Methyl viologen-induced changes in the Arabidopsis proteome implicate PATELLIN 4 in oxidative stress responses

Abstract The photosynthesis-induced accumulation of reactive oxygen species in chloroplasts can lead to oxidative stress, triggering changes in protein synthesis, degradation, and the assembly/disassembly of protein complexes. Using shot-gun proteomics, we identified methyl viologen-induced changes in protein abundance in wild-type Arabidopsis and oxidative stress-hypersensitive fsd1-1 and fsd1-2 knockout mutants, which are deficient in IRON SUPEROXIDE DISMUTASE 1 (FSD1). The levels of proteins that are localized in chloroplasts and the cytoplasm were modified in all lines treated with methyl viologen. Compared with the wild-type, fsd1 mutants showed significant changes in metabolic protein and chloroplast chaperone levels, together with increased ratio of cytoplasmic, peroxisomal, and mitochondrial proteins. Different responses in proteins involved in the disassembly of photosystem II–light harvesting chlorophyll a/b binding proteins were observed. Moreover, the abundance of PATELLIN 4, a phospholipid-binding protein enriched in stomatal lineage, was decreased in response to methyl viologen. Reverse genetic studies using patl4 knockout mutants and a PATELLIN 4 complemented line indicate that PATELLIN 4 affects plant responses to oxidative stress by effects on stomatal closure.


Introduction
Plants must adopt an optimal strategy to cope with an unfavorable external environment.The physiological and developmental adaptations require efficient control of transcription; post-transcriptional regulation; protein synthesis, processing, and degradation; post-translational protein modifications; and protein-protein interactions (Huang et al., 2019;Yu et al., 2020;Melicher et al., 2022b).Alteration in the photosynthetic electron transport chain during adverse external conditions leads to an array of complex preventive, protective, and repair mechanisms, including fast protein turnover of photosystem II (PSII) components (Li et al., 2018).Damaged proteins are subject to protein degradation by metalloproteases in chloroplasts and the proteasome complex and to autophagy in the cytosol, and are replaced by de novo synthesis (Xiong et al., 2007).This fast protein recycling preconditions the effective photoprotection and plant tolerance to high light and other stresses that affect the photosynthetic apparatus.
D1 is a PSII core protein showing high-speed turnover rates involving rapid degradation mediated by FtsH metalloproteases, fast protein synthesis, and subsequent incorporation of the newly synthesized D1 into PSII (Inagaki, 2022).Other proteins showing degradation rates similar to D1 include protein POST-ILLUMINATION CHLOROPHYLL FLUORESCENCE INCREASE (PIFI; an auxiliary subunit of the NAD(P)H DEHYDROGENASE (NDH) complex; Wang and Portis, 2007;Li et al., 2017), and CYTOCHROME b 6 f (Cyt b 6 f) COMPLEX SUBUNIT 4 (PetD), a scaffold protein and plastoquinone binding site component of the Cytochrome b 6 f complex (Cramer and Zhang, 2006;Li et al., 2017).Prolonged photoinhibition also alters protein profiles in other cellular compartments, including cytosol, mitochondria, and peroxisomes (Taylor et al., 2009;Melicher et al., 2022a).The signal generated by altered redox homeostasis in chloroplasts is transduced to other organelles and compartments by diverse biomolecules such as H 2 O 2 , tetrapyrrols, oxidized carotenoids, and lipids during retrograde signaling (Sierla et al., 2013;Gollan et al., 2015;Chan et al., 2016;Exposito-Rodriguez et al., 2017) or by redox sensing hubs such as thioredoxins (Dietz and Hell, 2015).
Proteomic and redox proteomic analyses have brought novel insight into the dynamics of protein abundance during photoinhibition or methyl viologen (MV)-induced inhibition of the photosynthetic electron transport chain.Exposure of Chlamydomonas reinhardtii cell culture to 0.66 µM MV for 6 h affects the abundance of PSI and PSII subunits, light-harvesting complex proteins, enzymes of the Calvin cycle, and antioxidant enzymes (Nestler et al., 2012).Another highly affected group of proteins is involved in protein turnover, mostly in protein synthesis and degradation (Nestler et al., 2012).In Chlorella vulgaris, 72 h of exposure to MV leads to a significant decrease in PsaB and RbcL transcript and protein abundance (Qian et al., 2009).Our recent comparative proteomics study of wild-type (WT) and fsd1 mutants deficient in IRON SUPEROXIDE DISMUTASE 1 (FSD1) showed that 8 h of MV treatment affected the proteins containing Fe-S clusters as well as components of both photosystems.Fe-S cluster proteins and PSI components were more affected in fsd1 mutants than WT, showing the importance of antioxidant defense for maintaining proteome homeostasis (Melicher et al., 2022a).
The early response of plants to MV treatment is linked to changes in the profile of oxidation-sensitive proteins.Thus, antioxidant and redox buffering enzymes and regulators of PSII oxygen-evolving complex (OEC) such as PsbO-1, PsbP-1, and PsbQ were identified after 30 min of MV treatment in Arabidopsis (Muthuramalingam et al., 2013).However, short, 30 min MV treatment may lead also to modulation of protein abundance.For example, cytosolic ASCORBATE PEROXIDASE (cAPX) significantly increased in abundance in WT, but not in fsd1 mutant (Melicher et al., 2022a).
Recently, we characterized the effects of MV on Arabidopsis WT and fsd1 mutants (Melicher et al., 2022a).MV caused an immediate decrease in PSII activity, which was more pronounced in two fsd1 mutants.The ascorbate concentration and the level of protein carbonylation were similar to the control after 1 h of MV treatment of WT and mutant plants.Moreover, overactivation of FSD1 and APX was encountered after 30 min MV treatment in Arabidopsis WT, but was not observed in fsd1 mutants.Thus, short-term MV treatment alters the PSII activity and antioxidant capacity in WT plants (Melicher et al., 2022a).
Early proteome remodeling has not yet been studied in response to MV in plants.Therefore, in this study, we attempted to examine early proteome remodeling associated with MV-induced reactive oxygen species (ROS) accumulation in WT and oxidative stress-sensitive fsd1 mutants.Our analyses revealed that chloroplastic proteins showed the most prominent sensitivity to MV in WT and fsd1 mutants.The FSD1 deficiency resulted in proteome remodeling by vigorous change in abundance in metabolic proteins, chloroplastic chaperones, and heat shock proteins.PATELLIN 4 (PATL4), a phospholipidbinding protein localized to plasma membrane, was detected as a protein quickly responding to MV treatment and affecting the oxidative stress response of Arabidopsis.Microscopic analyses showed that PATL4 is involved in the regulation of stomatal aperture during oxidative stress.

Proteome investigation
For proteomic analysis, 12-day-old WT, fsd1-1, and fsd1-2 plants were incubated for 30 min in liquid ½ MS medium supplemented with 1 µM MV under 180 μmol m -2 s -1 illumination.For mock control, plants were incubated in liquid ½ MS medium without MV under the same cultivation conditions.The analysis was performed in four biological replicates, while five whole plants were pooled into one replicate.
Protein extraction, in-solution trypsin digestion, and peptide purification were performed as described in Takáč et al. (2017).Briefly, proteins were extracted using phenol followed by protein precipitation in ammonium acetate-methanol.Proteins were dissolved in 6 M urea and digested in-solution using trypsin.Peptides were pre-cleaned on C18 gravitational cartridges (Bond Elut C18; Agilent Technologies, Santa Clara, CA, USA).Nano-LC-MS/MS, protein identification, and protein label-free quantification were performed as in Takáč et al. (2021) with modification-intact peptides were measured in the Orbitrap mass detector of the LTQ-Orbitrap hybrid mass spectrometer (Thermo Scientific), and consequently the precursor mass tolerance parameter was narrowed to 10 ppm, when the raw mass spectrometry data were interrogated by Proteome Discoverer 2.1 software (Thermo Scientific).
The differentially abundant proteins were selected based on the criteria ANOVA P < 0.05 and fold change >1.5.Proteins present in at least three out of four replicates corresponding to the control proteome and absent in all four replicates of the test proteome were considered as unique for the control proteome, and vice versa.

Microscopy analysis
To observe the effects of MV on green fluorescent protein (GFP)-PATL4 localization, 4-day-old seedlings were transferred to a microscope slide (containing a spacer from double-sided sticky tape) into liquid ½ MS medium and covered with a coverslip.After initial microscopic observation, plants were treated with 1 µM MV (diluted in ½ MS medium) by perfusion directly at the microscope stage, incubated for 30 min in light (180 μmol m -2 s -1 ) or dark (mock control), and documented.Image acquisition was performed using a confocal laser scanning microscope LSM 710 (Carl Zeiss, Jena, Germany) with a Plan-Apochromat ×20/0.8M27 or alpha Plan-Apochromat ×63/1.46Oil Korr M27 objective.GFP signal was excited by a 488 nm excitation laser and detected using BP420-480+BP495-550 emission filters.The image acquisition, post-processing, and fluorescence intensity measurement were performed using Zeiss ZEN software (Black and Blue versions, Carl Zeiss).Semi-quantitative evaluation of GFP-PATL4 fluorescence intensity was performed using profile measurement of petiole and leaf epidermal plasma membranes.Data shown as relative fluorescence intensities are obtained from petiole and leaf epidermal cells from five to six individual plants per treatment.Statistical significance was evaluated using Student's t-test.Images of GFP-PATL4 epidermal cells of developing first leaves were processed as orthogonal projections of acquired Z-stacks.

Stomatal aperture and stomatal size measurement
Cotyledons of 4-day-old WT, patl4-1, and patl4-2 were used for stomatal aperture analysis under MV-induced oxidative stress.At first, seedlings were incubated for 30 min in liquid ½ MS medium supplemented with 1 µM MV or without MV as mock control and incubated under 180 μmol m -2 s -1 illumination.Afterwards, plants were transferred to a microscope slide and into liquid ½ MS medium, covered with a coverslip, and immediately documented using a confocal laser scanning microscope LSM 710 (Carl Zeiss, Jena, Germany) with Plan-Apochromat ×40/1.4Oil DIC M27 objective in bright field conditions.
To measure stomatal size, cotyledons of 4-day-old WT, patl4-1, and patl4-2 seedlings were incubated in fixative chlorophyll-removing solution (ethanol, acetic acid and glycerol in ratio 3:1:1), which was replaced until the cotyledons were completely discolored.Afterwards, cotyledons were washed in a mixture of ethanol and glycerol (4:1), loaded on a glass slide in a drop of glycerol, covered with a coverslip and analysed with an epifluorescence microscope (Axio Imager.M2) using an EC Plan-Neofluar ×40/0.75M27 objective in bright field conditions.
The image acquisition, post-processing, stomatal aperture, and stomatal size measurements were performed using Zeiss ZEN software (Black and Blue versions).Stomatal aperture was measured as a ratio of aperture width and length, which is reduced upon closure.Stomatal size was measured as stomatal complex (guard cell pair) length multiplied by its width (Drake et al., 2013).The statistical significance was evaluated by one-way ANOVA with post-hoc Tukey honestly significant difference (HSD) test available at an online web statistical calculator (https://astatsa.com/OneWay_Anova_with_TukeyHSD/).

Immunoblotting analysis
The abundance of catalase in response to MV treatment was tested in WT, fsd1-1, and fsd1-2 mutants grown and treated as for proteomic analysis.Frozen powdered plant material (100 mg) was resuspended and homogenized with 200 μl extraction buffer (50 mM HEPES, pH 7.5, 75 mM NaCl, 1 mM EGTA, 1 mM MgCl 2 , 1 mM NaF, 1 mM dithiothreitol, 10% glycerol, and Complete™ EDTA-free Protease Inhibitor Cocktail, PhosSTOP™ (Roche Diagnostics, Mannheim, Germany).After 30 min incubation on ice, the extract was centrifuged at 13 000 g for 20 min at 4°C.The immunoblotting analysis was carried out using protein extract with 20 µg of total protein as described in Melicher et al. (2022a).Catalase (CAT) abundance was detected using anti-CAT primary antibody (AS09 501; Agrisera, Sweden, diluted 1:1000), recognizing all three catalase isoforms in Arabidopsis.The protein abundance was expressed as an average of band optical densities from three biological replicates.The statistical significance was evaluated by one-way ANOVA with post-hoc Tukey HSD test.

Phenotype and chlorophyll content analysis
The fresh weight of 15-19 seedlings of 14-day-old patl4 mutants grown in vitro was recorded per line and replicate.Three biological replicates were performed and the statistical significance was evaluated by one-way ANOVA with post-hoc Tukey HSD test.Seven-and 11-day-old plants were documented for comparison using a flat scanner (ImageScanner III, GE Healthcare, UK).
To examine plant phenotypic responses to MV, 5-day-old seedlings of WT, fsd1-1, patl4 mutants, and the GFP-PATL4 line growing on ½ MS medium were transferred to ½ MS medium with or without 2 µM MV.All measurements were performed in three repetitions.Seedlings were documented on the seventh day after the transfer, and the tolerance to MV was expressed as the percentage of green plants from 50 seedlings per replicate (150 in total).In addition, the primary root length of 30 MV-treated seedlings per replicate (90 in total) and 20 control seedlings per replicate (60 in total) was measured using ImageJ software (Schneider et al., 2012) on the seventh day after the seedling transfer.Statistical significance was evaluated by one-way ANOVA with post-hoc Tukey HSD test.
The relative amount of chlorophyll a and b was measured according to Barnes et al. (1992).The concentrations of chlorophyll a and b were measured using 30 seedlings for each line and the measurements were expressed per fresh weight of examined seedlings performed in triplicate (90 seedlings examined in total).The statistical significance was evaluated by one-way ANOVA with post-hoc Tukey HSD test.

Comparative shot-gun proteomic analysis
In this study, shot-gun quantitative proteomics revealed that 9, 11, and 19 proteins showed statistically significant changes in abundance in MV-treated as compared with mock-treated WT, fsd1-1, and fsd1-2 plants, respectively (Supplementary Tables S1-S3).Several other proteins showed unique occurrences in either mock-or MV-treated samples.We consider proteins identified only in mock-treated controls as downregulated in MV-treated samples, as they were below the detection threshold of the instrument.In accordance, proteins identified in MV-treated samples but not in mock-treated ones are considered up-regulated in MV-treated plants.Therefore, the uniquely identified proteins (14, 7, and 33 proteins in WT, fsd1-1, and fsd1-2 mutant, respectively) were evaluated together with the quantitatively identified differentially abundant proteins (Supplementary Tables S1-S3).
More than half of the MV-affected proteins were localized to the chloroplast in all three lines (Fig. 1).The number of cytoplasmic proteins was higher in the mutants compared with WT.Unlike mutants, MV affected two plasma membrane-localized proteins (PATL1 and PATL4) and two apoplastic peroxidases in WT.On the other hand, MV caused deregulation of mitochondrial and peroxisomal proteins in the mutants but not in WT plants (Figs 1, 2).
The profile of differentially regulated chloroplastic proteins differed in mutants and WT (Fig. 2).In WT plants, MV dramatically affected the PsbP-LIKE PROTEIN 1 (PPL-1), an extrinsic subunit of PSII and a PHOTOSYSTEM II FAMILY PROTEIN Psb27 localized in the thylakoid lumen.While PPL-1 showed substantial, 6.5-fold up-regulation, the abundance of Psb47 was considerably down-regulated (0.11-fold; Table 1).MV also negatively affected the LIGHT-HARVESTING COMPLEX PHOTOSYSTEM II Lhcb4.2, which was detected only in mock-treated WT samples.A component of the chloroplastic NDH complex, a PsbP-LIKE PROTEIN 2 (PPL-2) was uniquely found only in MV-treated WT plants.
Opposite to WT, Lhcb4.2 was up-regulated in fsd1-2 mutant.MV also disturbed the abundance of PSII components in this mutant, including PHOTOSYSTEM II SUBUNIT P-1 (0.30-fold down-regulated) and PHOTOSYSTEM II 44 kDa PROTEIN (CP43), showing full homology to PHOTOSYSTEM II CORE ANTENNA PROTEIN C (ATCG00280; uniquely found in the MV-treated fsd1-2 mutant).In addition to PSII, the fsd1-2 mutant exhibited changes in PSI subunits.According to quantitative evaluation, PHOTOSYSTEM I SUBUNIT F (PSI-F) was 2-fold up-regulated in the fsd1-2 mutant and PHOTOSYSTEM I SUBUNIT H-1 (PSI-H1) was uniquely found in fsd1-2 mocktreated sample.The fsd1-1 did not show significant changes in photosystem subunit abundances.
We also compared the abundances of proteins between MV-treated mutants with MV-treated WT plants (Supplementary Tables S4, S5).Both mutants had altered abundances of two PSI subunits.While PSI-N showed lower abundance in the mutants, the abundance of PSI-H1 was higher.PSII light harvesting complex protein B1B2 showed lower abundance in the fsd1-1 mutant.
The comparison of MV-treated mutants and WT plants showed altered abundances of proteins involved in chloroplast protein folding in the mutants (Supplementary Tables S4, S5).While CHAPERONIN 20 showed lower abundance in both fsd1 mutants, ATP-DEPENDENT CASEINOLYTIC (CLP) PROTEASE/CROTONASE FAMILY PROTEIN was more abundant in fsd1-1, and ROTAMASE CYP 4 in fsd1-2.TRIGGER FACTOR TYPE CHAPERONE FAMILY PROTEIN showed lower abundance in fsd1-2 mutant.The fsd1-1 mutant had lower abundance of TRANSLOCON AT THE OUTER ENVELOPE MEMBRANE OF CHLOROPLASTS 159 protein compared with WT.In addition, both mutants showed altered abundances of plastid ribosomal proteins (Supplementary Tables S4, S5).
Abundances of proteins involved in protein folding in cytosol were also altered by MV (Fig. 2).HSP 81-2, also known as HSP90.2,localized in cytosol, was down-regulated in WT plants, while the opposite regulation was found for another HSP isoform, HSP 81-3 (HSP 90.3) in fsd1-2 mutant.After the MV treatment, the levels of cytoplasmic proteins involved in protein folding did not show significant differences between the mutants and WT (Supplementary Tables S4, S5).
MV treatment disturbed the abundance of proteins involved in ROS decomposition in fsd1-2 mutant (Fig. 2), showing up-regulation of cytosolic MONODEHYDROASCORBATE REDUCTASE 1 (2.75-fold) and CATALASE 2 (2.5-fold) and down-regulation of APX1 (0.33-fold).In agreement, immunoblotting analysis using anti-CAT primary antibody recognizing all three Arabidopsis CAT isoforms showed an increase of total CAT abundance in fsd1 mutants, while  a decrease was observed in WT plants (Fig. 3).In WT, we encountered a higher abundance of NADPH-DEPENDENT THIOREDOXIN REDUCTASE C. Nevertheless, the levels of antioxidant enzymes did not differ significantly between fsd1 mutants and WT after MV treatment (Supplementary Tables S4, S5).
Some other stress-related proteins showed unique deregulation in WT (Fig. 2).Notably, we observed down-regulation of phospholipid binding, plasma membrane-localized PATL1 and PATL4 after MV treatment of WT plants.MV also affected the abundance of cytosolic DEHYDRIN FAMILY PROTEIN (ERD14), showing substantial 4.7-fold up-regulation, and a defense-related protein PYK10-BINDING-PROTEIN 1 localized to endoplasmic reticulum bodies (down-regulated).A nuclear protein, HISTONE SUPERFAMILY PROTEIN (putative H4-TYPE HISTONE), was dramatically up-regulated after MV treatment in WT.
Striking differences between WT and the mutants were found in metabolic proteins (Fig. 2).While negligible effects were found in WT, the metabolic proteins showed substantial abundance changes in both mutants.MV increased the abundance of proteins involved in amino acid biosynthesis, photorespiration, carbohydrate degradation, and detoxification of reactive carbonyls and methylglyoxal.In addition, abundances of proteins involved in mitochondrial electron transport chain, glycolysis, and energy homeostasis were also disturbed (Table 2).After MV treatment, the mutants differed from WT mainly in proteins involved in glycolysis/gluconeogenesis and amino acid biosynthesis (Supplementary Tables S4, S5).

PATL4 abundance drops shortly after methyl viologen treatment in Arabidopsis wild-type
Proteomic analysis indicated down-regulation of PATL4 abundance in response to short-term MV treatment in WT.Patellins are plasma membrane-localized proteins.Their possible involvement in plant oxidative stress has not been described yet.Therefore, the response of PATL4 to MV was evaluated by live-cell confocal laser scanning microscopy in patl4-1 mutants expressing the proPATL4::GFP:PATL4 construct.
We monitored the GFP-PATL4 fluorescence in epidermal cells of petioles and first true leaves.We found a decrease in GFP signal intensity after 30 min MV treatment at the plasma membranes of petiole and leaf epidermal cells (Fig. 4).Control MV treatment carried out in the dark did not affect GFP-PATL4 signal intensity (Supplementary Fig. S2).
As expected, fsd1-1 mutant showed hypersensitivity to MV treatment, exemplified by almost completely bleached cotyledons and the lowest chlorophyll content (Fig. 5A-C).It also showed the highest levels of root growth inhibition (Fig. 5D).On the other hand, both patl4 mutants show the highest tolerance to MV, indicated by the highest ratio of fully green cotyledons, chlorophyll content (Fig. 5A-C), and the highest increment of root length (Fig. 5D).WT plants and the GFP-PATL4 line showed intermediate phenotypes (Fig. 5A, B) and chlorophyll content (Fig. 5C), supporting that PATL4 complementation reversed phenotype of patl4 mutant.Under control conditions, the lines did not exhibit significant differences in root growth rate and primary root length (Fig. 5E; Supplementary Fig. S3A, B).However, patl4 mutants exhibited significantly higher fresh weight compared with WT, apparently due to the bigger size of leaf rosettes (Supplementary Fig. S3B, C).These data indicate that PATL4 affects the plant tolerance to MV-induced oxidative stress.

Measurement of stomata size and aperture after methyl viologen treatment
The prevalence of GFP-PATL4 in plasma membranes of stomatal precursors and stomata indicated its possible role in stomatal development and/or movement.To examine whether PATL4 may be involved in stomatal movement, we measured stomatal aperture of patl4 mutants under MV-induced oxidative stress.We found that MV caused more pronounced closure of stomata in patl4 mutants as confirmed by lower stomatal aperture of patl4 mutants (Fig. 6A, B).Moreover, the size of stomata in the mutants was significantly larger than in WT (Fig. 6A, C).Thus, PATL4 may have a regulatory role in stomatal development and movement during oxidative stress.

Discussion
Alterations of protein abundance detected by shot-gun proteomic analysis after stress treatments that last in the range of minutes require careful interpretation.The average protein half-life is estimated to be approximately 3 d and may differ under stress conditions (Nelson et al., 2014).Therefore, changes in abundance may reflect protein synthesis or degradation dynamics, as was most likely in the case of PHOTOSYSTEM II CP43 REACTION CENTER PROTEIN (up-regulated in fsd1-2 mutant), which undergoes fast repair during photodamage (Kato and Sakamoto, 2009).Alternatively, proteins displaying differential regulation may disassemble from complexes (such as Lhcb4.2;de Bianchi et al., 2011) or release from membranes, causing better accessibility to protein digestion during sample preparation.Many of the identified differentially regulated proteins were previously shown to be involved in photoprotection (Table 1), confirming that compositions of differential proteomes are relevant to oxidative stress generated in chloroplasts.Our approach provides an opportunity to get information on the abundance of proteins localized in diverse subcellular compartments and correlate them with various biochemical pathways.These proteins are likely involved in early molecular mechanisms of plant responses to oxidative stress generated in chloroplasts, while proteins identified in fsd1 mutants are likely associated with plant hypersensitivity to oxidative stress.Notably, the phenol extraction coupled to methanol precipitation used in this study can extract the hydrophobic proteins embedded in the membranes with lower efficiency.

fsd1 mutants differ from wild-type in photosystem II-light harvesting complex disassembly, chloroplastic protein folding, and metabolism
The different susceptibility of WT and the fsd1 mutants to MV-induced oxidative stress was reflected in the composition and the intraorganellar localization of differentially regulated chloroplastic proteins 2).The abundance of Lhcb4.2, a component of the light-harvesting complex of photosystem II, was down-regulated in WT but up-regulated in the fsd1-2 mutant.This protein is essential for PSII macro-organization and photoprotection (de Bianchi et al., 2011).Its phosphorylation by SERINE/THREONINE-PROTEIN KINASE STN7 in response to high light is required for PSII-light harvesting complex (LHC) disassembly (Fristedt and Vener, 2011), an integral step of the PSII repair mechanism (Nath et al., 2013).Furthermore, chloroplast SIGNAL RECOGNITION PARTICLE 54 (cpSRP54) is involved in sorting proteins of the LHC family, including Lhcb4.2, to the thylakoid membrane (Rutschow et al., 2008).cpSRP54 was found to be  PPL-2 (up-regulated;Ishihara et al., 2007;Ifuku et al., 2011) and Psb27 (down-regulated;Chen et al., 2006;Wei et al., 2010).Notably, PPL-1 (Che et al., 2020) shows over 6-fold increased abundance compared with the mock control.This unprecedented accumulation points to the importance of PPL-1 for the plant's early response to MV-induced oxidative stress.OUTER PLASTID ENVELOPE PROTEIN 16-1, also known as NADPH PROTOCHLOROPHYLLIDE OXIDOREDUCTASE A (PORA) translocation pore, is down-regulated in WT and prevents singlet oxygen production (Buhr et al., 2008).In addition, WT also showed down-regulation of VAR2, a protease involved in D1 degradation and repair cycle (Takechi et al., 2000;Kato et al., 2009).This VAR2 forms a complex with VAR1, both contributing to D1 protein degradation and to the PSII repair cycle (Sakamoto et al., 2003).Hence, the differential proteomes of fsd1 mutants contained some other photoprotective proteins including PHOTOSYSTEM II SUBUNIT P-1, which was down-regulated in fsd1-2 mutant and is required for normal thylakoid architecture in Arabidopsis (Yi et al., 2009), and THYLAKOID LUMEN 18.3 kDa PROTEIN involved in PSII repair cycle (Sirpiö et al., 2007;Järvi et al., 2016).CP43 undergoes repair during photodamage (Nelson et al., 2014) and regulates the accessibility of D1 protein to proteolysis by FtsH proteases (Krynická et al., 2015).The abundance of this protein was up-regulated after MV treatment in fsd1-2 mutant.In summary, these results indicate that FSD1 deficiency is associated with defects in LHC disassembly, but the processes connected to PSII photorepair show a pattern similar to WT.
Unlike in WT, two PSI components, PSI-F and PSI-H1, were up-regulated in fsd1 mutants (Fig. 2).PSI-F is essential for photoautotrophic growth and contributes to antenna function (Haldrup et al., 2000).This indicates an early alteration of PSI subunits in the sensitive mutant line.Interestingly, PSI subunits were affected also after longer MV treatment in the mutants (Melicher et al., 2022a), further supporting the importance of FSD1 for PSI integrity.
Remarkably, the abundances of chloroplastic chaperones, such as CASEIN LYTIC PROTEINASE B3 (Parcerisa et al., 2020), CYCLIN DELTA-3, CHAPERONIN-60 ALPHA, and chloroplastic HSP70 (Kessler and Schnell, 2009), were changed only in fsd1 mutants.CYCLIN DELTA-3 functions downstream of CHAPERONIN-60 in the assembly of chloroplast ATP SYNTHASE COUPLING FACTOR 1 (Mao et al., 2015).These results suggest that modulations in abundance of chloroplastic chaperones represent an early response to MV-induced oxidative stress in the mutant hypersensitive lines.
FSD1 deficiency caused modulation of metabolic protein abundance in chloroplasts, cytoplasm, mitochondria, and peroxisomes already after 30 min of MV treatment (Fig. 2).This implies that missing FSD1 and the associated deregulation of photosynthesis resulted in the transduction of signals to metabolic pathways occurring in other subcellular locations.The redox state of the chloroplast is transduced to the cellular environment via three major redox regulatory hubs, including the ferredoxin-thioredoxin system, the NADPH-NADPH THIOREDOXIN REDUCTASE C (NTRC) system and the glutathione-glutaredoxin system (Dietz and Hell, 2015).Of these systems, NTRC showed an increased abundance in WT.Given that the abundance of metabolic proteins remained almost unaltered in WT, we hypothesize that the change in NTRC abundance contributed to retaining the metabolic homeostasis in WT.
On the other hand, the altered metabolic homeostasis in fsd1 mutants might arise from a dramatic reduction in the abundance of cytosolic NAD-DEPENDENT MALATE DEHYDROGENASE 1 (MDH1).MDHs and malate/oxaloacetate translocators enable an indirect transfer of reducing equivalents between different subcellular compartments in plant cells (Selinski and Scheibe, 2019).MDH1 is active under reducing conditions, and its homodimerization through Cys330 disulfide formation protects the protein from overoxidation (Huang et al., 2018).In summary, provide a specific set of chloroplastic and proteins showing fast response to MV-induced oxidative stress in plants.

PATL is involved in plant response to methyl viologeninduced oxidative stress
PATLs are phosphoinositide binding/transfer proteins containing a SEC14 (lipid binding/transfer) and Golgi dynamics (GOLD) domain.The Arabidopsis PATL gene family consists of six isoforms (PATL1-6).PATL proteins exert a high degree of homology in their amino acid sequence, but with variable N-terminus.PATL1, PATL2, and PATL4 show high levels of similarity, because their N-terminus contains numerous EEK repeats (reminiscent of those found in the neurofilament triplet H proteins) and a coiled-coil, a common protein oligomerization/folding motif (Peterman et al., 2004).PATLs are localized at the plasma membrane and cell plate during cytokinesis (Peterman et al., 2004;Tejos et al., 2018).PATL1 is also localized in brefeldin A-sensitive endosomal compartments (Zhou et al., 2018), indicating its participation in the vesicular trafficking pathway.PATLs, including PATL4, are expressed in meristematic cells, including root apical meristem, lateral root primordia and cells entering a differentiation program (Tejos et al., 2018).In cotyledons, PATL4 localizes mainly to precursors of stomata and subcellularly to plasma membrane (Tejos et al., 2018).
PATLs have diverse functions.For example, PATL1 interacts with SALT-OVERLY-SENSITIVE 1 (SOS1) and controls its Na + /H + antiporter activity, thus negatively affecting Arabidopsis salt stress tolerance (Zhou et al., 2018).PATL1 also regulates plant freezing tolerance by interaction with Arabidopsis CALMODULIN 4 (Chu et al., 2018). is phosphorylated by MITOGEN-ACTIVATED PROTEIN KINASE 4 and contributes to the membrane regeneration during cell plate formation (Suzuki et al., 2016).Additionally, PATL2 affects Fe 2+ acquisition responses by interaction with IRON-REGULATED TRANSPORTER1 (IRT1).Arabidopsis patl2 knockout mutants show elevated lipid peroxidation and reduced levels of α-tocopherol, and PATL2 binds α-tocopherol through its Sec14 domain in vitro.It was suggested that PATL2 might present α-tocopherol at the membrane close to IRT1 and protect the membrane from detrimental Fe 2+ effects.PATL2, in addition, may bind several proteins responsive to oxidative stress, and thus it might attenuate cellular ROS stress or contribute to ROS signaling at the plasma membrane of root cells (Hornbergs et al., 2023).
Involvement of PATLs related to plant oxidative stress in chloroplasts has not been studied so far.Here we detected down-regulation of PATL4 preferentially enriched in plasma membrane of stomata precursors, after a short-term MV treatment, indicating its high sensitivity to ROS generated in chloroplasts (Fig. 4).The absence of PATL4 expression in mutants confers their higher tolerance to oxidative stress caused by MV (Fig. 5).It is known that MV negatively affects stomatal conductance and transpiration rate leading to intercellular CO 2 accumulation (Caverzan et al., 2014;Ozfidan-Konakci et al., 2023).Elevated CO 2 and its conversion to bicarbonate (HCO 3 -) leads to induction of anion efflux through the guard cell plasma membrane, positively affecting stomatal closure, mediated by Ca 2+ signaling (McAinsh et al., 1996;Ehonen et al., 2019).PLASMA MEMBRANE INTRINSIC PROTEIN 2;1 (PIP2;1), a potential interaction partner of PATL4 (Bellati et al.,

Fig. 2 .
Fig. 2. Schematic representation of the differential proteomes of Col-0 wild-type (A) and fsd1-1 and fsd1-2 mutant (B) plants exposed to 30 min methyl viologen treatment.Up-regulated and down-regulated proteins are depicted as red and blue circles, respectively.In (B), the abbreviations of differentially regulated proteins found in fsd1-1 mutant are shown in black and those found in fsd1-2 mutant are shown in green.Proteins with similar functions

Fig. 3 .
Fig. 3. Immunoblotting analysis of CATALASE (CAT) in Col-0 wild-type (WT) plants, fsd1-1, and fsd1-2 mutants subjected to 30 min methyl viologen (MV) treatment.(A) Immunoblot of CAT supplemented with respective control of protein loading visualized by Ponceau S staining.(B) Quantification of band optical densities in (A).Lane C, blot containing protein extracts from mock; lane MV, MV-treated plants.Raw band optical densities were normalized according to the total density of the specific bands on the membrane (mean ±SD, n=3).Asterisks indicate a statistically significant difference between mock control and MV treatment as revealed by one-way ANOVA with post-hoc Tukey HSD test (P<0.05).Uncropped, full original images of blot and Ponceau S stained membrane are provided in Supplementary Fig. S1.

Fig. 4 .
Fig. 4. GFP-PTL4 localization in epidermal cells of Arabidopsis petiole and leaf after methyl viologen (MV) treatment revealed by confocal laser scanning microscopy.(A) Plasma membrane localization of GFP-PTL4 in epidermal cells of petiole and first true leaf before and after 30 min treatment with MV in the light.Stomata are indicated by arrows and meristemoids by stars.(B) Semiquantitative analysis of plasma membrane GFP-PTL4 fluorescence intensity of petiole and leaf epidermal cells before and after 30 min treatment with MV in the light.Asterisks indicate statistical significance between treated and not treated cells (*P<0.05).Error bars show ±standard deviation.Scale bar, 10 µm.

Fig. 5 .
Fig. 5. Phenotypic response of Col-0 wild-type (WT), fsd1-1 mutant, patl4 mutants, and GFP-PATL4 complemented line to 2 μM methyl viologen (MV).(A) Representative images of seedlings documented on seventh day after the transfer to medium with MV. (B) Quantification of fully green seedlings from all examined seedlings (mean ±SD; 50 seedlings per n; n=3).(C) Quantification of relative amount of chlorophylls a and b in plants treated by MV (mean ±SD; n=3).(D) Quantification of root growth inhibition in response to MV, expressed as root length increment (30 seedlings per n; n=3).(E) Quantification of root growth in control conditions expressed as root length increment (20 seedlings per n; n=3).Statistically significant difference at a P<0.05 as determined by one-way ANOVA with post-hoc Tukey HSD test (mean ±SD) indicated by letters above the columns/boxes.

Fig. 6 .
Fig. 6. size and aperture measurements in patl4 (A) Representative images of epidermal cells of abaxial sides in wild-type (WT), patl4-1, and patl4-2 mutants under control conditions.(B) Representative images of stomata of WT, patl4-1, and patl4-2 mutants incubated in a liquid half-strength MS medium without (Control) or with 1 µM methyl viologen (MV).(C) Quantification of stomatal size under control conditions.Stomatal size was measured as stomatal width×length as depicted by white rectangles in (A) (n=200 stomata).(D) Quantification of stomatal aperture under control (C) and after MV treatment conditions.Stomatal aperture was quantified as a ratio of aperture width to aperture length as depicted by black (width) and red (length) lines in (B) (mean ±SD, 20 stomata per n, n=3).Statistically significant difference at a P<0.05 as determined by one-way ANOVA with posthoc Tukey HSD test is indicated by letters above the boxes/columns.Scale bar, 10 µm.

Table 2 .
Proteins involved in metabolic processes and differentially abundant in WT, fsd1-1, and fsd1-2 mutants after 30 min MV treatment