THE ROLE OF SA IN SAR
The role of SA in SAR has been discussed extensively in a number of reviews. As described above, in many plants SAR is preceded by an increase in SA concentration. However, some plants such as potato and rice have high endogenous levels of SA under noninducing conditions. Indeed, application of SA to potato does not protect it against Phytophthora infestans. However, expression of nahG in potato blocks resistance to P. infestans induced by arachidonic acid. This suggests that after treatment with arachidonic acid, instead of SA levels rising, the potato plants become more sensitive to SA. Thus, SA is an essential signal for SAR across a range of plants, although the mechanism by which SA induces SAR might differ.
SA Synthesis
It was previously assumed that SA for SAR is synthesized via the shikimate-phenylpropanoid pathway, although this was never proven. It has recently been shown that, like bacteria, plants can also synthesize SA from chorismate via isochorismate. Expression of the bacterial enzymes catalyzing these reactions, isochorismate synthase 1 (ICS1) and isochorismate pyruvate lyase 1 (IPL1), in tobacco and Arabidopsis results in increased SA accumulation and pathogen resistance.
Using HPLC, Nawrath & Métraux isolated the SA induction-deficient Arabidopsis mutants sid1 and sid2, which failed to accumulate SA after SAR induction. More alleles of sid1 and sid2, called eds5 and eds16, respectively, were identified independently by virtue of their enhanced disease-susceptibility phenotype. A recent breakthrough in our understanding of SA biosynthesis came when SID2/EDS16 was cloned by Wildermuth et al. and shown to encode a putative chloroplast-localized ICS. Mutations of the ICS1 gene, in sid2 and eds16, reduce SA accumulation after infection to only 5–10% of wild-type levels and compromise both basal and systemic resistance. This demonstrates that the isochorismate pathway in plants is the main source of SA synthesis during SAR. Consistent with this conclusion, ICS1 expression is induced by infection in both local and systemic tissues. Wildermuth et al. proposed that the phenylpropanoid pathway is responsible for the rapid production of SA associated with local cell death, whereas the isochorismate pathway is more important for sustained SA synthesis during development of SAR. Since SA synthesis is not completely abolished in sid2 plants, some SA must be produced either through the activity of another ICS-like protein, such as ICS2, or through the phenylpropanoid pathway.
Arabidopsis ICS1 contains a putative plastid transit sequence, suggesting that SA synthesis occurs in the plastid. Interestingly, EDS5/SID1 encodes another protein required for SA accumulation that has sequence similarity to the multidrug and toxin extrusion (MATE) family of transporter proteins. This suggests that EDS5 might be involved in moving SA or a phenolic precursor out of the plastid after synthesis.
Control of SA Synthesis
In plants such as tobacco and Arabidopsis, regulation of SA biosynthesis is an essential regulatory step in SAR activation. Therefore, identification of upstream regulatory components required for the induction of SA biosynthesis genes, especially ICS1, will be an important step toward understanding the control of SAR. The induction of ICS1 after infection by Erysiphe orontii and P. syringae pv. maculicola is not affected by depletion of SA in nahG plants, indicating that the ICS1 gene is not regulated by SA.
Many components upstream of ICS1 have been implicated in the regulation of SA synthesis, through characterization of various mutants with increased levels of SA. Most of these mutants form spontaneous HR-like lesions or have severe morphological phenotypes such as dwarfing. Expression of ICS1 is constitutively elevated in three such gain-of-resistance mutants, cpr1, cpr5, and cpr6 (constitutive expresser of PR genes). However, it is unclear whether these mutants directly affect SA synthesis, or whether SA levels are elevated as an indirect effect of cell death or disruption of cellular homeostasis. It has recently been shown that two mutants with elevated SA levels, ssi4 and snc1, have mutations in R genes that result in constitutive activation of a local defense response and therefore affect a step upstream of SA synthesis. Interestingly, snc1 plants do not form spontaneous lesions as observed in ssi4, suggesting that HR may not be required for activating SA biosynthesis.
SA synthesis induced by another R gene, RPS4, requires EDS1 and PAD4. The eds1 and pad4 mutants also block SA synthesis triggered by infection with virulent P. syringae. In eds1 and pad4, induction of EDS5, after infection with either virulent or avirulent P. syringae is blocked, places EDS1 and PAD4 upstream of EDS5 in the regulation of SA synthesis. Since EDS1 and PAD4 are required for resistance conferred by the same subset of R genes (TIR-NB-LRR) and have been shown to physically interact in planta, they are likely to function in the same pathway. However, the eds1 mutation significantly impedes the onset of HR and confers full susceptibility, whereas pad4 plants retain HR and show only intermediate susceptibility. This leads to the hypothesis that EDS1 contributes to initial SA accumulation and development of the HR downstream of TIR-NB-LRR type R genes, and then recruits PAD4 to drive amplification of the defense response by further increasing SA levels. Consistent with this hypothesis, EDS1 and PAD4 influence the expression of each other, with PAD4expression decreased more strongly in eds1 than EDS1 expression in pad4 . This suggests that EDS1 and PAD4 function in a positive feedback loop that amplifies their own expression and increases production of SA after infection. A role for a positive feedback loop in SA signaling is also supported by SA-mediated EDS1, PAD4, and EDS5expression. The similarity of EDS1 and PAD4 to lipases suggests that lipid metabolites may be involved in regulating the synthesis and/or accumulation of SA in local and systemic tissues.
Enhancement of the SA signal also occurs through a signal amplification loop involving ROS. The observation that SA binds the H2O2 scavenging enzymes catalase and ascorbate peroxidase (APX) and inhibits their activity led to the proposal that increases in H2O2 were responsible for signal transduction leading to PR gene induction and resistance. However, the concentrations of SA required for inhibition of catalase and APX are higher than those seen in systemic tissues after infection. Later studies suggested that H2O2 functions upstream of SA. Treatment of tobacco with high concentrations (>300 mM) of H2O2 leads to a dose-dependent accumulation of SA and PR-1 expression, which was suppressed in plants expressing nahG. Low concentrations of SA have also been shown to potentiate the production of ROS and HR cell death. In soybean cells inoculated with P. syringae, the addition of SA dramatically enhances the oxidative burst and cell death. It is hypothesized that in systemic tissues, the accumulation of low levels of SA together with the development of microbursts of ROS could amplify responses to secondary infections and contribute to SAR.
In addition to the signal amplification loops described above, there is evidence for negative feedback of SA synthesis. In the SA-insensitive npr1 mutant, levels of ICS1mRNA and SA are both elevated after infection compared to wild type. Furthermore, npr1 mutants show reduced tolerance to exogenous SA (0.5 mM), failing to develop beyond the cotyledon stage. The biological significance of such a feedback mechanism has yet to be determined; however, it might be utilized to shut off SAR once the pathogen challenge subsides. Many mutants with constitutively high levels of SA are dwarfs, and continuous spraying of wild-type plants with BTH also results in growth retardation (N. Weaver & X. Dong, unpublished observations), suggesting that accumulation of SA is detrimental to the plant's growth and development.