Plants can retain the memory of a prior encounter with a pest. This memory confers upon a plant the ability to subsequently activate defenses more robustly when challenged by a pest. In plants that have retained the memory of a prior, localized, foliar infection by a pathogen, the pathogen-free distal organs develop immunity against subsequent infections by a broad-spectrum of pathogens. The long-term immunity conferred by this mechanism, which is termed systemic acquired resistance (SAR), is inheritable over a few generations. Signaling mediated by the phenolic metabolite salicylic acid (SA) is critical for the manifestation of SAR. Recent studies have described the involvement of additional small metabolites in SAR signaling, including methyl salicylate, the abietane diterpenoid dehydroabietinal, the lysine catabolite pipecolic acid, a glycerol-3-phosphate-dependent factor and the dicarboxylic acid azelaic acid. Many of these metabolites can be systemically transported through the plant and probably facilitate communication by the primary infected tissue with the distal tissues, which is essential for the activation of SAR. Some of these metabolites have been implicated in the SAR-associated rapid activation of defenses in response to subsequent exposure to the pathogen, a mechanism termed priming. Here, we summarize the role of these signaling metabolites in SAR, and the relationship between them and SA signaling in SAR.
Systemic acquired resistance (SAR) is an inducible defense mechanism that immunizes the foliage against a broad spectrum of pathogens (Sticher et al., 1997; Spoel and Dong, 2012; Shah and Zeier, 2013). SAR is activated in the distal pathogen-free organs of a plant that has retained the memory of a prior infection in another organ. Tissues exhibiting SAR display a heightened state of preparedness that is associated with the priming of defenses for faster and stronger activation in response to subsequent pathogen inoculation (Conrath, 2011; Spoel and Dong, 2012; Shah and Zeier, 2013). SAR confers a fitness advantage to plants growing under high disease pressure (Traw et al., 2007). Furthermore, when inoculated with a pathogen, defenses are induced faster in the progeny of plants expressing SAR, thus resulting in higher level of disease resistance in the progeny, a phenomenon that has been termed ‘trans-generational SAR' (Luna et al., 2012). The memory associated with the inheritance of SAR is likely to be epigenetic in nature (Luna and Ton, 2012; Luna et al., 2012). As discussed below, the activation of salicylic acid (SA)-mediated defenses is essential for SAR-conferred immunity. Recent studies have implicated additional small metabolites in SAR signaling, which likely interact to provide better control of SAR under different environmental conditions (Park et al., 2007; Jung et al., 2009; Chanda et al., 2011; Chaturvedi et al., 2012; Návarová et al., 2012). This review combines current knowledge with the latest status of signaling by small metabolites in SAR.
A long distance-transported signal(s) is required for SAR
Critical for the activation of SAR is communication between the pathogen-inoculated organ and the distal organs, which involves the transport of a ‘SAR signal’ to the distal organs (Figure 1). Grafting between pathogen-inoculated and the pathogen-free plants and experiments in which the phloem path above the pathogen-inoculated organs was obstructed have suggested that the phloem is the likely conduit for the translocation of this long-distance transported ‘SAR signal’ (Jenns and Kuc, 1979; Guedes et al., 1980; Tuzun and Kuc, 1985). In Arabidopsis thaliana, which has served as an excellent model plant for the molecular-genetic characterization of SAR-associated signaling, the SAR signal is present in the phloem sap-enriched petiole exudates (Pexs) collected from avirulent (Avr) pathogen-inoculated leaves, thus further supporting the prospect of the phloem being a channel for the systemic movement of the SAR signal (Maldonado et al., 2002; Chaturvedi et al., 2008; Jung et al., 2009; Chanda et al., 2011). However, the phloem may not be the exclusive route for the transport of the long distance-translocated SAR signal (Kiefer and Slusarenko, 2003).
The SAR-inducing activity in Avr-Pex collected from Arabidopsis was sensitive to treatment with proteases (Chanda et al., 2011; Chaturvedi et al., 2012), thus suggesting that a protein(s) in Avr-Pex is likely to be involved in the accumulation and/or transport of the SAR signal. In Arabidopsis, the DIR1 (DEFECTIVE IN INDUCED RESISTANCE1) protein, which exhibits homology to non-specific lipid-transfer proteins (LTPs), is required for SAR (Maldonado et al., 2002; Chaturvedi et al., 2008, 2012; Chanda et al., 2011; Yu et al., 2013). Occasionally, under some circumstances, a DIR1-like protein may substitute for DIR1 (Champigny et al., 2013). Similarly in tobacco (Nicotiana tabacum), a DIR1 homolog is required for the activation of SAR (Liu et al., 2011b). While the Arabidopsis dir1 mutant was sensitive to the SAR-inducing activity contained in Avr-Pex from wild-type (WT) plants, Avr-Pex from dir1 plants was unable to induce SAR in WT plants (Maldonado et al., 2002; Chaturvedi et al., 2008). Thus, the SAR defect in the dir1 mutant is due to its inability to accumulate and/or transport the SAR signal from the pathogen-inoculated to the distal pathogen-free leaves. DIR1 is expressed in phloem tissues and the DIR1 protein, which contains a N-terminal secretory sequence, is targeted to the apoplast in Arabidopsis and can also be recovered in phloem sap-enriched Avr-Pex (Champigny et al., 2011, 2013) (Figure 1). In addition, locally expressed DIR1 protein was found to move systemically through the plant (Champigny et al., 2013; Yu et al., 2013). Taken together, these studies suggest that DIR1 might be a component of the long-distance SAR signal or be involved in the systemic translocation of the SAR signal. Although the predicted size of DIR1 is approximately 7 kDa, on western blots DIR1 from Avr-Pex was detected as 7- and 15-kDa bands (Champigny et al., 2013). Furthermore, when subjected to gel-filtration chromatography, DIR1 present in the Avr-Pex co-purified with a high molecular weight fraction (>100 kDa) (Figure 1) which contained a trypsin-sensitive SAR-inducing factor (Chaturvedi et al., 2012). These observations suggest that DIR1 in Avr-Pex is either oligomeric or associated with other proteins. The two SH3 domains in DIR1 (Lascombe et al., 2008) probably facilitate protein–protein interaction. Indeed, when transiently coexpressed in Nicotiana benthamiana leaves, DIR1 was recently shown to be capable of interacting with itself as well as with AZI1 (AZELAIC ACID-INDUCED 1) (Yu et al., 2013), another protein involved in SAR, which exhibits homology to LTPs (Jung et al., 2009; Chanda et al., 2011; Chaturvedi et al., 2012; Yu et al., 2013). Whether AZI1 is also present in the phloem sap, in association with DIR1, is not known.
The SAR-inducing activity in sap from Arabidopsis and cucumber is effective in conferring resistance in other plants (Jenns and Kuc, 1979; Chaturvedi et al., 2008, 2012; Chanda et al., 2011), thus suggesting that the systemically mobile SAR signal is not species specific. As discussed below, recent studies have implicated a variety of small metabolites (Figure 2), both primary and secondary, including methyl salicylate (MeSA), the abietane diterpenoid dehydroabietinal (DA), the lysine catabolite pipecolic acid (Pip), the 3C sugar alcohol glycerol-3-phosphate (G3P) and the dicarboxylic acid azelaic acid (AzA), in SAR signaling (Nandi et al., 2004; Park et al., 2007; Jung et al., 2009; Chanda et al., 2011; Chaturvedi et al., 2012; Návarová et al., 2012). Some of these metabolites, when exogenously applied, move systemically through the foliage and are suggested to participate in long-distance signaling (Park et al., 2007; Jung et al., 2009; Chanda et al., 2011; Chaturvedi et al., 2012). As depicted in Figure 3 and summarized below, networking between these signaling metabolites and SA is likely to provide the plant with flexibility in optimally regulating SAR under different environmental conditions (Shah, 2009; Dempsey and Klessig, 2012; Shah and Zeier, 2013).
Salicylic acid, a key signaling factor in the manifestation of SAR
Although long considered a secondary metabolite, SA (2-hydroxy benzoic acid) has important roles in a variety of plant processes, including flowering and thermogenesis (heat production) that accompanies flowering in certain angiosperms and in the reproductive structures of cycads (Raskin, 1992; Shah and Klessig, 1999; Dempsey et al., 2011). However, it is the involvement of SA in plant defense that has provided the best evidence of its essential role as a signaling metabolite in plants (Chaturvedi and Shah, 2007; Dempsey et al., 2011). SA and the ensuing signaling is required for controlling diseases, particularly those caused by pathogens that exhibit a biotrophic phase in their life cycle when they acquire nutrients from live plant cells. Levels of SA and its glucoside SAG (SA-glucoside) (Figure 4), and expression of SA-responsive genes like PR1 (PATHOGENESIS-RELATED1), increase in pathogen-infected organs. The upregulation of PR1 transcript accumulation has been utilized as a molecular marker to monitor the activation of SA signaling (Sticher et al., 1997). Genetic studies with mutant and transgenic plants that are compromised in their ability to accumulate SA have confirmed the importance of SA in controlling diseases caused by a variety of pathogens (Shah and Klessig, 1999; Chaturvedi and Shah, 2007; Dempsey et al., 2011).
Two distinctive pathways (Figure 4) that are dependent on chorismic acid are involved in the biosynthesis of SA (Dempsey et al., 2011). In the isochorismate pathway, which is the major SA biosynthesis pathway in Arabidopsis, SA is synthesized from isochorismic acid produced by the action of plastid-localized isochorismate synthase (ICS). An isochorismate pyruvate lyase activity is predicted to further convert isochorismic acid to SA. Arabidopsis contains two genes (ICS1 and ICS2) that encode ICSs. ICS1 is the major contributor to SA synthesis in pathogen-infected plants (Garcion et al., 2008). The second SA biosynthesis pathway is dependent on phenylalanine ammonia lyase (PAL) activity, which catalyzes the deamination of l-phenylalanine to yield trans-cinnamic acid, which can subsequently be converted into SA via the ortho-coumaric acid or benzoic acid routes (Figure 4). Transgene-mediated co-suppression of PAL expression or pharmacological inhibition of PAL activity with 2-aminoindan-2-phosphonic acid attenuated the pathogen-infection-associated accumulation of SA in tobacco, Arabidopsis and cucumber, thus confirming a role for the PAL pathway in SA synthesis (Meuwly et al., 1995; Mauch-Mani and Slusarenko, 1996; Pallas et al., 1996).
Contribution of SA signaling to SAR
SA also has an important function in SAR (Shah, 2009; Dempsey et al., 2011; Spoel and Dong, 2012; Shah and Zeier, 2013). Preventing the accumulation of SA in Arabidopsis and tobacco, either due to mutations (or suppression of expression) in genes (e.g. ICS1 and PAL) encoding enzymes involved in SA synthesis or by promoting the turnover of SA to catechol by expression of the bacterial nahG gene-encoded salicylate hydroxylase, resulted in the attenuation of SAR-conferred immunity (Vernooij et al., 1994; Lawton et al., 1995; Pallas et al., 1996; Wildermuth et al., 2001; Mishina and Zeier, 2007; Chaturvedi et al., 2008, 2012; Jung et al., 2009). Similarly, SAR is compromised in plants lacking a functional copy of PAD4 (PHYTOALEXIN-DEFICIENT4) (Mishina and Zeier, 2006), which is involved in the feed-forward amplification of SA accumulation (Figure 3) (Rietz et al., 2011). SAR is also attenuated in the sard1 cbp60g double mutant, which is defective in SA accumulation; the SARD1 and CBP60g proteins bind to the ICS1 promoter to regulate ICS1 expression, and thus SA accumulation (Zhang et al., 2010; Wang et al., 2011). Although SA accumulates at elevated levels in leaves inoculated with a SAR-inducting pathogen and in the phloem sap collected from these leaves (Malamy et al., 1990; Métraux et al., 1990), grafting experiments in tobacco between transgenic NahG-expressing or PAL-silenced plants that are both deficient in SA accumulation, and non-transgenic WT plants, indicated that the rise in SA content in the primary pathogen-inoculated leaf was not critical for the activation of SAR in the distal leaves (Vernooij et al., 1994; Pallas et al., 1996). However, SA was required in the distal leaves for SAR-conferred disease resistance.
During SAR, a significant increase in SA content occurs in the distal pathogen-free organs. Some of this SA in the distal leaves could be derived from SA or its derivatives (e.g. MeSA) transported from the primary pathogen-inoculated leaves (Figure 3). In Arabidopsis, systemic increase in SA requires the upregulation of ICS1 expression in the distal leaves (Attaran et al., 2009), thus suggesting that de novo synthesis is required for the accumulation of SA in the pathogen-free distal leaves. Further support for de novo synthesis of SA in the systemic leaves comes from experiments with the fld (flowering locus D) mutant, which is SAR-deficient (Singh et al., 2013). The SA content increased in the primary pathogen-inoculated leaves of the fld mutant and the SAR signal was generated, systemically transported and perceived by the distal leaves. However, the distal pathogen-free leaves of fld did not accumulate elevated levels of SA (Singh et al., 2013). Since FLD affects histone modifications (He et al., 2003; Liu et al., 2007; Yu et al., 2011), it is likely that upon perception of the SAR signal, FLD-dependent changes in histone modifications at genes involved in SA metabolism, and/or encoding regulators of SA biosynthesis, are associated with the systemic enhancement of SA content. Systemic accumulation of SA and SAR-conferred resistance were also attenuated in the fmo1 (flavin-dependent monooxygenase1) mutant (Mishina and Zeier, 2006; Chaturvedi et al., 2012), which as described later is also suggested to be a component of a SA amplification loop involving Pip. The exact function of the SAR-associated SA accumulation in the pathogen-free leaves is unclear. Levels of SA and downstream signaling in the systemic organs of plants exhibiting SAR are primed for robust increases in response to subsequent exposure to a pathogen (Jung et al., 2009; Návarová et al., 2012). This priming of SA signaling is diminished in the fld mutant (Singh et al., 2013). SA and its functional analog BTH (S-methyl-1,2,3-benzothiadiazole-7-carbothioate) can prime the rapid activation of defense signaling in response to subsequent stress (Conrath, 2011). This priming effect of SA and BTH, which is suggested to involve epigenetic changes (Conrath, 2011; Jaskiewicz et al., 2011), requires NPR1 (NON-EXPRESSER OF PR GENES1), a key regulator of SA signaling (described below), thus suggesting that SA signaling per se is required for this priming-conferred robust activation of defenses. A positive feed-forward loop involving PAD4 and SA is known to amplify the accumulation of SA in response to pathogen infection (Figure 3) (Rietz et al., 2011). Similarly, the SAR-associated increase in SA in the distal pathogen-free leaves could promote the subsequent robust accumulation of SA and the resultant defense signaling when these leaves are challenged with a pathogen.
Perception of SA and regulation of downstream defense signaling
The NPR1 gene, which encodes a transcription co-activator, has an important role in SA-mediated signaling during SAR, as well as basal resistance against pathogens (Durrant and Dong, 2004; Chaturvedi and Shah, 2007; Spoel and Dong, 2012). In the cytosol, NPR1 exists in an oligomeric form (Mou et al., 2003). It has been suggested that SA causes a conformational change in NPR1, thus promoting the disassembly of oligomeric NPR1 to facilitate nuclear localization of NPR1 monomers (Kinkema et al., 2000). In the nucleus, NPR1 interacts with other proteins, including a subset of the TGA family of DNA-binding proteins, to regulate expression of SAR-associated genes (Despres et al., 2000; Zhang et al., 2003). The targets of NPR1 include genes involved in the protein secretory pathway that are important for the secretion of PR proteins (Wang et al., 2005). SA also promotes the phosphorylation of NPR1 to facilitate ubiquitination and subsequent degradation of NPR1 by the proteasome, thus controlling the level of NPR1 protein in the nucleus. Spoel et al. (2009) have suggested that by promoting the removal of the ‘exhausted’ phosphorylated protein from the promoters of its target genes, SA facilitates the binding of fresh NPR1 to its target promoters and thus facilitates additional transcription cycles to occur, thus ensuring maximal expression of its target genes during SAR. Unlike Fu et al. (2012), Wu et al. (2012) were unable to detect SA-binding activity for NPR1. In contrast, Fu et al. (2012) observed that binding of SA to the NPR1 paralogs NPR3 and NPR4 promoted their interaction with NPR1 and the CUL3 ubiquitin ligase. Since the SA-promoted turnover of NPR1 was not observed in the npr3 npr4 double mutant, Fu et al. (2012) suggested that NPR3 and NPR4 promote the SA-dependent turnover of NPR1 by the proteasome and thus regulate the function of NPR1 in SA signaling. Readers are directed to a recent review (Pajerowska-Mukhtar et al., 2013) for additional details on the regulation/modulation of NPR1 function and signaling in plant defense.
Small metabolites in systemic signaling and priming of defense signaling
Although SAR was described in the early 20th century and the role of SA in SAR was described towards the end of the 20th century, the identity of the long-distance SAR signal remained elusive until recently, when significant strides were made towards the identification of small metabolites that have been implicated in SAR-associated long-distance signaling and/or priming of defenses. These include MeSA, DA, Pip, a G3P-dependent metabolite and AzA (Park et al., 2007; Jung et al., 2009; Chanda et al., 2011; Chaturvedi et al., 2012; Návarová et al., 2012). The function of these metabolites in SAR is dependent on SA signaling (Figure 3), suggesting that they all target steps upstream or at the level of SA accumulation/signaling. The discovery and our current understanding of the involvement of these metabolites in SAR signaling is discussed below.
Methyl salicylate, a systemically transported derivative of SA
In 1997, Shulaev et al. demonstrated that MeSA (Figure 2) produced by tobacco leaves infected with Tobacco mosaic virus (TMV) can function as an airborne signal that activates SA signaling to promote resistance against TMV in neighboring tobacco plants. Using 14C-SA they showed that MeSA could be synthesized from SA by tobacco leaves. Furthermore, exposure of tobacco plants to 14C-MeSA vapors resulted in the recovery of the 14C label in SA. More than 90% of the SA recovered from leaves exposed to 14C-MeSA vapors was 14C-labeled, leading them to suggest that the activation of defenses by MeSA vapors resulted from its conversion to SA in tobacco leaves. They further suggested that during SAR, vascular translocation of MeSA, which is a liquid at room temperature, from the primary TMV-inoculated leaves could provide another mechanism for the transport and accumulation of SA in the distal pathogen-free leaves. Ten years later, Park et al. (2007) provided genetic evidence that MeSA synthesized from SA in the TMV-infected leaves of tobacco is transported to the distal leaves, where it contributes to the accumulation of SA and SAR-conferred disease resistance. Methyl salicylate is synthesized by the transfer of a methyl group from S-adenosyl-l-methionine to SA by S-adenosyl-l-methionine: SA carboxyl methyltransferase (SA-methyltransferase; SAMT) (Figure 4). RNA interference (RNAi)-mediated silencing of SAMT1 expression in tobacco resulted in the inability to activate SAR (Park et al., 2007). Reciprocal grafting experiments between non-transgenic tobacco and the SAMT1-silenced transgenic plants further indicated that SAMT1 function was required in the primary TMV-inoculated leaves for the manifestation of SAR in the distal leaves. In the distal leaves the SA-binding protein SABP2, which possesses MeSA esterase activity that converts MeSA to SA, was required for the manifestation of SAR (Forouhar et al., 2005; Kumar et al., 2006; Park et al., 2007, 2009). Reciprocal grafting experiments showed that SAR-determined disease resistance was attenuated whenever the scion, which received the challenge inoculation, was derived from the SABP2-silenced lines, but not when SABP2 expression was silenced in the root-stock, which received the SAR-inducing pathogen treatment (Park et al., 2007). Pharmacological studies with 2,2,2,2′-tetra-fluoroacetophenone, a competitive inhibitor of the MeSA esterase activity of SABP2, as well as experiments with a missense Ser81 → Ala81 version of SABP2, which lacked MeSA esterase activity, confirmed the importance of the MeSA esterase activity of SABP2 in SAR (Park et al., 2007, 2009). The SAR was compromised in 2,2,2,2′-tetra-fluoroacetophenone-treated plants as well as in plants expressing the mutant version of SABP2. It has been suggested that MeSA synthesized by SAMT1 in the TMV-inoculated tobacco leaves is transported to the distal pathogen-free leaves, probably via the vasculature, where it is converted to SA by the esterase activity of SABP2 (Dempsey and Klessig, 2012). This MeSA-derived SA, in association with any de novo synthesized SA, contributes to the overall increase in SA content and downstream activation of defense signaling in the pathogen-free leaves of plants exhibiting SAR (Figure 3).
Subsequent studies have indicated that MeSA also has an important function in the manifestation of SAR in potato (Solanum tuberosum) plants treated with arachidonic acid (Manosalva et al., 2010). Local application of arachidonic acid resulted in the accumulation of SA, MeSA and PR1 transcript in the treated as well as the distal untreated leaves, which exhibited enhanced resistance against Phytophthora infestans. However, the arachidonic acid-induced systemic increase in SA, PR1 expression and resistance against P. infestans was compromised in transgenic potato plants in which expression of the potato METHYL ESTERASE 1 (StMES1) gene, which encodes a homolog of tobacco SABP2, was silenced. In comparison with the non-transgenic plants, the StMES1-silenced RNAi plants accumulated higher levels of MeSA in the distal leaves in response to treatment of another leaf with arachidonic acid, thus suggesting that the reduction in StMES1 transcript levels, and hence StMES1 activity, had an adverse impact on the conversion of MeSA to SA, and thus the activation of SAR. The induction of SAR by arachidonic acid was prevented in potato plants treated with 2,2,2,2′-tetrafluoroacetophenone as well, thus providing additional support for the involvement of MeSA esterase and MeSA in SAR in potato.
In Arabidopsis SAR is also attenuated by 2,2,2,2′-tetrafluoroacetophenone (Park et al., 2009). Arabidopsis contains several genes that encode potential MeSA esterases. Vlot et al. (2008) showed that the simultaneous knock-down of expression of multiple AtMES genes attenuated SAR in Arabidopsis. However, this attenuation of SAR was observed in only half the experiments. Similarly, Chaturvedi et al. (2012) observed that the attenuation of SAR in these AtMES-silenced lines was at best weak and observed only occasionally. Subsequently, Liu et al. (2010) reported that SAR-conferred enhanced disease resistance was attenuated in the Arabidopsis bsmt1 mutant, which lacks MeSA-synthesizing benzoic acid/SA methyl transferase 1 (BSMT1) activity. However, another study noted that the bsmt1 mutant plants were SAR competent despite their inability to produce MeSA (Attaran et al., 2009). Taken together, this variable requirement for MeSA in SAR suggests that other factors influence the requirement for MeSA in SAR. Liu et al. (2011a) reported that light is one of the environmental factors that influences the contribution of MeSA to SAR. They observed that when the SAR-inducing treatment was delivered close to the dark period, MeSA was required for SAR. However, when the SAR-inducing treatment was delivered during the early part of the light period, MeSA was less important for SAR.
Dehydroabietinal, a potent inducer of systemic SA accumulation and signaling
Abietane diterpenoids are major components of oleoresin, produced by conifers. Some of these diterpenoids possess activities against pests and also are valued precursors for industrial bioproducts and pharmacologically active molecules (Trapp and Croteau, 2001; Bohlmann and Keeling, 2008). Structurally the abietane diterpenoids, which have a tricyclic frame, are similar to the tetracyclic ent-kaurane diterpenes, which include the gibberellins (Hamberger et al., 2011). Like the gibberellins, in conifers the abietane diterpenoids are derived from geranylgeranyl diphosphate by the sequential action of diterpene synthases and P450 monooxygenases. The early steps in abietane diterpenoid synthesis occur in the chloroplast and involve the cyclization of geranylgeranyl diphosphate by diterpene synthase to yield abietadiene via a copalyldiphosphate intermediate (Figure 5a). Abietadiene can be further converted to dehydroabietadiene. The subsequent sequential oxidation steps, which involve a cytochrome P450 monooxygenase (abietane oxidase), occur in the cytosol to yield the corresponding alcohols, aldehydes and acids (Keeling et al., 2008, 2011). Although poorly studied in angiosperms, abietane diterpenoids are found in flowering plants (Hanson, 2009; Chaturvedi et al., 2012). Like soil-grown plants, the foliage of axenically cultivated Arabidopsis contain DA (Figure 5a), thus suggesting a plant origin for these diterpenoids in Arabidopsis. Arabidopsis contains a homolog of the Sitka spruce (Picea sitchensis) CYP720B4, which is a cytochrome P450 monooxygenase class enzyme that is capable of synthesizing DA and related compounds (Hamberger et al., 2011). However, whether Arabidopsis and other flowering plants synthesize abietane diterpenoids by a mechanism that is comparable to the biosynthesis machinery in conifers that begins with geranylgeranyl diphosphate (Figure 5a) remains to be determined.
Recently, DA was suggested to have a signaling function in SAR. Chaturvedi et al. (2012) reported that a SAR-inducing activity that is present in Arabidopsis Avr-Pex could be extracted in chloroform, thus suggesting that a hydrophobic factor is probably involved in long-distance signaling. Subsequent purification identified DA as the active component in Avr-Pex. DA is a very potent inducer of SAR. Picomolar solutions of DA applied to a couple of leaves were sufficient to induce SAR in Arabidopsis, tomato and tobacco. Experiments with 2H-DA indicated that DA is rapidly transported from the treated leaves to rest of the foliage. Localized application of DA to a few leaves resulted in a systemic increase in SA and upregulation of SA-responsive genes. Expression of ICS1 was induced in distal leaves of plants that had other leaves treated with DA, thus suggesting de novo synthesis of SA in the distal leaves of DA-treated plants. The ICS1 and NPR1 genes were required for DA-induced SAR, thus suggesting that SA accumulation and signaling are critical for DA-induced SAR. As depicted in Figure 3, DA-induced SAR and systemic SA accumulation also required FMO1, which is also needed for systemic accumulation of SA during SAR.
DA levels did not increase in leaves and Pex in response to the treatment of leaves with a SAR-inducing pathogen (Chaturvedi et al., 2012). However, compared with Pex collected from mock-inoculated leaves in which DA eluted in a low-molecular-weight (<30 kDa) fraction that was unable to induce SAR, in the Avr-Pex, DA was enriched in a trypsin-sensitive high-molecular-weight fraction (>100 kDa) that was active in inducing SAR, thus suggesting that the mobilization of DA from the biologically inactive low-molecular-weight pool into a signaling form (DA*) (Figure 3), which is associated with proteins in the high-molecular-weight fraction, is a rate-limiting step in SAR. In Avr-Pex, the DIR1 protein is also associated with the DA-containing high molecular weight fraction (>100 kDa) that is capable of inducing SAR (Figure 1). In contrast, DIR1 was undetectable in the corresponding low-molecular-weight fraction (<30 kDa) (Figure 1). DIR1, a lipid-binding protein which can be systemically translocated (Champigny et al., 2013; Yu et al., 2013), could be involved in transporting DA, a hydrophobic metabolite, from the primary pathogen-inoculated leaves to the distal leaves. Alternatively, DIR1 might be part of a ‘SAR signalosome’, a signaling complex that includes DA and is involved in long-distance signaling (Figure 3). Indeed, DIR1 was required for the full extent of SAR induction by DA, confirming an important function for DIR1 in DA-induced SAR. Although the SFD1/GLY1 and AZI1 genes (described below), which are involved in G3P- and AzA-mediated signaling, were not essential for DA-induced SAR, their presence enhanced the effectiveness of DA in inducing SAR (Chaturvedi et al., 2012), thus suggesting that G3P- and AzA-dependent activities are likely to enhance the effect of DA* in SAR. The role of G3P and AzA is discussed later in this review.
The application of DA also promotes flowering in Arabidopsis, which is accompanied by the upregulation of FLD and FT (FLOWERING LOCUS T) expression and a concomitant reduction in expression of FLC (FLOWERING LOCUS C) (Figure 6). In the flowering pathway, during the vegetative phase of growth, FLC, which encodes a MADS box family protein, represses the expression of FT to prevent flowering. The FT protein, which is considered to be the phloem-mobile ‘florigen’ released by leaves, signals to the shoot meristem to transition from the vegetative to the reproductive phase. The FLD protein is part of a repressor complex that influences the post-translation modification of histones at the FLC locus to suppress FLC expression and thus promote FT expression, resulting in the switch to flowering (Kim and Sung, 2012). FLD also has an essential role in SAR induction by DA and Avr pathogens, thus suggesting that FLD is a critical component of SAR signaling (Singh et al., 2013, 2014). As mentioned above, during the biological induction of SAR by inoculation with a pathogen, FLD is required in the distal leaves to respond to the SAR-inducing factor (Figure 3) (Singh et al., 2013). FLD function is required for the SAR-associated systemic accumulation of SA and priming of PR1 and the defense-regulatory WRKY6 and WRKY29 genes (Singh et al., 2013, 2014). In contrast, FLD is not required for the accumulation of SA in pathogen-treated leaves (Singh et al., 2013). Analysis of the fld flc double mutant indicated that unlike the involvement of FLD in flowering, the role of FLD in SAR is independent of its relationship with FLC. Although presence of the flc allele in the fld flc double mutant suppressed the flowering defect of the fld allele, the fld flc double mutant retained the fld-conferred SAR-deficient phenotype (Singh et al., 2013). FLD is also required for the induction of SAR by AzA (Singh et al., 2013), which as described below has been suggested to be involved in priming of SA-dependent defense signaling during SAR (Jung et al., 2009), thus further suggesting that FLD is a key player, functioning upstream of SA accumulation in SAR.
Pipecolic acid, a non-protein amino acid involved in signal amplification
The cyclic non-protein amino acid l-pipecolic acid (Pip; homoproline) (Figure 5b) is a catabolite of l-lysine that is found in plants and animals. Levels of Pip in plants increase in response to a variety of stresses (Yatsu and Boynton, 1959; Pálfi and Dézsi, 1968; Moulin et al., 2006). In addition, Pip also promotes flowering in the aquatic plant Lemna gibba (Fujioka et al., 1987). l-Pipecolic acid was recently shown to accumulate at elevated levels in leaves treated with a SAR-inducing pathogen as well as the distal pathogen-free leaves (Návarová et al., 2012). This increase in Pip content in the systemic leaves preceded the elevation in SA content. The accumulation of Pip in Arabidopsis requires the ALD1 (AGD2-LIKE DEFENSE RESPONSE PROTEIN1) gene that encodes an aminotransferase (Návarová et al., 2012) which in vitro has a strong substrate preference for lysine (Song et al., 2004b). ALD1 transcript accumulation is elevated in the pathogen-inoculated and the distal pathogen-free leaves (Song et al., 2004a). Furthermore, SAR as well as basal resistance against virulent and Avr pathogens is attenuated in the ald1 mutant (Song et al., 2004a; Návarová et al., 2012). The ald1 mutant fails to accumulate SA in the distal pathogen-free leaves following inoculation of other leaves with a SAR-inducing pathogen (Song et al., 2004a; Jing et al., 2011; Návarová et al., 2012). This defect was complemented by irrigating the ald1 plants with Pip prior to the inoculation with the SAR-inducing pathogen. Irrigating plants with Pip also resulted in a stronger induction of SA accumulation and defense gene expression upon subsequent treatment with a pathogen, thus suggesting that Pip is involved in the priming of defenses. The FMO1 gene, which is required for the SAR-associated systemic accumulation of SA (Mishina and Zeier, 2006; Chaturvedi et al., 2012), was also required for Pip-induced systemic resistance (Návarová et al., 2012), thus suggesting that FMO1 functions downstream of ALD1 and Pip. Furthermore, the ICS1 gene, and hence SA synthesis, was required for the accumulation of Pip in the distal leaves. Since the application of Pip enhances its own biosynthesis in the distal leaves, which is dependent on ALD1, FMO1 and ICS1, Návarová et al. (2012) have proposed, as depicted in Figure 3, the presence of a positive-feedback amplification loop that integrates ALD1-dependent accumulation of Pip and the FMO1-stimulated SA synthesis by ICS1. Levels of Pip also increase in Pex collected from the SAR-inducing pathogen-inoculated leaves (Návarová et al., 2012). However, whether Pip has a direct role in long-distance communication between the pathogen-infected and the distal uninfected organs remains to be determined.
A glycerol-3-phosphate-dependent factor in systemic signaling
Glycerol-3-phosphate (G3P) (Figure 2), a primary metabolite, provides the C backbone required for the synthesis of several biomolecules, including membrane and storage lipids. The involvement of a G3P-dependent factor in SAR was first suggested by Nandi et al. (2004) who demonstrated that the Arabidopsis SFD1 (SUPPRESSOR OF FATTY ACID DESATURASE DEFICIENCY1) gene, also known as GLY1, which encodes a dihydroxyacetone phosphate reductase that synthesizes G3P (Figure 5c), is required for the induction of SAR. Lack of sfd1 function had a significant effect on the composition of chloroplast-synthesized lipids (Nandi et al., 2003, 2004; Lorenc-Kukula et al., 2012), thus suggesting that SFD1 functions in the chloroplasts. Indeed, a SFD1–GFP fusion was shown to localize to the plastids and plastid localization was essential for the involvement of SFD1 in lipid metabolism and SAR (Lorenc-Kukula et al., 2012). The SAR-associated systemic accumulation of SA and upregulation of PR1 expression was lacking in the sfd1 mutant (Nandi et al., 2004; Chaturvedi et al., 2008). Although the sfd1 mutant was perceptive to the SAR signal contained in Avr-Pex collected from WT plants, Avr-Pex collected from sfd1 leaves lacked the SAR-inducing factor, thus leading to the suggestion that the sfd1 mutant is defective in the synthesis and/or accumulation of a long-distance signal required for SAR (Chaturvedi et al., 2008). Mutations in the GLI1 [GLYCEROL-INSENSITIVE 1; also known as NHO1 (NONHOST1)] gene, which encodes a glycerol kinase that synthesizes G3P in the cytosol, also attenuated SAR (Chanda et al., 2011; Yu et al., 2013).
More recently, Chanda et al. (2011) demonstrated that although SAR was compromised in the gly1 mutant, which is allelic with sfd1 but in a different accession, the gly1 mutant retained the SAR-associated systemic accumulation of SA and PR1 expression. They observed that during SAR, G3P levels increased in the leaves inoculated with the SAR-inducing Avr pathogen, as well as in Avr-Pex and the distal pathogen-free leaves. When co-applied with Pex, G3P was capable of enhancing resistance in the untreated leaves, thus suggesting that a G3P-dependent factor can systemically enhance disease resistance. Although SAR was accompanied by the accumulation of G3P in Avr-Pex and a systemic increase in the G3P content, 14C-labelled G3P when infiltrated into leaves could not be recovered from the distal leaves as 14C-G3P. Instead, the 14C was recovered in an unknown compound in the distal leaves. It is plausible that 14C-G3P infiltrated into the intracellular spaces does not access the phloem, thus explaining the inability to recover 14C-G3P in the distal leaves. The fact that the radiolabel was recovered in another compound indicates that G3P when infiltrated into the intracellular spaces is metabolized and that a G3P-dependent factor contributes to systemic signaling in SAR. These results also indicate that the systemic increase in G3P during SAR is probably due to de novo synthesis of G3P in the systemic leaves.
The application of G3P affected gene expression in the distal, untreated leaves of Arabidopsis, resulting in the upregulation of nearly 40 genes and the downregulation of approximately 130 genes (Chanda et al., 2011). Expression of MES9, which encodes the Arabidopsis homolog of the tobacco MeSA esterase, was upregulated in the distal leaves of G3P-treated plants and expression of BSMT1, which encodes a methyl transferase that is involved in the synthesis of MeSA, was simultaneously downregulated (Chanda et al., 2011). However, since the application of G3P to Arabidopsis leaves does not result in a corresponding increase in SA and SAG content in the distal untreated leaves (Chanda et al., 2011), the biological relevance of this impact of G3P on the expression of genes involved in MeSA metabolism is unclear. It is plausible that G3P is involved in priming. However, in plants that were pre-treated with G3P, the distal leaves, when challenged with a pathogen, did not respond with significantly faster or stronger increase in SA compared to the pathogen-treated leaves of plants that did not receive G3P pre-treatment (Yu et al., 2013). Whether the G3P biosynthesis- and SAR-defective sfd1/gly1 and gli1 mutants are also deficient in priming associated with SAR has not been reported. Thus, although G3P biosynthesis is required for SAR, how the infiltration of G3P into leaves promotes SAR is unclear. The proposed involvement of a G3P-dependent factor in SAR signaling and its relationship with other metabolites and genes involved in SAR is summarized in Figure 3.
Azelaic acid, a SA signaling priming dicarboxylic acid
The 9C dicarboxylic acid AzA (Figure 2) has been implicated in the SAR-associated priming of SA accumulation and signaling in Arabidopsis. AzA was identified as a metabolite that accumulated at elevated levels in Arabidopsis Avr-Pex compared with Mock-Pex (Jung et al., 2009). Jung et al. (2009), further noted that although the application of AzA per se did not promote SA accumulation and signaling, AzA-treated plants exhibited robust induction of SA accumulation and signaling upon pathogen inoculation that correlated with enhanced resistance against the pathogen. SA accumulation and signaling were required for AzA-promoted systemic immunity. Also required for AzA-promoted SAR in Arabidopsis were FMO1 and ALD1 (Jung et al., 2009), both of which are suggested to be involved in an amplification loop involving SA (Návarová et al., 2012). As described above, ALD1 is involved in the biosynthesis of Pip (Návarová et al., 2012), thus suggesting that the proposed involvement of AzA-induced priming of defense during SAR is probably dependent on Pip-mediated amplification of SA accumulation and signaling (Figure 3). Application of AzA had a minimal impact on gene expression in Arabidopsis (Jung et al., 2009), thus further suggesting that AzA per se does not induce defenses but is probably involved in priming of defenses. AZI1 was one gene that was transiently induced by AzA. The SAR-associated priming of PR1 expression was attenuated in the azi1 mutant, which is also SAR-deficient (Jung et al., 2009). AzA as well as G3P were also unable to induce SAR in the azi1 mutant (Jung et al., 2009; Yu et al., 2013). Furthermore, as noted above, the presence of AZI1 also sensitized Arabidopsis to DA (Chaturvedi et al., 2012).
Deuterium-labeled AzA applied locally to leaves was subsequently recovered in the Pex, as well as systemically in the untreated leaves, thus leading to the suggestion that AzA is a likely to be a systemic signal in SAR (Jung et al., 2009). However, more recently, Yu et al. (2013) reported that only a small fraction (<10%) of locally applied AzA could be recovered in the distal leaves. Whether this reflects the amount of AzA transported from the pathogen-inoculated leaf to the distal leaves during biologically induced SAR is not known. Yu et al. (2013) further showed that exogenously applied AzA stimulated the accumulation of G3P by promoting the accumulation of the GLY1 and GLI1 transcripts. In addition, AzA-induced systemic immunity required GLY1 and GLI1 function, thus leading Yu et al. (2013) to conclude that AzA functions in SAR by promoting G3P synthesis. If AzA functions to promote G3P synthesis during biologically (pathogen)-induced SAR, then accumulation of AzA should precede or parallel the accumulation of G3P. However, in an earlier study Chanda et al. (2011) showed that at 6 h post-inoculation with a SAR-inducing pathogen, although G3P content had increased eight-fold, there was no increase in AzA. Rather, a significant increase in AzA occurred later, at 24 h post-inoculation, when the G3P level had begun to decline. Moreover, Jung et al. (2009) had previously reported that AzA was capable of inducing SAR in the G3P biosynthesis-deficient sfd1 mutant, thus suggesting that the SFD1/GLY1-dependent factor is not required for AzA-induced SAR. Hence, whether AzA accumulation indeed promotes G3P accumulation during the biological induction of SAR, as opposed to that observed in pharmacological experiments involving exogenously applied AzA, needs further clarification.
Two recent studies suggest that AzA and AZI1 may not be important for SAR under all conditions (Návarová et al., 2012; Zoeller et al., 2012). Furthermore, unlike the priming effect of AzA on SA accumulation as noted by Jung et al. (2009), Yu et al. (2013) did not observe any significant level of priming of SA accumulation in AzA-treated plants. Similarly, as noted above, G3P treatment also did not have any significant priming effect on SA accumulation (Yu et al., 2013). In addition, Liu et al. (2011a) observed that, similar to the role of MeSA in SAR, the timing of the primary pathogen-treatment relative to the onset of the dark period determines the relative importance of SFD1/GLY1 in SAR (Liu et al., 2011a). Taken together, these studies indicate that the contribution of defense priming by AzA is not always important for the biological induction of SAR.
How AzA is synthesized by plants is unclear. Although, a pathway involving the oxidation of 9-oxononanoic acid, synthesized by the sequential action of 9-lipoxygenase (9-LOX) and hydroperoxide lyase, is a potential route for the biosynthesis of AzA, Zoeller et al. (2012) observed that the pathogen inoculation-induced accumulation of AzA was not compromised in the lox1 lox5 double mutant, which lacks the two known 9-LOXs in Arabidopsis, thus leading them to suggest that 9-LOXs are not required for the synthesis of AzA. Instead, Zoeller et al. (2012) suggest that AzA is synthesized non-enzymatically from galactolipids.
A feedback loop involving DIR1, AZI1 and G3P in SAR?
Although Avr-Pex from sfd1 and dir1 mutants were unable to induce SAR when applied individually to WT plants, when co-applied, sfd1 plus dir1 Avr-Pex induced SAR in WT plants (Chaturvedi et al., 2008), thus suggesting that a G3P-dependent factor and DIR1, or a DIR1-dependent factor, are together required for long-distance signaling in SAR. Indeed, although G3P and recombinant DIR1 when applied individually were relatively weak inducers of SAR, when co-applied they had a stronger inductive effect on systemic disease resistance in Arabidopsis (Chanda et al., 2011). The fact that recombinant DIR1 does not bind G3P in vitro, but DIR1 application enhances the systemic translocation of the radiolabel present in 14C-G3P, led Chanda et al. (2011) to suggest that DIR1 facilitates systemic transport of a G3P-dependent factor. These outcomes of Chanda et al. (2011) are based on the assumption of proper folding of recombinant DIR1 made in Escherichia coli. More recently, Yu et al. (2013) noted that DIR1 was also required for the accumulation of G3P during SAR. They further observed that in the dir1 mutant, SFD1/GLY1 transcript abundance was lower, thus suggesting that the influence of DIR1 on G3P accumulation is mediated through stabilization of the SFD1/GLY1 transcript. Similarly, DIR1 was also required for the stabilization of the GLI1 transcript (Yu et al., 2013). Thus, DIR1 affects G3P accumulation as well as the systemic translocation of a G3P-dependent SAR signal. By comparison, G3P was required for the stability of the DIR1 as well as AZI1 transcripts (Yu et al., 2013).
DIR1 is also required for AzA-induced immunity (Jung et al., 2009; Yu et al., 2013). Similarly, AZI1 is required for AzA- and G3P-induced immunity (Jung et al., 2009; Yu et al., 2013). When transiently coexpressed in N. benthamiana, AZI1 was capable of physically interacting with itself and DIR1. Furthermore, when overexpressed from the heterologous 35S promoter, DIR1 and AZI1 could complement the SAR deficiency of the azi1 and dir1 mutants, respectively, thus leading Yu et al. (2013) to suggest that DIR1 and AZI1 are likely to have a shared function in SAR and that they function as a unit. For DIR1 and AZI1 to function together as a single unit, it is expected that there should be an overlap in the temporal and spatial pattern of accumulation of these proteins during SAR, which, however, remains to be determined. Considering that expression from the 35S promoter does not reflect the natural expression pattern of DIR1 and AZI1, caution should be exerted in interpreting these cross-complementation experiments. Yu et al. (2013) have proposed a model involving a regulatory feedback loop in which the DIR1 and AZI1 proteins functioning together, are required for AzA-induced accumulation of G3P, which in turn is required for the stability of the DIR1 and AZI1 transcripts and hence the levels of the DIR1 and AZI1 proteins.
Chanda et al. (2011) showed that application of G3P promoted the systemic translocation of DIR1–GFP expressed in agroinfiltrated tissues of N. benthamiana. However, in these experiments conducted with transiently expressed DIR1–GFP, the DIR1–GFP chimera was described to localize to the plasmodesmata, in addition to the nuclear membrane and endoplasmic reticulum, leading Chanda et al. (2011) to suggest that the systemic translocation of DIR1 occurs through the symplast and not through the apoplast. This interpretation of symplastic translocation of DIR1 is based on the ability of the DIR1–GFP chimera to localize to the plasmodesmata in the presence of a viral movement protein, which was coexpressed in these experiments to mark the plasmodesmata (Chanda et al., 2011). However, the path of systemic translocation of DIR1 has not been tested in Arabidopsis in which DIR1 is present in the apoplast (Champigny et al., 2011, 2013). Furthermore, the infiltration of DIR1 into the apoplastic space was shown to be sufficient to complement the Arabidopsis dir1 defect and enhance systemic disease resistance (Chanda et al., 2011). Thus, additional clarity is needed on the biological function of the proposed symplast-translocated DIR1 as opposed to the apoplastic DIR1 and their relationship with the G3P-dependent factor in SAR.
For plants growing in an environment with a high disease pressure, the ability to activate SAR confers a fitness advantage that can also benefit subsequent generations. Studies with transgenic plants constitutively expressing NPR1 have indicated that the NPR1-mediated mechanism is evolutionarily conserved and can be targeted for enhancing resistance against a variety of pathogens in agronomically important plants (Chern et al., 2001; Lin et al., 2004; Makandar et al., 2006; Malnoy et al., 2007; Potlakayala et al., 2007; Quilis et al., 2008; Wally et al., 2009; Parkhi et al., 2010). However, SAR is an energy-driven process that results in the diversion of resources from growth and development (Heidel et al., 2004; Pajerowska-Mukhtar et al., 2012). Further, the inappropriate activation of SAR, in particular SA-dependent signaling, can limit the ability of a plant to resist infection by other pests, especially those that are targeted by jasmonate-dependent defenses (Koornneef and Pieterse, 2008). Pathogens have also evolved factors that can potentially limit the ability of the plant to activate SAR (Cui et al., 2005). The involvement of multiple interacting signals offers sufficient flexibility to optimally turn on SAR under a diverse array of environmental conditions. As summarized in Figure 3, a variety of signaling metabolites contribute to SAR by functioning upstream and/or by potentiating SA accumulation. Increasing evidence indicates that chromatin restructuring and epigenetic changes have an important function in SAR (Conrath, 2011; Jaskiewicz et al., 2011; Luna and Ton, 2012; Luna et al., 2012; Singh et al., 2013). A recent study indicates that signaling by DA and AzA requires FLD (Singh et al., 2013), thus suggesting that these signaling metabolites function by promoting histone modifications at genes involved in SA synthesis, or genes regulating SA synthesis and/or accumulation. What is the contribution of this process to SAR and trans-generational SAR? What is the extent of interaction between these signaling metabolites and what are the molecular components of this network? These are some important questions that need to be addressed in the coming years.
The authors would like to thank numerous colleagues for informal discussions, Robin Cameron for providing the DIR1 antibody and two anonymous referees for their comments. This work was supported by a grant from the National Science Foundation (IOS-1121570).