JXB Advance Access originally published online on December 23, 2005
Journal of Experimental Botany 2006 57(3):489-505; doi:10.1093/jxb/erj052
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
FOCUS PAPER |
NO way to live; the various roles of nitric oxide in plant–pathogen interactions
1University of Wales Aberystwyth, Institute of Biological Sciences, Aberystwyth, Ceredigion SY23 2DA, UK
2Institute of Grassland and Environmental Research, Aberystwyth, Ceredigion SY23 3EB, UK
3Instituto de Agricultura Sostenible IAS-CSIC; Alameda del Obispo Apdo 4084, E-14080 Córdoba, Spain
* To whom correspondence should be addressed. E-mail: lum{at}aber.ac.uk
Received 15 August 2005; Accepted 11 November 2005
 |
Abstract |
|---|
 Top Abstract IntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
Nitric oxide has attracted considerable interest from plant pathologists due its established role in regulating mammalian anti-microbial defences, particularly via programmed cell death (PCD). Although NO plays a major role in plant PCD elicited in response to certain types of pathogenic challenge, the race-specific hypersensitive response (HR), it is now evident that NO also acts in the regulation of non-specific, papilla-based resistance to penetration by plant cells that survive attack and, possibly, in systemic acquired resistance. Equally, the potential roles of NO signalling/scavenging within the pathogen are being recognized. This review will consider key defensive roles played by NO in living cells during plant–pathogen interactions, as well as in those undergoing PCD.
Key words: Defence, hypersensitive response, nitric oxide, plant–pathogen interactions, programmed cell death
|
Introduction |
|---|
|
TopAbstractIntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
Nitric oxide is emerging as an important signal in plant–pathogen interactions and many recent excellent reviews cover this area (Romero-Puertas et al., 2004
; Wendehenne et al., 2004; Delledonne, 2005). Although most work focuses on the role of NO in the plant hypersensitive response (HR), a localized programmed cell death (PCD) which confines the pathogen to the site of attempted infection (reviewed by Lam et al., 2001; Greenberg and Yao, 2004), NO has other roles in plant defence and within pathogens it may influence their virulence. In this review, current perceptions of roles of NO during cell death and in other aspects of plant–pathogen interactions are outlined. |
NO chemistry and signalling |
|---|
|
TopAbstractIntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
NO acts as a signalling molecule within species from every biological kingdom, a feature reflecting its physical properties which give it an exceptionally rich chemistry (Fig. 1). NO is highly reactive due to the presence of an unpaired electron and, as with oxygen, it can exist in a variety of reduced states, –NO (nitroxyl ion), NO and +NO (nitrosonium ion), with each reactive nitrogen intermediate (RNI) able to undergo specific interactions (Gow and Ischiropoulos, 2001
). ·NO can react with atmospheric O2 producing NO2 giving it a theoretical half-life of 6 s although, at the low ·NO concentrations seen under physiological conditions, this is likely to be extended to 
30 s (Wink et al., 1996). Further, NO is a remarkably mobile molecule being able to move freely across membranes and, with very low solubility in aqueous solutions (partition coefficient: 0.04=4.6 ml per 100 ml H2O at 20 °C, 1 atm.), it can readily enter the gaseous phase. This markedly enhances its potential as a signal.
|
to form the potent oxidant peroxynitrite ( 
is used as the electron donor (described in detail in Radi, 2004). (VI) Reduction of In biological systems, NO affects signalling through a range of actions (Fig. 1). Many NO effects are mediated by oxidative damage associated with the formation of the potent oxidant peroxynitrite via interaction with superoxide
A more subtle action is the electrophilic attack by ·NO on thiol groups, particularly cysteine residues, resulting in S-nitrosylation of molecules such as glutathione or proteins. Protein S-nitrosylation can modulate protein activity; for example, the monomeric GTP binding protein, p21ras, is activated by S-nitrosylation (Lander et al., 1996) while, by contrast, the kinase, JNK1, is inactivated (Park et al., 2000). Alternatively, NO can modify proteins by nitration, particularly of tyrosine residues (Radi, 2004). NO affinity for transition metals (e.g. Fe2+/Fe3+) to form iron–nitrosyl complexes is also exploited to regulate biological activities. This interaction is central to the classical NO signalling pathway where NO binding to the haem groups of soluble and membrane-associated guanylyl cyclase results in a configurational change to activate the enzymic production of cGMP from GTP (Murad, 1994; Denninger and Marletta, 1999). Elevated cGMP increases kinase (NO-G-kinase) and phosphodiesterase activity influences cGMP regulated ion channels. One well-documented target of NO-G-kinase is myosin light chains resulting in smooth muscle relaxation/vasodilation (reviewed by Beck et al., 1999). Another important interaction involves the binding of NO to haem groups within P450 enzymes, and this inhibits their activity (Khatsenko et al., 1993; Takemura et al., 1999). Especially interesting is the interaction of NO with haemoglobin (Hb) as this complex has the ability to deliver, store and to oxidize (detoxify) NO (Gow et al., 1999). Plant haemoglobins have been shown to react with NO to produce NO3 under aerobic conditions in a two-stage process. First, NO oxidizes the haem centre to form methaemoglobin whereupon NO3 is released after reduction by NADPH or a flavin-containing reductase. Such pathways represent mechanisms by which NO levels are rapidly modified in biological systems (Perazzolli et al., 2004).
NO/peroxynitrite can also reversibly disrupt iron–sulphur (4Fe-4S) enzymes via cluster co-ordination leading to iron loss 
which affects enzymatic activity. Well-characterized targets include the Krebs cycle enzyme aconitase which catalyses the isomerization of citrate to isocitrate (Drapier, 1997) and complexes I and II in the electron transport chain (Stuhr and Nathan, 1989). NO suppressed aconitase protein adopts a new role as an iron-regulatory protein by binding to RNAs encoding proteins which utilize or sequester iron, such as ferritin, thereby preventing their translation (Domachowske, 1997). Thus, NO/4Fe-4S protein interaction will increase oxidative stress through mitochrondrial disruption and by promoting the formation of the particularly reactive hydroxyl radical via Fenton reactions (Fe2++H2O2
Fe3++OH–+·OH).
|
NO generation in animals and plants |
|---|
|
TopAbstractIntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
In phytopathological studies, understanding NO generation will allow insight into elicitation mechanisms and will suggest how its production can be modulated by either plant or pathogen. Further, based on the NO-generating system(s), mutants, antagonists, and agonists will become available with which to assess the role of NO in particular plant–pathogen interactions.
In biological systems, NO can be generated enzymatically or non-enzymatically. The most extensively described NO-producing enzymes have been nitric oxide synthase (NOS) and nitrate reductase (NR). NOS (Moncada et al., 1991) catalyses the two-step oxidation of L-arginine to NO and citrulline (L-arginine+NADPH2+O2N
hydroxyarginine+NADPH2+O2L-citrulline+NO), a reaction that might also be catalysed by a cytochrome P450 (Boucher et al., 1992). NR generates NO from nitrite with NADPH as electron donor (Kaiser et al., 2002; Yamasaki and Sakihama, 2000). In mammalian systems, NOS occurs in three isoforms. Neuronal (nNOS, NOSI, 160 kDa) and endothelial (eNOS, NOSIII, 130 kDa) are responsible for low level constitutive production of NO that controls vasodilation and neurotransmission, respectively, and their activity is modulated by Ca2+ concentration (reviewed by Stuehr, 1999). Transcription of inducible NOS (iNOS, NOSII, 130 kDa) is unaffected by Ca2+ but is modulated by peptide hormones produced in response to infection and inflammation (MacMicking et al., 1995).
Much early effort by plant scientists focused on searching for a plant NOS. The enzymic oxidation of L-arginine to yield NO and L-citrulline has been reported in extracts from pea (Leshem and Harmaty, 1996), lupin (Cueto et al., 1996), soybean (Delledonne et al., 1998), tobacco (Durner et al., 1998), and maize (Ribeiro et al., 1999). Competitive inhibitors based on L-arginine have been used to suppress NO production in soybean, Arabidopsis and tobacco (Delledonne et al., 1998; Durner et al., 1998) implicating NOS activity. Further, anti-mammalian NOS antibodies were shown to bind to specific protein in plants (Barroso et al., 1999; Ribeiro et al., 1999) where antigenicity was localized to peroxisomes (Corpas et al., 2001). However, the specificity of these antibodies in plants has been questioned (Beligni and Lamattina, 2001) and, crucially, no gene with homology to mammalian NOS was detected within the Arabidopsis genome. However, Guo et al. (2003) identified a NO-generating enzyme, AtNOS1, which reconciles these apparently conflicting data. Although AtNOS1 (60 kDa) is much smaller than mammalian NOS and exhibits no sequence homology to NOS, it binds to nNOS antibodies and apparently produces NO by oxidizing arginine.
Despite the tentative identification of a plant NOS gene, clear evidence shows that plants can produce ·NO from nitrite via NADPH-dependent nitrate reductase (NR) (NO3NO2·NO+O2). The application of high nitrite levels under conditions of anoxia increased NO production (Rockel et al., 2002) and, most convincingly, NR mutants of Arabidopsis, nia1, and nia2 were perturbed in NO-mediated stomatal closure (Desikan et al., 2002). Furthermore, challenge of double nia1 nia2 mutants with a bacterial pathogen, Pseudomonas syringae pv. maculicola, established that NR was a major source of NO during this plant–pathogen interaction where a lesser contribution originated from NOS-like activity (Modolo et al., 2005). It should also be recognized that NO can be formed through the non-enzymatic reduction of nitrite 
but this is favoured only under acid conditions such as found in the barley aleurone apoplast (Bethke et al., 2004).
|
Die and let live: programmed cell death in plants and animals |
|---|
|
TopAbstractIntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
Perhaps the best-characterized plant response to pathogenic challenge is the HR conditioned by the recognition/interaction of a pathogen-encoded avirulence (avr) gene product with a plant resistance (R) gene product; failure of this recognition system allows disease development. Such interactions have been well studied in the Pseudomonas syringae (P. s. pv.) and Xanthomonas campestris pathogens where the bacterial avirulence protein is one of a large number of proteins delivered into the plant cell. Here, delivery of the avr protein is via a pilus-forming Type III secretion mechanism encoded by hrp genes, and the delivered proteins are collectively designated Hops (hrp outer proteins; reviewed by Bonas and Van den Ackerveken, 1999
; Collmer et al., 2000).
To understand the HR more clearly, many plant pathologists have sought mechanistic commonalities with the well-characterized processes of mammalian programmed cell death (PCD), termed apoptosis, particularly in attempting to resolve the role of ·NO. It is not appropriate here to consider apoptosis in detail; good overviews are provided by Strasser et al. (2000), Boatright and Salvesen (2003) and Jiang and Wang (2004). Briefly, however, apoptosis involves nuclear and cytoplasmic shrinkage and DNA cleavage, ultimately leading to a loss in cell integrity and the formation of apoptotic bodies which are phagocytosed by surrounding cells. Central to this process is the activation of cysteine-dependent aspartate-specific proteases (or caspases) which have a range of cellular targets (reviewed by Creagh et al., 2003). Pre-existing caspases are activated in response to a range of cues, including certain proteins released from disrupted mitochondria. Mitochondrial disruption follows the formation of permeability transition pores (PTP), organelle depolarization, loss of proton motive force, Ca2+ release, and increased oxidative stress. A key mitochondrial protein in caspase activation is cytochrome c which interacts with the chaperone protein Apaf1 to activate caspase 9 within an apoptosome complex (reviewed by Adams and Cory, 2002). NO can influence apoptosis at many points. High ONOO– levels can cause massive oxidative damage, especially to nucleic acids (Messmer et al., 1994), but at lower NO levels the mitochondrion appears to be a particular target for its proapoptotic effects. NO will reversibly bind to the haem group in cytochrome oxidase to inhibit electron transport (Brown and Copper, 1994) leading to an increase in 
which augments ONOO– production and damage (Packer et al., 1996). ONOO– induces a mitochondrial Ca2+ efflux to promote PTP formation, cytochrome c release, and thus caspase activation (reviewed Murphy, 1999).
The plant HR resulting from R/avr interaction shares many cytological features with apoptosis, most strikingly in mitochondrial changes (Lam et al., 2001). Cell death and its associated calcium influxes and oxidative stress results in the formation of PTP and the release of cytochrome c (Gottlieb, 2000; Tiwari et al., 2002). As yet, however, the link between cytochrome c release and the cell death mechanism is unclear because satisfactory plant gene homologues of Apaf1 are lacking (although homology with some R gene motifs has been suggested; van der Biezen and Jones, 1998) and caspases have not been identified. There is, however, evidence of caspase activity in plants (D'Silva et al., 1998) and transgenic plants expressing the caspase inhibitor proteins p35 and Op-IAP show a suppressed HR (Dickman et al., 2001; Del Pozo and Lam, 2003). At least two explanations have been suggested for this apparent contradiction. First, it is possible that metacaspases, which have been detected in plants and have only a distant homology to caspases, could be involved in the HR (Uren et al., 2000). Alternatively, proteases with caspase-like specificity, but dependent on a serine at their active site, ‘saspases’, may be involved (Chichkova et al., 2004). The targets of saspases/metacaspases are unclear, as is how, or indeed if, they are activated by cytochrome c release from mitochondria. van der Hoorn and Jones (2004) provide an excellent review of the range of proteases involved in plant defence.
NO has now emerged, with oxidative stress (Levine et al., 1994), as a major arbiter of plant PCD. NO was first measured in soybean cultures inoculated with avirulent but not virulent bacteria. Further, application of inhibitors of mammalian NOS suppressed HR cell death in Arabidopsis thaliana (Delledonne et al., 1998). Subsequently,·NO production has been shown during the HR elicited in suspension cultures of Arabidopsis inoculated with P. s. pv. maculicola (Clarke et al., 2000) and tobacco cultures challenged with P. s. pv. tomato (Conrath et al., 2004). Most in planta measurements have utilized NO-specific indicator dyes based on fluorescein, notably DAF-2DA (4, 5-diaminofluorescein diacetate) or DAF-FM (4-amino-5-methylamino-2', 7'-difluorofluorescein), which both produce fluorescent DAF-2 triazoles in the presence of NO. DAF dyes have revealed rapid NO production in tobacco peels treated with the necrosis-inducing elicitor cryptogein from Phytophthora cryptogea (Foissner et al., 2000) and within the context of plant–pathogenic interactions. For example, a significant but transient NO burst was observed in barley epidermal cells attacked by the powdery mildew fungus, Blumeria graminis f. sp. hordei just prior to their HR-associated collapse (Prats et al., 2005; Fig. 2). As in many studies, a NO-scavenger, cPTIO (2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide) was used to suppress the fluorescent signal and also delay cell death, suggesting a contribution of NO to the HR process. It should be noted that the reaction product of CPTIO and NO, CPTI (2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazole-1-oxy-3-oxide) itself suppressed cryptogein-elicited cell death in tobacco cultures without scavenging NO (Planchet et al., 2005). Hence, although CPTIO remains valuable in establishing that NO generation is being detected, more than suppression with CPTIO may be required if seeking to correlate a reduction in NO with a physiological effect. Thus, the role of NO in a particular phenomenon, requires confirmation through a multitude of approaches; actual NO measurements, the use of pharmaceutical agents which scavenge NO or suppress NO generation, as well as mutants exhibiting reduced or elevated NO levels. In another approach, genetic evidence of a role for NO in the HR was provided through the expression of a nitric oxide dioxygenase (NOD), encoded by hmp from E. coli in transgenic Arabidopsis (Zeier et al., 2004). NOD catalysed the dioxygenation of NO to nitrate and NOD-expressing transgenic lines challenged with avirulent P. s. pv. tomato avrB showed reduced NO production and, crucially, delayed cell death. Other work shows that treating plant tissues with NO donors initiates chromatin condensation and DNA fragmentation as reported by in situ terminal dioxynucleotide transferase-mediated dUTP nick end labelling (TUNEL, Clarke et al., 2000; Pedroso et al., 2000). Further, the initiation of NO-mediated cell death can be suppressed with a caspase 1 inhibitor (Clarke et al., 2000), and expression of a cysteine (cystatin-class) protease inhibitor (AtCYS1) in transgenic tobacco suppressed cell death initiated by NO or attack by avirulent bacteria (Belenghi et al., 2003). NO donors also initiate PCD associated features in plant mitochondria including dissipation of the electrical potential and the release of cytochrome c (Casolo et al., 2005). All of these data notwithstanding, other studies have failed to detect NO before cell collapse. For example, Zhang et al. (2003) using DAF-FM detected NO production only after the HR had been elicited by P. s. pv. tomato DC3000 avrB and avrRpt2; this suggested a role for NO in the propagation of cell death but not its initiation. Thus, there appear to be interesting parallels between the roles of NO in HR-associated plant PCD and mammalian apoptosis, although much remains to be understood.
|
It is worth noting here various NO assay methods, and their advantages as well as their limitations that complicate data collection and interpretation. The DAF stains are easily applied in vivo but react with NO+, not NO, to produce the fluorescent triazoles, there are limitations on the sensitivity of the detection methods, and the dyes may accumulate in particular organelles (Foissner et al., 2000
). A widely used alternative is the oxyhaemglobin assay which is based on NO interaction with the haem core to form methanohaemoglobin. Under conditions of oxidative stress this method can overestimate NO levels, but it has proved to be useful for assessing NO production in suspension cultures where anti-oxidant enzymes can readily be applied (Delledonne et al., 1998; Clarke et al., 2000). Electron paramagnetic resonance spectroscopy (EPR) is potentially very useful because it can distinguish between different RNI (–NO/·NO/NO+; Xia et al., 2000), but this method does not easily lend itself to on-line in planta monitoring. Perhaps the most satisfactory option involves mass spectrometry approaches, especially membrane inlet mass spectrometry/capillary inlet mass spectrometry (MIMS/RIMS) which offers a sensitive, on-line technique by which both gaseous and aqueous NO may be assessed (Conrath et al., 2004). It is noteworthy that all of the most-sensitive assays have detected NO in pre-necrotic plant tissue, thus implicating NO in the initiation of the HR.
To illustrate this point, our own in planta NO assay has utilized photoacoustic laser detection to sample NO in the gas phase (Mur et al., 2005b). This method provides a highly sensitive on-line method to assess the emission of NO although it does not assess intracellular NO levels or non-gaseous RNI. The HR elicited by P. s. pv. phaseolicola and P. s. pv. tomato harbouring the avirulence gene avrRpm1 were examined in tobacco (Fig. 3A) and Arabidopsis (Fig. 3B), respectively. In each interaction, NO production was initiated rapidly, between 30–45 min following bacterial inoculation. This early NO burst immediately preceded generation of H2O2 and occurred some 6 h before the first visible signs of cell death (Mur et al., 2005b). Further, no significant NO production was detected in plants that were inoculated with hrp mutant derivatives of the avirulent bacteria. As hrp mutants are unable to deliver avr protein into the plant, this suggested that the rapid NO production is avr-R dependent. This contrasts with the pathogenesis-associated calcium and ROS fluxes that are first induced by non-avr elicitors before a second avr-R dependent burst (Fig. 4; Lamb and Dixon, 1997; Grant et al., 2000). Further, as hrp genes are not expressed in the rich media used to grow these bacterial strains, the 30–45 min (Brown et al., 2001) interval before the NO burst would represent the period required for in planta hrp gene induction, hrp pilus assembly and avr delivery. Expression of hrpA, which encodes the pilus protein, occurs in conjunction with that of hrpZ (‘harpin’) which forms pores in the plant lipid bilayers and is one of the first Hops to be delivered (Lee et al., 2001). Harpin has proven to be an effective initiator of NO production (Krause and Durner, 2004), but early NO generation elicited by a disease forming strain of P. s. pv. tomato (and also producing hrpZ) did not match that of the strain harbouring avrRpm1 (Fig. 3B). Hence, it is suggested that the rapid NO burst reflects avr-R interaction whilst the NO produced during disease in this pathosystem results from HrpZ detection, possibly with a contribution by other Hop elicitors (Fig. 3C). Bennett et al. (2005) also suggested that NO is specific marker for avr-R interactions, noting that the avr-R-dependent emission of biophotons in Arabidopsis was suppressed by NO scavengers.
|
|
Accumulating evidence indicates that NO can affect the generation of ROS, which in plants is closely associated with the HR. Thus, Zeier et al. (2004)
demonstrated that the pathogen-challenged NOD transgenic lines exhibited not only reduced NO generation, but also a reduced oxidative burst. H2O2 acting alone affects mitochondrial function, including cytochrome c release (Tiwari et al., 2002), and induces chromatin condensation (Houot et al., 2001). Thus NO modulation of the oxidative burst would represent an additional mechanism by which the HR cell death could be affected. Although ROS may be generated from many sources, it seems likely that an NADPH oxidase is a major contributor to ROS generated during the HR (Torres et al., 2002). The NADPH oxidase is regulated by calcium influxes (Levine et al., 1996) or the action of a monomeric GTP-binding protein (MGBP, Ono et al., 2001), and both are likely sites for NO action. For example, the activity of the mammalian MGBP p21ras is regulated by S-nitrosylation (Lander et al., 1996) and in plants, cryptogein-elicited NO initiates the release of Ca2+ from intracellular stores as well as activating an oxidative burst (Lamotte et al., 2004). NO will also elevate oxidative stress by binding to the haem centres and inhibiting the activity of the antioxidant enzymes catalase and ascorbate peroxidase (Clarke et al., 2000). The effects of increased oxidative stress would be further augmented by NO inhibition of plant aconitase to increase iron levels within the cytoplasm (Navarre et al., 2000).
Interestingly, NO can influence apoptosis at many points but can be either pro- or anti-apoptotic (see below) depending mainly on its concentration (reviewed by Kim et al., 1998) or the period of NO generation (Hortelano et al., 2003). Various anti-apoptotic actions for NO include the induction of heat shock proteins (HSP32, HSP70, Kim et al., 1995, 1997a) and by increasing cGMP to suppress the release of Ca2+ from mitochondria (Genaro et al., 1995). A key anti-apoptotic mechanism is the S-nitrosylation of the essential cysteine within the caspases active site (Dimmeler et al., 1997; Kim et al., 1997b; Li et al., 1997; Mannick et al., 1999; Rossig et al., 1999). Similarly in plants, there have also been reports of NO suppressing cell death (Bethke et al., 2004). The relative concentrations of particular ROS and NO appear to be vital in the initiation or suppression in cell death. Delledonne et al. (2001) described results showing that conditions favouring the accumulation of superoxide over H2O2 or an excess of NO would reduce cell death. Application of NO donors has been shown to induce the expression of protective genes, such as alternative oxidase, which may aid in preventing the generation of ROS. The transcription of small HSPs is also increased by NO and these could protect key proteins through their role as chaperones (Krause and Durner, 2004). Alternatively, as lipid peroxidation is a feature of cell death in plants (Montillet et al., 2005), the role of.NO could be as an effective breaker of such free-radical chain reactions (R·, ROO· and ROR-NO, ROO-NO· and RO-NO; Beligni and Lamattina, 1999). The mechanisms of cell death suppression by NO will undoubtedly feature in much future research.
|
Live and let live: NO and non-PCD defence |
|---|
|
TopAbstractIntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
Although avr-R mediated events have received the vast majority of attention it is now being recognized that other pathogen-derived elicitors induce a highly effective non-cell death associated defence, variously known as innate, basic or non-specific resistance, though each term carries its own particular implications according to the system involved. In animals, such innate resistance is engaged by a range of bacterially-derived elicitors termed pathogen-associated molecular patterns (PAMPs, Janeway, 1989
) including lipopolysaccharides (LPS) and flagellin. In plant–pathogen interactions a similar range of PAMPs have been implicated in defence responses (Nurnberger et al., 2004). Both avirulent and virulent pathogens will elicit innate resistance, but in the latter case, defences dependent upon avr-R interaction are avoided, suppressed or neutralized. In a seminal paper, Zeidler et al. (2004), implicated NO in innate resistance in plants. Application of LPS to Arabidopsis cultures initiated a near-immediate burst in NO production and, crucially, this LPS-elicited NO production was suppressed in the AtNOS1 mutant. Such LPS-mediated resistance would be exhibited as a component of defences in disease scenarios where avr-R interactions play no role. Indeed, NO production has been observed during establishment of nominally compatible plant–bacterial interactions (Conrath et al., 2004; Mur et al., 2005a) and innate host defences against disease development were compromised in the AtNOS1 mutant and also in tobacco treated with cPTIO and NOS inhibitors (Zeidler et al., 2004; Mur et al., 2005b). Nevertheless, it should also be noted that little NO production was elicited by hrp-compromised bacteria where the extracellular polysaccharides are unaffected, but avirulence and Hop proteins are not delivered into the plant cell (Mur et al., 2005b; Fig. 3A, B). This raises the question of whether Hop-induced NO is of greater significance than LPS during compatible interactions (Fig. 3C). It will be interesting to investigate NO production elicited by bacterial LPS mutants and assess the contribution of other PAMPs, for example, the important defence elicitor flagellin (flg22; Felix et al., 1999), in the elicitation of NO.
The interaction between PAMP and avr/R-conditioned resistance is an area of increasing research (Parker, 2003). Within the context of calcium influxes and the oxidative burst, these interactions apparently lead to biphasic generation patterns with an early but transient PAMP-dependent rise followed by a later, persistent, avr-R-dependent rise (Lamb and Dixon, 1997; Grant et al., 2000). Given the significance of early generation of NO to avr-R interactions (Fig. 3A, B) and the positive role of NO in the generation of calcium fluxes and oxidative bursts (Lamotte et al., 2004; Zeier et al., 2004), NO could acts as an avr/R-dependent input (perhaps as the main input) modulating persistent calcium fluxes and generation of ROIs (Fig. 4).
Many mechanisms of innate resistance are associated with particular patterns of defence genes expression (Zeidler et al., 2004). NO signalling to defence gene expression in plants often involves guanylyl cyclase activation and an increase in cGMP levels (Durner et al., 1998). The NO-induced defence genes PR1 (pathogenesis-related protein 1) and PAL1 have been linked to the synthesis of salicylic acid (SA; Uknes et al., 1992; Mauch-Mani and Slusarenko, 1996) and NO was shown to cause SA accumulation (Durner et al., 1998), but to suppress defence signalling regulated by jasmonic acid (JA; Orozco-Cardenas and Ryan, 2002; Huang et al., 2004). SA plays a central role in HR-mediated resistance (Delaney et al., 1994; Cao et al., 1997; Mur et al., 1997, 2000) partially through the induction of PR proteins but also through the potentiation of a wider range of defence genes and the oxidative burst (Kauss and Jeblick, 1995; Mur et al., 2000; Conrath et al., 2002). By initiating SA synthesis, NO is likely to participate in this potentiation mechanism (Van Camp et al., 1998, Mur et al., unpublished results). However, SA-mediated defence is not effective against all plant–pathogenic organisms. For example, defence against tissue-macerating necrotrophic pathogens is mediated by JA (Thomma et al., 1998) and, predictably, resistance against the necrotrophic fungus Botrytis cinerea, was compromised by the application of NO donors (Ma
olepszá and Ró
alska, 2005).
SA was first characterized as a key signal in the establishment of systemic acquired resistance (SAR; reviewed by Durrant and Dong, 2004). SAR is a form of whole-plant ‘immunity’ that is exhibited against a wide range of pathogens and is independent of avr-R gene interaction and could be considered as an extended example of induced innate resistance. SAR has been linked to the systemic expression of genes such as PR proteins and the potentiation of plant defence (Mur et al., 1996). It was first thought that the dispersal of SA through the phloem led to the establishment of SAR. However, although SA accumulation both at sites of the HR and systemically is important, the nature of the mobile signal remains obscure. Amongst the possible candidates are NO derived S-nitrosylated proteins or S-nitrosoglutathione which acts as a long-distance signal in blood (Jia et al., 1996) and hence could act as the mobile SAR signal (Durner and Klessig, 1999). As evidence for this, localized application of a range of NO-donors, including S-nitrosoglutathione conferred SAR against TMV through an SA-dependent mechanism (Song and Goodman, 2001). It is possible that S-nitrosoglutathione formed in the vicinity of the HR is loaded into phloem, systemically dispersed, and unloaded to initiate systemic SA synthesis. This is an attractive model but requires the integration of DIR1, a putative lipid transfer protein, which has been shown to be essential for loading systemic SAR signal(s) in Arabidopsis (Maldonado et al., 2002). To a certain extent this model also goes against the results recently described by Feechan et al. (2005). S-nitrosothiol levels are regulated by S-nitrosoglutathione reductase (GSNOR) and, accordingly, a T-DNA mutation within an Arabidopsis GSNOR (atgnor1-3) shows elevated concentrations of S-nitrosothiols. However, this was not associated with increased resistance, in fact, the virulence of P. s. pv. tomato and wheat powdery mildew (B. graminis f. sp. tritici) was increased and SA levels were reduced (Feechan et al., 2005). These data would argue for S-nitrosothiols suppressing plant defence and presumably SAR, however, given the data from Song and Goodman (2001), it seems likely that further analysis will reveal particular roles for specific S-nitrosothiols which could include defence activation.
One very distinctive example of innate resistance is the formation of papillae. Papillae are localized apoplastic wall appositions induced in a non-avr/R dependent manner in response to pathogen attack and they confer broad-spectrum penetration resistance. Papillae have been extensively studied as a response to fungal pathogen attack, particularly in barley (Hordeum vulgare) challenged with the powdery mildew fungus, B. graminis f. sp. hordei (Zeyen et al., 2002). Germinating B. graminis spores first form a short-primary germ tube (PGT) that attaches rapidly to the plant epidermal surface but does not penetrate the host cell wall (Edwards, 2002). Subsequently, a second tube emerges, elongates, and then differentiates a swollen apical appressorium from beneath which a penetration peg emerges and attempts to breach the host epidermal cell wall. If penetration succeeds, the tip of the peg enters the cell lumen and differentiates a specialized digitate feeding structure, the haustorium (Green et al., 2002). Contact with both the PGT and appressorium can stimulate the formation of a host cell papilla subadjacent to the germ tube tip (Fig. 5A). Papillae are chemically complex, containing callose, proteins, and autofluorescent phenolic compounds. Their formation follows reorganization of the cell cytoskeleton and deposition involves oxidative cross-linking and immobilization of constituent compounds via the localized production of H2O2: this has been considered one of the earliest features of their construction (reviewed by Collins et al., 2002; Zeyen et al., 2002; Hückelhoven and Kogel, 2003). Recently, however, Prats et al. (2005) observed a significant NO burst associated with papilla deposition sites (Fig. 5B) which occurred well before the accumulation of autofluorogenic compounds and even earlier than others have detected H2O2 generation (Thordal-Christensen et al., 1997; Vanacker et al., 2000). Crucially, suppression of this NO production led to increased fungal penetration i.e. increased susceptibility to attack (Prats et al., 2005). Similarly, the use of actin-polymerization inhibitors in Arabidopsis (Yun et al., 2003) and cowpea (Mellersh et al., 2002) has indicated that cytoskeletal arrangement also plays a central role in both H2O2 accumulation and callose deposition. Recently a series of breakthrough papers have shown the key role played by SNARE (soluble N-ethylamalemide-sensitive factor receptor) complexes in papilla formation (Collins et al., 2003; Assaad et al., 2004). SNARE proteins located on vesicle and target membranes mediate correct vesicular trafficking (reviewed by Bock et al., 2001). H2O2 and, probably, other papilla components are delivered to sites of papilla formation within vesicles. In Arabidopsis and barley, respectively, PEN1 and ROR2 proteins are located at the plasma membrane and are likely to interact with a vesicle-located vSNARE protein, SNAP25 (a SNAP34 homologue) to promote fusion (Collins et al., 2003). Interestingly, in mammalian systems, ONOO–-mediated nitration of SNAP25 was shown to promote SNARE complex formation (Di Stasi et al., 2002; Fig. 5C). It is also possible that NO could influence this process at other points as S-nitrosylation regulates other mammalian SNARE complex components, not as yet implicated in plant papilla formation. These include nitrosylating the Sec1 chaperone to cause release of free SNARE for further interaction (Meffert et al., 1996) and suppressing activity of N-ethylamalemide-sensitive factor (NSF) which separates components of the SNARE complex (Matsushita et al., 2003).
|
An interesting question is posed by the tight association of DAF2-DA fluorescence at sites of papilla formation (Fig. 5B). Given the free diffusibility of NO, as seen for instance during cells exhibiting incipient HR (Fig. 2A), this appears to be an odd observation. It may be that, as with H2O2 (Yun et al., 2003
), the site of NO generation is influenced by cytoskeletal organization. Significantly, in mammalian phagocytes, NOS is located within vesicles which co-localize with the cortical actin lying immediately adjacent to the peripheral cell membrane (Webb et al., 2001). |
NO and the pathogen |
|---|
|
TopAbstractIntroductionNO chemistry and signallingNO generation in animals...Die and let live:...Live and let live:...NO and the pathogenFuture targetsReferences |
|---|
All too often, the roles of NO within the pathogen are ignored when considering plant–pathogen interactions although recent evidence indicates its potential importance within plant pathogens. However, NO effects have been well documented within denitrifying bacterial species where, under anaerobic conditions, they can reduce nitrate to atmospheric nitrogen through the action of four membrane bound reductases (
[nitrate reductase; NAR]NO2 (nitrite reductase [NIR])·NO (nitric oxide reductase [NOR])N2O (nitrous oxide reductase [N2OR])N2) where NO acts a terminal acceptor for anaerobic respiration (de Vries and Schröder, 2001). NOR enzymes have been characterized from a range of bacteria and exist as either two-component cytochrome complexes (NorCB; van der Oost et al., 1994) or single component qNor proteins which lack a cytochrome moiety and receive electrons directly from quniols (Cramm et al., 1999). NO detoxification in aerobic bacteria has also been intensively investigated since NO plays a major role in, for example, macrophage-mediated microbial killing (Hibbs et al., 1987). Thus, the ability to scavenge or detoxify NO has proven to be a survival factor in, for example, Salmonella enterica (De Groote et al., 1996). Under aerobic conditions NO detoxification occurs through the flavohaemoglobin, nitric oxide deoxygenase (NOD; 
; Gardner et al., 1998), although this enzyme can act as a NOR under conditions of low partial pressures of oxygen (Poole and Hughes, 2000). NO-detoxifying gene expression is mediated by specific NO sensors, for example NorR (Arai et al., 1995; Membrillo-Hernández et al., 1998; Pohlmann et al., 2000). Recent microarray studies in E. coli under anaerobic conditions have demonstrated that such NO sensors have wider roles including regulating large-scale changes in global gene expression (e.g. iron–sulphur cluster assembly mechanisms), as well as detoxification/stress response regulators following NO treatment (Justino et al., 2005).
Examining the genomes of several bacterial phytopathogens has revealed the conservation of the NO-detoxifying NOD (hmp) gene and NorR (Fig. 6A). Interestingly the highest level of conservation compared with E. coli homologues was observed in necrotrophic bacteria from the genus Erwinia. In addition, the genes from hemibiotrophic P. syringae and Xanthomonas pathovars shared homology only in the NAD+/FAD binding sites which have no globulin domains (Fig. 6A, B). Mutation of the hmp X gene from Erwinia chrysanthemi resulted in a significant increase in NO levels accumulating during attempted tissue maceration of its host and this led to a resistance response of the HR type. Correspondingly, introduction of HmpX into P. s. pv. tomato avrB suppressed the HR elicited in Arabidopsis (Boccara et al., 2005). Such data further emphasize the role of NO in mediating an HR, but also demonstrate that suppression of NO levels is a requirement for successful infection in certain pathogens. In another example, microarray experiments have suggested that arginase (which competes for L-arginine with NOS) is required for virulence in Xylella fastidiosa (Koide et al., 2004).
|
The suppression of cell death with hmpX expressing P. s. pv. tomato avrB begs the question, why such hemibiotrophic bacteria did not acquire such a function? One possibility must be that these pathogens require NO to combat some other feature of plant defence or for a virulence function. OxyR is a transcription factor that regulates anti-oxidant defences and is conserved amongst bacterial species including plant pathogenic bacteria. Only in its oxidized state will OxyR bind to its cognate promoters to express genes coding for antioxidant enzymes (Kim et al., 2002
). OxyR activity is dependent on six cysteine residues of which the most important is Cys199 which is flanked by the S-nitrosylation motif, suggesting regulation by NO (Marshall et al., 2000). This key cysteine residue and flanking S-nitrosylation motif is conserved in Oxy proteins from many plant pathogenic bacteria (Fig. 6C). As considerable oxidative stress and some NO production have been associated with developing disease symptoms due to P. syringae (Mur et al., 2005a, b), it may be that host NO is required by these pathogens to contribute towards OxyR-mediated anti-oxidant defences. Another role for NO in plant pathogen virulence was described by Kers et al. (2004) who reported a NOS homologue in Streptomyces turgidiscabies and showed that one function of derived NO was the nitration of a dipeptide phytotoxin required for plant pathogenicity. NO production has also been noted within phytopathogenic fungi. Using the MIMS/RIMS technique, Conrat