Arabidopsis thaliana (L.) Heynh (ecotype Columbia-0 and Wassilewskija) wild type seeds and knock-out mutants of the GSNOR gene (At5g43940) were obtained from GABI-kat (GABI-Kat 315D11, background Columbia-0) and FLAG T-DNA collections (Versailles Genomic Resource Centre; FLAG_298F11, background Wassilewskija). After vernalisation for 2 days (4°C in dark), plants were cultivated for 4 weeks in a climate chamber at 60% relative humidity under long-day condition (16 h light/8 h dark cycle, 20°C day/18°C night regime, 100 μmol m^-2 s^-1 photon flux density). For seed germination analyses, Arabidopsis seeds were surface sterilized and grown on half strength MS medium containing 1% sucrose in a climate chamber under long-day condition.

Sterile seeds were germinated and grown on half strength MS plates containing different paraquat (methyl viologen, Sigma–Aldrich, Steinheim, Germany) concentrations (0.25–10 μM) for 1 to 2 weeks. To test tyrosine nitration, Western blot was made using anti-nitrotyrosine antibody (Merck Millipore, Darmstadt, Germany) as described (Holzmeister et al., 2015). For spray application, 1, 10, and 50 μM paraquat or water (control) was sprayed onto the leaf surface of 4-week-old plants. Leaves were collected after 1-day of treatment, frozen and kept at -80°C until use. For Western-blot analysis of total protein extract made from paraquat-treated leaves, polyclonal antibody against Arabidopsis GSNOR (Agrisera, Sweden) was used.

Four-week-old plants were placed in an incubator and fumigated with 80 ppm gaseous NO or with synthetic air without NO for 20 min. The experimental setup consisted of controlled-environment cabinets as well as equipment to adjust and control gaseous NO treatment. NO concentration was monitored with a Chemiluminescence Nitrogen Oxides Analyzer AC32M (Ansyco, Karlsruhe, Germany). After fumigation, the plants were placed in the growth chamber until sample collection.

Total RNA isolated from wild type Arabidopsis leaves was used to produce cDNAs by SuperScript II Reverse transcriptase (Invitrogen, Carlsbad, CA, USA). For amplification of coding sequence of AtGSNOR for gateway cloning, the first PCR reaction was made using gene-specific primers (ADH2-ATG-for: 5′-ATGGCGACTCAAGGTCAG-3′; ADH2-TGA-rev: 5′-TCATTTGCTGGTATCGAGGAC-3′). Afterward, the second PCR reaction was performed to introduce recombination sequences (att) at the 5′- and 3′-end using the following primers (ADH2-GW-forward: 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTCATGGCGACTCAAGGTC-3′; ADH2-GW-reverse: 5′- GGGGACCACTTTGTACAAGAAAGCTGGGTCTCATTTGCTGGTATCGAG-3′). The resulting PCR products were cloned into pDONR221 vector (Invitrogen, Carlsbad, CA, USA) by recombination reaction using BP Clonase enzyme mixture according to the instructions of the manufacturer. After sequencing, the correct clone was transferred into the expression vector pDEST17 by recombination using LP Clonase enzyme mixture (Invitrogen, Carlsbad, CA, USA).

For expression of the recombinant N-terminal His₆ fusion proteins the wild type pDEST17-GSNOR and the cysteine mutants GSNOR^C47S, GSNOR^C177S, GSNOR^C271S were transformed into the E. coli strain BL21 DE3 pLysS., The LB cultures at A₆₀₀ ∼0.6 were induced with 0.1 mM isopropyl-β-D-thiogalactopyranoside and further incubate for 4 h at 28°C. After induction the bacterial cells were harvested by centrifugation and stored frozen. For protein isolation the cells were resuspended in a lysis buffer [25 mM Tris-HCl, pH 8.0, 1 mM EDTA, 0.5% Triton X-100, 20% (v/v) glycerol, 1 mM β-mercaptoethanol] and disrupted by sonication. Cellular debris was removed by centrifugation and the soluble fraction was purified by affinity chromatography using Ni-NTA agarose (Qiagen, Hilden, Germany). Adsorbed proteins were eluted from the matrix with elution buffer containing 25 mM Tris-HCl pH 8.0, 250 mM imidazole, 10 mM DTT, 20% (v/v) glycerol. The eluates were aliquoted, frozen in liquid nitrogen and stored at -80 °C until analysis. The purity was checked by SDS-PAGE and the protein concentration was measured by Bradford assay (Bio-Rad). Before the further use the elutions were desalted by Zeba-Spin column (Thermo Scientific, Rockford, IL, USA).

Static light scattering (SLS) experiments on recombinant GSNOR were performed at 30°C using a Viscotek TDA 305 triple array detector (Malvern Instruments) downstream to an Äkta Purifier (GE Healthcare) equipped with an analytical size exclusion column (Superdex 200 10/300 GL, GE Healthcare) at 4°C. GSNOR protein was purified under reducing condition, followed by Äkta purification, than 200 μg GSNOR was oxidized by 1 mM H₂O₂. The reduced and oxidized GSNOR samples were run in 50mM Tris-HCl pH 8.0, 200 mM NaCl, with or without 10 mM DTT, respectively, at a flow rate of 0.5 ml/min. The molecular masses of the samples were calculated from the refractive index and right-angle light-scattering signals using Omnisec (Malvern Instruments). The SLS detector was calibrated with a 4 mg/ml BSA solution with 66.4 kDa for the BSA monomer and a dn/dc value of 0.185 ml/g for all protein samples.

Purified GSNOR protein was re-buffered using Zeba Spin column equilibrated with 20 mM Tris-HCl pH 8.0 buffer. GSNOR activity was determined by measuring the reaction rate of NADH usage at 340 nm in Ultrospec 3100 pro (Amersham Biosciences) spectrophotometer. The reaction buffer contained 20 mM Tris-HCl pH 8.0, 0.5 mM EDTA, 0.2 mM NADH. Serial dilutions (50–2000-fold) of recombinant His-tagged GSNOR proteins (wild type and cysteine mutants) were prepared and the reaction was started to add GSNO (Enzo Life Sciences) at a final concentration of up to 0.5 mM. Water was used instead of GSNO in the reference sample. The reaction was monitored for 5 min and the linear rate was corrected with a reference rate without GSNO. There was no detectable NADH oxidation without enzyme. Specific activity was calculated using a molar extinction coefficient for NADH 6.22 mM^-1cm^-1. Effect of H₂O₂, PN or NEM on GSNOR activity was analyzed by incubation of recombinant GSNOR with these compounds for 20 min in 20 mM Tris-HCl pH 8.0. Excess H₂O₂, PN or NEM was removed by gel filtration using Zeba desalting columns (Thermo Fisher Scientific, Waltham, MA, USA). Zeba Spin columns were equilibrated with 20 mM Tris-HCl pH 8.0. The eluates were used to determine GSNOR activity. To analyze reversibility of H₂O₂-dependent inhibition of GSNOR, excess H₂O₂ was removed (Zeba Spin, 20 mM Tris-HCl pH 8.0) and inhibited GSNOR was divided into two fractions. One fraction was treated with water, the other one was treated with 10 mM DTT for 10 min at room temperature. Before measuring the activity, both samples were desalted using Zeba Spin columns equilibrated with 20 mM Tris-HCl pH 8.0. To analyze the effect of H₂O₂ in presence of excess Zn²⁺, GSNOR was treated with 0.5 mM H₂O₂ in presence of 0.5 μM ZnSO₄ for 20 min. Excess H₂O₂ and ZnSO₄ was remove using Zeba Spin columns (20 mM Tris-HCl pH 8.0) and GSNOR activity was determined.

To measure GSNOR activity from plant tissue, total soluble proteins were extracted from treated seedlings or leaves in buffer of 0.1 M Tris–HCl pH 7.5, 0.1 mM EDTA, 0.2% TritonX-100, 10% glycerol. The homogenate was centrifuged twice at 14.000 g for 20 min at 4°C and total protein concentration of the supernatant was measured according to Bradford using BSA as a standard. GSNOR activity was determined by incubating 100 μg of protein extract in 1 ml reaction buffer as described above.

Glutathione reductase activity assay is based on the NADPH-dependent reduction of GSSG to GSH. GR activity was measured by the rate of NADPH oxidation at 340 nm. Proteins from 2-week-old seedlings were extracted with extraction buffer (50 mM potassium phosphate pH 7.8, 0.1 mM EDTA, 0.5% Triton X-100, 0.5% PVP-40). 50 μg total protein was incubated in 1 ml GR reaction buffer (100 mM potassium phosphate pH 7.8, 1 mM EDTA, 0.2 mM NADPH). The reaction was started by addition of 100 μl of 5 mM GSSG and was monitored at 340 nm for 5 min. The linear rate of reaction was corrected with a reference rate without GSSG (molar extinction coefficient for NADPH 6.22 mM^-1cm^-1).

Glutathione-S-Transferase (GST) activity was measured by spectrophotometrically using the artificial substrate 1-chloro-2,4-dinitrobenzene (CDNB) for GSH attachment. 50 μg of total protein was incubated in 1 ml GST reaction buffer (100 mM potassium phosphate pH 7.8, 1 mM CDNB in 80% ethanol) and the reaction was started by adding 100 μl of 10 mM GSH. The production of GSH-CDNB conjugate was monitored at 340 nm [molar extinction coefficient (𝜀) = 9600 M^-1 cm^-1].

Free Zn²⁺ ions were detected by a metallochromic indicator 4-(2-pyridylazo)-resorcinol (PAR), which binds Zn²⁺ in a 2:1 complex and turn its color from yellow to orange with strong absorbance at 490 nm (Crow et al., 1997). Recombinant GSNOR (50–100 μg) was oxidized by increasing molar excess (100–3000 to protein) of H₂O₂ for 1 h in 50 mM Tris-HCl pH7.2 and mix with 100 μM PAR. The absorbance of PAR₂-Zn complex was measured at 490 nm and the Zn content was calculated using ZnCl₂ standard curve.

Twenty microliter of H₂O₂-treated (molar ratio of H₂O₂ to protein was 100:1 or 1000:1) or water-treated (as control) recombinant GSNOR (corresponding to approximately 8 μg of protein) was incubated with 20 μl of non-reducing SDS-gel loading buffer (20% glycerol, 4% SDS, 0.125 M Tris-HCl pH 6.8) containing 55 mM 2-iodoacetamide (IAA). To remove the excess of IAA completely, samples were separated under non-reducing conditions on SERVAGel TG PRiME 4-12% (SERVA, Heidelberg, Germany). Then the gel was stained with Coomassie overnight and after de-staining the bands were excised and transferred into 1.5 mL reaction tubes. Reduction of cysteine residues was performed for 30 min at 55°C using 45 mM DTT in 50 mM NH₄HCO₃. After removal of DTT solution, blocking was performed using 50 mM S-methyl-methanethiosulfonate (MMTS) for 30 min, then the gel slices were washed three times using 50 mM NH₄HCO₃. The in-gel digestion was performed overnight using 500 ng bovine chymotrypsin (Roche Diagnostics, Mannheim, Germany) or 70 ng trypsin (Promega, Fitchburg, WI, USA). Peptides were separated on NanoLC Ultra chromatography system (Eksigent, Redwood City, CA, USA) coupled to an LTQ Orbitrap XL mass analyzer (Thermo Fisher Scientific, San Jose, CA, USA). Mobile phase A was 0.1% formic acid and mobile phase B was 84% acetonitrile/0.1% formic acid. For separation, a reversed phase nano-column (Reprosil-Pur C18 AQ, 2.4 μm; 150 mm × 75 μm, Dr. Maisch, Ammerbuch-Entringen, Germany) at a flow rate of 280 nl/min was used. The separation method consisted of two linear gradients (1–30% B in 120 min and 30–60% B in 10 min). Mass spectra were acquired in cycles of one MS Orbitrap scan, followed by five data dependent ion trap MS/MS scans (CID, collision energy of 35%). MS spectra were searched using MASCOT 2.4 (Matrix Science, London, UK) using the A. thaliana subset of the SwissProt Database and the following parameters: (a) Variable modifications: Dioxidation (C), Trioxidation (C), Methylthio (C), Carbamidomethyl (C), Oxidation (M); (b) Enzyme: none, (c) Peptide charge: 1+, 2+, and 3+; (d) Peptide tol. ± : 10 ppm; (e) MS/MS tol. ± : 0.8 Da. MASCOT DAT files were imported into the Scaffold software package (Proteome Software Inc., Portland, OR, USA) and filtered for hits with a confidence of 99% at the protein level and 95% for individual peptides.

Total RNA from 4 to 5-week old rosette leaves of Wassilewskija Arabidopsis WT and gsnor was isolated using RNeasy Plant Mini Kit (Qiagen, Germany) according to the manufacturer’s instructions. Quality checking and quantification of RNA isolates were carried out using Agilent RNA 6000 Nano kit on Agilent 2100 BioAnalyzer. Microarray analysis was performed on Agilent platform using the technique “One color Microarray-based Gene Expression Analysis” according to the protocol described in the Agilent manual. The raw expression data of three biological replicates per genotype was analyzed using GeneSpring GX software tool. Statistical analysis were carried out to identify the differentially expressed genes (p < 0.05) between the two genotype using One Way Anova analysis with the Benjamini-Hochberg multiple test correction (FDR) and SNP Post hoc test. From the gene list, those ones regulated at least by twofold differences were selected for downstream analysis. Microarray data are available in the ArrayExpress database¹ under accession number E-MTAB-4756.

The amount of the reduced and oxidized glutathione was determined in 2-week-old seedlings using the GR-based recycling assay described previously (Queval and Noctor, 2007). Furthermore, total glutathione, cysteine and γ-glutamyl-cysteine content were measured by reverse-phase HPLC after NO fumigation experiment. Briefly, 200 mg leaf material was extracted in 0.1 M HCl. Subsequently, all low molecular weight thiols were reduced by addition of DTT and then derivatized with 10 mM monobromobimane as described previously (Wirtz et al., 2004). Samples were analyzed by reverse-phase HPLC and fluorescence excitation at 380 nm.

The amount of the reduced and total ascorbate was determined as described by Queval and Noctor (2007). Four weeks old Arabidopsis plants were sprayed with 0, 1, 10, or 50 μM of paraquat, harvested and stored in -80°C until use. The amount of reduced ascorbate (ASC) was measured at 265 nm in a Tecan plate reader before and after incubation with ASC oxidase. ASC oxidase converts the reduced ASC to the non-absorbing oxidized form. For determination of total ASC, the oxidized ASC was first reduced to ASC by adding 1 mM DTT for 30 min, then total ASC was measured as above. ASC standard was used to calculate the amount of ASC.

For detection of H₂O₂, paraquat or water treated Arabidopsis seedlings were vacuum infiltrated with 0.1% 3,3′-Diaminobenzidine (DAB) in 10 mM MES pH 6.5 solution, washed three times with water and incubated for 45 min at RT in light. After staining, plants were destained with 90% ethanol at 60°C. The brown precipitate shows the presence of H₂O₂ in the cell and tissue.

Arabidopsis plants were vacuum infiltrated with nitroblue tetrazolium (NBT) [50 mM potassium phosphate/pH 6,4; 10 mM NaN₃; 0,1% (w/v) NBT] solution, incubated for 45 min in dark and washed three times with water. Afterward, plants were de-stained with 90% ethanol at 60°C.

Total nitrite, nitrate, and nitrosothiol content were measured using a Sievers 280i nitric oxide analyser (GE Analytical Instruments, Boulder, CO, USA). Proteins were extracted from rosettes using extraction buffer (137 mM NaCl, 0,027 mM KCl, 0,081 mM Na₂HPO₄.2H₂O, 0,018 mM NaH₂PO₄). Leaf protein extract was injected into the purging vessel containing 3.5 ml of acidified KI/I₃^- solution (reducing agent) at 30°C. The recorded mV signals were plotted against a calibration curve produced using known concentrations of sodium nitrite solution to quantify the nitrite level. To estimate the S-nitrosothiol content (RSNO), the above protocol was repeated by pre-treating the leaf protein extract with 20 mM sulphanilamide (in 1 M HCl) at the ratio of 9:1. For nitrate quantification, the reducing agent was replaced with vanadium chloride at 95°C. The recorded mV signals were plotted against a calibration curve produced using known concentrations of sodium nitrate solution to quantify the nitrate levels.

For multiple comparisons, analysis of variance was performed by Anova (one way or two way) followed by Holm-Sidak test. For pairwise comparison, Student’s t-test was used. The level of significance is indicated in each figure.

S-nitrosoglutathione-reductase controls intracellular levels of SNOs and thereby this enzyme is important for NO homeostasis. Paraquat (1,1′-dimethyl-4,4′-bipyridinium dichloride) is an herbicide, which induces the accumulation of ROS, such as superoxide, H₂O₂ and other deleterious oxygen radicals (Dodge, 1971; Babbs et al., 1989) and is a good tool to investigate the effect of ROS on GSNOR activity in vivo. Four week old Arabidopsis plants (Col-0) were sprayed with this herbicide. Paraquat treatment for 24 h significantly decreased GSNOR activity in a dose-dependent manner (Figure 1A). Application of 50 μM paraquat resulted in 40% enzyme inhibition, which could be restored with 10 mM of the reducing agent 1,4-dithiothreitol (DTT) suggesting that oxidative modification(s) are responsible for inhibition of GSNOR activity. Moreover, immunoblot analysis using GSNOR-specific antibody demonstrated only a slight change of GSNOR protein amount during paraquat treatment (Figure 1B). The accumulation after treatment with 50 μM paraquat is around 1.3-fold in comparison to the control sample and might partly compensate for the paraquat-induced inhibition of GSNOR.

To study the physiological function of paraquat/ROS-induced GSNOR inhibition, plants lacking GSNOR function were analyzed for their response to paraquat treatment. Interestingly, Chen et al. (2009) previously observed that GSNOR plays a role in regulating paraquat-induced cell death in plant cells through modulating intracellular NO level. We used two T-DNA insertion alleles for GSNOR (background Col-0 and Wassilewskija, named gsnor and WS/gsnor, respectively) to test their germination and growth in presence of paraquat. Seeds of WT and gsnor plants were cultivated on MS media containing 0–1 μM paraquat and the germination rate was determined by counting fully germinated seedlings with two open cotyledons. The germination rate of WT plants was strongly reduced to 20% in presence of 1 μM paraquat (Figure 1C). In contrast, the germination rate of gsnor plants was significantly higher (72%) at this paraquat concentration (Figure 1C). Similar results could be observed using the T-DNA insertion line in Wassilewskija background (WS/gsnor; Supplementary Figure S1A). Paraquat induced cell death phenotype of WT seedlings was obvious using higher paraquat concentration (1 and 10 μM) by the yellowish-brown colored cotyledons and restricted growth (Figure 1D). Interestingly, gsnor mutant showed an enhanced tolerance even in the presence of 10 μM paraquat demonstrated by green viable seedlings.

Paraquat-induced O₂^- can react with NO resulting in ONOO^- production, which can oxidize cysteine residues or nitrate tyrosine residues of proteins. Protein tyrosine nitration is a marker for pathological processes in cell death. WT seedlings germinated on 0.5 μM paraquat showed stronger tyrosine nitration than gsnor seedlings (Supplementary Figure S1B) suggesting a weaker cell death phenotype for gsnor mutant, which correlates with the observed visible effect of more green seedlings (Figure 1D).

Since GSNOR activity is important for SNO homeostasis, intracellular SNO levels were analyzed after paraquat treatment. WT and gsnor plants were sprayed with 1, 10, and 50 μM paraquat and total cellular SNO content was determined. SNO levels increased in paraquat-treated WT plants from 15 pmol mg^-1 up to 40 pmol mg^-1 (50 μM paraquat) (Figure 2A). The gsnor mutant has already elevated SNO content under control condition (around 40 pmol mg^-1) correlating to the loss of GSNOR function. Paraquat application did not result in significant further accumulation of SNOs in gsnor plants (Figure 2A). In the living cell NO can be oxidized to give nitrite, which would indicate freshly produced NO. Therefore, we measured nitrite content from WT and gsnor plants treated with paraquat. The nitrite level increased in both plant lines in a similar degree in the presence of increasing herbicide concentrations (Figure 2B); however, the nitrite content was significantly higher over the treatment in gsnor plants.

Resistance against oxidative stress is often related to an enhanced cellular antioxidant capacity. Therefore, the amount of reduced and oxidized glutathione (GSH and GSSG, respectively) was measured in WT and gsnor seedlings germinated on media with and without 0.5 μM paraquat for 14 days. GSH level of untreated gsnor plants was about twofold higher than of WT plants and the content increased upon paraquat treatment about four and threefold in WT and gsnor plants, respectively (Figure 3A). About 10% of the glutathione pool was oxidized under control conditions in both plant lines demonstrating that the redox status is the same in WT and gsnor plants (GSH:GSSG ratio, Figure 3A). Germination on paraquat-containing media for 14 days resulted in increased total glutathione content in both plant lines; however, the glutathione pool was more reducing in the gsnor mutant than in WT (GSH:GSSG ratio around 30 and 20, respectively, Figure 3A). This result indicates that gsnor may cope better with a long-term paraquat treatment than WT plants to keep cellular redox condition more reducing.

In correlation with the increased level of glutathione, we measured around 20% higher activity for glutathione reductase (GR) enzyme in untreated gsnor plants in comparison to WT plants (Figure 3B). GR reduces GSSG to GSH in a NADPH-dependent manner. In both lines, GR activity increased by around 30% after paraquat treatment. The glutathione S-transferase (GST) activity was also higher (with 15%) in untreated gsnor plants than in WT and the GST activity increased in both plant lines after paraquat-treatment by around 10% (Figure 3B). Ascorbate is another important cellular reductant inter-connected to the anti-oxidative response in chloroplast. However, although the levels of reduced ascorbate were higher in gsnor plants after paraquat-treatment, the differences were significant only at high paraquat concentrations (Supplementary Figure S2). All these results suggest that the enhanced GSH levels and enhanced activities of GSH-dependent enzymes contribute to the paraquat tolerant phenotype of gsnor plants.

Taken into consideration that loss of GSNOR function results in elevated SNO/NO and GSH levels, which are both important for resistance against paraquat-induced oxidative stress, we analyzed the interplay between both components. Four-week-old WT and gsnor plants were fumigated with 80 ppm NO gas for 20 min to mimic NO burst and levels of low molecular weight thiols of the GSH biosynthesis pathway such as cysteine, γ-glutamylcysteine and total glutathione were determined. NO fumigation resulted in an increase of these compounds in both plant lines (Figure 3C). In addition, untreated gsnor plants had around twofold higher levels of cysteine and glutathione than WT plants, indicating a connection between SNO/NO and the GSH biosynthesis pathway.

To further demonstrate the presence of a pre-induced antioxidant system in gsnor plants, a microarray analysis was performed of 4-week-old rosettes of gsnor and WT plants. Out of 2159 genes, which were differentially regulated, 1407 genes were significantly upregulated and 752 genes were significantly downregulated by at least twofold in gsnor mutant (Supplementary Table S1) (ArrayExpress accession number E-MTAB-4756). Gene enrichment analysis of the regulated genes using VirtualPlant 1.3 platform (Katari et al., 2010) revealed that the most significantly enriched functional categories of the upregulated genes were the catalytic-, hydrolase-, oxidoreductase-, and glutathione transferase-activities (Supplementary Table S2). Among these functional categories, we focused on genes related to processes metabolizing H₂O₂. Within the upregulated genes, 16 genes encoding for peroxidases were identified. Peroxidases are heme-containing enzymes that use hydrogen peroxide as the electron acceptor to catalyze a number of oxidative reactions (Table 1). Moreover, genes encoding for thioredoxins (H-type 7, H-type 8, and atypical ACHT5) and thioredoxin-dependent peroxidases were upregulated in gsnor plants (Table 1). The third subgroup of upregulated genes are encoding for 13 members of the Tau subfamily of GSTs (Table 1). The list of downregulated genes contains only a few transcripts related to redox-regulation, for example one putative peroxidase and a glutathione peroxidase 7 (Table 1). Moreover, four member of the Fe(II) and 2-oxoglutarate-dependent dioxygenase family with an oxidoreductase activity was found to be downregulated. In sum, the transcript profile analysis of gsnor plants suggests a pre-induced antioxidant system under normal growth condition, which can help to defend plants against subsequent oxidative stress.

In line with the transcriptional data, we have analyzed the O₂^- and H₂O₂ level in WT and gsnor plants after paraquat treatment. Ten days old seedlings grown on MS media were treated with 25 μM paraquat and vacuum infiltrated with either nitroblue tetrazolium (NBT) for O₂^- detection (Doke, 1983) or DAB for H₂O₂ accumulation (Thordal-Christensen et al., 1997) (Supplementary Figure S3). No obvious difference in O₂^- accumulation could be observed in paraquat-treated WT and gsnor plants (Supplementary Figure S3A). In contrast, H₂O₂ accumulation was lower in leaves of gsnor plants in comparison to WT plants (Supplementary Figure S3B) suggesting a higher capacity to metabolize H₂O₂ in gsnor line.

To analyze whether H₂O₂ affect Arabidopsis GSNOR activity, recombinant protein was incubated with increasing concentrations of H₂O₂ (0.01, 0.1, and 1 mM). A dose-dependent reduction of the GSNOR activity was observed (Figure 4A). Treatment with 10 μM H₂O₂ caused already 15% inhibition, which further decreased to 35% in the presence of 1 mM H₂O₂. Incubation of inhibited GSNOR with 10 mM of the reducing agent DTT partly restored the activity, concluding that enzyme inhibition was a result of reversible and irreversible modifications of one or several redox-sensitive amino acid residues. Treatment with the sulfhydryl-blocking agent N-ethylmaleimide also inhibited the activity of GSNOR to 40% demonstrating that cysteine residues are important for its activity (Figure 4B). Peroxynitrite (ONOO^-) is formed from the reaction of superoxide (O₂^-) and NO and acts as a potent oxidant on cysteine residues and as a nitrating agent on tyrosine residues of proteins (Arasimowicz-Jelonek and Floryszak-Wieczorek, 2011). Treatment of GSNOR with 10 μM of peroxynitrite reduced the activity by 70% (Figure 4C). Western blot analysis using 3-nitrotyrosine specific antibodies to detect tyrosine nitration of peroxynitrite-treated GSNOR protein did not show any nitration (data not shown) indicates that the peroxynitrite-dependent loss of GSNOR activity is probably due to the oxidation of cysteine(s) (Zeida et al., 2013).

As known from the crystal structure, GSNOR is a homodimer protein. To analyze, whether oxidative conditions cause changes in the native structure of GSNOR SLS experiments with reduced and oxidized recombinant GSNOR were performed. Figure 5 shows that GSNOR protein is detected as dimer under both reducing (10 mM DTT) and oxidizing (1 mM H₂O₂) conditions with no significant difference between both states. The determined molecular weight was 90.1 and 88.4 kDa for reduced and oxidized GSNOR, respectively, which corresponds to the calculated weight of dimer (86 kDa for His₆-GSNOR).

Shifts in the electrophoretic mobility of proteins are diagnostic for the presence of oxidative modifications, like cysteine oxidations or formation of disulfide bridges (Benezra, 1994; Mahoney et al., 1996; Despres et al., 2003). Therefore, we treated recombinant GSNOR with 10–500 μM H₂O₂ and DTT and investigated its running behavior by non-reducing SDS PAGE (Supplementary Figure S4). We could not observe a defined shift in the mobility of the oxidized proteins, but the more diffuse protein bands with increasing concentration of H₂O₂ may indicate the presence of several different oxidative modifications. Subsequent treatment of oxidized GSNOR with 10 mM DTT resulted in a sharp single band demonstrating the reversibility of the oxidative modifications.

The thiol group of cysteine residues is susceptible to oxidation resulting in a formation of sulfenic, sulfinic, or sulfonic acids, the latter two are irreversible modifications. Moreover, disulfide bridges can be formed. Since GSNOR has 15 cysteine residues, we analyzed their oxidative modifications by nano LC-MS/MS spectrometry (Supplementary Figure S5). Recombinant GSNOR was purified under reducing condition, than was oxidized with 500 μM H₂O₂ (equivalent to 100:1 molar ratio of H₂O₂ to GSNOR protein) and with 5 mM H₂O₂. Afterward, free cysteine residues were blocked with iodoacetamide and reversibly modified cysteine residues were reduced with DTT and labeled with S-methyl-methanethiosulfonate (MTHIO). After chymotryptic digestion, the cysteine containing peptides were analyzed for their modifications by nano-LC-MS/MS. Peptides containing Cys94, Cys99, Cys102, and Cys105 could not be detected at all. All other cysteine residues could be identified in its reduced form and the average amount of free thiols (-SH) was 90% in the water treated samples (Figure 6A). Increasing concentrations of H₂O₂ increased the amount of oxidized cysteine residues. We identified both reversibly (MTHIO-labeled) and irreversibly (sulfinic or sulfonic acids, SO_xH) modified cysteine residues (Supplementary Figure S5). According to the increasing concentration of H₂O₂ the average frequency of MTHIO-labeled peptides increased from 17% to around 40% and the SO_xH-modified cysteines from 8 to 55% (Figure 6A). Three cysteine residues (Cys47, Cys177, and Cys271) are located in the substrate-binding site of GSNOR highlighted in the three-dimensional structure of Arabidopsis GSNOR (Figure 6B). Cys47 and Cys177 are coordinating the catalytic Zn²⁺ together with His69 and water molecule. H₂O₂ treatment increased the amount of oxidative modifications for both Zn²⁺-coordinating cysteines (Figure 6C). Around 35% of Cys47 was already oxidized in water-treated GSNOR and the abundance of both MTHIO and SO_xH-modification increased significantly in the presence of 0.5 and 5 mM H₂O₂ (66 and 80%, respectively). This indicates that Cys47 is very sensitive for oxidation. The other Zn²⁺-coordinating cysteine Cys177 also showed increased reversible oxidation up to 55% after 5 mM H₂O₂ treatment. Interestingly, Cys271, which is located in the NAD⁺ cofactor-binding site, was mainly found in its reduced form independent of the treatment. To analyze the importance of these three cysteines, we have generated GSNOR mutants, where these cysteine residues were exchanged by serine resulting GSNOR^C47S, GSNOR^C177S, and GSNOR^C271S. GSNOR^C47S and GSNOR^C177S showed drastically reduced specific activity (100-fold less and 50-fold less, respectively) compared to WT GSNOR (Figure 6D) demonstrating the importance of these two cysteines. In contrast, the mutation of Cys271 resulted in twofold increase of the specific activity.

S-nitrosoglutathione-reductase contains two Zn²⁺ per subunit; one is located in the active center of protein called catalytic Zn²⁺, the other one is structural Zn²⁺. The catalytic Zn²⁺ is coordinated by Cys47 and Cys177 and both are sensitive for oxidation (Figure 6C). Therefore, we tested whether oxidative inhibition of enzyme correlates with Zn²⁺-release. Treatment of recombinant GSNOR protein with H₂O₂ resulted in Zn²⁺-release in a H₂O₂ concentration-dependent manner (Figure 7A). We observed a maximum release of 0.82 Zn²⁺ (±0.16) calculated for one subunit of the dimer with the highest molar excess of H₂O₂ (3000-fold excess corresponded to 20 mM H₂O₂). The activity measurement of the H₂O₂-treated GSNOR showed that the enzyme was completely inhibited by 1000-fold molar excess of H₂O₂ (Figure 7B). This corresponds to a release of 0.4 Zn²⁺/subunit (Figure 7B) suggesting a correlation between activity loss and Zn²⁺-release. We could not measure higher Zn²⁺-release than 0.98 Zn²⁺/subunit indicating that the second Zn²⁺ atom (the structural one) is probably not affected by oxidation. Interestingly, excess of external Zn²⁺ prevented H₂O₂-caused inhibition of GSNOR activity (Supplementary Figure S6), further confirming, that loss of Zn²⁺ results in loss of GSNOR activity.