Since nearly three decades, the importance of nitric oxide is well recognized in animal systems. In animals, NO plays an important role as neurotransmitter, and is involved in the regulation of immunological responses, muscle relaxation, oxygen sensing and respiratory energy production (Schmidt and Walter 1994; Ignarro 2000, Hagen et al. 2003, Berchner-Pfannschmidt et al. 2010). Also the occurrence of NO in bacteria is well investigated (Sudhamsu and Crane 2009). At present, the free-radical NO is well established as a key signaling molecule and nitric oxide research became a hot and challenging topic also in plant biology. Since the discovery of a role of NO in plants at the very end of the 20th century (Delladonne et al. 1998, Durner et al. 1998) the role of NO in plant development, metabolism and disease responses has been studied extensively. Examples of processes shown to be regulated by NO include seed germination, root growth, respiration, stomatal closure and adaptive responses to biotic and abiotic stresses (Besson Bard et al., 2009; Borisjuk et al. 2007, Moreau et al., 2010). Despite these discoveries, our understanding of the mechanisms underlying the biosynthesis of NO in plants and the signaling activities governed by it are still rudimentary. In this review, we describe the current knowledge on NO biosynthetic pathways in plants and discuss the functional aspects of NO signaling in relation to the various production pathways in a plant cell.
Plants posses various pathways for NO synthesis
In plants, several metabolic pathways for the production of NO from various substrates have been described during the past decades. NO biosynthetic pathways can be classified as either oxidative or reductive in operation. The well-documented routes via nitrate reductase (NR), and mitochondrial or plasmamembrane associated NO production are all reductive reactions, whereas NO production from L-arginine, or polyamine are oxidative routes. Also the recently discovered NO-production pathway that uses hydroxylamine as substrate is oxidative in nature (Rümer et al. 2009a).
All reductive pathways that lead to NO production known by now depend on nitrite as primary substrate. Nitrite on its turn is primarily made from nitrate by NR and therefore, reductive NO formation is generally assumed to depend on NR activity. However, NO production has been demonstrated in plants that lack of either measurable NR activity (Corpas et al., 2009; Rümer et al., 2009a) or NR plus NOS activtiy (Lozano-Juste and Leon 2010) indicating that other reaction pathways must exist. Although clear evidence was brought up for the existence of such alternative reactions in vivo, no enzymes have yet been identified that could catalyze this. The current knowledge about the various biosynthetic pathways that generate NO is summarized in figure 1 and described in detail below.
Reductive pathways
Nitrate reductase (NR)
The best characterized production pathway for NO in plants is the nitrate reductase (NR) pathway, which is localized in the cytosol and primarily catalyses the reduction of nitrate to nitrite using NADH as major electron donor. Apart from this reaction, NR can also catalyze the reaction from nitrite to NO (Yamasaki and Sakihama, 2000, Rockel et al., 2002) via the reaction NAD(P)H + 3H3O+ + 2NO2- → NAD+ + 2NO + 5H2O. However, under normal growth conditions the percentage of in vivo activity of NR involved in nitrite reduction is estimated to be about 1% of the nitrate reducing capacity only (Rockel et al., 2002; Planchet et al., 2005). Biochemical and genetic approaches using either NR-deficient nia double mutants with reduced nitrite and NO content or Nitrite reductase (NiR) antisense lines that accumulate both nitrite and NO (Morot-Gaudry Talarmain et al., 2002; Planchet et al., 2005) indicate that NR is an important source of enzymatic NO production in plants (Botrel et al. 1996).
Various parameters are known to affect NO production via NR. Firstly, NO synthesis by NR requires low nitrate concentrations but the accumulation of nitrite as the Km-value for nitrite (100mM) is relatively high as compared to the inhibition constant (Ki) of nitrate (50mM) (Rockel et al. 2002). Second, NR is activated by a decrease in the cellular pH (Kaiser and Brendle-Behnisch 1995). Notably, the production of NO at low pH is even further stimulated as the plastidial NiR is inhibited, leading to the accumulation of nitrite (Ferrari and Varner 1971). Furthermore, posttranslational modifications of the NR enzyme were shown to affect NO production both in vitro and in vivo. Phosphorylation of a conserved serine in the NR enzyme by NR-kinase prepares NR for binding to 14-3-3 proteins which in a magnesium and calcium dependent manner ultimately leads to inactivation and the proteolytic degradation of NR (Lillo et al. 2004).
NR-mediated NO production is induced by various (a)biotic factors, such as chemical elicitors from fungal plant pathogens (Shi and Li 2008, Srivastava et al., 2009; Yamamoto-Katou et al. 2006), osmotic stress (Kolbert et al., 2010) water stress (Sang et al., 2008), floral transition (Seligman et al., 2008), auxin induced lateral root formation (Kolbert et al., 2008) and hypoxia (Benamar et al. 2008). Various roles have been proposed for NR mediated NO. Upon pathogen infection NO likely acts as a signaling molecule that is required to induce defence mechanism. Upon water-deficit, a role for NO production by NR was demonstrated for the induction of Crassulacean acid metabolism (CAM) in pineapple . (Freschi et al. 2010) as well as water stress induced antioxidant defence mechanisms (Sang et al. 2008). The closure of stomata appeared, among other factors, to be induced by the production of NO (Neill et al., 2008). In addition to a signaling function for NO, it is discussed that during hypoxia NO could inhibit the activity of the respiratory enzyme Cytochrome-c oxidase, thereby regulating the rate of oxygen consumption and the plant internal oxygen concentration during hypoxic stress (Benemar et al. 2008, Gupta et al. 2009).
Plasma membrane-bound Nitrite:NO reductase (NiNOR)
Nitrite derived NO production was also determined in membrane fractions isolated from tobacco roots, but interestingly no NO production was detected in membranes isolated from leaves (Stöhr et al. 2001). Native size-exclusion chromatography separated the protein fraction that contained plasmamembrane-bound NR protein from the heavier fraction that showed NO synthesis activity. Therefore, the existence of a root specific, membrane-bound NiNOR enzyme was suggested. However, the molecular identity of NI-NOR has still to be revealed.
Enzyme activity studies revealed that NAD(P)H is probably not the source for electrons of the reduction reaction catalyzed by NiNOR. In vitro, the artificial electron donor methyl viologen as well as cytochrome-c stimulated the NO production (Stöhr et al. 2001), but it is not likely that cytochrome-c acts as primary electron donor in vivo (Stöhr and Stremlau 2006). Therefore, the true source for electrons has to be revealed yet. The nitrite as substrate for NiNOR is likely to be directly provided by the plasmamembrane-bound NR.
The activity of NiNOR has its optimum at pH 6.1 which is close to the actual pH in the apoplast (Stöhr et al. 2001). The NiNOR capacity under these conditions is expected to be sufficient to reduce all nitrite that is likely to be produced in the apoplast by the plasmamembrane-bound NR. The NO that is produced so can easily enter the cell and trigger various processes that would allow the plant to react on signals from outside (Figure1) Probably, the plasmamembrane-bound NR:NiNOR system is involved in the sensing of nitrate availability in the soil (Meyer and Stöhr 2002). Furthermore, evidence is provided that NiNOR mediated NO production plays a role in the regulation of root infection by mycorrhizal fungi (Moche et al. 2010)
Xantine oxidoreductase in plant peroxisomes
Nitrite reduction to NO can also be catalyzed by the peroxisomal enzyme Xantine oxidoreductase (XOR). Redox regulation of XOR can reversibly change the function of the enzyme from a Xantine dehydrogenase (XDH) into a Xantine oxidase (XOD) (Corpas et al. 2008). The predominant reaction products of XOD under aerobic conditions are uric acid and superoxide. Under anaerobic conditions, purified XOD from bovine milk mainly reduces nitrite to NO using NADH or xanthine as reducing substrate (Godmer et al. 2000). In pea leaves, its activity is strongly focused within the peroxisomes and a strong interaction between the production of reactive oxygen and reactive nitrogen species has been discussed (del Rio et al. 2004). This enyme probably play role in phosphate deficiency induced cluster root formation in white lupin (Wang et al., 2010)
NO production in mitochondria
In the absence of molecular oxygen, mitochondria have been shown to exhibit nitrite reductase activity and thereby become a significant source of NO in various organisms, such as mammals, yeast, algae and plants (Shiva 2010; Tischner et al., 2004; Planchet et al. 2005). In some fungi, a mitochondrial nitrite reductase uses nitrite as terminal electron acceptor of the mitochondrial electron transport chain thus providing a way to generate ATP in the absence of oxygen (Tielens et al 2002). Interestingly root mitochondria also able to produce ATP under anoxic conditions by nitrite driven mechanism. However, the inhibition of nitrite-dependent ATP production under anoxia by the respiratory inhibitors myxothiazol and potassium cyanide (KCN) indicates that nitrite reduction is likely to occur by complexes of the mitochondrial electron transport chain downstream of ubiquinol (Stoimenova et al. 2007). In contrast to the inhibitory effect of myxothiazol, mitochondrial nitrite reduction appeared to be insensitive to antimycin-A.. Both drugs interact with different sites of Cytochrome-c reductase (Complex III) and for mammals it was shown that under anoxia the blockage of the ubiquinone cycle in Complex III by antimycin-A could indeed be bypassed by a nitrite reduction step (Kozlov et al. 1999, Muller et al. 2002), whereas inhibition by myxothiazol blocks the electron transfer from ubiquinol to the iron-centre of Complex III thereby preventing any subsequent redox reaction of the complex. Furthermore, also Cytochrome-c oxidase from yeast and rat mitochondria revealed nitrite reductase activity under hypoxia (Castello 2006). Interestingly, this nitrite reductase activity increased when the pH decreased which is a physiological state when cells imposed to hypoxia
The kinetic parameters of mitochondrial NO-production determine that this can only occur at very low oxygen concentrations, when respiration is already limited by the oxygen concentration. The affinity for nitrite of the mitochondrial nitrite reductase reaction is rather poor (Km = 175 µM), indicating that the reaction requires a significant accumulation of nitrite, which occurs for example during hypoxic stress due to the inhibition of the plastidial NiR. Moreover, the oxygen inhibition coefficient for mitochondrial NO production in plants is very low (Ki oxygen = 0.15mM), and close to the Km oxygen of Cytochrome-c oxidase (Gupta et al. 2009).
Oxidative pathways
Existence of Nitric oxide synthase (NOS) remains elusive
In the late 1980s, it was shown for animal cells that arginine oxidizes into citrulline and NO by the enzyme NO synthase (NOS; Palmer et al., 1987). After the discovery of a role for NO in plants at the very end of the last century (Delledonne et al. 1998; Durner et al. 1998) many researchers started to search for NOS activity in plants even though the Arabidopsis genome did not reveal any gene with significant homology to animal NOS. Since then, many studies provided evidence for the involvement of arginine derived NO in the regulation of various processes such as development (Corpas et al., 2006; Wang et al., 2009a; Corpas et al., 2009 Wang et al 2010), cadmium stress (Besson Bard 2009b; De Michele 2009), pathogen responses (Asai et al., 2009, Bessen Bard et al., 2008, Delledonne et al., 1998) and protection against UV-B radiation (Tossi et al., 2009).
The first claims about the identification of the enzyme responsible for NOS-activity in plant cells were postulated in the late 1990s by determining the conversion of radio-labeled arginine into a product that was expected to be citrulline by a crude enzyme extract from various plant tissues (Cueto et al., 1996; Delledonne et al.,1998, Durner et al., 1998, Bredt and Snyder 1989). Subsequently, antibodies against mammalian NOS enzymes detected immunoreactive proteins in various plant-cell compartments including the cytosol and nucleus (Ribeiro et al. 1999). NOS activity was also demonstrated in peroxysomes (Barroso et al., 1999), in mitochondria (Guo and Crawford 2005) and chloroplasts (Jasid et al., 2006). Furthermore, Corpas et al. (2009) summarized the increasing evidence for L-Arginine dependent NO production in various plant tissues. However, a few years after the first description of plant-NOS, Butt et al., (2003) demonstrated that the plant proteins that were recognized by the antibodies against animal-NOS were in fact not related to animal NOS at all but rather resulted from non-specific cross-reactions of the antibodies. The existence of NOS enzymes in plants became even more disputable when the method to measure NOS activity via the arginine / citrulline assay was seriously questioned as it was shown that enzyme extracts from Arabidopsis do convert arginine into arginosuccinate, rather than to citrulline and NO (Tischner et al., 2007). Also, the purification of pathogen-inducible NOS-like activity in glycine decarboxylase comeplex from virus-infected tobacco (Nicotiana tabacum) leaves (Chandok et al., 2003, 2004) was doubted by the authors themselves shortly and retracted after its first publication due to concerns about the reliability of the published data (John Travis, 2004).
The quest for plant NOS revived with the identification of AtNOS1, an Arabidopsis protein exhibiting 39.5% similarity to a protein from snail that was reported to produce NO in response to hormonal signals (Guo et al., 2003; Zemojtel et al., 2004). Neither the snail protein nor AtNOS1 have homology to any other animal NOS enzyme, but increased arginine dependent NO synthesis was demonstrated when the protein was heterologously expressed in Escherichia coli. These results suggested that AtNOS1 might belong to a novel class of NOS enzymes that is divergent from its animal homologues. Unfortunately, the detection of NOS activity of the recombinant atNOS1 protein could not be reproduced, probably due to the artifacts induced by the arginine / citrulline assay as discussed before (Crawford et al., 2006; Zemojtel et al., 2006a; Moreau et al., 2008, Tischner et al., 2007). Furthermore, the absence of NOS activity in bacterial homologs of the AtNOS1 protein (Sudhamsu et al., 2008) and the inability of AtNOS1 to bind and oxidize arginine and produce NO (Moreau et al., 2008) led to the conclusion that AtNOS1 is in fact not a NO synthase enzyme. Nevertheless, it was shown that AtNOS1 protein is still somehow associated with NO accumulation as the pale green and dwarfish phenotype of Atnos1 knockout lines was rescued by supplying NO exogenously. Therefore, AtNOS1 was renamed to AtNOA1, for NITRIC OXIDE ASSOCIATED PROTEIN1 and it was proven that AtNOA1 hydrolyses GTP to GDP (Moreau et al., 2008).
Apparently, AtNOA1 belongs to a family of circularly permuted GTPases (cGTPase), the function of which is linked to RNA binding to ribosomes (Anand et al., 2006). The intracellular localization of AtNOA1 is unclear, but both chloroplast and mitochondrial localization have been reported from GFP-fusion protein localisation studies (Guo and Crawford 2005, Flores-Perez et al., 2008). In both organelles, highly active electron transport occurs, thus an increased ROS production is likely to result from a defect in organellar function due to the disruption of ribosome biogenesis in a Atnoa1 knockout line. This increased ROS production could explain why an Atnoa1 T-DNA knockout line had a reduced NO accumulation during several kinds of stress conditions (Bright et al., 2006, He et al., 2004, Zeidler et al., 2004, Zottini et al., 2007) as NO can be scavenged via oxidation by ROS (Moreau et al., 2008, Sudhamsu et al., 2008) which would explain why NO levels are reduced in the Atnoa1 line.
Despite the multiple unsuccessful attempts to indentify NOS in plants, numerous reports still point to the presence of arginine dependent, NOS-like activity in plants. Correlations between NO production and the supply of L-Arginine as well as substrates and cofactors of mammalian NOS such as NADPH, FAD, FMN, calcium, calmodulin and tetrahydrobiopterin (BH4) are suggesting NOS-like activity in plants. However, the existence of BH4 in plants is not clear yet, although tetrahydrofolate was detected plants and this could probably form a substitute for BH4. Research to identify such putative precursors for NO in plants would help to better interpret findings about NOS-like activity in plants in the future.
An alternative approach to analyze arginine-dependent NO synthesis is to use L-arginine analogs that are widely used as NOS inhibitors in animals, such as PBITU, AET, L-NAME, L-NMMA, L-NIL. Reduction in NO production after the application of arginine analogues is interpreted to be indicative for the involvement of NOS-like enzymes in specific physiological or developmental process (Corpas et al., 2006, 2009). However, the concentrations at which these analogues are supplied are sometimes very high. For instance in some studies up to 300 mM of L-NAME (Mackerness et al., 2001; Lum et al., 2002) was applied. Interestingly such concentrations are much higher than those regularly used in animal research which normally uses concentrations in the micromolar range (Kubes et al., 1991, Legrand et al., 2009). These high concentrations may be required because of the low uptake rate of the arginine analogues by plant cells or alternatively because endogenous arginine concentrations are already relatively high in plants as compared to animal cells.
Arginine dependent, polyamine mediated NO production
An alternative pathway to produce NO from arginine was suggested by Tun et al. (2006). Arginine is a substrate for the biosynthetic pathway that leads to the production of polyamines such as spermine and spermidine. Exogenous supply of these polyamines to Arabidopsis seedlings provoked immediate release of NO both in the elongation zone of the root tip and in the veins and trichomes of primary leaves. In planta, polyamine synthesis depends on the availability of arginine as substrate for the enzyme Arginine decarboxylase. In transgenic plants in which the arginine availability was affected by either overexpression or knocking out of the genes encoding for the enzyme Arginase (also: Arginine amidohydrolase) the NO production was modulated accordingly: a decrease in Arginase activity resulted in an increased production of NO, whereas upregulation of Arginase activity reduced the release of NO (Flores et al., 2008). The latter phenotype could be rescued by providing spermine to the plants, indicating that the polyamine synthesis from arginine is indeed involved in the production of NO. Nevertheless, the biochemical mechanism by which NO is released from polyamines has not been revealed yet (Yamasaki and Cohen, 2006).
Hydroxylamine mediated NO pathway
To date various biochemical pathways have been suggested to produce NO from hydroxylamine in bacterial and animal systems: either via the enzyme Hydroxylamine oxidoreductase (EC 1.7.3.4) of Nitrosomonas which oxidizes hydroxylamine to NO (Hooper and Terry, 1979) or hydroxylamine reacts directly with superoxide to form NO and also by catalase (Vetrovsky et al. 1996, Craven et al., 1979). In nature, hydroxylamine is described as a possible intermediate of the bacterial nitrification reactions of ammonia to nitrate and nitrite (Lees 1952). Further, it has been hypothesized that hydroxylamine can be an intermediate in the conversion reaction of L-arginine to citrulline (Demaster et al., 1989) Another possibility of hydroxylamine formation is that of the bacterial ammonia oxidation reaction (Hooper et al. 1997).
Exogenous supply of hydroxylamine to NR deficient tobacco cell cultures released considerable amounts of NO under aerobic conditions only, thus providing evidence that hydroxylamine oxidation to NO is feasible in plants as well. The NO formation rates were between 0.2 and 0.8 nmoles g-1 fresh weight h-1 which is several orders of magnitude lower than NO production via NR or the mitochondrial electron transport chain which is between 5-50 nmoles g-1 fresh weight h-1 for both the pathways. Addition of hydrogen peroxide increased the NO production from hydroxylamine whereas catalase reduced NO release, indicating that the reaction is ROS dependent (Rümer et al. 2009a).
Although the biochemical pathway of NO conversion from hydroxylamine apparently exists in plants, its physiological significance remains to be uncovered, as to date the natural occurrence of hydroxylamine in plants has not been unquestionably proven. Hydroxylamine was suggested as intermediate product of the reduction reaction of S-nitrosoglutathione (GSNO) to glutathionedisulphide (GSSG) by GSNO reductase (GSNOR) also known as GSNO terminase, a class III alcohol dehydrogenase homologue, especially in the presence of high concentrations of glutathione (Jensen et al. 1998). Other products that occur at simultaneously, though at very low abundance, are ammonium and nitrite. However, this reaction scheme was disputed by Hedberg et al. (2003). In either case, it is not likely that the GSNOR reaction would ever lead to hydroxylamine concentrations that are suggested to be required to produce considerable amounts of NO in vivo. Given these many uncertainties, it is clear that further research is required to elucidate the role of hydroxylamine in NO production of plants.
Conclusions and future perspectives
Although NO is now generally recognized as an important messenger and signaling compound of various physiological processes also, the fragmentary knowledge about NO biosynthetic pathways in plants still hampers the understanding of NO as regulating agent in plants. By now, only detailed knowledge is available for Nitrate reductase and its role for NO production under various conditions. As long as the other enzymes that are involved in the various reductive and oxidative pathways that lead to the production of NO are not identified and characterized, it will remain very hard to design proper experiments that can tell us more about the various roles of NO as a regulatory compound. Identification of these enzymes should therefore get high priority.
The lack of transgenic plant lines that are affected in their ability to produce NO via the natural production pathways provokes scientists to fall back onto the use of artificial NO donors and synthetic scavengers. A frequently used NO donors is sodium nitroprusside (SNP) (Xiong et al., 2009, , Singh et al., 2009,). However, upon dissociation into NO, SNP also releases cyanide (Betke et al., 2006), which is a potent inhibitor of various enzymes such as NR and Cytochrome-c oxidase. Since both these enzymes are described to be involved in various in vivo NO production pathways, the use of SNP as artificial NO donor does interfere with the endogenous NO reaction as well.
An alternative NO donor is diethylamine/NO (DEA/NO;, , Fu et al., 2010). However, it should be considered that the dissociation DEA into NO is very rapid: the addition of DEA leads to a strong NO burst within seconds to minutes that fades out rapidly as well. In contrast, SNP generates a low abundant but long lasting (up to several hours) production of NO. Unfortunately, this time course effect is often neglected when artificial NO donors are applied, although it is a relevant aspect of NO application studies that should be considered when interpreting the results (Planchet and Kaiser 2006). The importance of local variation in NO signals was nicely demonstrated by a study in which Arabidopsis plants were used in which the NO scavenging enzyme hemoglobin was overexpressed (Perazzolli et al. 2004). The increased abundance of hemoglobin lead to a reduction of the amount of NO as triggered by hypoxic stress, but in contrast the NO burst upon pathogen infection was not affected. Local differentiation, intensity and other free radicals those react with NO signals should therefore get more attention in the future.
Summarizing, multiple NO production pathways are recognized in plants. Evidence exist for the occurrence of both reductive and oxidative NO production. Only for the NR pathway clear molecular knowledge is available about the enzymes involved. Further understanding of the role and regulation of NO as a signaling molecule and biochemical regulator in plants requires a more profound knowledge about the molecular mechanisms involved in the various production pathways of NO.
Figure 1. Schematic diagram illustrating reactions and locations where NO is produced within the plant cell. NO can be produced from nitrite via nonenzymatic or enzymatic pathways, catalyzed by nitrate reductase (NR) Mitochondria electron transport chain, nitrite-NO reductase (Ni-NOR) or via a still uncharacterized L-Arg-dependent pathway that involves a non identified nitric oxide synthase (NOS)-like enzyme, and via an uncharacterized process that uses polyamines (PAs) as substrates.
Box1. Listing of the known characteristics of the various NO production pathways and the processes in which the NO that is produced is involved in.
