Salt Stress Activation of Wound-Related Genes in Tomato Plants

Plants in the field are exposed to multiple stresses, and their response to these various stresses determines their capacity to survive. Plants can use multiple signaling pathways and signals to mediate their response; for example, at least four different signaling pathways have been identified for water-deficit stress (Shinozaki and Yamaguchi-Shinozaki, 1997; Xiong et al., 2002). Different forms of stress may activate or utilize the same components, including proteins and other signaling molecules. Transgenic tobacco (Nicotiana tabacum) plants expressing constitutively active NPK1, a mitogen-activated protein kinase kinase kinase (MAPKKK) showed an enhanced tolerance to cold, heat shock, and salt stress (Kovtun et al., 2000). Signaling molecules such as jasmonic acid (JA) are involved in multiple stress responses and development in plants (Creelman and Mullet, 1995, 1997; Turner et al., 2002). However, it is the specific combination of various components of the signaling network coupled with spatial and temporal factors that allows the plant to mount a directed response to any given stress factor.

The phenomenon of cross-tolerance where a plant's resistance to one stress resulted in resistance to another form of stress is a growing area of research (Genoud and Metraux, 1999; Bowler and Fluhr, 2000). Increased tolerances to multiple stresses can involve the interaction between abiotic and biotic stresses. For example, plants pretreated with a sub-lethal dose of ozone or UV irradiation exhibited tolerance to a virulent pathogen (Yalpani et al., 1994; Sharma et al., 1996). Inoculation of Arabidopsis with a rhizobacterium made the plants more tolerant to drought and the pathogen Erwinia carotovora (Timmusk and Wagner, 1999). It has been shown that drought and UV radiation act synergistically to induce protective mechanisms in pea (Pisum sativum) and wheat (Triticum aestivum) plants (Alexieva et al., 2001). Grasses infected with endophytic fungi display enhanced tolerance to a variety biotic and abiotic stresses (Bacon et al., 1997; Malinowski and Belesky, 2000). Currently, very little is known on how different stresses interact with one another, or how their interaction will effect the plant's ability to respond to its environment or a given stress.

Wounding in tomato (Lycopersicon esculentum) plants is a well-characterized stress response. Tomato plants respond to mechanical wounding or herbivorous insect attack by inducing the synthesis of a wide array of defense-related proteins at the wound site and systemically throughout the distal portions of the plant (Bergey et al., 1996). Wounding causes the release and mobilization of an 18-amino acid polypeptide hormone called systemin (Pearce et al., 1991) and other wound-signaling peptides (Ryan et al., 2002). The binding of systemin to a cell surface receptor kinase (Scheer and Ryan, 2002) causes alterations in ion transport (Felix and Boller, 1995; Schaller and Oecking, 1999), the formation of hydrogen peroxide (Orozco-Cárdenas and Ryan, 1999; Orozco-Cárdenas et al., 2001), the activation of an MAPK (Stratmann and Ryan, 1997; Stratmann et al., 2000a), an increase of intracellular calcium (Moyen et al., 1998) and calmodulin (Bergey and Ryan, 1999), and the activation of a phospholipase A₂ (Lee et al., 1997; Narváez-Vásquez et al., 1999). The phospholipase acts on plant membranes to release linolenic acid, which is subsequently converted to the biologically active oxylipins, including 12-oxo-phytodienoic acid and JA via the octadecanoid pathway (for review, see Vick and Zimmerman, 1984; Farmer and Ryan, 1992; Peña-Cortés et al., 1993; Blechert et al., 1995; Doares et al., 1995a, 1995b; Bergey et al., 1996). In addition to the elevated levels of oxylipins, wounding induces the synthesis of ethylene, which appears to act synergistically with JA to induce the expression of the defense genes (Felix and Boller, 1995; O'Donnell et al., 1996).

The wound response in tomato plants has been shown to be inhibited as well as activated by other stresses. The signaling molecule salicylic acid produced in response to infection with biotrophic pathogens was shown to inhibit the synthesis of proteinase inhibitors in response to wounding and herbivory (Doherty et al., 1988; Peña-Cortés et al., 1993; Doares et al., 1995b). However, when tomato plants were challenged with the pathogen Pseudomonas syringae pv. tomato (Pst), the causal agent for bacterial speck disease, it increased plants resistance to chewing insects (Bostock et al., 2001). This increased resistance may be due to elevated levels of JA present in plant tissues as a result of the infection by the necrotrophic pathogen (Pieterse and van Loon, 1999). The exposure of tomato seedlings to a form of abiotic stress, UVC radiation, resulted in a JA-mediated activation of wound-related genes and the synthesis of proteinase inhibitors (Conconi et al., 1996). However, when tomato plants were exposed to UVB/UVA radiation, they did not accumulate proteinase inhibitors, but showed a strong potentiation of the systemic induction of proteinase inhibitor synthesis in response to wounding (Stratmann et al., 2000b). It has also been reported that the wound-inducible defense gene Leu aminopeptidase is activated by water deficit, salinity, and abscisic acid (Chao et al., 1999).

In this report, we have investigated the effects of salt stress on the wound response cascade in tomato plants. Salt stress alone was found to induce wound-related genes and this gene activation was mediated via the octadecanoid pathway. The salt-dependent activation of the octadecanoid pathway was found to be independent of the wound prohormone, prosystemin, but prosystemin (PS) activity was necessary to achieve maximal accumulation of proteinase inhibitors. In addition, salt stress was found to strongly enhance the plant's ability to respond to wounding. This analysis describes how an abiotic stress activates the wound response in tomato, and provides further insight into defense gene signaling and the interaction between stress responses in plants.

RESULTS

DISCUSSION

Wounding in tomato plants is a well-characterized stress response. Tomato plants respond to mechanical wounding and herbivore attack by inducing the local and systemic synthesis of a wide variety of defense-related proteins, including proteinase inhibitors (Bergey et al., 1996; Ryan, 2000). The levels of accumulation of wound-inducible proteinase inhibitors in the leaves reflects the magnitude of the defense response. The activation and enhancement of the wound response by UV irradiation prompted us to investigate the effect of another abiotic stress, salt on the wound response. Our analysis revealed that a 24-h exposure to salt stress alone was able to induce the accumulation of the wound-inducible Inh II protein in a dose-dependent manner (Fig. 1).

Prolonged exposure to salt stress over a 5-d period resulted in the accumulation of high levels of proteinase inhibitors (Fig. 2). Salt-induced Inh II levels were as high as 350 to 450 μg mL^-1 leaf juice, in contrast to a typical strong wound response that produces proteinase inhibitor levels of 100 to 120 μg mL^-1 leaf juice. These increased levels were even more dramatic in sense-PS plants, obtaining levels of almost 600 μg mL^-1 in leaf juice. This extremely high level in sense-PS plants is probably even higher because the concentrations of Inh II are at the limiting range of the radial immunodiffusion assay. All plants treated with 200 mm NaCl for 5 d grew slowly, with darker green leaves and were roughly one-half the size of untreated plants. It has been shown that prolong exposure to salt stress and drought results in lower water content in the leaves of tomato plants (Cuartero and Fernandez-Muñoz, 1999). We used northern-blot analysis to determine whether the increase in Inh II protein concentration was a result of an up-regulation in the gene expression or as a result of diminished water content in the leaf (Fig. 2). Although it is likely that dehydration of the plant contributes to elevated Inh II concentration in the leaf, the sustained high levels of Inh II transcript over the 5 d shown in Figure 2B suggests active protein synthesis. In addition, the 1.75-fold increase in Inh II transcript level on d 5 correlated very closely with the increase in protein accumulation observed in the leaf. Therefore, the extremely high Inh II concentrations observed in the leaves of the salt-treated plants were probably a result of a combination of factors: a high level of gene activation and protein synthesis, diminished water content in the leaf, and protein turnover during the 5-d period.

Northern-blot analysis revealed that salt stress elevated the transcript levels for the wound-signaling gene LOX, suggesting the involvement of the octadecanoid pathway in mediating the response. This was further confirmed when the def-1 mutant subjected to salt stress showed a severe impairment in the accumulation of the Inh II protein as compared with similarly treated wild-type plants (Fig. 3), and its failure to respond to both wounding and salt stress (Fig. 7). The def 1 mutant is impaired in the octadecanoid pathway, compromising its ability to accumulate JA in response to wounding and elicitors (Howe et al., 1996). This indicates that an intact octadecanoid pathway is necessary for the salt-induced accumulation of the wound-inducible Inh II protein. However, when exposed to salt stress, the def-1 mutant did accumulate low levels of Inh II protein, 10% to 20% of wild-type levels. Since the def-1 mutant has been described as a “leaky” mutant (Howe et al., 1996), the observed Inh II accumulation is likely due to either low-level JA biosynthesis or the action of another component associated with water stress, such as abscisic acid (Peña-Cortés et al., 1989, 1996; Wasternack et al., 1996).

In addition to salt stress, we found that wilting alone also induced the accumulation of proteinase inhibitors, and this response was also mediated via the octadecanoid pathway (data not shown). These data support the findings from a microarray analysis in Arabidopsis showing that a subset of wound-inducible genes was activated in dehydrating leaves (Reymond et al., 2000). However, because the analysis was conducted on excised leaves and not in whole plants, it was unclear if the increased transcript levels resulted from dehydration or a wound stimulus.

To further investigate the effects of salt stress on activation of wound-related genes, we investigated how salt stress that is initially perceived in the roots would affect the transmission of the wound signal in the leaves. This allowed us to examine the interaction of two distinct stresses that are initially perceived in different tissues and at distant locations. The data in Figure 5 show that tomato plants subjected to wounding and salt stress exhibit an enhanced response to wounding when compared with plants subjected to wounding alone. This salt-induced enhancement of the wound response occurs in the local wounded leaves, but was found to be more pronounced in the systemic unwounded leaves.

The increased accumulation of Inh II due to wounding and salt treatment cannot be explained solely by the additive effect of salt stress and wounding alone. We found that the simultaneous exposure to 100 mm NaCl and wounding increased Inh II levels an average of 60% higher in the wounded leaves, and 50% higher in the systemic leaves from what we would have expected if it was merely an additive effect. This enhancement was even more pronounced in plants subjected to 200 mm NaCl stress, where the increases were found to be nearly 100% higher in wounded and 180% in systemic leaves above the expected values for an additive effect. These data indicated that salt stress not only activates the wound response, but it also strongly enhances the response. These results are similar to the effects shown by tomato plants subjected to wounding and UVB/A irradiation (Stratmann et al., 2000b).

PS was another wound-signaling gene that was found to be up-regulated by salt stress. However, we showed that PS activity was not required for salt stress-induced accumulation of Inh II protein. This finding was based on the analysis of the effect of salt stress on the spr-1 mutant and antisense-PS tomato plants. The transgenic antisense-PS plant has a strongly suppressed PS expression (McGurl et al., 1992), resulting in a substantially reduced ability to systemically accumulate proteinase inhibitors in response to wounding, whereas the spr-1 tomato mutant is defective in PS perception and the production of a systemic wound signal (Lee and Howe, 2003). Exposure of the spr-1 mutant and antisense-PS tomato plants to salt stress resulted in a significant accumulation of Inh II, roughly 50% of wild-type levels. These results suggested that PS was not necessary for salt-induced accumulation of the Inh II protein. To further investigate the role of PS in the salt-induced accumulation of proteinase inhibitors, we subjected plants overexpressing the cDNA for PS (sense-PS) to salt stress. Sense-PS plants exhibit a constitutive wound phenotype, synthesizing and accumulating wound-related proteins to high levels (McGurl et al., 1994; Bergey et al., 1996). Just as with wounding, we wanted to examine if salt treatment could further enhance the accumulation of Inh II in sense-PS plants. We found that sense-PS plants exposed to salt stress had dramatically increased levels of Inh II in the leaves (Fig. 8). These results combined with data that showed reduced levels of Inh II in the spr-1 and antisense-PS plants (Fig. 4) indicate that although PS activity is not required for salt activation of wound response genes, it is necessary to achieve maximum levels of accumulation.

When we investigated the effects of salt stress on wounding in the spr-1 and antisense-PS backgrounds, we were surprised to discover that salt stress had appeared to restore the systemic accumulation of Inh II, although at 50% the level observed in wild-type plants (Fig. 7). The systemically accumulated Inh II levels cannot be explained by just salt stress induction alone because the Inh II levels were at least two times higher than what is observed in the unwounded salt treated controls (Figs. 1 and 4). However, it is unclear how this restoration of the systemic response occurs.

What insights can we derive from our analysis of salt stress activation of the wound response with respect to signaling? We found that the initial activation of Inh II gene occurs 4 h after wounding or salt treatment alone. This suggests that the initial salt-induced signal from the root travels (Fig. 1) at the same velocity to the leaf as the wound signal from wounded to systemic leaf (Fig. 6). However, these signals were generated differently. In leaves, a single transient wound across the mid-vein was made, whereas intact roots where exposed to a continual salt stimulus. Another possibility could be that the salt-induced signal travels faster but the concentrations are very low, requiring 4 h to reach levels that can induce the gene. This salt-induced gene activation appears to occur in two phases. One can speculate that the initial signal leading to gene activation originates from the roots (Jackson, 1997), whereas the second may be produced in the leaf. Because the octadecanoid pathway mediates the salt-induced accumulation of Inh II, it follows that the signal generated may be JA or some other oxylipin. There is evidence from other studies that show increased JA accumulation in the roots and leaves of salt-stressed plants (Moons et al., 1997; Wang et al., 2001). These salt-induced increases in JA levels could occur as a by-product of alterations in membrane composition and structure, or as a result of the activation of fatty acid desaturases, which lead to release and accumulation of linolenic acid, the precursor to JA biosynthesis (Khaware et al., 1995; Surjus and Durand, 1996; Nishiuchi and Iba, 1998).

There is evidence suggesting that JA acts as the long-distance wound signal (Zhang and Baldwin, 1997; Li et al., 2002). In addition, the recent analysis of the spr-1 mutant indicates that a synergistic interaction between JA and systemin is necessary to generate and transduce the systemic signal, and that the mutant displayed reduced JA accumulation in wounded leaves (Lee and Howe, 2003). Our results appear to be consistent with the latest findings with respect to wound signaling, and the synergistic interaction of JA and PS.

Based on these recent findings, we can propose a mechanism for accumulation of Inh II in response to salt stress. Salt stress leads to the synthesis and release of JA. JA induces the synthesis of the PS and Inh II proteins. The newly synthesized PS can activate the wound-response pathway (Dombrowski et al., 1999), resulting in the further release of linolenic acid from the membranes leading to JA biosynthesis and the accumulation of additional Inh II and PS, thereby creating a positive feedback loop that amplifies the salt-induced wound response. Coupling this feedback loop with a salt-induced, PS-independent JA biosynthesis results in the accumulation of very high levels of proteinase inhibitors.

Whether the strong induction of the wound response by salt stress is relevant to the success of the plant in the field is not known. In addition, it is unclear if this activation provides the plant any increased tolerance to water-deficit stress. Sense-PS plants were found to survive higher initial concentrations of salt than wild-type plants (J.E. Dombrowski and C.A. Ryan, unpublished data). One possible explanation for their increased survival rate could be that in addition to accumulating wound-related proteins, there is a concurrent activation of water deficit-tolerance genes. In support of this hypothesis, microarray analysis revealed that mechanical wounding activated osmotic stress-related genes in Arabidopsis (Reymond et al., 2000; Cheong et al., 2002). The survivability of the sense-PS plants could also be attributed to the presence of high levels of proteinase inhibitors in the leaves or even PS itself, a highly hydrophilic protein (McGurl et al., 1992), which could function as osmoprotectants.

JA has been implicated in a wide range of stress responses, as well as development in plants (Creelman and Mullet, 1995, 1997; Turner et al., 2002). It has been reported that pretreatment of barley (Hordeum vulgare) seedlings with JA improved their performance to salt stress (Tsonev et al., 1998). Sorbitol stress in barley led to a low and transient rise in JA levels in the leaves (Kramell et al., 2000). Salt stress increased JA levels in the roots of rice (Oryza sativa) plants and in the leaves of Iris hexagona (Moons et al., 1997; Wang et al., 2001). However, it is still unclear exactly what role JA plays in the water-deficit response, and more research is needed to better define its function.

Why would a plant redirect energy and resources to activate a defense pathway and accumulate such high levels of proteinase inhibitors during a period of water deficit? One possible rationale could be that plants under water stress display reduced growth, which results in decreases in the overall biomass for herbivorous insects to feed on. The activation of the wound response by water-deficit stress would protect the plant against defoliating chewing insects during periods of low growth and would preserve limited foliage until water resources are no longer limiting, thereby improving their survivability. Alternatively, the water deficit-mediated activation of the wound response could simply be a serendipitous event due to functional redundancy in stress signaling networks.

We are beginning to discover that responses to stress are not linear pathways, but are complicated integrated circuits involving multiple pathways and specific cellular compartments, tissues, and the interaction of additional cofactors and/or signaling molecules to coordinate a specified response to a given stimulus. The work described here provides another important step toward unraveling the interactions of stress-activated pathways in plants.

MATERIALS AND METHODS