NO synthesis and signaling in plants – where do we stand?

Nitrate reductase

The primary function of nitrate reductase (NR) is the reduction of nitrate to nitrite in an NAD(P)H-dependent manner (Crawford 1995). In vitro experiments have shown that NR is also capable of reducing nitrite to NO (Dean and Harper 1988, Rockel et al. 2002, Yamasaki and Sakihama 2000). The reduction efficiency is low (∼1% of NR activity); it is more pronounced in a low-oxygen environment and requires nitrite levels to be in excess of the natural substrate nitrate (Rockel et al. 2002). Approximately 20 studies provide evidence supporting NR's role in NO synthesis in planta. For example, it has been demonstrated that NR is responsible for NO production during stomatal closure (Bright et al. 2006, Desikan et al. 2002, Neill et al. 2008), in response to defense elicitors (Shi and Li 2008, Srivastava et al. 2009, Wu et al. 2009), in response to abiotic stress (Sang et al. 2008) and during developmental processes such as flowering (Seligman et al. 2008) or lateral root induction (Kolbert et al. 2008). The evidence for NR's role in NO production in planta is both pharmacological and genetic. Addition of nitrite to plant extracts induces NO formation that coincides with NR activity (Rockel et al. 2002, Sakihama et al. 2002). NR inhibitors like tungstate impair stomatal closure induced by abscisic acid (ABA) in Arabidopsis or by a fungal elicitor in pea (Bright et al. 2006, Srivastava et al. 2009). Furthermore, the NR inhibitors sodium azide and potassium cyanide inhibit NO production in planta (Liu et al. 2007, Sang et al. 2008, Wu et al. 2009). However, these effects alone cannot rule out the involvement of other heme proteins, as these compounds are general heme ligands.

Genetic evidence using NR knockout mutants or silenced plants also supports a role for NR in NO production, as these plants fail to accumulate NO or otherwise mediate NO effects upon elicitation (Bright et al. 2006, Desikan et al. 2002, Seligman et al. 2008, Shi and Li 2008, Yamamoto-Katou et al. 2006). More specifically, one of the two Arabidopsis NR isoforms, NIA1, is the functional NO-producing enzyme involved in ABA-induced stomatal closure (Ribeiro et al. 2009). However, results involving nia mutants should be interpreted with caution. While the lower NO content observed in these mutants could be attributed to the loss of the NO-producing enzyme, it could also correspond to more extensively altered metabolism. In this light, it is not surprising that the Arabidopsis mutant nia1nia2 showed not only a lower nitrite content but also a lower amino acid content, particularly arginine (Modolo et al. 2006). Arginine supplementation did not rescue the nia1nia2 phenotype, contrary to nitrite addition, suggesting that the observed impairment in NO production and resistance to pathogens observed in nia mutants are more likely related to their lower nitrite levels (Oliveira et al. 2009). The production of NO upon nitrite addition to nr/nia mutants suggests that NR is not the only player in nitrite-derived NO formation in planta.

Other nitrite-dependent NO sources

A plasma membrane-bound nitrite:NO reductase, distinct from the plasma membrane NR, has been shown to convert nitrite to NO in tobacco (Stohr and Stremlau 2006). Mitochondrial electron transport-dependent reductase is also capable of reducing nitrite to NO (Planchet et al. 2005). However, these latter routes seem to be localized in the roots of higher plants, where the oxygen tension is low (Gupta et al. 2005). Lastly, non-enzymatic reduction of nitrite is also possible but it requires an acidic pH, thus restricting this reaction to the apoplast of seeds and roots (Bethke et al. 2004).

Plant NOS but not AtNOS1

The search for a plant NOS has been complicated and the identity of this protein remains elusive. Initially, a variant form of the P protein of the glycine decarboxylase complex was described as a potential plant NOS but when this activity was found to be non-reproducible and unreliable, the paper was retracted (Chandok et al. 2003). In 2003, Crawford et al. identified a putative Arabidopsis NOS gene, AtNOS1, through characterization of an Arabidopsis mutant that was defective in NO accumulation in roots (Guo et al. 2003). AtNOS1 encodes a protein with sequence similarity to a hypothetical snail NOS, or NOS partner, that cross-reacts with mammalian NOS antibodies. Several groups have now shown independently that this protein is not an NOS per se but might be associated with NO accumulation (Crawford et al. 2006, Moreau et al. 2008, Zemojtel et al. 2006a). For this latter reason, AtNOS1 has been renamed NO-associated protein 1 (AtNOA1).

AtNOA1 belongs to a family of circularly permuted GTPases (cGTPases) and has been shown to hydrolyze GTP to GDP (Moreau et al. 2008). All members of the bacterial cGTPase family contain RNA/ribosome binding domains (Anand et al. 2009, Daigle and Brown 2004, Matsuo et al. 2006, Saveanu et al. 2001). Because the physiological defect of atnoa1 mutant is complemented by overexpression of the bacterial closest homolog YqeH (43.6% similarity), AtNOA1 also might function by binding RNA/ribosomes (Sudhamsu et al. 2008).

The relationship between NO accumulation and AtNOA1 function is currently unclear. Atnoa1 displays lower NO accumulation under several stress conditions (Bright et al. 2006, He et al. 2004, Zeidler et al. 2004, Zottini et al. 2007) but accumulates as much NO as wild-type plants in other situations (Arnaud et al. 2006, Besson-Bard et al. 2009, Bright et al. 2006, Kolbert et al. 2008, Shi and Li 2008, Shi et al. 2009, Tun et al. 2008). In addition, the atnoa1 mutant was identified in a screen for resistance to fosmidomycin, an inhibitor of the isoprenoid biosynthetic pathway, where it was designated rif1. The rif1/atnoa1 mutant displayed a chloroplastic protein synthesis defect and elevated expression of methylerythritol phosphate (MEP) pathway enzymes (Flores-Perez et al. 2008). Exogenous NO application partially rescued the pale yellowish phenotype of atnoa1/rif1, but did not alter the other physiological traits associated with this mutant (e.g. chloroplastic protein synthesis, MEP pathway regulation). As the main argument in favor of AtNOA1 being involved in NO production is based on the ability of an NO donor to complement the mutant phenotype, this incomplete chemical complementation might indicate that the connection between NO and AtNOA1 is indirect. Whereas the localization of the protein is unclear, the two possible locations chloroplastic and mitochondrial that have been reported so far (Flores-Perez et al. 2008, Guo and Crawford 2005) are both associated with electron transfers, which if not properly controlled can lead to ROS production. Based on structural and complementation data, it was proposed that a defect in organellar function in atnoa1, caused by disruption of ribosome/RNA binding, could lead to a high content of NO-scavenging ROS thus lowering NO bioavailability (Moreau et al. 2008, Sudhamsu et al. 2008).

AtNOA1's homologs might also participate in proper regulation of NO formation in other organisms. In plants, NO accumulation depends on the presence of an AtNOA1 homolog in Nicotiana benthamiana and ortholog in rice (Kato et al. 2008, Qiao et al. 2009). In mice, mAtNOA1 (23.2% similarity) is localized in mitochondria and, as such, seems to regulate mitochondrial NO production and tyrosine nitration, a hallmark of concomitant NO and ROS production (Parihar et al. 2008a,b, Zemojtel et al. 2006b). A human homolog, associated with the inner mitochondrial membrane, was shown to interact with complex I and to be a positive regulator of apoptosis (Tang et al. 2009). Finally, a close homolog in diatoms, PtNOA (44% similarity), is located in chloroplasts, contributes to increased NO production and sensitizes the diatom to stress-inducing aldehydes when overexpressed (Vardi et al. 2008).

S-nitrosylation of cysteine residues

The majority of all NO-affected proteins seem to be regulated by S-nitrosylation, which occurs by oxygen-dependent chemical reactions or by the transfer of NO from a nitrosothiol to a sulfhydryl group of a cysteine residue (trans-nitrosylation). In animals and/or bacteria, NO-mediated S-nitrosylation has been shown to regulate enzymatic activities, ion channels and transcription factor activity. Furthermore, S-nitrosylation can modify cellular signaling by influencing potential protein–protein interactions (Hess et al. 2005).

Research into the role of S-nitrosylation in plants was sparked by the identification of putative candidates for this NO-dependent redox modification. More than 100 Arabidopsis proteins that are targets for S-nitrosylation have been identified from cell culture extracts treated with the NO donor S-nitrosoglutathione (GSNO) or from NO-treated plants (Lindermayr et al. 2005). Additionally, 16 S-nitrosylated proteins have been identified in Arabidopsis undergoing a hypersensitive response to pathogens (Romero-Puertas et al. 2008).

Analyses of knock-out mutants that are impaired for GSNO metabolism have further demonstrated the importance of nitrosothiols in plant disease resistance, thermotolerance and plant growth (Chaki et al. 2009, Feechan et al. 2005, Lee et al. 2008). GSNO is thought to be a form of NO storage and transport. GSNO reductase (GSNOR) is able to breakdown GSNO and thus regulates the cellular level of nitrosothiols (Liu et al. 2001). Consequently, plants lacking GSNOR activity contain higher levels of nitrate and nitroso species. Interestingly, these plants are compromised for acclimating to heat, fail to grow on nutrient plates and exhibit increased reproductive shoots and reduced fertility (Lee et al. 2008). Furthermore, GSNOR knock-out plants display compromised plant defense responses, while plants with enhanced GSNOR activity are more resistant to virulent microbial pathogens (Feechan et al. 2005). GSNOR activity in these plants directly correlates with the accumulation of salicylic acid (SA), a critical signal for disease resistance, and the expression of a defense-associated gene, pathogenesis-related-1 (PR-1), following pathogen infection. Thus, GSNOR appears to be a positive regulator of the SA network and thereby influences disease resistance. In contrast with this result, Rusterucci et al. reported that Arabidopsis with reduced GSNOR expression due to antisense technology showed enhanced basal resistance against Hyaloperonospora arabidopsidis; this correlated with higher levels of intracellular SNOs and constitutive activation of PR-1 (Rusterucci et al. 2007). Furthermore, systemic acquired resistance was impaired in plants overexpressing GSNOR and enhanced in the antisense plants. While the disparity between the results of these two studies has yet to be resolved, these findings reinforce the physiological importance of fine-tuning SNO levels in plants and indicate that S-nitrosylation is as an important redox-dependent regulation mechanism in plants as in animals.

In addition to this more general survey about the importance of SNO in plants, several detailed analyses of S-nitrosylation of specific proteins have been published, including studies on hemoglobin 1, S-adenosylmethionine synthetase, metacaspase, Rubisco, non-expressor of PR-1 (NPR1), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), peroxiredoxin II E (PrxIIE) and the SA-binding protein 3 (Abat et al. 2008, Belenghi et al. 2007, Grun et al. 2006, Holtgrese et al. 2008, Lindermayr et al. 2006, Perazzolli et al. 2004, Romero-Puertas et al. 2007, Tada et al. 2008, Wang et al. 2009b).

Although GAPDH is part of the glycolytic pathway, it also fulfils important regulatory functions. In the animal system, S-nitrosylated GAPDH stabilizes an E3 ubiquitin ligase (Siah1). This complex is translocated into the nucleus, where Siah1 degrades nuclear proteins which leads to apoptosis (Hara et al. 2005). Similar events–including S-nitrosylation and nuclear translocation–have also been observed for cytosolic plant GAPDH, suggesting that S-nitrosylation of this enzyme could also play a role in cell death regulation (Holtgrefe et al. 2008).

The activity of PrxIIE is also regulated by S-nitrosylation (Romero-Puertas et al. 2007). Plant PrxIIE is able to degrade peroxynitrite, which is the major toxic reactive nitrogen species in cells and is responsible for lipid oxidation and tyrosine nitration. S-nitrosylation of PrxIIE inhibits its peroxynitrite detoxification activity, resulting in increased peroxynitrite-dependent formation of nitrotyrosine residues. Thus, NO regulates the effect of its own reactive species through S-nitrosylation of crucial components of the antioxidant defense system that function as common triggers for ROS- and NO-mediated signaling events.

Another very interesting example of the regulatory function of NO is S-nitrosylation of AtSABP3 (Wang et al. 2009b). AtSABP3 exhibits SA binding and carbonic anhydrase (CA) activities, and S-nitrosylation of AtSABP3 at Cys280 suppresses both. Interestingly, only the CA activity is required for establishing disease resistance, suggesting that inhibition of CA activity could contribute to a negative feedback loop that modulates plant defense responses.

NPR1 is a master positive regulator of SA-mediated defense genes, which are known to be redox-regulated (Fobert and Despres 2005). In uninfected cells, this protein is located in the cytoplasm as an oligomer stabilized by intermolecular disulfide bridges. SA-mediated activation of defense responses leads to changes in the cellular redox status resulting in the reduction of NPR1 to its active monomer form. In the monomeric form, NPR1 is translocated into the nucleus, where it interacts with the reduced form of the transcription factor TGA1 (Pieterse and Van Loon 2004). Surprisingly, S-nitrosylation of Cys156 of NPR1 by GSNO facilitates its oligomerization, maintaining protein homeostasis upon SA induction (Tada et al. 2008). These authors concluded that this highly specific modification of a distinct cysteine residue results in conformational changes to NPR1, which allows disulfide bridge formation.

Tyr nitration

Formation of 3-nitrotyrosine in proteins results from the substitution of a tyrosine hydrogen atom for a nitro group. In animals, this modification affects a small number of proteins and a few tyrosines within each protein (Souza et al. 2008). NO and superoxide are required for nitrotyrosine formation. They either rapidly react together to yield the oxidant and nitrating agent peroxynitrite or decay into NO₂ and H₂O₂, respectively, and, together with heme peroxidases, mediate tyrosine oxidation and nitration (van der Vliet et al. 1997). The addition of the nitro group decreases tyrosine's pK_a nearly 3 pH units and enhances tyrosine hydrophobicity. These changes can lead to either a loss of function, such as with GSNOR (Savvides et al. 2002) or a gain of function, as with cytochrome c (Cassina et al. 2000). The pK_a shift also results in signal modulation, as it decreases the extent of tyrosine phosphorylation (Gow et al. 1996).

Tobacco nitrite reductase antisense plants were the first to display a high content of protein nitration (Morot-Gaudry-Talarmain et al. 2002). The non-symbiotic hemoglobins in Arabidopsis (AtGLB) are the only plant proteins identified as Tyr nitration targets. After treatment with nitrite, AtGLB can catalyze its own nitration and nitrate other proteins as well (Sakamoto et al. 2004). A growing body of evidence now shows that Tyr nitration occurs in plants following biotic or abiotic stresses (Chaki et al. 2009, Corpas et al. 2008, Romero-Puertas et al. 2007, Saito et al. 2006, Valderrama et al. 2007). The identity of the nitrated proteins is unknown, but the majority have molecular weights between 20 and 70 kDa. Plant–pathogen interactions are prone to the concomitant release of NO and ROS, thus creating conditions favorable for nitration. Interestingly, in the sunflower–mildew pathosystem, only the susceptible cultivar experiences an increase in protein nitration (Chaki et al. 2009). The authors thus propose that Tyr nitration constitutes a molecular marker for nitrosative stress in plants, as it does in animals. Recent evidence even suggests that Tyr nitration might be a reversible modification, thereby serving as a signaling mechanism (Souza et al. 2008).