Ethylene Modulates the Role of NONEXPRESSOR OF PATHOGENESIS-RELATED GENES1 in Cross Talk between Salicylate and Jasmonate Signaling

The plant hormones salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) play crucial roles in the signaling network that regulates induced defense responses against biotic stresses. Antagonism between SA and JA operates as a mechanism to ﬁne-tune defenses that are activated in response to multiple attackers. In Arabidopsis ( Arabidopsis thaliana ), NONEXPRESSOR OF PATHOGENESIS-RELATED GENES1 (NPR1) was demonstrated to be required for SA-mediated suppression of JA-dependent defenses. Because ET is known to enhance SA/NPR1-dependent defense responses, we investigated the role of ET in the SA-JA signal interaction. Pharmacological experiments with gaseous ET and the ET precursor 1-aminocyclopropane-1-carboxylic acid showed that ET potentiated SA/NPR1-dependent PATHOGENESIS-RELATED1 transcription, while it rendered the antago- nistic effect of SA on methyl jasmonate-induced PDF1.2 and VSP2 expression NPR1 independent. This overriding effect of ET on NPR1 function in SA-JA cross talk was absent in the npr1-1/ein2-1 double mutant, demonstrating that it is mediated via ET signaling. Abiotic and biotic induction of the ET response similarly abolished the NPR1 dependency of the SA-JA signal interaction. Furthermore, JA-dependent resistance against biotic attackers was antagonized by SA in an NPR1-dependent fashion only when the plant-attacker combination did not result in the production of high levels of endogenous ET. Hence, the interaction between ET and NPR1 plays an important modulating role in the ﬁne tuning of the defense signaling network that is activated upon pathogen and insect attack. Our results suggest a model in which ET modulates the NPR1 dependency of SA-JA antagonism, possibly to compensate for enhanced allocation of NPR1 to function in SA-dependent activation of PR genes.

The plant hormones salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) play crucial roles in the signaling network that regulates induced defense responses against biotic stresses. Antagonism between SA and JA operates as a mechanism to finetune defenses that are activated in response to multiple attackers. In Arabidopsis (Arabidopsis thaliana), NONEXPRESSOR OF PATHOGENESIS-RELATED GENES1 (NPR1) was demonstrated to be required for SA-mediated suppression of JA-dependent defenses. Because ET is known to enhance SA/NPR1-dependent defense responses, we investigated the role of ET in the SA-JA signal interaction. Pharmacological experiments with gaseous ET and the ET precursor 1-aminocyclopropane-1-carboxylic acid showed that ET potentiated SA/NPR1-dependent PATHOGENESIS-RELATED1 transcription, while it rendered the antagonistic effect of SA on methyl jasmonate-induced PDF1.2 and VSP2 expression NPR1 independent. This overriding effect of ET on NPR1 function in SA-JA cross talk was absent in the npr1-1/ein2-1 double mutant, demonstrating that it is mediated via ET signaling. Abiotic and biotic induction of the ET response similarly abolished the NPR1 dependency of the SA-JA signal interaction. Furthermore, JA-dependent resistance against biotic attackers was antagonized by SA in an NPR1-dependent fashion only when the plant-attacker combination did not result in the production of high levels of endogenous ET. Hence, the interaction between ET and NPR1 plays an important modulating role in the fine tuning of the defense signaling network that is activated upon pathogen and insect attack. Our results suggest a model in which ET modulates the NPR1 dependency of SA-JA antagonism, possibly to compensate for enhanced allocation of NPR1 to function in SA-dependent activation of PR genes.
Plants have a broad spectrum of mechanisms to cope with adverse conditions such as abiotic stress (e.g. flooding and drought) and biotic stress (e.g. pathogen and insect attack). With regard to biotic stress, plants possess both physical and chemical barriers to prevent harmful attackers from causing damage. When these constitutively active layers of defense are overcome, inducible defense systems are recruited to counteract the attacker (Walters et al., 2007). The phytohormones salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) have emerged as key players in regulating the activation of the induced defense responses involved (Dong, 1998;Howe, 2004;Pozo et al., 2004;Grant and Lamb, 2006;Von Dahl and Baldwin, 2007;Vlot et al., 2008). Their production varies greatly, depending on the nature of the attacking pathogen or insect. The quantity, composition, and timing of the hormonal blend produced results in the activation of a specific set of defense-related genes that eventually determines the nature of the defense response that is triggered by the attacker encountered (De Vos et al., 2005;Mur et al., 2006). Other plant hormones, including abscisic acid (Mauch-Mani and Mauch, 2005;de Torres-Zabala et al., 2007;Asselbergh et al., 2008), brassinosteroids (Nakashita et al., 2003), gibberellins (Navarro et al., 2008), and auxins (Navarro et al., 2006;Wang et al., 2007), have also been reported to play a role in the plant's immune response, but their significance is less well studied.
In Arabidopsis (Arabidopsis thaliana), it was shown that SA-, JA-, and ET-dependent pathways regulate defense responses that are differentially effective against specific types of attackers (Thomma et al., 2001;Glazebrook, 2005;Thatcher et al., 2005). Pathogens with a biotrophic lifestyle, such as Pseudomonas syringae and Hyaloperonospora arabidopsidis, are generally more sensitive to SA-dependent responses, whereas necrotrophic pathogens, such as Botrytis cinerea and Alternaria brassicicola, and herbivorous insects, such as Pieris rapae (small cabbage white) and Frankliniella occidentalis (western flower thrips), are commonly deterred by JA-and/or ET-dependent defenses Kessler and Baldwin, 2002;Ton et al., 2002;De Vos et al., 2006;Abe et al., 2008). In nature, plants often deal with simultaneous or subsequent invasion by multiple aggressors, which can influence the primary induced defense response of the host plant ( Van der Putten et al., 2001;Bezemer and Van Dam, 2005;Stout et al., 2006;Poelman et al., 2008). Activation of plant defense mechanisms is associated with ecological fitness costs (Heil and Baldwin, 2002;Heidel et al., 2004;Van Hulten et al., 2006). Hence, plants need regulatory mechanisms to effectively and efficiently adapt to changes in their complex hostile environment. Cross talk between induced defense signaling pathways provides the plant with such a powerful regulatory potential (Reymond and Farmer, 1998;Spoel and Dong, 2008). Signaling interactions can be either (mutually) antagonistic or synergistic, resulting in negative or positive functional outcomes. Cross talk helps the plant to minimize fitness costs and create a flexible signaling network that allows the plant to fine-tune its defense response to the invaders encountered (Reymond and Farmer, 1998;Pieterse et al., 2001;Kunkel and Brooks, 2002;Bostock, 2005). Yet, it seems that insect herbivores and pathogens have also evolved to manipulate plants for their own benefit by suppressing induced defenses through modulation of the plant's defense signaling network (Pieterse and Dicke, 2007;Robert-Seilaniantz et al., 2007;Walling, 2008).
One of the best studied examples of defense-related signal cross talk is the interaction between the SA and the JA response pathways (Kunkel and Brooks, 2002;Thaler et al., 2002;Glazebrook et al., 2003;Beckers and Spoel, 2006;Spoel and Dong, 2008). Many studies have demonstrated that endogenously accumulating SA antagonizes JA-dependent defenses, thereby prioritizing SA-dependent defenses over JA-dependent ones (Doherty et al., 1988;Peñ a-Cortés et al., 1993;Gupta et al., 2000;Spoel et al., 2003). As a result of the negative interaction between SA and JA signaling, activation of the SA response should render a plant more susceptible to attackers that are resisted via JA-dependent defenses and vice versa. Indeed, many examples of trade-offs between SA-dependent resistance against biotrophic pathogens and JA-dependent defense against insect herbivory and necrotrophic pathogens have been reported (Pieterse et al., 2001;Bostock, 2005;Stout et al., 2006). In Arabidopsis, Spoel et al. (2007) showed that SA-mediated defenses that are triggered upon infection by a virulent strain of the hemibiotrophic pathogen P. syringae rendered infected tissues more susceptible to infection by the necrotrophic pathogen A. brassicicola by suppressing the JA signaling pathway. Similarly, infection by the biotrophic pathogen H. arabidopsidis strongly suppressed JA-mediated defenses that were activated upon feeding by caterpillars of the small cabbage white P. rapae . Conversely, JA signaling can act antagonistically on SA-dependent defenses. For instance, P. syringae produces the phytotoxin coronatine, which functions as a JA mimic and suppresses effectual SA-dependent defenses, thereby promoting susceptibility of the plant to this pathogen (Zhao et al., 2003;Brooks et al., 2005;Cui et al., 2005;Nomura et al., 2005;Uppalapati et al., 2007). Although many reports describe an antagonistic interaction between SA-and JA-dependent signaling, synergistic interactions have been described as well (Schenk et al., 2000;Van Wees et al., 2000;Mur et al., 2006). For example, application of low concentrations of both SA and JA (10-100 mM) led to enhanced JA/ET response in the combination treatment compared with JA alone, suggesting that hormone concentration is important for the final output during plant-microbe interactions (Mur et al., 2006).
Pharmacological experiments with Arabidopsis revealed that transcription of JA-responsive marker genes, such as PDF1.2 and VSP2, is highly sensitive to suppression by exogenous application of SA. This SA-mediated suppression of JA-responsive gene expression (hereafter referred to as SA-JA cross talk) was observed in a large number of Arabidopsis accessions, highlighting the potential significance of this phenomenon in the regulation of induced plant defenses in nature . Several lines of evidence point to a role for SA-mediated redox changes in the regulation of SA-JA cross talk (Ndamukong et al., 2007;. In Arabidopsis, the redox-sensitive protein NPR1 (for NONEXPRESSOR OF PATHOGENESIS-RELATED GENES1), an important transducer of SA-induced redox changes (Mou et al., 2003;Dong, 2004;Pieterse and Van Loon, 2004;Tada et al., 2008), was shown to be a key regulator of SA-mediated suppression of JA signaling (Spoel et al., 2003). Induction of the SA response, either by pathogen infection or by exogenous application of SA, strongly suppressed JA-responsive genes, such as PDF1.2 and VSP2. However, in mutant npr1-1 plants, this antagonistic effect was completely abolished (Spoel et al., 2003). The npr1-1 mutant shows enhanced resistance against Trichoplusia ni (Cabbage looper) and Spodoptera littoralis (Egyptian cotton worm; Cui et al., 2002;Stotz et al., 2002), indicating that blocking the NPR1-dependent SA signaling pathway resulted in enhanced JA-dependent defenses against these insect herbivores. Nuclear localization of NPR1, which is essential for SA-mediated defense gene expression (Kinkema et al., 2000), is not required for the suppression of JA-responsive genes, indicating that the antagonistic effect of SA on JA signaling is modulated through a function of NPR1 in the cytosol (Spoel et al., 2003). In rice (Oryza sativa), a similar cytosolic function of NPR1 in SA-JA cross talk was reported (Yuan et al., 2007): overexpression of cytosolic OsNPR1 suppressed JA-responsive gene transcription and enhanced the level of susceptibility to insect herbivory, whereas NPR1-mediated suppression of the JA response was no longer present in plants expressing OsNPR1 that was constitutively targeted to the nucleus.
Besides SA and JA, ET has also been demonstrated to play an important role in the plant's defense response to pathogen and insect attack (Broekaert et al., 2006;Adie et al., 2007;Von Dahl and Baldwin, 2007). In addition to effects on the level of pathogen or insect resistance, ET was shown to function as an important modulator of the plant's response to other hormones, such as JA, SA, and abscisic acid (Adie et al., 2007). For instance, ET enhanced the response of Arabidopsis to SA, resulting in a potentiated expression of the SA-responsive marker gene PATHOGENESIS-RELATED1 (PR-1; Lawton et al., 1994;De Vos et al., 2006). Moreover, in tobacco (Nicotiana tabacum), ET was shown to be essential for the onset of SA-dependent systemic acquired resistance (SAR) against Tobacco mosaic virus (Verberne et al., 2003). Also, the synergistic interaction between ET and JA has been well established. Many defense-related genes, such as PDF1.2, are regulated via a signaling pathway that requires both ET and JA Broekaert et al., 2006;Adie et al., 2007). Similarly, the corequirement of ET and JA has been demonstrated for the onset of broad-spectrum induced systemic resistance, which is triggered after colonization of plant roots by beneficial microorganisms (Pieterse et al., 1998;Van Wees et al., 2008), highlighting the important modulating role of ET in plant defense.
In many plant-attacker interactions, ET is part of the signal signature that is produced upon pathogen or insect attack (De Vos et al., 2005). The established role of ET in modulating SA-and JA-dependent defense responses prompted us to investigate the effect of ET on the interaction between SA and JA signaling. Here, we demonstrate that ET bypasses the NPR1 dependency of the SA-mediated antagonistic effect on JA signaling, thereby shaping the final outcome of the plant defense signaling network that is activated upon pathogen or insect attack.

ET Modulates the NPR1 Dependency of the SA-JA Signal Interaction
In Arabidopsis, pharmacological experiments revealed that SA can antagonize the expression of JA-responsive genes, such as PDF1.2 and VSP2 (Spoel et al., 2003;. To investigate whether ET affects this SA-JA cross talk, we analyzed the effect of the ET precursor 1-aminocyclopropane-1carboxylic acid (ACC) on SA-and NPR1-dependent suppression of JA-responsive gene expression. To this end, we made use of the mutant npr1-1, which contains a missense mutation that alters a key ankyrin repeat in the NPR1 protein and disrupts NPR1-dependent regulation of both SA-and JA-dependent genes (Cao et al., 1997;Glazebrook et al., 2003). Twelve-day-old seedlings of wild-type accession Columbia (Col-0) and mutant npr1-1 plants were grown on Murashige and Skoog (MS) agar medium (Murashige and Skoog, 1962) with or without increasing concentrations of ACC and 0.5 mM SA, 0.02 mM methyl jasmonate (MeJA), or a combination of both chemicals. Two days later, the expression of the SA-responsive marker gene PR-1 and the JA-responsive marker gene PDF1.2 was analyzed by northern-blot analysis ( Fig. 1). In the absence of ACC, the single treatments of Col-0 with SA or MeJA activated PR-1 and PDF1.2, respectively. In addition, the combination treatments with SA and MeJA resulted in effective SA-mediated suppression of MeJA-induced PDF1.2 expression. As expected, neither PR-1 induction nor PDF1.2 suppression was apparent in the npr1-1 mutant, supporting previous findings that SA-JA cross talk is dependent upon wildtype NPR1 function (Spoel et al., 2003). However, addition of ACC into the medium at different concentrations resulted in effective SA-mediated suppression of MeJA-induced PDF1.2 expression in both Col-0 and npr1-1 plants, suggesting that ET relieved the NPR1 dependency of the SA-JA signal interaction. Similar results were obtained with the JA-responsive gene VSP2 (Supplemental Fig. S1).
In order to corroborate our observation with medium-grown seedlings that ET overrules the NPR1 dependency of SA-JA cross talk, we investigated the effect of ET on SA-JA cross talk in 5-week-old, soilgrown plants using both ACC and gaseous ET. The plants were treated with 1 mM SA, 0.1 mM MeJA, or a combination of both chemicals and either 0.1 mM ACC or gaseous ET (2 mL L 21 ) and harvested 6 h later for northern-blot analysis. As reported previously (Lawton et al., 1994;De Vos et al., 2006), ACC and ET both enhanced the SA-induced expression of PR-1 in adult wild-type plants (Fig. 2). However, these chemicals failed to restore PR-1 expression in the npr1-1 mutant, suggesting that ET stimulates SA signaling through the wild-type function of NPR1. In addition, Figure 2 shows that in the absence of ACC or ET, MeJA-induced PDF1.2 gene expression was effectively suppressed by SA in SA/MeJA-treated Col-0 plants but not in mutant npr1-1 plants. However, as observed in seedlings, addition of ACC ( Fig. 2A) or gaseous ET (Fig. 2B) into the SA/MeJA treatment resulted in a partial restoration of SA-mediated suppression of MeJA-induced PDF1.2 gene expression in the npr1-1 background. Together, these results indicate that ET affects the dependency of SA-mediated suppression of JA-responsive gene expression on wildtype NPR1 function. The observation that neither ACC nor ET treatment bypassed the NPR1 dependency of SA-induced PR-1 expression (Figs. 1 and 2) indicates that NPR1 has a dual role in the suppression of JAdependent genes on the one hand and in the activation of SA-dependent gene expression on the other.

Modulation of the NPR1 Dependency of SA-JA Cross Talk by ET Is EIN2 Dependent
To test if the modulation of the NPR1 dependency of SA-JA cross talk by ET is governed by the ET signaling pathway, we performed cross talk experiments with an npr1-1/ein2-1 double mutant (Clarke et al., 2000). Because the ein2-1 mutation completely blocks the ET signaling pathway (Alonso et al., 1999) and PDF1.2 expression requires an intact response to both JA and ET , we performed these ex-periments with the JA-responsive marker gene VSP2, which is similarly sensitive to the antagonistic effect of SA (Spoel et al., 2003;. Fiveweek-old Col-0, ein2-1, npr1-1, and npr1-1/ein2-1 plants were treated with SA, MeJA, or a combination of both chemicals in the absence or presence of ACC. In the absence of ACC, MeJA-induced expression of VSP2 was effectively suppressed by SA in Col-0 and ein2-1 but not in the npr1-1 and npr1-1/ein2-1 backgrounds (Fig. 3), confirming the critical role of wild-type NPR1 in SA-JA cross talk under low-ET conditions. In ACCtreated plants, the NPR1 dependency of SA-JA cross talk was again relieved, as demonstrated by the SAmediated suppression of MeJA-induced VSP2 expression in the npr1-1 background. Compared with npr1-1 plants, however, SA-JA cross talk remained blocked upon ACC treatment of the npr1-1/ein2-1 double mutant. These data indicate that the modulation of the NPR1 dependency of SA-JA cross talk by ET is dependent upon EIN2 and is thus regulated by the ET signaling pathway.

Abiotic Induction of Endogenous ET Relieves the NPR1 Dependency of SA-JA Cross Talk
To test the biological relevance of the effect of ET on the role of NPR1 in SA-JA cross talk, we performed SA-JA cross talk experiments under abiotic conditions in which Arabidopsis produces enhanced levels of ET. To this end, 5-week-old plants were placed in trays with open or closed lids. As shown in Figure 4A, Col-0 and npr1-1 plants grown in trays with closed lids showed a typical hyponastic response, which is a phenomenon demonstrated to be mediated by ET (Millenaar et al., 2005). Mutant ein2-1 did not display this hyponastic response, confirming the ET dependency of this phenomenon. Besides the ET-dependent hyponastic response, plants grown in trays with closed lids produced more ET (Fig. 4B) and accumulated enhanced transcript levels of the ET-responsive genes ERS2 and EBF2 ( Fig. 4C; Millenaar et al., 2005;, indicating that growth of the plants in closed trays results in enhanced ET signaling. To investigate the effect of endogenously produced ET on the NPR1 dependency of SA-JA cross talk, plants grown in trays with open or closed lids were treated with SA, MeJA, or a combination of both chemicals and harvested 24 h later for northern-blot analysis of PR-1 and PDF1.2 gene expression. Figure 4D shows that the antagonistic effect of SA on MeJA-induced expression of PDF1.2 is blocked in npr1-1 mutant plants when grown in trays with open lids (basal ET signaling). However, when the cross talk experiment was performed with plants grown in trays with closed lids (enhanced ET signaling), the level of SA-mediated suppression of PDF1.2 in npr1-1 plants was similar to that observed in Col-0 plants. Hence, abiotic induction of the ET response relieves the dependency of SA-JA cross talk on wild-type NPR1 function. Figure 1. ACC modulates the NPR1 dependency of SA-JA cross talk in Arabidopsis seedlings. Northern-blot analysis of PR-1 and PDF1.2 mRNA levels in Col-0 and npr1-1 seedlings that were treated with SA, MeJA, or a combination of both chemicals in the absence or presence of the ET precursor ACC. Pharmacological assays were performed with seedlings that were grown for 12 d on MS medium, after which they were transferred to fresh MS medium supplemented with increasing concentrations of ACC and 0.5 mM SA, 0.02 mM MeJA, or a combination of both chemicals. Seedlings were harvested after 2 d for RNA extraction. rRNA is presented as a loading control.
Next, we wanted to investigate whether ET produced during a plant-attacker interaction affects the NPR1 dependency of SA-JA cross talk. To this end, we made use of two JA-inducing attackers: the necrotrophic fungal pathogen A. brassicicola and the herbivorous insect F. occidentalis. A. brassicicola stimulates the biosynthesis of both JA and ET, while F. occidentalis induces only JA production (De Vos et al., 2005). Inoculation of Col-0 and npr1-1 plants with A. brassicicola indeed resulted in a strong increase in the production of ET, whereas infestation with thrips had no effect (Fig. 5A).
To investigate the NPR1 dependency of SA-mediated suppression of JA-responsive gene expression during both Arabidopsis-attacker combinations, Col-0 and npr1-1 plants were infested with F. occidentalis or infected with A. brassicicola 24 h prior to SA treatment. Twenty-four hours later, leaf material was harvested for northern-blot analysis of PR-1 and PDF1.2 transcript levels. Figure 5B shows that F. occidentalis and A. brassicicola both induced the expression of PDF1.2 and that this expression was strongly suppressed by SA in Col-0 plants. In the npr1-1 mutant, this SA-mediated suppression of PDF1.2 transcription was not observed when the JA response was activated by the non-ET-inducer F. occidentalis, indicating that in this plant-attacker combination SA-JA cross talk is NPR1 dependent. However, when the JA response was activated by the JA-and ET-inducer A. brassicicola, PDF1.2 transcription was suppressed in the npr1-1 mutant background. These results indicate that attackerinduced ET largely overrules the NPR1 dependency of SA-JA cross talk and hence potentially affects the outcome of the defense response that is induced upon attack by multiple invaders.

NPR1 Is Not Required for SA-Mediated Suppression of JA-Dependent Resistance against ET-Inducing Attackers
In Arabidopsis, resistance against F. occidentalis and A. brassicicola has been demonstrated to be mediated by the JA response pathway Abe et al., 2008). To investigate the role of NPR1 in the antagonistic effect of SA on the JA-dependent resistance against these attackers, we performed resistance assays in Col-0 and npr1-1 plants. We hypothesized that the antagonistic effect of SA on JA-dependent resistance against ET-noninducing thrips would be NPR1 dependent, while the negative effect of SA on JA-dependent resistance against the ET-inducing fungal pathogen would function independently of NPR1.
In the thrips resistance assays, Col-0 and npr1-1 plants were pretreated with 1 mM SA, 0.1 mM MeJA, or a combination of both. Twenty-four hours later, leaf discs from this material were taken and infested with F. occidentalis. Two days later, the level of thrips Leaf tissue was harvested 6 h after chemical treatment for RNA analysis. Equal loading of RNA samples was checked using a probe for 18S rRNA. Signal intensities of the depicted northern blots were quantified using a phosphor imager (right panels). PDF1.2 transcript levels in the single MeJA treatments were set to 100%. resistance was determined by measuring the level of feeding scar damage that was inflicted by thrips feeding (Abe et al., 2008). As shown in Figure 6A, SA treatment had no significant effect on the basal level of thrips resistance in Col-0 plants. However, MeJAtreated Col-0 plants showed a significantly reduced area of feeding scars, indicating that MeJA treatment enhanced the level of resistance to thrips feeding. Col-0 plants treated with both SA and MeJA showed a basal level of thrips resistance that was not significantly different from that in control plants, suggesting that SA suppressed the level of MeJA-induced resistance against thrips feeding. In mutant npr1-1 plants, MeJA and SA/MeJA treatments both led to a significant increase in the level of thrips resistance, indicating that the SA-mediated suppression of MeJA-induced resistance to F. occidentalis is controlled by NPR1.
Wild-type Col-0 plants are highly resistant to A. brassicicola infection, but this resistance is lost in JAinsensitive coi1-1 mutant plants ( Thomma et al., 1998), indicating that JA is an important regulator of basal resistance against this pathogen. Previously, Spoel et al. (2007) demonstrated that SA suppresses this JA-dependent resistance against A. brassicicola, resulting in enhanced susceptibility of Col-0 plants to A. brassicicola infection. Indeed, exogenous application of SA to Col-0 plants broke the JA-dependent resistance to A. brassicicola (Fig. 6B). However, treatment of npr1-1 plants with SA only moderately reduced the level of JA-dependent resistance against this pathogen. These results suggest that the SA-mediated suppression of JA-dependent resistance against A. brassicicola is functioning, at least partly, independently of NPR1. Since A. brassicicola-infected tissues produced high levels of ET and thrips-infested tissues did not (Fig. 5A), it is likely that the regulatory role of NPR1 in the antagonism between SA and JA is determined by the presence or absence of ET in the signal signature of the plant-attacker combination.

ET Modulates the Role of NPR1 in Cross Talk between SA and JA Signaling
Cross talk between defense signaling pathways is thought to play an important role in the regulation and fine-tuning of the defense responses that are activated upon pathogen and insect attack. The antagonism between SA and JA signaling emerged as one of the Figure 3. NPR1-independent SA-JA cross talk facilitated by ET is EIN2 dependent. Northern-blot analysis of PR-1 and VSP2 transcript levels in 5-week-old Col-0, ein2-1, npr1-1, and npr1-1/ein2-1 plants that were treated with 1 mM SA, 0.1 mM MeJA, or a combination of both chemicals in the absence (2) or presence (+) of 0.1 mM ACC. Leaf tissue was harvested 6 h after treatment for RNA analysis. Equal loading of RNA samples was checked using a probe for 18S rRNA. Signal intensities of the depicted northern blots were quantified using a phosphor imager (right panels). VSP2 transcript levels in the single MeJA treatments were set to 100%.  most prominent of all signal interactions studied to date Spoel and Dong, 2008). Pharmacological experiments revealed that the suppression of JA-responsive genes such as PDF1.2, VSP2, and LOX2 by SA is regulated by NPR1 (Spoel et al., 2003). Following a whole-genome transcript profiling approach to identify Arabidopsis genes that are sensitive to SA-JA cross talk, we recently identified 258 MeJA-responsive genes of which the expression was significantly affected by SA (A. Koornneef and C.M.J. Pieterse, unpublished data). Sixty percent of the JA-responsive genes that were suppressed by SA displayed this suppression in an NPR1-dependent manner, demonstrating that NPR1 is involved in the SA-mediated down-regulation of a large number of MeJA-responsive genes. Because ET is an important modulator of plant defense and a major constituent of the blend of defense signals that is produced during many plant-attacker interactions (Broekaert et al., 2006;Adie et al., 2007;Von Dahl and Baldwin, 2007), we investigated the effect of ET on the SA-JA signal interaction. Here, we demonstrate that ET strongly affects the requirement of wild-type NPR1 in the antagonistic effect of SA on JAdependent defenses. Exogenous application of the ET precursor ACC or gaseous ET (Figs. 1-3), as well as endogenously produced ET during induction of the hyponastic response (Fig. 4) or pathogen attack (Fig.  5), bypassed the NPR1 dependency of SA-JA cross talk. Experiments in the mutant ein2-1 background showed that this ET effect is EIN2 dependent and thus mediated through the ET signaling pathway (Fig. 3). These findings indicate that the final outcome of the SA-JA signal interaction during the complex interaction of plants with their attackers can be shaped by ET. Indeed, the antagonistic effect of SA on MeJA-induced resistance against feeding by ET-noninducing thrips was controlled by NPR1. By contrast, SA-mediated suppression of JA-dependent resistance against the JA-and ET-inducing necrotroph A. brassicicola functioned independently of NPR1 (Fig. 6), highlighting the modulating role of ET in the SA-JA signal interaction. In Figure 7, we present a schematic model of the interplay between SA, JA, ET, and NPR1 in the Arabidopsis-attacker interactions studied. Figure 5. Attacker-induced ET enables NPR1independent SA-JA cross talk. A, ET production in Col-0 and npr1-1 after infection with the necrotrophic fungus A. brassicicola or infestation with larvae of the western flower thrips F. occidentalis. FW, Fresh weight. B, Northern-blot analysis of PR-1 and PDF1.2 transcript levels in Col-0 and npr1-1 plants that were infested with F. occidentalis or inoculated with A. brassicicola and treated (+) or not (2) with 1 mM SA. Leaf tissue was harvested for RNA analysis 24 h after application of SA. Equal loading of RNA samples was checked using a probe for 18S rRNA. Signal intensities of the depicted northern blots were quantified using a phosphor imager (right panels). PDF1.2 transcript levels in the single MeJA treatments were set to 100%.

Dual Role of NPR1
NPR1 is a regulatory protein that was originally identified in Arabidopsis through several genetic screens for SAR-compromised mutants (Cao et al., 1994;Delaney et al., 1995;Glazebrook et al., 1996;Shah et al., 1997). Mutant npr1-1 plants are not only compromised in SAR but also in basal resistance against many types of pathogens that are sensitive to SAdependent defenses (Dong, 2004). In addition, mutant npr1-1 plants appeared to be blocked in the activation of induced systemic resistance by beneficial rhizobacteria, an induced defense response that requires regulators of ET and JA signaling (Pieterse et al., 1998;Van Wees et al., 2008). Moreover, NPR1 has been implicated in JA-and ET-dependent resistance against the soil-borne fungus Verticillium longisporum (Johansson et al., 2006). The fact that NPR1 also functions as an important regulator of SA-JA cross talk (Spoel et al., 2003;Yuan et al., 2007) demonstrates that NPR1 plays a central role in the induced defense signaling network that is controlled by SA, JA, and ET (Dong, 2004;Pieterse and Van Loon, 2004). Our finding that the requirement of NPR1 in SA-JA cross talk is bypassed under conditions in which ET production is induced provides a direct link between ET and NPR1 function.
In this study, we demonstrate that ET bypasses the need for NPR1 in SA-JA cross talk, while it enhances NPR1-dependent, SA-responsive PR-1 expression. This clearly indicates that NPR1 plays a dual role in regulating SA-mediated suppression of JA-responsive gene expression on the one hand and SA-mediated activation of SA-responsive PR gene expression on the other hand. This raises the question: how does ET signaling differentially affect the NPR1 dependency of these two SA-dependent cellular responses? The differential effect of ET on NPR1 function may be caused by the fact that the role of NPR1 in SA-JA antagonism is mediated by a cytosolic function of NPR1 (Spoel et al., 2003;Yuan et al., 2007), whereas the role of NPR1 as a coactivator of SA-responsive PR gene expression is exerted in the nucleus (Kinkema et al., 2000;Dong, 2004). Previously, Glazebrook et al. (2003) demonstrated that two different alleles of the npr1 mutant (npr1-1 and npr1-3) behaved differently in terms of transcriptome changes upon infection by P. syringae. The npr1-1 mutant, which has a mutation in a key ankyrin-repeat domain, was affected in the expression of SA-dependent as well as JA-and ET-dependent genes. However, the npr1-3 mutant, which produces a truncated cytoplasmatically localized NPR1 protein Figure 6. Antagonistic effect of SA on JA-dependent resistance against F. occidentalis and A. brassicicola in Col-0 and npr1-1. A, F. occidentalis resistance assay with 3-week-old Col-0 and npr1-1 plants that were pretreated for 24 h with 1 mM SA, 0.1 mM MeJA, or a combination of both chemicals. Presented are means 6 SD (n = 10) of the relative area of feeding scars (control treatment is set to 1) on Col-0 and npr1-1 leaf discs after 2 d of thrips feeding. Different letters indicate statistically significant differences between treatments (Tukey-Kramer honestly significant difference test; P , 0.05). B, A. brassicicola resistance assay with 5-week-old Col-0 and npr1-1 plants that were treated or not with 1 mM SA. Data represent the percentage of leaves (n = 80) that developed spreading lesions after inoculation with A. brassicicola. Figure 7. Working model illustrating the role of ET in modulating the NPR1 dependency of SA-JA cross talk. Attack of Arabidopsis by the necrotrophic fungus A. brassicicola and the herbivorous insect F. occidentalis results in the biosynthesis of JA and the activation of the JA signaling pathway in which the E3 ubiquitin ligase SCF COI1 and jasmonate ZIM-domain (JAZ) proteins that repress the transcription of JA-responsive genes are central components (Chini et al., 2007;Thines et al., 2007). Activation of the JA signaling cascade leads to the activation of JA-responsive genes such as PDF1.2 and VSP2. SA suppresses JA-responsive gene expression in an NPR1-dependent manner. However, when ET signaling is stimulated, such as upon infection by the ET-inducer A. brassicicola, the NPR1 dependency of SA-JA cross talk is bypassed, resulting in wild-type levels of suppression of JA signaling in the npr1-1 mutant background. that misses the C-terminal domain with the nuclear localization signal (Dong, 2004), was only affected in SA-dependent gene expression, suggesting that the cytoplasmatic function of NPR1 plays a role in the control of JA-and ET-dependent responses. In agreement with this, the antagonistic effect of SA on JAresponsive gene expression was much less affected in npr1-3 than in npr1-1 (Supplemental Fig. S2). These results suggest a model in which the cytosolic function of NPR1 plays a role in SA-JA cross talk and can be bypassed by ET and in which the nuclear function of NPR1 plays a role in the activation of SA-responsive genes and can be stimulated by ET.
Previously, the glutaredoxin GRX480 and the transcription factor WRKY70 were identified as important players in SA/NPR1-dependent suppression of JAresponsive gene expression (Li et al., 2004;Ndamukong et al., 2007). In wild-type plants, transcription of GRX480 and WRKY70 was activated by SA in an NPR1-dependent manner, indicating that the roles of GRX480 and WRKY70 in the suppression of JA-responsive genes are downstream of the NPR1dependent induction of GRX480 and WRKY70 by SA. However, the fact that SA/NPR1-dependent gene expression is hampered in mutant npr1-3, while SA/ NPR1-dependent suppression of JA-responsive gene expression is still intact in this mutant, suggests that the antagonistic effect of SA on JA signaling can function independently of GRX480 or WRKY70. This is corroborated by previous findings that grx480 and wrky70 knockout mutants showed wild-type levels of SA-mediated suppression of MeJA-induced PDF1.2 gene expression (Ndamukong et al., 2007;A. Leon-Reyes and C.M.J. Pieterse, unpublished data).

Interaction between ET and NPR1
NPR1 is an important transducer of the SA signal. In uninduced cells, NPR1 is present as an oligomer formed through intermolecular disulfide bonds (Mou et al., 2003). SA mediates a change in the cellular redox potential, resulting in the reduction of the NPR1 oligomer to its active monomeric form. Monomeric NPR1 is then translocated into the nucleus, where it functions as a coactivator of SA-responsive genes, such as PR-1, by enhancing the binding of TGA transcription factors to SA-responsive promoter elements (Després et al., 2003;Mou et al., 2003;Rochon et al., 2006;Tada et al., 2008). Recently, we demonstrated that SA-mediated redox modulation also plays an important role in the SA-mediated attenuation of the JA signaling pathway . Hence, it is plausible that the cytosolic function of NPR1 in SA-JA cross talk is controlled by active NPR1 monomers that are produced upon SA-mediated changes in the redox state.
With our current knowledge of NPR1 function, we can only speculate on how ET affects the NPR1 dependency of the SA-JA signal interaction. On the one hand, ET potentiates the NPR1-dependent expression of the SA-responsive marker gene PR-1 in Arabidopsis (Figs. 2 and 3;Lawton et al., 1994;De Vos et al., 2006). On the other hand, our study clearly shows that ET bypasses the need for NPR1 in SA-JA cross talk. These results suggest a model in which ET modulates the allocation of NPR1's positive and negative functions. Since SA-activated NPR1 functions in the nucleus to activate PR genes and in the cytosol to suppress JAresponsive genes, it is tempting to speculate that ET signaling allocates more NPR1 to the nucleus to support SA signaling, thereby making less NPR1 available in the cytosol for SA-JA cross talk. At the same time, possible negative effects of this trade-off on SA-JA cross talk are compensated, because in combination with ET, SA can suppress JA-responsive gene expression in an NPR1-independent manner.
So how could ET modulate the NPR1 dependency of SA-JA cross talk? In the absence of ET, SA-activated NPR1 monomers may bind a positive regulator of JAresponsive gene expression in the cytosol, which is then prevented from entering the nucleus, resulting in the suppression of JA-responsive gene expression. Alternatively, NPR1 may activate a negative regulator of the JA pathway. A simple explanation for the role of ET in these scenarios may be that ET signaling results in a similar effect on the putative positive or negative regulator, rendering the function of NPR1 redundant in SA-JA cross talk. However, other scenarios are plausible as well. For instance, various genetic screens revealed mutations that restored the SAR-compromised phenotype of the npr1-1 mutant. Mutations in genes such as SNI1, SSI1, and CPR6 were demonstrated to restore SA-mediated PR gene expression and SAR in the absence of a functional NPR1 protein (Clarke et al., 1998;Li et al., 1999;Shah et al., 1999;Durrant et al., 2007). This clearly indicates that the NPR1 dependency of important SA-mediated cellular responses can be bypassed by inactivation of proteins such as SNI1, SSI1, and CPR6. Future research will be focused on elucidating the targets of ET through which this hormone is able to affect NPR1 function during SA-JA cross talk.

Alternaria brassicicola Assays
For induction of JA-responsive gene expression and ET production, Col-0 plants were inoculated with Alternaria brassicicola strain MUCL 20297 as described previously (De Vos et al., 2005). Briefly, the fungus was grown on potato dextrose agar for 2 to 3 weeks at 22°C. Spores were collected as described by Broekaert et al. (1990). Five-week-old plants were inoculated with 5-mL drops of 50% potato dextrose broth containing 5 3 10 5 spores mL 21 . For assessing the effect of SA on the level of resistance against A. brassicicola, leaves of 5-week-old plants were pressure infiltrated with a solution of 10 mM MgSO 4 supplemented with or without 1 mM SA (Spoel et al., 2007). After 24 h, the treated leaves were inoculated with A. brassicicola by applying a 3-mL drop of 50% potato dextrose broth containing 1 3 10 6 spores mL 21 . At 4 d after inoculation, the percentage of leaves with spreading lesions was assessed.

Frankliniella occidentalis Assays
For induction of JA-responsive gene expression, thrips infestations were performed on 5-week-old plants by transferring 20 larvae of Frankliniella occidentalis to each plant using a fine paintbrush (De Vos et al., 2005). For determination of thrips resistance, the leaf disc assay described by Abe et al. (2008) was used. Briefly, isolated leaf discs from 3-week-old plants that were pretreated for 24 h with 1 mM SA, 0.1 mM MeJA, or a combination of both chemicals (see below) were floated on 1.5 mL of distilled water in wells of a white 1.5-mL sample tube stand. A single adult female that had been starved for 2 to 3 h was placed on a single leaf disc. Thrips were allowed to feed for 1 or 2 d at 22°C. The area of thrips feeding scars on the surface of each leaf disc was measured by ImageJ software (Abramoff et al., 2004) on digitized images.

Chemical Treatments
Plants were treated with SA, MeJA, and/or ACC by dipping the leaves into a solution of 0.015% (v/v) Silwet L77 (Van Meeuwen Chemicals) containing 1 mM SA (Mallinckrodt Baker), 0.1 mM MeJA (Serva, Brunschwig Chemie), 0.1 mM ACC (Sigma), or a combination of these chemicals as described previously (Spoel et al., 2003;. Control treatments were dipped into a solution containing 0.015% (v/v) Silwet L77. Chemical induction of plants grown on MS medium was performed by transferring 12-d-old seedlings to fresh MS medium supplemented with 0.5 mM SA, 0.02 mM MeJA, 0.1 to 10 mM ACC, or a combination of these chemicals (Spoel et al., 2003). MeJA was added to the solutions from a 1,000-fold-concentrated stock in 96% ethanol. To the solutions without MeJA, a similar volume of 96% ethanol was added.
Application of gaseous ET to the plants was performed as described by Millenaar et al. (2005). In brief, gaseous ET (100 mL L 21 ; Hoek Loos) and air (70% relative humidity) were mixed using flow meters (Brooks Instruments) to generate an output concentration of 2 mL L 21 ET, which was flushed continuously through glass cuvettes (13.5 3 16.0 3 29.0 cm) at a flow rate of 75 L h 21 and then vented to the outside of the building. The concentration of ET in the air flow was verified using gas chromatography as described by Millenaar et al. (2005). For the duration of the gaseous ET treatment, 5-weekold plants were placed in the cuvette, which were placed under climate chamber conditions as described above. Control plants were treated in a similar manner but without ET in the air flow.

ET Measurements
To measure ET production in plants challenged with either A. brassicicola or F. occidentalis, rosettes of inoculated or infested plants were detached from the roots, weighed, and placed individually in 35-mL gas-tight serum flasks (n = 10) that were subsequently incubated under climate chamber conditions. At different time intervals, 1-mL gas samples were withdrawn through the rubber seal. The concentration of ET was measured by gas chromatography as described by De Laat and Van Loon (1982) To measure the ET production by plants grown in trays with open and closed lids, plants were removed from the trays and whole rosettes of about 300 mg were immediately transferred into a syringe with a volume of 1.5 mL. ET was allowed to accumulate in the syringe for 15 min, after which the head space was analyzed for ET levels using gas chromatography as described by Millenaar et al. (2005).

RNA Extraction and Northern-Blot Analysis
For RNA extraction, at least five plants per treatment were harvested at the time points indicated. RNA isolation was performed as described previously by Van Wees et al. (2000). For RNA-blot analysis, 15 mg of RNA was denatured using glyoxal and dimethyl sulfoxide (Sambrook et al., 1989), electrophoretically separated on a 1.5% agarose gel, and blotted onto Hybond-N + membranes (Amersham) by capillary transfer. The buffers used for electrophoresis and blotting were 10 and 25 mM sodium phosphate (pH 7.0), respectively. RNA blots were hybridized with probes for PR-1, PDF1.2, and VSP2 as described previously by Pieterse et al. (1998). To check for equal loading, ribosomal RNA (rRNA) bands were stained with ethidium bromide or the blots were stripped and hybridized with a probe for 18S ribosomal RNA. The Arabidopsis Genome Initiative numbers for the genes studied are At2g14610 (PR-1), At5g44420 (PDF1.2), and At5g24770 (VSP2). After hybridization with [a-32 P]dCTP-labeled probes, blots were exposed for autoradiography. Signal intensities of PDF1.2 or VSP2 mRNA on the northern blots were quantified using a Bio-Rad Molecular Imager FX with Quantity One software (Bio-Rad). The PDF1.2 and VSP2 mRNA levels of the MeJA treatment were set to 100% and compared with PDF1.2 and VSP2 mRNA levels of the rest of the treatments. All gene expression analyses were repeated with similar results.

Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure S1. ACC modulates the NPR1 dependency of SAmediated suppression of MeJA-induced VSP2 expression in Arabidopsis seedlings.