New Role for Thioredoxins in Plants: Implication of TRXo1 in Protein Depersulfidation

New role for thioredoxins in plants: Implication of TRXo1 in protein depersulfidation Sabrina De Brasi-Velasco a,1 , Angeles Aroca b,1 , Luis C. Romero b , Cecilia Gotor b,** , Francisca Sevilla a , Ana Jim´ enez a,* a Centro de Edafología y Biología Aplicada del Segura-CSIC, Murcia, Spain b Instituto de Bioquímica Vegetal y Fotosíntesis (CSIC-US), Sevilla, Spain ARTICLE INFO Keywords: Ascorbate peroxidase Dehydroascorbate reductase Monodehydroascorbate reductase Persulfidation Redox regulation Thioredoxin o1 ABSTRACT Persulfidation, a posttranslational modification of cysteines to persulfides, is the best characterized molecular mechanism of H 2 S signaling. This study is focused on new functions for thioredoxins (TRXs) in plants beyond those of thiol disulfide (S–S) exchange, including the regulation of protein persulfidation as it has been described in animal systems. To elucidate the impact of TRXo1 deficiency on the protein persulfidation pattern in plants of Arabidopsis thaliana L. wild type (WT) and two Attrxo1 T-DNA insertion mutants grown under non stress conditions, a quantitative proteomic approach was performed. The proteomic analysis revealed a higher number of proteins that were more persulfidated in the mutants compared to WT plants, suggesting a role for TRXo1 in protein depersulfidation. Interestingly, most of the differentially persulfidated proteins were located in the chloroplast, implying a coordination between chloroplast H 2 S-dependent persulfidation and mitochondrial TRXo1 depersulfidation. Among the differentially persulfidated proteins located in mitochondria, the antioxidant enzymes sAPX, DHAR1 and MDAR6 were selected for further studies. The effect of H 2 S-dependent persulfidation on their enzymatic activities and its reversibility by the NADPH/thioredoxin reductase (NTRB)/TRXo1 system was analyzed, as well as their persulfidation levels were quantified. Sulfide treatment brought about increases in the activity levels of the enzymes, that match with a raise on the persulfidation levels. Interestingly, both activations declining after treatment with the thioredoxin system, indicate the regulation of their persulfidation by TRXo1. These results point to a positive effect of persulfidation on the enzymatic activities and also to a new depersulfidase activity for TRXo1. All together these results give a new insight of the mechanism of elimination of –SSH groups in plants exerted by TRXo1, and the involvement of a redox regulation on the protein persulfidation. 1. Introduction Hydrogen sulfide (H 2 S) is nowadays a well-stablished signaling molecule like others, such as carbon monoxide (CO), nitric oxide (NO), and hydrogen peroxide (H 2 O 2 ), and regulates essential processes in both animals and plants [1–4]. H 2 S is produced endogenously by cells through different enzymes involved in cysteine metabolism. In plants, the main source of H 2 S is the chloroplast, where it is produced in the photosynthetic sulfate assimilation pathway [5], although in the cytosol, H 2 S is metabolically generated from cysteine through several types of cysteine-degrading enzymes, and in mitochondria is generated during cyanide detoxification [6]. In recent years, intensive research on H 2 S has been carried out in different organisms, and the number of biological functions in which sulfide is known to be involved has increased rapidly [7,8]. In plants, H 2 S has been shown to be essential for the regulation of plant responses to a wide range of environmental conditions, mainly abiotic stresses, ranging from metal stresses to drought, salinity, hypoxia, heat, and many others. In this way, H 2 S allows plant adaptability and viability, and its beneficial effects play a role in important aspects of development [5,9–11]. H 2 S also regulates processes that are critical for adequate plant performance, including abscisic acid (ABA)-dependent stomatal movement and autophagy. In particular, the regulation of these processes by H 2 S has been the subject of intense investigation in recent years [12–20]. The mechanism of action of H 2 S is related to its chemical * Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (C. Gotor), [email protected] (A. Jim´ enez). 1 These authors have contributed equally to this work and share first authorship. Contents lists available at ScienceDirect Redox Biology journal homepage: www.elsevier.com/locate/redox https://doi.org/10.1016/j.redox.2025.103627 Received 24 February 2025; Received in revised form 26 March 2025; Accepted 4 April 2025 Redox Biology 82 (2025) 103627 Available online 7 April 2025 2213-2317/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/bync/4.0/ ). properties, and three mechanisms have been proposed. H 2 S can bind metal centers in metalloproteins, can also react with biological oxidants and thus acting as antioxidant, and can modify protein cysteine residues to form persulfides, which is the posttranslational modification (PTM) named persulfidation (also known as S-sulfhydration), and that has been the most studied so far [21–23]. In plants, different high-throughput proteomic analyses have revealed that protein persulfidation is widely distributed in the Arabidopsis proteome, representing almost 13 % of the whole annotated proteome [24–29]. Furthermore, specific studies of persulfidated proteins have shown that the outcomes of this modification are the alteration of protein functions, causing important physiological effects, and several target cysteine residues have been identified (reviewed in Ref. [30]). Despite all this research, the precise mechanism that leads to the formation of the persulfide has been subjected to extensive debate, due to H 2 S (or HS − ). Therefore, it has been proposed that H 2 S reacts with oxidized cysteine residues, and the reaction with sulfenylated residues (R–SOH) is the most favorable process [31], and consequently, protein persulfidation should require the previous formation of protein sulfenylation. Temporal dynamics of protein sulfenylation and persulfidation have been shown in mammalian cells and in Arabidopsis leaves [26,32]. Based on these data and other studies, it has been concluded the existence of an interplay between the H 2 O 2 and H 2 S signaling pathways. Insight into the reversibility of protein persulfidation is still scanty, largely in plants. Protein persulfidation was proposed to be a mechanism that prevents the overoxidation of cysteine residues, resulting in loss of protein function. Under persistent oxidation, thiol groups originate sulfinic (-SO 2 H) or sulfonic (-SO 3 H) motifs that are mainly irreversible, whereas persulfide groups generate oxidized perthiol residues (-SSOH, –SSO 2 H and –SSO 3 H) that can be reduced back to thiol by the action of reductant agents [21,33]. Enzymatic mechanism that reverses in vivo the persulfidation by the thioredoxin system has been demonstrated in animal systems [32,34], highlighting thioredoxin as depersulfidase; however, in plants, this reversion of persulfidation has not been proven to take place so far. Mitochondria are highly dynamic and metabolically active cell organelles that have essential functions in plant energy conversion, which play a central role in a variety of primary and secondary metabolic pathways. A key feature of mitochondrial biochemistry is the unavoidable production of ROS, with complex I and complex III being the major sites of superoxide (O 2 .- ) generation, which is scavenged by superoxide dismutase (Mn-SOD), producing hydrogen peroxide (H 2 O 2 ), being the most relevant ROS due to its reactivity, diffusivity, and prolonged halflife. H 2 O 2 may be reduced by the peroxidase systems such as glutathione peroxidases (GPXs) and peroxiredoxins (PRXs) and by antioxidant enzymes in the ascorbate-glutathione (ASC-GSH) cycle, using the reducing power of NAD(P)H. It is composed by the enzymes ascorbate peroxidase (APX), monodehydroascorbate reductase (MDAR), dehydroascorbate reductase (DHAR) and glutathione reductase (GR), as well as by the antioxidants ASC and GSH [35,36]. Redox regulation and ROS metabolism are interlinked and involved in optimizing the function of cell organelles. The mitochondrial antioxidant system has a key role in the detoxification of O 2 .- and peroxides and thus plays a crucial role in controlling redox signaling. H 2 O 2 oxidizes the thiolate anion to the sulfenic form (Cys-SOH) which can react with another thiolate to form an intraor intermolecular disulfide bond (S–S) (reviewed by Ref. [37]). The cellular redox system is composed of proteins sensitive to oxidative changes classified as redox sensors, including PRXs and GPXs, as well as redox transmitters such as thioredoxins (TRXs) and glutaredoxins (GRXs) [38–40]. Thus, the cellular redox state will depend on the thiol groups and the oxidized forms of antioxidants such as GSH and among others, the cited proteins TRX, PRX, GRX, and also sulfiredoxins (SRX), the last being able to reverse the sulfinic forms of certain PRXs [41,42]. TRXs are considered key components of cellular redox balance through the redox regulation of a multitude of targets proteins involved in various metabolic pathways in different cellular compartments [37, 43–46]. In this way, the redox state of the TRXs can reversibly affect the activity of their target proteins. TRXs are in turn regenerated from their oxidized forms by NADPH-dependent thioredoxin reductases (NTR) such as NTRA in the cytoplasm, NTRB in the mitochondria and nucleus, and NTRC in plastids, or ferredoxin-dependent FTR in chloroplasts [47–51]. In A. thaliana, there are 21 genes of typical TRXs, and among them, two type o thioredoxins have been located in mitochondria: TRXo1 and TRXo2 [43,52]. Regarding TRXo1 studied in this work, it showed a strong interaction with pea leaf mitochondrial PsPRXIIF, which together with PsSRX, may constitute a mitochondrial redox system [42,53]. In addition to mitochondria, TRXo1 is also located in the nucleus in pea and in Arabidopsis plants, and regulates diverse proteins involved in processes such as ATP synthesis, protein translation, respiration, photorespiration, tricarboxylic acid cycle, cell proliferation, cell cycle progression and ABA perception, among others [43–45,54,55]. Likewise, in Arabidopsis and in pea mitochondria, the 3-mercaptopyruvate sulfurtransferase (MST) protein, named as STR1 in Arabidopsis, has been described as target of TRXo1, while the cytosolic isoform STR2 interacts with TRXh1 in this compartment [43,56]. The biochemical characterization of human MSTs has confirmed that TRXs are physiological persulfide acceptors contributing to the generation of H 2 S and, in this way, TRXs can act as regulators of protein persulfide levels in the Abbreviations ABA abscisic acid ASC reduced ascorbate CAM chloroacetamide CAT catalase Cl–NBF 4-chloro-7-nitrobenzofurazan Daz-2 dimedone azida-2 DCP-bio 3-(2,4-dioxocyclohexyl) propyl-biotin DHAR dehydroascorbate reductase EDTA ethylenediaminetetraacetic acid ETHE ethylmalonic encephalopathy GPX glutathione peroxidase GR glutathione reductase GSH reduced glutathione GRX glutaredoxin KO knock out LC-MS/MS liquid chromatography-mass spectrometry LFQ label-free quantitative MDAR monodehydroascorbate reductase MST 3-mercaptopyruvate sulfurtransferase NRX nucleoredoxin NTR NADPH-dependent thioredoxin reductase PRX peroxiredoxin POD peroxidase PTM post-translational modification ROS reactive oxygen species sAPX stromal ascorbate peroxidase SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis SnRK SNF1-related protein kinase SOD superoxide dismutase TRPC indole-3-glycerol phosphate synthase TRX thioredoxin WT wild type S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 2 cells [57]. Similarly, in plants it was reported that STR1 could produce H 2 S by interacting with mitochondrial TRXo1 and TRXo2 [44], and the interaction between STR2 and TRXh1 could represent a pathway for H 2 S synthesis in the cytosol. All these proteins may then play a key role in the regulation of persulfide levels [56,58], although the mechanism occurring in vivo needs to be understood. Unlike animals, there is scarce knowledge about the role of TRXmediated modulation of Cys PTMs such as glutathionylation, nitrosation, or persulfidation in plant biology. In order to deepen on the possible new function for redox regulation of protein persulfidation mediated by TRX, in this work we examine the effect of mitochondrial redox imbalance caused by Arabidopsis TRXo1 deficiency on the plant persulfidation pattern. For this purpose, a comparative and quantitative proteomic analysis was performed in wild type (WT) and KO Attrxo1 plants, identifying mitochondrial, chloroplastic and nuclear differentially persulfidated proteins. Among them, antioxidant recombinant proteins were produced and treated with NaHS and NADPH//NTRB/ TRXo1 system, identifying the specific modified cysteine residues responsible for sulfide regulation of the enzymatic activities, and evidencing a new depersufidase activity for TRXo1. 2. Material and methods 2.1. Plant material and growth conditions Plants of Arabidopsis thaliana L. wild type ecotype Columbia-0 and Attrxo1 T-DNA insertion mutants (trxo1-1, SALK_ 143294C, KO1 and trxo1-2, SALK_042792, KO2) [59] were grown in soil during 21 days in controlled growth chambers under a photoperiod of 16 h of LED light (120 μ mol m −2 s −1 ) at 24 ◦C and 8 h of dark at 18 ◦C, with 65 % humidity. Plant leaves were frozen in liquid nitrogen till analysis. 2.2. In-gel persulfidation detection The in-gel persulfidation detection was performed using 500 mg of Arabidopsis leaves harvested from 21-day-old wild type and trxo1 mutant plants, following the previously described dimedone switch detection method [32,60]. Briefly, in a first step, protein extracts were incubated with 5 mM 4-chloro-7-nitrobenzofurazan (Cl–NBF, Merck) which reacts with all Cys residues and amino groups. Then, the persulfide adducts in the proteins were selectively labeled with the mix dimedone azida-2 (Daz-2, Cayman Chemical)–Cy5-alkyne (Lumiprobe) and proteins were assessed by SDS-PAGE measuring fluorescence of the bands on an Ettan DIGE imager (GE Healthcare). Intensity of the fluorescence of NBF and Cy5 was analyzed with IMAGEJ software, and the Cy5/NBF ratio was calculated for the quantification of persulfidation. 2.3. Dimedone-switch method and proteomics A total of 500 mg of Arabidopsis leaves from wild type and trxo1 mutants of 21-day old plants were subjected to the dimedone-switch method as previously described [60]. Briefly, protein extracts were first labeled with 5 mM Cl–NBF and then incubated with 100 μ M DCP-Bio1 (Merck) which reacts with persulfide adducts. Persulfidated proteins were enriched using Sera-Mag Magnetic Streptavidin beads (Cytiva). A total of 400 μ g of proteins were trypsinized and analyzed by LC-MS/MS using a TIMS Tof Pro spectrometer (Bruker, Germany). The processed data were analyzed with the MQ (Max Quant) search engine and label-free quantification was performed using PEAKS Studio (BSI, Canada). The search settings were set as previously described [60], including modifications of lysine: NBF (mass shift:163.0012), modifications of cysteines hydrolyzed DCP-Bio1 (mass shift:168.0786), NBF (mass shift:163.0012), carbamidomethyl (C) and methionine oxidation (M) as variable modifications. The functional enrichment analysis and functional annotation of gene lists were performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) web server [61]. 2.4. Expression and purification of recombinant his-tagged proteins The complete complementary DNAs of stromal ascorbate peroxidase (sAPX, AT4G08390), dehydroascorbate reductase 1 (DHAR1, AT1G19570), and monodehydroascorbate reductase 6 (MDAR6, AT1G63940) provided by ABRC (The Ohio State University), were cloned into the pDEST17 vector (Invitrogen) to express the N-terminal 6His-tagged proteins using the Escherichia coli expression system with Gateway Technology (Invitrogen). Proteins were expressed in E. coli BL21 (DE3) cell cultures treated with 0.1 mM isopropyl-β-D-thiogalactopyranoside for 16 h at 30 ◦C, and were purified using Ni-NTA purification system (Invitrogen) according to the manufacturer’s recommendations. Recombinant protein production was assessed by SDSPAGE. 2.5. Identification of persulfidated cys residues of recombinant proteins Persulfidated residues of trypsin-digested recombinant proteins were identified by liquid chromatography and mass spectrometry analysis. Before digestion, proteins were incubated with 10 mM chloroacetamide (CAM) in dark, room temperature for 30 min 1 μ g aliquot from each sample was subjected to 1D-nano LC ESI-MSMS analysis using a nano liquid chromatography system (Eksigent Technologies nanoLC 425, (SCIEX) coupled to a high-speed TripleTOF 5600 plus mass spectrometer (SCIEX) with a Nanospray III source. The analytical column was a silicabased reversed-phase column C18, Acclaim PepMap 100C18, 3 μ m, 100 Å, 75 μ m id ×250 mm, Thermo Fisher Scientific, with an emitter, PicoTip (F360-20-10-N-20_C12, New Objective). The trap column was Acclaim PepMap 100C18, 5 μ m, 100 Å, 100 μ m id ×20 mm, Thermo Fisher Scientific. The injection volume was equivalent to 1 μ g of protein. Data acquisition and processing of the spectra were performed as previously described [24]. Searches were done with the A. thaliana protein database from UniProt (at February 16, 2024) which contains 39281 protein-coding genes and their corresponding reversed entries using the Mascot Server v. 2.6 (Matrix Science, London, UK) including the database from E. coli due to some proteins were expected in the extract of recombinant protein purification. Search parameters were set as follows: enzyme, trypsin, CAM (Cysteine), CAM-Sulfide (S–S-CAM) (Cysteine), Sulfide (Cysteine), Oxidation (Methionine) as variable modifications. The peptide mass tolerance was set to 50 ppm and 0.6 Da for-fragment masses, and 1 missed cleavage was allowed. False discovery rates (FDR≤1 % at the PSM level) for peptide identification were manually calculated. 2.6. Determination of enzymatic activities Recombinant sAPX, MDAR6 and DHAR1 were assayed for their enzymatic activities after different treatments such as, incubation with NaHS, reduction of the NaHS-treated proteins by the TRXo1/NTRB/ NADPH system, and only NTRB and NADPH (as control). All the treatments were done at 25 ◦C incubating 30 min with slight movement and in darkness. Recombinant proteins (around 600 μ g) were treated with 1 μ M NaHS and after dialysis by Amicon 3K, protein concentration was measured. An aliquot of 5 μ M of the protein was treated with 5 μ M PsTRXo1 (produced by Abyntek, Spain), 0.12 μ M AtNTRB (kindly provided by Dr. J.J. L´ azaro, EEZ-CSIC, Granada, Spain) and 1.25 mM NADPH and incubated as specified above. TRXo1 activity was measured by the insulin-disulfide reduction assay as previously described [43], using 0.5 mM NADPH and 50 nM AtNTRB. Ascorbate peroxidase (APX; EC 1.11.1.11), monodehydroascorbate reductase (MDAR; EC 1.1.5.4) and dehydroascorbate reductase (DHAR; EC 1.8.5.1) activities were assayed as previously described [35]. Briefly, APX activity was measured in 50 mM Hepes-NaOH buffer, pH 7.6, following the decrease in A 290 due to the ascorbate (0.2 mM) oxidation S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 3 by H 2 0 2 (0.09 mM) (Ɛ 290 =2.8 mM −1 cm −1 ). MDAR activity was assayed in Tris-HCl 50 mM buffer, pH 7.2, following the decrease in A 340 , due to the NADH (0.2 mM) oxidation (Ɛ 340 6.22 mM −1 cm −1 ), generating monodehydroascorbate by the ascorbate (1 mM)/ascorbate oxidase (0.5 units) system. DHAR activity was measured in 100 mM potassium phosphate buffer, pH 6.5, containing 1 mM EDTA (ethylenediaminetetraacetic acid) saturated with N 2 , following the increase in A 265 due to ascorbate formation (Ɛ 290 =14 mM −1 cm −1 ), using 0.2 mM DHA and 2.5 mM GSH. 2.7. Immunochemical detection of persulfidated recombinant proteins The level of persulfidation of the recombinant proteins subjected to different treatments as described above, was analyzed by the dimedoneswitch method [60], followed by immunoblotting. For that, treated proteins were blocked with 5 mM Cl–NBF in 50 mM Tris-HCl buffer, pH 8.0 containing 2 % SDS at 37 ◦C for 30 min in darkness as described previously and kept at −20 ◦C overnight. Next, the samples were filtrated by Amicon to eliminate the excess of NBF–Cl and treated with 200 μ M of DCP-Bio1 at 37 ◦C for 2 h to be analyzed by Western blot after SDS-PAGE in 12 % gels (Miniprotean Bio-Rad) and transference to Fig. 1. In-gel detection of total protein persulfide (PSSH) levels. (A) Fire pseudo-coloring was used to visualize Cy5 signal. Green coloring corresponds to the total protein load (7-nitrobenzofurazan (NBF)-protein adducts). Ratio of Cy5: NBF signals is used for quantification (n =6). (B) Quantification of PSSH changes (n = 6). Values are presented as a mean ±SD and letters indicates statistically significant difference (ANOVA, Turkey test, P <0.001). S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 4 nitrocellulose membranes, as previously described [55]. Biotinylation was detected using a monoclonal anti-biotin primary antibody (Merck, ref. B7653, 1:8000) and goat anti-mouse secondary antibody (Abcam, ref ab205719, 1:10000). Super Sigma west Pico Plus kit (Thermofisher) was used to detect proteins, and membranes were analyzed using an image analyzer (Amersham Imager 600, GE Healthcare). Ponceau staining allowed us to check and correct the loading using Image J software (https://imagej.nih.gov/ij/) to quantify bands. 2.8. Statistical analysis The experiments were conducted in a completely randomized design. Three samples of each genotype with 8 plants per sample were processed for analysis of in vivo protein persulfidation. Results of persulfidation level are shown as the mean ±SD of three biological and two technical replicas, and the data were analyzed by ANOVA and Turkey test, P < 0.001. All results related to activity assays of recombinant proteins are shown as the mean ±SD of three independent experiments, doing at least four replicates for each treatment. Data were analyzed using the Student’s t-test (P <0.05). 3. Results 3.1. Comparison of total protein persulfidation profiles To elucidate the impact of TRXo1 deficiency on total protein persulfidation, the in-gel detection of persulfidation level was first analyzed and quantified in leaves from trxo1 (KO1 and KO2) mutant plants compared to WT plants grown under physiological conditions (Fig. 1A). A significant higher level of persulfidation was determined in both trxo1 mutants, being the level of persulfidation in KO2 the highest (Fig. 1B). 3.2. Identification and quantitative comparison of persulfidated proteins To assess which are the specific proteins whose level of persulfidation could be regulated by thioredoxin, we used a label-free quantitative (LFQ) approach combined with the dimedone switch method to measure protein persulfidation in leaf samples in wild type and the trxo1 mutant alleles [32,60]. Protein samples from three biological replicates (independent pools) of leaf tissues from wild type (WT) and two trxo1 mutant alleles (KO1 and KO2) were isolated and subjected to the dimedone Fig. 2. Proteomic analysis of protein persulfidation in wild type and the trxo1 mutant alleles. (A) Hierarchical clustering of persulfidated proteins in wild type and the trxo1 mutant alleles KO1 and KO2. (B) Venn diagram showing the intersection of the differentially persulfidated proteins identified in the pairwise comparison between wild type and each of the mutant alleles. Number in the white boxes indicates the number of proteins. S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 5 switch procedure. The proteins eluted from streptavidin beads were digested and the peptide solution analyzed by liquid chromatography tandem mass spectrometry with a DIA approach. A total of 2224 proteins were identified with at least two peptides at FDR <1 % as susceptible to persulfidation in the analyzed samples (Dataset 1). Hierarchical clustering analysis of the three plant lines correctly assigned replicates to the experimental groups and showed a high correlation between each of the control samples and between the replicas of the two alleles among them (Fig. 2A), finding a higher number of proteins with a higher level of persulfidation in the mutants than in WT plants, in concordance with that obtained in the in-gel analysis (Fig. 1). The LFQ proteomic approach led to the identification and quantification of 439 proteins that were differentially persulfidated in the KO1 mutant line compared to wild type (Dataset 2) and 350 in the KO2 (Dataset 3). Of these, were common 203 in all lines, 236 were regulated in KO1 and 147 in KO2 (Fig. 2B). This represents a total of 586 differentially persulfidated proteins in both mutant alleles compared to wild type, of which 377 proteins showed higher level of persulfidation in the mutants while 209 proteins were lesser persulfidated in them (Dataset 4). The level of persulfidation in both mutant alleles was very high up to 10-fold in several common proteins such as the COP9 signalosome subunit CSN5A, the nucleoredoxin 1 NRX1, the indole-3glycerol phosphate synthase TRPC, a receptor-like protein kinase or the 50S ribosomal protein L20 with diverse subcellular localization such as cytosol, chloroplast, nuclei, or cellular membrane. Among the set of differentially persulfidated proteins, 50 showed subcellular location in mitochondria and included thiosulfate/3-mercaptopyruvate sulfurtransferase 1, STR1, and persulfide dioxygenase ETHE1 homolog with 2.77 and 2.42-fold-increase in persulfidation level, respectively. Among antioxidants, ascorbate peroxidase (sAPX), dehydroascorbate reductase (DHAR1), peroxiredoxin IIF and Mn superoxide dismutase (Mn-SOD) were found to be more persulfidated in the mutants (1.8, 1.4, 3.6 and 1.8-fold-increase, respectively). Other 27 proteins also showed subcellular location in the nucleus but the main group of differentially persulfidated proteins were plastid located, with 189 proteins (Dataset 4). The 586 differentially persulfidated proteins were further analyzed, and Gene Ontology by biological process (GO_BP) and KEGG pathway enrichment and clustering were performed to classify the functions/ pathways that were more affected by the trxo1 mutation (Table 1). Six clusters of functions and pathways with significant enrichment scores containing the up-persulfidated proteins in the trxo1 mutants were identified. Based on the GO_BP function, the most significant term corresponded to protein folding and cellular response to unfolded proteins and comprised several chaperonins and heat shock 70 kDa proteins such as BIP-1and BIP-2 (Table 1). The second group was light harvesting in photosystem I and included 6 chlorophyll a-b binding proteins. Based on the KEGG pathway annotation, 4 terms were significantly enriched, 3 of them related to glycolysis, carbon metabolism, and metabolic pathways in general. Interestingly, the fourth term corresponded to sulfur metabolism and included four proteins of the sulfate assimilatory pathway (APR1, APS2, OASA1 and SIR) and the 3-mercaptopyruvate sulfurtransferase STR1, whose persulfide group formed in the catalytic cysteine of the protein can eventually be transferred to GSH for sulfite detoxification in the mitochondria or could lead to sulfide production by reaction with TRXo1 [62]. The functional classification analysis of the down-persulfidated proteins revealed only one significantly enriched term that corresponded to embryo development ending in seed dormancy, which comprised 20 proteins and included TIM9 that encodes a small zinc finger-like protein that is a component of the mitochondrial protein import apparatus [63]. 3.3. Analysis of persulfidated cys residues of target proteins In the proteomic analysis performed in WT and trxo1 mutant plants, a large number of proteins were revealed as targets of persulfidation, Table 1 Functional annotation clustering of the overrepresent terms based on GOBiological Process and KEGG pathway of the 377 up-persulfidated and 209 down-persulfidated proteins. Count: number of proteins in the annotation term. P-adj: Benjamini adjuted p-value. Cluster Description of term Count p-adj Proteins in the cluster Up-persulfidated proteins 1 GO:0006457: Protein folding/cellular response to unfolded proteins 25 9.7 × 10 −8 AIP3; ALDH5F1; ANNAT1; ANNAT2; BIP1; BIP-2; Hsp70-2; CNX1; CPN60B; CPN60B2; CR88; CRT1b; FTSH11; HSP7C; HSP702; HSP60; HSP602; HSP81-2; MTHSC702; PBP1; PDIL1-1; PEN3; ROC3; SGT1B 2 GO:0009768: Photosynthesis, light harvesting in PS I/ Protein-chromophore linkage 8 1.7 × 10 −3 LHCA1; LHCA3; LHCB2.2; LHCB3; LHCB4.1; LHCB4.2; LIL3:2; LIL3:1 3 ath01100: Metabolic pathways 149 1.6 × 10 −9 6PGD1; ACCO2; ACO1; ADK2; AGAL1; AL2B4; AL3I1; ALFC4; ALFC5; ALFC6; ALFP3; AOC1; AOC2; AOC4; APR1; APS2; APT1; APXS; APY5; ASD1; ATPG1; BAM5; BCA2; BGL20; BGL23; BGL26; BGL37; BGL39; CAB6; CADH4; CAPP2; CARA; CB3; CB4A; CB4B; CHI5; COQ3; CORI3; CP74A; CX5B1; CYSK1; DHAR1; DHE2; DLDH2; DPE1; DPNP1; E134; ENO2; ETHE1; FAH; FNRR1; FQR1; FQRL2; G3PA1; G6PIP; GAGT1; GATP; GCSP2; GGT1; GLO2D; GLX1; GLYM2; GPX2; GSTF2; GSTF8; GSTF9; GSTFA; GSTU5; GSTZ1; GUN25; HEM3; HEXO3; HIBC6; HIS2; HXK1; IDI1; ILVH1; INVA4; ISPF; ITPA; KASC2; LACS4; LEU12; LEU31; LEUD3; LHCA3; LHCB2; LIP2M; MDHC1; MDHNP; METC; MOC2B; MTBC; MTHR2; MTNA; NDADB; NDUA2; NDUS1; NILP1; NRL1; ODO2A; ODP23; ODPB1; OPR3; OTC; PDX11; PDX13; PDXK; PER12; PER30; PFPB2; PGKH2; PGMC2; PGMP; PHS1; PLPD1; PMG1; PMG2; PMK; PPA4; PSA; PSAH1; PSAN; SAHH1; SEBP1; SEBP2; SGAT; SIR; SOX; SPDS1; SQD1; SSDH; SSL12; STAD5; STR1; THI4; THRC1; TL29; TRPC; UGHY; USP; VATD; 4 ath00010: Glycolisis/ Gluconeogenesis 19 5.5 × 10 −5 ALDH2B4; ALDH3I1; CPGK2; FBA4; FBA5; FBA6; HXK1; iPGAM1; iPGAM2; LOS2; LPD1; MAB1; MEE51; MTE2-3; mtLPD2; PDE345; PGI1PGM1; PGM2 (continued on next page) S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 6 and some of the mitochondrial antioxidant components were selected for further analysis, such as sAPX, MDAR and DHAR1 (Dataset 1). To elucidate the target site for persulfidation in these proteins, we obtained the recombinant A. thaliana proteins and carried out liquid chromatography (LC)–tandem mass spectrometry (MS/MS) analysis on them. Proteins were purified from an E. coli extract and trypsin digested under nonreducing conditions to avoid the reduction of persulfide residues, but chloroacetamide was used to carbamidomethylate thiol and persulfide groups. As disulfide bridges between digested peptides cannot be avoided, one missed cleavage was allowed in the search. The digested peptides were analyzed using LC-MS/MS for a 32-Da mass increase plus carbamidomethylation (SS-CAM) in the fragmentation spectrum. sAPX, DHAR1, and MDAR6 were identified with a sequence coverage of 59 %; 90 %; and 65 %; respectively. All of them showed only one peptide containing an SS-CAM modification (Fig. 3). In the case of sAPX, the persulfidated peptide was VDASGPEDCPEEGR containing the Cys 223 modified (Fig. 3A). For DHAR1, the only peptide identified with SS-CAM was AAVGAPDHLGDCPFSQR at the Cys 20 (Fig. 3B). The Cys 25 of MDAR6 in the peptide SGLSLWCPSSPSLAR was exclusively identified as persulfidated and modified by SS-CAM (Fig. 3C). Only in DHAR1, all cysteine-containing peptides were identified, while in sAPX one cysteine was missed and in MDAR6 only 3 out of 5 cysteines were identified in the LC-MS/MS. The score values from the MASCOT search engine for these peptides VDASGPEDC (SS-CAM) PEEGR, AAVGAPDHLGDC (SSCAM) PFSQR and SGLSLWC (SS-CAM) PSSPSLAR were 98, 113 and 62, respectively. 3.4. Reversible effect of persulfidation on enzymatic activities In order to elucidate whether the cysteine persulfidation exerted any effect on enzymatic activities, and if this effect is in a reversible manner, we measured the activity of the selected enzymes described previously under different treatments. First, the specific activity of non-treated proteins (Control) was measured as nmol min −1 mg −1 protein and considered as 100 %. After persulfidation with NaHS, a fraction of the protein was treated with the TRXo1/NADPH/NTRB system to analyze the effect of the possible depersulfidation activity of this system. The sAPX specific activity showed a statistically significant activation by NaHS-induced persulfidation of around 2.1-fold and the treatment with the TRX system abolished that increase (Fig. 4). DHAR1 also showed a significant increased activity by NaHS-induced persulfidation (1.6-fold) and likewise the TRX system reverted this effect (Fig. 4). The effect of persulfidation was observed as well in MDAR6 activity, increasing 2.6fold its activity, and the TRX system provoked the decrease of the activity, reaching similar values to those of the reduced protein. As a control of reversibility, we checked the effect of treatment with only NADPH or NTRB without TRXo1 on persulfidated enzymes and no effect was observed on their activities (Fig. 4). These results point to a positive effect of NaHS-induced persulfidation on the enzymatic activities and interestingly, to a depersulfidation activity by TRXo1. We also performed the insulin-reduction method to measure TRXo1 activity after NaHS-induced persulfidation of the protein, finding that the treatment with 1, 10 or 100 μ M NaHS did not have any effect on the activity, although insulin reduction occurred slowly (Fig. S1). 3.5. Immunodetection of the persulfidation induced by NaHS and the depersulfidation by TRXo1 system To confirm regulation of persulfidation/depersulfidation of the recombinant enzymes treated as indicated in the previous section, we performed the dimedone-switch method followed by immunoblot analysis of biotinylated proteins. Signal of persulfidation of sAPX without any treatment appeared mainly in the dimeric form of around 84 kDa and with less intensity in the monomeric form of around 42 kDa (Fig. 5A). This indicated that the recombinant sAPX was purified from bacteria with endogenous level of persulfidation, as we routinely detected in other proteins [25]. After the NaHS treatment, the dimeric form showed an increase in the persulfidation level, which strongly decreased when it was treated with the thioredoxin system (NADPH/TRXo1/NTRB). Longer time exposure of the membrane (see magnification in Fig. 5A) revealed that the monomeric form of sAPX showed a lower increase of persulfidation level when treated with the NaHS, although decreased with the TRX system treatment in a similar way to the dimeric form. A signal of persulfidation was also observed in the monomer and dimer of NTRB. For recombinant DHAR1, the monomer with a molecular weight of around 25 kDa was the form presented mainly persulfidated with a low amount of persulfidated dimer (Fig. 5B), the latter better detected under longer time of exposure of the membrane. The NaHS treatment increased the persulfidation level in both monomeric and dimeric forms, and the TRX system efficiently decreased this level also in both forms. Interestingly, the TRX system significantly decrease the level of persulfidation of all proteins forms of both sAPX and DHAR1 proteins, even to lower levels than the observed in non-treated proteins. As persulfidated NTRB monomeric and dimeric forms were observed (Fig. 5A), we analyzed the persulfidation level in the recombinant NTRB and also in TRXo1, finding that both of them were persulfidated even without the NaHS treatment in the monomeric and oligomeric forms, as it is shown for purified proteins from bacteria (Fig. S2). In addition, the treatment of TRXo1 with 100 μ M NaHS increased the signal in the monomer and dimer forms of the protein and the NADPH/NTRB treatment of the NaHS-treated TRXo1 did not eliminate the persulfidation of the monomer or dimer (Fig. S2). Therefore, our findings demonstrate that the TRXo1 efficiently reverts persulfidation of plant proteins, acting as depersulfidase. Moreover, we have revealed the effect of persulfidation/depersulfidation on the enzyme activities of the antioxidative proteins studied in a positive/ negative manner, respectively. 4. Discussion The control of ROS generated during cellular metabolism and especially in stressful situations is performed by the antioxidant system located in several cellular compartments. Different components of the antioxidant system have been identified and characterized [35,38,42, 43,45,53,66]. In particular, the mitochondrial/nuclear located thioredoxin TRXo1 has been the focus of our previous studies and has been demonstrated to play a key role in processes such as germination, development, stomatal opening, respiratory and photosynthetic metabolism, cell cycle, regulation of carbon metabolism and plant response Table 1 (continued) Cluster Description of term Count p-adj Proteins in the cluster 5 ath01200: Carbon metabolism 29 1.6 × 10 −4 ACO1; AGT; c-NADMDH1; Clp; CPGK2; E2OGDH2; FBA4; FBA5; FBA6; GAPA; GDH2; GGT1; GLDP2; HXK1; iPGAM1; iPGAM2; LOS2; LPD1; MAB1; MTE2-3; MTHFR2; mtLPD2; NADP-MDH; OASA1; PDE345; PGD1; PGI1; PPC2; SHM2 6 ath00920: Sulfur metabolism 9 3.3 × 10 −3 APR1; APS2; GLY3; OASA1; SBP1; SBP2; SIR; STR1 Downpersulfidated proteins 1 GO:0009793: Embryo development ending in seed dormancy 20 1.9 × 10 −3 6PGL3; ACCT2; ACOX4; COG7; CPNA1; CRK17; DHQSD; IF62; ODP25; PIN1; PP273; R18A2; RS16A; SKP1B; SUFE1; SYGM2; SYRM; TIM9; VATE1; Y5848 S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 7 Fig. 3. Analysis of sAPX (A), DHAR1 (B) and MDAR6 (C) using mass spectrometry. Proteins were identified with a sequence coverage of 59 %; 90 %; and 65 %, respectively. The identified peptides are shown in red, and the peptide containing persulfidated Cys (Cys 223 , Cys 20 and Cys 25 , respectively) are shown in bold and underlined. LC-MS/MS analysis of the tryptic peptide containing persulfidated Cys are shown, and the tables contain the predicted ion types for the modified corresponding peptides. The ions detected in the spectrums are highlighted in red color. Nomenclature of the fragment ions and types corresponds to that proposed by Roepstorff and Fohlman [64] and modified by Biemann [65]. S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 8 to salinity [43,45,59,67–70]. Recently, there has been a growing interest in persulfidation, the PTM of cysteine residues to form persulfides, which is the wellestablished molecular mechanism involved in H 2 S signaling in both mammals and plants [1,5,10]. Interestingly, close interconnections between H 2 S and H 2 O 2 signaling at different levels have been described. Due to the sulfur oxidation states of H 2 S and the thiol group being the same, a previous oxidation is required to obtain a persulfidated cysteine. Different thermodynamic studies have suggested the reaction of H 2 S with the oxidized form of thiol group such as sulfenic acid (R–SOH) should occur [21,29]. In this way, H 2 O 2 -induced sulfenylation is fine-tuned regulated to take place before H 2 S-induced persulfidation as demonstrated [26,32]. Consequently, persulfidation has emerged as a protective mechanism against irreversible oxidation that negatively affects protein structure and function, thus modulating protein oxidation which occurs in stressful situations. Despite the extensive antioxidant system, it may be insufficient to control high levels of ROS in specific scenarios leading to thiol overoxidation that cannot be reversed to reduced thiol [33]. However, in the same situation, the oxidized form of persulfides can be reduced back to thiol by the TRX system as shown playing a role as depersulfidase activity in the animal system [32,34,71]. In recent years, several proteomic analyses have identified an extensive catalog of persulfidated proteins in plants, describing the effect on proteins and identifying the specific cysteine residues target of persulfidation [5,9]. In all cases, in vitro reversibility of the effect of persulfidation has been shown by incubating with reducing agents. Thus, persulfidation has been shown to inhibit the activity of antioxidant enzymes such as catalase, but increased that of cytosolic APX1 in A. thaliana, being reversed by reducing conditions, although the physiological reductant is unknown [25,72]. Interestingly, the in vitro reversibility of the effect on the activities implies the existence of a redox regulation of the process, and it has been suggested that TRXs in plants may play a role similar to that demonstrated in animal systems [32,34, 71]. This study aims to specifically involve TRXo1 in redox regulation of the persulfidation process, by analyzing this PTM in the proteins extracted from WT plants and Attrxo1 mutants, as discussed next. In the absence of TRXo1, the total persulfidation level was significantly higher in both mutant plants than in WT, which pointed to the role of TRXo1 in the reversibility of the process acting as a depersulfidase. This increased persulfidation levels was supported by the LFQ proteomic data. In general, the basal level (in WT plants under normal growth conditions) of protein persulfidation is relatively high [24,28], so a significant increase in the absence of TRXo1 is a clear indication of its involvement in protein depersulfidation. Interestingly, the functional characterization and enrichment analysis of the 586 differentially persulfidated proteins highlighted biological processes comprising proteins located in different compartments apart from mitochondrial proteins, outstanding a significant number of chloroplast proteins. Thus, the deficiency of the mitochondrial TRXo1 provoked a significant increase of persulfidation levels of proteins involved in photosynthesis, carbon and sulfur metabolisms, being the cellular response to unfolded proteins the most enriched process, indicating that mitochondrial imbalance redox homeostasis is sensed by other organelles. Plant mitochondria are accounted hubs in redox metabolism and communication between chloroplasts and mitochondria optimizes chloroplast function while maintaining cell redox homeostasis [73,74]. Given that the main source of H 2 S production is the plant chloroplast [5], clearly, in this study we observe a coordination between chloroplast sulfide-induced persulfidation and mitochondria TRXo1 depersulfidation. Due to the mitochondrial localization of TRXo1, a more detailed analysis of the persulfidome in the two Attrxo1 mutant lines focused on mitochondrial antioxidants, showed that not only some components of this defense system could be regulated by persulfidation, but also that TRXo1 could play a role in this regulation. Attrxo1 mutants were previously reported to show higher oxidative parameters, such as increased lipid peroxidation and lower redox state of the antioxidants ascorbate and glutathione compared to wild type plants. At the mitochondrial level, these mutants presented lower ascorbate peroxidase activity regulating H 2 O 2 level [69,70]. In this sense, the high level of persulfidation in the mutants may be due to both an increase in oxidative protein and the absence of an effective depersulfidase. Among antioxidants, we found enzymes such as mitochondrial sAPX, DHAR1, Mn-SOD, and PRXIIF that appeared to be more persulfidated in trxo1 mutants than wild type under control conditions. Different antioxidant enzymes have been described as persulfidated in several proteomics studies [26–28, 75] and recombinant enzymes as tomato SlCAT1, SlcAPX1 and SlPOD5 treated with NaHS showed that some of their cysteines were persulfidated, while for SOD the persulfidation site has not been identified [76]. Not only the lack of TRXo1 is associated in our study with a higher protein persulfidation in the reported antioxidant enzymes, but also with persulfidation of several known targets of TRXo1 in the organelle. Fig. 4. Regulation of sAPX, MDAR6 and DHAR1 enzymatic activities by persulfidation and depersulfidation by TRXo1/NADPH/NTRB system. Untreated purified proteins (Control) were treated with 1 μ M NaHS for persulfidation (+NaHS), and after persulfidation were treated with the complete TRXo1 system (+NaHS +NADPH/NTRB +TRXo1), or without TRXo1 (+NaHS + NADPH/NTRB). Results are the mean ±SD (n =12) and asterisks indicate significant differences compared to the control using the t-Student’s test (p <0.05). S. De Brasi-Velasco et al. Redox Biology 82 (2025) 103627 9