Abstract
S-Nitrosylation of Cys residues is one of the molecular mechanisms driven by nitric oxide (NO) for regulating biological functions of key proteins. While the studies on S-nitrosylation of Cys residues have served for identifying SNO proteomes, the physiological relevance of protein S-nitrosylation/denitrosylation remains poorly understood. In this study, it is shown that auxin influences the balance of S-nitrosylated/denitrosylated proteins in roots of Arabidopsis seedlings. 2D-PAGE allowed the identification of ascorbate peroxidase 1 (APX1) as target of auxin-induced denitrosylation in roots. Auxin causes APX1 denitrosylation and partial inhibition of APX1 activity in Arabidopsis roots. In agreement, the S-nitrosylated form of recombinant APX1 expressed in Escherichia coli is more active than the denitrosylated form. Consistently, Arabidopsis apx1 mutants have increased H2O2 accumulation in roots, shorter roots, and less sensitivity to auxin than the wild type. It is postulated that an auxin-regulated counterbalance of APX1 S-nitrosylation/denitrosylation contributes to a fine-tuned control of root development and determination of root architecture.
Introduction
Redox signalling is pivotal in stress perception, photosynthesis regulation, pathogen defence responses, and plant growth and developmental processes (Apel and Hirt, 2004; Gapper and Dolan, 2006; Eckardt, 2010; Tognetti et al., 2012). Nitric oxide (NO) is a free radical active in preserving cellular reduction/oxidation homeostasis. As a consequence, NO is involved in numerous physiological processes in the plant life (Lamattina et al., 2003; Leitner et al., 2009). Regarding plant growth, NO is required during adventitious and lateral root formation, root hair elongation, and gravitropic responses, all processes controlled by the phytohormone auxin. Auxin induces NO production in roots via nitrate reductase and nitric oxide synthase-like activities (Pagnussat et al., 2002; Correa-Aragunde et al., 2004; Hu et al., 2005; Lombardo et al., 2006; Kolbert et al., 2007).
As a diffusible free radical, NO reacts with sulphhydryl groups in Cys residues and transition metals to form S-nitrosothiols and metal nitrosyls, respectively. Protein S-nitrosylation represents a ubiquitous mechanism for post-translational modification that is emerging with relevant biological functions as O-phosphorylation does (Lane et al., 2001; Mannick and Schonhoff, 2002). Proteomic assays in plants have detected S-nitrosothiol modification of enzymes, ion channels, and transcription factors (Lindermayr et al., 2005; Abat et al., 2008; Romero-Puertas et al., 2008). Several target proteins of NO are related with the enzymes involved in reactive oxygen species production and metabolism. S-Nitrosylation of peroxiredoxin II E inhibits hydroperoxide-reducing peroxidase activity (Romero-Puertas et al., 2007). In peroxisomes, the H2O2-degrading enzymes catalase and glycolate oxidase are S-nitrosylated proteins and its activities are inhibited by NO donors (Ortega-Galisteo et al., 2012). In addition, NADPH oxidase have been described as a new target of S-nitrosylation involved in plant immunity (Yun et al., 2011). The regulation of NO level is tightly mutually linked with H2O2; NO regulates the levels of H2O2 and vice versa (Neill et al., 2002). Both NO and H2O2 are involved in programmed cell death and pathogen defence responses (Delledonne et al., 1998), stomatal closure (Bright et al., 2006), and lateral root growth (Wang et al., 2010a,b).
Auxin is the main hormone controlling root growth. The primary mechanisms controlling auxin action involve auxin biosynthesis, conjugation, catabolism, and transport. The auxin receptors are the F-box TIR1 and its paralogues AFB1, AFB2, and AFB3, subunits of SCF E3-ligase complex. Auxin binds to SCFTIR1 and triggers the ubiquitination of the auxin repressors Aux/IAA proteins and its degradation by the proteasome. The degradation of Aux/IAA leads to the release of the auxin transcription factors ARFs with the consequent induction of auxin-responsive genes (Calderon-Villalobos et al., 2010). Very recently, it has been demonstrated the S-nitrosylation of the auxin receptor TIR1 (Terrile et al., 2012).
This work studied whether auxin and NO could regulate the balance of S-nitrosylated/denitrosylated proteins that take part of the auxin signalling during root growth. Through biotin switch assays, it was shown that auxin treatment affects the level of S-nitrosylated proteins. Ascorbate peroxidase 1 (APX1) was found to be denitrosylated and its enzyme activity partially inhibited in auxin-treated Arabidopsis roots. APX1 is a cytosolic enzyme involved in H2O2 degradation using ascorbate as the electron donor. APX1 is highly expressed in Arabidopsis roots (Fourcroy et al., 2004). To date, studies on the APX1 role in Arabidopsis have been focused on the reactive oxygen species-scavenging mechanism triggered by stress conditions (Davletova et al., 2005; Vanderauwera et al., 2011; Maruta et al., 2012). In this report, the participation of APX1 in auxin-mediated root growth process is investigated.
Materials and methods
Plant material and growth conditions
Arabidopsis thaliana Columbia (Col-0) ecotype seeds were surface-sterilized in 30% (v/v) bleach and 0.02% (v/v) Triton-X100 for 15min, rinsed with sterile water, and stratified at 4 °C for 2 days. Arabidopsis mutant tir1-1 afb2-3 is in the Columbia ecotype and apx1 is in Wassilewskija (WS) ecotype. Seeds were plated on ATS medium with 1% (w/v) sucrose and 0.8% (w/v) agar (Wilson et al., 1990) and placed vertically in a growth chamber at 25 °C under 100 μmol photons m–2 s–1 and a 16/8 light/dark cycle. For growth measurements, seedlings were grown for 4 days and then transferred to the auxin naphthyl acetic acid (NAA) or sodium nitroprusside provided by Sigma (St Louis, MO, USA). The density of lateral roots and the length of the primary roots were determined after 4 days of growth. For APX activity assays and 2D analysis, seedlings were grown for 7 days and then transferred to the treatments for 1 day.
Detection of H2O2
Roots were incubated in 15 μM of the peroxide-sensitive dye 2,7ʹ-dichlorodihydrofluorescein diacetate (H2DFDA) (Molecular Probes, Eugene, OR, USA) in 10mM MES buffer (pH 6) and 10mM KCl for 20min followed by 20min wash in fresh buffer. Roots were visualized under bright field and epifluorescent microscopy (Nikon).
APX and TrxR activity assays
Root proteins were extracted with 50mM phosphate buffer (pH 7.5), 1mM EDTA, and 1mM ascorbic acid. APX activity was measured following the H2O2-dependent oxidation of ascorbic acid at 290nm (extinction coefficient = 2.8mM–1 cm–1) as described by Amako et al. (1994). To measure the activity of S-nitrosylated recombinant APX1, protein extracts from bacteria expressing APX1 were treated with the S-nitrosylating agent nitrosocysteine (CysNO) for 20min. Then proteins were desalted by Micro Bio-Spin columns with Bio-Gel P-6 (Bio-Rad). Thioredoxin reductase (TrxR) activity was measured in 100mM phosphate buffer (pH 7.0), 10mM EDTA, and 200 μM NADPH. Reactions were initiated by the addition of 300 μM of the substrate 5,5ʹ-dithiobis(2-nitrobenzoic) acid and followed at 412nm. For inhibition experiments of TrxR activity, samples were incubated with 5 μM auranofin (ANF) for 15min previous to the measurement of TrxR activity.
Biotin switch
The biotin switch was performed according to Jaffrey and Snyder (2001) with minor modifications. Arabidopsis roots were homogenized in HEN buffer (25mM HEPES pH 7.7, 1mM EDTA, 0.1mM neocuproine) containing complete protease inhibitor cocktail (Sigma). For in vitro S-nitrosylation, protein extracts were incubated with CysNO in the dark for 1h. Samples were incubated with 30mM methyl methanethiosulphate (MMTS, Sigma) and 3.3% (w/v) SDS in HEN buffer at 50 °C for 20min and precipitated with 2 volumes of acetone. Samples were resuspended in HEN buffer, 1% (w/v) SDS, 20mM ascorbic acid, and 1mM biotin-HPDP (Pierce Chemical) and incubated for 1h. Proteins were then subjected to immunoblot using an anti-biotin antibody (Sigma).
Purification of S-nitrosylated proteins
Purification of S-nitrosylated proteins was performed according to Forrester et al. (2009). For proteomic assays, the starting material was 8mg protein. Biotinylation of S-nitrosylated proteins was carried out as described above. Samples were incubated with 30 μl of neutravidin agarose resin (Thermo Scientific) per milligram of initial protein with three volumes of neutralization buffer (25mM HEPES, 100mM NaCl, 1mM EDTA, 0.5%, v/v, Triton X-100, pH 7.5) and overnight at 4 °C. The resin was washed four times with 1ml neutralization buffer containing 600mM NaCl. Bound proteins were eluted with HEN/10 buffer containing 1% (v/v) of β-mercaptoethanol.
2D-PAGE and MS analyses
Purified proteins were resuspended in a buffer containing 8M urea, 2% (w/v) CHAPSO, 20mM DTT, 2% (v/v) carrier ampholytes (pH 4–7), and 0.002% (w/v) bromophenol blue. IEF was performed as applying a voltage gradient up to 10,000V with a limiting current of 50 μA per strip. The strips were loaded onto the second dimension and run at 15 mA/gel. Spots were excised and subjected to in-gel trypsin digestion. Mass spectrometric data were obtained using a MALDI-TOF-TOF spectrometer, Ultraflex II (Bruker), in the mass spectrometry facility at the CEQUIBIEM mass spectrometry centre, Argentina.
Cloning and expression of AtAPX1 in Escherichia coli
The ORF of AtAPX1 was amplified from PUNI51 vector containing the full-length cDNA provided by the Arabidopsis Biological Resource Center. Primers were designed with an EcoRI site at the 5ʹ-end (5ʹ-CGAATTCATGACGAAGAACTACCCAACCGTGAG CGAAGATTACAAGAAGGCTGTTGAG-3ʹ) and with a c-Myc tag and XhoI site followed by a stop codon at the 3ʹ end (5ʹ-GCTCGAGTCACAGGTCCTCCTCTGAGATCAGC TTCTGCTCAGCATCAGCAAACCCAAG-3ʹ). The amplified pro duct was ligated in pET24b vector (Novagen) and transformed BL21 Rosetta strain. Soluble expressed proteins were extracted by sonication in 50mM potassium phosphate buffer (pH 7.5), 1mM EDTA, 1mM ascorbic acid, and 1mM PMSF. Recombinant APX1 was detected by immunoblot using anti c-Myc antibody (Sigma) or cytosolic APX antibody (Agrisera, Sweden).
Homology modeling and alignment
Structural models were built with the program @TOME (Pons and Labesse, 2009). The figures were drawn using Web Lab ViewerLite 3.20 software (Molecular Simulations). Cytosolic APX sequences were aligned using Clustal X software (version 1.81) and edited with GeneDoc software (version 2.5.010).
Densitometry and statistical analysis
Densitometry analyses of gel bands were performed with Matrox Inspector v2.2 (Matrox, Quebec, Canada). Root fluorescence was quantified by optical densitometry using ImageJ software. Statistical analysis was performed employing SigmaStat (Jandel Scientific, San Rafael, CA, USA), using ANOVA and Tukey tests for multiple comparison analyses and the t-test for pair-wise comparisons.
Results
Detection of S-nitrosylated proteins in auxin-treated Arabidopsis roots
The effect of NO and auxin on root growth was studied in Arabidopsis seedlings with the NO donor sodium nitroprusside and the synthetic auxin NAA. As previously reported (Casimiro et al., 2001), Fig. 1A and B show that the treatment with 0.1 μM NAA inhibits primary root elongation and stimulates lateral root formation. Similarly but to a lesser extent, the treatment with 10 μM sodium nitroprusside induces 2-fold increase in lateral root density and inhibit primary root growth (Fig. 1A and B). The treatment of Arabidopsis with the NO scavenger CPTIO inhibits the NAA-mediated lateral root formation, suggesting that NO is required for the auxin response (Fig. 1B). As was previously described, NO production is stimulated by auxin in roots of several plant species (Correa-Aragunde et al., 2004; Kolbert et al., 2007; Terrile et al., 2012).
Auxin affects lateral root density, primary root length, and the pattern of S-nitrosylated proteins in Arabidopsis. (A) Four-day-old Arabidopsis seedlings treated with 0.1 μM naphthyl acetic acid (NAA) with or without 0.5mM of the NO scavenger CPTIO or 10 μM of the NO donor sodium nitroprusside (SNP) for 4 days; bar, 1cm. (B) Primary root length and lateral root density; data represent mean and SE of three experiments (n = 6); letters represent significant difference (ANOVA, P < 0.05). (C) Effect of auxin in the pattern of S-nitrosylated proteins; 7-day-old Arabidopsis seedlings treated with NAA for 1 day; root extracts were S-nitrosylated in vitro with 0.5mM CysNO for 1h; S-nitrosylated proteins were detected by the biotin switch method using an anti-biotin antibody (IB: anti-biotin); Coo Blue, Coomassie blue staining.
To analyse the effect of auxin in the pattern of S-nitrosylated protein in roots, Arabidopsis seedlings were treated with increasing concentrations of NAA and then root extracts were S-nitrosylated with the nitrosylating agent nitrosocysteine (CysNO). Fig. 1C shows that the treatment with NAA modifies the pattern of S-nitrosylated proteins in roots. NAA treatment increases the level of S-nitrosylated protein among 35–53kDa whereas diminishes the level of S-nitrosylated proteins with molecular weights higher than 53kDa and below 35kDa. This result suggests that while some proteins are favoured to be S-nitrosylated other target proteins seem to be denitrosylated by the auxin treatment. To improve the analysis, S-nitrosylated proteins were separated in a 2D-PAGE and identified by MS.
APX1 is denitrosylated by auxin treatment in Arabidopsis roots
To identify the S-nitrosylated proteins affected by auxin treatment, approximately 8mg of root protein extracts from H2O (control)- or NAA-treated Arabidopsis were subjected to the biotin-switch method, purified by neutravidin affinity chromatography, and separated by 2D-PAGE. Fig. 2A shows more than 40 polypeptides clearly detected in the range pH 4–7 in control roots. Supplementary Fig. S1 (available at JXB online) shows the comparative intensity of spots in the 2D-PAGE from the control and NAA treatments. The most abundant 15 spots were chosen for MS. Four out of 15 gave protein identification (Fig. 2A, right). The other 11 spots were unable to be identified. The analysis of the proteins identified reveals that three of them have already been detected as S-nitrosylated targets in other biotin-switch proteomic assays in Arabidopsis (Supplementary Table S1). Enolase (2-phospho-d-glycerate hydrolase), triose-phosphate isomerase, and ascorbate peroxidase 1 (APX1) have been already identified as target proteins for S-nitrosylation in Arabidopsis (Lindermayr et al., 2005; Romero-Puertas et al., 2008; Fares et al., 2011). In contrast, a copper-binding MLP-like protein is now reported as a new target of S-nitrosylation in Arabidopsis roots (Fig. 2A, right). The four identified proteins diminished its S-nitrosylation levels in NAA treatment (Supplementary Fig. S1).
Ascorbate peroxidase 1 (APX1) is S-nitrosylated in Arabidopsis roots. (A) Seven-day-old Arabidopsis seedlings treated with H2O (control) or 1 μM naphthyl acetic acid (NAA) for 1 day; root extracts were subjected to the biotin switch method; biotinylated proteins were purified with neutravidin beads and analysed by 2D-PAGE; the experiment was repeated three times with similar results; numbers indicate identified proteins by MS. (B) Four-week-old seedlings treated with H2O (control), 1 μM NAA, or 1 μM NAA plus 2 μM auranofin (ANF); root extracts were subjected to the biotin switch method; biotinylated proteins were purified with neutravidin beads (IP: anti-biotin), and APX1 was detected by immunoblot (IB: anti-APX); to verify protein loading, APX1 was detected by immunoblot before neutravidin purification. (C) Root extracts treated with 0.5mM CysNO with or without 50mM DTT; APX1 was detected by immunoblot (IB: anti-APX). Arrowheads indicate high-molecular-weight forms of APX1; M, monomer; Ponceau red staining was used to confirm equal loading.
APX1 was chosen to continue the study since many reports have pointed out the correlation between oxidative stress and auxin signal transduction (Vernoux et al., 2000; Bashandy et al., 2010; Wang et al., 2010a,b; Tognetti et al., 2012). However, the information about the mechanism of how auxin regulates the redox changes leading to root development is scarce. Fig. 2A (centre) shows that the level of S-nitrosylated APX1 diminished 35% in NAA-treated roots with respect to control. Consequently, this study was interested in knowing the connection between auxin actions on root growth through the control of APX1 activity.
To confirm the S-nitrosylation of APX1, Arabidopsis root proteins were subjected to the biotin switch method and biotinylated proteins were purified by a neutravidin affinity chromatography. APX1 was detected by immunoblot using an anti-APX antibody. Fig. 2B shows that, under these conditions, the S-nitrosylated forms of APX1 are found as high-molecular-weight forms, probably including the homodimer form of APX1. Cytosolic APX is a dimer consisting of identical subunits with a molecular mass of 28kDa (Mittler and Zilinskas, 1991). The NAA treatment does not induce a significant reduction of the level of S-nitrosylated APX1 (Fig. 2B). However, when one of the enzymatic activities responsible of protein denitrosylation the thioredoxin reductase (TrxR)-thioredoxin (Trx) system (Benhar et al., 2009) is inhibited with the TrxR inhibitor ANF, the level of S-nitrosylated APX1 increases up to 5-fold with respect to the control (Fig. 2B). The inhibitory effect of ANF on the plant TrxR activity was confirmed in Arabidopsis roots (Supplementary Fig. S2). To check the effect of NO in the transition of the APX monomer to the high-molecular-mass form of APX, protein root extracts were treated in vitro with CysNO. Fig. 2C shows that the treatment with CysNO stabilizes the high-molecular-mass form of APX1. Furthermore, the addition of the thiol reducing agent DTT to root extracts switches the high-molecular-mass form of APX1 to the monomer (Fig. 2C).
APX1 contains five Cys that could be potential target of S-nitrosylation (Fig. 3A). Cys32 and Cys168 are almost conserved among all described cytosolic APXs in plants (Fig. 3B and Supplementary Fig. S3). Cys32 is located near the ascorbate binding site (Fig. 3C; Mandelman et al., 1998) while Cys168 is located near the haem-binding motif. The distance between SH group of Cys168 and Fe ion of the haem group is 6.72 Å (Fig. 3C). The Cys32 was already reported to be S-nitrosylated in Arabidopsis (Fares et al., 2011). Local hydrophobicity within primary or tertiary protein structure might promote S-nitrosylation since it increases thiol nucleophilicity and concentrate nitrosylating equivalents (Hess et al., 2005; Marino and Gladyshev, 2010). The hydrophobicity value by Kyte-Doolittle scale showed that it is considerably high (value of 0.615) in the sequence flanking Cys32 in APX1, while the average of Cys168 is rather low (–1.308; Supplementary Fig. S4). At structural level, both Cys32 and Cys168 are surrounded by hydrophobic amino acids within a 6 Å radius from the sulphur atom (Supplementary Fig. S4). When longer distances (up to 8 Å) from the sulphur atom of Cys32 and Cys168 are considered, charged amino acids appear (Supplementary Fig. S4). These characteristics are shared among most CysNO of S-nitrosylated proteins reported so far (Marino and Gladyshev, 2010).
Structure and activity of ascorbate peroxidase 1 (APX1) from Arabidopsis roots. (A) Amino acid sequence of Arabidopsis APX1; amino acids that make contact with haem or ascorbic acid (ASC) are depicted in blue or green, respectively. (B) Cytosolic APX alignment showing the conservation in Cys32 and Cys168; black boxes indicate identical amino acids among all sequences; grey boxes indicate identical residues among 12 or 13 sequences. (C) Model of Arabidopsis APX1 according to the coordinates of APX from Glycine max (PDB code 2VCF); Cys32, Cys168, and amino acids that make contact to haem and ASC are shown. (D) APX activity from roots of Arabidopsis wild type and tir1 afb2 mutant treated with H2O (control) or different concentrations of naphthyl acetic acid (NAA) or 1 μM NAA plus 2 μM auranofin (ANF) for 1 day; values are fold increase with respect to control; APX activity=1 corresponds to 8.58 μmol ascorbic acid (ASC) min–1 (mg protein)–1; bars, 1 SE (n = 3); different letters indicate significant difference among treatments (ANOVA, P < 0.05). APX1 protein levels were detected by immunoblot.
To study the effect of auxin on APX1 activity in roots, Arabidopsis seedlings were treated with increasing concentrations of NAA. Fig. 3D shows that APX activity is inhibited up to 30% in NAA-treated Arabidopsis roots. APX1 activity restores to control levels when NAA-treated roots are incubated in the presence of ANF. Monomeric APX1 protein levels are unaffected by these treatments (Fig. 3D, top), while the dimeric form could not be detected in this assay. NAA treatment was unable to inhibit APX1 activity in the Arabidopsis auxin-receptor mutant tir1 afb2, indicating that an active TIR1/AFB2-mediated auxin signalling is required for the auxin-induced inhibition of APX1 activity (Fig. 3D, right).
S-Nitrosylated recombinant AtAPX1 is more active than the denitrosylated form
The ORF of AtAPX1 was cloned into pET24b expression vector with a c-Myc tag in the C-terminal. E. coli BL21 was transformed with pET24b-APX1 and induced with IPTG. Fig. 4A shows that the maximal APX1 activity was reached at 5h of induction with 0.1mM IPTG. The expression of IPTG-induced recombinant APX1 was verified by immunoblot and the biological activity of the recombinant APX1 was analysed in bacteria by examining the tolerance to oxidative stresses (Supplementary Fig. S5). Supplementary Fig. S5B shows that bacteria expressing recombinant APX1 are more tolerant to treatment with either NaCl or H2O2 than bacteria carrying the empty vector.
Expression analysis of the recombinant APX1 in Escherichia coli. (A) APX1 activity in E. coli expressing pET24b-APX1. (B) APX1 activity from bacterial extracts treated with CysNO in the presence or absence of 1mM DTT and desalted by bio-gel p6 filtration; APX1 activity = 1 corresponds to 0.706 μmol ascorbic acid (ASC) min–1 (mg protein)–1. (C) In vitro S-nitrosylation of APX1. Protein extracts from bacteria expressing APX1 were S-nitrosylated with CysNO for 1h. Samples were subjected to the biotin switch method and biotinylated protein were purified with neutravidin beads (IP: antibiotin). The monomeric form of APX1 was identified by immunoblot (IB: anti-Myc). The proteins that were not retained by the column were loaded on gels and stained with Coomassie Blue (Coo Blue) as loading control. (D) APX1 purified using an anti-Myc agarose column (IP: anti-Myc), subjected to the biotin switch method, and detected by immunoblot (IB: anti-biotin).
Soluble proteins from bacteria expressing APX1 were S-nitrosylated in vitro with CysNO, and then the CysNO was removed by gel filtration to prevent APX1 inhibition by reversible binding of NO to the haem group. Fig. 4B shows that the treatment with 50 μM CysNO increases APX1 activity 1.45-fold with respect to untreated APX1. Addition of 1mM DTT to CysNO-treated extract abolishes any increase of APX1 activity (Fig. 4B). Treatment with CysNO was able to S-nitrosylate the monomeric form of APX1 (Fig. 4C). This study did not observe the high-molecular-form of recombinant APX in the bacterial extracts, probably due to the reducing environment of bacterial cytoplasm. Very low levels of S-nitrosylated APX1 detected in samples treated with 0.5mM Cys, indicate that endogenous S-nitrosylation of APX1 could occur (Fig. 4C). Supplementary Fig. S6 also shows that the addition of the thiol-reducing agent DTT after S-nitrosylation with CysNO results in the removal of the S-NO bonds.
To detect the S-nitrosylated APX1 high-molecular-weight form, recombinant APX1 was purified from bacteria extracts, S-nitrosylated in vitro with CysNO, and subjected to the biotin switch method. S-Nitrosylated APX1 was recognized by an anti-biotin antibody. This strategy allows detecting the S-nitrosylation of the high-molecular-weight APX1 form (Fig. 4D).
Arabidopsis mutant in APX1 results in changes in root architecture
To analyse the role of APX1 in root development, this study used the apx1 Arabidopsis mutant. Immunoblot confirmed the absence of APX1 in the homozygous line (Fig. 5A). To analyse the level of H2O2 production in apx1 plants, the peroxide-sensitive dye 2ʹ,7ʹ- dichlorodihydrofluorescein diacetate (H2DFDA) was used. Fig. 5B and C shows that the treatment with auxin induces H2O2 production in Arabidopsis roots. Under basal conditions, roots of apx1 mutants accumulate 2-fold-more H2O2 than to the wild type (Fig. 5C). To analyse the root phenotype of apx1 mutant, Arabidopsis seedlings were treated with different concentrations of NAA for 4 days. Fig. 5D shows that increasing concentrations of exogenous NAA inhibited primary root elongation in the wild type. Intriguingly, the effect of auxin in the apx1 background is not statistically significant, suggesting that the mutant is less sensitive to auxin. When lateral root density was quantified, the Arabidopsis apx1 background was shown to be less sensitive than wild type too. Increased concentrations of NAA did not increase lateral root density in apx1 as much as in the wild type (Fig. 5D).
Root development and H2O2 accumulation are affected in apx1 Arabidopsis mutant. (A) Detection of cytosolic APX1 in wild type (WS) and apx1 mutant seedlings using anti-APX. (B) Production of H2O2 in apx1 mutant; Arabidopsis seedlings were treated with H2O (control) or 0.1 μM naphthyl acetic acid (NAA) for 4 days; roots were then incubated in 15 μM of the peroxide-sensitive dye H2DFDA for 20min; bar, 1mm. (C) Densitometric analysis of H2O2 production using ImageJ software; asterisks denote statistical difference from the respective control value (t-test; P < 0.05) (D) Four-day-old seedlings were treated with 0.05 or 0.1 μM NAA for 4 days. Primary root length and lateral root density were measured; data are mean and SE of three experiments (n = 6); different letters indicate significant difference (ANOVA, P < 0.05); asterisks denote statistical difference from the respective control value (t-test; *, P < 0.05; **, P < 0.01).
Discussion
Reactive oxygen species are involved in auxin-induced root growth in Arabidopsis
Auxin is the main plant morphogen that regulates developmental processes, including embryo morphogenesis and root, shoot, and flower development. Although the described auxin signalling pathway is apparently complete (Dharmasiri et al., 2005; Kepinski and Leyser, 2005), the modulator pathways remain to be integrated in a more comprehensive model. Redox molecules have been involved in auxin transport, metabolism, and signalling during plant development and stress responses (Gapper et al., 2006; Feher et al., 2008; Tognetti et al., 2012). The results presented here indicate that auxin treatment increases H2O2 accumulation in Arabidopsis roots and that this accumulation may be due to the auxin-mediated regulation of APX1 activity through its partial denitrosylation.
The participation of H2O2 in the auxin signalling was already reported in other developmental processes. H2O2 plays a role as a downstream component in the auxin-mediated signalling pathway leading to root gravitropic response in Arabidopsis (Joo et al., 2001). According to this, the impairment of auxin signalling in the tir1 afb2 Arabidopsis mutants resulted in a reduced accumulation of H2O2 and superoxide anion (Iglesias et al., 2010). Wang et al. (2010b) showed that the Arabidopsis carrying a mutation in the prohibitin gene PHB3, defective in H2O2-induced NO accumulation, was altered in auxin-induced lateral root formation. Alterations in the redox intermediates ASC and GSH are associated with changes in the quiescent centre (Kerk and Feldman, 1995; Jiang et al., 2003) and are involved in the G1 to S transition in the plant cell cycle (Potters et al., 2010). Altogether, the current knowledge indicates that a fine-tuned and spatial-temporal regulation of redox molecules is required for initiation and maintenance of cell division during root development.
Auxin treatment induces denitrosylation and inhibition of APX1 activity in Arabidopsis roots
APX is a haem-containing enzyme that reduces H2O2 to water with ascorbate as specific electron donor (Asada, 1992; Shigeoka et al., 2002). APX1 is a cytosolic enzyme as well as APX2 and APX6. Expression of APX1 in leaves is relatively high with respect to APX2 and APX6 (Panchuk et al., 2005) and also in roots according to the Genevestigator database (Hruz et al., 2008). The current work demonstrates the S-nitrosylation of APX1. High-molecular-weight forms of APX1 were detected to be S-nitrosylated in vivo. Even though this study cannot discard the possibility of interaction between APX1 and other proteins (i.e. Trxs) and/or macromolecules, one of these forms may be the homodimer of APX1, since it also appears when the recombinant APX1 purified from bacterial extracts is S-nitrosylated. The intensity of the bands corresponding to the high-molecular-weight forms of APX1 in Arabidopsis extracts are relatively low compared to monomeric APX1 form. It cannot be discarded to a low recognition of the monomeric form by the specific APX1 antibody. Blue native gels should probably help to get more insights about the identity of the high-molecular-S-nitrosylated complex.
Auxin induces denitrosylation and partial inactivation of APX1 in roots. Two cellular systems have emerged as physiologically relevant denitrosylases: the nitrosoglutathione reductase (GSNOR) and the Trx system. The Trx system is involved in auxin signalling, transport, and homeostasis (Bashandy et al., 2010). APX1 has been clearly described as target of plant Trxs and Trx-SH2 causes the inactivation of APX1 (Marchand et al., 2004; Gelhaye et al., 2006). According to this, the treatment with the TrxR inhibitor ANF augments the level of S-nitrosylated APX1 and increases its activity in roots. In mammalian systems, ANF was shown to inhibit TrxR at nanomolar concentrations. At higher concentrations (1–50 μM) ANF could also inhibit glutathione reductase (Gromer et al., 1998).
Few examples of protein denitrosylation have been reported in plants so far. Denitrosylation of the salicylic acid (SA) regulator NPR1 catalysed by Trx facilitates its monomerization, nuclear translocation, and activation of genes coding for pathogenesis-related proteins (Tada et al., 2008). On the other hand, NO production is induced by SA treatment (Zottini et al., 2007) and required for the expression of SA-regulated pathogenesis-related protein genes (Klessig et al., 2000). The proposed model for the SA-mediated regulation of NPR1 is then through the opposing and balancing action of GSNO and Trx on NPR1 (Tada et al., 2008; Vlot et al., 2009). In a similar way, auxin treatment induces NO production in roots and at the same time provokes protein denitrosylation. Taking all available data, it is suggested that a counterbalance effect of NO production and APX1 denitrosylation are coordinately operating downstream auxin in root developmental processes.
S-Nitrosylated APX1 is more active than denitrosylated form
The stimulation of APX activity by S-nitrosylation may explain the effect of NO donors on APX activity described in several plant organs and physiological conditions. For example, three isoforms of APX are activated by NO donors in soybean root nodules (Keyster et al., 2011). NO protects against UV irradiation by stimulating the activity of the antioxidant enzymes catalase and APX (Shi et al., 2005). In addition, NO induces S-nitrosylation of APX in Anticaria toxicaria seeds, triggering a positive effect on APX activity during seed desiccation (Bai et al., 2011). Also reported is the reversible in vitro inactivation of APX by NO donors in tobacco crude extract (Clark et al., 2000). Since NO binds to and reversibly inhibits some mammalian haem-containing enzymes, the effect of NO donors on tobacco APX activity was attributed to the binding of NO to the metal Fe-group (Clark et al., 2000). Thereby, two mechanisms might be operating on NO-mediated regulation of APX activity: (i) S-nitrosylation of Cys residue/s and (ii) binding of NO to the haem group. Proteins that bind transition metals as Fe2+ and Cu2+ close to a S-nitrosylated target thiol can catalyse the transnitrosylation to reactive Cys residues (Hess et al., 2005). In APX1, the proximity of Cys168 to the haem group is 6.72 Å and makes Cys168 a strong candidate for S-nitrosylation through transnitrosylation reaction.
APX1 is involved in auxin-mediated root growth
Cytosolic APX1 was characterized as a central component regulating redox homeostasis in Arabidopsis. Knockout apx1 Arabidopsis plants have an increased H2O2 accumulation in leaves under high light conditions and higher sensitivity to paraquat (Davletova et al., 2005). Regarding the growth behaviour, apx1 plants have delayed development phenotype, late flowering, and perturbed stomatal responses (Pnueli et al., 2003). The current data show that primary root elongation and lateral root formation is altered in response to auxin in the apx1 mutant. This mutant has increased levels of H2O2 in roots (Suzuki et al., 2013; this work). However, Arabidopsis apx1 root phenotype differs from exogenous application of H2O2 to roots. Exogenous H2O2 results in inhibition of primary root elongation and induction of lateral root formation, reminiscent of the phenotype found in auxin-treated roots (Wang et al., 2010a). Thus, APX1 may participate in the appropriate redistribution of H2O2 accumulation during root growth and lateral root development in Arabidopsis. The implicances of the crosstalk between auxins and NO in the regulation of APX1 activity by S-nitrosylation/denitrosylation in roots and the effect on the H2O2 accumulation and signalling began to be elucidated. A finely tuned and controlled counterbalance of the redox cell status seems to be a hub in auxin signalling control of root architecture.
Supplementary material
Supplementary data are available at JXB online.
Supplementary Fig. S1. Bidimensional PAGE of S-nitrosylated proteins in Arabidopsis roots.
Supplementary Fig. S2. Effect of auranofin (ANF) on TrxR activity in Arabidopsis roots.
Supplementary Fig. S3. Alignment of plant cytosolic APXs and the conservation of Cys residues in APX sequences.
Supplementary Fig. S4. Microenvironment of Cys32 and Cys168 in Arabidopsis APX1.
Supplementary Fig. S5. Expression of recombinant APX1 in bacteria.
Supplementary Fig. S6. Recombinant APX1 is specifically S-nitrosylated in vitro.
Supplementary Table S1. Protein and peptide sequences identified from selected spots by MALDI-TOF/TOF MS.
Acknowledgements
The authors thank Ron Mittler for providing the apx1 Arabidopsis mutants and Claudia Casalongué for providing PUNI51-APX1 vector, the tir1 afb2 Arabidopsis mutants. They acknowledge María Cecilia Terrile and María José Iglesias for critical reading of the manuscript. This work was supported by Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT, PICTs 2383 to LL), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET, PIP112 20110100903 to LL and NC-A) and Universidad Nacional de Mar del Plata (UNMdP), Argentina. NC-A acknowledges the support of the EMBO Fellowship. LL and NC-A are members of the research staff and NF is a postgraduate fellow from CONICET, Argentina.
References
Comments