Nitric oxide influences auxin signaling through S-nitrosylation of the Arabidopsis TRANSPORT INHIBITOR RESPONSE 1 auxin receptor

Authors

  • María C. Terrile,

    1. Instituto de Investigaciones Biológicas, UE-CONICET-UNMDP, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata, Funes 3250, CC 1245, 7600 Mar del Plata, Argentina
    Search for more papers by this author
  • Ramiro París,

    1. Instituto de Investigaciones Biológicas, UE-CONICET-UNMDP, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata, Funes 3250, CC 1245, 7600 Mar del Plata, Argentina
    Search for more papers by this author
  • Luz I. A. Calderón-Villalobos,

    1. Section of Cell and Developmental Biology, University of California, San Diego, La Jolla, CA 92093, USA
    Search for more papers by this author
    • Present address: IPB-Leibniz Institute of Plant Biochemistry, 06120 Halle (Saale), Germany.

  • María J. Iglesias,

    1. Instituto de Investigaciones Biológicas, UE-CONICET-UNMDP, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata, Funes 3250, CC 1245, 7600 Mar del Plata, Argentina
    Search for more papers by this author
  • Lorenzo Lamattina,

    1. Instituto de Investigaciones Biológicas, UE-CONICET-UNMDP, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata, Funes 3250, CC 1245, 7600 Mar del Plata, Argentina
    Search for more papers by this author
  • Mark Estelle,

    1. Section of Cell and Developmental Biology, University of California, San Diego, La Jolla, CA 92093, USA
    Search for more papers by this author
  • Claudia A. Casalongué

    Corresponding author
    1. Instituto de Investigaciones Biológicas, UE-CONICET-UNMDP, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata, Funes 3250, CC 1245, 7600 Mar del Plata, Argentina
    Search for more papers by this author

(fax +54 223 4724143; e-mail casalong@mdp.edu.ar).

Summary

Previous studies have demonstrated that auxin (indole-3-acetic acid) and nitric oxide (NO) are plant growth regulators that coordinate several plant physiological responses determining root architecture. Nonetheless, the way in which these factors interact to affect these growth and developmental processes is not well understood. The Arabidopsis thaliana F-box proteins TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX (TIR1/AFB) are auxin receptors that mediate degradation of AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA) repressors to induce auxin-regulated responses. A broad spectrum of NO-mediated protein modifications are known in eukaryotic cells. Here, we provide evidence that NO donors increase auxin-dependent gene expression while NO depletion blocks Aux/IAA protein degradation. NO also enhances TIR1-Aux/IAA interaction as evidenced by pull-down and two-hybrid assays. In addition, we provide evidence for NO-mediated modulation of auxin signaling through S-nitrosylation of the TIR1 auxin receptor. S-nitrosylation of cysteine is a redox-based post-translational modification that contributes to the complexity of the cellular proteome. We show that TIR1 C140 is a critical residue for TIR1–Aux/IAA interaction and TIR1 function. These results suggest that TIR1 S-nitrosylation enhances TIR1–Aux/IAA interaction, facilitating Aux/IAA degradation and subsequently promoting activation of gene expression. Our findings underline the importance of NO in phytohormone signaling pathways.

Introduction

Auxin (indole-3-acetic acid) coordinates many plant growth processes by modulating gene expression, which leads to changes in cell division, elongation and differentiation. It activates transcription by stimulating the degradation of transcriptional repressors called AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA) proteins through the action of the E3-ubiquitin ligase complex, SCFTIR1/AFB (Ruegger et al., 1998; Gray et al., 2001). The substrate receptor subunits of the complex, the TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX (TIR1/AFB) proteins, also function as auxin receptors. Auxin acts as a ‘molecular glue’ to stabilize the interaction between TIR1/AFBs and Aux/IAA proteins, thus promoting their ubiquitination and degradation (Dharmasiri et al., 2005a; Kepinski and Leyser, 2005; Tan et al., 2007). TIR1 is a member of a small gene family that contains five additional AFB proteins (AFB1–AFB5), that collectively mediate auxin response and are essential for the growth and development of Arabidopsis thaliana (Dharmasiri et al., 2005b; Parry et al., 2009).

Nitric oxide (NO) is a hydrophobic, highly diffusible gaseous molecule with a broad spectrum of regulatory functions involved in controlling growth, developmental and patho-physiological processes (Wendehenne et al., 2004; Delledonne, 2005). Frequently under the control of hormonal stimuli, NO acts as a second messenger implicated in many plant cell signaling events. In particular, NO is an important molecule in the auxin-regulated signaling cascade determining root morphology during growth and development (Correa-Aragunde et al., 2007). Accumulation of NO was reported in response to auxin treatment during adventitious root (AR), lateral root (LR) and root hair formation (RH), as well as asymmetric accumulation of NO in the root tip during the gravitropic response (Pagnussat et al., 2002; Correa-Aragunde et al., 2004; Hu et al., 2005; Lombardo et al., 2006). Auxin-induced AR, LR and RH formation as well as root gravitropic response were prevented by application of the specific NO scavenger carboxy-PTIO (cPTIO), suggesting a key role for endogenous NO in the control of those processes (Pagnussat et al., 2002; Correa-Aragunde et al., 2004; Hu et al., 2005; Lombardo et al., 2006).

In general, the different ways by which NO interacts with proteins have been described. Besides its capacity to bind metal ions of heme groups, NO participates in important post-translational modifications through S-nitrosylation and nitration. These events are currently under study in plants (Lindermayr and Durner, 2009).

S-nitrosylation is the reversible binding of a NO moiety to a reactive cysteine residue of a protein to form an S-nitrosothiol (Hess et al., 2005). Even though NO redox-based modification of target proteins by S-nitrosylation is emerging as an efficient regulatory mechanism in plant and mammalian signal transduction, few S-nitrosylated proteins are known in plants (Lindermayr and Durner, 2009; Lindermayr et al., 2010; Palmieri et al., 2010; Yun et al., 2011).

In this study, we demonstrate that NO modulates auxin-dependent gene expression through the stimulation of TIR1/AFB–Aux/IAA interaction. The results presented here provide evidence for a link between the auxin signaling pathway and NO through S-nitrosylation of TIR1. Our findings emphasize the physiological impact of two key cysteine residues in TIR1 that are responsible, at least in part, for the action of NO in auxin signaling.

Results

NO activates auxin signaling

To explore the role of NO in auxin signaling, we examined auxin-induced NO production in Arabidopsis roots. Seedlings were pre-loaded with the permeable NO-sensitive fluorophore diaminofluorescein-FM diacetate (DAF-FM DA) and then exposed to 1 μm IAA for 1.5 h. We observed increased NO-associated fluorescence in IAA-treated roots (Figure 1). Further, we analyzed the auxin response after NO donor application in BA3:GUS transgenic seedlings (Oono et al., 1998). This line carries auxin response elements fused to the β-glucuronidase encoding gene (GUS). Seedlings were treated with a low dose of IAA (10 nm). While GUS was expressed poorly in IAA-treated seedlings, simultaneous application of IAA and the NO donor sodium nitroprusside (SNP) caused a substantial increase in GUS reporter expression (Figure 2a). Treatment of BA3:GUS seedlings with SNP alone did not induce significant GUS activity (Figures 2a and S1a in Supporting Information). Furthermore, we used the specific NO scavenger cPTIO to study the NO requirement for IAA-dependent gene expression. BA3:GUS seedlings treated with 50 nm IAA showed GUS activity, but the application of cPTIO prior to IAA treatment prevented such induction (Figure 2b). Hemoglobin, another NO scavenger, and cPTIO inhibited auxin-dependent BA3:GUS induction in a dose-dependent manner (Figure S1c). Moreover, the auxin-regulated reporter DR5:GUS (Ulmasov et al., 1997) displayed a similar response to BA3:GUS to NO-donor and NO-scavenger treatments (Figures S1b,d). We also directly analyzed the expression of auxin-responsive genes in response to NO donor treatment. The expression of both IAA1 and IAA5 genes seems to be potentiated by SNP in IAA-treated wild-type seedlings (Figure 2c) suggesting that NO might indeed be required for auxin-dependent gene expression in Arabidopsis.

Figure 1.

 Indole-3-acetic acid (IAA) increases endogenous nitric oxide (NO) in Arabidopsis roots.
Wild-type seedlings grown on Arabidopsis thaliana salts (ATS) medium were pre-loaded with the probe diaminofluorescein-FM diacetate (DAF-FM DA), and exposed to 1 μm IAA for 1.5 h at 23°C in the dark. Accumulation of NO is shown as green fluorescence in representative roots. Bar = 100 μm.

Figure 2.

 Nitric oxide (NO) is required for indole-3-acetic acid (IAA)-dependent gene expression and auxin (Aux)/IAA protein degradation.
(a) Arabidopsis BA3:GUS seedlings were treated with sodium nitroprusside (SNP) with or without 10 nm IAA, and (b) pretreated for 45 min with carboxy-PTIO (cPTIO; 0.5 mm) or hemoglobin (50 μm) prior treatment with 50 nm IAA and them subjected to GUS staining.
(c) Wild-type seedlings were treated with different concentrations of IAA with or without increasing concentrations of SNP. Total RNA was isolated and transcript levels of IAA1 and IAA5 were analyzed using specific 32P-labeled probes. Ribosomal RNA is shown as a loading control.
(d) HS:AXR3NT-GUS seedlings were treated with S-nitrosoglutathione (GSNO) with or without 50 nM IAA or (e) with 1 μm IAA in combination with cPTIO (1 mm) or hemoglobin (50 μm) and stained for GUS activity.

Considering that auxin-mediated gene expression is regulated via degradation of Aux/IAA repressors, we hypothesized that the action of NO may result in destabilization of Aux/IAA. To address this possibility, we tested the stability of the reporter protein AXR3NT-GUS. This reporter is a fusion between the amino terminus of the Aux/IAA protein AXR3/IAA17 (AXR3NT) and GUS under control of a heat-shock inducible promoter (HS) (Gray et al., 2001). After heat shock treatment, Arabidopsis seedlings were treated with IAA in the presence of the NO donor S-nitrosoglutathione (GSNO) and the NO scavengers cPTIO and hemoglobin, and subsequently stained for GUS activity. It is important to note that we used low- or fast-release NO donors according to the treatment times required for each type of experiment (Noble and Williams, 2000; Floryszak-Wieczorek et al., 2006). The IAA treatment (50 nm) caused a decrease in AXR3NT-GUS stability that was substantially enhanced with 10 μm GSNO (Figure 2c). Moreover, seedlings treated with cPTIO and hemoglobin and in combination with high IAA concentration (1 μm) exhibited much stronger GUS staining (Figure 2d). Taken together, all these findings show that auxin-dependent gene expression and Aux/IAA degradation might rely on the availability of NO.

TIR1 protein undergoes S-nitrosylation

Since NO enhances auxin signaling by increasing degradation of Aux/IAA, a possible mechanism that could explain this modulation is via the S-nitrosylation of cysteine residues in target proteins. In this scenario, members of the Aux/IAA and TIR1/AFB families are attractive candidates for S-nitrosylation. We therefore searched for cysteine residues proposed to constitute the acid–basic S-nitrosylation motif described by Stamler et al. (2001) in Aux/IAA and TIR1-AFBs. None of Aux/IAA protein members contain an obvious putative cysteine that matched the consensus. However, two cysteine residues within TIR1, C140 and C480, are putative candidates to undergo S-nitrosylation. Interestingly, both cysteine residues are highly conserved within the TIR1/AFB receptor family (Figure 3a). The C140 residue is located at the LRR 4 loop and C480 is placed at the end of the α-helix of LRR16 in TIR1/AFBs (Figure 3b).

Figure 3.

 Conservation of C140 and C480 in the TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX (TIR1/AFB) auxin receptor family and location in the TIR1 crystal structure.
(a) Alignment of amino acid sequences surrounding C140 and C480 in the TIR1/AFB receptor family using the ClustalX program. Asterisks stand for the cysteine residues.
(b) Two views of the TIR1 structure shown as a ribbon diagram. Carbon alphas of C140 (red) and C480 (blue) are represented by spheres. IP6 (colored by element) and the IAA7 substrate peptide (yellow) molecules are shown as stick models. Indole-3-acetic acid (IAA) is represented by a space filling model (green).

To determine if TIR1 could be S-nitrosylated, baculovirus-expressed and -purified TIR1 protein was incubated with the physiological NO donor GSNO and subjected to the biotin-switch assay (Jaffrey and Snyder, 2001). In this method, nitrosothiol groups in nitrosylated proteins are substituted for a more stable biotin moiety via chemical reduction by ascorbate, and then identified by western blot. As shown in Figure 4, biotinylated TIR1 was detected upon the biotin switch and the immunoreactivity was GSNO-dependent. Moreover, TIR1 signal was drastically diminished when DTT was added after GSNO treatment. This result strongly supports S-nitrosylation of TIR1.

Figure 4.

S-nitrosylation of TRANSPORT INHIBITOR RESPONSE 1 (TIR1) recombinant protein.
Purified TIR1 protein was incubated with S-nitrosoglutathione (GSNO) or glutathione (GSH) for 30 min and subjected to biotin-switch assay. S-nitrosylated TIR1 was detected with an anti-biotin antibody (upper panel). As a negative control, GSNO-treated TIR1 was incubated with 20 mm DTT before the biotin switch. Silver-stained TIR1 is shown as a loading control.

S-nitrosylation of TIR1 affects TIR1–Aux/IAA interaction

To analyze whether S-nitrosylation of TIR1 affects auxin-dependent TIR1-Aux/IAA interaction, we performed glutathione S-transferase (GST) pull-down experiments. c-Myc-tagged TIR1 was in vitro transcription and translation (TNT) synthesized in wheat germ extracts and incubated with GST-IAA3 in the presence of a NO source. The proteins were incubated simultaneously with the NO donor S-nitrosocysteine (CysNO) and IAA. At a low IAA concentration (5 μm), the addition of CysNO resulted in a significant increase in recovery of TIR1-Myc protein as compared with the control (Figures 5a and S2a). On the other hand, incubation with cPTIO completely abolished recovery of TIR1. Carboxy-PTIO exerts its effect by scavenging endogenous NO that could be present in the TNT-wheat germ extract. The fact that IAA, as a nitrate compound may produce NO in an aqueous solution or cell-free system must not be excluded (Feelisch and Noack, 1987; Pataricza et al., 1998). The NO-dependent interaction was also observed for the AFB2–IAA3 pair (Figures 5b and S2b). Next, we investigated the effect of NO on in vivo TIR1/AFB2–Aux/IAA interactions using the LexA yeast two-hybrid system (Prigge et al., 2010). In yeast cells, the NO donor, SNP, enhanced the interaction of TIR1 and AFB2 with IAA7 in a dose-dependent fashion (Figure 5c). Altogether, in vitro and in vivo experiments indicate that NO is required and enhances both TIR1/AFB2–Aux/IAA interactions.

Figure 5.

 Nitric oxide (NO) enhances interaction of TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX 2 (TIR1/AFB2)–AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA).
Pull-down reactions were performed using (a) 35S-methionine in vitro synthetized TIR1-Myc or (b) in vitro synthetized AFB2-Myc and recombinant glutathione S-transferase (GST) IAA3 proteins. Reactions were carried out in the presence of the indicated IAA concentrations and the addition of 1 mmS-nitrosocysteine (CysNO) or 1 mm carboxy-PTIO (cPTIO). TIR1 and AFB2 proteins were detected using Phosphoimager and anti-Myc antibody, respectively. Coomassie blue-stained GST–IAA3 was used as a loading control.
(c) Yeast two-hybrid assays were carried out with cells co-transformed with the indicated constructs and grown on SD–U–H–T selective media plus the addition of 50 μm IAA or increasing concentrations of sodium nitroprusside (SNP) and 5-bromo-4-chloro-indolyl-β-d-galactopyranoside (X-Gal) to develop β-galactosidase activity.

Two TIR1 cysteines are critical for auxin-dependent TIR1–Aux/IAA interaction

Since C140 and C480 may act as putative targets for S-nitrosylation in TIR1, we expected that mutation of these residues may interfere with TIR1–Aux/IAA interaction. To address this, we generated single TIR1 mutant versions, tir1 C140A and tir1 C480A, in which each cysteine was independently changed to an alanine residue by site-directed mutagenesis. TIR1 and the mutated versions were in vitro-TNT synthesized in wheat germ extracts and tested in pull-down assays with GST-tagged IAA3 protein. As Figure 6(a) shows, the C140A mutation abolished the auxin-induced recovery of the tir1–IAA3 complex, whereas C480A severely reduced it. Moreover, yeast two-hybrid assays with C140A and C480A mutated versions of TIR1 validated our in vitro assays. Even with the addition of IAA, tir1 C140A–IAA7 interaction was completely abolished and tir1 C480A–IAA7 interaction was strongly impaired compared with the control (Figure 6b). These results indicated that the TIR1–Aux/IAA interaction is dependent on C140 and, to a lesser extent, on C480. Interestingly, mutations on other single amino acids surrounding C140, such as S138A, S139A and E141A had no effect on their binding to GST–IAA3 (Figure 6c). This result supports the specificity of S-nitrosylation in C140 and C480 TIR1 residues and ruled out the possibility that these mutations simply disrupted the overall structure of TIR1 independently of any effect of NO.

Figure 6.

 Interaction of TRANSPORT INHIBITOR RESPONSE 1 (TIR1)–AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA) is dependent on C140 and C480 of TIR1.
(a) The mutated proteins tir1-Myc C140A and tir1-Myc C480A were 35S-methionine in vitro synthesized and used in pull-down reactions with the recombinant glutathione S-transferase (GST)–IAA3 protein. Reactions were carried out in the presence or absence of 50 μm IAA and TIR1-Myc and mutant proteins were detected using Phosphoimager. Coomassie blue-stained GST-IAA3 was used as a loading control.
(b) Yeast two-hybrid assays were carried out with cells co-transformed with the indicated constructs and grown on SD–U–H–T selective media plus the addition of 50 μm IAA.
(c) Mutations in amino acids surrounding C140 do not affect TIR1–Aux/IAA interaction. The TIR1-Myc mutated versions on the S-nitrosylation consensus motif around C140 were used to carry out pull-down reactions. Experiments were performed as in (a).

The C140 residue has a pivotal role in in vivo auxin-mediated TIR1 activity

To assess the functional relevance of C140 and C480 of TIR1 in vivo, we introduced the single mutants, tir1 C140A and tir1 C480A, under the control of the constitutive cauliflower mosaic virus 35S, into both Arabidopsis wild-type and tir1-1 backgrounds (Figures S3a–b and 7). Auxin-dependent induction of LR and inhibition of primary root elongation is affected in the tir1-1 mutant plants, due to the inability of the receptor to sense the hormone. Overexpression of the wild-type TIR1 protein in the tir1-1 background rescued the normal root sensitivity to the synthetic auxin 2,4-dichlorophenoxyacetic acid (2,4-D) (Figures 7a–c). We assayed two different overexpressing transgenic lines of both tir1 C140A (L1 and L3) and tir1 C480A (L2 and L4) for auxin resistance. Although the introduction of tir1 C480A in the tir1-1 background partially restores auxin sensitivity, tir1 C140A did not recover TIR1 functionality and seedlings remained resistant to 2,4-D treatment. These data provide further evidence about the relevance of C140 as putative S-nitrosylation site in TIR1 as well as NO-mediated modification for proper in vivo TIR1 function.

Figure 7.

 tir1 C140A overexpression does not rescue the tir1-1 auxin resistant phenotype.
Four-day-old seedlings grown on Arabidopsis thaliana salts (ATS) medium were transferred to media containing 2,4-dichlorophenoxyacetic acid (2,4-D).
(a) Representative seedlings after 6 days under 2,4-D treatment.
(b) Root elongation was measured after 2 days of 2,4-D treatment.
(c) Lateral root formation was measured after 4 days of 2,4-D treatment.
Values represent the mean of at least three independent experiments ± SE.

Discussion

In this study we present evidence that NO acts as a positive modulator of the TIR1/AFB auxin signaling pathway. As previously reported in other plant species, IAA induces the accumulation of NO in Arabidopsis roots (Pagnussat et al., 2002; Correa-Aragunde et al., 2004; Hu et al., 2005). In addition, NO dependence of auxin-responsive gene expression together with NO-mediated Aux/IAA degradation and NO enhancement of TIR1–Aux/IAA interaction provide further support for this idea. We were able to show the putative S-nitrosylation of the TIR1 protein by biotin-switch assay, indicating an important role for a redox-based mechanism in the control of TIR1 action by NO. In support of our data, Kepinski and Leyser (2004) reported that N-ethylmaleimide (NEM) and 5-hydroxy-1,4-naphthoquinone (juglone), which form cysteine adducts, blocked the interaction of TIR1 and Aux/IAA. However, treatments with oxidative stressors or with skewed ratios of glutathione (GSH)/oxidized GSH had only a modest effect. More recently, Yan et al. (2009) demonstrated that CORONATINE INSENSITIVE1 (COI1), which is an F-box protein essential for all the jasmonate responses depends to a great extent on a reducing environment for its function. Similarly, S-nitrosylation of TIR1 may occur in a very particular cell-redox environment. For example, Jiang et al. (2003) reported that the formation of the quiescent center in maize is correlated with an auxin-oxidizing environment. In this sense, auxin modulates NO production and because of its reported capacity to influence redox status of tissues, it probably provides an environment suitable for S-nitrosylation.

The mechanism of auxin perception by the E3-ubiquitin ligase SCFTIR1/AFB has recently been elucidated. Since our results reveal redox-based post-translational modification of a plant E3-ubiquitin-ligase by NO, one important further issue could be to get insights into the action of S-nitrosylation on the binding between TIR1 and its ligand, auxin. Interestingly, the discovery that inositol hexakiphosphate (IP6) is associated with the TIR1 protein (Tan et al., 2007) suggests that TIR1 activity might be regulated by additional cofactors and that a ubiquitin ligase might integrate diverse signals. Moreover, our experiments showed that AFB2–IAA3 interaction is also modulated by NO. A weak TIR1/AFB–Aux/IAA interaction occurs in the absence of exogenous auxin but it is possible that NO enhances it even under this former condition (Greenham et al., 2011). S-nitrosylation could also modify the ability of TIR1 to bind auxin. However, this issue is worth addressing in a future work. S-nitrosylation of the TIR1/AFB proteins could be a versatile point of control of auxin signaling. Nevertheless, an additional mechanism of activation of auxin signaling by NO could not be discarded. Since other proteins involved in auxin signaling may also be subjected to a redox-based modification by NO, a complete understanding of its action requires the direct identification of S-nitrosylated targets. Remarkably, it was recently demonstrated that salicylic acid signaling requires regulation of the proteins NON-EXPRESSOR OF PATHOGENESIS-RELATED GENES 1 (NPR1) and TGACG MOTIF BINDING FACTOR 1 (TGA1) by NO for a proper systemic acquired resistance in plants (Lindermayr et al., 2010). Much more recently, S-nitrosylation of Arabidopsis NADPH oxidase and its consequence for activity have been also demonstrated (Yun et al., 2011). It is likely that these S-nitrosylation events are important in plant defense mechanisms. Nevertheless, S-nitrosylated-TIR1 could exert its action in plant growth and developmental responses. Thus, S-nitrosylation in plants appeared to be a general mechanism implicated in a broad spectrum of plant processes.

In TIR1 protein, the mutation of two putative S-nitrosylated residues, C140 and C480, resulted in reduced interaction with Aux/IAA proteins in vitro and in vivo. In contrast, substitution of residues adjacent to C140 had no effect on TIR1 activity. The physiological relevance of TIR1 S-nitrosylation is supported by the analysis of tir1 C140A and tir1 C480A function in tir1-1 receptor mutant plants. When primary root growth and LR formation were analyzed, tir1 C140A transgenic seedlings displayed an auxin-resistant phenotype. However, mutations in C140 and C480 of TIR1 have different effects on root growth. This fact allows us to suggest that each cysteine residue could be important for the specific redox-based post-translational modification by NO, depending on the intracellular oxidation/reduction status in the cell. As mention before, C140 is located at the LRR 4 loop in TIR1 protein and it is not close to other cysteine residues, making the formation of disulfide bonds unlikely. Notably, the attenuation of TIR1–IAA3 interaction by the C140A mutation was comparable in magnitude with that shown in the absence of NO (e.g. NO scavenger treatments) suggesting that the effect of the C140A mutation could most likely be ascribed to S-nitrosylation of C140 rather than to a structural disruption. Experimental determination for mapping S-nitrosylated residues in TIR1 will consolidate our hypothesis. Likewise, different cysteines may be selective for diverse modifications, and this fact may affect protein function in distinct ways. In Arabidopsis, C260 and C266 of the TGA1 protein were found to be S-nitrosylated and also S-glutathionylated by GSNO, whereas C172 was only S-glutathionylated under low GSNO concentrations (Lindermayr et al., 2010). Moreover, the mammalian ryanodine receptor channel, RyR1, contains different reactive thiols with selectivity for S-nitrosylation or S-glutathionylation, each leading to specific functional consequences (Martinez-Ruiz and Lamas, 2007).

It has been well documented that one of the effects of S-nitrosylation is to influence protein stability via modulation of ubiquitination and proteasome-dependent degradation (Hess et al., 2005). In human cells, ubiquitin ligases themselves have also been identified as targets for S-nitrosylation. For instance, S-nitrosylation of Parkin, a ubiquitin E3-ligase important for the survival of dopamine neurons in Parkinson’s disease, controls its E3-ubiquitin ligase activity and thereby affects its neuroprotective function (Chung et al., 2004; Yao et al., 2004). Taking into account all these findings and our data, we propose that NO might operate in multiple ways, including the negative but also positive regulation of ubiquitin-dependent protein degradation.

In summary, our results reveal an important new aspect of TIR1/AFB-mediated auxin signaling. We present evidence that S-nitrosylation of TIR1 promotes interaction between TIR1 and Aux/IAA proteins facilitating Aux/IAA degradation. Certainly, the characterization of GSNO reductase (GSNOR) mutants, atgsnor1-1 and atsgnor1-3, with altered levels of S-nitrosothiols (Feechan et al., 2005) is the most suitable genetic tool that could strongly complement our chemical approach. However, in support of our hypothesis, previous studies at our laboratory using Arabidopsis mutants which display perturbed NO levels evidenced changes in different auxin-mediated physiological responses. Flores et al. (2008) for instance, reported that the arginase-negative mutants argah1-1 and argah2-1, with high NO levels, show enhanced formation of lateral and adventitious roots. The nia1 nia2 double mutant, on the other hand, which has low NO levels, has been shown to be affected in root hair elongation, which is a well-known auxin-mediated response (Lombardo et al., 2006).

Finally, our findings pave the way for studies on the potential of F-box ubiquitin ligases to sense and integrate signals in a transduction network and to explore the role of NO on other phytohormone signaling pathways.

Experimental Procedures

Plant material, growth conditions and treatments

Arabidopsis transgenic lines BA3:GUS, DR5:GUS and HS:AXR3NT-GUS have been previously described (Ulmasov et al., 1997; Oono et al., 1998; Gray et al., 2001). Seeds were surface-sterilized and stratified at 4°C for 2–4 days in the dark and plated on Arabidopsis thaliana salts (ATS) medium (Wilson et al., 1990) containing 1% sucrose with 0.8% agar, and vertically grown at 23°C under 120 μmol photons m−2 sec−1 with 16 h:8 h light:dark cycles. Where necessary, 8- to 10-day-old seedlings were transferred to soil and grown at 23°C under the same photoperiod.

For auxin growth assays, 5-day-old seedlings growing on minimal medium on vertical agar plates were transferred onto fresh media ± 2,4-D for an additional 2 days, after which the length of roots upon transfer was measured. Lateral root assays were performed in a similar manner, except that the number of emerged lateral roots was measured after an additional 4 days. Emerged lateral roots were counted using a dissecting microscope.

Generation of transgenic lines

Arabidopsis Pro35S:tir1-Myc C140A and Pro35S:tir1-Myc C480A transgenic lines were created by introducing the point mutation in the TIR1 sequence using a QuikChange Site-Directed Mutagenesis kit (Stratagene, http://www.stratagene.com/) and pENtr-TIR1 (Invitrogen, http://www.invitrogen.com/) vector as a template. After LR Clonase reaction the sequences were cloned into the destination vector pGWB17 (Nakagawa et al., 2007). The primer sequences used for the mutagenesis reactions were as follows (altered residues underlined):

tir1 C140A fw: 5′-GTTCTTGTGCTTTCTTCCGCCGAAGGCTTCTCCACC-3′

tir1 C140A rv: 5′-GGTGGAGAAGCCTTCGGCGGAAGAAAGCACAAGAAC-3′

tir1 C480A fw: 5′-CATCATGTTTTGTCCGGGGCCGATAGCTTGAGGAAACTAG-3′

tir1 C480A rv: 5′-CTAGTTTCCTCAAGCTATCGGCCCCGGACAAAACATGATG-3′.

We transformed each destination binary vector into electrocompetent Agrobacterium tumefaciens strain GV3101 and then transformed wild-type (Col-0) and tir1-1 plants using the floral dip method (Clough and Bent, 1998). Selection of transgenic seedlings was performed by growth on hygromycin-containing plates (35 μg ml−1).

To analyze mutated tir1-Myc C140A and tir1-Myc C180A protein levels, 10-day-old seedlings were collected from wild-type, tir1-1 tir1-Myc C140A and tir1-Myc C480A transgenic lines, homogenized in ice-cold buffer [50 mm 2-amino-2-(hydroxymethyl)-1,3-propanediol (TRIS) pH 7.5, 200 mm NaCl, 10% glycerol, 0.1% Tween-20, containing 1 mm phenylmethylsulfonyl fluoride (PMSF) and complete protease inhibitor cocktail (Roche, http://www.roche.com/)], and centrifuged twice at 10 000g for 15 min at 4°C. Equal amounts of protein were loaded onto SDS-PAGE and blotted onto nitrocellulose membranes. Membranes were incubated with α-Myc antibody (Sigma, http://www.sigmaaldrich.com/) and visualized using the ECL kit (Amersham, http://www.amersham.com/).

Measurement of NO production

For determination of NO production, 5-day-old seedlings were loaded with 5 mm of the cell-permeable fluorescent probe DAF-FM DA (Calbiochem, http://www.merck-chemicals.com/) in 20 mm HEPES–NaOH pH 7.5 for 30 min in the dark. Then, seedlings were washed with fresh buffer and incubated with 1 μm IAA for an additional 1.5 h in the same conditions. After three washes, seedlings were examined by epi-fluorescence (DAF-FM DA excitation 490 nm, emission 525 nm) in an Eclipse E200 microscope (Nikon, http://www.nikon.com/) connected with a high-resolution digital camera (Nikon).

GUS staining and analysis

Five-day-old seedlings from BA3:GUS and DR5:GUS transgenic lines were transferred into liquid ATS medium containing the different compounds and incubated for 6 h at 23°C. For NO scavenger treatments, seedlings were pre-treated with these compounds 45 min before the addition of IAA. After treatment, seedlings were fixed in 90% acetone at −20°C for 1 h, washed twice in 50 mm sodium phosphate buffer pH 7 and incubated in staining buffer [50 mm sodium phosphate buffer pH 7, 1 mg ml−1 5-bromo-4-chloro-3-indolyl-β-d-glucuronide (GBT, http://www.goldbio.com/)] and stained for GUS activity from 2 h to overnight at 37°C. Seedlings were cleared by ethanol series (70, 50, 30 and 10%). Bright-field images were taken using a Nikon SMZ800 magnifier.

Six-day-old HS:AXR3NT-GUS seedlings were heat shocked at 37°C for 2 h in liquid ATS medium. The seedlings were transferred into new ATS medium containing the indicated compounds, incubated for 1 h at 23°C and stained for GUS activity as indicated above.

RNA gel blot analysis

Ten-day-old seedlings were transferred from vertical agar plates growing on minimal medium into fresh ATS medium with the indicated compounds. Seedlings were incubated with mild shaking for 3 h and ground in liquid nitrogen. Total RNA was extracted using TRIzol reagent (Invitrogen, http://www.invitrogen.com/). The RNA gel blots were undertaken using standard techniques. 32P-labeled specific DNA probes were produced using a Megaprime® DNA labeling system (Amersham Biosciences, http://www.gelifesciences.com/).

Protein expression and pull-down reactions

The tir1-Myc S138A, tir1-Myc S139A, tir1-Myc C140A, tir1-Myc E141A and tir1-Myc C480A proteins were obtained by introducing the point mutation in the TIR1 sequence using QuickChange Site-Directed Mutagenesis kit (Stratagene) and pTNT-TIR1-Myc vector as the template. The primers used for the mutagenesis reactions were as followed (altered residues underlined):

tir1 S138A fw: 5′-GGTTCTTGTGCTTGCTTCCTGCGAAGGCTTC-3′

tir1 S138A rv: 5′-GAAGCCTTCGCAGGAAGCAAGCACAAGAACC-3′

tir1 S139A fw: 5′-GTTCTTGTGCTTTCTGCCTGCGAAGGCTTCTCCACC-3′

tir1 S1390A rv: 5′-GGTGGAGAAGCCTTCGCAGGCAGAAAGCACAAGAAC-3′

tir1 E141A fw: 5′-GTGCTTTCTTCCTGCGCAGGCTTCTCCAC-3′

tir1 E141A rv: 5′-GGTGGAGAAGCCTGCGCAGGAAGAAAGCAC-3′.

For C140A and C480A point mutations, the set of primers used were the same as indicated above. TIR1, AFB2 and mutated versions of the proteins were obtained by in vitro translation using a TNT-coupled wheat germ extract system (Promega, http://www.promega.com/) in the presence of [35S] translabeled methionine where indicated. The GST–IAA3 protein was expressed in Escherichia coli and purified using GSH-sepharose according to the manufacturer’s instructions.

For pull-down assays, 20 μl of TIR1-Myc, tir1-Myc S138A, tir1-Myc S139A, tir1-Myc C140A, tir1-Myc E141A, tir1-Myc C480A and AFB2-Myc proteins were incubated for 2 h at 4°C with >10 μg of GSH-agarose immobilized GST–IAA3 protein in 200 μl of lysis buffer (20 mm TRIS pH 8.0, 200 mm NaCl, 5 mm DTT) in the presence of the indicated compounds. After washing samples with 10 bed volumes of lysis buffer, beads were resuspended in one bed volume of sample buffer, denatured and separated on 12% SDS-PAGE. Products were detected by autoradiography or by immunoblotting with anti-Myc antibody coupled to peroxidase and visualized using the ECL kit (Amersham Biosciences).

Yeast two-hybrid analysis

The cDNAs of TIR1 and AFB2 were cloned into pGILDA (Clontech, http://www.clontech.com/) and the cDNA of IAA7 was cloned into pB42AD (Clontech). To obtain mutated versions of TIR1 protein, site-directed mutagenesis was performed using the pGILDA-TIR1 vector as the template and the primers indicated above. Yeast two-hybrid vectors pGILDA and pB42AD containing the different cDNAs were transformed into the yeast strain EGY48 [pSH18-34] (Clontech) by the lithium acetate method (Gietz et al., 1992). Handling of yeast cultures, plate growth assays and β-galactosidase assays were done as described in the Clontech Yeast Protocols Handbook®.

Biotin switch assay

Recombinant TIR1 was co-expressed with ASK1 as a GST fusion protein, in High Five® (Invitrogen) suspension insect cells as previously published (Tan et al., 2007). The TIR1–ASK1 complex was isolated from the soluble cell lysate by glutathione affinity chromatography. After cleavage by tobacco etch virus (TEV) protease, the complex was further purified by anion exchange and gel filtration chromatography. TIR1 was S-nitrosylated with the stated concentration of GSNO, a NO donor, for 30 min in the dark and then subjected to the biotin-switch assay (Jaffrey and Snyder, 2001) including controls according to Sell et al. (2008) and Forrester et al. (2009). Subsequently, proteins were subjected to western blot analysis using an anti-biotin antibody (Sigma).

Acknowledgements

This work was supported by grants from the NIH (GM43644), the Howard Hughes Medical Institute, and the Gordon and Betty Moore Foundation to ME from USA; Agencia Nacional de Promoción Científica y Técnica (ANPCyT), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) and Universidad Nacional de Mar del Plata (UNMDP) to MCT, CAC and LL from Argentina. MCT, RP, CAC and LL are members of the research staff of CONICET. MJI is a graduate fellow of the same institution. MCT and CAC were recipients of Wood-Whelan and Fulbright fellowships, respectively.

Accession Numbers: The Arabidopsis Genome Initiative accession number for the genes and gene products mentioned in this article are as follows: TIR1 (At3g62980), AFB2 (At3g26810), IAA3 (At1g04240), IAA1 (At4g14560) and IAA5 (At1g15580). The Protein Data Bank accession number for TIR1 is 2p1q.

Ancillary