PubMed Kabsch W . 2010a . Integration, scaling, space-group assignment and post-refinement . Acta Crystallographica. Section D, Biological Crystallography 66 , 133 – 144 . Google Scholar CrossRef Search ADS Kabsch W . 2010b . XDS . Acta Crystallographica. Section D, Biological Crystallography 66 , 125 – 132 . Google Scholar CrossRef Search ADS Krissinel E , Henrick K . 2007 . Inference of macromolecular assemblies from crystalline state . Journal of Molecular Biology 372 , 774 – 797 . Google Scholar CrossRef Search ADS PubMed Laloi C , Rayapuram N , Chartier Y , Grienenberger JM , Bonnard G , Meyer Y . 2001 . Identification and characterization of a mitochondrial thioredoxin system in plants . Proceedings of the National Academy of Sciences, USA 98 , 14144 – 14149 . Google Scholar CrossRef Search ADS Lemaire SD , Michelet L , Zaffagnini M , Massot V , Issakidis-Bourguet E . 2007 . Thioredoxins in chloroplasts . Current Genetics 51 , 343 – 365 . Google Scholar CrossRef Search ADS PubMed Lindermayr C , Saalbach G , Durner J . 2005 . Proteomic identification of S-nitrosylated proteins in Arabidopsis . Plant Physiology 137 , 921 – 930 . Google Scholar CrossRef Search ADS PubMed McCoy AJ , Grosse-Kunstleve RW , Adams PD , Winn MD , Storoni LC , Read RJ . 2007 . Phaser crystallographic software . Journal of Applied Crystallography 40 , 658 – 674 . Google Scholar CrossRef Search ADS PubMed Meyer Y , Buchanan BB , Vignols F , Reichheld J-P . 2009 . Thioredoxins and glutaredoxins: unifying elements in redox biology . Annual Review of Genetics 43 , 335 – 367 . Google Scholar CrossRef Search ADS PubMed Meyer Y , Belin C , Delorme-Hinoux V , Reichheld J-P , Riondet C . 2012 . Thioredoxin and glutaredoxin systems in plants: molecular mechanisms, crosstalks, and functional significance . Antioxidants and Redox Signaling 17 , 1124 – 1160 . Google Scholar CrossRef Search ADS PubMed Mhamdi A , Hager J , Chaouch S , et al. 2010a . Arabidopsis GLUTATHIONE REDUCTASE1 plays a crucial role in leaf responses to intracellular hydrogen peroxide and in ensuring appropriate gene expression through both salicylic acid and jasmonic acid signaling pathways . Plant Physiology 153 , 1144 – 1160 . Google Scholar CrossRef Search ADS Mhamdi A , Queval G , Chaouch S , Vanderauwera S , Van Breusegem F , Noctor G . 2010b . Catalase function in plants: a focus on Arabidopsis mutants as stress-mimic models . Journal of Experimental Botany 61 , 4197 – 4220 . Google Scholar CrossRef Search ADS Murphy KP , Freire E . 1992 . Thermodynamics of structural stability and cooperative folding behavior in proteins . Advances in Protein Chemistry 43 , 313 – 361 . Google Scholar CrossRef Search ADS PubMed Murshudov GN , Skubák P , Lebedev AA , Pannu NS , Steiner RA , Nicholls RA , Winn MD , Long F , Vagin AA . 2011 . REFMAC5 for the refinement of macromolecular crystal structures . Acta Crystallographica. Section D, Biological Crystallography 67 , 355 – 367 . Google Scholar CrossRef Search ADS PubMed O’Brien R , Haq I . 2004 . Applications of biocalorimetry: binding, stability and enzyme kinetics . In: Ladbury JE , Doyle ML , eds. Biocalorimetry 2: applications of calorimetry in the biological sciences . Chichester, UK : John Wiley & Sons, Ltd , 1 – 34 . Google Scholar CrossRef Search ADS Oger E , Marino D , Guigonis JM , Pauly N , Puppo A . 2012 . Sulfenylated proteins in the Medicago truncatula–Sinorhizobium meliloti symbiosis . Journal of Proteomics 75 , 4102 – 4113 . Google Scholar CrossRef Search ADS PubMed Pyr Dit Ruys S , Wang X , Smith EM , Herinckx G , Hussain N , Rider MH , Vertommen D , Proud CG . 2012 . Identification of autophosphorylation sites in eukaryotic elongation factor-2 kinase . Biochemical Journal 442 , 681 – 692 . Google Scholar CrossRef Search ADS PubMed Queval G , Issakidis-Bourguet E , Hoeberichts FA , Vandorpe M , Gakiere B , Vanacker H , Miginiac-Maslow M , Van Breusegem F , Noctor G . 2007 . Conditional oxidative stress responses in the Arabidopsis photorespiratory mutant cat2 demonstrate that redox state is a key modulator of daylength-dependent gene expression, and define photoperiod as a crucial factor in the regulation of H2O2-induced cell death . The Plant Journal 52 , 640 – 657 . Google Scholar CrossRef Search ADS PubMed Reichheld J-P , Khafif M , Riondet C , Droux M , Bonnard G , Meyer Y . 2007 . Inactivation of thioredoxin reductases reveals a complex interplay between thioredoxin and glutathione pathways in Arabidopsis development . The Plant Cell 19 , 1851 – 1865 . Google Scholar CrossRef Search ADS PubMed Reichheld J-P , Meyer E , Khafif M , Bonnard G , Meyer Y . 2005 . AtNTRB is the major mitochondrial thioredoxin reductase in Arabidopsis thaliana . FEBS Letters 579 , 337 – 342 . Google Scholar CrossRef Search ADS PubMed Ruelland E , Johansson K , Decottignies P , Djukic N , Miginiac-Maslow M . 1998 . The autoinhibition of sorghum NADP malate dehydrogenase is mediated by a C-terminal negative charge . Journal of Biological Chemistry 273 , 33482 – 33488 . Google Scholar CrossRef Search ADS PubMed Scheibe R . 2004 . Malate valves to balance cellular energy supply . Physiologia Plantarum 120 , 21 – 26 . Google Scholar CrossRef Search ADS PubMed Scheibe R , Dietz KJ . 2012 . Reduction–oxidation network for flexible adjustment of cellular metabolism in photoautotrophic cells . Plant, Cell and Environment 35 , 202 – 216 . Google Scholar CrossRef Search ADS Selinski J , König N , Wellmeyer B , Hanke GT , Linke V , Neuhaus HE , Scheibe R . 2014 . The plastid-localized NAD-dependent malate dehydrogenase is crucial for energy homeostasis in developing Arabidopsis thaliana seeds . Molecular Plant 7 , 170 – 186 . Google Scholar CrossRef Search ADS PubMed Suh JJ , Savisky RM . 2011 . Intrinsic fluorescence studies of metal ion binding effects on the thermodynamic stability of human α-lactalbumin . International Journal of Biophysics 1 , 1 – 6 . Google Scholar CrossRef Search ADS Vignols F , Bréhélin C , Surdin-Kerjan Y , Thomas D , Meyer Y . 2005 . A yeast two-hybrid knockout strain to explore thioredoxin-interacting proteins in vivo . Proceedings of the National Academy of Sciences, USA 102 , 16729 – 16734 . Google Scholar CrossRef Search ADS Waszczak C , Akter S , Eeckhout D , et al. 2014 . Sulfenome mining in Arabidopsis thaliana . Proceedings of the National Academy of Sciences, USA 111 , 11545 – 11550 . Google Scholar CrossRef Search ADS Waszczak C , Kerchev PI , Mühlenbock P , et al. 2016 . SHORT-ROOT deficiency alleviates the cell death phenotype of the Arabidopsiscatalase2 mutant under photorespiration-promoting conditions . The Plant Cell 28 , 1844 – 1859 . Google Scholar CrossRef Search ADS PubMed Willems P , Mhamdi A , Simon S , Storme V , Kerchev P , Noctor G , Gevaert K , Van Breusegem F . 2016 . The ROS wheel: refining ROS transcriptional footprints . Plant Physiology 171 , 1720 – 33 . Google Scholar CrossRef Search ADS PubMed Yamazaki D , Motohashi K , Kasama T , Hara Y , Hisabori T . 2004 . Target proteins of the cytosolic thioredoxins in Arabidopsis thaliana . Plant and Cell Physiology 45 , 18 – 27 . Google Scholar CrossRef Search ADS PubMed Yoshida K , Hisabori T . 2016 . Adenine nucleotide-dependent and redox-independent control of mitochondrial malate dehydrogenase activity in Arabidopsis thaliana . Biochimica et Biophysica Acta 1857 , 810 – 818 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2017. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Experimental Botany Oxford University Press

Loading next page...
 
/lp/ou_press/self-protection-of-cytosolic-malate-dehydrogenase-against-oxidative-yIgtpeMeq0
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
0022-0957
eISSN
1460-2431
D.O.I.
10.1093/jxb/erx396
Publisher site
See Article on Publisher Site

Abstract

Abstract Plant malate dehydrogenase (MDH) isoforms are found in different cell compartments and function in key metabolic pathways. It is well known that the chloroplastic NADP-dependent MDH activities are strictly redox regulated and controlled by light. However, redox dependence of other NAD-dependent MDH isoforms have been less studied. Here, we show by in vitro biochemical characterization that the major cytosolic MDH isoform (cytMDH1) is sensitive to H2O2 through sulfur oxidation of cysteines and methionines. CytMDH1 oxidation affects the kinetics, secondary structure, and thermodynamic stability of cytMDH1. Moreover, MS analyses and comparison of crystal structures between the reduced and H2O2-treated cytMDH1 further show that thioredoxin-reversible homodimerization of cytMDH1 through Cys330 disulfide formation protects the protein from overoxidation. Consistently, we found that cytosolic thioredoxins interact specifically with cytMDH in a yeast two-hybrid system. Importantly, we also show that cytosolic and chloroplastic, but not mitochondrial NAD-MDH activities are sensitive to H2O2 stress in Arabidopsis. NAD-MDH activities decreased both in a catalase2 mutant and in an NADP-thioredoxin reductase mutant, emphasizing the importance of the thioredoxin-reducing system to protect MDH from oxidation in vivo. We propose that the redox switch of the MDH activity contributes to adapt the cell metabolism to environmental constraints. CytMDH1, dimerization, H2O2-triggered oxdation, overoxidation, sulfenic acid, thioredoxin Introduction NAD(P)-dependent malate dehydrogenases (MDHs, l-malate:NAD oxidoreductases; EC 1.1.1.37) reversibly catalyze the oxidation (dehydrogenation) of l-malate to oxaloacetate (OAA) while reducing NAD(P)+ to NAD(P)H. In plants, MDH isoforms play a role in central metabolism and redox homeostasis between organellar compartments (Faske et al., 1995; Scheibe, 2004; Scheibe and Dietz, 2012; Selinski et al., 2014). As in other plants, Arabidopsis MDH isoforms can be classified into different groups based on their localizations: three cytosolic NAD-dependent MDHs, cytMDH1, cytMDH2, and cytMDH3; two mitochondrial NAD-dependent MDHs, mitMDH1 and mitMDH2; one chloroplastic NAD-dependent MDH, chlMDH; two peroxisomal NAD-dependent MDHs, perMDH1 and perMDH2; and one NADP-dependent MDH (NADP-MDH) (see Supplementary Fig. S1 at JXB online). NADP-MDH, localized in the chloroplast, is redox regulated through two sets of redox-sensitive cysteine pairs, located at the N- and C-terminus, which form two intramolecular disulfides (Cys77–Cys82 and Cys418–Cys430) that inhibit the enzymatic activity (Issakidis et al., 1992, 1994; Faske et al., 1995; Ruelland et al., 1998; Johansson et al., 1999; Carr et al., 1999). The inactive disulfide-containing state of NADP-MDH can be reduced to its active form through thiol/disulfide exchange with specific classes of disulfide reductases, namely thioredoxins (TRXm and TRXf), which themselves are reduced by a light-dependent ferredoxin-TRX reductase (reviewed by Jacquot et al., 1997; Lemaire et al., 2007; Meyer et al., 2009). Active NADP-MDH oxidizes NADPH to convert OAA into malate. Malate is then transported to the cytosol, where the cytMDH converts it back to OAA, while reducing NAD+ to NADH in the process. This so-called ‘malate valve’ is the first step in the redistribution of chloroplastic reducing equivalents to other parts of the plant cell (Scheibe, 2004; Hossain and Dietz, 2016). Two decades ago, chlMDH was reported as a non-redox-regulated protein (Berkemeyer et al., 1998), and recently mitMDH was shown to be redox independent (Yoshida and Hisabori, 2016). As the redox-sensitive cysteine pairs of the NADP-MDH are absent in cytMDH isoforms, it had not been considered as a redox-sensitive enzyme. However, cytMDH was captured on aTRX affinity chromatography column (Yamazaki et al., 2004), and found to be S-nitrosothiolated upon gaseous nitric oxide treatment in Arabidopsis (Lindermayr et al., 2005). More recently, cytMDH was detected in a sulfenylated state (–SOH) in Medicago truncatula (Oger et al., 2012), and in Arabidopsis cell suspensions (Waszczak et al., 2014; Akter et al., 2015). In an in vitro study, CuCl2 treatment was shown to inactivate cytMDH and a disulfide-linked homodimer was formed. Moreover, this disulfide bond was efficiently reduced by a cytosolic thioredoxin h1 (TRXh1) (Hara et al., 2006). In Arabidopsis, cytosolic TRXs are encoded by multigenic families for which TRXh members are the major isoforms (reviewed by Meyer et al., 2009). Most of these members are reduced by NADPH-dependent TRX reductases (NTRA and NTRB) that are dually located in the cytosol and mitochondria (Laloi et al., 2001; Reichheld et al., 2005). Some discrete TRXh members are alternatively reduced by the glutathione reductase (GR)/glutaredoxin (GRX) pathway, another major cytosolic thiol reduction pathway dependent on NADPH (Gelhaye et al., 2003). Here, we re-examined the redox dependence of cytMDH by combining in vitro biochemical, structural, and biophysical approaches. We reveal that oxidation triggers overoxidation of several cysteine residues through sulfenylation. Moreover, we show that under oxidative stress, cytMDH switches from a non-covalent dimer to a covalent disulfide-linked dimer, probably affecting the structure, the kinetics, and the stability of the enzyme. We indicate that the dimerization via Cys330–Cys330 disulfide acts as a redox switch, protecting cytMDH1 from overoxidation, and can be reduced by TRXs. Importantly, we show that such regulation also occurs in planta and might be conserved in other cytMDH isoforms. Therefore, based on both in vitro and in planta data, we propose a mechanism of how cytosolic NAD-MDH is redox regulated. Materials and methods Plant material and growth conditions Arabidopsis thaliana plants were grown in soil in a controlled growth chamber (180 µE m–2 s–1, 16 h day/8 h night, 22 °C 55% relative humidity day, 20 °C 60% relative humidity night) for up to 3 weeks. Plant mutant lines ntra ntrb, cat2, gr1, and gr1 cat2 were previously described (Queval et al., 2007; Reichheld et al., 2007; Mhamdi et al., 2010a). In vitro protein-based complementation and TRX activity assays For in vitro protein-based complementation assays, 5 µg of protein extracts (200 ng µl–1) were incubated with 1 mM NADPH, 4.59 μM TRXh3, and 3.12 µMNTRA in 25 µl 50 mM Tris–HCl (pH 7.5) for 2 h on ice. This reaction mixture was diluted 40 times in the same buffer, and the NAD-MDH activity assay was performed as described above. Yeast two-hybrid assay For yeast expression, full-length TRXh and MDH cDNA were respectively cloned in the pDEST32 and pDEST22 plasmids using the Gateway cloning system. Yeast two-hybrid assays were performed according to the method described previously (Vignols et al., 2005). Briefly, the yeast strain CY306 (a Δtrx1Δtrx2 mutant) was co-transformed with both pDEST22 and pDEST32 constructs and plated on YNB medium with the required amino acids and bases (His, Ura, Lys, Ade, and Met). Transformed colonies were subcultured and diluted using dilution series (diluted at 5 × 10–2, 5 × 10–3, and 5 × 10–4 at an optical density of 600 nm) prior to dotting on YNB –His –Trp –Leu in Petri dishes containing 20 mM 3-amino-1,2,4-triazole (3AT). Three independent transformations per binary assay were performed, of which 2–4 colonies were assayed. Images were taken 4 d post-dotting. Cloning, expression, and purification of recombinant cytMDH1 The cytMDH1-coding sequence (At1g04410), with attL and PreScission (EVLFQ/GP) sites at the N-terminal end, was synthesized by the Gen9 company. The gene fragment was inserted into the pDEST17 vector (Life Technologies). The constructs were transferred into Escherichia coli C41(DE3) strain. The expression and purification of recombinant His-tagged cytMDH1 was done as previously described (Hara et al., 2006). The purified His-tagged cytMDH1 was incubated with PreScission HRV3C enzyme (20:1, w/w) at 4 °C overnight to remove the His tag. The cleaved protein sample was loaded onto an Ni2+-Sepharose column equilibrated with 50 mM Tris–HCl, pH 7.5, 200 mM NaCl. The flow-through was collected and evaluated on an SDS–polyacrylamide gel as pure cleaved recombinant protein. Purified protein samples were pooled and flash-frozen in liquid nitrogen and then stored at –80 °C. Protein extracts from Arabidopsis plants and cytMDH1 enzymatic assay Purified cytosolic, mitochondrial, and chloroplastic protein extracts were prepared as described before (Daloso et al., 2015). NAD-MDH activity was measured from 3–5 µg of protein extracts at 25 °C in a reaction medium containing 50 mM Tris–HCl (pH 7.5), 250 µM NADH, and 2.5 mM OAA. Catalase activity was measured for 20 µg of protein extracts at 25 °C in a reaction medium containing 50 mM potassium phosphate buffer (pH 7.0) and 10 mM H2O2. The activity was monitored as a change in absorbance at 240 nm due to breakdown of H2O2. Purified recombinant cytMDH1 protein was incubated with 20 mM DTT at room temperature for 1 h to obtain reduced cytMDH1 (cytMDH1Red). cytMDH1Red (8 μM) was incubated with 10 mM H2O2 at 30 °C for 45 min to obtain oxidized cytMDH1 (cytMDH1Ox). For oxidation by diamide, 8 µM cytMDH1Red was incubated with 40 µM or 1 mM diamide at 30 °C for 45 min. For oxidation of the iodoacetamide (IAM)-alkylated cytMDH1, 20 µM cytMDH1Red was incubated with 10 mM IAM at room temperature for 20 min. Then 8 μM cytMDH1IAM was incubated with 10 mM H2O2 at 30 °C for 45 min. After each treatment, the protein samples were passed through a micro bio-spin column (Bio-Rad) to remove chemicals. The concentration of modified cytMDH1 was measured by absorbance at 280 nm. For cytMDH1 enzymatic assay, the reaction was performed in a 96-well plate at 30 °C with a final volume of 200 µL (ThermoFisher Scientific), containing 50 mM Tris–HCl pH 7.5, 2.5 mM OAA, 0.25 mM NADH, and 2.5 nM cytMDH1. The decrease in NADH absorbance at 340 nm was monitored using a SpectraMax 340PC spectrophotometer (Molecular Devices). The molar extinction coefficient for NADH of 6220 M–1 cm–1 was used for the calculation. To obtain the KM for OAA or NADH, progress curves were recorded using varying concentrations of OAA (0–2.5 mM) or NADH (0.0125–0.25 mM). The initial velocity (vi) for each substrate concentration was measured, and the vi/E0 values were plotted and fitted with the Michaelis–Menten equation to obtain the kinetic parameters. Three independent replicates of vi were measured for each substrate concentration. For the time course experiment with H2O2 incubation, 4.5 μM cytMDH1Red was incubated with 50 mM Tris–HCl (pH 7.5)/10 mM DTT/10 mM H2O2 at 30 °C for 2 h. CytMDH1 was diluted to 2.5 nM to measure its activity. For MDH reactivation by TRXh, cytMDH1Ox was incubated with the indicated concentrations of recombinant TRXh in the presence of NADPH (0.125 mM) and NTRA (0.5 µM). After 30 min incubation, the mix was diluted 40-fold and used for the NAD-MDH activity assay, as described above. In vitro sulfenylation Dimedone (500 mM; Sigma-Aldrich) was prepared freshly in DMSO. Then 8 µM cytMDH1Red was incubated with 2 mM dimedone in the presence of 10 mM DTT or 10 mM H2O2 at 30 °C for 45 min. The excess was removed on an equilibrated Micro Bio-Spin® Column Bio-Gel® P-6. The eluted samples were treated with 2 mM IAM at room temperature for 20 min. The samples were run on an SDS–polyacrylamide gel and the protein was transferred onto a polyvinylidene difluoride (PVDF) membrane. The western blot was done with 1:10 000 dilution rabbit anti-Cys-SOH antibody (Millipore); the result was visualized using Western Lightning Plus-ECL, Enhanced Chemiluminescence Substrate (PerkinElmer). LC–MS/MS The dimedone-treated cytMDH1Red and cytMDH1Ox samples were examined by SDS–PAGE and the bands were excised after Coomassie Brilliant Blue (CBB) staining. After destaining, the protein samples were digested by trypsin and subjected to LC-MS/MS analysis. The peptides were analyzed by LC-MS/MS as described (Pyr Dit Ruys et al., 2012). The mass spectrometer was operated in the data-dependent mode and switched automatically between MS, Zoom Scan for charge state determination and MS/MS for the most abundant ion. Each MS scan was followed by a maximum of five MS/MS scans using a collision energy of 30%. Dynamic exclusion was enabled to allow analysis of co-eluting peptides. The following parameters were used: trypsin was selected with cleavage only after lysine and arginine; the number of internal cleavage sites was set to 1; the mass tolerance for precursor and fragment ions was 1.1 Da and 1.0 Da, respectively; and the considered dynamic modifications on cysteine residues were +138.0 Da for sulfenic-dimedone, +32.0 Da for sulfinic, +48.0 Da for sulfonic, and +57.0 Da for carbamidomethyl modifications. For peptide identification, peak lists were generated using the application spectrum selector in the Proteome Discoverer 1.4 package. The resulting peak lists were searched using Sequest against an Arabidopsis protein database. Peptide matches were filtered using Percolator within Proteome Discoverer and manually validated. The mixed disulfide peptide between C330 and C330 of MDH was identified by the use of DBond software (Choi et al., 2010). All the annotated MS/MS spectra of modified peptides were manually evaluated. Melting and overall secondary structure assessed by circular dichroism (CD) The far CD spectra for cytMDH1Red and cytMDH1Ox were recorded in a JASCO J-715 spectropolarimeter equipped with a Peltier temperature control unit. The measurements were carried out in a 0.1 cm path length cuvette with an average of three scans for each spectrum, a 100 nm min–1 scan speed, band width of 2 nm, data pitch of 4 nm s–1, and the recorded range was from 190 nm to 260 nm. The buffer used was 50 mM ammonium bicarbonate, pH 7.4, and the protein concentration was in the range of 3–4.5 µM. Thermal unfolding was carried out at 4.5 μM protein concentration by increasing the temperature of the cuvette from 15 °C to 75 °C, and monitoring the intensity peak shift at 222 nm. Thermodynamic parameters and melting temperature of cytMDH1Red and cytMDH1Ox unfolding were calculated as described previously (Suh and Savisky, 2011). The melting temperature was calculated assuming that the unfolding process occurred following a simple two-state model. The peak shift at 222 nm over the specified temperature range was fitted into the following dose–response equation: y=A+Z−A1+1 0(logX0−X).h (1) The terms of Equation 1 are as follows: A is the folded state, Z is the unfolded state, logX0 is the melting temperature, and h is the Hill coefficient. The thermodynamic parameters were estimated based on the following equations: ΔG=2.303×R×h×(T×Tm–T2) (2) ΔH=2.303×R×T2×h (3) ΔG=ΔH–TΔS (4) where ΔG is the free energy Gibbs variation, ΔH is the enthalpy variation, ΔS is the entropy variation and T is the temperature (30 °C). Protein reduction and oxidation were performed as described before. Crystallization, X-ray data collection, and structure solution Crystals of native cytMDH1 were grown by hanging-drop vapor diffusion at 20 °C in droplets with 1:1 addition of 18 mg ml–1 cytMDH1 to a precipitant of 0.18 M ammonium sulfate, 0.09 M sodium acetate trihydrate, pH 4.6, 27% polyethylene glycol monomethyl ether 2000, and 10% glycerol. Crystals intended for oxidation were also obtained by hanging-drop vapor diffusion, with a crystallization solution of 1.5 M ammonium sulfate, 0.1 M Tris–HCl, pH 8.0. Crystallization droplets were composed of 1 µl of cytMDH1 and 1 µl of precipitant solution, with NAD+ present in 10× molar excess. For X-ray data collection, native cytMDH1 crystals were cryoprotected by supplementation of 80% mother liquor with 20% ethylene glycol. Diffraction data were collected using an in-house MicroMax-007HF X-ray generator with a Saturn 944+ detector. Crystals of cytMDH1 were oxidized through inclusion of 30 mM H2O2 in the cryoprotectant of 0.75 M ammonium sulfate, 50 mM Tris–HCl, pH 8.0, 16% ethylene glycol, 10% glycerol, 10% 1,2-propanediol, with a soak duration of 5 min. Diffraction data from these H2O2-treated crystals were collected at the Proxima 2a beamline at SOLEIL synchrotron (https://www.synchrotron-soleil.fr/en). Indexing and integration of reflections was performed in XDS (Kabsch, 2010a, b), and scaling and merging using AIMLESS (Evans and Murshudov, 2013). Initial phases were determined in Phaser (McCoy et al., 2007) of the PHENIX software suite (Adams et al., 2002, 2010) with Sus scrofa MDH as search model (PDB ID: 4MDH, 62% sequence identity). Manual model building was performed in COOT (Emsley and Cowtan, 2004), with maximum-likelihood refinement in Phenix.refine (Afonine et al., 2012), and REFMAC5 (Murshudov et al., 2011) of the CCP4 suite (Collaborative Computational Project, 1994). Structural stereochemistry was checked in MolProbity (Chen et al., 2010). The native and peroxide-treated crystal structures of cytMDH1 have been submitted to the PDB under the accession codes 5NUF and 5NUE, respectively. Statistics of data reduction and refinement are summarized in Supplementary Table S3. Biomolecular interfaces were analyzed, and biological assemblies assigned using the PISA web service (Krissinel and Henrick, 2007). All structural figures were prepared in PyMOL. Accession numbers Assigned accession numbers for the genes used in this work are as follows: At1g04410 (cytMDH1), At5g43330 (cytMDH2), At5g56720 (cytMDH3), At1g53240 (mtMDH1), At3g15020 (mtMDH2), At2g22780 (perMDH1), At5g09660 (perMDH2), At3g47520 (chlMDH), At5g58330 (NADP-MDH). Reduced and oxidized cytMDH1 structures were deposited at the PDB database (www.PDBe.org) with the respective IDs: 5NUF and 5NUE. The RNA sequencing (RNA-Seq) (Waszczak et al., 2016) and microarray data (Willems et al., 2016) discussed in this article have been deposited in the GEO repository (http://www.ncbi.nlm.nih.gov/geo/) and are accessible through the GEO accession nos GSE77017 and GSE80200, respectively. Results H2O2-treated cytMDH1 forms a disulfide-linked homodimer In order to characterize redox modifications of cytMDH1, we produced cytMDH1 in E. coli, and purified the protein to homogeneity. To validate cysteine sulfenylation in vitro, we applied dimedone, a chemical compound that forms a thioether bond with the electrophilic sulfur of sulfenylated proteins. CytMDH1Red migrated as an ~37 kDa band, which is very close to the expected size of 35.7 kDa based on the amino acid sequence. However, we detected an additional ~70 kDa band of cytMDH1Ox (10 mM H2O2) on the non-reducing CBB-stained SDS–PAGE gel (Fig. 1A), suggesting that cytMDH1 forms a covalently linked dimer upon oxidation. On immunoblots with anti-dimedone antibodies (Fig. 1A), we only observed sulfenylation of the ~37 kDa band of cytMDH1Ox. Remarkably, the dimer was not sulfenylated, suggesting that the covalent dimerization protected cytMDH1Ox from cysteine sulfenylation. We further studied whether the covalent dimer and the sulfenylated form were reduced by DTT and/or the NADPH-dependent TRX system (NTS). While the covalent homodimer was readily reduced either by DTT or NTS, consistent with previous data of Hara et al. (2006), the sulfenylated MDH monomer was only could be reduced by DTT (Fig. 1B). Fig. 1. View largeDownload slide cytMDH1Ox is partially sulfenylated in vitro and dimerizes via a TRX-reducible intermolecular disulfide bond. (A) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence of 2 mM dimedone. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed on a western blot with an anti-cysteine sulfenic acid antibody. (B) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence or absence of the NTS system. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed by immunoblot with an anti-cysteine sulfenic acid antibody. Fig. 1. View largeDownload slide cytMDH1Ox is partially sulfenylated in vitro and dimerizes via a TRX-reducible intermolecular disulfide bond. (A) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence of 2 mM dimedone. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed on a western blot with an anti-cysteine sulfenic acid antibody. (B) cytMDH1Red (8 μM) was incubated with 10 mM DTT (cytMDH1Red+DTT) or treated with 10 mM H2O2 (cytMDH1Ox) in the presence or absence of the NTS system. Samples were analyzed by SDS–PAGE followed by Coomassie blue staining. cytMDH1-SOH formation was analyzed by immunoblot with an anti-cysteine sulfenic acid antibody. To confirm the molecular size of the reduced and oxidized cytMDH1, we evaluated their elution position on a size exclusion chromatography (SEC) column (Supplementary Fig. S2). Both cytMDH1Red and cytMDH1Ox eluted as a single peak of ~70 kDa, indicating that both cytMDH1Red and cytMDH1Ox are in a dimeric form in solution. We further examined cytMDH1 dimerization by migrating cytMDH1Red and cytMDH1Ox on native gels (Supplementary Fig. S3). Both cytMDH1Red and cytMDH1Ox migrated at an apparent molecular mass of ~100 kDa in either non-reducing or reducing conditions. The observed molecular mass is slightly higher than that expected for a dimer (~72 kDa), which could be due to the charge of the protein. However, consistent with the SEC data, no monomeric forms were observed under reducing or oxidizing conditions, supporting the fact that both dimeric forms co-exist (covalent and non-covalent) depending on the redox environment of the protein. cytMDH1Red dimer is probably organized by non-covalent interactions, while cytMDH1Ox is likely to be a mixture of non-covalent and covalent dimers (Fig. 1; Supplementary Figs S2, S3). We also detected several bands close to each other in cytMDH1Ox on the native gel (Supplementary Fig. S3), probably due to oxidation of other cysteines, which gives extra charges to the protein. Notably, the upper band (in between 100 kDa and 130 kDa) almost completely disappeared under reducing conditions. Whether this form corresponds to the covalent dimer cannot be fully confirmed. Collectively, our data suggest that cytMDH1 undergoes a partial conformational change from a non-covalent homodimer to a disulfide-linked homodimer under H2O2 stress. Cysteine and methionine sulfurs are sensitive to oxidation To confirm the sulfenylation of cytMDH1 in vitro and the disulfide bond formation of cytMDH1Ox, we analyzed the purified recombinant protein treated with or without H2O2 in the presence of dimedone by MS. After treatment with 10 mM H2O2 and 2 mM dimedone, we blocked all free thiols by IAM, and the cytMDHOx sample was loaded on a non-reducing SDS–polyacrylamide gel. The 37 kDa and 70 kDa bands (Fig. 1A) were in-gel trypsin digested prior to MS analysis. In the 37 kDa band, a dimedone adduct on the Cys330-containing peptide was found, resulting in a mass increase of 138 Da (Fig. 2A), showing that Cys330 was sulfenylated. Cys330 was also partially present as a free thiol (carbamidomethyl modification by IAM) (Supplementary Table S1) and overoxidized to a sulfonic acid (Fig. 2B). Furthermore, Cys79 was also sulfenylated and overoxidized to a sulfinic acid and sulfonic acid (Supplementary Table S1). All other cysteines were identified as free thiols (Supplementary Table S1). In the 70 kDa band, an intermolecular disulfide bond between Cys330 was found (Fig. 2C), as described previously (Hara et al., 2006). Besides these cysteine modifications, we also observed that methionines were oxidized to a methionine sulfoxide in both 35 kDa (Met24/30/56/87/313) and 70 kDa (Met24/30/41/56/87/270/313) bands (Supplementary Table S1). Among these oxidized methionines, Met313 had been identified previously as a methionine sulfoxide in the cat2 mutant when compared with wild-type Arabidopsis plants (Jacques et al., 2015). All together, we showed that several Cys and Met residues of cytMDH1 are sensitive to oxidation. Fig. 2. View largeDownload slide Identification of oxidative cysteine modifications of cytMDH1Ox by LC-MS/MS. (A) Sulfenylation identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=1009.3. The y and b fragments indicate that Cys330 is present as a dimedone group-modified sulfonic acid (+138.0681 Da). (B) Sulfonic acid identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=919.3. The y and b fragments indicate that Cys330 is present as a sulfonic acid (+48 Da). (C) Disulfide lined peptide identification by using Dbond software. A doubly charged parent ion of [M+2H]2+=871.5 Da shows fragmentation characteristics of a disulfide linkage between two Cys330s of MDH. The y and b series of ions allow the exact localization of the disulfide bridge, p* one strand of the dipeptide. Fig. 2. View largeDownload slide Identification of oxidative cysteine modifications of cytMDH1Ox by LC-MS/MS. (A) Sulfenylation identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=1009.3. The y and b fragments indicate that Cys330 is present as a dimedone group-modified sulfonic acid (+138.0681 Da). (B) Sulfonic acid identified on the Cys330-modified peptide. MS/MS spectrum of the singly charged parent ion [M+H]1+=919.3. The y and b fragments indicate that Cys330 is present as a sulfonic acid (+48 Da). (C) Disulfide lined peptide identification by using Dbond software. A doubly charged parent ion of [M+2H]2+=871.5 Da shows fragmentation characteristics of a disulfide linkage between two Cys330s of MDH. The y and b series of ions allow the exact localization of the disulfide bridge, p* one strand of the dipeptide. Kinetics are affected by H2O2 and TRXh reactivates cytMDH1 We next investigated the effect of cytMDH1 oxidation on its enzymatic activity. The cytMDH1 activity progressively decreased upon H2O2 treatment. No decrease in activity could be observed for the DTT-treated sample over the same time range and temperature, indicating that oxidation of cytMDH1 is the major factor leading to loss of activity (Fig. 3A). To understand how oxidation changes the activity of cytMDH1, we compared the kinetic parameters of cytMDH1Red and cytMDH1Ox (Fig. 3B, C; Table 1). The kcat is 2-fold higher for cytMDH1Red than that of cytMDH1Ox, indicating a decrease of the rate of substrate conversion when the enzyme is oxidized. Interestingly, the KM of cytMDH1Ox for OAA is 143 ± 15 µM, which is 40% lower than that of 238 ± 21 µM for cytMDH1Red. Similarly, the KM of cytMDH1Ox for NADH (50 ± 5 µM) is also 30% lower than that for cytMDH1Red (72 ± 7 µM) (Fig. 3B, C; Table 1), indicating small increases in affinity for both substrates when cytMDH1 is oxidized. Overall, the kcat/KM values for OAA are similar for cytMDH1Red and cytMDH1Ox. The kcat/KM value of cytMDH1Ox for NADH is 6.9 × 106 M–1 s–1, which is in the same range as the 9.0 × 106 M–1 s–1 value obtained for cytMDH1Red. This suggests that a KM adjustment occurs, to compensate for the kcat decrease between cytMDH1Red and cytMDH1Ox. Fig. 3. View largeDownload slide CytMDH1 is sensitive to oxidation and reactivated by TRXh in vitro. (A) The time course of cytMDH1 activity was measured under reducing or oxidizing conditions. cytMDH1Red at 8 μM was incubated with buffer (cytMDH1Red), with 10 mM DTT (cytMDH1Red+DTT), or with 10 mM H2O2 (cytMDH1Red+H2O2) at 30 °C. The plot of k (s–1) versus time of treatment is shown. Data presented are means ±SD (n >3). (B and C) CytMDH1 activity was measured at different concentrations of substrates: (B) OAA (0–2.5 mM) or (C) NADH (0–0.25 mM). The initial velocity (vi) for each substrate concentration was measured, and the k (s–1) values versus [OAA] or [NADH] were plotted and fitted with the Michaelis–Menten equation to obtain the kinetic parameters KM and specific activity. At least three independent replicates of k were measured for each substrate concentration. (D) CytMDH1Ox was incubated with increasing concentrations of recombinant TRXh3 in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). (E) Reduced or H2O2-oxidized cytMDH1 was incubated with different TRXhs (2 µM) in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). An asterisk indicates values significantly different from cytMDH1Red by the Student’s t-test at 5‰ (*P<5 × 10–3). Fig. 3. View largeDownload slide CytMDH1 is sensitive to oxidation and reactivated by TRXh in vitro. (A) The time course of cytMDH1 activity was measured under reducing or oxidizing conditions. cytMDH1Red at 8 μM was incubated with buffer (cytMDH1Red), with 10 mM DTT (cytMDH1Red+DTT), or with 10 mM H2O2 (cytMDH1Red+H2O2) at 30 °C. The plot of k (s–1) versus time of treatment is shown. Data presented are means ±SD (n >3). (B and C) CytMDH1 activity was measured at different concentrations of substrates: (B) OAA (0–2.5 mM) or (C) NADH (0–0.25 mM). The initial velocity (vi) for each substrate concentration was measured, and the k (s–1) values versus [OAA] or [NADH] were plotted and fitted with the Michaelis–Menten equation to obtain the kinetic parameters KM and specific activity. At least three independent replicates of k were measured for each substrate concentration. (D) CytMDH1Ox was incubated with increasing concentrations of recombinant TRXh3 in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). (E) Reduced or H2O2-oxidized cytMDH1 was incubated with different TRXhs (2 µM) in the presence of NADPH and NTRA. After 30 min incubation, the mix was diluted 40-fold and used for NAD-MDH activity assay. Data presented are means ±SD (n >3). An asterisk indicates values significantly different from cytMDH1Red by the Student’s t-test at 5‰ (*P<5 × 10–3). Table 1. Steady-state parameters for reduced and oxidized cytMDH1 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 View Large Table 1. Steady-state parameters for reduced and oxidized cytMDH1 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 KM OAA (µM) KM NADH (µM) kcat OAA (s–1) kcat NADH (s–1) kcat/KM OAA (M–1 s–1) kcat/KM NADH (M–1 s–1) cytMDH1Red 238 ± 21 72 ± 7 608 ± 22 677 ± 24 2.6 × 106 9.0 × 106 cytMDH1Ox 143 ± 15 50 ± 5 351 ± 11 343 ± 12 2.5 × 106 6.9 × 106 View Large Previously, it has been shown that TRXh1 reactivates cytMDH1Ox by reducing a disulfide bond at the Cys330 residue (Hara et al., 2006). The Arabidopsis genome encodes five cytosolic TRX isoforms (TRXh1–TRXh5) that might reduce cytMDH1. In order to test if TRXhs could reduce cytMDH1, we first tested protein interactions between recombinant TRXh1–TRXh5 and MDH isoforms in a yeast two-hybrid system (Vignols et al., 2005). In order to stabilize redox-sensitive interactions, the resolving cysteine of the TRX active site was mutated to a serine. All five TRXh isoforms interacted with cytMDH1. However, no interaction between the cytosolic TRXhs and mitMDH1, chlMDH, and NADP-MDH was observed, indicating a specificity of the TRXhs to interact with cytMDH1 (Supplementary Fig. S4). To test if MDHs are able to interact with TRXs other than TRXh, we performed a yeast two-hybrid test using the chloroplastic TRXf1 and f2 as bait. All tested MDHs (mitMDH1, chlMDH, cytMDH1, and NADP-MDH) were able to interact with TRXf1 and f2, suggesting that TRXfs have a larger interaction spectrum than TRXhs for MDH isoforms (Supplementary Fig. S5). We further tested whether recombinant TRXhs are able to restore the activity of H2O2-oxidized cytMDH1. All TRXh1–TRXh5 isoforms identified as interactors of cytMDH1 by yeast two-hybrid assay were able to restore the activity of cytMDH1Ox to a level similar to that of cytMDH1Red (Fig. 3D, E), implying that the TRXh–MDH interaction occurs in a redox-dependent manner. The thermodynamic stability of cytMDH1 is affected by H2O2 In order to understand how H2O2 affects the activity of cytMDH1, the secondary structures of reduced and oxidized cytMDH1 were compared with far UV-CD (Fig. 4A). cytMDH1Red and cytMDH1Ox are characterized by a typical α-helical and β-sheet secondary structure with negative peaks around 210–220 nm and a positive peak around 190 nm. However, the spectrum of cytMDH1Ox shows a pronounced modification of the secondary structure, indicating a more disordered protein. We further determined the influence of oxidation on the protein stability by following the unfolding of oxidized and reduced cytMDH1 at 222 nm from 15 °C to 75 °C by CD. We obtained a melting temperature (Tm) of cytMDH1Red of 54.74 °C, while the Tm of cytMDH1Ox dropped by almost 5 °C to 49.88 °C (Fig. 4B), indicating that cytMDH1Ox is less stable than cytMDH1Red. Thermodynamic processes, such as the thermal unfolding of cytMDH1, are characterized by a variation in entropy (ΔS), enthalpy (ΔH), and Gibbs free energy (ΔG) (Murphy and Freire, 1992). Thermal unfolding parameters specify a higher need for energy in order to disrupt intramolecular interactions such as Van der Waals forces, hydrogen bonds, and ionic salt bridges for cytMDH1Red, based on the ΔΔH value of 8.6 kcal mol–1 (Table 2). These interactions are well known as the main intramolecular forces that keep proteins structured, indicating that the amount of energy to unfold a protein is directly proportional to the interactions previously mentioned (O’Brien and Haq, 2004). The more flexible a protein is, the higher are the number of possible conformations. This effect is represented by conformational entropy (O’Brien and Haq, 2004). The ΔΔS value of 23 cal mol–1 KJ–1 suggests that cytMDH1Red undergoes a larger conformational change on the way to reach its unfolded conformation than its oxidized counterpart (Table 2). This result indicates that before thermal denaturation, cytMDH1Red is in a more restricted number of conformations maintained by a larger amount of interactions when compared with cytMDH1Ox. Fig. 4. View largeDownload slide CD of reduced and oxidized cytMDH1. (A) The far CD spectra for the cytMDH1Red and cytMDH1Ox were obtained with an average of three scans for each spectrum with a range from 190 nm to 260 nm. (B) Thermal unfolding monitored at 222 nm was carried out for cytMDH1Red and cytMDH1Ox by increasing the temperature from 15 °C to 75 °C. Fig. 4. View largeDownload slide CD of reduced and oxidized cytMDH1. (A) The far CD spectra for the cytMDH1Red and cytMDH1Ox were obtained with an average of three scans for each spectrum with a range from 190 nm to 260 nm. (B) Thermal unfolding monitored at 222 nm was carried out for cytMDH1Red and cytMDH1Ox by increasing the temperature from 15 °C to 75 °C. Table 2. Tm and thermodynamic parameters of reduced and oxidized cytMDH1 thermal denaturation ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔΔ stands for reduced minus oxidized. a Thermodynamic constants calculated at 30 °C. View Large Table 2. Tm and thermodynamic parameters of reduced and oxidized cytMDH1 thermal denaturation ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔH (kcal mol–1) ΔS (cal mol–1 KJ–1) ΔG (kcal mol–1) ΔΔHa (kcal mol–1) ΔΔSa (cal mol–1 KJ–1) ΔΔGa (kcal mol–1) Tm (°C) cytMDH1Red 63.4 ± 0.2 192 ± 1 5.10 ± 0.01 8.6 23 1.5 54.7 ± 0.2 cytMDH1Ox 54.8 ± 0.6 169 ± 1 3.62 ± 0.02 49.8 ± 0.1 ΔΔ stands for reduced minus oxidized. a Thermodynamic constants calculated at 30 °C. View Large Crystal structure of cytMDH1 In an effort to gain structural insights into the effects of oxidative stress on cytMDH1, the crystal structures of cytMDH1Red and H2O2-soaked cytMDH1 were determined to a resolution of 1.8 Å and 1.35 Å, respectively (www.PDBe.org, ID: 5NUF and 5NUE). CytMDH1 crystallizes as a homodimer with an interfacing area of 1817 Å2 as determined by PISA analysis, and exhibited the structural fold common to the cytosolic MDH from porcine heart (Fig. 5) (Birktoft et al., 1989). NAD+ was modeled into the available density in the mFO–DFC omit map at the nucleotide-binding cleft of cytMDH1. A sulfate ion was found to occupy the malate/OAA-binding site proximal to the nicotinamide head of NAD+, binding with five water molecules and the side-chain groups of residues Asn132, Arg163, His188, and Ser243. Fig. 5. View largeDownload slide Crystal structures of cytMDH1 and sulfoxidation. (A) The conserved dimeric structural arrangement of MDH places the C-termini at opposite ends. Here in the crystal structure of Arabidopsis cytMDH1, each protomer is colored separately, and the side chain of Cys330 and NAD+ are colored orange and green, respectively. (B) Met97 is located on a flexible loop proximal to the OAA/malate-binding site, which in the crystal structure is occupied by a sulfate ion. Superposition of two subunits of the cytMDH1 homodimer shows that this flexible loop can be in either a closed (pink) or an open (blue) conformation independent of the loop conformation of the neighboring dimer subunit. Sulfoxidation of Met97 observed in the crystal structure of cytMDH1 is indicated, though it should be noted that this flexible loop region can adopt a closed/open conformation irrespective of Met97 sulfoxidation. Fig. 5. View largeDownload slide Crystal structures of cytMDH1 and sulfoxidation. (A) The conserved dimeric structural arrangement of MDH places the C-termini at opposite ends. Here in the crystal structure of Arabidopsis cytMDH1, each protomer is colored separately, and the side chain of Cys330 and NAD+ are colored orange and green, respectively. (B) Met97 is located on a flexible loop proximal to the OAA/malate-binding site, which in the crystal structure is occupied by a sulfate ion. Superposition of two subunits of the cytMDH1 homodimer shows that this flexible loop can be in either a closed (pink) or an open (blue) conformation independent of the loop conformation of the neighboring dimer subunit. Sulfoxidation of Met97 observed in the crystal structure of cytMDH1 is indicated, though it should be noted that this flexible loop region can adopt a closed/open conformation irrespective of Met97 sulfoxidation. Superposition of the individual protomers of the cytMDH1 homodimer revealed significant differences in the conformation of a flexible loop of Pro92–Val102 at the opening to the substrate-binding site. In one conformation, this loop is in a ‘closed’ conformation, and co-ordinates a sulfate ion through the backbone amide and side-chain head of Arg99. The alternative ‘open’ conformation of this flexible loop region was found to be more disordered, with elevated B-factors and relatively poor coverage of electron density. No significant structural differences were observed between the structures of cytMDH1 and the H2O2-soaked cytMDH1 crystal, both structures aligning with a root mean square deviation (RMSD) of 0.266 Å, and the largest structural deviations arising in the flexible loop region described above (Fig. 5A). Evidence of partial sulfoxidation was identified on Met56 and Met97. Sulfoxidation of Met56 appears to have no direct effect on the local structural environment. The Met97 is located on the flexible loop proximal to the malate/OAA-binding site, which has also been previously identified to form methionine sulfoxide in Arabidopsis plants (Jacques et al., 2015). Sulfoxidation of Met97 could conceivably affect mobility of the Pro92–Val102 loop region, and thereby possibly influence substrate delivery to the active site (Fig. 5B). However, we found no significant difference in activity after we oxidized the IAM-pretreated cytMDH1 with 10 mM H2O2 (Supplementary Fig. S6), indicating that methionine oxidation does not play a major role in the effect of oxidation on the enzymatic capacity of cytMDH1. From a mFO–DFC difference map, additional positive density adjoining the sulfur of Cys330 was observed in two of the three molecules of the asymmetric unit of cytMDH1. Into this extra density, partial-occupancy sulfenylation or sulfinylation at Cys330 was modeled for the respective molecules of cytMDH1 (Fig. 6). The partial occupancy of the cysteine sulfenylation and sulfinylation may be due to the short duration of the H2O2 crystal soak, resulting in an unequal distribution of peroxide throughout the crystal unit cells. Fig. 6. View largeDownload slide Cysteine oxidation observed in the crystal structure of H2O2-treated cytMDH1. The presence of additional density arising in the FO–FC map at Cys330 of chain B (A) and chain C (C) prompted modeling of a partial-occupancy sulfinic and sulfenic acid (B and D, respectively). Shown are 2mFO–DFC (gray) and mFO-DFC (green) electron density maps contoured at 0.33 e Å–3 and 0.34 e Å–3, respectively, for (C), and 0.41 e Å–3 and 0.295 e Å–3 for both (B) and (D). The 2mFO–DFC map for (A) is contoured at 0.25 e Å–3. Fig. 6. View largeDownload slide Cysteine oxidation observed in the crystal structure of H2O2-treated cytMDH1. The presence of additional density arising in the FO–FC map at Cys330 of chain B (A) and chain C (C) prompted modeling of a partial-occupancy sulfinic and sulfenic acid (B and D, respectively). Shown are 2mFO–DFC (gray) and mFO-DFC (green) electron density maps contoured at 0.33 e Å–3 and 0.34 e Å–3, respectively, for (C), and 0.41 e Å–3 and 0.295 e Å–3 for both (B) and (D). The 2mFO–DFC map for (A) is contoured at 0.25 e Å–3. Redox sensitivity of NAD-MDHs varies in different plant cell compartments To obtain an overview of the redox sensitivity of NAD-MDH in the plant cell, we measured the NAD-MDH activity in purified cytosol, mitochondrion, and chloroplast fractions before and after the 10 mM H2O2 treatment. Both the cytosolic and chloroplastic NAD-MDH activities were significantly decreased after the H2O2 treatment (40% and 20%, respectively) (Fig. 7A). No significant difference of the mitochondrial NAD-MDH activities between the non-treated and treated samples was observed. Fig. 7. View largeDownload slide NAD-MDH activity in different cell compartments and mutants. (A) The NAD-MDH activity was measured for different cell compartments in wild-type (Col-0) plant extracts. A 3 µg aliquot of protein extracts was used for activity assays. Data presented are means ±SD (n ≥3). An asterisk indicates values significantly different from the wild type by the Student’s t-test (*P<0.05, **P<0.01). (B) The NAD-MDH activity was measured in wild-type (Col-0) and different mutant plants grown in soil for 2 weeks (white bars) or 3 weeks (black bars). A 5 µg aliquot of protein extracts was used for activity assays. (C) Protein-based complementation assays were performed in the same mutants as in (A). A 5 µg aliquot of protein extracts was untreated (black bars) or incubated in 1 mM NADPH in the presence of 4.59 μM TRXh3 and 3.12 µM NTRA for 2 h on ice (white bars). This reaction mixture was diluted 40 times before performing the NAD-MDH activity assay. Data presented are means ±SD (n=4). An asterisk indicates values significantly different from the wild type by the Student’s t-test at 5‰ (*P<5 × 10–3). Fig. 7. View largeDownload slide NAD-MDH activity in different cell compartments and mutants. (A) The NAD-MDH activity was measured for different cell compartments in wild-type (Col-0) plant extracts. A 3 µg aliquot of protein extracts was used for activity assays. Data presented are means ±SD (n ≥3). An asterisk indicates values significantly different from the wild type by the Student’s t-test (*P<0.05, **P<0.01). (B) The NAD-MDH activity was measured in wild-type (Col-0) and different mutant plants grown in soil for 2 weeks (white bars) or 3 weeks (black bars). A 5 µg aliquot of protein extracts was used for activity assays. (C) Protein-based complementation assays were performed in the same mutants as in (A). A 5 µg aliquot of protein extracts was untreated (black bars) or incubated in 1 mM NADPH in the presence of 4.59 μM TRXh3 and 3.12 µM NTRA for 2 h on ice (white bars). This reaction mixture was diluted 40 times before performing the NAD-MDH activity assay. Data presented are means ±SD (n=4). An asterisk indicates values significantly different from the wild type by the Student’s t-test at 5‰ (*P<5 × 10–3). In order to study the impact of H2O2 on the MDH gene expression, we analyzed the microarray and RNA-Seq data for all MDH genes (Supplementary Fig. S7). Microarray analyses were performed on 7-day-old wild-type (Col-0) Arabidopsis seedlings without and with 10 mM H2O2 treatment for 24 h (Willems et al., 2016). RNA-Seq analysis was done on 21-day-old Col-0 and cat2 mutant Arabidopsis seedlings under photorespiratory stress (Waszczak et al., 2016). No major change was found for mRNA levels of all MDHs upon stresses from both data sets. From the RNA-Seq data set, similar mRNA levels of MDHs were found between the Col-0 wild-type and cat2 mutant Arabidopsis plants. Collectively, these data indicate that H2O2 has little influence on the transcriptional level of MDHs, and that the impact of H2O2 on NAD-MDH activities is probably due to post-translational modifications. Mutations in cat2 and ntra ntrb lower the in vivo NAD-MDH activity We further evaluated the NAD-MDH activity in a cat2 loss-of-function mutant, in which the major catalase isoform in leaves is inactivated, retaining only 8% of extractible catalase activity (Mhamdi et al., 2010b) (Supplementary Fig. S8). The NAD-MDH activity was significantly decreased in the cat2 mutant when compared with wild-type plants (25–33%), indicating that the NAD-MDH activity is affected by increased in vivo H2O2 levels (Fig. 7B). As we showed that the reduction in cytMDH1Ox depends on cytosolic TRXh, we also measured the NAD-MDH activity in the TRX reductase double mutant ntra ntrb in which TRXhs are partially inactive (Reichheld et al., 2007). As previously observed (Daloso et al., 2015), we also found a significant decrease (17–20%) in this mutant. However, decreasing the NTR gene expression with RNAi in cat2 (cat2 ntra ntrb) had no additional impact on the NAD-MDH activity, suggesting that the residual NTR activity in the RNAi lines is efficient enough to reduce the oxidized MDH (Fig. 7B; Supplementary Figs S8, S9). We further tested whether the glutathione-reducing system could be part of the cytMDH redox regulation mechanism. Therefore, the NAD-MDH activity was measured in a cytosolic glutathione reductase mutant (gr1) with a more oxidized glutathione redox state (Supplementary Table S2). No decrease in NAD-MDH activity was observed, and in the gr1 cat2 double mutant, the decrease in NAD-MDH activity was similar to that observed in the cat2 mutant. Finally, to confirm that the decreased NAD-MDH activities in cat2 and the ntra ntrb double mutants are linked to TRX reduction, we performed NAD-MDH activity assays for the plant extracts in the presence of recombinant cytosolic NTRA and TRXh3 proteins (Fig. 7C). After the addition of the NTR/TRXh3 system in the wild-type plant extracts, the NAD-MDH activity remained the same. However, the activity was significantly increased in the plant extracts from cat2 and ntra ntrb mutants, confirming a TRX-dependent activity of the MDH pool. Discussion Plant NAD(P)-dependent MDHs are part of a multigenic family (Supplementary Fig. S1). The redox regulation of chloroplastic NADP-MDH has been extensively studied in the last decades. However, a redox regulation of other isoforms of NAD-MDH is unlikely to act in a similar way. Accordingly, the mitochondrial NAD-MDH isoforms that are involved in the tricarboxylic acid (TCA) cycle are not prone to redox regulation (Yoshida and Hisabori, 2016), though other enzymes of the TCA cycle have been shown to be redox regulated (Daloso et al., 2015). Our NAD-MDH activity assays performed on purified subcellular fractions also support redox-independent mitochondrial NAD-MDH activities. However, the H2O2 sensitivity of the chloroplastic NAD-MDH activity might indicate a redox dependence of chloroplastic NAD-MDH in vivo, which was not previously found in vitro (Berkemeyer et al., 1998). A potential redox regulation of the chloroplastic NAD-MDH activity will need to be further explored. In clear contrast to mitochondrial isoforms, different arguments are in favor of a redox regulation of the Arabidopsis cytMDH1 (Hara et al., 2006; Hara and Hisabori, 2013; this work): (i) cytMDH1 was identified as a sulfenylated protein under oxidative stress and during symbiosis (Waszczak et al., 2014; Akter et al., 2015), which has been validated in vitro in our study (Fig. 1; Supplementary Fig. S10); (ii) not only cysteine sulfenylation, but also cysteine overoxidation and methionine sulfoxidations occur under oxidizing conditions (Figs 3, 6, 7; Supplementary Table S2); (iii) modifications in enzymatic kinetics, secondary structure, and folding characteristics are observed under H2O2 treatment (Figs 3–5); and (iv) NAD-MDH activities are affected in H2O2-supplied purified cytosolic fractions, and in mutants accumulating high levels of H2O2 (cat2) or impaired in the TRX reduction systems (ntra ntrb) (Fig. 7). Therefore, our present work extends in vitro data demonstrating a redox regulation of cytMDH1, and we provide clear evidence that this regulation occurs in vivo. All eukaryotic cytosolic NAD-MDHs exist as homodimers (Berkemeyer et al., 1998; Birktoft et al., 1989), which was confirmed for the cytMDH1 isoform by SEC (Hara et al., 2006). Our structural data also confirm this observation (Fig. 6). Moreover, we show that a major conformational change occurs between cytMDH1Red and dimeric cytMDH1Ox. Whereas the cytMDH1Red dimer is mainly stabilized by electrostatic and hydrophobic interactions, as visualized by SEC and native gels (Supplementary Figs S2, S3), oxidation causes the formation of a disulfide-linked dimer (Figs 2, 3). It has been hypothesized that an intermolecular Cys330–Cys330 bond co-ordinates a tetrameric structure (Hara et al., 2006), but not supported by our SEC and native gel profiles, because no tetrameric structure was observed (Supplementary Figs S2, S3), possibly due to different oxidation conditions between the two studies: we used H2O2-induced oxidation, while CuCl2-induced oxidation was used in the previous study (Hara et al., 2006). The H2O2-mediated cytMDH1 oxidation inducing a conformational modification is reminiscent of the conformational change reported for oxidized chloroplastic NADP-MDH (Ruelland et al., 1998). The formation of the C-terminal intramolecular disulfide bond upon oxidation of this enzyme triggers the binding of the C-terminal domain to OAA. In contrast, cytMDH1 oxidation has a minor effect on the affinity for the substrate, even slightly decreasing the KM for OAA and NADH. Therefore, a similar redox regulatory mechanism between the two isoforms can be ruled out. However, the observed decrease in the turnover rate (kcat) of the cytMDH1Ox suggests that this modification of the conformation affects the catalytic activity of the enzyme. Surprisingly, the comparison of the crystal structures of reduced and oxidized cytMDH1 does not show extensive changes in the global conformational fold of both structures. However, we should note that the crystal structure of oxidized cytMDH1 is a crystal of cytMDH1Red incubated with H2O2. Therefore, the constraints of the crystal lattice preclude the possibility of large-scale conformational changes. As both cytMDH1Red and cytMDH1Ox form homodimers, it is difficult to separate cytMDH1Ox from the cytMDH1Red for cytMDH1Ox crystallization trials. Although Met97 at the substrate-binding site was sensitive to oxidation (Fig. 6), the enzyme kinetics show no major impact of oxidation on the affinity of cytMDH1 towards OAA and NADH (Table 1). The activity of alkylated cytMDH1 is not sensitive to H2O2 oxidation, whereas cytMDH1Red is very sensitive to diamide (Supplementary Fig. S6), a reagent that introduces disulfide bonds. Altogether, our results indicate that the decrease in activity is not due to methionine oxidation, but rather to disulfide bond formation. Among the cysteine residues of cytMDH1 undergoing oxidative modifications, Cys79 is prone to overoxidation, while Cys330 tends preferentially to form a Cys330–Cys330 disulfide bond (Supplementary Table S2), suggesting that disulfide formation protects Cys330 from overoxidation (Fig. 7). We also demonstrated that sulfenylation of Cys does not by itself impact the activity of cytMDH1 (Supplementary Fig. S10). Therefore, we conclude that in contrast to another sulfenylated residue, Cys79, the sulfenylated Cys330 forms a disulfide bond with free Cys330 thiol (Fig. 8), which causes the reversible conformational change and a decrease of the activity. Fig. 8. View largeDownload slide How cytMDH1 copes with H2O2 oxidation. Oxidation of homodimeric cytMDH1 leads to sulfenylation of Cys330 and the formation of an intermolecular Cys330–Cys330 disulfide. Therefore, a major conformational change is required that leads to the formation of a thermodynamically less stable homodimer with an altered kinetic behavior but with similar kcat/KM values. The S–S linked dimer protects the Cys330 from irreversible overoxidation and becomes a substrate for the TRX pathway. Methionine sulfoxidation and sulfenylation of cysteines other than Cys330 do not affect the kinetic behavior. Fig. 8. View largeDownload slide How cytMDH1 copes with H2O2 oxidation. Oxidation of homodimeric cytMDH1 leads to sulfenylation of Cys330 and the formation of an intermolecular Cys330–Cys330 disulfide. Therefore, a major conformational change is required that leads to the formation of a thermodynamically less stable homodimer with an altered kinetic behavior but with similar kcat/KM values. The S–S linked dimer protects the Cys330 from irreversible overoxidation and becomes a substrate for the TRX pathway. Methionine sulfoxidation and sulfenylation of cysteines other than Cys330 do not affect the kinetic behavior. Intriguingly, our crystal structure data show that Cys330 residues from two interacting subunits are located at opposite sides in the cytMDH1 dimeric molecule (Fig. 6). A hypothesis is that upon H2O2 oxidation, a major conformational change brings Cys330 of both subunits in close proximity to permit Cys330–Cys330 disulfide bond formation. As mentioned above, our oxidized crystal structure does not allow the visualization of such conformational changes and we would need to obtain a crystal structure after oxidation of the protein in solution, but have failed as yet. Another more likely hypothesis is that the disulfide bond is formed between Cys330 of two individual non-covalent dimers. As a tetrameric cytMDH1 state was not detected, we suggest that Cys330–Cys330 formation between dimers leads to structural changes that disrupt the native non-covalent dimer, leading to a population of covalent homodimers (Fig. 8). This scenario is supported by the CD analyses demonstrating that cytMDH1Ox is significantly more disordered (Fig. 4A) than the cytMDH1Red, comprising a larger number of possible states prior to thermal denaturation (Fig. 4B; Table 2), and that cytMDH1Red is in a more restricted number of possible conformations. Additionally, ΔΔG and Tm values also indicate that cytMDH1Red is significantly more stable than cytMDH1Ox (Table 2). CD and steady-state kinetic data reinforce the final conclusion that oxidation of cytMDH1 induces a conformational disorder and increases instability of the protein, possibly caused both by Cys330–Cys330 disulfide formation and overoxidation. Furthermore, the disulfide bond dimer can be efficiently reactivated by TRXh proteins, providing full reversibility to the system and further demonstrating the role of the disulfide bond in protecting the enzyme activity (Fig. 8). Interestingly, the Cys330 is also conserved in the other two cytMDH isoforms encoded by the Arabidopsis genome (Supplementary Fig. S1) as well as in cytMDH orthologs of other plants, suggesting that this mechanism is conserved in the green lineage. It is also likely that the decrease in NAD-MDH activities observed in the cat2 mutant is reporting the inactivation of all three cytosolic isoforms. Altered activity is also found in the ntra ntrb mutants, where the TRX system is partially inactivated (Reichheld et al., 2007). Decreased NAD-MDH activity is probably due to the lower capacity of the ntra ntrb mutant to regenerate oxidized MDH in vivo. The Arabidopsis genome harbors multiple cytosolic TRXs (Meyer et al., 2012). We have tested five isoforms of TRXh with cytMDH1 as substrate, and all of them are able to restore the cytMDH1 activity after oxidation. It is likely that these TRX isoforms are redundant, which also explain the decreased NAD-MDH activity that is only observed in the ntra ntrb mutant, and not with several TRXh single mutants (trxh1, trxh2, trxh3, trxh4, and trxh5) tested so far (data not shown). Interestingly, our yeast two-hybrid test showed that cytosolic TRXhs interact specifically with cytMDH1. As Cys330 is only found in cytosolic MDHs, it is tempting to speculate that TRXh/cytMDH forms a mixed disulfide between the catalytic Cys of TRXh and Cys330 of cytMDH1, although other structural features could mediate this specific interaction. Cytosolic MDH plays a key role in equilibrating reducing equivalents between the organelles as well as in providing an important pool of OAA for further metabolic activities. Therefore, its activity needs to be preserved under stress conditions leading to reduced availability of its substrates. Our data clearly suggest that the redox regulation of cytMDH is playing an efficient protective role to avoid irreversible overoxidation of the enzyme. Interestingly, our kinetic data show that the cytMDH1Ox catalytic efficiency (kcat/KM) is not profoundly perturbed compared with that of cytMDH1Red. While the turnover rate (kcat) is 2-fold decreased, the affinity (KM) for the substrates is slightly increased, suggesting that cytMDH1Ox is adapting its activity to low availability of the substrate. Such a buffering mechanism as well as the reversibility of the regulation by the TRX system probably acts as a safety mechanism to minimize oxidative inactivation of cytMDH1 under stress conditions. Such a fine-tuned redox regulation is also consistent with the fact that the cytMDH1 protein is abundantly maintained under oxidative stress, in order to adapt the cell metabolism under changing environmental conditions. Supplementary data Supplementary data are available at JXB online. Fig. S1. Alignment of Arabidopsis thaliana MDHs. Fig. S2. Dimeric forms of reduced and oxidized cytMDH1 identified by size exclusion chromatography. Fig. S3. Reduced and oxidized cytMDH1 in native gel. Fig. S4. Cytosolic TRXhs interact with cytMDH1. Fig. S5. Plastidial TRXf1 and TRXf2 interact with members of the MDH family. Fig. S6. Redox sensitivity of cytMDH1 is regulated by TRXh on the disulfide bridge. Fig. S7. Transcriptional analysis of MDHs under stress conditions. Fig. S8. Catalase activity in mutants. Fig. S9. NTR accumulation is decreased in RNAi lines. Fig. S10. Sulfenylation does not affect the activity of cytMDH1. Table S1. Mass spectral summary of post-translational oxidation modifications on cysteines and methionines. Table S2. Glutathione levels in the gr1 mutant. Table S3. Crystallographic data collection and refinement statistics Abbreviations: Abbreviations: CD circular dichroism chlMDH chloroplastic NAD-dependent MDH cytMDH cytosolic NAD-dependent MDH IAM iodoacetamide MDH malate dehydrogenase mitMDH mitochondrial NAD-dependent MDH NADP-MDH NADP-dependent MDH NTRA/B NADPH-dependent thioredoxin reductase A/B OAA oxaloacetate perMDH peroxisomal NAD-dependent MDH SEC size exclusion chromatography –SOH sulfenic acid TCA tricarboxylic acid TRX thioredoxin. Acknowledgements This work was supported by grants from (i) the Centre National de la Recherche Scientifique and the Agence Nationale de la Recherche (ANR-Blanc Cynthiol 12-BSV6-0011); (ii) VIB (to JM); (iii) the SPR34 project of the Vrije Universiteit Brussel (VUB) (to JM); (iv) Research Foundation Flanders (grant no. G0D7914N) (to JM and FVB); and (v) Flanders Hercules Foundation (grant no. HERC16) for the purification platform (to JM). We thank the SOLEIL beamline scientists at Proxima 2a for their technical assistance in data collection; Inge Van Molle from VIB-VUB Center for Structure Biology for help with structural data collection; Barbara De Smet and Valerie Van Ruyskensvelde from VIB-UGent Center for Plant Systems Biology for the genotype and phenotype screening; Jing Yang from Beijing Proteome Research Center (BPRC) for the mass spectrometry support; and Patrick Willems from VIB-UGent Center for Plant Systems Biology for bioinformatics analysis. References Adams PD , Afonine PV , Bunkóczi G , et al. 2010 . PHENIX: a comprehensive Python-based system for macromolecular structure solution . Acta Crystallographica. Section D, Biological Crystallography 66 , 213 – 221 . Google Scholar CrossRef Search ADS PubMed Adams PD , Grosse-Kunstleve RW , Hung LW , et al. 2002 . PHENIX: building new software for automated crystallographic structure determination . Acta Crystallographica. Section D, Biological Crystallography 58 , 1948 – 1954 . Google Scholar CrossRef Search ADS PubMed Afonine PV , Grosse-Kunstleve RW , Echols N , et al. 2012 . Towards automated crystallographic structure refinement with phenix.refine . Acta Crystallographica. Section D, Biological Crystallography 68 , 352 – 367 . Google Scholar CrossRef Search ADS PubMed Akter S , Huang J , Bodra N , et al. 2015 . DYn-2 based identification of Arabidopsis sulfenomes . Molecular and Cellular Proteomics 14 , 1183 – 1200 . Google Scholar CrossRef Search ADS PubMed Berkemeyer M , Scheibe R , Ocheretina O . 1998 . A novel, non-redox-regulated NAD-dependent malate dehydrogenase from chloroplasts of Arabidopsis thaliana L . Journal of Biological Chemistry 273 , 27927 – 27933 . Google Scholar CrossRef Search ADS PubMed Birktoft JJ , Rhodes G , Banaszak LJ . 1989 . Refined crystal structure of cytoplasmic malate dehydrogenase at 2.5-Å resolution . Biochemistry 28 , 6065 – 6081 . Google Scholar CrossRef Search ADS PubMed Carr PD , Verger D , Ashton AR , Ollis DL . 1999 . Chloroplast NADP-malate dehydrogenase: structural basis of light-dependent regulation of activity by thiol oxidation and reduction . Structure 7 , 461 – 475 . Google Scholar CrossRef Search ADS PubMed Chen VB , Arendall WB 3rd , Headd JJ , Keedy DA , Immormino RM , Kapral GJ , Murray LW , Richardson JS , Richardson DC . 2010 . MolProbity: all-atom structure validation for macromolecular crystallography . Acta Crystallographica. Section D, Biological Crystallography 66 , 12 – 21 . Google Scholar CrossRef Search ADS PubMed Choi S , Jeong J , Na S , Lee HS , Kim HY , Lee KJ , Paek E . 2010 . New algorithm for the identification of intact disulfide linkages based on fragmentation characteristics in tandem mass spectra . Journal of Proteome Research 9 , 626 – 635 . Google Scholar CrossRef Search ADS PubMed Collaborative Computational Project, Number 4 . 1994 . The CCP4 suite: programs for protein crystallography . Acta Crystallographica. Section D, Biological Crystallography 50 , 760 – 763 . CrossRef Search ADS PubMed Daloso DM , Müller K , Obata T , et al. 2015 . Thioredoxin, a master regulator of the tricarboxylic acid cycle in plant mitochondria . Proceedings of the National Academy of Sciences, USA 112 , E1392 – E1400 . Google Scholar CrossRef Search ADS Emsley P , Cowtan K . 2004 . Coot: model-building tools for molecular graphics . Acta Crystallographica. Section D, Biological Crystallography 60 , 2126 – 2132 . Google Scholar CrossRef Search ADS PubMed Evans PR , Murshudov GN . 2013 . How good are my data and what is the resolution ? Acta Crystallographica. Section D, Biological Crystallography 69 , 1204 – 1214 . Google Scholar CrossRef Search ADS PubMed Faske M , Holtgrefe S , Ocheretina O , Meister M , Backhausen JE , Scheibe R . 1995 . Redox equilibria between the regulatory thiols of light/dark-modulated chloroplast enzymes and dithiothreitol: fine-tuning by metabolites . Biochimica et Biophysica Acta 1247 , 135 – 142 . Google Scholar CrossRef Search ADS PubMed Gelhaye E , Rouhier N , Jacquot J-P . 2003 . Evidence for a subgroup of thioredoxin h that requires GSH/Grx for its reduction . FEBS Letters 555 , 443 – 448 . Google Scholar CrossRef Search ADS PubMed Hara S , Hisabori T . 2013 . Kinetic analysis of the interactions between plant thioredoxin and target proteins . Frontiers in Plant Science 4 , 508 . Google Scholar CrossRef Search ADS PubMed Hara S , Motohashi K , Arisaka F , Romano PG , Hosoya-Matsuda N , Kikuchi N , Fusada N , Hisabori T . 2006 . Thioredoxin-h1 reduces and reactivates the oxidized cytosolic malate dehydrogenase dimer in higher plants . Journal of Biological Chemistry 281 , 32065 – 32071 . Google Scholar CrossRef Search ADS PubMed Hossain MS , Dietz KJ . 2016 . Tuning of redox regulatory mechanisms, reactive oxygen species and redox homeostasis under salinity stress . Frontiers in Plant Science 7 , 548 . Google Scholar PubMed Issakidis E , Miginiac-Maslow M , Decottignies P , Jacquot J-P , Crétin C , Gadal P . 1992 . Site-directed mutagenesis reveals the involvement of an additional thioredoxin-dependent regulatory site in the activation of recombinant sorghum leaf NADP-malate dehydrogenase . Journal of Biological Chemistry 267 , 21577 – 21583 . Google Scholar PubMed Issakidis E , Saarinen M , Decottignies P , Jacquot J-P , Crétin C , Gadal P , Miginiac-Maslow M . 1994 . Identification and characterization of the second regulatory disulfide bridge of recombinant sorghum leaf NADP-malate dehydrogenase . Journal of Biological Chemistry 269 , 3511 – 3517 . Google Scholar PubMed Jacques S , Ghesquière B , De Bock PJ , Demol H , Wahni K , Willems P , Messens J , Van Breusegem F , Gevaert K . 2015 . Protein methionine sulfoxide dynamics in Arabidopsis thaliana under oxidative stress . Molecular and Cellular Proteomics 14 , 1217 – 1229 . Google Scholar CrossRef Search ADS PubMed Jacquot J-P , Lancelin JM , Meyer Y . 1997 . Thioredoxins: structure and function in plant cells . New Phytologist 136 , 543 – 570 . Google Scholar CrossRef Search ADS Johansson K , Ramaswamy S , Saarinen M , Lemaire-Chamley M , Issakidis-Bourguet E , Miginiac-Maslow M , Eklund H . 1999 . Structural basis for light activation of a chloroplast enzyme: the structure of sorghum NADP-malate dehydrogenase in its oxidized form . Biochemistry 38 , 4319 – 4326 . Google Scholar CrossRef Search ADS PubMed Kabsch W . 2010a . Integration, scaling, space-group assignment and post-refinement . Acta Crystallographica. Section D, Biological Crystallography 66 , 133 – 144 . Google Scholar CrossRef Search ADS Kabsch W . 2010b . XDS . Acta Crystallographica. Section D, Biological Crystallography 66 , 125 – 132 . Google Scholar CrossRef Search ADS Krissinel E , Henrick K . 2007 . Inference of macromolecular assemblies from crystalline state . Journal of Molecular Biology 372 , 774 – 797 . Google Scholar CrossRef Search ADS PubMed Laloi C , Rayapuram N , Chartier Y , Grienenberger JM , Bonnard G , Meyer Y . 2001 . Identification and characterization of a mitochondrial thioredoxin system in plants . Proceedings of the National Academy of Sciences, USA 98 , 14144 – 14149 . Google Scholar CrossRef Search ADS Lemaire SD , Michelet L , Zaffagnini M , Massot V , Issakidis-Bourguet E . 2007 . Thioredoxins in chloroplasts . Current Genetics 51 , 343 – 365 . Google Scholar CrossRef Search ADS PubMed Lindermayr C , Saalbach G , Durner J . 2005 . Proteomic identification of S-nitrosylated proteins in Arabidopsis . Plant Physiology 137 , 921 – 930 . Google Scholar CrossRef Search ADS PubMed McCoy AJ , Grosse-Kunstleve RW , Adams PD , Winn MD , Storoni LC , Read RJ . 2007 . Phaser crystallographic software . Journal of Applied Crystallography 40 , 658 – 674 . Google Scholar CrossRef Search ADS PubMed Meyer Y , Buchanan BB , Vignols F , Reichheld J-P . 2009 . Thioredoxins and glutaredoxins: unifying elements in redox biology . Annual Review of Genetics 43 , 335 – 367 . Google Scholar CrossRef Search ADS PubMed Meyer Y , Belin C , Delorme-Hinoux V , Reichheld J-P , Riondet C . 2012 . Thioredoxin and glutaredoxin systems in plants: molecular mechanisms, crosstalks, and functional significance . Antioxidants and Redox Signaling 17 , 1124 – 1160 . Google Scholar CrossRef Search ADS PubMed Mhamdi A , Hager J , Chaouch S , et al. 2010a . Arabidopsis GLUTATHIONE REDUCTASE1 plays a crucial role in leaf responses to intracellular hydrogen peroxide and in ensuring appropriate gene expression through both salicylic acid and jasmonic acid signaling pathways . Plant Physiology 153 , 1144 – 1160 . Google Scholar CrossRef Search ADS Mhamdi A , Queval G , Chaouch S , Vanderauwera S , Van Breusegem F , Noctor G . 2010b . Catalase function in plants: a focus on Arabidopsis mutants as stress-mimic models . Journal of Experimental Botany 61 , 4197 – 4220 . Google Scholar CrossRef Search ADS Murphy KP , Freire E . 1992 . Thermodynamics of structural stability and cooperative folding behavior in proteins . Advances in Protein Chemistry 43 , 313 – 361 . Google Scholar CrossRef Search ADS PubMed Murshudov GN , Skubák P , Lebedev AA , Pannu NS , Steiner RA , Nicholls RA , Winn MD , Long F , Vagin AA . 2011 . REFMAC5 for the refinement of macromolecular crystal structures . Acta Crystallographica. Section D, Biological Crystallography 67 , 355 – 367 . Google Scholar CrossRef Search ADS PubMed O’Brien R , Haq I . 2004 . Applications of biocalorimetry: binding, stability and enzyme kinetics . In: Ladbury JE , Doyle ML , eds. Biocalorimetry 2: applications of calorimetry in the biological sciences . Chichester, UK : John Wiley & Sons, Ltd , 1 – 34 . Google Scholar CrossRef Search ADS Oger E , Marino D , Guigonis JM , Pauly N , Puppo A . 2012 . Sulfenylated proteins in the Medicago truncatula–Sinorhizobium meliloti symbiosis . Journal of Proteomics 75 , 4102 – 4113 . Google Scholar CrossRef Search ADS PubMed Pyr Dit Ruys S , Wang X , Smith EM , Herinckx G , Hussain N , Rider MH , Vertommen D , Proud CG . 2012 . Identification of autophosphorylation sites in eukaryotic elongation factor-2 kinase . Biochemical Journal 442 , 681 – 692 . Google Scholar CrossRef Search ADS PubMed Queval G , Issakidis-Bourguet E , Hoeberichts FA , Vandorpe M , Gakiere B , Vanacker H , Miginiac-Maslow M , Van Breusegem F , Noctor G . 2007 . Conditional oxidative stress responses in the Arabidopsis photorespiratory mutant cat2 demonstrate that redox state is a key modulator of daylength-dependent gene expression, and define photoperiod as a crucial factor in the regulation of H2O2-induced cell death . The Plant Journal 52 , 640 – 657 . Google Scholar CrossRef Search ADS PubMed Reichheld J-P , Khafif M , Riondet C , Droux M , Bonnard G , Meyer Y . 2007 . Inactivation of thioredoxin reductases reveals a complex interplay between thioredoxin and glutathione pathways in Arabidopsis development . The Plant Cell 19 , 1851 – 1865 . Google Scholar CrossRef Search ADS PubMed Reichheld J-P , Meyer E , Khafif M , Bonnard G , Meyer Y . 2005 . AtNTRB is the major mitochondrial thioredoxin reductase in Arabidopsis thaliana . FEBS Letters 579 , 337 – 342 . Google Scholar CrossRef Search ADS PubMed Ruelland E , Johansson K , Decottignies P , Djukic N , Miginiac-Maslow M . 1998 . The autoinhibition of sorghum NADP malate dehydrogenase is mediated by a C-terminal negative charge . Journal of Biological Chemistry 273 , 33482 – 33488 . Google Scholar CrossRef Search ADS PubMed Scheibe R . 2004 . Malate valves to balance cellular energy supply . Physiologia Plantarum 120 , 21 – 26 . Google Scholar CrossRef Search ADS PubMed Scheibe R , Dietz KJ . 2012 . Reduction–oxidation network for flexible adjustment of cellular metabolism in photoautotrophic cells . Plant, Cell and Environment 35 , 202 – 216 . Google Scholar CrossRef Search ADS Selinski J , König N , Wellmeyer B , Hanke GT , Linke V , Neuhaus HE , Scheibe R . 2014 . The plastid-localized NAD-dependent malate dehydrogenase is crucial for energy homeostasis in developing Arabidopsis thaliana seeds . Molecular Plant 7 , 170 – 186 . Google Scholar CrossRef Search ADS PubMed Suh JJ , Savisky RM . 2011 . Intrinsic fluorescence studies of metal ion binding effects on the thermodynamic stability of human α-lactalbumin . International Journal of Biophysics 1 , 1 – 6 . Google Scholar CrossRef Search ADS Vignols F , Bréhélin C , Surdin-Kerjan Y , Thomas D , Meyer Y . 2005 . A yeast two-hybrid knockout strain to explore thioredoxin-interacting proteins in vivo . Proceedings of the National Academy of Sciences, USA 102 , 16729 – 16734 . Google Scholar CrossRef Search ADS Waszczak C , Akter S , Eeckhout D , et al. 2014 . Sulfenome mining in Arabidopsis thaliana . Proceedings of the National Academy of Sciences, USA 111 , 11545 – 11550 . Google Scholar CrossRef Search ADS Waszczak C , Kerchev PI , Mühlenbock P , et al. 2016 . SHORT-ROOT deficiency alleviates the cell death phenotype of the Arabidopsiscatalase2 mutant under photorespiration-promoting conditions . The Plant Cell 28 , 1844 – 1859 . Google Scholar CrossRef Search ADS PubMed Willems P , Mhamdi A , Simon S , Storme V , Kerchev P , Noctor G , Gevaert K , Van Breusegem F . 2016 . The ROS wheel: refining ROS transcriptional footprints . Plant Physiology 171 , 1720 – 33 . Google Scholar CrossRef Search ADS PubMed Yamazaki D , Motohashi K , Kasama T , Hara Y , Hisabori T . 2004 . Target proteins of the cytosolic thioredoxins in Arabidopsis thaliana . Plant and Cell Physiology 45 , 18 – 27 . Google Scholar CrossRef Search ADS PubMed Yoshida K , Hisabori T . 2016 . Adenine nucleotide-dependent and redox-independent control of mitochondrial malate dehydrogenase activity in Arabidopsis thaliana . Biochimica et Biophysica Acta 1857 , 810 – 818 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2017. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Experimental BotanyOxford University Press

Published: Nov 29, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off