The allene oxide cyclase (AOC)-catalyzed step in jasmonate (JA) biosynthesis is important in the wound response of tomato. As shown by treatments with systemin and its inactive analog, and by analysis of 35S::prosysteminsense and 35S::prosysteminantisense plants, the AOC seems to be activated by systemin (and JA) leading to elevated formation of JA. Data are presented on the local wound response following activation of AOC and generation of JA, both in vascular bundles. The tissue-specific occurrence of AOC protein and generation of JA is kept upon wounding or other stresses, but is compromised in 35S::AOCsense plants, whereas 35S::AOCantisense plants exhibited residual AOC expression, a less than 10% rise in JA, and no detectable expression of wound response genes. The (i) activation of systemin-dependent AOC and JA biosynthesis occurring only upon substrate generation, (ii) the tissue-specific occurrence of AOC in vascular bundles, where the prosystemin gene is expressed, and (iii) the tissue-specific generation of JA suggest an amplification in the wound response of tomato leaves allowing local and rapid defense responses.
Plants have to adjust to numerous biotic and abiotic stresses, and many of the accompanied responses are mediated by jasmonates (Wasternack and Hause, 2002). The term jasmonate is used for the cyclopentanone compound jasmonic acid (JA), its methyl ester (JAME) and derivatives such as JA amino acid conjugates. In the last decade, jasmonates were recognized as central elements in a lipid-based signalling cascade, which has a pivotal role in plant defense reactions against herbivore and pathogen attack (Kessler and Baldwin, 2002; Wasternack and Hause, 2002). As well as JA, its precursor 12-oxophytodienoic acid (OPDA) and related compounds, collectively named octadecanoids, were suggested to function as independent signals (Blechert et al., 1999; Kramell et al., 2000; Stintzi et al., 2001).
Initially, expression of defense genes such as those coding for proteinase inhibitors (PINs) by airborne JAME or wounding was observed in tomato leaves (Farmer and Ryan, 1990). Later on, numerous biotic and abiotic stimuli were found to induce expression of specific genes via an endogenous rise of octadecanoids and jasmonates. Beside wounding, such stimuli are elicitation of cell suspension cultures (Parchmann et al., 1997), pathogen attack (Penninckx et al., 1996), osmotic stress (Kramell et al., 2000), or chitosan and oligogalacturonide treatment (Doares et al., 1995). Other important aspects of JA action are its role in seedling and flower development (Hause et al., 1996, 2000) and the downregulation of house-keeping proteins such as Rubisco (Reinbothe et al., 1993).
The wound response in tomato is one of the best-studied jasmonate signaling pathways (Ryan and Pearce, 1998; Ryan, 2000). In tomato, local wounding leads to expression of the prosystemin gene within vascular bundles (Jacinto et al., 1997; McGurl et al., 1994). The encoded pro-peptide of 200 amino acids is processed into a peptide of 18 amino acid residues, called systemin (Ryan et al., 2002), which can activate the recently cloned membrane-located receptor (Scheer and Ryan, 2002), whereas its analog with alanine at position 17 (Sys-17) is inactive (Dombrowski et al., 1999). Downstream events of the signal perception are activation of a MAPK, a Ca2+ ion flux, and rise in linolenic acid and jasmonate levels (Conconi et al., 1996; O'Donnell et al., 1996). Finally, expression of various defense genes such as that coding for PINs is followed by a systemic response due to loading of systemin in the phloem and transport to upper leaves (Ryan, 2000).
The biosynthesis of JA, elucidated by Vick and Zimmerman (1983), proceeds via an oxylipin pathway (Figure 1). As a first step, α-linolenic acid (α-LeA) is released from chloroplast membranes as indicated by JA deficiency of the mutant of Arabidopsis thaliana delayed anther dehiscence1 (dad1) which is defective in a chloroplast-located phospholipase of the A1-type (Ishiguro et al., 2001). Subsequently, molecular oxygen is inserted by a lipoxygenase (LOX)-catalyzed reaction at position 13 of α-LeA. The resulting fatty acid hydroperoxide ((13S)-hydroperoxy(9Z,11E,15Z)-octadecatrienoic acid, 13-HPOT) is dehydrated by an allene oxide synthase (AOS) to an allene oxide. This highly unstable compound can rapidly decay by chemical hydrolysis to α- and γ-ketol and racemic OPDA (Figure 1) or, in presence of the allene oxide cyclase (AOC), can be cyclized exclusively to the cis-(+)-enantiomer (9S,13S) of OPDA. Only this enantiomer carries the stereochemistry of the naturally occurring (+)-7-iso-JA. OPDA reductase (OPR) and three β-oxidation steps lead to JA.
Most of the genes coding for the enzymes of JA biosynthesis are cloned. Several cDNAs coding for different isoforms of 13-LOX were isolated from various plants including tomato (Heitz et al., 1997). However, there is no functional proof so far, to indicate which of the various 13-LOX forms acts in JA biosynthesis. The LOX-product 13-HPOT can serve as a substrate for various branches of the LOX pathway (Feussner and Wasternack, 2002). In the AOS-catalyzed reaction, 13-HPOT is directed to the formation of signaling compounds such as octadecanoids and jasmonates. cDNAs coding for AOS were isolated from various plants including tomato (Howe et al., 2000; Sivasankar et al., 2000). Recently, the first cDNA coding for AOC was isolated from tomato (Ziegler et al., 2000). Finally, among three Arabidopsis sequences coding for OPRs, only the OPR3 is active in JA biosynthesis (Müssig et al., 2000; Schaller et al., 2000).
Recently, two types of mutants were isolated both of which suppressing 35S::prosystemin-mediated signaling (Li et al., 2001). Mutants of one group are affected in JA biosynthesis, e.g. spr-2, whereas another group contains JA- and systemin-insensitive mutants, such as jai-1 (Li et al., 2002). Grafting experiments with both mutants and wild type led to the final proof that a functional JA biosynthetic pathway is required for the local wound response, but not for the systemic response (Li et al., 2002). It is less understood, however, how the local rise of JA is regulated. In tomato, JA biosynthetic enzymes such as LOX, AOS, and AOC are transcriptionally upregulated upon wounding (Heitz et al., 1997; Howe et al., 2000; Sivasankar et al., 2000; Ziegler et al., 2000). In case of AOC, an increase in mRNA within the first 30 min upon wounding coincided with the exclusive formation of cis(+)OPDA (Ziegler et al., 2000), whereas in case of AOS1 and AOS2 the mRNA accumulation appeared later (Howe et al., 2000; Sivasankar et al., 2000; Stenzel et al., 2002). In untreated tomato leaves, however, the AOC protein is specifically detectable in parenchymatic cells of vascular bundles (Hause et al., 2000), the tissue of prosystemin gene expression and systemin generation (Jacinto et al., 1997), suggesting a spatial and temporal role of AOC as an intermediate step of the wound response pathway downstream of systemin.
Here, we show that the constitutive occurrence of AOC in vascular bundles (Hause et al., 2000) contributes to JA formation if the AOC substrate is generated upon wounding. The vascular bundle-specific occurrence of AOC is kept upon wounding or other treatments. Consequently, JA levels are preferentially elevated in main veins compared to surrounding areas. Inspection of the corresponding wild types, 35S::AOCsense (AOCs), 35S::AOCantisense (AOCas), 35S::prosysteminsense (PrSs) and 35S::prosysteminantisense (PrSas) plants for levels of AOC mRNA, AOC protein, AOC activity, for tissue-specific location of AOC protein as well as for levels of OPDA, JA and PIN2 mRNA suggests a systemin-dependent AOC expression and an AOC-dependent amplification of the local wound response via vascular bundle-specific generation of jasmonates.
AOC mRNA accumulates upon systemin treatment and precedes pin2 mRNA accumulation
The recently cloned AOC of tomato was found to be wound inducible (Ziegler et al., 2000), and specifically expressed in vascular bundles (Hause et al., 2000), the tissue in which the prosystemin gene is expressed (Jacinto et al., 1997). This prompted us to inspect whether AOC expression depends on systemin. As shown in Figure 2(a), AOC mRNA accumulated upon wounding and systemin treatment with similar kinetics. Both mRNA accumulations could be decreased significantly by pre-treatment with the inactive systemin analog Sys17. These data suggest that at least locally wound-induced expression of AOC may occur via systemin. A final proof, however, is still missing showing that accumulation of mRNAs coding for JA biosynthesis enzymes such as AOS (Howe et al., 2000) or AOC (Figure 2) is necessary and sufficient for the rise in JA known to occur in tomato leaves following wounding (Heitz et al., 1997; Ziegler et al., 2000). Therefore, kinetics of mRNA accumulation of AOC and the JA-inducible gene PIN1 were recorded. Treatments such as wounding, or floating of detached leaves on 50 µm JAME led to AOC mRNA accumulation peaking at about 1–2 h which clearly preceded PIN1 mRNA accumulation (Figure 2b).
No rise in JA and OPDA, however, could be found upon treatment with 12-OH-JA, although this compound induced AOC mRNA accumulation (Figure 2c, lane 2). Also, upon treatment with the ethyl-indanoyl isoleucine conjugate, which mimics the action of JA without rise in JA (Koch et al., 1999), AOC mRNA appeared and the lack of rise of JA was confirmed (Figure 2c). Obviously, this compound can induce per se jasmonate-responsive gene expression, since PIN2 mRNA accumulated (Figure 2c). These data suggest that there is no causal link between AOC mRNA accumulation and rise in JA formation within the first hour of treatment.
AOC protein was found to be confined to vascular bundles and its localization was not altered upon wounding and JA treatment
Inspection of AOC protein level following wounding or treatment with JAME revealed constant levels until 24 h of treatment (Figure 3a). This contrasts with the strong and transient changes in AOC mRNA levels following identical treatments (Figure 2b). Immunocytochemical analysis revealed that the previously shown confinement of AOC protein to vascular bundles (Hause et al., 2000) is kept upon treatment of detached wild-type leaves with water or JA for 24 h, or upon wounding for 30 min (Figure 3c). These results indicate that the AOC protein detected by Western blot analysis is preferentially, if not exclusively, confined to vascular bundles. Interestingly, wounding performed locally by puncture with a needle and indicated by autofluorescence of damaged cells (see arrow in Figure 3c), did not lead to AOC accumulation in surrounding mesophyll cells. A confinement of AOC protein in vascular bundles was also observed upon treatment of wild-type leaves with systemin and in wounded leaves of 35S::prosysteminsense plants (data not shown). In contrast, the LOX protein was detected in all leaf tissues of an unwounded (Figure 3d) and wounded leaf (data not shown), suggesting that LOX products are generated everywhere. The AOS protein was also detected in all tissues of an unwounded leaf (Figure 3e). Immunolabeling experiments with the corresponding pre-immune serum did not show any label (Figure 3b).
Amplification of the wound response by AOC
To analyze the suggested systemin dependence of AOC expression/JA generation, we compared the wild-type Better Boy (BB) with PrSs and PrSas plants.
Due to the constitutive expression of the prosystemin gene in these plants (McGurl et al., 1994), constitutive elevation of AOC mRNA, AOC protein, AOC activity leading to elevated levels of JA or OPDA was expected in freshly harvested leaves. This was not the case (Figure 4d). There was, however, a higher and earlier rise of JA in wounded transgenic leaves of 6-week-old plants than that of the wild-type Better Boy (Figure 4d). Furthermore, detached unwounded leaves of 3-week-old PrSs plants accumulated already about 1.2 nmol g−1 FW JA and 1.4 nmol g−1 FW OPDA. Such levels are sufficient to induce PIN2 expression, and might be the reason for the observed constitutive PIN2 expression in 2-week-old PrSs plants (McGurl et al., 1994). In contrast, in 6-week-old plants, the JA levels are not different in unwounded leaves (Figure 4d). However, differences can be observed between wild-type and PrSs or PrSas plants after wounding.
Presumably caused by the low differences in constitutive prosystemin mRNA levels between wild-type and PrSas plants (Orozco-Cardenas et al., 1993), both types of leaves exhibited no difference in AOC mRNA levels without wounding. There was, however, less AOC mRNA level in PrSas than in wild-type leaves following wounding (data not shown), and wound-induced JA levels were significantly lower in PrSas leaves than in wild-type leaves indicating systemin dependence (Figure 4d).
To inspect the role of AOC and its vascular bundle-specific occurrence in the local wound response of tomato, we generated AOCs and AOCas plants and compared them with the corresponding wild-type Lukullus (Lu).
These plants exhibited a strong constitutive accumulation of AOC mRNA and AOC protein (Figure 4a). Immunocytochemical analysis revealed AOC expression in all leaf tissues of AOCs plants, thereby contrasting with the tissue-specific expression in wild-type leaves (Figures 3 and 5). JA and OPDA levels in freshly harvested (0-time) leaves of wild-type and AOCs plants were identical (Figure 4b,c), indicating that the high levels of AOC mRNA and AOC protein (Figure 4a) did not cause an increase in these compounds. In vitro AOC activity (Table 1) of freshly harvested leaves was 10-fold higher in AOCs plants than in wild-type plants, indicating that the capacity to form OPDA and JA is elevated. Only upon wounding, however, an increase could be detected for both compounds (Figure 4b,c). This wound-induced rise was more than twofold higher in AOCs leaves than in wild-type leaves.
Table 1. In vitro AOC activity in crude extracts of unwounded and wounded tomato leaves
nmol cis(+)-OPDA per mg protein after wounding
AOC activity was detected by measuring formation of cis(+)-OPDA as described in Experimental procedures.
AOC mRNA was undetectable in freshly harvested leaves and no AOC protein was detected by Western blot analysis (Figure 4a). A residual amount of AOC protein was found by immunocytochemical analysis (Figure 5) which might be responsible for the AOC activity detected in vitro (Table 1). Upon wounding, only a weak accumulation of AOC mRNA occurred, indicating that AOC expression is strongly diminished. Consequently, in wounded AOCas leaves, no increase was found for OPDA, and only a minor increase was detected for JA, indicating a loss of AOC function. The lack of PIN2 mRNA accumulation following wounding and systemin treatment clearly indicates an AOC dependence of the wound response. Furthermore, upon systemin treatment of AOCas leaves for 30 min, only 0.25 nmol g−1 FW of JA were found, whereas the corresponding wild-type leaves accumulated 1.75 nmol g−1 FW.
Finally, we analyzed the JA-deficient mutant def1 which has reduced wound-induced JA accumulation (Howe et al., 1996) compared with the wild-type Castlemart.
Upon wounding of detached def1 leaves a similar kinetics of induction but lower AOC mRNA accumulation was observed compared to the wild-type Castlemart (Figure 4a). Again, levels of AOC protein remained unchanged in both cases. The data suggest that the diminished JA formation of def1 in response to wounding observed previously (Howe et al., 1996) is not due to an impaired AOC gene. To prove this, we sequenced AOC cDNAs isolated from def1 (Acc. No. AJ308482), Lycopersicon esculentum cv. Moneymaker (Acc. No. AJ308481) and L. pennellii and compared the sequence with that of L. esculentum cv. Lukullus (Ziegler et al., 2000). AOC sequences of def1, Moneymaker and Lukullus were identical (data not shown). Another possible explanation for the signaling defect in def1 plants might be found in the tissue-specific expression of AOC. However, the immunocytochemical inspection of def1 leaves revealed exclusive occurrence of AOC protein in parenchymatic cells of vascular bundles similar to that of the wild type (Figure 5). Def1 leaf protein extracts exhibited much lower in vitro AOC activity compared to the wild type even upon wounding (Table 1). In def1 leaves, the wound-induced levels of OPDA (200 pmol g−1 FW at 40 min) and JA (450 pmol g−1 FW at 40 min) were also lower than that of wounded wild-type leaves (350 pmol g−1 FW OPDA and 750 pmol g−1 FW JA). Taken together, these data indicate that def1 plants possess an intact AOC gene expression and the AOC protein has an identical localization as that of the wild type, but the mutant exhibits a deficiency in AOC activity.
The data can be summarized as follows:
1PrSs plants are apparently pre-conditioned by overexpression of the prosystemin which in turn may contribute to the rapid rise in JA levels upon wounding.
2A systemin dependence was suggested by less wound-induced AOC mRNA accumulation and JA formation in PrSas than in wild-type leaves.
3Overexpression of AOC increased AOC protein and AOC in vitro activity constitutively, whereas levels of JA and OPDA and the PIN2 transcript increased only upon wounding.
4AOCantisense plants lack any wound response in terms of PIN2 expression, whereas AOC protein and OPDA were below the detection limit, and the in vitro AOC activity and JA levels were strongly diminished.
5In all plants except the AOCs plants, AOC protein was confined to vascular bundles.
Preferential generation of jasmonates in main veins of tomato leaves
The specific occurrence of AOC protein in vascular bundles suggests that elevation of jasmonates upon wounding or other treatments might occur predominantly in veins. Indeed, in leaf areas of the wild-type Castlemart containing predominantly main veins, the JA level increased within 30 min upon wounding from 0.3 to 2.2 nmol g−1 FW, whereas in the remaining leaf lamina, the levels increased only from 0.2 to 1.0 nmol g−1 FW (Table 2). Also, OPDA increased preferentially in main vein-containing leaf areas, whereas dinorOPDA was neglected in all measurements due to its minor occurrence. In contrast, the JA level and the OPDA level differed only slightly between main vein-containing leaf area and the remaining leaf tissues in AOCs plants (Table 2), thus reflecting the abundant and equal occurrence of AOC in all leaf tissues (Figure 5). A residual amount of JA in both types of wounded leaf areas were found in the AOCas plants.
Table 2. Preferential accumulation of OPDA and JA in main veins of wild type and wounded leaves of AOCs and AOCas plants
Line/time of wounding
Preferential accumulation of OPDA and JA occurs in main veins of wild-type leaves following wounding, but not in wounded leaves of AOCs plants and AOCas plants. Main vein-containing leaf areas were cut off from the remaining tissues within 20 sec, and both leaf tissues were frozen by liquid nitrogen and subjected to GC/MS analysis as described in Experimental procedures. Data of three independent sampling and extractions were obtained and deviated by 15%. One representative set of data is shown in nmol g−1 FW.
The preferential biosynthesis of jasmonates in the main vein compared to the leaf lamina is also indicated by the amount of (+)-7-iso-JA, the initial product in JA biosynthesis, thereby indicating de novo synthesis. In the main vein-containing tissues of wounded wild-type leaves, 70% of all jasmonates were (+)-7-iso-JA, whereas only 15% were in the leaf lamina. In contrast, in wounded AOCs plants, both types of tissues contained about 60% (+)-7-iso-JA of total amount of jasmonates, indicating identical capacity for de novo synthesis of JA.
Among the lipid-derived compounds, octadecanoids and jasmonates have a crucial role in plant responses to biotic and abiotic stresses such as wounding (León et al., 2001; Wasternack and Hause, 2002; Weber, 2002). Jasmonic acid is suggested to be one of several interacting signals which, in tomato, lead to expression of PINs and other plant defense genes (Howe and Schilmiller, 2002; Ryan, 2000). JA can act as an intra- and intercellular as well as interorganismic signal (Farmer, 2001). Upon wounding, JA levels increase locally and systemically (Herde et al., 1996), but it is unclear how this is regulated. Recent grafting experiments (Li et al., 2002) with the JA-insensitive tomato mutant jai-1 and the spr2 mutant, which is defective in JA biosynthesis, revealed strong evidence that JA or a derivative may act as a long-distance transmissible signal in wound signaling (Ryan and Moura, 2002). It is still unknown, however, how JA levels are elevated. Here, we show that an amplification of the local wound response may occur by activation of AOC and generation of jasmonates in vascular bundles presumably in a systemin-dependent manner.
In tomato, the AOC is encoded by a single-copy gene (Ziegler et al., 2000). Thus, the immunological detection of AOC protein in vascular bundles (Hause et al., 2000) indicates that all AOC is confined in this tissue. This is reminiscent of a vascular bundle-specific location of the JA-inducible vegetative storage proteins in soybean (Huang et al., 1991), which is de-localized following jasmonate treatment. The confinement of AOC to vascular bundles of tomato is kept upon JA- or wound-induced AOC expression (Figure 3c). As a consequence, the capacity of the different leaf areas to form JA may differ, preferentially in leaf areas differing in the types of veins. In the minor veins, the AOC protein was found exclusively in the cells of the bundle sheet (Figures 3 and 5). In contrast, main veins and leaf stalks, both of them composed of many vascular bundles (Figure 5), contain AOC protein within the parenchymatic cells. Therefore, leaf stalks and main veins carry a higher total amount of AOC protein than a comparable leaf area containing minor veins and mesophyll cells. Since the LOX protein and the AOS protein were equally distributed throughout the leaf tissues (Figure 3d,e), mainly the AOC protein might be of special importance in this preferential generation of jasmonates in main veins compared to surrounding leaf areas (Table 2). The immunological analysis of LOX and AOS, however, was hampered by the fact that the antibodies did not distinguish between the various LOX and AOS forms appearing in tomato leaves (Heitz et al., 1997; Howe et al., 2000). Thus, there is still the possibility that a LOX and/or an AOS form occurs in a tissue-specific manner. Nevertheless, leaves of AOCs plants exhibiting constitutively high level of AOC protein in all leaf tissues did not show such a preferential JA accumulation (Figure 5; Table 2). Thus, localized generation of OPDA and JA might be preferentially caused by the tissue specificity of AOC. Moreover (+)-7-iso-JA accumulated preferentially in the main vein-containing leaf area of a wounded wild-type leaf, but was equally distributed in wounded AOCs leaves. (+)-7-iso-JA is the initial product in JA biosynthesis which isomerizes after synthesis into the more stable (–)-JA. Thus, the appearance of (+)-7-iso-JA above the basal level indicates de novo synthesis.
The preferential generation of jasmonates and exclusive localization of AOC protein in the same tissue, in which systemin is generated from its precursor prosystemin, substantiate the recently proposed model on local wound signaling (Orozco-Cardenas et al., 2001; Ryan, 2000). According to this model, the prosystemin gene expression occurs in vascular bundles and is JA and wound inducible (Jacinto et al., 1997; Ryan and Moura, 2002). Subsequently, prosystemin is processed to systemin which is released into the vascular system. Here, systemin can be a mobile signal (Narváez-Vásquez et al., 1995) and/or can be perceived locally by the systemin receptor (Scheer and Ryan, 2002), possibly in the vascular parenchyma cells (Orozco-Cardenas et al., 2001; Ryan and Moura, 2002). Due to AOC expression in these cells (Figure 5), the following scenario is suggested: The preferential generation of JA in main veins occurring in the first hour following wounding is the consequence of wound-induced generation of systemin and vascular bundle-specific occurrence and activiation of AOC protein. Subsequently, the generated JA may contribute to the JA-inducible prosystemin gene expression in the same tissue early after wounding. Such spatial and temporal coincidence in the generation of cumulatively acting signals would amplify the wound response. Consistent with this model is that (i) in comparison to the corresponding wild type, young PrSs leaves exhibited elevated levels of JA and OPDA, constitutive occurrence of PIN2, and rapid generation of JA upon wounding (Figure 4d); (ii) PrSas leaves exhibited less AOC mRNA accumulation and JA formation upon wounding than wild-type leaves; and (iii) systemin treatment elevates in vitro AOC activity in wild-type leaves but much less in AOCas leaves (data not shown). The data are consistent with a systemin-dependent AOC activation, but additional triggering by JA cannot be excluded. Finally, JA is essential for the local wound response in terms of PIN2 expression, since PIN2 is absent in wounded AOCas leaves (Figure 4a) and cannot be induced by systemin treatment of these leaves (C. Kutter and C. Wasternack, unpublished).
Another scenario is highlighted by the vascular bundle-specific occurrence and activation of AOC and the local generation of JA adjacent to the veins. The paraveinal cells adjacent to the veins exhibit a more rapid oxidative burst in the incompatible plant pathogen interaction of Arabidopsis (Alvarez et al., 1998), and tomato leaves treated with race-specific elicitors of Cladosporium also show rapid and preferential cell death in paraveinal cells (Hammond-Kosack et al., 1994). Generation of reactive oxygen species has been repeatedly proposed to be linked to JA. Indeed, such a cell type-specific generation of H2O2 was detected in tomato and discussed as a consequence of polygalacturonase gene expression which is induced by wounding or treatment with systemin and JA in the first 2 h (Bergey and Ryan, 1999). Non-mobile oligogalacturonides are generated by this enzyme accompanied with the formation of mobile H2O2 in parenchymatic cells of vascular bundles about 4 h after wounding (Orozco-Cardenas et al., 2001). Accordingly, systemin potentiates the oxidative burst in cultured tomato cells (Stennis et al., 1998). Thus, the early activation of AOC and generation of JA coincides spatially and temporally with the generation of H2O2 which is suggested to diffuse into mesophyll tissue (Orozco-Cardenas et al., 2001), where PIN2 expression is located (Narváez-Vásquez et al., 1993).
Obviously, compartmentalization of prosystemin and AOC in vascular bundles keeps the tissue in a pre-activated state which allows immediate response upon wounding. In this context, regulation of JA biosynthesis by substrate availability would contribute to its rapid generation. In transgenic tobacco plants overexpressing AOS, elevated level of JA were found only upon stimulation such as wounding, suggesting that wound-induced substrate generation is necessary for JA biosynthesis (Laudert et al., 2000; Wang et al., 1999). Substrate-dependent JA formation was also found in non-transgenic Arabidopsis plants, where LOX, AOS, and AOC occur abundantly without tissue specificity (Stenzel et al., 2003), and in transgenic Arabidopsis plants overexpressing AOS (Park et al., 2002). In contrast, in transgenic potato plants overexpressing the flax AOS elevated JA levels without constitutive PIN2 expression were found (Harms et al., 1995).
The data shown here confirm a control by substrate availability. The basic level of JA and OPDA in untreated tomato leaves is low even under constitutive overexpression of AOC (Figure 4b,c). Upon wounding, however, a rapid transient rise was observed, indicating that AOS is not limiting in tomato, even under constitutive overexpression of AOC. This is consistent with metabolite profiling data of 10 free oxylipins in tomato leaves, where the initial substrate of the AOS branch, 13-HPOT, could not be detected (Stenzel et al., 2003). It is possible that the activation of pre-existing JA biosynthetic enzymes by substrate generation or other mechanisms allow the leaf to generate JA immediately. The 10-fold higher in vitro AOC activity in total leaf extracts of AOCs plants and much less activity in def1 plants are in line with such a post-translational control. Furthermore, in situ hybridization of AOC mRNA (data not shown) and immunocytochemical localization of AOC (Figure 3c) revealed that AOC protein, but not AOC mRNA, appeared in a tissue-specific manner, suggesting post-translational control which might be overridden in the AOCs leaves (Figure 5).
The vascular bundle-specific AOC expression and JA generation combined with a post-translational control in JA biosynthesis at the AOC level may allow the plant to respond locally and rapidly leading to a burst in JA and/or related compounds. Such a burst within the first hour of elicitation, wounding or other external factors is regularly found (Koch et al., 1999; Ziegler et al., 2001) (Figure 4c). In tomato, amplification of JA and systemin formation may lead to a level sufficient for a systemic response. Indeed, transport of JA was found (Zhang and Baldwin, 1997), and elevation of JA level was detected in systemic leaves (Herde et al., 1996). Most convincingly, the recent grafting experiments with mutants in JA biosynthesis and signalling revealed a wound response in the systemic unwounded leaf with blocked JA biosynthesis but intact JA signaling (Li et al., 2002; Ryan and Moura, 2002), suggesting that JA or related compounds act as systemic signal.
Plant materials, treatments and reagents
Lycopersicon esculentum Mill. cv. Moneymaker, cv. Lukullus, cv. Castlemart or Better Boy were grown as described (Wasternack et al., 1998). Identical cultivation conditions were used for the homozygous def1 plants, and transgenic lines. For treatments, leaves were cut at the petiole and were floated on 50 µm (+/–)-JAME in Petri dishes containing 50 ml solution per leaf and were incubated at 25°C under continuous white light (120 µmol m−2 sec−1) provided by fluorescent lamps (Narva, Berlin, NC 250/01). In some experiments indicated in the legends, 100 nm systemin, 100 nm Sys17, 50 µm (+/–)-JA, 50 µm 12-hydroxy-JA, or 50 µm ethyl-indanoyl isoleucine conjugate were supplied via the petiole or applied to non-detached leaves.
All jasmonates and octadecanoids were prepared and purchased, respectively, checked on purity and used as described (Hause et al., 2000; Maucher et al., 2000). 13(S)-HPOT was prepared from α-linolenic acid by incubation with soybean LOX (Sigma Chemical Co., St. Louis, USA) (Hamberg and Gotthammar, 1973). 12-OH-JA was prepared according to Kitahara et al. (1984). The ethyl-indanoyl isoleucine conjugate recently described (Schüler et al., 2001) was kindly provided by Prof. W. Boland (Jena). Systemin and its analog systemin-17 were synthesized and kindly provided by Dr T. Nürnberger (Halle).
Extraction of RNA and Northern blot analysis
Total RNA was purified from frozen tissues by treatments with buffered phenol:chloroform:isoamyl alcohol 25 : 24 : 1 (v/v/v) and 20 µg per lane were subjected to RNA gel blot analysis according to Sambrook et al. (1989). Blots were hybridized at 60°C for 16 h with 32P-labeled cDNAs of tomato AOC (full length), tomato AOS (near full length), tomato PIN1 (1200-bp fragment containing 1 intron), tomato PIN2 (800-bp fragment), and potato PR1b (300-bp fragment). Loading control was performed by ethidium bromide staining of rRNA or by hybridization with a tomato cDNA of ATP/ADP carrier (600-bp fragment), and is exemplified as shown. cDNA fragments of PIN1 and PIN2 were isolated by PCR with genomic tomato DNA and primers deduced from database sequences (NO. KO3290 for LePIN1 and NO. KO3291 for LePIN2).
Extraction of proteins, immunoblot analysis and assay of AOC activity
For protein analysis, all tissues were subjected to extraction procedures described by Meyer et al. (1988). Proteins were solubilized in SDS sample buffer followed by SDS–PAGE and immunoblot analysis according to standard protocols. Immunodetection was performed with a polyclonal antibody raised against the recombinant AOC (Ziegler et al., 2000) as primary antibody in a dilution of 1 : 5000 and antirabbit IgG conjugated with alkaline phosphatase (Roche Diagnostics, Mannheim, Germany) as secondary antibody. Immunodecorated AOC was stained with p-nitroblue tetrazoline chloride (NBT) and 5-bromo-4-chloro-3-indolylphosphate (BCIP). With the anti-AOC antibody, only one band was detected in total protein extracts of tomato leaves or tomato flowers (Hause et al., 2000). AOC activity tests were performed as follows: Leaf tissue (0.5 g) was ground under liquid nitrogen and the resulting powder was extracted with 50 mm Na-phosphate buffer, pH 7, containing 2% PVPP and 0.1% Triton X-100. The homogenate was centrifuged at 15 000 × g for 10 min and the supernatant was used as the enzyme extract. The enzyme assay consisted of excess of purified recombinant AOS (Maucher et al., 2000) with an activity of 4 nkat, 50 mm Na-phosphate buffer, pH 7, and the enzyme extract in a total volume of 250 µl. The reaction was started by the addition of 5 µl of 10 mm 13-HPOT and was terminated by extraction of the reaction products with ether. Subsequently, the extract was incubated with 0.1 m NaOH in order to achieve cis→trans isomerization of OPDA. OPDA was purified by reversed phase HPLC on Nucleosil 120-5 C18 (Macherey-Nagel, Düren, Germany) with isocratic elution (63% v/v MeOH in water containing 0.1% acetic acid). The trans enantiomers were separated on Nucleodex-β-PM (250 mm × 4.6 mm, 5 µm; Macherey-Nagel, Düren, Germany) by isocratic elution with MeOH/H2O (0.1% triethylammoniumacetate pH 4) 60/40 (v/v) at a detection wavelength of 225 nm. The peak area of (9S,13R)-OPDA, which originated from the (9S,13S)-enantiomer, exceeding 50% detected after chemical hydrolysis, was used to calculate AOC activity.
Extraction and quantitative analysis of jasmonates and octadecanoids
Plant material from three different plants was pooled to minimize biological differences and was immediately frozen in liquid nitrogen. One gram (FW) was homogenized in a mortar and extracted with 5 ml of 80% (v/v) methanol. Appropriate amounts of (2H6)-JA were added as internal standards for quantitative analysis of JA. For purification and fractionation before GC–MS analysis, all jasmonate-containing compounds were subjected to ion exchange chromatography on DEAE Sephadex A-25 cartridges and RP-HPLC as described (Kramell et al., 2000). For quantitative analysis of OPDA, 0.5 g (FW) plant tissue was extracted and prepared for GC–MS analysis as described by Mueller and Brodschelm (1994). GC–MS analysis was performed with a Finnigan GCQ as described (Hause et al., 2000).
Construction of vectors and plant transformation
Transformation of tomato plants was performed with 35S::AOCsense and 35S::AOCantisense constructs (Maucher et al., in preparation). Briefly, the SmaI/SmaI–ApaI/SmaI 1.1-kb fragment of the tomato AOC-cDNA (full-length coding region; cf. Ziegler et al., 2000) was introduced in sense (s) and antisense (as) orientation by blunt end ligation into the multiple cloning site of the binary vector pBin-Hyg-Tx derived from pBin19 (Bevan, 1984). Recombinant plasmids were maintained and selected in Escherichia coli XL1-blue MRF1. Plasmids pBin-sAOC and pBin-asAOC were transfected into Agrobacterium tumefaciens strain LBA 4404. For transformation, cotyledons of tomato plants (L. esculentum cv. Lukullus) were used according to standard transformation protocol (Ling et al., 1998). Five independent sense and nine independent antisense plants were selected from 170 primary explants followed by isolation of homozygous lines in the T2 generation (Maucher et al., in preparation). These lines were used here.
Immunocytochemical analysis was performed as described (Hause et al., 2000) using cross-sections (2 µm thick) from PEG-embedded material. The anti-AOC antibody raised against recombinant tomato AOC (Ziegler et al., 2000) was used in a dilution of 1 : 2000. The LOX antibody raised against lipid body LOX of cucumber (Feussner et al., 1995) recognizes one LOX form in the Western blot analysis of total leaf extracts of untreated tomato leaves (Wasternack et al., 1998), and was used here in a dilution of 1 : 500. It cannot be excluded, however, that more than one of the chloroplast-located LOX forms of tomato were recognized in the immunocytological analysis. The tomato AOS antibody tomato AOS1 recognized in the Western blot analysis of total leaf extracts of untreated tomato leaves one of the two AOCs known to occur in leaves (Howe et al., 2000; Sivasankar et al., 2000) and was used in a dilution of 1 : 1000. As secondary antibody, antirabbit IgG conjugated with alkaline phosphatase and antirabbit IgG conjugated with Alexa488 (Molecular Probes Eugene, OR), respectively, were used.
We thank Prof. L.Varin (Montreal) for critical reading of the manuscript, B. Ortel, and S. Vorkefeld for technical assistance, and C. Dietel for typing the manuscript. Systemin and Sys-17 was kindly supplied by Dr. T. Nürnberger (Halle) and ethyl-indanoyl isoleucine conjugate by Prof. W. Boland (Jena). The LOX antibody was a gift from Prof. I. Feussner (Göttingen, Germany), and the tomato AOS antibody was a gift from Prof. G. Howe (East Lansing, USA). The research was supported by a grant (project C5 of the SFB 363) of the Deutsche Forschungsgemeinschaft to C.W. and O.M., and grant No. IBN 0090766 of the US National Science Foundation to C.A.R.
EMBL nucleotide sequence database accession numbers AJ 308481 for AOC of L. esculentum cv. Moneymaker and AJ 308482 for AOC of L. esculentum cv. Castlemart, def1 mutant.